The Journal of Pediatrics

Volume 151, Issue 3, Pages.223-328 (September 2007)

The Editors' Perspectives 1.

Safety of bumper pads Page A1 Alan H. Jobe

2.

Treatment of childhood obesity: What works? Page A1 Stephen R. Daniels

3.

Staphylococcal anti-adhesive antibodies fail to protect premature infants from bloodstream infection (BSI) Page A2 Sarah S. Long

4.

Vocal cord dysfunction and feeding after cardiac surgery Page A2 Stephen R. Daniels

5.

Kawasaki disease and subsequent risk for atherosclerosis Page A2 Stephen R. Daniels

6.

Prebiotics and weight gain Page A3 Stephen R. Daniels

7.

Tonsillectomy as treatment of PFAPA syndrome Page A3 Sarah S. Long

8.

In vivo magnetic resonance spectroscopy of muscle Page A3 Robert W. Wilmott

9.

Table of contents Pages A4-A7

10.

Editorial Board Page Page A8

11.

Information for Readers Page A9

12.

Announcements Page A10

Notes from the Association of Medical School Pediatric Department Chairs, Inc. 13.

American Pediatric Academia: The Looming Question Pages 223-224 Scott A. Rivkees and Myron Genel

Editorials 14.

Are Patients with Kawasaki Disease at Risk for Premature Atherosclerosis? Pages 225-228 Elif Seda Selamet Tierney and Jane W. Newburger

15.

Non! to Non-Steroidal Anti-Inflammatory Therapy for Inflammatory Lung Disease in Cystic Fibrosis (at Least at the Moment) Pages 228-230 Andrew Bush and Jane Davies

16.

What is the Role of Cystic Fibrosis Transmembrane Conductance Regulator Dysfunction in Primary Sclerosing Cholangitis? Pages 230-232 Dennis D. Black

17.

Specific Immune Globulin Therapy for Prevention of Nosocomial Staphylococcal Bloodstream Infection in Premature Infants: Not What We Hoped for! Pages 232-234 M. Teresa de la Morena

18.

Acute Viral Bronchiolitis: To Treat or Not to Treat—That Is the Question Pages 235-237 Claudia Calogero and Peter D. Sly

19.

“And Things that Go Bump in the Night”: Nothing to Fear? Pages 237-238 Rachel Y. Moon

Original Articles 20.

Subclinical Atherosclerosis, but Normal Autonomic Function after Kawasaki Disease Pages 239-243 Robert Dalla Pozza, Susanne Bechtold, Simon Urschel, Rainer Kozlik-Feldmann and Heinrich Netz

21.

Are Patients after Kawasaki Disease at Increased Risk for Accelerated Atherosclerosis? Pages 244-248.e1 Brian W. McCrindle, Susan McIntyre, Christopher Kim, Tammy Lin and Khosrow Adeli

22.

High-Dose Ibuprofen in Cystic Fibrosis: Canadian Safety and Effectiveness Trial Pages 249-254 Larry C. Lands, Ruth Milner, André M. Cantin, David Manson and Mary Corey

23.

Primary Sclerosing Cholangitis in Childhood is Associated with Abnormalities in Cystic Fibrosis–Mediated Chloride Channel Function Pages 255-259 Harpreet Pall, Julian Zielenski, Maureen M. Jonas, Deborah A. DaSilva, Kimberly M. Potvin, Xiao-Wei Yuan, Qiuju Huang and Steven D. Freedman

24.

Clinical Trial of Safety and Efficacy of IHN-A21 for the Prevention of Nosocomial Staphylococcal Bloodstream Infection in Premature Infants Pages 260-265.e1 Mitchell DeJonge, David Burchfield, Barry Bloom, Maria Duenas, Whit Walker, Mark Polak, Elizabeth Jung, Dietra Millard, Robert Schelonka, Fabien Eyal, et al.

25.

Nebulized Hypertonic Saline in the Treatment of Viral Bronchiolitis in Infants Pages 266-270.e1 Brian A. Kuzik, Samim A. Al Qadhi, Steven Kent, Michael P. Flavin, Wilma Hopman, Simon Hotte and Sarah Gander

26.

Deaths and Injuries Attributed to Infant Crib Bumper Pads Pages 271-274.e3 Bradley T. Thach, George W. Rutherford Jr and Kathleen Harris

27.

Insulin Resistance in Adolescents Pages 275-279 Ann M. Rodden, Vanessa A. Diaz, Arch G. Mainous III, Richelle J. Koopman and Mark E. Geesey

28.

A Cognitive Behavioral Therapy Program for Overweight Children Pages 280-283 Erica L.T. van den Akker, Patrycja J. Puiman, Mieke Groen, Reinier Timman, Mieke T.M. Jongejan and Wim Trijsburg

29.

Socioeconomic Position, Maternal IQ, Home Environment, and Cognitive Development Pages 284-288.e1 Shilu Tong, Peter Baghurst, Graham Vimpani and Anthony McMichael

30.

A Randomized, Controlled Trial of Tonsillectomy in Periodic Fever, Aphthous Stomatitis, Pharyngitis, and Adenitis Syndrome Pages 289-292 M. Renko, E. Salo, A. Putto-Laurila, H. Saxen, P.S. Mattila, J. Luotonen, O. Ruuskanen and M. Uhari

31.

50 Years Ago in The Journal of Pediatrics: Steroid therapy for rheumatic fever Page 292 Eli M. Eisenstein

32.

Effect of Prebiotic Supplementation and Calcium Intake on Body Mass Index Pages 293-298 Steven A. Abrams, Ian J. Griffin, Keli M. Hawthorne and Kenneth J. Ellis

33.

Long-Term Follow-Up in 12 Children with Pulmonary Arteriovenous Malformations: Confirmation of Hereditary Hemorrhagic Telangiectasia in all Cases Pages 299-306 Aurore Curie, Gaëtan Lesca, Vincent Cottin, Patrick Edery, Gabriel Bellon, Marie E. Faughnan and Henri Plauchu

34.

50 Years Ago in The Journal of Pediatrics: Niemann-Pick disease in a boy of 16 months Page 306 Hans C. Andersson

35.

Duodenogastro-Esophageal Reflux in Children with Refractory Gastro-Esophageal Reflux Disease Pages 307-311 Ilse Hoffman, Alexander Tertychnyy, Nadine Ectors, Toon De Greef, Nancy Haesendonck and Jan Tack

36.

Vocal Cord Dysfunction and Feeding Difficulties after Pediatric Cardiovascular Surgery Pages 312-315.e2 Ritu Sachdeva, Elora Hussain, M. Michele Moss, Michael L. Schmitz, Richard M. Ray, Michiaki Imamura and Robert D.B. Jaquiss

Grand Rounds 37.

Growing Skull Fracture after Minor Closed-Head Injury Pages 316-318 Jean-Rodolphe Vignes, N.U. Owase Jeelani, Ashfaq Jeelani, Michel Dautheribes and Dominique Liguoro

Clinical and Laboratory Observations 38.

In Vivo Proton Magnetic Resonance Spectroscopy Assessment for Muscle Metabolism in Neuromuscular Diseases Pages 319-321 Tsyh-Jyi Hsieh, Chien-Kuo Wang, Hung-Yi Chuang, Yuh-Jyh Jong, Chun-Wei Li and Gin-Chung Liu

39.

ABCA3 Deficiency Presenting as Persistent Pulmonary Hypertension of the Newborn Pages 322-324 Anette M. Kunig, Thomas A. Parker, Lawrence M. Nogee, Steven H. Abman and John P. Kinsella

Insights 40.

Severe Cerebellar Hypoplasia Associated with Osteogenesis Imperfecta Type III Page 325 B. Tabarki, S. Al-Malki and H. Al-Ghamdi

Current Best Evidence 41.

Adenotonsillectomy less beneficial for sleep apnea in older and obese children Page 326 Sarah Lacy

42.

Predictive value of rapid influenza tests varies with prevalence Pages 326-327 Vidya Sharma

43.

Failure to respond to name is indicator of possible autism spectrum disorder Pages 327-328 John G. Frohna

44.

Also Noted Page 328 John G. Frohna

45.

Also Noted Page 328 John G. Frohna

Letters to the Editor 46.

Transfusion threshold in anemic premature infants Page e10 Arie L. Alkalay and Charles F. Simmons

47.

Reply Page e10 H. Kirpalani, R. Whyte and R. Roberts

48.

The natural history of thyroid autoimmunity and thyroid function in children with type 1 diabetes Pages e10-e12 Terri H. Lipman, Iraj Rezvani and Angelo M. DiGeorge

49.

Reply Page e12 Giorgio Radetti, Elena Gottardi, Gianni Bona, Andrea Corrias, Silvana Salardi and Sandro Loche

50.

Extreme obesity among children in Mexico Pages e12-e13 Arturo Jimenez Cruz, Montserrat Bacardi-Gascon and Elizabeth Jones

This Month in

THE JOURNAL OF

PEDIATRICS September 2007 • Volume 151 • Number 3 Copyright © 2007 by Mosby, Inc.

THE EDITORS’ PERSPECTIVES Safety of bumper pads As the Back to Sleep program has effectively decreased the incidence of SIDS, the residual causes of accidental deaths in infancy become more apparent. The recent discussions about the characteristics of bedding material, bed sharing and associations with smoking are examples of the scrutiny given to accidental deaths. In this issue of The Journal, Thach et al identify bumper pads used in cribs as another source of risk for injury and accidental death for infants. Their conclusions are based on the databases of the US Consumer Product Safety Commission, which depend on what gets reported. Thus, there is neither an accurate numerator for the actual number of attributable injuries and deaths nor a denominator about frequency of use of bumper pads. Nevertheless, the probably low estimates do raise significant concerns that need to be recognized. —Alan H. Jobe, MD, PhD page 271 (article) page 237 (editorial)

Treatment of childhood obesity: What works? The epidemic of childhood obesity has left clinicians with important questions concerning effective management strategies. In this issue, van den Akker et al present the 1-year follow-up results of a multidisciplinary cognitive behavioral therapy approach to treat obese children. The program uses group therapy with a variety of behavioral therapy techniques. They found a weight loss of approximately 19% after 1 year in those who completed follow-up. Unfortunately, 33% of the patients dropped out. It is clear that behavioral therapy can be beneficial for some children who are overweight and obese. Further research is needed to determine more effective strategies for those children who drop out of therapy. In this study, the dropouts were older, had higher BMI at baseline, and were less successful in weight management at the early stages of therapy. —Stephen R. Daniels, MD, PhD page 280

The Journal of Pediatrics (ISSN 0022-3476) is published monthly by Elsevier Inc., 360 Park Avenue South, New York, NY 10010. Business and Editorial Offices: 1600 John F. Kennedy Blvd., Suite 1800, Philadelphia, PA 19103-2899. Accounting and Circulation Offices: 6277 Sea Harbor Drive, Orlando, FL 32887-4800. Periodicals postage paid at New York, NY, and additional mailing offices. POSTMASTER: Send address changes to The Journal of Pediatrics, Elsevier Periodicals Customer Service, 6277 Sea Harbor Drive, Orlando, FL 32887-4800.

The Journal of Pediatrics

September 2007

1A

Staphylococcal anti-adhesive antibodies fail to protect premature infants from bloodstream infection (BSI) Very low birth weight (ⱕ 1500 g VLBW) infants comprise 1.4% of births in the United States, and survival rates are increasing. So too, are complications of neonatal intensive care such as late onset septicemia, which occurs in 16 to 25% of VLBW infants. DeJonge et al in this issue of The Journal report the results of a phase III, randomized, double-blind, placebo-controlled multicenter trial of prophylactic use of immunoglobulin selected for high titers of staphylococcal antiadhesive antibodies (IHN-A21) and prepared for intravenous use (IGIV). The phase III trial was conducted on the heels of a phase II trial that showed a trend toward protection from staphylococcal and candidal bloodstream infection (BSI), with adequate sample size and power in the Phase III trial. Almost 2000 neonates in 95 centers in the United States and Canada received at least one infusion of study drug (750 mg/kg IHN-A21 or saline placebo). Although there was no safety issue, there was no effect on rate or timing of staphylococcal or other BSIs. In the accompanying editorial, de la Morena puts this definitively negative study result in the context of other disappointments of transiently promising interventions, and speculates on what premature neonates really need. —Sarah S. Long, MD page 260 (article) page 232 (editorial)

Vocal cord dysfunction and feeding after cardiac surgery Pediatric patients may have feeding difficulties after cardiovascular surgery. Cardiothoracic surgery can also be associated with vocal cord dysfunction. In this issue of the Journal, Sachdeva et al evaluated the relationship of vocal cord dysfunction and feeding difficulties in a group of patients after cardiovascular surgery. Approximately 2% had vocal cord dysfunction. They found that many patients with vocal cord dysfunction had concomitant feeding difficulties, some of whom required feeding by gastrostomy tube. The authors recommend surveillance for vocal cord dysfunction in patients undergoing surgery of the aortic arch in which the recurrent laryngeal nerve can be damaged. —Stephen R. Daniels, MD, PhD page 312

Kawasaki disease and subsequent risk for atherosclerosis A major complication of Kawasaki disease is inflammation and vasculitis of the coronary arteries. This may lead to coronary artery aneurysms. An important question is whether those without aneurysms have increased long-term risk of atherosclerosis. In this issue of The Journal two studies address this question. McCrindle et al found that patients, after Kawasaki disease, have some abnormalities in cardiovascular risk factors, but have no evidence for systemic arterial endothelial dysfunction. This is reassuring. However, in a separate study, Dalla Pozza et al found that carotid artery intimal-medial thickness was greater in patients after Kawasaki disease than controls and that those with coronary artery involvement after Kawasaki disease had the largest intimal-medial thickness. These findings suggest potentially accelerated atherosclerosis in patients with Kawasaki disease. In an editorial, Selamet Tierney and Newburger point out that we currently must rely on relatively small studies, with differing patient characteristics, different length of followup and intermediate outcomes in which interpretation of the clinical relevance of results may be difficult. They emphasize that long-term international studies with clinical outcomes will be optimum to assess the impact of Kawasaki disease on vascular health. In the meantime, appropriate management of known risk factors, such as hypertension and dyslipidemia, is important to maximize vascular health in patients who have had Kawasaki disease. —Stephen R. Daniels, MD, PhD page 244 (McCrindle) page 239 (Dalla Pozza) page 225 (editorial)

2A

September 2007

The Journal of Pediatrics

Prebiotics and weight gain There have been few studies of the impact of prebiotics on weight gain in children and adolescents. In this issue of The Journal, Abrams et al evaluated the effect of a daily prebiotic supplement which consisted of a co-spray dried 1:1 mixture of oligofructose and long chain inulin compared to a control supplement. They found that subjects randomized to the prebiotic had a smaller increase in BMI compared to controls, and maintained their BMI Z-score while controls had increased BMI Z-score. Of interest is that the prebiotic appeared to have a greater impact on BMI in subjects who had higher intake of calcium. This study did not focus on subjects who were overweight or randomly allocate the level of calcium intake, so the results may not be applicable to treatment of overweight and could be confounded by other aspects of diet and physical activity. Further research will be needed to evaluate this product and other prebiotics in the treatment of obesity.

In vivo magnetic resonance spectroscopy of muscle Proton magnetic resonance spectroscopy was used in vivo to evaluate muscle metabolism in patients with neuromuscular diseases. The results show that children with Duchenne muscular dystrophy and spinal muscular atrophy have reduced tri-methyl-amid (TMA) peaks. TMA is involved in the metabolism of phospholipids and the decrease must represent a reduced rate of cell membrane synthesis, decreased cell turnover or a decreased cell number in these diseases that are characterized by fatty degeneration. —Robert W. Wilmott, MD page 319

—Stephen R. Daniels, MD, PhD page 293

Tonsillectomy as treatment of PFAPA syndrome Periodic-Fever-Aphthous stomatitis-Pharyngitis and cervical Adenitis (PFAPA) syndrome is a curious, troublesome and not uncommon pediatric diagnosis. Diagnosis is entirely clinical, with the primary feature of clockwork periodicity of fever every 3-6 weeks for 3-5 days with little else. The cause is unknown, but the course is known likely to be persistent over years before spontaneous, full resolution. Case series have all reported a curious, apparent, immediate, curative effect of tonsillectomy in many affected children. In this issue of The Journal, Renko et al report results of a randomized controlled trial of tonsillectomy versus follow-up alone in 26 children with PFAPA enrolled at mean age of 4.1 years after at least 5 predictable, periodic episodes of fever. Six months after randomization, all 14 children who underwent tonsillectomy were free of symptoms, whereas 6 of 12 children randomized to follow-up alone continued to have periodic fever; 5 of these 6 children then underwent tonsillectomy and were promptly “cured.” With caveats of small sample size, inability to blind and inability to ascertain specificity of diagnosis of PFAPA (and noting that 29% had exudative tonsillitis, which is distinctly unusual in case series from the United States), these results are impressive, and “cure” is as curious as is the problem. —Sarah S. Long, MD page 289

The Journal of Pediatrics

September 2007

3A

THE JOURNAL OF

PEDIATRICS September 2007 • Volume 151 • Number 3 Copyright © 2007 by Mosby, Inc.

NOTES FROM THE ASSOCIATION OF MEDICAL SCHOOL PEDIATRIC DEPARTMENT CHAIRS, INC.

American Pediatric Academia: The Looming Question

223

Scott A. Rivkees, MD, and Myron Genel, MD, New Haven, Connecticut

EDITORIALS

Are Patients with Kawasaki Disease at Risk for Premature Atherosclerosis?

225

Elif Seda Selamet Tierney, MD, and Jane W. Newburger, MD, MPH, Boston, Massachusetts

Non! to Non-Steroidal Anti-Inflammatory Therapy for Inflammatory Lung Disease in Cystic Fibrosis (at Least at the Moment)

228

Andrew Bush, MB, BS (Hons), MA, MD, FRCP, FRCPCH, and Jane Davies, MB, ChB, MRCP, MRCPCH, MD, London, United Kingdom

What is the Role of Cystic Fibrosis Transmembrane Conductance Regulator Dysfunction in Primary Sclerosing Cholangitis?

230

Dennis D. Black, Memphis, Tennessee

Specific Immune Globulin Therapy for Prevention of Nosocomial Staphylococcal Bloodstream Infection in Premature Infants: Not What We Hoped for!

232

M. Teresa de la Morena, MD, Dallas, Texas

Acute Viral Bronchiolitis: To Treat or Not to Treat—That Is the Question

235

Claudia Calogero, MD, and Peter D. Sly, MBBS, MD, DSC, FRACP, Florence, Italy, and Subiaco, Australia

“And Things that Go Bump in the Night”: Nothing to Fear?

237

Rachel Y. Moon, MD, Washington, DC

ORIGINAL ARTICLES

Subclinical Atherosclerosis, but Normal Autonomic Function after Kawasaki Disease

239

Robert Dalla Pozza, MD, Susanne Bechtold, MD, Simon Urschel, MD, Rainer Kozlik-Feldmann, MD, and Heinrich Netz, MD, PhD, Munich, Germany

Are Patients after Kawasaki Disease at Increased Risk for Accelerated Atherosclerosis?

244

Brian W. McCrindle, MD, MPH, Susan McIntyre, RN, Christopher Kim, Tammy Lin, and Khosrow Adeli, PhD, Toronto, Ontario, Canada

4A

September 2007

The Journal of Pediatrics

High-Dose Ibuprofen in Cystic Fibrosis: Canadian Safety and Effectiveness Trial

249

Larry C. Lands, MD, PhD, Ruth Milner, PhD, André M. Cantin, MD, David Manson, MD, and Mary Corey, PhD, Montreal and Sherbrooke, Quebec, Vancouver, British Columbia, and Toronto, Ontario, Canada

Primary Sclerosing Cholangitis in Childhood is Associated with Abnormalities in Cystic Fibrosis– Mediated Chloride Channel Function

255

Harpreet Pall, MD, Julian Zielenski, PhD, Maureen M. Jonas, MD, Deborah A. Dasilva, RN, Kimberly M. Potvin, Xiao-Wei Yuan, MSc, Qiuju Huang, MD, and Steven D. Freedman, MD, PhD, Boston, Massachusetts, and Toronto, Ontario, Canada

Clinical Trial of Safety and Efficacy of IHN-A21 for the Prevention of Nosocomial Staphylococcal Bloodstream Infection in Premature Infants

260

Mitchell DeJonge, MD, David Burchfield, MD, Barry Bloom, MD, Maria Duenas, MD, Whit Walker, MD, Mark Polak, MD, Elizabeth Jung, MD, Dietra Millard, MD, Robert Schelonka, MD, Fabien Eyal, MD, Amy Morris, MBA, Barry Kapik, MS, Destrey Roberson, RN, Karen Kesler, PhD, Joe Patti, PhD, and Seth Hetherington, MD, Grand Rapids and Detroit, Michigan, Gainesville, Florida, Wichita, Kansas, Greenville, South Carolina, Morgantown, West Virginia, St. Louis, Missouri, Evanston, Illinois, Birmingham and Mobile, Alabama, Alpharetta, Georgia, and Chapel Hill, North Carolina

Nebulized Hypertonic Saline in the Treatment of Viral Bronchiolitis in Infants

266

Brian A. Kuzik, MD, MSc, FRCP(C), Samim A. Al Qadhi, MD, MBChB, Steven Kent, BSc(med), MD, FRCP(C), Michael P. Flavin, MB, MRCP(UK), FRCP(C), Wilma Hopman, MA, Simon Hotte, MD, and Sarah Gander, MD, Abu Dhabi, United Arab Emirates, Victoria, British Columbia, and Kingston, Ontario, Canada

Deaths and Injuries Attributed to Infant Crib Bumper Pads

271

Bradley T. Thach, MD, George W. Rutherford, Jr, MS, and Kathleen Harris, St. Louis, Missouri

Insulin Resistance in Adolescents

275

Ann M. Rodden, DO, Vanessa A. Diaz, MD, MS, Arch G. Mainous III, PhD, Richelle J. Koopman, MD, MS, and Mark E. Geesey, MS, Charleston, South Carolina

A Cognitive Behavioral Therapy Program for Overweight Children

280

Erica L. T. van den Akker, MD, Patrycja J. Puiman, MD, Mieke Groen, MSc, Reinier Timman, PhD, Mieke T. M. Jongejan, MD, PhD, and Wim Trijsburg, PhD, Rotterdam, the Netherlands

Socioeconomic Position, Maternal IQ, Home Environment, and Cognitive Development

284

Shilu Tong, PhD, Peter Baghurst, PhD, Graham Vimpani, PhD, and Anthony McMichael, PhD, Kelvin Grove, Adelaide, Newcastle, and Canberra, Australia

A Randomized, Controlled Trial of Tonsillectomy in Periodic Fever, Aphthous Stomatitis, Pharyngitis, and Adenitis Syndrome

289

M. Renko, MD, PhD, E. Salo, MD, PhD, A. Putto-Laurila, MD, PhD, H. Saxen, MD, PhD, P. S. Mattila, MD, PhD, J. Luotonen, MD, PhD, O. Ruuskanen, MD, PhD, and M. Uhari, MD, PhD, Oulu, Helsinki, and Turku, Finland

50 Years Ago in The Journal of Pediatrics—Steroid Therapy for Rheumatic Fever

292

Eli M. Eisenstein, MD, Mount Scopus, Jerusalem, Israel

continued on page 6A The Journal of Pediatrics

September 2007

5A

Effect of Prebiotic Supplementation and Calcium Intake on Body Mass Index

293

Steven A. Abrams, MD, Ian J. Griffin, MBChB, Keli M. Hawthorne, MS, and Kenneth J. Ellis, PhD, Houston, Texas

Long-Term Follow-Up in 12 Children with Pulmonary Arteriovenous Malformations: Confirmation of Hereditary Hemorrhagic Telangiectasia in all Cases

299

Aurore Curie, MD, Gaëtan Lesca, MD, Vincent Cottin, MD, PhD, Patrick Edery, MD, PhD, Gabriel Bellon, MD, PhD, Marie E. Faughnan, MD, MSc, and Henri Plauchu, MD, PhD, Lyon, France and Toronto, Canada

50 Years Ago in The Journal of Pediatrics—Niemann-Pick Disease in a Boy of 16 Months

306

Hans C. Andersson, MD, FACMG, New Orleans, Lousiana

Duodenogastro-Esophageal Reflux in Children with Refractory Gastro-Esophageal Reflux Disease

307

Ilse Hoffman, MD, Alexander Tertychnyy, MD, Nadine Ectors, PhD, Toon De Greef, Nancy Haesendonck, and Jan Tack, PhD, Leuven, Belgium, and Moscow, Russia

Vocal Cord Dysfunction and Feeding Difficulties after Pediatric Cardiovascular Surgery

312

Ritu Sachdeva, MD, Elora Hussain, MD, M. Michele Moss, MD, Michael L. Schmitz, MD, Richard M. Ray, MD, Michiaki Imamura, MD, PhD, and Robert D. B. Jaquiss, MD, Little Rock, Arkansas

GRAND ROUNDS

Growing Skull Fracture after Minor Closed-Head Injury

316

Jean-Rodolphe Vignes, MD, PhD, N. U. Owase Jeelani, MRCS, MBA, MPhil, Ashfaq Jeelani, MD, MSc, MRCPCH, Michel Dautheribes, MD, and Dominique Liguoro, MD, PhD, Bordeaux, France, and London and Wickford, Essex, UK

CLINICAL AND LABORATORY OBSERVATIONS

In Vivo Proton Magnetic Resonance Spectroscopy Assessment for Muscle Metabolism in Neuromuscular Diseases

319

Tsyh-Jyi Hsieh, MD, Chien-Kuo Wang, MD, Hung-Yi Chuang, MD, ScD, Yuh-Jyh Jong, MD, MMS, Chun-Wei Li, PhD, and Gin-Chung Liu, MD, Kaohsiung, Taiwan

ABCA3 Deficiency Presenting as Persistent Pulmonary Hypertension of the Newborn

322

Anette M. Kunig, MD, Thomas A. Parker, MD, Lawrence M. Nogee, MD, Steven H. Abman, MD, and John P. Kinsella, MD, Denver, Colorado, and Baltimore, Maryland

INSIGHTS

Severe Cerebellar Hypoplasia Associated with Osteogenesis Imperfecta Type III

325

B. Tabarki, MD, S. Al-Malki, MD, and H. Al-Ghamdi, MD, Taif, Kingdom of Saudi Arabia

CURRENT BEST EVIDENCE

Clinical Research Abstracts for Pediatricians

6A

September 2007

326

The Journal of Pediatrics

LETTERS

The following section is available in the online version of The Journal.

Transfusion Threshold in Anemic Premature Infants

e10

Arie L. Alkalay, MD, and Charles F. Simmons, MD, Los Angeles, California

Reply

e10

H. Kirpalani, MSc, FRCP(UK), R. Whyte, MB, FRCP(C), and R. Roberts, M.Tech, Hamilton, Ontario, Canada

The Natural History of Thyroid Autoimmunity and Thyroid Function in Children with Type 1 Diabetes

e10

Terri H. Lipman, PhD, CRNP, Iraj Rezvani, MD, and Angelo M. DiGeorge, MD, Philadelphia, Pennsylvania

Reply

e12

Giorgio Radetti, MD, Elena Gottardi, MD, Gianni Bona, MD, Andrea Corrias, MD, Silvana Salardi, MD, and Sandro Loche, MD, Bolzano, Novara, Torino, Bologna, and Cagliari, Italy

Extreme Obesity among Children in Mexico

e12

Arturo Jimenez Cruz, MD, PhD, Montserrat Bacardi-Gascon, MD, EdD, and Elizabeth Jones, RD, EdD, Mexico

READER SERVICES

Information for Readers

9A

Announcements Guide for Authors

The Journal of Pediatrics

10A Available at www.jpeds.com

September 2007

7A

THE JOURNAL OF

PEDIATRICS Copyright © 2007 by Mosby, Inc.

EDITOR William F. Balistreri Cincinnati, Ohio ASSOCIATE EDITORS Stephen R. Daniels Denver, Colorado Alan H. Jobe Cincinnati, Ohio Sarah S. Long Philadelphia, Pennsylvania Thomas R. Welch Syracuse, New York Robert W. Wilmott St. Louis, Missouri MANAGING EDITOR Alice M. Landwehr Cincinnati, Ohio EDITORIAL BOARD Steven H. Abman Denver, Colorado Pasquale J. Accardo Richmond, Virginia Hans C. Andersson New Orleans, Louisiana

John G. Frohna Editor, Current Best Evidence

Madison, Wisconsin Thomas P. Green Chicago, Illinois Terrie E. Inder St. Louis, Missouri

James F. Padbury Providence, Rhode Island Janet H. Silverstein Gainesville, Florida Charles Stanley Philadelphia, Pennsylvania Bonita Stanton

James Brown East Syracuse, New York

M. Susan Jay Milwaukee, Wisconsin

Editor, AMSPDC Section

Ronald I. Clyman San Francisco, California

Nancy F. Krebs Denver, Colorado

Marshall L. Summar Nashville, Tennessee

Paul G. Fisher Palo Alto, California

Mary B. Leonard Philadelphia, Pennsylvania

Russell E. Ware Memphis, Tennessee

Wayne J. Morgan Tucson, Arizona

Reginald L. Washington Denver, Colorado

Detroit, Michigan

INTERNATIONAL ADVISORY PANEL Anita Aperia Stockholm, Sweden

8A

September 2007

Michael J. Lentze Bonn, Germany

Albert M. Li Shatin, Hong Kong

Philip M. Sherman Toronto, Canada

The Journal of Pediatrics

THE JOURNAL OF

PEDIATRICS Copyright © 2007 by Mosby, Inc.

Manuscript Submissions. Submit all manuscripts to William F. Balistreri, MD, The Journal of Pediatrics, Editor-in-Chief, at www.ees.elsevier.com/jpeds. New users will be asked to register and create a user account in the online system. Authors who have submitted to the Journal in the past may have existing accounts in the system. The online system includes an author tracking feature that monitors the publication process from initial submission, through screening, review, revision, and publication. Paper submissions cannot be accepted. Please send editorial questions to [email protected]. Correspondence. Communications regarding original articles and editorial management should be addressed to William F. Balistreri, MD, The Journal of Pediatrics, Children’s Hospital Medical Center, 3333 Burnet Ave, MLC 3021, Cincinnati, OH 45229-3039; telephone 513-636-7140. Customer Service. (orders, claims, online, change of address): Elsevier Periodicals Customer Service, 6277 Sea Harbor Drive, Orlando, FL 32887-4800. Tel: (800) 654-2452 (U.S. and Canada); (407) 345-4000 (outside U.S. and Canada). Fax: (800) 225-6030 (U.S. and Canada); (407) 363-9661 (outside U.S. and Canada). E-mail: [email protected] Address changes must be submitted six weeks in advance. Subscriptions. United States and possessions: Individual $189.00; Institution $520.00; Student/Resident: $95.00. All other countries (prices include airspeed delivery): Individual $235.00; Institution $566.00; Student/Resident $120.00. Single Issues $48.00. To receive student/resident rate, orders must be accompanied by name of affiliated institution, date of term and the signature of program/residency coordinator on institution letterhead. Orders will be billed at the individual rate until proof of status is received. Current prices are in effect for back volumes and back issues. Further information on this journal is available from the Publisher or from this journal’s website (http://www.jpeds.com). Information on other Elsevier products is available through Elsevier’s website (http://www.elsevier.com).

Advertising Information. Advertising orders and inquiries can be sent to: USA, Canada, and South America, Pat Hampton, Elsevier, 360 Park Avenue South, New York, NY, 10010; phone: 212-633-3181; fax: 212-633-3820. Classified advertising orders and inquiries can be sent to Traci Peppers, Elsevier Inc., 360 Park Ave South, New York, NY 10010-1710; phone: 212-633-3766; fax: 212-633-3820. Europe and the rest of the world, Julie Toop; phone ⫹44 (0) 1865 843016; fax ⫹44 (0) 1865 843976; e-mail media@ elsevier.com.

Reprints. For queries about author offprints, e-mail [email protected]. To order 100 or more reprints for educational, commercial, or promotional use, contact the Commercial Reprints Department, Elsevier Inc., 360 Park Avenue South, New York, NY 10010-1710. Fax: (212) 462-1935; e-mail [email protected]. Reprints of single articles available online may be obtained by purchasing Pay-Per-View access for $30 per article on the journal website, www.jpeds.com. Microform edition

The Journal of Pediatrics

available from ProQuest Information and Learning, 300 N Zeeb Rd, Ann Arbor, MI 48106-1346. © 2007 Mosby, Inc. All rights reserved. This journal and the individual contributions contained in it are protected under copyright by Mosby, Inc., and the following terms and conditions apply to their use:

Photocopying. Single photocopies of single articles may be made for personal use as allowed by national copyright laws. Permission of the Publisher and payment of a fee is required for all other photocopying, including multiple or systematic copying, copying for advertising or promotional purposes, resale, and all forms of document delivery. Special rates are available for educational institutions that wish to make photocopies for nonprofit educational classroom use. Permissions may be sought directly from Elsevier’s Global Rights Department in Oxford, UK: phone (⫹1) 215 239 3804 or ⫹44 (0) 1865 843830, fax ⫹44 (0) 1865 853333, e-mail healthpermissions@ elsevier.com. Requests may also be completed online via the Elsevier homepage (http://www.elsevier.com/permissions). In the USA, users may clear permissions and make payments through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; phone: (⫹1) (978) 7508400, fax: (⫹1) (978) 7504744, and in the UK through the Copyright Licensing Agency Rapid Clearance Service (CLARCS), 90 Tottenham Court Road, London W1P 0LP, UK; phone: (⫹44) 20 7631 5555; fax: (⫹44) 20 7631 5500. Other countries may have a local reprographic rights agency for payments. Derivative Works. Subscribers may reproduce tables of contents or prepare lists of articles including abstracts for internal circulation within their institutions. Permission of the Publisher is required for resale or distribution outside the institution. Permission of the Publisher is required for all other derivative works, including compilations and translations.

Electronic Storage or Usage. Permission of the Publisher is required to store or use electronically any material contained in this journal, including any article or part of an article. Except as outlined above, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission of the Publisher. Address permissions requests to: Elsevier Rights Department, at the fax and e-mail addresses noted above.

Notice. No responsibility is assumed by the Publisher for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions or ideas contained in the material herein. Because of rapid advances in the medical sciences, in particular, independent verification of diagnoses and drug dosages should be made. Although all advertising material is expected to conform to ethical (medical) standards, inclusion in this publication does not constitute a guarantee or endorsement of the quality or value of such product or of the claims made of it by its manufacturer.

September 2007

9A

ANNOUNCEMENTS

September 2007 Youth Mental Health: Bridging Research and Clinical Practice. September 17-18, 2007, University of Minnesota Office of Continuing Medical Education, Radisson University Hotel, Minneapolis, Minnesota. Review the most common pediatric disorders (psychosis, anxiety and depression, substance abuse, trauma) in clinical practice. Connect the science and practice of prevention and treatment. Go over current knowledge, discuss the neurobiology, and get practical advice for prevention and treatment. For more information: Office of Continuing Medical Education, University of Minnesota; phone 612-626-7600 or 800-776-8636; E-mail: [email protected]; Register online: www.cmecourses.umn.edu.

countries. The symposium will be comprised of expert presentations providing an overview of the problems, issues and instances of work that is being done; oral presentations from selected abstracts on related issues; and structured panel discussions and open forums focused on determining research that is needed. A full list of speakers and topics and details of a call for abstracts are available on the conference website: www.chinamed.com.cn/pgpr2007. For more information, contact Alvin Zipursky, MD, Chair and Scientific Director, The Programme for Global Paediatric Research, The Hospital for Sick Children, Toronto, Canada; phone (001) 416813-8762; E-mail: [email protected]; Website: www.globalpaediatricresearch.org.

October 2007 New Insights into Childhood Functional Abdominal Pain and IBS, a one-day thematic educational and research conference associated with the 2007 NASPGHAN Meeting and Postgraduate Course. October 24, 2007, Grand American Hotel, Salt Lake City, Utah. Sponsored by Columbus Children’s Hospital’s Division of Gastroenterology, Hepatology and Nutrition. More than 20 multidisciplinary health care professionals will present and serve as moderators for the day-long program, scheduled the day before the NASPGHAN event. Carlo Di Lorenzo, MD, chief of the GI Division at Columbus Children’s Hospital will be one of the lead presenters. Focusing on functional abdominal pain and IBS, the program will incorporate successive moderated discussions that include etiology and pathophysiology, lessons learned from other specialties, Rome III, treatment and research. Eight CME hours will be available. For complete registration, contact NASPGHAN National Office: phone 215-233-0808; E-mail: [email protected]; Website: www.naspghan.org. The Effect of Environmental Pollutants on Foetal and Child Development: A Global Issue. October 26 and 27, 2007, Hangzhou, China. A Programme for Global Paediatric Research Symposium, presented by The Programme for Global Paediatric Research and the Chinese Pediatric Society of The Chinese Medical Association; in cooperation with The Children’s Hospital of Zhejiang University School of Medicine, Shanghai Children’s Medical Center and Xinhua Hospital, affiliated with Shanghai Jiao Tong University School of Medicine. PGPR’s sixth symposium will be held October 26 and 27, 2007 in association with the Chinese Society of Pediatrics of The Chinese Medical Association. The sessions will focus on the effects of environmental pollution on foetal and child development. Particular emphasis will be placed on child health in developing

December 2007 Hot Topics in Neonatology. December 2-4, 2007, Omni Shoreham Hotel, Washington, DC. Sponsored by Neonatal Research and Technology Assessment, Inc. (NRTA). Premier, exciting, interactive annual conference for neonatologists from around the world. Average 1400 attendees, 25⫹ speakers and guest discussants. Chairman, Dr. Jerold F. Lucey, Wallace Professor of Neonatology, Burlington, Vermont, Editor-in-Chief, Pediatrics. For more information, contact Gail Murphy, Neonatal Research and Technology Assessment, Inc; phone 802-865-2283; E-mail: [email protected]; Website: www.hottopics.org.

10A

September 2007

May 2008 9th Congress of the European Society for Pediatric Dermatology (ESPD). May 15-17, 2008, Athens Hilton Hotel, Athens, Greece. Sponsored by The European Society for Pediatric Dermatology. For more information, contact Mrs. Penelope Mitroyianni, Erasmus Conferences Tours & Travel; phone 0030 210 7257693; E-mail: info@ espd2008.com; Website: www.espd2008.com. 2007-2008 Certifying Examinations of the American Board of Pediatrics 111 Silver Cedar Court, Chapel Hill, NC 27514-1513 telephone: 919-929-0461 fax: 919-918-7114 or 919-9299255 Website: www.abp.org All applicants for certifying examinations must complete applications online during the registration periods. The final month of each registration requires payment of a late fee. The requirements for online applications may be found on the ABP Website (www.abp.org) or may be obtained by contacting the ABP. Additional information including eligibility requirements and registration dates may be found on the ABP Website.

The Journal of Pediatrics

NOTES FROM THE ASSOCIATION OF MEDICAL SCHOOL PEDIATRIC DEPARTMENT CHAIRS, INC.

EDITOR’S NOTE: Periodically in the AMSPDC section, we publish a perspective on an important topic in academic pediatrics. The opinions expressed in such commentaries reflect those of the individual author(s) and not necessarily those of AMSPDC as an organization. —Bonita Stanton, MD, Section Editor, The Journal of Pediatrics

American Pediatric Academia: The Looming Question

Department of Pediatrics, Yale University School of Medicine, New Haven, CT

(J Pediatr 2007;151:223-4)

223

C D P

Concomitant with the decline of the physician-scientist has been large growth in the ranks of academic clinicians in pediatric departments. Financial pressures have the potential to limit the ability of this growing faculty component to contribute to scholarly activities if faculty compensation is tied to relative value unit-based clinical productivity alone.9 Proper compensation and time for education, research, publication, and non-clinical departmental activities is also needed, because these activities give a medical center enthusiasm and purpose. The importance of the pediatric clinician-scientist is recognized at all levels of academia. Developing clinicianscientists for departments of pediatrics requires time and capital outlay that tops $1 million per junior faculty member. It costs $150,000 to $250,000, including salary, fringe benefits, and laboratory support, to train a fellow for 3 years. After completion of a fellowship, another $650,000 to $800,000 is needed to support a junior faculty member in the early years of academic growth.10 After academic independence is realized, junior and senior pediatric physician-scientists are hard-pressed to achieve sustained funding to provide substantial salary support in the current funding climate. Both junior and established senior clinician-scientists will need episodic financial support to cover funding shortfalls. Recognized for intellectual and scientific contributions, these individuals can quickly become a financial liability to departments. In several pediatric subspecialties, including pediatric surgery, neonatology, gastroenterology, cardiology, and critical care medicine, the maximum NIH-funded salary is significantly less than the typical salaries of active practitioners.11 Although academically desirable, fully funded clinician-scientists in these subspecialties may be unaffordable, because salary supplementation needed for academic retention hurts the bottom line.

S

A

MYRON GENEL, MD

M

cademic conferences and editorial columns have been marked in the past decade by addresses lamenting the decline of the physician-scientist, asking, “Where have all the young ones gone?”1-3 Further into the decline of the physician-scientist, coupled with progressive shortfalls of research funding and clinical revenues that have been very problematic for pediatrics, we wonder whether the next tough question will be, “Where have American academic pediatric departments gone?” From 1998 to 2003, the National Institutes of Health (NIH) budget doubled under visionary political and research leadership. In 1997, 32 departments of pediatrics had ⬎10 NIH grant awards.4 By 2005, ⬎45 departments of pediatrics had ⬎10 NIH research grants.5 Yet, as the NIH budget doubled, pediatric research growth did not keep up. The pediatric portfolio of the NIH represented 14.1% of funding in 1994.6 By 2000, this mark slipped to 12.6%.6 In 2005, the pediatric portfolio was 11.3% of NIH funding—25% less than 10 years earlier.6 Perhaps reflecting funding and faculty development issues, or perhaps not, pediatric research is shrinking on the center scientific stage. If one looks at the published reports in top-tier journals, including Science, Nature, The Proceeding of the National Academy of Sciences, The Journal of Clinical Investigation, and The New England Journal of Medicine, there has been a decline in publications from American departments of pediatrics; American pediatric departments contributed 35% fewer reports to these top journals in 2006 than in 2000.7 It is important to note that this decline is not unique to pediatrics; there has been a general decline in America’s contribution to the most meritorious literature.8

AND

A

SCOTT A. RIVKEES, MD,

We are concerned that pediatric physician-scientists are not being trained in basic or clinical science with the same rigor and depth as PhD graduate students, with whom physician-scientists will be competing for funding. Many of us serving on grant review panels observe a “research-quality gap” between junior pediatric MD and PhD scientists. The research component of pediatric fellowship training will not succeed if viewed as a hobby or if mentors are chosen by fellows on the basis of collegiality. Reflected by the success of graduates of programs like the national Pediatric Scientist Development Program,12 pediatric academicians do know how to train physician-scientists—place talented individuals with great mentors and provide protected time to be creative and thrive. As is being done at a few pediatric centers, similar programs can be developed.

A

Pediatric departments will need to rethink their missions over the coming decade.

M S P D C

Serious manpower and leaderships issues that limit academic growth are present in many departments of pediatrics. In pediatric endocrinology, for example, ⬎70 positions are posted on the job-listing site of the Lawson Wilkins Pediatric Endocrinology Society,13 and many prestigious section-chief positions remain unfilled after lengthy searches. Perhaps we should establish distinct clinical and research training tracks in substitute of the “3-years of training for all” approach in place for 2 decades now. Shortening the duration of fellowship training for a clinical track to 2 years (or less) will speed the entry of needed clinicians to the workforce— helping one problem—and free precious departmental dollars for physician-scientist support— helping another. Pediatric departments will need to rethink their missions in the coming decade. Biomedical research and faculty development is expensive. It may not be financially feasible to support clinical operations, medical education, the research enterprise, and faculty development and growth. Cultivating clinical faculty to develop clinical programs of excellence with sound revenue streams may be a legitimate alternative to a 4-part mission and an important form of faculty development. Such a model will be far richer if pediatric-based clinical programs are linked with institutional research programs, which can be directed toward elucidating disease mechanisms and optimizing treatment. Placing fellows and junior faculty members in non-pediatric departments during periods of training will cultivate the needed broad-based multidisciplinary ties among pediatric, basic, and other clinical departments.

224

Rivkees

Considering the current funding and medical economic climate, pediatric research may contract at many individual institutions. This change in the pediatric research base will heighten the need for federally funded center programs that can provide the needed infrastructure to maintain and enhance research in childhood diseases and disorders. In structuring such program, it will be crucial that additional funds be earmarked for pediatric research, rather than reallocating an already tight NIH pediatric research portfolio. Failing the aforementioned, uncovering dollars to maintain academic strength and expansion may turn out to be the major challenge to academic pediatrics in the next decade. The future of academic pediatric growth in the United States may fall on the shoulders of those departments able to secure substantial philanthropic dollars, corporate investment, or medical school leaders willing to invest precious funds for the next decade. Ten years from now, we should not be surprised if pediatric academic medical prowess and leadership in the United States is concentrated in a handful of institutions that have been the beneficiaries of past and current philanthropy and commit themselves to developing the resources needed for academic achievement. Institutions unable to make strong commitments to the years ahead with real and substantive dollars may find themselves moving from asking “Where have all our young ones gone?” to asking “Where has our once vital and thriving department gone?” And this question may arrive much sooner than we imagine. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.04.018

REFERENCES 1. Goldman E, Marshall E. Research funding. NIH grantees: where have all the young ones gone? Science 2002;298:40-1. 2. Marks AR. Physician-scientist, heal thyself. J Clin Invest 2007;117:2. 3. Rossini AA. Passing the baton-to whom? J Clin Invest 2007;117(2):285-8. 4. Feigin RD. American Pediatric Society Presidential Address 1998: what is the future for academic pediatrics? Pediatr Res 1998;44:958-63. 5. http://grants.nih.gov/grants/award/awardtr.htm. 6. Gitterman DP, Greenwood RS, Kocis KC, Mayes BR, McKethan AN. Did a rising tide lift all boats? The NIH budget and pediatric research portfolio. Health Aff (Millwood) 2004;23:113-24. 7. www.ncbi.nlm.nih.gov/entrez/. Accessed April 12, 2007. 8. Olefsky JM. The US’s changing competitiveness in the biomedical sciences. J Clin Invest 2007;117:270-6. 9. North MA. Missions, monies and metrics. A funds flow to preserve the academic mission. 2005. In: Academic Practice Compensation and Production Survey for Faculty and Management: 2005 report based on 2004 data. 10. Jobe AH, Abramson JS, Batshaw M, Boxer LA, Lister G, McCabe E, et al. Recruitment and development of academic pediatricians: departmental commitments to promote success. Pediatr Res 2002;51:662-4. 11. Rangel SJ, Moss RL. Recent trends in the funding and utilization of NIH career development awards by surgical faculty. Surgery 2004;136:232-9. 12. www.med.yale.edu/pediat/pedsci/. Accessed April 12, 2007. 13. www.lwpes.org/. Accessed April 12, 2007.

The Journal of Pediatrics • September 2007

EDITORIALS

Are Patients with Kawasaki Disease at Risk for Premature Atherosclerosis?

awasaki, disease (KD) is a childhood vasculitis of unknown cause characterized by fever, rash, enanthem, conjunctival injection, extremity changes, cervical adenopathy, and laboratory test results reflecting intense systemic inflammation.1 First described in Japan in 1967, KD has been described worldwide among children of all races and ethnicities. In the United States, more than 4000 hospitalizations associated with KD were reported in 2000,2 and KD has replaced rheumatic fever as the leading cause of acquired heart disease in children.3 Clinical and epidemiologic features suggest an infectious trigger, with expression of clinical disease likely modified by genetic susceptibility.2 Conventional therapy for KD includes administration of aspirin and intravenous gamma globulin (IVIG) within the first 10 days of illness, and ideally within the first week. The goal of therapy in the acute phase of KD is to reduce inflammation in the coronary artery wall and prevent coronary artery thrombosis. The acute signs and symptoms of KD are self limited, and the disease only rarely recurs. However, the vascular inflammation that accompanies this disease is diffuse and may have long-term sequelae. The most severely affected children have coronary artery aneurysms that can lead to myocardial infarction, ischemic cardiomyopathy, and sudden death.4,5 Coronary artery aneurysms defined by Japanese Ministry of Health criteria occur in up to 25% of untreated children; treatment with high-dose IVIG in the acute phase of the disease reduces the risk of aneurysms by approximately 5-fold.6 More subtle coronary artery dilation occurs among those who do not meet Japanese Ministry of Health criteria, however. When coronary artery dimensions are adjusted for body surface area, more than 1 in 4 children classified as having normal coronary arteries by Japanese Ministry of Health criteria have at least 1 coronary artery dimension more than 2 standard deviations above the expected mean.7 Thus coronary artery dilation in KD is even more prevalent than originally suspected. Clinical or subclinical inflammation of the coronary and systemic arteries may form the substrate for longer-term functional and structural abnormalities and increase the risk of premature atherosclerosis.8-10 A number of noninvasive methods have been developed to study endothelial function and structural changes suggestive of atherosclerosis. Brachial artery flow–mediated dilation has been studied

K

IMT IVIG KD

Editorials

Intima-medial thickness Intravenous gamma globulin Kawasaki disease

widely and can be safely applied to large and varied groups of patients including children.11 The brachial artery dilation response to increased shear stress is mainly due to endothelial release of nitric oxide and correlated with coronary endothelial function.12 An alternative noninvasive method is measurement of arterial stiffness by pulsed-wave analysis or arterial tonometry, which is now recognized as important in predicting coronary artery disease.13 Structural arterial abnormalities are indicated by increased thickness of the intimal-medial portion of the carotid artery measured by B-mode ultrasonography. Increased carotid artery intima-medial thickness (IMT) has been shown to reliably indicate the presence of atherosclerosis.14 Tests of arterial structure and function have been applied to patients with a history of KD with and without detectable coronary artery aneurysms in the acute phase of the illness. This issue of The Journal of Pediatrics includes 2 small studies with conflicting inferences about arterial health after KD. Dalla Pozza et al15 compared carotid artery IMT among 48 patients with KD and 28 control subjects of similar age and sex. Carotid artery IMT, expressed as both unadjusted dimension and z-score, was greater among patients with KD than control subjects; within the KD group, the 15 patients with a history of coronary artery aneurysms had greater carotid artery IMT than the 5 children without coronary artery lesions. Patients with KD and control subjects had similar baroreceptor sensitivity and levels of established risk factors for adult atherosclerotic heart disease, including body mass index, blood pressure, and lipid profile. These authors infer that patients with KD have subclinical atherosclerosis and may be at risk even in the absence of persistent coronary artery abnormalities. In contrast, McCrindle et al16 report that 52 patients with KD, compared with 60 healthy control subjects, had similar systemic endothelial function, assessed by flow-mediated brachial artery reactivity. Furthermore, flow-mediated dilation was not significantly related to either patient or KD characterSee related articles, istics, similar findings to p 239 and p 244 those in the authors’ earlier report with fewer paReprint requests: not available. tients.17 In the past, these J Pediatr 2007;151:225-8 authors reported that pa0022-3476/$ - see front matter tients with KD had a Copyright © 2007 Mosby Inc. All rights more adverse cardioreserved. vascular risk profile, 10.1016/j.jpeds.2007.05.011 225

with higher blood pressure and greater adiposity, compared with control children.17 In the current study, few differences in atherosclerotic risk factors were found between patients with KD and healthy control subjects, with the exception that patients with KD in this study had significantly lower apolipoprotein A1 and hemoglobin A1c levels, as well as lower blood pressure, but with less nocturnal decline. Markers of the systemic inflammatory response were not measured. To help to interpret the importance of these contradictory manuscripts, it is useful to place them in the context of prior work in the field. We will review evidence for vascular changes according to coronary artery status, because the severity of vasculitis might be expected to affect the future risk of atherosclerotic vascular disease. Patients with persistent coronary artery aneurysms have suffered the most severe arterial insult.18 In such patients, compared with control patients, the carotid arterial wall has been reported to have a higher IMT and lower distensibility,19,20 although others have not confirmed these findings.21 Abnormalities of arterial function have also been reported. Ikemoto et al21 demonstrated endothelial dysfunction, as indicated by decreased brachial artery flow–mediated dilation, in patients with persistent coronary artery lesions. Two earlier studies on endothelial function in patients with KD reported similar results.22,23 Patients with persistent aneurysms have been shown to have ongoing systemic inflammation years after disease onset, as evidenced by C-reactive protein levels that are significantly higher than those seen in normal age-matched children or among patients with KD without aneurysms or with regressed aneurysms.24 Inflammatory mediators, such as C-reactive protein, may themselves promote atherosclerosis.25 Patients whose aneurysms have regressed to normal diameter represent an intermediate group. By 2 years after disease onset, approximately half of coronary artery aneurysms will have regressed to normal lumen diameter on angiography.4,26 Regressed aneurysms are characterized by fibrous intimal thickening on histopathologic examination27-29 and by marked symmetric or asymmetric myointimal thickening on intravascular ultrasonography.30,31 Indeed, initial coronary artery dimension has been shown to be highly related to coronary artery IMT by intravascular ultrasonography more than 10 years later.32 In addition to abnormal vascular structure, regressed coronary artery aneurysms have abnormal endothelial function, with reduced vascular reactivity to isosorbide dinitrate and constriction with acetylcholine.33-36 The proximal and peripheral arterial beds have also been reported to be stiffer among patients with KD with persistent or regressed aneurysms than in normal control subjects,20,37,38 with aortic pressure waveforms late after illness onset resembling those in the elderly.37 In the era of IVIG therapy, most children with KD do not have development of coronary artery aneurysms. With careful late clinical follow-up, such patients have morbidity and mortality rates that are similar to those in the normal population.39 However, data are conflicting on preclinical vascular changes in patients with KD in whom coronary abnormalities were never detected. Some studies in this subgroup have shown preclinical 226

Editorials

abnormalities in endothelial function, arterial stiffness, and myocardial flow reserve.20-22,40-43 For example, compared with normal subjects, they have been reported to have depressed endothelium-dependent brachial artery reactivity,22,23 as well as higher brachial-radial artery mean pulse wave velocity, suggesting increased arterial stiffness.44,45 Others have reported endothelial dysfunction only among patients with persistent coronary artery lesions,21 and that endothelial dysfunction is worst among those with coronary artery aneurysms.23 With respect to structural abnormalities, data are once again conflicting. Some investigators have found no difference in carotid artery IMT between patients with KD and control subjects,21,23 consistent with the hypothesis that functional abnormalities might precede those of structure. In contrast, Cheung et al20 found increased carotid artery IMT even among patients with KD with normal coronary arteries, compared with control subjects. Cardiac catheterization studies also have been conflicting with regard to whether endothelium-dependent relaxation is impaired in “normal” epicardial coronary artery segments of patients with KD.46,47 Of note, patients with KD without a history of coronary artery dilation appear to have lower myocardial flow reserve and higher total coronary artery resistance than control subjects.42,43 The only immunohistochemical study of the coronary arteries of a patient with KD without coronary dilation was performed in a child who died of unrelated causes48; compared with control subjects, the coronary artery intima was mildly thickened, and platelet-derived growth factor–␣, transforming growth factor–␤1, and inducible nitric oxide synthase were expressed in the intimal smooth muscle cells. How can we reconcile the conflicting literature on longterm vascular health among patients who have had KD, including the two most recent contributions in The Journal?15,16 Studies of arterial structure and function in KD are handicapped by small sample sizes and limited power; similar studies in adults characteristically include hundreds and even thousands of patients.49-51 Statistical significance can be reached only when differences between groups are large or sources of variance other than KD-related vascular changes are small. Unfortunately, potential sources of variation are numerous and include both technical factors associated with test performance and patient characteristics.11 Among patient characteristics that influence vascular health, dyslipidemia is prevalent in patients with KD with or without overt coronary artery sequelae well beyond the time that the clinical disease has resolved.38,52,53 Other patient factors influencing vascular health include hypertension, diabetes mellitus, smoking, obesity, systemic inflammation, age, pubertal status, and sex. It is impossible to adjust for all of these factors in a small series of patients. Furthermore, the higher prevalence of risk factors for future atherosclerotic coronary artery disease among youth in North America compared with in Japan and other Asian countries could have affected the ability of McCrindle et al16 to detect vascular changes related to KD. It is unlikely that a large international study of vascular health in children with KD will be performed in the near future. Thus multiple small studies must be viewed in the aggregate to The Journal of Pediatrics • September 2007

assess the arterial health and guide management of patients with KD. Among those with persistent or regressed coronary artery aneurysms, coronary artery structure and function are well documented to be impaired; therefore the presence or absence of abnormalities in other systemic arteries does not affect their need for aggressive management of other risk factors. Further investigation is needed, however, before conclusions can be reached regarding the impact of KD on vascular health among those in whom coronary artery changes were never detected. Indeed, we will not know with certainty whether “always normal” patients with KD are at higher risk for atherosclerosis until early Japanese cohorts reach middle and older age.54 Until published data allow evidence-based practice, all patients with a history of KD should be carefully assessed for risk factors for future atherosclerotic heart disease, including dyslipidemia, hypertension, smoking, obesity, diabetes mellitus, and sedentary lifestyle. Guidance for clinicians is provided by recent American Heart Association recommendations for cardiovascular risk reduction, with thresholds for counseling and pharmacologic management in patients with KD tailored to the degree of coronary artery involvement.55 Elif Seda Selamet Tierney, MD Jane W. Newburger, MD, MPH Children’s Hospital Department of Cardiology Boston, Massachusetts

REFERENCES 1. Kawasaki T. Acute febrile mucocutaneous syndrome with lymphoid involvement with specific desquamation of the fingers and toes in children. Arerugi 1967; 16:178-222. 2. Burns JC, Glode MP. Kawasaki Syndrome. Lancet 2004;364:533-44. 3. Taubert KA, Rowley AH, Shulman ST. Nationwide survey of Kawasaki disease and acute rheumatic fever. J Pediatr 1991;119:279-82. 4. Kato H, Sugimura T, Akagi T, Sato N, Hashino K, Maeno Y, et al. Long-term consequences of Kawasaki disease. A 10- to 21-year follow-up study of 594 patients. Circulation 1996;94:1379-85. 5. Dajani AS, Taubert KA, Gerber MA, Shulman ST, Ferrieri P, Freed M, et al. Diagnosis and therapy of Kawasaki disease in children. Circulation 1993;87:1776-80. 6. Newburger JW. Kawasaki disease. Curr Treat Options Cardiovasc Med 2000;2:227-36. 7. de Zorzi A, Colan SD, Gauvreau K, Baker AL, Sundel RP, Newburger JW. Coronary artery dimensions may be misclassified as normal in Kawasaki disease. J Pediatr 1998;133:254-8. 8. Ross R. Atherosclerosis—an inflammatory disease. N Engl J Med 1999; 340:115-26. 9. Libby P, Ridker PM, Maseri A. Inflammation and atherosclerosis. Circulation 2002;105:1135-43. 10. Pearson TA, Mensah GA, Alexander RW, Anderson JL, Cannon RO, III, Criqui M, et al. Markers of inflammation and cardiovascular disease: application to clinical and public health practice: a statement for healthcare professionals from the Centers for Disease Control and Prevention and the American Heart Association. Circulation 2003;107:499-511. 11. Deanfield JE, Halcox JP, Rabelink TJ. Endothelial function and dysfunction: testing and clinical relevance. Circulation 2007;115:1285-95. 12. Anderson TJ, Uehata A, Gerhard MD, Meredith IT, Knab S, Delagrange D, et al. Close relation of endothelial function in the human coronary and peripheral circulations. J Am Coll Cardiol 1995;26:1235-41. 13. Weber T, Auer J, O’Rourke MF, Kvas E, Lassnig E, Berent R, et al. Arterial stiffness, wave reflections, and the risk of coronary artery disease. Circulation 2004;109:184-9. 14. Heiss G, Sharrett AR, Barnes R, Chambless LE, Szklo M, Alzola C. Carotid atherosclerosis measured by B-mode ultrasound in populations: associations with cardiovascular risk factors in the ARIC study. Am J Epidemiol 1991;134:250-6. 15. Dalla Pozza R, Bechtold S, Urschel S, Kozlik-Feldmann R, Netz H. Subclinical atherosclerosis, but normal autonomic function after Kawasaki syndrome. J Pediatr 2007;151:239-43.

Editorials

16. McCrindle BW, McIntyre S, Kim C, Lin T, Adeli K. Are patients after Kawasaki Disease at increased risk for accelerated atherosclerosis? J Pediatr 2007;151:244-8. 17. Silva AA, Maeno Y, Hashmi A, Smallhorn JF, Silverman ED, McCrindle BW. Cardiovascular risk factors after Kawasaki disease: a case-control study. J Pediatr 2001;138:400-5. 18. Newburger JW, Takahashi M, Gerber MA, Gewitz MH, Tani LY, Burns JC, et al. Diagnosis, treatment, and long-term management of Kawasaki disease: a statement for health professionals from the Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease, Council on Cardiovascular Disease in the Young, American Heart Association. Pediatrics 2004;114:1708-33. 19. Noto N, Okada T, Yamasuge M, Taniguchi K, Karasawa K, Ayusawa M, et al. Noninvasive assessment of the early progression of atherosclerosis in adolescents with Kawasaki disease and coronary artery lesions. Pediatrics 2001;107:1095-9. 20. Cheung YF, Wong SJ, Ho MH. Relationship between carotid intima-media thickness and arterial stiffness in children after Kawasaki disease. Arch Dis Child 2007;92:43-7. 21. Ikemoto Y, Ogino H, Teraguchi M, Kobayashi Y. Evaluation of preclinical atherosclerosis by flow-mediated dilatation of the brachial artery and carotid artery analysis in patients with a history of Kawasaki disease. Pediatr Cardiol 2005;26:782-6. 22. Dhillon R, Clarkson P, Donald AE, Powe AJ, Nash M, Novelli V, et al. Endothelial dysfunction late after Kawasaki disease. Circulation 1996;94:2103-6. 23. Kadono T, Sugiyama H, Hoshiai M, Osada M, Tan T, Naitoh A, et al. Endothelial function evaluated by flow-mediated dilatation in pediatric vascular disease. Pediatr Cardiol 2005;26:385-90. 24. Mitani Y, Sawada H, Hayakawa H, Aoki K, Ohashi H, Matsumura M, et al. Elevated levels of high-sensitivity C-reactive protein and serum amyloid-A late after Kawasaki disease: association between inflammation and late coronary sequelae in Kawasaki disease. Circulation 2005;111:38-43. 25. Libby P, Ridker PM. Inflammation and atherosclerosis: role of C-reactive protein in risk assessment. Am J Med 2004;116(Suppl 6A):9S-16S. 26. Takahashi M, Mason W, Lewis AB. Regression of coronary aneurysms in patients with Kawasaki syndrome. Circulation 1987;75:387-94. 27. Tanaka N, Naoe S, Masuda H, Ueno T. Pathological study of sequelae of Kawasaki disease (MCLS). With special reference to the heart and coronary arterial lesions. Acta Pathol Jpn 1986;36:1513-27. 28. Fujiwara H, Hamashima Y. Pathology of the heart in Kawasaki disease. Pediatrics 1978;61:100-7. 29. Sasaguri Y, Kato H. Regression of aneurysms in Kawasaki disease: a pathological study. J Pediatr 1982;100:225-31. 30. Sugimura T, Kato H, Inoue O, Fukuda T, Sato N, Ishii M, et al. Intravascular ultrasound of coronary arteries in children. Assessment of the wall morphology and the lumen after Kawasaki disease. Circulation 1994;89:258-65. 31. Suzuki A, Yamagishi M, Kimura K, Sugiyama H, Arakaki Y, Kamiya T, et al. Functional behavior and morphology of the coronary artery wall in patients with Kawasaki disease assessed by intravascular ultrasound. J Am Coll Cardiol 1996;27:291-6. 32. Tsuda E, Kamiya T, Kimura K, Ono Y, Echigo S. Coronary artery dilatation exceeding 4.0 mm during acute Kawasaki disease predicts a high probability of subsequent late intima-medial thickening. Pediatr Cardiol 2002;23:9-14. 33. Kurisu Y, Azumi T, Sugahara T, Igarashi Y, Takamiya M, Kozuka T. Variation in coronary arterial dimension (distensible abnormality) after disappearing aneurysm in Kawasaki disease. Am Heart J 1987;114:532-8. 34. Matsumura K, Okuda Y, Ito T, Hirano T, Takeda K, Yamaguchi N. Coronary angiography of Kawasaki disease with the coronary vasodilator dipyridamole: assessment of distensibility of affected coronary arterial wall. Angiology 1988;39:141-7. 35. Sugimura T, Kato H, Inoue O, Takagi J, Fukuda T, Sato N. Vasodilatory response of the coronary arteries after Kawasaki disease: evaluation by intracoronary injection of isosorbide dinitrate. J Pediatr 1992;121:684-8. 36. Iemura M, Ishii M, Sugimura T, Akagi T, Kato H. Long term consequences of regressed coronary aneurysms after Kawasaki disease: vascular wall morphology and function. Heart 2000;83:307-11. 37. Senzaki H, Chen CH, Ishido H, Masutani S, Matsunaga T, Taketazu M, et al. Arterial hemodynamics in patients after Kawasaki disease. Circulation 2005;111: 2119-25. 38. Cheung YF, Ho MH, Tam SC, Yung TC. Increased high sensitivity C reactive protein concentrations and increased arterial stiffness in children with a history of Kawasaki disease. Heart 2004;90:1281-5. 39. Nakamura Y, Yanagawa H, Kato H, Harada K, Kawasaki T. Mortality among patients with a history of Kawasaki disease: the third look. The Kawasaki Disease Follow-up Group. Acta Paediatr Jpn 1998;40:419-23. 40. Albisetti M, Chan AK, McCrindle BW, Wong D, Vegh P, Adams M, et al. Fibrinolytic response to venous occlusion is decreased in patients after Kawasaki disease. Blood Coagul Fibrinolysis 2003;14:181-6. 41. Deng YB, Li TL, Xiang HJ, Chang Q, Li CL. Impaired endothelial function in the brachial artery after Kawasaki disease and the effects of intravenous administration of vitamin C. Pediatr Infect Dis J 2003;22:34-9.

227

42. Furuyama H, Odagawa Y, Katoh C, Iwado Y, Ito Y, Noriyasu K, et al. Altered myocardial flow reserve and endothelial function late after Kawasaki disease. J Pediatr 2003;142:149-54. 43. Muzik O, Paridon SM, Singh TP, Morrow WR, Dayanikli F, Di Carli MF. Quantification of myocardial blood flow and flow reserve in children with a history of Kawasaki disease and normal coronary arteries using positron emission tomography. J Am Coll Cardiol 1996;28:757-62. 44. Cheung YF, Yung TC, Tam SC, Ho MH, Chau AK. Novel and traditional cardiovascular risk factors in children after Kawasaki disease: implications for premature atherosclerosis. J Am Coll Cardiol 2004;43:120-4. 45. Ooyanagi R, Fuse S, Tomita H, Takamuro M, Horita N, Mori M, et al. Pulse wave velocity and ankle brachial index in patients with Kawasaki disease. Pediatr Int 2004;46:398-402. 46. Mitani Y, Okuda Y, Shimpo H, Uchida F, Hamanaka K, Aoki K, et al. Impaired endothelial function in epicardial coronary arteries after Kawasaki disease. Circulation 1997;96:454-61. 47. Yamakawa R, Ishii M, Sugimura T, Akagi T, Eto G, Iemura M, et al. Coronary endothelial dysfunction after Kawasaki disease: evaluation by intracoronary injection of Acetylcholine. J Am Coll Cardiol 1998;31:1074-80. 48. Suzuki A, Miyagawa-Tomita S, Komatsu K, Nakazawa M, Fukaya T, Baba K, et al. Immunohistochemical study of apparently intact coronary artery in a child after Kawasaki disease. Pediatr Int 2004;46:590-6. 49. Splansky GL, Corey D, Yang Q, Atwood LD, Cupples LA, Benjamin EJ, et al. The Third Generation Cohort of the National Heart, Lung, and Blood Institute’s

Framingham Heart Study: design, recruitment, and initial examination. Am J Epidemiol 2007 [Epub ahead of print]. 50. Vita JA, Keaney JF Jr, Larson MG, Keyes MJ, Massaro JM, Lipinska I, et al. Brachial artery vasodilator function and systemic inflammation in the Framingham Offspring Study. Circulation 2004;110:3604-9. 51. Ingelsson E, Sullivan LM, Murabito JM, Fox CS, Benjamin EJ, Polak JF, et al. Prevalence and prognostic impact of subclinical cardiovascular disease in individuals with the metabolic syndrome and diabetes. Diabetes 2007 [Epub ahead of print]. 52. Cabana VG, Gidding SS, Getz GS, Chapman J, Shulman ST. Serum amyloid A and high density lipoprotein participate in the acute phase response of Kawasaki disease. Pediatr Res 1997;42:651-5. 53. Newburger JW, Burns JC, Beiser AS, Loscalzo J. Altered lipid profile after Kawasaki syndrome. Circulation 1991;84:625-31. 54. Nakamura Y, Aso E, Yashiro M, Uehara R, Watanabe M, Tajimi M, et al. Mortality among persons with a history of Kawasaki disease in Japan: can paediatricians safely discontinue follow-up of children with a history of the disease but without cardiac sequelae? Acta Paediatr 2005;94:429-34. 55. Kavey RE, Allada V, Daniels SR, Hayman LL, McCrindle BW, Newburger JW, et al. Cardiovascular risk reduction in high-risk pediatric patients: a scientific statement from the American Heart Association Expert Panel on Population and Prevention Science; the Councils on Cardiovascular Disease in the Young, Epidemiology and Prevention, Nutrition, Physical Activity and Metabolism, High Blood Pressure Research, Cardiovascular Nursing, and the Kidney in Heart Disease; and the Interdisciplinary Working Group on Quality of Care and Outcomes Research: endorsed by the American Academy of Pediatrics. Circulation 2006;114:2710-38.

Non! to Non-Steroidal Anti-Inflammatory Therapy for Inflammatory Lung Disease in Cystic Fibrosis (at Least at the Moment)

here is generally a “love-hate” relationship between the inflammatory response and the human organism. On one hand, a major congenital or acquired defect in the recognition of, or response to, an environmental pathogen leads to recurrent severe infections and often rapid death. Conversely, failure to control the inflammatory response can lead to death and destruction from “friendly fire,” the most clear-cut examples being the formation of tissue-specific antibodies such as the anti-glomerular basement membrane antibody in Goodpasture’s syndrome.1 In the case of cystic fibrosis (CF), the relationship has been perceived as being more one of “hate-hate” rather then “lovehate.” The dogma has been that chronic airway infection has resulted in an over-exuberant inflammatory response, with the recruitment of excessive amounts of neutrophils, failure of clearance of the micro-organisms, neutrophil necrosis instead of apoptosis, and the release of tissue-damaging enzymes, with resultant tissue destruction disproportionate to the actual burden of infection.2 This idea led to the seemingly paradoxical concept that immunosuppression might be beneficial in the setting of chronic airway infection. The initial choice was prednisone, with the first study, using a huge dose (2 mg/kg on alternate days) reporting benefits, but apparently no adverse effects.3 This was

T

CF FEV1 FVC

228

Cystic fibrosis Forced expired volume Forced vital capacity

Editorials

not confirmed in the first of the great series of multicenter CF trials emanating from North America.4 In this study, 2 mg/kg of prednisone on alternate days was compared with 1 mg/kg on alternate days and with placebo. The well-known findings were that the benefit of prednisone was confirmed, but the adverse effects necessitated discontinuing the high- and low-dose prednisone arms after 2 and 4 years, respectively. An important clue, the significance of which has been under-appreciated, was that the benefits of prednisone were confined to those patients chronically infected with Pseudomonas aeruginosa. The subsequent history of steroids, in brief, is that after a number of contradictory studies of inhaled steroids,5 a trial of withdrawal of inhaled steroids (CF WISE study) showed that they are largely ineffective in CF.6 Furthermore, the appreciation has grown of the importance of systemic complications of See related article, p 249 CF, such as diabetes 7 mellitus and bone disease,8 which may be Reprint requests: Andrew Bush, Department of Paediatric Respiratory Medicine, worsened by steroids, and Royal Brompton Hospital, Sydney Street, so enthusiasm for steLondon SW3 6NP, UK. E-mail: a.bush@ rbh.nthames.nhs.uk. roids in CF (other than J Pediatr 2007;151:228-30 when mandated by, for 0022-3476/$ - see front matter example, allergic bronCopyright © 2007 Mosby Inc. All rights chopulmonary aspergilreserved. losis9) has waned. 10.1016/j.jpeds.2007.06.019 The Journal of Pediatrics • September 2007

However, 2 important lessons can be drawn from the history of steroids in CF. The first is the credulity of physicians who treat CF in their willingness to prescribe treatments of no value; CF WISE had to study withdrawal of inhaled steroids, not prescription of them, because so few patients could be found who were not already taking these medications. Second, from the multicenter prednisone study4 is that the stage of the disease may determine the response to antiinflammatory therapy. This makes biological sense; the child with CF at birth almost certainly has a sterile and normal airway and initially probably meets and repels pathogens, until eventually the defences are overwhelmed, and chronic infection supervenes. Thus it is likely, but unproven, that very early immunosuppression actually might accelerate the onset of chronic infection. The concept that immunosuppression may cause harm, not good, was further strengthened by the large multicenter study of the leukotriene (LT) B4 receptor antagonist B11L 284 BS.10 The trial was stopped by the datamonitoring committee because of an increase in serious adverse events (infective exacerbations) in the treatment limb. What then are the lessons for those wishing to study anti-inflammatory medications in CF? Key (and most difficult) must be to have a focussed hypothesis about the type of patient with CF who may benefit from the intervention, rather than trying it out on all comers. This is doubly difficult; we are as yet struggling to generate these hypotheses, and, even when we do so, finding enough patients to do an adequately powered study becomes even more problematic. In this context, how do we interpret the results of the use of ibuprofen as an anti-inflammatory medication in CF, in particular in the light of the study by Lands et al in this issue of The Journal?11 The logic for the studies is impeccable; if steroids are effective but have adverse effects, why not use non-steroidal anti-inflammatory medications? The initial carefully controlled study showed that ibuprofen slowed the rate of deterioration of first second forced expired volume (FEV1).12 The placebo group had an annual decline of ⫺3.60% ⫾ 0.55% against ⫺2.17% ⫾ 0.57% in the active group. The results were more dramatic in the group that was compliant with medication, and in those who were ⬍13 years old at the start of the trial. Despite this study, ibuprofen has only enjoyed patchy use in the clinic13; whether this is because of worries of albeit rare, but not trivial, adverse effects such as gastrointestinal haemorrhage and acute renal failure,14,15 the narrow therapeutic window necessitating monitoring of levels,16 or more cynically, because big pharmaceutical companies were not promoting it lavishly and assiduously, is unclear. However, 12 years later, one could be forgiven for questioning the relevance of the study because the rate of decline in lung function in the placebo group is much higher than would be acceptable now,17-19 and the intervention group results are inferior to those currently seen in some clinics that do not use much ibuprofen. This study11 recruited more patients (n ⫽ 142), but studied them during a shorter period (2 years). Patients with relatively mild impairment of lung function (FEV1 ⬎60% predicted) were recruited, but children who were ⬎13-year age cutoff of the earlier study were included. Editorials

Their power calculation was based on a high expected rate of decline in the placebo group (⌬FEV1 4%/year, higher than was actually seen), set the bar low at 80% power, and concluded that substantially more patients were needed than were recruited, despite a monumental effort by the investigators. They found no change in either their primary end point (rate of change of FEV1) or of another variable that one would expect to be affected by obstructive lung disease, mid-expiratory flow (FEF25-75). However, what they did find was a beneficial effect on forced vital capacity (FVC). In the ibuprofen group, FVC actually did not change in 2 years. After a prolonged post hoc pas de deux with the data, the authors also managed to torture out a small, statistically significant benefit for the patients who were treated in days spent in hospital. How then should ibuprofen be positioned in the therapeutic armamentarium? Lands et al11 have put in a tremendous effort to make an impeccably designed study work, for which unreserved congratulations are due. However, they fail to convince us that they have shown a biologically likely benefit. Their study was under-powered and failed to show an effect in their primary end point or a biologically plausible secondary end point. Are the changes in FVC biologically plausible in an obstructive lung disease or an artefact of a relatively small study? We do not believe that the case for widespread use of ibuprofen can be made on these data alone. What of the future? One lesson from this study is that the better we get at conventional treatment, the harder it will be to show an improvement in outcome with a novel therapy. The relative insensitivity of lung function, or rather, the huge numbers of patients needed if spirometry is to be an outcome, has been highlighted20,21; reliable surrogates, changes in which truly reflect the course of the disease, need to be found. In manipulating CF inflammation, we need to know at what time in the disease we should be repressing instead of boosting the host defenses; we need to know what part of the inflammatory cascade is responsible for the host damage; and we need to be able to measure it easily, repeatedly, and non-invasively. Only at that point will we be able to design rational studies of anti-inflammatory therapies with appropriate end points; after all, the asthma doctors have eventually realized that anti-inflammatory therapy is most effective when the therapeutic target, inflammation, is measured.22-26 Until that time, trials are likely to include patients who have no chance of benefiting from the proffered treatment, thus diluting any effect, and continue to be underpowered anyway, because lung function and inflammation may be only very loosely related. The concept of treating a genetic defect (premature stop codon) rather than a disease has recently gained practical currency27; we need almost certainly to move to treating specific pathways, not a global mish-mash of “inflammation.” However, it must be stressed that the absence of evidence of benefit is not the same as evidence of non-benefit. It may be that ibuprofen is the answer for some patients with CF, but we do not yet have the data. Although we disagree with their conclusions, we most readily acknowledge that Lands et al11 have done a signal service by keeping ibuprofen at the forefront of the debate about optimal treatment in CF, by conducting an impressively designed study, and above all, by reminding us of 229

the importance of clinical trial work in children, who may gain from therapies that are useless in adults. If pediatric trial work like this is not done, then substantial therapeutic benefits in the fight against CF may be frittered away. Andrew Bush, MB, BS (Hons), MA, MD, FRCP, FRCPCH Professor of Paediatric Respirology Imperial School of Medicine at National Heart and Lung Institute Honorary Consultant Paediatric Chest Physician Royal Brompton Hospital London, United Kingdom

Jane Davies, MB, ChB, MRCP, MRCPCH, MD Senior Lecturer in Gene Therapy Imperial College Honorary Consultant Paediatric Chest Physician Royal Brompton Hospital London, United Kingdom

REFERENCES 1. Borza DB, Neilson EG, Hudson BG. Pathogenesis of Goodpasture syndrome: a molecular perspective. Semin Nephrol 2003;23:522-31. 2. Konstan MW, Berger M. Infection and inflammation of the lungs in cystic fibrosis. In: Davis PB, editor. Cystic fibrosis. New York: Marcel Dekker; 1993. p. 219-76. 3. Williams HS, Auerbach M, Kirkpatrick JA, Colten HR. Alternate-day prednisone reduces morbidity and improves pulmonary function in cystic fibrosis. Lancet 1985;2:686-8. 4. Eigen H, Rosenstein BJ, FitzSimmons S, Schidlow DV. A multicenter study of alternate-day prednisone therapy in patients with cystic fibrosis. J Pediatr 1995;126:515-23. 5. Dezateux C, Walters S, Balfour-Lynn I. Inhaled corticosteroids for cystic fibrosis. Cochrane Database Syst Rev 2000;(2):CD001915. 6. Balfour-Lynn IM, Lees B, Hall P, Phillips G, Khan M, Flather M, et al. Multicenter randomized controlled trial of withdrawal of inhaled corticosteroids in cystic fibrosis. Am J Respir Crit Care Med 2006;173:1356-62. 7. Solomon MP, Wilson DC, Corey M, Kalnins D, Zielenski J, Tsui LC, et al. Glucose intolerance in children with cystic fibrosis. J Pediatr 2003;142:128-32. 8. Elkin SL, Vedi S, Bord S, Garrahan NJ, Hodson ME, Compston JE. Histomorphometric analysis of bone biopsies from the iliac crest of adults with cystic fibrosis. Am J Respir Crit Care Med 2002;166:1470-4. 9. Stevens DA, Moss RB, Kurup VP, Knutsen AP, Greenberger P, Judson MA, et al. Allergic bronchopulmonary aspergillosis in cystic fibrosis—state of the art: Cystic Fibrosis Foundation Consensus Conference. Clin Infect Dis 2003;37(Suppl 3):S225-64.

10. Konstan M, Doring G, Lands LC. Results of a phase 11 clinical trial of B11L 284 BS (an LTB4 receptor antagonist) for the treatment of CF lung disease. Pediatr Pulmonol 2005;(Suppl 28):125-6. 11. Lands LC, Milner R, Cantin AM, Manson M, Corey M. High-dose ibuprofen in cystic fibrosis: Canadian safety and effectiveness trial. J Pediatr 2007;151:249-54. 12. Konstan MW, Byard PJ, Hoppel CL, Davis PB. Effect of high dose ibuprofen in patients with cystic fibrosis. N Engl J Med 1995;322:848-54. 13. Balfour-Lynn IM, Dezateux C. Corticosteroids and ibuprofen in cystic fibrosis. Thorax 1999;54:657. 14. Kovesi TA, Swartz R, MacDonald N. Transient renal failure due to simultaneous ibuprofen and aminoglycoside therapy in children with cystic fibrosis. N Engl J Med 1998;338:65-6. 15. Scott CS, Retsch-Bogart GZ, Henry MM. Renal failure and vestibular toxicity in an adolescent receiving gentamicin and standard-dose ibuprofen. Pediatr Pulmonol 2001;31:314-6. 16. Rinaldo JE, Pennock B. Effects of ibuprofen on endotoxin-induced alveolitis: biphasic dose response and dissociation between inflammation and hypoxemia. Am J Med Sci 1986;291:29-38. 17. Merkus PJ, Tiddens HA, de Jongste JC. Annual lung function changes in young patients with chronic lung disease. Eur Respir J 2002;19:886-91. 18. Merkus PJ, Govaere ES, Hop WC, Stam H, Tiddens HA, de Jongste JC. Preserved diffusion capacity in children with cystic fibrosis. Pediatr Pulmonol 2004;37:56-60. 19. Que C, Cullinan P, Geddes D. Improving rate of decline of FEV1 in young adults with cystic fibrosis. Thorax 2006;61:155-7. 20. Davis PB, Byard PJ, Konstan MW. Identifying treatments that halt progression of pulmonary disease in cystic fibrosis. Pediatr Res 1997;41:161-5. 21. Alton EW, Davies JC, Geddes DM. Biomarkers for cystic fibrosis: are we progressing? Am J Respir Crit Care Med 2007;175:750-1. 22. Green RH, Brightling CE, McKenna S, Hargadon B, Parker D, Bradding P, et al. Asthma exacerbations and sputum eosinophil counts: a randomised controlled trial. Lancet 2002;360:1715-21. 23. Pijnenburg MW, Bakker EM, Hop WC, De Jongste JC. Titrating steroids on exhaled nitric oxide in children with asthma: a randomized controlled trial. Am J Respir Crit Care Med 2005;172:831-6. 24. Smith AD, Cowan JO, Brassett KP, Herbison GP, Taylor DR. Use of exhaled nitric oxide measurements to guide treatment in chronic asthma. N Engl J Med 2005;352:2163-73. 25. Zacharasiewicz A, Wilson N, Lex C, Erin EM, Li AM, Hansel T, et al. Clinical use of noninvasive measurements of airway inflammation in steroid reduction in children. Am J Respir Crit Care Med 2005;171:1077-82. 26. Pijnenburg MW, Hofhuis W, Hop WC, De Jongste JC. Exhaled nitric oxide predicts asthma relapse in children with clinical asthma remission. Thorax 2005;60:215-8. 27. Welch EM, Barton ER, Zhuo J, Tomizawa Y, Friesen WJ, Trifillis P. PTC124 targets genetic disorders caused by nonsense mutations. Nature 2007;447:87-91.

What is the Role of Cystic Fibrosis Transmembrane Conductance Regulator Dysfunction in Primary Sclerosing Cholangitis?

rimary sclerosing cholangitis (PSC) is a devastating and insidiously progressive cholestatic liver disease resulting from progressive inflammation, fibrosis, and obliteration of the intrahepatic and extrahepatic bile ducts.1 It is a relatively uncommon disorder, with an approximate annual incidence of 1 per 100,000. Most adult patients (⬎70%) have or will develop

P CF CFTR DHA IBD NTPD PPAR-␣ PSC STOPSC

230

Cystic fibrosis Cystic fibrosis transmembrane conductance regulator Docosahexaenoic acid Inflammatory bowel disease Nasal transmembrane potential difference Peroxisome proliferator-activated receptor-␣ Primary sclerosing cholangitis Studies of Primary Sclerosing Cholangitis

Editorials

inflammatory bowel disease (IBD), usually ulcerative colitis, and approximately 5% of patients with IBD may develop PSC. Ultimately, PSC leads to cirrhosis and end-stage liver disease necessitating transplantation. Cholangiocarcinoma is a dreaded and often fatal complication.

See related article, p 255 Reprint requests: Dennis D. Black, Department of Pediatrics, University of Tennessee Health Science Center, Children’s Foundation Research Center of Memphis, Le Bonheur Children’s Medical Center, Room 401, West Tower, 50 North Dunlap, Memphis, TN 38103. E-mail: [email protected]. J Pediatr 2007;151:230-2 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.05.025

The Journal of Pediatrics • September 2007

Although predominantly an adult disease, PSC affects children as well.2 The prognosis may be somewhat better in children than in adults, because dominant strictures, recurrent cholangitis, and cholangiocarcinoma are uncommon in children. However, approximately 1/3 of pediatric patients require transplantation by adulthood. There is no satisfactory treatment for PSC. Although high-dose oral ursodeoxycholic acid therapy improves biochemical measures, it does not appear to alter clinical outcome.3 PSC is associated with autoantibodies, occurs in the setting of IBD, and may occasionally present as an overlap syndrome with autoimmune hepatitis. However, PSC does not behave as a typical autoimmune disease and generally responds poorly to immunosuppressive therapy, although a subset of pediatric patients may demonstrate a response.4 The etiology of PSC remains a mystery but is probably multifactorial. Several lines of evidence from animal models and in vitro studies suggest a process of immune dysregulation in the setting of genetic predisposition. The process is initiated by an acute or chronic insult (eg, portal bacteremia in IBD), which triggers an immune response within the liver targeting the cholangiocytes, with resultant chronic inflammation.1 Although hepatic immune cells, such as Kupffer cells, may be major players in this process, it is now clear that the cholangiocyte itself is susceptible to activation. This “reactive” cholangiocyte phenotype acquires the ability to secrete proinflammatory cytokines and chemotactic molecules and is an active participant in the inflammatory process.5 Cholangiocytes generate nitric oxide in response to proinflammatory cytokines, which in turn inhibits cAMP-dependent secretion, including that mediated by cystic fibrosis (CF) transmembrane conductance regulator (CFTR), further contributing to decreased bile flow.6 Of recent interest are genes that may predispose to PSC, participate in the disease process, modulate disease severity, and/or influence the response to therapy and prognosis. Candidate genes include HLA haplotypes, biliary transporter genes (eg, mdr3), genes that modulate host–bacteria interactions (eg, nod2), liver disease– modifying genes (eg, alpha-1-antitrypsin), and inflammatory mediator genes (eg, tumor necrosis factor-␣ gene-promoter polymorphisms). However, the association of CFTR dysfunction and mutations in the CFTR gene with PSC is of particular interest. The CFTR gene product is a cAMP-regulated chloride channel expressed in diverse tissues, including respiratory tract, intestine, pancreas, sweat glands, male reproductive tract, and the hepatic canalicular and cholangiocyte membranes. Depending on their location and zygosity, CFTR gene mutations may have a clinical spectrum ranging from asymptomatic to severe illness, as well as a differing predilection for specific organs. There is compelling clinical and experimental evidence linking CFTR dysfunction to PSC. There are similarities between liver disease seen in patients with CF and PSC, including chronic inflammation, bile duct injury, and progressive fibrosis. Presumably, thickened, inspissated bile in CF causes obstruction and inflammation, with resultant injury of bile duct epithelium. In cftr⫺/⫺ mice with experimentally induced acute colitis, elevated serum alkaline phosphatase levels and histological bile duct injury developed.7 Interestingly, Editorials

although both cftr⫺/⫺ and wild-type mice exhibited suppressed peroxisome proliferator-activated receptor-␣ (PPAR-␣) expression in liver with colitis induction, mRNA levels later increased in the wild-type but not in the cftr⫺/⫺ animals, concomitant with development of bile duct injury. PPAR-␣ recently has been recognized as an important antiinflammatory immunomodulator. Treatment with the longchain polyunsaturated fatty acid docosahexaenoic acid (DHA) restored PPAR-␣ expression in the cftr⫺/⫺ mice and prevented bile duct injury.8 The protective effects of DHA may be related to its role as a PPAR-␣ agonist, as well as to other anti-inflammatory properties. In a study of a cftr⫺/⫺ mouse model that develops CF-like disease of all organs, including liver, DHA treatment specifically and significantly reduced hepatic periportal inflammation without effect on other organs.9 A human pilot study to assess the impact of DHA treatment in adults with CFTR mutations and PSC is currently underway. The association of CFTR mutations and PSC has been studied in adults with conflicting results.10-13 However, the negative studies tended to use a small sample size or to screen for only a limited number of mutations. The article by Pall et al14 in this issue of The Journal is the first report in a pediatric population of patients with PSC. Their data demonstrate that CFTR function in these patients with PSC, as assessed by sweat chloride analysis and nasal transmembrane potential difference (NTPD), is intermediate between nonPSC IBD disease control and classic CF values. A major strength of this study is the evaluation of CFTR function by 2 methods, the classic sweat chloride test and the more sensitive (and technically challenging) NTPD measurement. Few gene products are this accessible in living humans for in vivo functional analysis, but NTPD testing cannot be performed easily in young children. The comprehensive genetic analysis is another strength, although the results were not conclusive despite identification of various mutations (CF-causing), variants (associated with decreased CFTR function and/or non-CF CFTR-defective diseases), and polymorphisms (not linked to specific diseases) in a high percentage of both PSC and disease control patients. This finding differs from that of an adult study by this same group showing significantly higher frequencies of mutations and variants in PSC patients.12 There are several possible explanations for the discrepancy. First, it is possible that some of the IBD controls may have early, clinically silent PSC or may be predisposed to develop PSC later. Hopefully, this cohort will be followed into adulthood. Second, other genes undoubtedly contribute to the CFTR-deficient PSC phenotype, including those that modulate the inflammatory response as well as liver disease modifiers. Expanded genetic analysis to include these genes in the future may help clarify this issue. Finally, the present study involves a relatively small sample; a much larger number of subjects is needed to provide a more definitive answer. Another intriguing relationship is that between CFTR function and IBD, given the strong association of IBD with PSC. Such factors as mucosal permeability and bacterial flora are thought to be important for the portal access of bacteria and their products, such as LPS and CpG DNA, to possibly 231

contribute to PSC pathogenesis. In the intestine, CFTR is a modulator of permeability, mucus production, and interactions with bacteria. Therefore, CFTR dysfunction in colonic mucosa, as well as biliary epithelium, may contribute to development of PSC. A recent study from Europe demonstrated an association of heterozygosity of the CFTR ⌬F508 mutation with a reduced incidence of Crohn’s disease, especially right-sided colitis.15 Therefore, studies of CFTR gene mutations and PSC should include non-IBD controls, as well as both normal and non-PSC liver disease controls. A recent PSC conference jointly sponsored by NIDDK, the Office of Rare Diseases, and the Morgan Foundation stressed the need for genetic studies in PSC to increase our understanding of the pathogenesis, disease course and prognosis, and response to potential new therapies.1 Both pediatric and adult studies are crucial, because there may be age-related differences in factors such as the relative role of autoimmunity, clinical course, prognosis, and response to treatment, especially immunomodulatory therapy. Larger-scale genetic studies are needed to identify other relevant genetic markers and clinical associations. A newly launched North American PSC registry and DNA repository, Studies of Primary Sclerosing Cholangitis (STOPSC) (www.STOPSC.org), comprises 19 pediatric and adult hepatology programs in 12 major medical centers. Hopefully, STOPSC will provide the number of subjects needed to power studies of this uncommon, but clinically important, disease. The report by Pall et al14 in this issue of The Journal is a groundbreaking beginning, but much more remains to be done. Dennis D. Black Department of Pediatrics University of Tennessee Health Science Center Children’s Foundation Research Center of Memphis Le Bonheur Children’s Medical Center Memphis, Tennessee

REFERENCES 1. LaRusso NF, Shneider BL, Black D, Gores GJ, James SP, Doo E, et al. Primary sclerosing cholangitis: summary of a workshop. Hepatology 2006;44:746-64. 2. Feldstein AE, Perrault J, El-Youssif M, Lindor KD, Freese DK, Angulo P. Primary sclerosing cholangitis in children: a long-term follow-up study. Hepatology 2003;38:210-7. 3. Olsson R, Boberg KM, de Muckadell OS, Lindgren S, Hultcrantz R, Folvik G, et al. High-dose ursodeoxycholic acid in primary sclerosing cholangitis: a 5-year multicenter, randomized, controlled study. Gastroenterology 2005;129:1464-72. 4. Gregorio GV, Portmann B, Karani J, Harrison P, Donaldson PT, Vergani D, et al. Autoimmune hepatitis/sclerosing cholangitis overlap syndrome in childhood: a 16-year prospective study. Hepatology 2001;33:544-53. 5. Lazaridis KN, Strazzabosco M, Larusso NF. The cholangiopathies: disorders of biliary epithelia. Gastroenterology 2004;127:1565-77. 6. Spirli C, Fabris L, Duner E, Fiorotto R, Ballardini G, Roskams T, et al. Cytokine-stimulated nitric oxide production inhibits adenylyl cyclase and cAMP-dependent secretion in cholangiocytes. Gastroenterology 2003;124:737-53. 7. Blanco PG, Zaman MM, Junaidi O, Sheth S, Yantiss RK, Nasser IA, et al. Induction of colitis in cftr⫺/⫺ mice results in bile duct injury. Am J Physiol Gastrointest Liver Physiol 2004;287:G491-6. 8. Pall H, Zaman MM, Andersson C, Freedman SD. Decreased peroxisome proliferator activated receptor alpha is associated with bile duct injury in cystic fibrosis transmembrane conductance regulator⫺/⫺ mice. J Pediatr Gastroenterol Nutr 2006;42:275-81. 9. Beharry S, Ackerley C, Corey M, Kent G, Heng YM, Christensen H, et al. Long-term docosahexaenoic acid therapy in a congenic murine model of cystic fibrosis. Am J Physiol Gastrointest Liver Physiol 2007;292:G839-48. 10. McGill JM, Williams DM, Hunt CM. Survey of cystic fibrosis transmembrane conductance regulator genotypes in primary sclerosing cholangitis. Dig Dis Sci 1996; 41:540-2. 11. Girodon E, Sternberg D, Chazouilleres O, Cazeneuve C, Huot D, Calmus Y, et al. Cystic fibrosis transmembrane conductance regulator (CFTR) gene defects in patients with primary sclerosing cholangitis. J Hepatol 2002;37:192-7. 12. Sheth S, Shea JC, Bishop MD, Chopra S, Regan MM, Malmberg E, et al. Increased prevalence of CFTR mutations and variants and decreased chloride secretion in primary sclerosing cholangitis. Hum Genet 2003;113:286-92. 13. Gallegos-Orozco JF, C EY, Wang N, Rakela J, Charlton MR, Cutting GR, et al. Lack of association of common cystic fibrosis transmembrane conductance regulator gene mutations with primary sclerosing cholangitis. Am J Gastroenterol 2005;100:874-8. 14. Pall H, Zielenski J, Jonas MM, Dasilva DA, Potvin KM, Yuan X-W, et al. Primary sclerosing cholangitis in childhood is associated with abnormalities in cystic fibrosis-mediated chloride channel function. J Pediatr 2007;151:255-9. 15. Bresso F, Askling J, Astegiano M, Demarchi B, Sapone N, Rizzetto M, et al. Potential role for the common cystic fibrosis deltaF508 mutation in Crohn’s disease. Inflamm Bowel Dis 2007;13:531-6.

Specific Immune Globulin Therapy for Prevention of Nosocomial Staphylococcal Bloodstream Infection in Premature Infants: Not What We Hoped for!

dvances in medical knowledge coupled with the proliferation of capable neonatal intensive care units have been instrumental in the survival of infants born prematurely. Yet, premature birth remains an important contributor to overall infant death (⬎30% of all infant deaths).1 In this issue of The

A CoNS IVIG LOS MRSA NRN VLBW

232

Coagulase-negative staphylococci Intravenous immune globulin Late-onset sepsis Methicillin-resistant Staphylococcus aureus Neonatal Research Network Very low birth weight

Editorials

Journal, DeJonge et al2 describe the important negative findings of a study that attempted to use a pathogen-specific antibody for prevention of nosocomial infections in very low birth weight infants (VLBW; ⬍1500 g; ⬍32 weeks gestation).

See related article, p 260 Reprint requests: M. Teresa de la Morena, MD, Division of Allergy and Immunology, University of Texas Southwestern Medical Center in Dallas, 1935 Motor St, Dallas, TX 75235. E-mail: maite.delamorena@ utsouthwestern.edu. J Pediatr 2007;151:232-4 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.06.031

The Journal of Pediatrics • September 2007

In 1986, in response to the growing population of preterm infants, the National Institutes of Child Health and Human Development created the Neonatal Research Network (NRN), a group of academic centers working in collaboration, with the goal of reducing neonatal morbidity and mortality rates and improving outcomes. Spurred on by this initiative, a series of carefully conducted observational studies and randomized clinical trials have been performed in the past decade that have not only allowed the recognition of risk factors but also provided an infrastructure to systematically analyze potential therapeutic interventions with adequate numbers of subjects. Consequently, as expected, infections are recognized as an important problem for VLBW infants.3 Since the early 1990s the NRN has been monitoring both early-onset sepsis (occurring before 72 hours of life) and late-onset sepsis (LOS; positive blood culture after 72 hours of life) in VLBW infants and its impact on comorbidities, survival, and long-term outcomes. The incidence of sepsis occurring early is 2% versus 25% if the child survives past 72 hours of life. The most common organisms causing LOS are Gram-positive pathogens (70%), and coagulase-negative staphylococci (CoNS) represented 48% of all infections.4,5 Taken together, VLBW infants with LOS have longer hospitalizations, high morbidity and mortality rates, and care costs resulting in a societal economic burden that reached $26.2 billion in 2005.6 Addressing prevention and treatment of LOS in these infants is imperative. Antibodies are the effector molecules of the adaptive humoral immune response. Their physiological function is defense against extracellular microbes and microbial toxins through mechanisms of opsonization, neutralization, complement activation, and antibody-dependent cellular toxicity. Recognizing that transplacental transfer of maternal antibodies occurs after 32 to 35 weeks gestation, it was reasonable to hypothesize a benefit for intravenous administration of immune globulin (IVIG) to VLBW infants to prevent or treat infection. Since the late 1980s, more than a dozen studies and meta-analyses on the benefits of immune globulin therapy have been published. In general, studies agree on the safety of appropriate doses of IVIG in such small infants. However, benefits and efficacy varied. Baker et al7 studied 588 infants (500 to 1750 g) who were randomized to receive periodic infusions of IVIG. An important initial observation was a decrease in risk of first nosocomial infection in IVIG recipients compared with control subjects (RR 0.7; 95% CI) and trend to less necrotizing enterocolitis. Two years later, Weisman et al8 studied 753 babies (500-2000 g; ⬍34 weeks;⬍12 hours of life) who were randomized to receive a single infusion of IVIG. Despite maintaining higher immunoglobulin G levels, bloodstream infection, and overall morbidity and death caused by sepsis were not affected by IVIG. In 1994, Fanaroff et al9 in a collaborative NRN effort, published the results of a randomized placebo-controlled trial of IVIG (phase I/II) for 2416 infants (501-1500 g). IVIG failed to reduce the incidence of nosocomial infections. SubseEditorials

quently, studies and meta-analyses of use of IVIG for adjunctive therapy in infants with suspected/confirmed infection were published. These studies varied in IVIG product, age, birth weight, and sample size. When death was reported as an outcome, borderline statistical significance was noted in groups treated with IVIG compared with placebo (RR 0.63; 95% CI; 0.40, 1.00).10 In spite of the lack of clear benefit of IVIG, previous successful experience with pathogen-specific immune globulin for respiratory syncytial virus prophylaxis11 suggested a potential role for the development of pathogen-specific IVIG preparations. Because CoNS is the most common organism causing LOS, staphylococcal binding proteins C1fA and SdrG were selected as possible targets. These proteins are surface adhesins, present on ⬎98% of Staphylococcus aureus and most strains of Staphylococcus epidermidis and play a critical role in the attachment of bacteria to host tissue, an important pathogenic step for entry. In 2005, Bloom et al12 published a phase II, multicenter, double-blind clinical trial of INH-A21, a plasma-derived, donor-selected polyclonal antistaphylococcal human immune globulin. The study was sponsored by the manufacturer and the Office of Orphan Product Development of the US Food and Drug Administration. Plasma concentrations of the staphylococcal-specific antibody necessary to achieve protection were unknown, so the trial was designed to identify a dose for future analysis. A total of 512 infants (500-1250 grams; ⬎72 hours to 7 days of age) were randomized to 250, 500, and 750 mg/kg/dose of INH-A21. No differences in incidence of any staphylococcal (CoNS or S. aureus) infections was identified. The study by DeJonge et al2 is the subsequent phase III analysis: A multicenter (95 different centers within the United States and Canada), randomized double-blind, placebo-controlled clinical trial. A total of 1983 neonates (500-1250 g) received either placebo or study drug INH-A21. Because the phase II study only demonstrated possible activity of INH-A21 against S. aureus LOS, this study’s primary outcome was LOS caused by S. aureus, with CoNS infections as a secondary outcome. Sepsis was clearly defined by the authors. Incidence of LOS due to S. aureus was no different between the groups (5% in the placebo vs 6% in the INHA21 group). No association was found between the number of infusions and infections. The lack of clinical benefit is disappointing, particularly as an increase in methicillin-resistant S. aureus (MRSA) strains emerged within the study neonatal intensive care units (23% of S. aureus were MRSA2). For the past decades, knowledge of the neonatal human immune response has derived from analyses of the immunologic status of the mother, the role of the placenta in the transfer of antibodies, the in vitro analysis of cord and newborn peripheral blood cell, from study of patients and inferred from studies using mouse models.13 This knowledge underestimates the complexity of the immune response. For example, the B-lymphocyte system is fully developed at birth; fetal bone marrow B lymphocyte pools are similar in size to those of adults and preterm infants are capable of forming specific antibodies with comparable 233

isotype diversity as adults.14 Control of staphylococcal infections requires not only humoral, but cellular and phagocytic responses for effective killing. The lack of efficacy of antibody therapies may relate to confounding mechanisms necessary for clearance of pathogens by the host. The extent to which neonatal deficiencies of neutrophil function, complement, or antibody contribute to the increased risk of infection remains unknown, even though these factors are important in vitro for opsonophagocytic killing of Escherichia coli, group B Streptococcus, and Candida species. Answers may lie in expanding our knowledge and therapeutic approaches to stimulate/enhance innate components of the immune response. Antimicrobial products, receptors capable of recognizing pathogen-associated molecular patterns, phagocytic cells, complement proteins, dendritic and natural killer cells are all essential components of the innate immune response and constitute the first line of defense against invading pathogens. The skin is the most important barrier against pathogens invasion. Epithelial cells are capable of secreting 2 classes of antimicrobial peptides: defensins and cathelicidins.15 Interestingly, during the third trimester of pregnancy, the fetus becomes covered by the vernix caseosa that contains antimicrobial peptides including ␣-defensins and LL-37, a human cathelicidin. Vernix extracts exhibited both antibacterial activity against gram-negative bacteria, and antifungal properties against Candida albicans.16 Intriguingly, psoriasis and atopic dermatitis, are 2 inflammatory skin conditions associated with skin breakdown. However, although infection is rarely associated with psoriasis, patients with atopic dermatitis are commonly infected with S. aureus. Human ␤-defensin 2 and the cathelicidin LL-37 appear to be strongly expressed in psoriasis and not in eczematous skin. Interleukin 13, produced under atopic conditions, suppresses the induction of these antimicrobial peptides.17 Furthermore, LL-37 has also been identified in the ductal epithelium of salivary and sweat glands, suggesting a role in the protection of the gland itself from microbial invasion.18 Finally, the molecular biology of staphylococcal infections provides hints for the interactions of innate and adaptive immune responses. In vitro and in vivo experiments demonstrate that exposure of S. aureus to host cells induces the antimicrobial products ␤-defensins and LL37/CAP-18 but vary among stains of S. aureus, with MRSA exhibiting lower susceptibility.19 The publication by The Journal of this well-conducted study with negative results is essential as the accumulation of new findings that do support perceived knowledge advance the field.

234

Editorials

M. Teresa de la Morena, MD Division of Allergy and Immunology University of Texas Southwestern Medical Center Dallas, Texas

REFERENCES 1. Callaghan WM, MacDorman MF, Rasmussen SA, Qin C, Lackritz EM. The contribution of preterm birth to infant mortality rates in the United States. Pediatrics 2006;118:1566-73. 2. DeJonge M, Bloom BD, et al. Phase III, randomized, double-blind, placebocontrolled multi-center clinical trial of safety and efficacy of INH-A21 for the prevention of nosocomial staphylococcal sepsis in premature infants. J Pediatr 2007;151:260-5. 3. Stoll BJ, Hansen N. Infections in VLBW infants: studies from the NICHD Neonatal Research Network. Semin Perinatol 2003;27:293-301. 4. Stoll BJ, Gordon T, Korones SB, Shankaran S, Tyson JE, Bauer CR, et al. Late-onset sepsis in very low birth weight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network. J Pediatr 1996;129:63-71. 5. Stoll BJ, Hansen N, Fanaroff AA, Wright LL, Carlo WA, Ehrenkranz RA, et al. Late-onset sepsis in very low birth weight neonates: the experience of the NICHD Neonatal Research Network. Pediatrics 2002;110(Pt 1):285-91. 6. Behrman RE, Butler AS, eds. Committee on Understanding Premature Birth and Assuring Healthy Outcomes, Board on Health Science Policy, Institute of Medicine. Preterm Births: Causes, Consequences, and Prevention. Brief Report July 2006. http:// www.iom.edu/object.file/Master/35/975/pretermbirth.pdf. Accessed August 8, 2007. 7. Baker CJ, Melish ME, Hall RT, Casto DT, Vasan U, Givner LB. Intravenous immune globulin for the prevention of nosocomial infection in low-birth-weight neonates. The Multicenter Group for the Study of Immune Globulin in Neonates. N Engl J Med 1992;327:213-9. 8. Weisman LE, Stoll BJ, Kueser TJ, Rubio TT, Frank CG, Heiman HS, et al. Intravenous immune globulin prophylaxis of late-onset sepsis in premature neonates. J Pediatr 1994;125(Pt 1):922-30. 9. Fanaroff AA, Korones SB, Wright LL, Wright EC, Poland RL, Bauer CB, et al. A controlled trial of intravenous immune globulin to reduce nosocomial infections in very-low-birth-weight infants. National Institute of Child Health and Human Development Neonatal Research Network. N Engl J Med 1994;330:1107-13. 10. Ohlsson A, Lacy JB. Intravenous immunoglobulin for suspected or subsequently proven infection in neonates. Cochrane Database Syst Rev 2004(1): p. CD001239. 11. Reduction of respiratory syncytial virus hospitalization among premature infants and infants with bronchopulmonary dysplasia using respiratory syncytial virus immune globulin prophylaxis. The PREVENT Study Group. Pediatrics 1997;99:93-9. 12. Bloom B, Schelonka R, Kueser T, Walker W, Jung E, Kaufman D, et al. Multicenter study to assess safety and efficacy of INH-A21, a donor-selected human staphylococcal immunoglobulin, for prevention of nosocomial infections in very low birth weight infants. Pediatr Infect Dis J 2005;24:858-66. 13. Bona C. Neonatal immunity. In: Rose NR, editor. Contemporary immunology. Totowa, NJ: Humana Press. 2005. 1-389. 14. Ballow M, Cates KL, Rowe JC, Goetz C, Desbonnet C. Development of the immune system in very low birth weight (less than 1500 g) premature infants: concentrations of plasma immunoglobulins and patterns of infections. Pediatr Res 1986;20:899-904. 15. Yang D, Chertov O, Bykovskaia SN, Chen Q , Buffo MJ, Shogan J, et al. Beta-defensins: linking innate and adaptive immunity through dendritic and T cell CCR6. Science 1999;286(5439):525-8. 16. Yoshio H, Tollin M, Gudmundsson GH, Lagercrantz H, Jornvall H, Marchini G, et al. Antimicrobial polypeptides of human vernix caseosa and amniotic fluid: implications for newborn innate defense. Pediatr Res 2003;53:211-6. 17. Ong PY, Ohtake T, Brandt C, Strickland I, Boguniewicz M, Ganz T, et al. Endogenous antimicrobial peptides and skin infections in atopic dermatitis. N Engl J Med 2002;347:1151-60. 18. Murakami M, Ohtake T, Dorschner RA, Schittek B, Garbe C, Gallo RL. Cathelicidin anti-microbial peptide expression in sweat, an innate defense system for the skin. J Invest Dermatol 2002;119:1090-5. 19. Komatsuzawa H, Ouhara K, Yamada S, Fujiwara T, Sayama K, Hashimoto K, et al. Innate defenses against methicillin-resistant Staphylococcus aureus (MRSA) infection. J Pathol 2006;208:249-60.

The Journal of Pediatrics • September 2007

Acute Viral Bronchiolitis: To Treat or Not to Treat—That Is the Question

cute viral bronchiolitis is one of the most common conditions caused by respiratory viruses in infants and young children. For decades, controversy has surrounded both the treatment of bronchiolitis in early life and its sequelae. Part of the confusion in the literature comes from there being no common definition of acute viral bronchiolitis that is used internationally. In the United Kingdom, Australia, and New Zealand, acute viral bronchiolitis is a term used for a condition characterized by the presence of tachypnea, infhyperinflation of the chest, and widespread fine end-inspiratory crackles (also called crepitations) heard on auscultation. Wheeze on expiration may or may not be present. This typical clinical pattern is generally seen in the first year of life, with most children requiring admission to hospital in the first 6 months of life. It is largely caused by obstruction of respiratory and terminal bronchioles by mucosal edema and mucus production caused by the viral infection with formation of fluid menisci in the bronchioles. The fine crackles are caused by the “popping open” of these small airways in late inspiration.1 The developmental stage of the lung in the first year of life, with poorly developed collateral ventilation between adjacent lung units, facilitates the development of widespread airway obstruction. In North America and parts of Europe, however, the term bronchiolitis is commonly used to describe any lower respiratory viral infection occurring in the first 2 years of life.2 In these older children, wheeze and bronchospasm may play a greater role in the disease pathogenesis. Traditionally, most studies of infants with acute viral bronchiolitis have involved infants requiring hospitalization, and the virus responsible for most cases has been the respiratory syncytial virus (RSV).3-5 Studies following populations of such infants have shown a substantial rate of respiratory problems up to the age of 5 to 6 years,6,7 but longer term follow-up suggests that these children do not have an increased rate of atopy or of persistent asthma in later life.7-9 The situation appears to be somewhat different when community-based cohorts are studied. Two recent studies10,11 have shown the major contribution of rhinoviruses (RV) to lower respiratory infection associated with wheeze (wLRI) in the first year of life. In both of these studies, RV was responsible for approximately 3-times as many wLRI as was RSV in the first year of life. The children in both of these studies have been observed to the age of 5 years, and wLRI associated with RV in the first year of life was a major risk factor for asthma at the age of 5 years. These recent reports raise significant doubts about

A

AAP RSV RV wLRI

Editorials

American Academy of Pediatrics Respiratory syncytial virus Rhinoviruses Lower respiratory infection associated with wheeze

RSV having a “special” role in inducing asthma and favor the “susceptible host” theory. Despite decades of study, the mechanisms underlying viral-induced airway obstruction are not clear. Some authors have suggested that RSV picks out susceptible hosts. Two cohort studies that measured lung function in early life before any significant viral infections had occurred reported that low pre-morbid lung function was a major risk factor for wheezing during a lower respiratory tract infection in early life.12,13 A genetic predisposition or environmental exposures, such as maternal smoking during pregnancy, that result in sluggish maturation of the fetal and neonatal immune systems14 may increase the risk of contracting infection with RSV and other viruses in early life.15 Although there is no doubt that admission to hospital with acute viral bronchiolitis is associated with recurrent respiratory problems during early childhood and is a major risk factor for asthma at 5 to 6 years of age, the association with long-term persistent asthma is much less clear. Acute viral bronchiolitis is associated with considerable acute morbidity and mortality, with associated economic and social imposts on the community. The cost of hospitalization for acute viral bronchiolitis in children ⬍1 year of age was estimated to exceed $700 million per year in the United States in 2001.16 Although mortality has been falling in the past decades,17,18 young children still die from acute viral bronchiolitis. The aim of treatment should therefore be to reduce mortality, to reduce the economic and social burden (decrease the length of stay in hospital and associated costs), and to reduce the long-term sequelae (recurrent respiratory problems and maybe persistent asthma). Part of the problem in designing and implementing effective treatments is a lack of understanding of the underlying disease pathogenesis. Treatment with ␤-adrenergic agents, including albuterol and epinephrine, anti-cholinergic agents, corticosteroids, and, more recently, leukotriene receptor antagonists during the acute or recovery phase have been tried with See related article, p 266 varying success. Acute viral bronchiolitis is Reprint requests: Professor Peter Sly, Telecharacterized by acute thon Institute for Child Health Research, Division of Clinical Sciences 100 Roberts inflammation of the reRd, Subiaco, WA6008, Australia. E-mail: spiratory and terminal [email protected]. bronchioles; the process J Pediatr 2007;151:235-7 includes edema, necro0022-3476/$ - see front matter sis of epithelial cells, Copyright © 2007 Mosby Inc. All rights reserved. production of mucus, 10.1016/j.jpeds.2007.05.041

235

and possibly some degree of bronchospasm. A variety of treatments designed to overcome the acute airway obstruction have been championed throughout the years. In the 1980s, bronchodilators, in particular ␤2-agonists, were championed. Numerous articles debated their benefits and adverse effects. Studies measuring lung function during the acute disease phase claimed improvements19 or deterioration in lung function, referred to as a “paradoxical effect.”20-22 Other studies showed no effect on lung function either way.23 The potential for a decrease in arterial oxygen saturation relating to a worsening of the already disturbed ventilation-perfusion balance in the lungs was recognized.24 Despite these physiological disturbances, bronchodilators were reported to result in an acute improvement in clinical score.25 The use of nebulized epinephrine has also been controversial. A multicenter randomized double-blind controlled trial conducted in Australia that included 194 children reported that the length of stay in the hospital was not reduced in the group treated with epinephrine.26 A recent meta-analysis published by the Cochrane Collaboration27 concluded that bronchodilators improved clinical scores in the short term, but at a penalty of increased costs and increased adverse effects. This analysis included studies using albuterol, ipratropium bromide, and epinephrine. The rate and duration of hospitalization was not significantly reduced in the group treated with bronchodilators versus the control group.27 The American Academy of Pediatrics (AAP) recommends that “bronchodilators should not be used routinely in the management of bronchiolitis.”28 Corticosteroids have also been favored as an acute treatment for infants hospitalized with acute viral bronchiolitis. This practice is more common in some parts of the world than others; it is uncommon in Australia,29 but common in many parts of the world.30 Despite an earlier review of 6 trials of steroid therapy supporting a small reduction (mean, 0.43 days) in hospital length of stay compared with placebo,31 a more recent and larger review has disputed this conclusion. A meta-analysis that included data from 1198 children between the ages of 0 and 30 months concluded that there was no decrease in length of stay or clinical score in infants and young children treated with systemic glucocorticoids when compared with children treated with placebo,30 thus questioning the efficacy of this mode of treatment. In addition, the use of inhaled steroids during the acute phase of bronchiolitis to reduce post-bronchiolitis wheeze has been questioned, with a systematic review not able to demonstrate a significant benefit.32 The AAP recommends that “corticosteroid medication should not be used routinely in the management of bronchiolitis.”28 In this issue of The Journal, Kuzik et al33 report a clinically relevant 26% reduction in the length of hospitalization in children with acute viral bronchiolitis treated with nebulized 3% hypertonic saline (2.6 ⫾ 1.9 days compared with 3.5 ⫾ 2.9 days with normal saline; P ⫽ .05). Although this study has some methodological deficiencies, it does provide some hope of a way forward. The use of nebulized hypertonic saline is not new. This treatment is commonly used in children with cystic fibrosis as an aid to physiotherapy by improving muco-ciliary clearance.34 It has also been previously shown to reduce the length of hospital236

Editorials

ization for infants with acute viral bronchiolitis when combined with epinephrine35 and to reduce symptom scores more rapidly in ambulatory patients with bronchiolitis when added to inhaled terbutaline.36 Kuzik et al33 randomized 47 infants to receive nebulized 3% hypertoninc saline and 49 infants to receive nebulized normal saline in addition to “regular therapy” prescribed by the infant’s attending physician. A major strength of this study was that although the age range included children as old as 12 months, most children (38/47 and 35/49) were aged 0 to 6 months. This makes the data presented by Kuzik et al33 directly relevant to those treating acute viral bronchiolitis in other parts of the world. Unfortunately, the study did not have sufficient power to demonstrate conclusively a treatment effect in this younger age group alone. The major weakness of this study was that most infants also received other, discredited treatments, including albuterol (37%), racemic epinephrine (23%), or inhaled steroids (3%). Despite the clear recommendations from the AAP on the basis of solid evidence from Cochrane reviews that these treatments should not be used routinely in the management of acute viral bronchiolitis, they clearly are still being used. The authors do discuss the potential for adverse effects from treatment with nebulized 3% hypertonic saline and correctly state that this is not likely to be a major problem in infants with acute viral bronchiolitis. The other issue they raise— whether there is additional benefit to combining a bronchodilator with 3% hypertonic saline— cannot be addressed by their study. The authors have, correctly, not reported the length of hospitalization separately in those children who received the study solution alone or in combination with other treatments; they would not have had sufficient power to make any valid comparisons. The authors highlight the need for further definitive studies. The results of this study, with the background literature, provide sufficient rationale for a study of nebulized 3% saline versus normal (0.9%) saline, excluding other treatments. However, the question of a synergistic interaction between 3% hypertonic saline and albuterol is worth considering. To this end, a study design with a double randomization (3% versus 0.9% saline and albuterol versus placebo-0.9% saline) would be feasible. We urge that such a study be restricted to children aged ⱕ6 months and that steroids, epinephrine, and antibiotics be “banned’ during the study period. The question is worth pursuing because, if treatment with nebulized 3% saline does reduce the length of hospitalization in infants with acute viral bronchiolitis, the economic and social benefits gained from an inexpensive therapy would be worthwhile. Claudia Calogero, MD Telethon Institute for Child Health Research Subiaco, WA Australia School of Pediatrics University of Florence Florence, Italy

Peter D. Sly, MBBS, MD, DSC, FRACP Telethon Institute for Child Health Research Subiaco, WA Australia

The Journal of Pediatrics • September 2007

REFERENCES 1. Sly PD, Collins RA. Physiological basis of respiratory signs and symptoms. Paediatr Respir Rev 2006;7:84-8. 2. Everard ML. Acute viral bronchiolitis and pneumonia in infancy resulting from the respiratory syncytial virus. In: Taussig LM, Landau LI, Le Souef PN, Morgan WJ, Martinez FD, Sly PD, editors. Pediatric respiratory medicine. St. Louis: Mosby; 1999. p. 580-95. 3. Henderson FW, Clyde WJ, Collier AM. The etiologic and epidemiologic spectrum of bronchiolitis in pediatric practice. J Pediatr 1979;95:183-90. 4. Kim H, Arrobio J, Brandt C. Epidemiology of respiratory syncytial virus infection in Washington, DC, importance of the virus in different disease syndromes and temporal distribution of infection. Am J Epidemiol 1973:98:216-25. 5. Abzug M, Beam A, Gyorkos E, Levin M. Viral pneumonia in the first of month of life. Pediatr Infect Dis J 1990;9:881-5. 6. Sly PD, Hibbert ME. Childhood asthma following hospitalization with acute viral bronchiolitis in infancy. Pediatr Pulmonol 1989;7:153-8. 7. Stein RT, Sherrill D, Morgan WJ, Holberg CJ, Halonen M, Taussig LM, et al. Respiratory syncytial virus in early life and risk of wheeze and allergy by age 13 years. Lancet 1999;354:541-5. 8. Pullan CR, Hey EN. Wheezing, asthma and pulmonary dysfunction 10 years after infection with respiratory syncytial virus in infancy. Br Med J 1982;284:1665-9. 9. Welliver RC, Duffy L. The relationship of RSV-specific immunoglobulin E antibody responses in infancy, recurrent wheezing, and pulmonary function at age 7-8 years. Pediatr Pulmonol 1993;15:19-27. 10. Kusel MMH, de Klerk NH, Holt PG, Kebadze T, Johnston SL, Sly PD. Role of respiratory viruses in acute upper and lower respiratory tract illness in the first year of life a birth cohort study. Pediatr Infect Dis J 2006;25:680-6. 11. Lemanske RF, Jackson DJ, Gangnon RE, Evans MD, Li Z, Shult PA, et al. Rhinovirus illness during infancy predicts subsequent childhood wheezing. J Allergy Clin Immunol 2005;116:571-7. 12. Young S, O’Keeffe PT, Arnott J, Landau LI. Lung function, airway responsiveness, and respiratory symptoms before and after bronchiolitis. Arch Dis Child 1995;72:16-24. 13. Martinez FD, Wright AL, Taussig LM, Holberg CJ, Halonen M, Morgan WJ. Asthma and wheezing in the first six years of life. N Eng J Med 1995;332:133-8. 14. Macaubas C, de Klerk N, Holt BJ, Wee C, Kendall G, Firth M, et al. Association between antenatal cytokine production and the development of atopy and asthma at age 6 years. Lancet 2003;362:1192-7. 15. Rowe J, Macaubas C, Monger T, Holt BJ, Harvey J, Poolman JT, et al. Heterogeneity in diptheria-tetanus-acellular pertussis vaccine-specific cellular immunity during infancy: relationship to variations in the kinetics of postnatal maturation of systematic Th1 function. J Infect Dis 2001;184:80-8. 16. Stang P, Brandenburg N, Carter B. The economic burden of respiratory syncytial virus-associated bronchiolitis hospitalizations. Arch Ped Adolesc Med 2001;155:95-6. 17. Mullins JA, Lamonte AC, Bresee JS, Anderson LJ. Substantial variability in community respiratory syncytial virus season timing. Pediatr Infect Dis J 2003; 22:857-86.

18. Leader S, Kohlhase K. Recent trends in severe respiratory syncytial virus (RSV) among US infants, 1997-2000. J Pediatr 2003;143:S127-32. 19. Soto ME, Sly PD, Uren E, Taussig LM, Landau LI. Bronchodilator response during acute viral bronchiolitis in infancy. Pediatr Pulmonol 1985;2:85-90. 20. Hughes DM, LeSouef PN, Landau LI. Effect of salbutamol on respiratory mechanics in bronchiolitis. Pediatr Res 1987;22:83-8. 21. O’Callaghan C, Milner AD, Swarbnick A. Paradoxical deterioration in lung function after nebulised salbutamol in wheezy infants. Lancet 1988;2:1424-5. 22. Prendiville A, Green A, Silverman M. Paradoxical response to nebulised salbutamol, assessed by partial expiratory flow volume curves. Thorax 1987;42:81-91. 23. Mallol J, Hibbert ME, Robertson CF, Olinsky A, Phelan PD, Sly PD. Inherent variability of pulmonary function tests in infants with bronchiolitis. Pediatr Pulmonol 1988;5:152-7. 24. Ho L, Collis G, Landau LI, LeSouef PN. Effects of salbutamol on oxygen saturation in bronchiolitis. Arch Dis Child 1991;60:1061-4. 25. Mallol J, Barrueto L, Girardi G, Munoz R, Puppo H, Ulloa V, et al. Use of nebulised bronchodilators in infants under 1 year of age: analysis of four forms of therapy. Pediatr Pulmonol 1987;3:298-303. 26. Wainwright C, Altimirano L, Cheney M, Cheney J, Barber S, Price D, et al. A multicenter, randomized, double-blind, controlled trial of nebulized epinephrine in infants with acute bronchiolitis. N Eng J Med 2003;349:27-35. 27. Gadomski AM, Bhasale AL. Bronchodilators for bronchiolitis. Cochrane Database of Systematic Reviews 2006; issue 3. Art No: CD001266. DOI: 10.1002/14651858. 28. Subcommittee on diagnosis and management of bronchiolitis. Diagnosis and management of bronchiolitis. Pediatrics 2006;118:1774-93. 29. Dawson K, Kennedy D, Asher I, Cooper D, Cooper P, Francis PP, et al. Consensus view: the management of acute bronchiolitis. J Paediatr Child Health 1993;29:335-7. 30. Patel H, Platt R, Lorenzo JM, Wang EEL. Glucocorticoids for acute viral bronchiolitis in infants and young children. Cochrane Database of Systematic Reviews 2004; issue 3. Art No: CD004878. DOI: 10.1002/14651858. 31. Garrison MM, Christakis DA, Harvey E, Cummings P, Lavis RL. Systemic corticosteroids in infant bronchiolitis: a meta-analysis. Pediatrics 2000;105:e44. 32. Blom D, Ermers M, Bont L, van Aaldren WMC, van Woensel JMB. Inhaled corticosteroids during acute bronchiolitis in the prevention of post-bronchiolitis wheezing. Cochrane Database of Systematic Reviews 2007; issue 1. Art. No: CD004881. DOI: 10.1002/14651858. 33. Kuzik BA, Al Qadhi SA, Kent S, Flavin MP, Hopman W, Hotte S, Gander S. Nebulized hypertonic saline in the treatment of viral bronchiolitis in infants. J Pediatr 2007;151:266-70. 34. Bye PTP, Elkins MR. Mini-symposium: airway clearance in cystic fibrosis. Other mucoactive agents for cystic fibrosis. Pediatr Respir Rev 2007;8:30-9. 35. Mandelberg A, Tal G, Witzling M, Someck E, Houri S, Balin A, et al. Nebulized 3% hypertonic saline solution treatment in hospitalized infants with viral bronchiolitis. Chest 2003;123:481-7. 36. Sarrell EM, Tal G, Witzling M, Someck E, Houri S, Cohen HA, et al. Nebulized 3% hypertonic saline solution treatment in ambulatory children with viral bronchiolitis decreases symptoms. Chest 2002;122:2015-20.

“And Things that Go Bump in the Night”: Nothing to Fear?

ver since the first crib bumper pads were sold, they have held a seemingly irresistible appeal to new parents. All parents, after all, are protective of their children and do their best to keep them from harm. This includes lumps, bumps, and other injuries. Childhood rhymes, such as “he went to bed and bumped his head, and couldn’t get up in the morning,” and traditional prayers equating “things that go bump in the night” with “ghoulies and ghosties” provide rein-

E AAP CPSC SIDS

Editorials

American Academy of Pediatrics Consumer Product Safety Commission Sudden Infant Death Syndrome

forcement that bumps are dangerous and to be avoided at all cost. Protection from these injuries is often provided in the form of a soft surface that can cushion a fall or bump. Thus, it is no surprise that parents are often unable to resist providing a soft envi-

See related article, p 271 Reprint requests: Rachel Y. Moon, MD, Goldberg Center for Community Pediatric Health, Children’s National Medical Center, 111 Michigan Avenue, NW, Washington, DC 20010. E-mail: [email protected]. J Pediatr 2007;151:237-8 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.05.028

237

ronment that will protect their vulnerable infant’s head and body from bumping up against the hard dangers of the wooden crib slats. And in fact, crib bumper pads initially became popular as a means to protect infants from injury at a time when crib slats were spaced to permit wedging of the head between the slats. However, since 1986, crib safety requirements have mandated that crib slats be no more than 2-3/8 inches apart, to prevent a head from slipping between the slats. Nonetheless, 20 years after crib bumper pads were made obsolete by crib safety standards, they continue to be extremely popular. Anecdotal reports from my practice and from my perusal of parenting websites suggest that parents buy and continue to use crib bumper pads for 1 of 3 main reasons: (1) The infant likes to sleep with his or her head in the corner of the crib, and the bumper pads provide a soft surface; (2) the infant’s extremities might become wedged between the slats or the infant will be bruised by bumping up against the crib; and (3) bumper pads look adorable and make the crib a “cozy” environment for the infant. The use of crib bumper pads has recently become more controversial. The American Academy of Pediatrics (AAP) Task Force on Sudden Infant Death Syndrome (SIDS) has not made a recommendation about bumper pad use, except to recommend that bumper pads be “thin, firm, well-secured, and not pillow-like.”1 In contrast, the Canadian Paediatric Society and Health Canada issued recommendations in 2004 against using bumper pads, because of the concern that the softness could create a potential SIDS or suffocation risk for the infant.2 Now a study reported by Thach, Rutherford, and Harris3 in this issue of The Journal confirms these concerns. Using Consumer Product Safety Commission (CPSC) data, the authors describe 3 distinct mechanisms by which bumper pads can contribute to sudden infant death: strangulation by ties, suffocation against the bumper pad, and entrapment between the bumper pads and another object. In addition, resourceful children can use the bumper pads to step on and raise themselves up in an effort to reach outside of or climb out of the crib. Thach et al provide data that can be used to make a stronger case against bumper pads to families reluctant to give them up. For those parents who use bumper pads to provide a soft surface because their infants like to wedge themselves in the crib corner, the descriptions and photographs clearly demonstrate the risk of wedging between the bumper pad and the mattress. In fact, the authors consider the “firm” bumper pads deemed acceptable by the AAP

238

Editorials

Task Force on SIDS to have the highest potential for wedging accidents. Parents are very concerned about the potential for injury from the infant bumping up against the crib or getting an extremity wedged between crib slats. They should be reassured by the CPSC data that clearly show that infants suffer essentially no long-term injuries from contact with the crib slats themselves, making bumper pads unnecessary for infant safety. Another argument against crib bumper pads not mentioned by Thach et al is that they obscure visibility. Particularly now with the new emphasis on room-sharing without bed-sharing as the preferred sleeping arrangement for parents and their infants, improved visibility of the infant in the crib may provide an additional impetus to avoid bumper pads. Perhaps those parents who find crib bumper pads adorable can be convinced of the risks of maintaining such a “cute” or “cozy” crib environment. This is clearly more difficult than it may sound, however. Many parents ask me why stores that carry baby merchandise are selling so-called “dangerous” blankets and comforters, with the implied assumption that if an item were truly dangerous, then stores would stop selling it. However, if “truly dangerous” were actually a criterion for determining whether or not an item should be sold, then many items no longer would be on the shelves. We cannot necessarily expect merchants to stop selling an item as long as there is consumer demand for it. Instead, we as a medical community need to be more proactive in alerting parents to the dangers of soft bedding in the infant sleep environment. Parents often confuse safety for their child with objects that are soft. Although it is true that a soft surface can help cushion a fall, we must continue to remind parents that when it comes to sleep time for their infants, soft and cozy do not equal safe. Rachel Y. Moon, MD Goldberg Center for Community Pediatric Health Children’s National Medical Center Washington, DC

REFERENCES 1. Kattwinkel J, Hauck FR, Keenan ME, Malloy MH, Moon RY. American Academy of Pediatrics Task Force on Sudden Infant Death Syndrome. The changing concept of sudden infant death syndrome: diagnostic coding shifts, controversies regarding the sleeping environment, and new variables to consider in reducing risk. Pediatrics 2005;116:1245-55. 2. Canadian Paediatric Society, Community Paediatrics Committee. Recommendations for safe sleeping environments for infants and children. Paediatr Child Health 2004;9:659-63. 3. Thach BT, Rutherford GW, Harris K. Deaths and injuries attributed to infant crib bumper pads. J Pediatr 2007;151:271-4.

The Journal of Pediatrics • September 2007

ORIGINAL ARTICLES

Subclinical Atherosclerosis, but Normal Autonomic Function after Kawasaki Disease ROBERT DALLA POZZA, MD, SUSANNE BECHTOLD, MD, SIMON URSCHEL, MD, RAINER KOZLIK-FELDMANN, MD, HEINRICH NETZ, MD, PHD

AND

Objective To compare the carotid artery intima-media-thickness (IMT) of children with Kawasaki disease with normative data for Western children. Study design Forty-eight children (20 patients after Kawasaki disease, mean age 12.1 ⴞ 4.7 years; 28 age- and sex-matched healthy controls, mean age 12.0 ⴞ 3.1 years) were studied. Results Mean (IMT differed significantly (0.449 ⴞ 0.02 vs 0.424 ⴞ 0.01, P < .001) as well as IMT standard deviation score (1.2 ⴞ 0.6 vs 0.3 ⴞ 0.1, P < .001). Patients with coronary arterial involvement (n ⴝ 15) showed a further increase of the IMT (0.459 ⴞ 0.01 vs 0.436 ⴞ 0.01, P < .05). There was no difference regarding short-term blood pressure regulation. Conclusions In this small patient group, signs of subclinical atherosclerosis after Kawasaki disease have been detected. These preliminary data indicate that these patients may be at risk for cardiovascular disease even in the absence of permanent alterations of the coronary arteries. (J Pediatr 2007;151:239-43) awasaki syndrome is considered an acute, generalized vasculitis.1 Treatment with ␥-globulin has shown to be effective to reduce cardiac sequelae; ⬍5% of the children who are treated develop coronary arterial aneurysms.2,3 Late sequelae, especially in patients with large aneurysms, include myocardial ischemia and infarction with a frequency of 30% within 4 years after onset of the disease.1 Beside cardiac involvement, disturbed lipid metabolism as well as endothelial dysfunction have been described in patients with Kawasaki disease persisting after the acute phase has completely resolved.4 This finding is of particular importance as changes at the endothelial level during the inflammatory process in concomitance with the following atherogenic lipid profile and endothelial dysfunction may enhance the progression of atherosclerosis. The carotid artery wall in patients with coronary artery lesions 6 to 20 years after Kawasaki disease has been found to be less distensible and thicker than that in control patients.1 Coronary artery aneurysms that have regressed are not only histopathologically abnormal, but they also show reduced vascular reactivity to vasodilators and vasoconstrictors indicating persistent endothelial dysfunction.1 A disturbed homeostasis of the endothelium may result in a reduced antiinflammatory capacity, thus promoting an acute vasculitis to become chronic.5 Atherosclerosis has been shown to be the result of the interaction of chronic inflammation See editorial, p 225 and and of serum lipoproteins that may start with early atherosclerotic lesions (fatty streaks) 5 related article, p 244 and end with more complex vascular changes. Such atherosclerotic processes at the endothelial level begin in childhood and are accelerated in the presence of risk factors.6,7 Common carotid artery intima-media From the Department of Pediatric Cardiology (R.D.P., S.U., R.K.-F., H.N.) and the Dithickness (IMT) is a non-invasive marker of subclinical atherosclerosis.8-10 An increased vision of Pediatric Rheumatology (S.B.), IMT has been correlated to an increased relative risk for stroke and myocardial infarction University Children’s Hospital, Ludwig9 Maximilians-University, Munich, Germany. in adults. In children, a significant thickening of the endothelial wall has been demonSubmitted for publication Nov 7, 2006; last strated in obesity, hypertension, diabetes mellitus, and familial hypercholesterolrevision received Dec 29, 2006; accepted emia.9,11–22 March 22, 2007. 23,24 Atherosclerotic changes may be present in children with Kawasaki disease. Reprint requests: Dr R. Dalla Pozza, Department of Pediatric Cardiology, University However, normative values published recently for the IMT were not adopted for comChildren’s Hospital, Ludwig-Maximilians-Uniparison of the results of patients and controls.25 Furthermore, increased arterial stiffening versity, Lindwurmstr. 4, 80336 Munich, Ger-

K

BMI BRS CRP ESR

Body mass index Baroreceptor sensitivity C-reactive protein Erythrocyte sedimentation rate

HDL IMT LDL WBC

High-density lipoprotein cholesterol Intima-media thickness Low-density lipoprotein cholesterol White blood cell count

many. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.057

239

Table I. Patients and controls characteristics

Sex Age (y) Weight (kg) Height (cm) Body mass index (BMI) (kg/m2) BMI-SDS Mean follow-up interval (y) Heart rate at rest (beats/min) Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Total cholesterol (mg/dL) LDL-cholesterol (mg/dL) HDL-cholesterol (mg/dL) Triglycerides (mg/dL) Intima-media thickness (IMT) (mm) IMT-SDS BRS (msec/mm Hg)

Study group (n ⴝ 20)

Control group (n ⴝ 28)

12 m, 8 f 12.1 ⫾ 4.7 (range 6–23) 42.2 ⫾ 18.0 (20.4–83) 146 ⫾ 17.5 (120–178) 17.9 ⫾ 5.5 (14.1–28.2) 1.0 ⫾ 1.1 (0.5–1.8) 4.1 ⫾ 3.6 (0.3–9.6) 77.9 ⫾ 11.3 (56.9–104.7) 108 ⫾ 13.9 (92.4–134.5) 66.1 ⫾ 11.9 (53.9–92) 169.4 ⫾ 16.7 (140–182) 94.3 ⫾ 22.4 (65–117) 48.5 ⫾ 11.2 (28–62) 123.6 ⫾ 55.6 (67–189) 0.449 ⫾ 0.023 (0.417–0.495) 1.2 ⫾ 0.6 (0.1–3.4) 20.5 ⫾ 9.5 (5.84–47.49)

10 m, 18 f 12.0 ⫾ 3.1 (range 8.5–22.4) 51.7 ⫾ 14.6 (29.8–83)* 160 ⫾ 15.1 (125–185) 19.8 ⫾ 3.5 (14.5–30.0) 1.3 ⫾ 0.8 (0.9–1.6) — 74.5 ⫾ 11.6 (55.0–101.6) 110.8 ⫾ 8.6 (94.4–137.4) 65.7 ⫾ 7.5 (54.0–82.0) 167.3 ⫾ 18.4 (145–190) 92.5 ⫾ 16.4 (63–121) 47.7 ⫾ 17.9 (27–67) 130.5 ⫾ 65.3 (62–198) 0.424 ⫾ 0.010 (0.400–0.445)** 0.3 ⫾ 0.1 (0.1–0.4)** 24.4 ⫾ 8.3 (7.86–43.72)

IMT-SDS, standard deviation score of the intima-media thickness. *P ⬍ .05. **P ⬍ .001.

in children after Kawasaki disease has been detected. Data from animal experiments indicate that reduced compliance of the carotid arteries may lead to decreased short-term blood pressure regulation (baroreceptor sensitivity, BRS).26 Patients with reduced BRS in turn are at increased risk for the development of systemic hypertension.27 Investigating the IMT and BRS of children after Kawasaki disease may be helpful to identify those patients with an elevated risk for atherosclerosis and for systemic hypertension.

METHODS Twenty children and adolescents who had previous Kawasaki disease were enrolled (Table I). All patients were recruited consecutively during their regular visits as outpatients at a tertiary healthcare center. Subjects were excluded if they had evidence or history of clinically relevant systemic disease (ie, malignancies, hypertension) other than Kawasaki disease. Medical records were available for all patients for the entire follow-up period. The diagnosis of Kawasaki disease was based on the current classification.1 All patients fulfilled the diagnostic criteria for Kawasaki disease: fever persisting for at least 5 days and the presence of at least four of the five principal features. The treatment of the acute phase of the disease included the administration of intravenuous immunoglobulines (IVIG) in all patients at a dose of 400 mg/kg/day for five consecutive days in those patients who presented before 1998 and at a dose of 2 g/kg in those patients presenting after 1998. Aspirin was administrated in all patients at a dose of 82 to 95 mg/kg/day (mean dose 89.2 mg/kg/day). The control group consisted of 28 children, in part friends of the patients and in part children presenting for cardiac evaluation at our institution. After exclusion of an underlying chronic or cardiovascular disease by medical history, clinical assessment, 240

Dalla Pozza et al

electrocardiography, and echocardiography, the measurements were performed. Family history of coronary artery disease and stroke was determined by questionnaire. Written informed consent was obtained for all participants from their legal guardians. The study was performed according to the Declaration of Helsinki; the study protocol was approved by the local ethics committee. The ultrasonographic study was performed with the patients supine for at least 10 minutes in a quiet room at 22°C. For data acquisition, a Philipps IE33 was used, equipped with a linear 11.0 MHz transducer (Philipps, Germany). All studies were done according to a standardized scanning protocol for the right and left common carotid arteries. The common carotid artery bulb was identified, and the segments of the common carotid arteries 1 to 2 cm proximal to the bulb region were scanned. The image was focussed on the posterior (far) wall. Two angles were used at each side for scanning the common carotid artery IMT: lateral and anterior-oblique.9,28,29 All images were obtained by one single, experienced ultrasonographer. The images were stored digitally for subsequent, offline analysis. The ultrasonographer and the reader were blinded to the subject’s clinical details. For the measurement of the IMT, the distance between the leading edges of the lumen—intima interface and the media—adventitia interface of the B-mode frame was considered. Software (Qlab, Philipps, Germany) analyzed the IMT distance automatically at 64 points within a segment of 10 mm. The value given was the arithmetic mean IMT calculated. A manual overreading of the border detection during computed analysis was performed in all images obtained. The calculation of the mean IMT was performed for two separate scans of each side; the mean IMT for each The Journal of Pediatrics • September 2007

Table II. Comparison of patients with and without coronary artery involvement

Age (y) Weight (kg) Height (cm) Body mass index (BMI) (kg/m2) BMI-SDS Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) HbA1c (%) Total cholesterol (mg/dL) LDL-cholesterol (mg/dL) HDL-cholesterol (mg/dL) Triglycerides (mg/dL) Intima-media thickness (IMT) (mm) IMT-SDS BRS (msec/mm Hg)

Coronary involvement (n ⴝ 15)

Without coronary involvement (n ⴝ 5)

11.9 ⫾ 5.1 (6–19) 41.3 ⫾ 20.7 (20.8–73.2) 143.8 ⫾ 19.5 (120–169.2) 18.6 ⫾ 4.2 (15.2–28.2) 0.9 ⫾ 0.6 (0.3–1.6) 108.2 ⫾ 13.7 (100.8–134.5) 64.1 ⫾ 11.6 (53.9–86) 4.9 ⫾ 0.9 (4–5.8) 168.0 ⫾ 28.4 (140–182) 90.4 ⫾ 24.7 (65–116.2) 45.3 ⫾ 12.4 (28–58.7) 120.7 ⫾ 64.2 (67–185.2) 0.459 ⫾ 0.019 (0.44–0.49) 1.7 ⫾ 0.6 (0.8–3.4) 21.05 ⫾ 9.6 (10.2–37.4)

11.7 ⫾ 2.9 (8–23) 37.9 ⫾ 10.9 (20.4–83) 146.3 ⫾ 14.3 (134.3–178) 17.3 ⫾ 2.1 (14.1–26.7) 0.7 ⫾ 0.4 (0.3–1.3) 106.5 ⫾ 12.2 (92.4–128.2) 64.9 ⫾ 12.9 (54–92) 4.3 ⫾ 0.9 (3.3–5.3) 164.3 ⫾ 18.4 (140.4–181.1) 101.2 ⫾ 25.2 (72.2–117) 52.2 ⫾ 14.6 (32.1–62) 125.7 ⫾ 68.2 (67–189) 0.436 ⫾ 0.018 (0.41–0.44)* 1.2 ⫾ 0.6 (0.1–2.2)* 22.1 ⫾ 8.8 (5.8–47.4)

BRS, baroreceptor sensitivity; IMT-SDS, standard deviation score of the intima-media thickness. *P ⬍ .05.

patient was calculated from the four mean IMT values. Five patients and 10 controls presented twice for the determination of intraobserver variability within 8 weeks, and the IMT was measured without knowledge of the former values. The intraobserver variability was 4.1%. For the calculation of the standard deviation score of the IMT, the sex- and heightdependent normative values from the literature were adopted.25 Beat-to-beat blood pressure was measured by a fingercuff recording online beat-to-beat blood pressure on the second and third finger of the right hand using the vascular unloading technique (Task Force Monitor, CNS-Systems, Graz, Austria). This technique has been validated by Tanaka et al.30 The quality of the pulse signal obtained was controlled visually by one experienced reader. The calibration of the system occurred with a conventional non-invasive blood pressure cuff (Dinamap, GE Systems, Germany) positioned on the left upper arm and performed automatically every 2 minutes. A standard three-lead electrocardiogram was used for determination of heart rate. Visual control of the electrocardiogram quality was performed similarly as described above for the pulse wave form. The entire measurement was conducted over a time interval of 10 minutes. When premature beats were noted, the test was stopped and started again. For the calculation of the BRS, the relative changes of blood pressure (expressed in mm Hg) and of the heart rate (expressed as the distance between the corresponding QRScomplexes: RR-interval in milliseconds) were used according to the sequence method with a cutoff point of 1 mm Hg and 3 milliseconds.31,32 Blood samples were taken during the patient’s follow-up visit, when the above-mentioned measurements were performed. HbA1c, triglycerides, total cholesterol, high-density lipoprotein (HDL) cholesterol and low-density lipoprotein (LDL) cholesterol were obtained by standard laboratory methods.

Calculations were performed using the Statistical Package for the Social Sciences for Windows (version 14.0, SPSS, Chicago, Ill). Differences within the patient group and between the patients and the control group were tested using the independent sample t test and the nonparametric MannWhitney test. Correlations were analysed using Pearson’s correlation coefficient. All significance testing was fixed at P ⬍ .05 (two-sided).

RESULTS The anthropometric characteristics of the groups are reported in Table I. Controls showed significantly increased weight, but not body mass index (BMI) or BMI-SDS compared with the patient group. There was no family history of coronary artery disease. The mean time interval between the onset of the disease and the time of testing was 4.1 ⫾ 3.6 years (range 0.3 ⫾ 9.6 years). Absolute IMT and IMT-SDS were significantly different between the groups. Compared with normal values, the controls were within the normal range. In both groups, atherogenic measures such as the BMI, systolic blood pressure, and total and LDL cholesterol did not exceed normal limits enough to influence the IMT. In the patient group, a direct correlation of the IMT to factors characterizing the severity of the Kawasaki disease (ie, the C-reative protein [CRP], erythrocyte sedimentation rate [ESR], duration of illness before therapy, white blood cell count [WBC]) was not found. However, the IMT was significantly increased in those patients with involvement of the coronary arteries at the time of illness (Table II). Patients with coronary arterial involvement showed increased CRP and ESR at the time of presentation of the illness (Table III) indicating more pronounced inflammatory activity. There was no direct correlation between the time course after Kawasaki disease and the IMT-SDS. However, a slightly, but not significantly, increased correlation was present in those patients with coronary arterial involvement.

Subclinical Atherosclerosis, but Normal Autonomic Function after Kawasaki Disease

241

Table III. Clinical characteristics at the time of presentation of patients with and without coronary involvement

Days of fever before the beginning of therapy ESR (mm/h) WBC (cells/␮L) CRP (mg/dL)

With coronary involvement (n ⴝ 15)

Without coronary involvement (n ⴝ 5)

5.3 ⫾ 1.9 (3–8) 81.6 ⫾ 38.8 (40–163) 17885 ⫾ 1501.6 (14362–22340) 14.0 ⫾ 6.2 (5.3–22.7)

4.3 ⫾ 0.5 (3–5) 30.3 ⫾ 21.3 (12–65)* 9366 ⫾ 2640 (6230–15327) 3.0 ⫾ 3.9 (0.1–10.3)*

CRP, C-reactive protein; ESR, erythrocyte sedimentation rate; WBC, white blood cell count.

There was no difference between patients and controls in BRS. Impairment of the baroreflex sensitivity was not detectable. Blood pressure levels were within normal limits in all patients and healthy controls.

DISCUSSION The present study revealed signs of subclinical atherosclerosis in a group of children after Kawasaki disease. However, impairment of the short-term blood pressure regulation was not found during medium-term follow-up. Because the carotid IMT in adults is considered a valuable tool for the determination of cardiovascular risk,9 its use in the pediatric field is growing. The method of measurement of the IMT has been validated in children. Studies have revealed increased IMT in children after Kawasaki disease as well, but data on IMT values compared with the normal population have been lacking. In our patient group after Kawasaki disease, the significantly increased IMT suggests subclinical atherosclerosis. Children, in whom a dilatation of the coronary arteries had been diagnosed at the time of acute inflammation show an increased IMT compared with the remaining patients. In these children, the inflammatory process was more severe than in the remaining children (increased CRP and ESR). However, a direct correlation of the inflammatory signs to the level of IMT increase could not be found. We suspect that because of the wide range of the CRP and ESR, in our small patient group statistically significant correlations could not be found. We hypothesize that the increased IMT may be caused by a high inflammatory activity resulting in a generalized vasculitis. This relationship has been shown in patients with Takayasu’s arteritis and Behcet’s disease.33,34 In contrast to the changes in the coronary arteries that may be transient, traces of the vasculitis may be permanent in other vessels. The direct effect of a vasculitis may be enhanced by an altered lipid profile, which has been found after Kawasaki disease. It may be that vasculitis and an atherogenic lipid profile interact. We could not detect a significant correlation between the IMT-SDS and the time interval between the onset of Kawasaki disease and testing. We suppose that our patient and subgroups are too small to reveal statistical significance. However, a slightly increased correlation coefficient in those patients with coronary arterial involvement may be indicative of a relationship. In our patient group, there was a high incidence of those patients presenting with coronary arterial 242

Dalla Pozza et al

involvement at the time of inflammation when compared with data from the literature.1 The explanation may be that the patients studied were recruited from a pediatric cardiologic tertiary healthcare center, which may preferentially attract patients in whom coronary arterial changes are suspected or had been detected previously. The process of atherosclerosis starts in childhood and is enhanced in the presence of risk factors.7 Beyond a normal increase of the IMT because of an increase of the carotid arterial diameter, IMT may be further increased pathologically during life. Our patients after Kawasaki disease may be at an increased risk for atherosclerotic disease in adulthood. Atherosclerosis-promoting environmental factors such as smoking and obesity may have an additional deleterious impact. Lifelong follow-up visits may be necessary, and special counselling may be wise in patients after Kawasaki disease. In contrast to the increased IMT values, the short-term blood pressure regulation was normal in our patients. An impairment of the BRS has been associated with the development of systemic hypertension in the normal population.26 As patients after Kawasaki disease showed an increased vascular stiffness, we suspected an impaired BRS in this group. However, the short-term blood pressure regulation is completely normal in these patients. Study limitations include the small number of patients investigated. As we present a retrospective study, a prospective assessment of the IMT during the different phases of Kawasaki disease is warranted. Additionally, further studies are necessary to determine if increase in the IMT has the same prognostic value on morbidity and mortality as in adult patients. We propose regular follow-up visits in patients after Kawasaki disease. Longitudinal IMT measurements may help to identify patients at particular risk for cardiovascular disease. Those children need to be followed and counselled with regard to atherosclerosis-promoting factors such as obesity, hypercholesterolemia, systemic hypertension, and smoking. We appreciate the enthusiastic help of Ms. G. Walter.

REFERENCES 1. Newburger JW, Takahashi M, Gerber MA, Gewitz MH, Tany LY, Burns JC, et al. Diagnosis, treatment and long-term management of Kawasaki disease: a statement for health professionals from the committee on rheumatic fever, endocarditis, and Kawasaki disease, council on cardiovascular disease in the young, American Heart Association. Pediatrics 2004;114:1708-33. 2. Benseler S, McCrindle BW, Silverman ED, Tyrrell PN, Wong J, Yeung RSM.

The Journal of Pediatrics • September 2007

Infections and Kawasaki disease: implications for coronary artery outcome. Pediatrics 2005;116:760-6. 3. Fulton DR, Newburger JW. Long-term cardiac sequelae of Kawasaki disease. Curr Rheumatol Rep 2000;2:324-9. 4. Newberger JW, Burns JC, Beiser AS, Loscalzo J. Altered lipid profile after Kawasaki Syndrome. Circulation 1991;84:625-31. 5. Fan J, Watanabe T. Inflammatory reactions in the pathogenesis of atherosclerosis. J Atheroscler Thromb 2003;10:63-71. 6. Järvisalo MJ, Putto-Laurila A, Jartti L, Lehtimäki T, Solakivi T, Rönnemaa T, et al. Carotid intima-media thickness in children with type 1 diabetes. Diabetes 2002;51:493-8. 7. Berenson GS for the Bogalusa Heart Study Research Group. Childhood risk factors predict adult risk associated with subclinical cardiovascular disease: the Bogalusa heart study. Am J Cardiol 2002;90(suppl):3L-7L. 8. Parikh A, Danemann D. Is carotid ultrasound a useful tool in assessing cardiovascular disease in individuals with diabetes? Diabetes Technol Ther 2004;6:65-9. 9. De Groot E, Hovingh K, Wiegman A, Duriez P, Smit AJ, Fruchart JC, et al. Measurement of arterial wall thickness as a surrogate marker for atherosclerosis. Circulation 2004;109:33-8. 10. Pignoli P, Tremoli E, Poli A, Oreste P, Paoletti R. Intimal plus medial thickness of the arterial wall: a direct measurement with ultrasound imaging. Circulation 1986;74:1399-1406. 11. Stabouli S, Kotsis V, Papamichael C, Constantinopoulos A, Zakopoulos N. Adolescent obesity is associated with high ambulatory blood pressure and increased carotid intimal-medial thickness. J Pediatr 2005;147:651-6. 12. Koeijvoets KCMC, Rodenburg J, Hutten B, Wiegman A, Kastelein JJP, Sijbrands EJG. Low-density lipoprotein receptor genotype and response to pravastatin in children with familial hypercholesterolemia. Circulation 2005;112:3168-73. 13. Litwin M, Niemirska A, Sladowska J, Antoniewicz J, Daszkowska J, Wierzbicka A, et al. Left ventricular hypertrophy and arterial wall thickening in children with essential hypertension. Pediatr Nephrol 2006;21:811-19. 14. Stakos DA, Schuster DP, Sparks EA, Wooley CF, Osei K, Boudoulas H. Cardiovascular effects of type 1 diabetes mellitus in children. Angiology 2005; 56:311-17. 15. Atabek ME, Kurtoglu S, Pirgon O, Baykara M. Arterial wall thickening and stiffening in children and adolescents with type 1 diabetes. Diabetes Res Clin Pract 2006;74:33-40. 16. Atabek ME, Pirgon O, Kurtoglu S, Imamoglu H. Evidence for an association between type 1 diabetes and premature carotid atherosclerosis in childhood. Pediatr Cardiol 2006;27:428-33. 17. Järvisalo MJ, Raitakari M, Toikka JO, Putto-Laurila A, Rontu R, Laine S, et al. Endothelial dysfunction and increased arterial intima-media thickness in children with type 1 diabetes. Circulation 2004;109:1750-5. 18. Parikh A, Sochett EB, McCrindle BW, Dipchand A, Danemann A, Danemann D. Carotid artery distensibility and cardiac function in adolescents with type 1 diabetes. J Pediatr 2000;137:465-9. 19. Yavuz T, Akcay A, Omeroglu RE, Bundak R, Sukur M. Ultrasonic evaluation of early atherosclerosis in children and adolescents with type 1 diabetes mellitus. J Pediatr Endocrinol Metab 2002;15:1131-6.

20. Gunczler P, Lanes R, Lopez E, Esaa S, Villarroel O, Revel-Chion R. Cardiac mass and function, carotid artery intima-media thickness and lipoprotein (a) levels in children and adolescents with type 1 diabetes mellitus of short duration. J Pediatr Endocrinol Metab 2002;15:181-6. 21. Singh TP, Groehn H, Kazmers A. Vascular function and carotid intimal-medial thickness in children with insulin-dependent diabetes mellitus. J Am Coll Cardiol 2003;41:661-5. 22. Krantz JS, Mack WJ, Hodis HN, Liu CR, Liu CH, Kaufman FR. Early onset of subclinical atherosclerosis in young persons with type 1 diabetes. J Pediatr 2004;145:452-7. 23. Noto N, Okada T, Yamasuge M, Taniguchi K, Karasawa K, Aywusawa M, et al. Noninvasive assessment of the early progression of atherosclerosis in adolescents with Kawasaki disease and coronary artery lesions. Pediatrics 2001;107:1095-9. 24. Cheung Y, Wong SJ, Ho MH. Relationship between carotid intima-media thickness and arterial stiffness in children after Kawasaki disease. Arch Dis Child 2007; 92:43-7. 25. Jourdan C, Wühl E, Litwin M, Fahr K, Trelewicz J, Jobs K, et al. Normative values for intima-media thickness and distensibility of large arteries in healthy adolescents. J Hypertens 2005;23:1707-15. 26. Trasher TN. Baroreceptors, baroreceptor unloading, and the long-term control of blood pressure. Am J Physiol Regul Integr Comp Physiol 2005;288:R819-27. 27. Lucini D, Mela GS, Malliani A, Pagani M. Impairment in cardiac autonomic regulation preceding arterial hypertension in humans. Circulation 2002;106:2673-9. 28. Aggoun Y, Szezepanski I, Bonnet D. Noninvasive assessment of arterial stiffness and risk of atherosclerotic events in children. Pediatr Res 2005;58:173-8. 29. Touboul PJ, Hennerici MG, Meairs S, Adams H, Amarenco P, Desvarieux A, et al. Mannheim intima-media thickness consensus on behalf of the advisory board of the 3rd Watching the Risk Symposium 2004, 13th European Stroke Conference. Cerebrovasc Dis 2004;18:346-9. 30. Tanaka H, Thulesius O, Yamaguchi H, Mino M, Konishi K. Continuous noninvasive finger blood pressure monitoring in children. Acta Paediatr 1994;83:646-52. 31. Parati G, Ognaro G, Bilo G, Glavina F, Castiglioni P, Di Rienzo M, et al. Non-invasive beat-to-beat blood pressure monitoring: new developments. Blood Press Monit 2003;8:31-6. 32. Parati G, Omboni S, Frattola A, Di Rienzo M, Zanchetti A, Mancia G. Dynamic evaluation of the baroreflex in ambulant subject. In: Di Rienzo S, Parati G et al, eds. Blood Pressure and Heart Rate Variability. Amsterdam: IOS Press; 1992:123-37. 33. Keser G, Aksu K, Tamsel S, Ozmen M, Kitapcioglu G, Kabaroglu C, et al. Increased thickness of the carotid artery intima-media assessed by ultrasonography in Behcet’s disease. Clin Exp Rheumatol 2005;23:S71-S76. 34. Seth S, Goyal NK, Jagia P, Gulati G, Karthikeyan G, Sharma S, et al. Carotid indima-medial thickness as a marker of disease activity in Takayasu’s arteritis. Int J Cardiol 2006;108:385-90. 35. Dalla Pozza R, Kleinmann A, Bechtold S, Netz H. Hypertension in heart and heart-lung transplanted children: does impaired baroreceptor function play a role? Transplantation 2006;81:71-5. 36. Dalla Pozza R, Bechtold S, Putzker S, Bonfig W, Netz H, Schwarz HP. Young adults born small for gestational age: is reduced baroreceptor sensitivity a risk factor for hypertension? Clin Cardiol 2006;29:215-18.

Subclinical Atherosclerosis, but Normal Autonomic Function after Kawasaki Disease

243

Are Patients after Kawasaki Disease at Increased Risk for Accelerated Atherosclerosis? BRIAN W. MCCRINDLE, MD, MPH, SUSAN MCINTYRE, RN, CHRISTOPHER KIM, TAMMY LIN,

AND

KHOSROW ADELI, PHD

Objective To assess whether patients after Kawasaki disease (KD) have increased risk factors and abnormalities suggestive of early atherosclerosis in systemic arteries. Study design In a case-control study, we compared 52 patients after typical Kawasaki disease with varying coronary artery involvement (67% males; mean time from illness episode 11.2 ⴞ 3.7 years) studied between 10 and 20 years of age with 60 healthy control subjects (50% males). Brachial artery reactivity (BAR) was assessed using vascular ultrasonography, and atherosclerosis risk assessment was performed. Differences between cases and controls and factors associated with endothelial function in cases were determined. Results Case patients had lower resting systolic blood pressure (P < .001), lower apolipoprotein AI levels (P < .05), and higher levels of glycosylated hemoglobin (P ⴝ .007). There were no significant differences in BAR between case patients and control subjects in response to increased flow (P ⴝ .60) and nitroglycerine (P ⴝ .93). For case patients, significant factors in multivariable analysis for lower flow-mediated BAR included higher fasting triglyceride levels (P ⴝ .04) and lower free fatty acid levels (P < .001). No significant relationship was noted with past or current coronary artery involvement. Conclusion Patients with KD have some abnormalities for risk factors for atherosclerosis, but systemic arterial endothelial dysfunction is not present in the long term. (J Pediatr 2007;151:244-8) ecent reports have suggested that patients with Kawasaki disease (KD) may be at increased risk for accelerated atherosclerosis. This may be on the basis of ongoing functional and structural abnormalities of affected arteries,1-4 an abnormal profile of known risk factors for atherosclerosis,5-7 or the presence of a state of chronic inflammation.6,8,9Assessment of the structure and function of the coronary arteries has suggested the presence of long-term abnormalities in segments not previously believed to have been affected.1,4 Non-invasive assessment of other systemic arteries has shown inconsistent abnormalities, and an inconsistent relationship with the degree of coronary artery involvement.7,10-14 We sought to assess whether patients after KD have increased risk factors and abnormalities suggesSee editorial, p 225 and tive of early atherosclerosis in systemic arteries.

R

related article, p 239

METHODS Study Subjects Patients were selected from a database of all patients with KD seen at the Hospital for Sick Children between 1982 and 1998, randomly sampling from three groups based on current coronary artery involvement. Healthy control subjects of similar age were concurrently recruited from community groups. All subjects were between 10 and 20 years of age. Participants gave appropriate informed consent, as approved by the Research Ethics Board of the Hospital for Sick Children. Measurements The medical records of the patients with KD were reviewed to determine characteristics of the acute KD episode, including initial coronary artery involvement, disease management, and current cardiovascular findings. Standardized atherosclerosis risk factor BAR FMD HDL

244

Brachial artery reactivity Flow-mediated dilation High-density lipoprotein

KD LDL

Kawasaki disease Low-density lipoprotein

From the Division of Cardiology, Department of Pediatrics (B.M., S.M., C.K., T.L.), and the Division of Clinical Biochemistry, Department of Laboratory Medicine and Pathobiology (K.A.), University of Toronto, The Hospital for Sick Children, Toronto, Ontario, Canada. Funded by Physicians’ Services Incorporated Foundation, Ontario, and CIBC World Markets Children’s Miracle Foundation. Submitted for publication Dec 16, 2006; last revision received Jan 22, 2007; accepted Mar 22, 2007. Reprint requests: Dr Brian McCrindle, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario, Canada. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.056

assessments were performed for both case and control subjects. Assessment included recent and past medical history, smoking and smoke exposure, and medication use. The family history of each subject was reviewed for risk factors and cardiovascular disease in primary relatives and parents. Physical activity and dietary questionnaires were administered to both study groups, as well as 3-day dietary food recall records. All participants underwent physical examinations that included Tanner staging and measurement of skinfold thicknesses. Laboratory assessments on all subjects included fasting blood work to assess serum electrolytes, glucose, creatinine, urea, serum hemoglobin A1c, free fatty acids, insulin, Cpeptide, lipoprotein profile, apolipoproteins AI, B, and E, lipoprotein (a), homocysteine, and fibrinogen, and 24-hour urinary collection for microalbuminuria. Subjects also underwent 24-hour ambulatory blood pressure monitoring (for further information about methods, see the Appendix; available at www.jpeds.com). Brachial artery reactivity (BAR) in response to flowmediated dilation (FMD) and nitroglycerine was assessed in a standardized manner using vascular ultrasonography according to the published methodology of Dhillon et al (Appendix).7,10

Data Analysis Data are described as frequencies, means with standard deviations, or medians with ranges as appropriate. Significant differences between the case patients and control subjects were sought using Fisher’s exact tests, ␹2 tests, student’s t tests, and Kruskal-Wallis analysis of variance. For patients with KD, relationships between FMD and the category of coronary artery involvement after adjustment for cardiovascular risk factors were sought using multiple linear regression analysis. All analyses were performed with Statistical Anaylsis Systems software version 9.1 (SAS Institute, Inc., Cary, NC) with default settings.

RESULTS Characteristics of Study Subjects We enrolled 52 patients with KD (67% males). The mean age at the KD episode was 4 ⫾ 3 years, with 96% having typical KD. Treatment included aspirin for 92% and intravenous gamma globulin for 64%. The initial coronary artery involvement showed aneurysms in 37%, ectasia only in 16%, and 47% with no involvement. The mean age at the time of the current study was 15.5 ⫾ 2.3 years, with a mean time since the KD episode of 11.2 ⫾ 3.7 years. The coronary artery involvement at the time of the study showed no involvement in 30 patients, regressed aneurysms in 16, and persistent aneurysms in six patients. One patient was taking warfarin and aspirin, five were taking aspirin only, and one patient was taking atenolol. The mean age at study participation of the 60 healthy control subjects (50% males) was 14.9 ⫾ 2.4 years. Two patients and three control subjects

Table I. Comparison of patients with Kawasaki disease and normal control subjects regarding demographics and atherosclerosis risk factor assessment

Characteristic

Control subjects (n ⴝ 60)

Patients with KD (n ⴝ 52)

P value

Sex (male:female) 30:30 35:17 .09 Mean age (years) 14.9 ⫾ 2.4 15.5 ⫾ 2.3 .17 Adiposity measures Mean Z score of BMI ⫹0.71 ⫾ 1.16 ⫹0.55 ⫾ 1.45 .52 Mean percentage of ideal 105 ⫾ 20% 107 ⫾ 19% .59 weight for height 55 ⫾ 28% 55 ⫾ 28% .97 Mean percentile for skin fold thicknesses 5% 13% .19 Current smoking/smoke exposure Family history Premature CVD in first32% 26% .54 degree relative Mother: hypertension 7% 10% .73 ¡ on meds 2% 8% .18 obese 28% 10% .02 hyperlipidemia 7% 12% .51 ¡ on meds 0% 6% .10 Father: hypertension 7% 12% .70 ¡ on meds 2% 4% .59 obese 17% 8% .26 hyperlipidemia 10% 38% ⬍.001 ¡ on meds 2% 4% .59 Median physical activity (hours per week) Total activity 9.5 (0, 45) 9.5 (0, 75) .90 Total sedentary pursuits 15 (2.5, 50) 25 (0, 75) .07 Total activity of other 8 (0, 22) 7 (0, 37) .78 family members BMI, body mass index; CVD, cardiovascular disease or event; KD, Kawasaki disease; meds, medications.

were taking oral contraceptives, and five control subjects were taking methylphenidate.

Comparison Regarding Atherosclerosis Risk Factors A comparison of demographic variables between patients and control subjects showed patients to be significantly less likely to have an obese mother, and more likely to have a hyperlipidemic father (Table I). The groups did not differ regarding height for weight measures, skinfold thicknesses, or self-reported levels of physical activity. There was a nonsignificant trend toward greater time spent in sedentary pursuits (television viewing, computer or videogame use) for the patients with KD. There were no significant differences regarding dietary composition (data not shown). Regarding laboratory assessments, significant findings for the patients with KD versus control subjects included

Are Patients after Kawasaki Disease at Increased Risk for Accelerated Atherosclerosis?

245

Table II. Comparison of patients with Kawasaki disease and normal control subjects regarding fasting bloodwork and 24-hour urinary collection assessment

Characteristic

Control subjects (n ⴝ 60)

Patients with KD (n ⴝ 52)

Lipid profiles (mmol/L): Mean total cholesterol 4.08 ⫾ 0.70 4.16 ⫾ 0.62 Mean LDL cholesterol 2.43 ⫾ 0.57 2.52 ⫾ 0.56 Mean HDL cholesterol 1.19 ⫾ 0.31 1.14 ⫾ 0.26 Mean triglycerides 1.00 ⫾ 0.41 1.10 ⫾ 0.42 Apolipoproteins (g/L): Mean apo AI 1.28 ⫾ 023 1.19 ⫾ 0.26 Mean apo B 0.64 ⫾ 0.18 0.61 ⫾ 0.15 Median lipoprotein(a) ⬍10 (⬍10, 83) ⬍10 (⬍10, 69) (mg/dL) Mean blood glucose 4.9 ⫾ 0.5 4.8 ⫾ 0.4 (mmol/L) Mean hemoglobin A1c 0.050 ⫾ 0.003 0.052 ⫾ 0.004 fraction Mean free fatty acids 0.49 ⫾ 0.25 0.51 ⫾ 0.27 (mmol/L) Mean insulin (pmol/L) 68 ⫾ 57 60 ⫾ 31 Mean insulin C-peptide 3.2 ⫾ 1.0 3.2 ⫾ 1.0 Median urinary 5.9 (1.2, 77.5) 4.2 (1.8, 58.3) microalbumin (mg/dL) Mean fibrinogen (g/L) 2.74 ⫾ 0.70 2.91 ⫾ 0.80 Mean homocysteine 6.13 ⫾ 1.80 6.54 ⫾ 2.44 (␮mol/L)

Table III. Comparison of patients with Kawasaki disease and normal control subjects regarding 24-hour ambulatory blood pressure monitoring

P value .52 .43 .40 .22 .047 .35 .47 .22 .007 .66 .38 .98 .12 .27 .32

apo, apolipoprotein; dL, deciliter; g, gram; HDL, high-density lipoprotein; KD, Kawasaki disease; L, liter; LDL, low-density lipoprotein; mg, milligram; mmol, millimole; pmol, picomole; ␮mol, micromole.

lower levels of apolipoprotein AI and higher levels of hemoglobin A1c (Table II). Compared with local normal values, 7% of patients with KD versus 10% of control subjects had an abnormally low apolipoprotein AI level, and no patient or subject had an abnormally high hemoglobin A1c level. There were no significant differences in lipid profiles; however, patients with KD demonstrated a tendency toward a more atherogenic profile, with lower high-density lipoprotein (HDL) and higher total and low-density lipoprotein (LDL) cholesterol, as well as higher triglycerides. Baseline resting systolic and diastolic blood pressures were significantly lower for patients with KD, and there was a trend toward lesser nighttime fall in systolic pressure on 24-hour ambulatory blood pressure monitoring, but neither overall daytime nor nighttime blood pressure values were significantly different (Table III). Compared with published normal values, only three control subjects (5%) and no patient with KD had resting systolic blood pressure above the 95th percentile based on height, and only one control subject and no patient with KD had resting diastolic blood pressure above the 95th percentile based on height.15 Regarding the assessment of BAR, there were no clinically or statistically significant differences with regards to 246

McCrindle et al

Characteristic

Control subjects (n ⴝ 60)

Patients with KD (n ⴝ 52)

Mean resting blood pressure (mm Hg): Systolic 113 ⫾ 10 107 ⫾ 12 Diastolic 64 ⫾ 7 61 ⫾ 9 Mean of daytime values (mm Hg): Systolic 114 ⫾ 9 113 ⫾ 10 Diastolic 68 ⫾ 6 67 ⫾ 5 Mean of nighttime values (mm Hg): Systolic 107 ⫾ 7 108 ⫾ 8 Diastolic 61 ⫾ 4 61 ⫾ 5 Mean % change daytime to nighttime: Systolic ⫺6.4 ⫾ 5.4% ⫺4.4 ⫾ 5.3% Diastolic ⫺10.4 ⫾ 7.6% ⫺8.8 ⫾ 6.0%

P value

.001 .04

.40 .26

.54 .90

.052 .22

KD, Kawasaki disease; mm Hg, millimeters of Mercury.

FMD or nitroglycerine-mediated dilatation between patients and control subjects (Figure 1).

Factors Associated with Endothelial Function for Patients with KD Within the group of patients with KD, factors associated with endothelium-dependent FMD were assessed, particularly any association with coronary artery involvement category. There was no significant relationship to either the age of the patient or the date at diagnosis, sex, number of days of fever, number of diagnostic criteria, the maximum erythrocyte sedimentation rate and the maximum platelet counts, and treatment with intravenous gamma globulin or aspirin. BAR was not significantly related to coronary artery involvement category (Figure 2). Likewise, there was no relationship between FMD and age or Tanner staging at the time of the study, and adiposity as assessed as a percentage of actual weight for ideal weight based on height percentiles, body mass index, or skinfold thickness percentiles. Regarding the relationship between FMD and the assessed atherosclerosis risk factors, there was no significant association with family history, self-reported physical activity and sedentary pursuit times, current smoking and smoke exposure, or dietary composition. Lower FMD was significantly associated with dietary intake, including higher caloric intake (r ⫽ ⫺0.31; P ⬍ .05), higher percent calories from fat (r ⫽ ⫺0.33; P ⫽ .04) and higher intake of saturated fat (r ⫽ ⫺0.31; P ⬍ .05). There was no significant relationship with fasting total cholesterol, LDL or HDL cholesterol, or apolipoproteins AI or B, or lipoprotein (a). Lower FMD was significantly associated with higher fasting triglyceride levels (r ⫽ ⫺0.28; P ⬍ .05), with 23% of patients The Journal of Pediatrics • September 2007

Figure 1. Comparison of patients with Kawasaki disease with healthy control subjects regarding brachial artery reactivity. Box plots depict the minimum, 25th percentile, median, 75th percentile, and the maximum value.

Figure 2. Effect of current coronary artery involvement on brachial artery reactivity in patients with Kawasaki disease. “Regressed” and “persistent” refer to coronary artery aneurysms. Box plots depict the minimum, 25th percentile, median, 75th percentile, and the maximum value.

having a fasting triglyceride level above 1.30 mmol/L. Regarding other laboratory variables, FMD was not significantly associated with fasting insulin or insulin C-peptide levels, hemoglobin A1c, homocysteine or fibrinogen levels, or urinary microalbumin. Lower FMD was significantly associated with higher fasting serum glucose (r ⫽ ⫺0.30; P ⫽ .04) and lower free fatty acid levels (r ⫽ 0.56; P ⬍ .001), with no patient having abnormally high glucose and 10% having high free fatty acid levels. In multivariable analysis (model R2 ⫽ 0.37), only higher fasting triglyceride levels (P ⫽ .04) and lower free fatty acid levels (P ⬍ .001) were independently related to lower FMD. After adjustment for these significant factors, as well as for age and sex, there remained no significant relationship between coronary artery involvement category and FMD (P ⫽ .18). This remained true even when the categories of persistent and regressed coronary artery involvement were grouped together (P ⫽ .12).

DISCUSSION Patients who have had KD may be at increased risk for accelerated atherosclerosis because of a number of potential mechanisms. First, the arterial damage secondary to the disease process itself may alter the arterial structure and function in such a way as to initiate and propagate the atherosclerotic disease process. Second, possible ongoing inflammation secondary to the disease process could further promote atherosclerosis, as well as cause alterations in traditional and non-

traditional atherosclerosis risk factors. Third, patients who have had KD may be predisposed to having other types of atherosclerosis risk factors. We noted few significant differences in atherosclerosis risk factors between patients with KD and healthy control subjects. Adiposity measures were similar, with both groups showing higher than normal body mass index. Patients tended to be more sedentary, had lower apolipoprotein AI and higher hemoglobin A1c levels, and despite lower resting blood pressure levels they had a tendency toward less nighttime fall. Earlier studies have shown that HDL cholesterol levels may be depressed acutely,16,17 but there is controversy whether this persists in the long term. Mitani and colleagues, as a part of study of inflammation late after KD, noted no differences between patients with KD and control subjects regarding body mass index, systolic blood pressure, and total cholesterol and HDL cholesterol levels.6 Cheung et al noted no significant differences regarding body mass index, systolic and diastolic blood pressure, and total cholesterol and LDL cholesterol levels.5 Patients both with and without coronary artery aneurysms had significantly higher levels of apolipoprotein B and lower levels of HDL cholesterol and apolipoprotein AI compared with the normal control subjects. Noto et al noted that patients with KD had higher levels of glycosylated hemoglobin.11 The observation of HDL cholesterol and glycosylated hemoglobin abnormalities may be evidence for the presence of a persistent state of inflammation,18,19 and inflammation has been implicated as a mechanism for the development of atherosclerosis.20 Suzuki et al noted evidence of inflammation at autopsy in the otherwise normal coronary arteries of a patient 13 months after KD.8 Recent reports have noted higher levels of inflammatory markers only in those patients with KD with persistent coronary artery lesions.6,9 A limitation of our study was that we did not measure inflammatory markers. Systemic endothelial dysfunction did not appear to be present after KD, and FMD was not significantly related to patient and KD characteristics, similar to our previous report with smaller numbers of subjects.7 Dhillon et al first reported decreased FMD, but they did not find any relationship to coronary artery involvement.10 In contrast, Ikemoto et al recently reported decreased FMD only in those patients with KD with moderate or severe coronary artery abnormalities.13 Abnormalities of carotid intima-media thickness, arterial stiffness, and impaired fibrinolytic capacity have also been reported in patients with KD with and without coronary artery complications.11,12,14 Discrepancies between our results and other reports may relate to differences in the patient with KD and control subject populations, or the ultrasonography assessments. FMD was found to be significantly related to blood glucose, triglyceride, and free fatty acid levels, an observation that has not been previously reported. The positive correlation between free fatty acid concentrations and FMD is surprising, considering that previous studies have suggested that elevated

Are Patients after Kawasaki Disease at Increased Risk for Accelerated Atherosclerosis?

247

levels cause endothelial dysfunction.21,22 However, it should be noted that Steinberg et al could not find any relationship between fasting free fatty acid concentrations and FMD.21 Fasting free fatty acid levels exhibit so much variability in concentration and in composition that it is difficult to reconcile conflicting results of studies. Although most studies have shown that high free fatty acid levels have an inverse correlation with FMD, some have shown contradictory results. Observations that saturated and trans-fatty acids do not impair FMD are attributed to older subjects, positioning of the cuff used to measure FMD, and time of measurement.23 In a separate investigation, long-chain fatty acids were found to attenuate FMD.24 Fatty acids have also been observed to play a role in increased FMD in hand veins of study subjects via a cyclooxygenase-dependant mechanism.25 No consensus has been reached as to the effect of free fatty acids on FMD, and our findings indicate that further study is indicated. For patients who have had KD, the degree of coronary artery involvement does not appear to be significantly associated with systemic endothelial function, even after adjustment for atherosclerosis risk factors. General assessment and counseling regarding healthy lifestyles is indicated.

REFERENCES 1. Furuyama H, Odagawa Y, Katoh C, Iwado Y, Ito Y, Noriyasu K, et al. Altered myocardial flow reserve and endothelial function late after Kawasaki disease. J Pediatr 2003;142:149-54. 2. Furuyama H, Odagawa Y, Katoh C, Iwado Y, Yoshinaga K, Ito Y, et al. Assessment of coronary function in children with a history of Kawasaki disease using (15)O-water positron emission tomography. Circulation 2002;105:2878-84. 3. Iemura M, Ishii M, Sugimura T, Akagi T, Kato H. Long term consequences of regressed coronary aneurysms after Kawasaki disease: vascular wall morphology and function. Heart 2000;83:307-11. 4. Muzik O, Paridon SM, Singh TP, Morrow WR, Dayanikli F, Di Carli MF. Quantification of myocardial blood flow and flow reserve in children with a history of Kawasaki disease and normal coronary arteries using positron emission tomography. J Am Coll Cardiol 1996;28:757-62. 5. Cheung YF, Yung TC, Tam SC, Ho MH, Chau AK. Novel and traditional cardiovascular risk factors in children after Kawasaki disease: implications for premature atherosclerosis. J Am Coll Cardiol 2004;43:120-4. 6. Mitani Y, Sawada H, Hayakawa H, Aoki K, Ohashi H, Matsumura M, et al. Elevated levels of high-sensitivity C-reactive protein and serum amyloid-A late after Kawasaki disease: association between inflammation and late coronary sequelae in Kawasaki disease. Circulation 2005;111:38-43.

248

McCrindle et al

7. Silva AA, Maeno Y, Hashmi A, Smallhorn JF, Silverman ED, McCrindle BW. Cardiovascular risk factors after Kawasaki disease: a case-control study. J Pediatr 2001;138:400-5. 8. Suzuki A, Miyagawa-Tomita S, Komatsu K, Nakazawa M, Fukaya T, Baba K, et al. Immunohistochemical study of apparently intact coronary artery in a child after Kawasaki disease. Pediatr Int 2004;46:590-6. 9. Cheung YF, Ho MH, Tam SC, Yung TC. Increased high sensitivity C reactive protein concentrations and increased arterial stiffness in children with a history of Kawasaki disease. Heart 2004;90:1281-5. 10. Dhillon R, Clarkson P, Donald AE, Powe AJ, Nash M, Novelli V, et al. Endothelial dysfunction late after Kawasaki disease. Circulation 1996;94:2103-6. 11. Noto N, Okada T, Yamasuge M, Taniguchi K, Karasawa K, Ayusawa M, et al. Noninvasive assessment of the early progression of atherosclerosis in adolescents with Kawasaki disease and coronary artery lesions. Pediatrics 2001;107:1095-9. 12. Albisetti M, Chan AK, McCrindle BW, Wong D, Vegh P, Adams M, et al. Fibrinolytic response to venous occlusion is decreased in patients after Kawasaki disease. Blood Coagul Fibrinolysis 2003;14:181-6. 13. Ikemoto Y, Ogino H, Teraguchi M, Kobayashi Y. Evaluation of preclinical atherosclerosis by flow-mediated dilatation of the brachial artery and carotid artery analysis in patients with a history of Kawasaki disease. Pediatr Cardiol 2005;26:782-6. 14. Cheung Y, Wong SJ, Ho MH, Relationship between carotid intima-media thickness and arterial stiffness in children after Kawasaki disease. Arch Dis Child 2007;92:43-7. 15. The fourth report on the diagnosis, evaluation, and treatment of high blood pressure in children and adolescents. Pediatrics 2004;114:555-76. 16. Newburger JW, Burns JC, Beiser AS, Loscalzo J. Altered lipid profile after Kawasaki syndrome. Circulation 1991;84:625-31. 17. Salo E, Pesonen E, Viikari J. Serum cholesterol levels during and after Kawasaki disease. J Pediatr 1991;119:557-61. 18. Esteve E, Ricart W, Fernandez-Real JM. Dyslipidemia and inflammation: an evolutionary conserved mechanism. Clin Nutr 2005;24:16-31. 19. Wu T, Dorn JP, Donahue RP, Sempos CT, Trevisan M. Associations of serum C-reactive protein with fasting insulin, glucose, and glycosylated hemoglobin: the Third National Health and Nutrition Examination Survey, 1988-1994. Am J Epidemiol 2002;155:65-71. 20. Lord RS, Bobryshev YV. Hallmarks of atherosclerotic lesion development with special reference to immune inflammatory mechanisms. Cardiovasc Surg 2002;10:405-14. 21. Steinberg HO, Tarshoby M, Monestel R, Hook G, Cronin J, Johnson A, et al. Elevated circulating free fatty acid levels impair endothelium-dependent vasodilation. J Clin Invest 1997;100:1230-9. 22. Tripathy D, Mohanty P, Dhindsa S, Syed T, Ghanim H, Aljada A, et al. Elevation of free fatty acids induces inflammation and impairs vascular reactivity in healthy subjects. Diabetes 2003;52:2882-7. 23. de Roos NM, Siebelink E, Bots ML, van TA, Schouten EG, Katan MB. Trans monounsaturated fatty acids and saturated fatty acids have similar effects on postprandial flow-mediated vasodilation. Eur J Clin Nutr 2002;56:674-9. 24. Steer P, Basu S, Lithell H, Vessby B, Berne C, Lind L. Acute elevations of medium- and long-chain fatty acid have different impacts on endothelium-dependent vasodilation in humans. Lipids 2003;38:15-9. 25. Stepniakowski KT, Lu G, Davda RK, Egan BM. Fatty acids augment endothelium-dependent dilation in hand veins by a cyclooxygenase-dependent mechanism. Hypertension 1997;30:1634-9.

The Journal of Pediatrics • September 2007

Protocol for Assessment of Brachial Artery Reactivity 1. Imaging and analysis equipment. The brachial artery was scanned above the antecubital fossa of the right arm using high-resolution vascular ultrasonography (ATL 3000 ultrasound machine, 7-15-MHz linear-array transducer, Advanced Technology Laboratories, Bothel, Wash). Longitudinal, electrocardiogram-gated, end-diastolic images were acquired of the brachial arterial diameter over a 1- to 2-cm segment, and computer-assisted edge detection brachial analysis software (DEA, Vasometrix, Montreal) was used to measure the brachial artery diameters. 2. Stimulus protocol. All patients were assessed after an overnight fast, with measurements taken between 8 and 10 AM and after the patient had rested for at least 10 minutes in the supine position. Brachial artery imaging was recorded on Super VHS videotapes, for 30 seconds (baseline), and after 5 minutes of reduced blood flow (induced by inflation of a pneumatic cuff placed at the mid upper arm to ⬎20 mm Hg above resting systolic blood pressure), recorded for 3 minutes after release of the cuff. Flow-mediated (endothelium-dependent) vasodilatation (FMD) was assessed as percentage change from baseline to maximal diameter of the brachial artery with reactive hyperemia. After a further 10-minute supine rest, the brachial artery imaging was recorded for a further 30 seconds (repeat baseline), then for 1 minute, at 3 to 4 minutes after a single 400-␮g dose of sublingual glyceryl trinitrate (GTN). GTN-mediated (endothelium-independent) vasodilatation was assessed as percentage change from baseline to maximal diameter of the brachial artery

post-GTN. All scans were performed by experienced vascular sonographers, and measurements were obtained afterward using the automatic edge-detection algorithms with pre-defined acceptable confidence intervals by a single observer.

Protocol for Assessment of 24-hour Ambulatory Blood Pressure Monitoring 1. Measurement protocol and settings. A Spacelab ambulatory blood pressure monitor model #90217 (www.spacelabs.com) was used. An appropriate-sized blood pressure cuff was placed on the nondominant arm, and recordings started in the morning after clinical assessment. Monitors were set to record blood pressure every 15 minutes from 7 AM to 10 PM, and every 30 minutes from 10 PM to 7 AM. Analysis of recordings was performed using computer software specific to the monitor. 2. Patient instructions. All patients were given written instructions and a diary to record events hourly (eg, walking, sleeping, watching television, doing exercise). If symptoms occurred, they were instructed to push the event button to initiate a recording. Patients were instructed that the monitor would beep before a recording would occur, and then they were to keep still, try to relax, and stop any activities while a recording was being made. They were to check occasionally to ensure that tubing was not kinked. During sleep the monitor was to be kept under a pillow to minimize noise. Patients were instructed to turn the monitor off at the end of the 24-hour monitoring period, and return it.

Are Patients after Kawasaki Disease at Increased Risk for Accelerated Atherosclerosis?

248.e1

High-Dose Ibuprofen in Cystic Fibrosis: Canadian Safety and Effectiveness Trial LARRY C. LANDS, MD, PHD, RUTH MILNER, PHD, ANDRÉ M. CANTIN, MD, DAVID MANSON, MD,

AND

MARY COREY, PHD

Objective To assess the effectiveness and safety of high-dose ibuprofen when used as part of routine therapy in patients with cystic fibrosis (CF). Study design In this multicenter, double-blinded, placebo-controlled trial, a total of 142 patients age 6 to 18 years with mild lung disease (forced expiratory volume in 1 minute [FEV1] > 60 predicted) were randomized to receive either high-dose ibuprofen (70 subjects, 20 to 30 mg/kg/twice daily, adjusted to a peak serum concentration of 50 to 100 ␮g/mL) or placebo (72 subjects) for a 2-year period. The primary outcome was the annualized rate of change in FEV1% predicted. Results The patients in the high-dose ibuprofen group exhibited a significant reduction in the rate of decline of forced vital capacity percent predicted (0.07 ⴞ 0.51 vs –1.62 ⴞ 0.52; P ⴝ .03), but not FEV1%. The ibuprofen group also spent fewer days in hospital after adjusting for age (1.8 vs 4.1 days per year; P ⴝ .07). A total of 11 patients (4 in the ibuprofen group and 7 in the placebo group) withdrew due to adverse events. Conclusions High-dose ibuprofen has a significant effect on slowing the progression of lung disease in CF and generally is well tolerated. (J Pediatr 2007;151:249-54) ystic fibrosis (CF) is the most common lethal genetic disorder affecting the Caucasian population, with an incidence of about 1:3200 live births in North America.1 Although median survival in CF is now in the fourth decade of life, most patients eventually succumb to progressive respiratory disease. Lung disease in CF is characterized by an exuberant neutrophilic inflammation. A study of alternate-day prednisone administration demonstrated beneficial effects on lung function.2 However, the presence of adverse events, including growth retardation and cataracts, limits the ability to use systemic corticosteroids therapy for prolonged periods. A recent 6-month trial with azithromycin, a macrolide, demonstrated beneficial pulmonary effects.3 The effect of azithromycin on the rate of decline in pulmonary function, which will have a long-term impact on survival,4 was not evaluated, however. A 48-week trial of inhaled nebulized hypertonic saline solution failed to slow the rate of decline in lung function.5 In 1995, Konstan et al6 demonstrated that high-dose ibuprofen therapy could slow the progression of lung disease in CF, especially in children. Analyzed as a change in the rate of progression of lung disease (percent of forced expiratory volume in 1 minute [FEV1%] predicted), this therapy has shown promise. In the earlier study, however, the sample size was only 49 children under age 13 years, and 90% came from 1 center. Furthermore, concerns over the safety of long-term use of high-dose ibuprofen has limited its use.7 To investigate the long-term use of ibuprofen in a larger population, we undertook a multicenter, double-blinded, randomized clinical trial to examine the effectiveness and safety of high-dose ibuprofen in children with CF when used as part of routine care.

C

METHODS This was a multicenter, double-blinded, placebo-controlled trial.

ACTG CF

AIDS Clinical Trial Group Cystic fibrosis

FEV1 FVC

Forced expiratory volume in 1 minute Forced vital capacity

See editorial, p 228 From the Department of Pediatrics, Montreal Children’s Hospital-McGill University Health Center, Montreal, Quebec, Canada (L.L.); Center for Clinical Epidemiology and Evaluation, University of British Columbia, Vancouver, British Columbia, Canada (R.M.); Department of Medicine, University of Sherbrooke, Sherbrooke, Quebec, Canada (A.C.); Department of Diagnostic Imaging (D.M.), and Department of Pediatrics (M.C.), Toronto Hospital for Sick Children, Toronto, Ontario, Canada. Supported by the Canadian Institutes of Health Research and the Canadian Cystic Fibrosis Foundation. L. Lands, R. Milner, and M. Corey all participated in the study design and the acquisition and interpretation of the data. A. Cantin participated in the study design. D. Manson performed the chest radiograph scoring. All authors read and approved the final version of the manuscript. L. Lands wrote the first draft and received no honorarium or payment of any kind. Submitted for publication Sep 30, 2006; last revision received Dec 29, 2006; accepted Apr 9, 2007. Reprint requests: Larry C. Lands, Montreal Children’s Hospital, Room D-380, 2300 Tupper Street, Montreal, Quebec, Canada H3H 1P3. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.04.009

249

Patients The diagnosis of CF was based on a sweat chloride concentration ⬎60 mmol/L and a compatible clinical history. Children were recruited from 12 CF centers across Canada between September 1998 and August 2000. These centers were estimated to be able to enroll at least 6 patients who met the inclusion and exclusion criteria for the trial. Children were eligible if they were age 6 to 18 years and had an FEV1 of ⬎60% predicted8 at the time of entry into the trial, with no hospitalizations in the previous 2 months. Children were excluded if they had taken systemic corticosteroids or nonsteroidal anti-inflammatory agents for more than 1 month in the past year; had a hepatic, renal, or hematologic disorder or coagulopathy; had documented evidence of peptic ulcer disease (endoscopy) or allergic bronchopulmonary aspergillosis; or had a history of hypersensitivity reaction to nonsteroidal anti-inflammatory agents. The study received ethical approval in each of the participating centers, and written informed consent was received for all participants. Sample Size Our sample size was based on the effect of placebo treatment on the slope of the FEV1 in the intention-to-treat pediatric group of Konstan et al.6 Because post hoc subgroup analysis in that study may have overestimated the degree of response, we based our sample size estimate on a more conservative expected difference between the groups (4% for control vs 2% for treated). Using a mixed-model regression analysis9 and a standard deviation for the slope of 7.5%/year, we found that a 2-year study with 80% power required 220 patients in each study arm (2-sided ␣ ⫽ 0.05). Allocation Within each clinic, patients were allocated using a predefined block-randomization schedule. Randomization tables were prepared by the central pharmacy in Toronto, and coded packages of ibuprofen or placebo tablets were shipped to the recruiting clinic to be dispensed to the patients. Patients, caregivers, and study personnel were all blinded to treatment assignment. The central pharmacy kept the coded treatment assignment list on which individual assignments were sealed in a special form. Breaking of the code involved tearing off a cover sheet for the individual study subject by 1 of 2 designated research pharmacists. During the study, the code was broken for an individual patient only if requested by the Safety and Monitoring Committee. It was broken for all patients at study termination. No stratification was performed, and no subgroup analysis was planned. Intervention Dose. All patients underwent a baseline pharmacokinetic study (baseline concentration and every hour for 3 hours), using 200-mg tablets (Upjohn-Pharmacia) at a dose of 20 to 30 mg/kg to a maximum of 1600 mg. All ibuprofen levels were analyzed by high-pressure liquid chromatography 250

Lands et al

in the laboratory of J.V. Aranda (Division of Clinical Pharmacology and Toxicology, Children’s Hospital of Michigan, Detroit, MI). The number of assigned pills was then adjusted by the coordinating pharmacologist to provide a peak plasma concentration of 50 to 100 ␮g/mL for each patient in the study. The number of pills was adjusted regardless of whether the patient was subsequently assigned to active treatment or to placebo, and dosages were not rechecked. The patient was then instructed to take the prescribed number of pills (ibuprofen or placebo) twice daily. Participating centers were advised about possible interactions between ibuprofen and intravenous aminoglycosides, resulting in renal insufficiency.10 Centers were advised that study medication could be stopped during intravenous aminoglycoside administration. Advice on adjusting for other concomitant medications was not given. If a center decided to continue study medication during this time, then close monitoring of renal function was advised. Pharmacokinetic studies were to be repeated if any patient experienced a weight change of ⬎25% during the 2-year study period; this did not occur, however.

Measurements Patients were examined for the trial every 6 months for 2 years, but could be seen more often at the discretion of the treating center. At each visit, standing height was measured with the patient in stocking feet using a stadiometer, and weight was measured on an electronic balance with the patient lightly dressed. Height and weight were used to calculate standardized z scores and percent of ideal body weight, using the Centers for Disease Control 2000 growth standards.11 Expiratory spirometry was measured at each center using the same device at each visit and reported as a percentage of predicted value.8 Each center was responsible for conducting these tests using American Thoracic Society guidelines.12 The raw values were processed centrally to calculate the percent predicted values. At every visit, blood work was conducted to screen for complications; liver function (aspartate aminotransferase [serum glutamic oxaloacetic transaminase], alanine aminotransferase [serum glutamic pyruvic transaminase], gamma glutamyl transferase, total bilirubin), renal (serum electrolytes, blood urea nitrogen, creatinine, microscopic urinalysis), hematologic function (hemoglobin, white blood cell count, platelet count), and coagulation (prothrombin time/partial thromboplastin time) were assessed in each center. Values were considered abnormal as defined by AIDS Clinical Trial Group (ACTG) criteria (available at http://www.rcc.tech-res.com). Chest radiographs were performed at each visit. The films were read independently by a radiologist (D.M.) using the standard CF chest radiograph scoring system developed by Brasfield et al.13 Compliance At each visit, the pill containers were returned to the local pharmacist and new pills were issued for the next 6 The Journal of Pediatrics • September 2007

months of the study. The containers were returned to each center’s pharmacist for counting, at which point compliance was assessed. We estimated that 60% of the pills would be consumed; Konstan et al6 found compliance rates of 68% in the ibuprofen group and 72% in the placebo groups, with no variation by age. For a study of real-world clinical effectiveness, we considered that compliance assessment using pill counts was simpler and less expensive than random blood testing for ibuprofen.

Adverse Event Recording At each visit after randomization, an adverse event questionnaire was completed by interview with the local study coordinator, using the diary to assist memory. The coordinator also recorded the number of hospital admissions and length of stay, along with any concomitant therapy, including antibiotics and inhaled anti-inflammatory agents, such as corticosteroids. The common adverse events related to ibuprofen are primarily gastrointestinal symptoms, such as epigastric discomfort, heartburn, cramping, and gastric or duodenal ulceration. However, many of these symptoms are known to occur in patients with CF as part of the natural history of the disease. We used the adverse symptom questionnaire of Konstan et al6 and recorded the incidence of symptoms at each visit. Patient Withdrawal A patient could withdraw from the trial at his or her discretion, or could be withdrawn at the discretion of the treating physician if any of the following occurred: 1. Abdominal discomfort resolving with drug removal but recurring with drug reintroduction. 2. Upper gastrointestinal bleeding, documented by radiology or endoscopy. Monitoring for fecal occult blood was not included because although it is a relatively simple and inexpensive test, it is overly sensitive and most likely would trigger many further, more complex investigations. 3. Visual disturbances (eg, blurred or diminished vision, scotomata, or changes in color perception), resulting in immediate drug cessation and ophthalmologic examination. Medication could be restarted if the ophthalmologic examination suggested that the disturbance could be explained by a cause other than study medication. 4. If changes in renal, hepatic, or hematologic function occurred that could be classified in the mildly toxic range (according to ACTG Pediatric Toxicity tables), and remained in this range two weeks later while the subject remained on the study medication. If the abnormalities persisted, then the drug was discontinued, if no reasonable alternative explanation existed. If values were above those in the ACTG mild range, then the study drug was stopped immediately. The Safety Monitoring Committee decided a priori to stop the trial if (1) the number of subjects voluntarily with-

drawing from the trial exceeded 20% of those enrolled; (2) after breaking the blinding code for the above-mentioned adverse effects, the treatment group had a 20% higher incidence of these effects compared with controls; or (3) after breaking the code, the incidence of life-threatening adverse events requiring admission to an intensive care unit or hospital were 5% higher in the treatment group. No interim analysis was planned a priori.

Data Analysis An intention-to-treat analysis was performed on all available information on each subject, including those subjects who did not complete 2 years of study follow-up. Baseline data were assessed for balance between groups. The primary outcome variable, the annual rate of change in FEV1%, was analyzed using a mixed-model analysis of variance. Mixedmodel analysis allows for efficient use of all data, because the mean slopes are not distorted by subjects with incomplete data. This model effectively weights the contributions of individuals with shorter follow-up or missed observations, ensuring that the estimates of mean and variance of the slope are not biased. The mixed model was also used to analyze secondary outcomes, including predicted FVC1%, anthropometric data, and chest radiograph scores. The number of hospitalizations, adverse effects, and use of concomitant therapy were compared by ␹2 analysis. The number of days in the hospital per year of follow-up was analyzed by Poisson regression, with Pearson adjustment for overdispersion where indicated. Role of Funding Sources The sponsors of the study and the company supplying the study medication played no role in study design, data collection, data analysis, data interpretation, or writing of this report.

RESULTS A total of 142 patients (70 in the ibuprofen group and 72 in the placebo group) were enrolled in the trial (Figure 1). Recruitment was stopped at that point, because a survey of participating centers revealed that prolonging the recruitment period would produce no significant increase in enrollments. The groups had similar baseline characteristics (Table I). At study enrollment, the patients had mild lung disease (mean FEV1 ⬎90%) and were reasonably well nourished, with mean z scores for height and weight approximately 1/2 standard deviation below the normal median values for age and sex. The difference in mean annual rate of decline in FEV1% predicted was not statistically significant (⫺2.69 ⫾ 0.57 for placebo vs ⫺1.49 ⫾ 0.57 for ibuprofen; P ⫽ .14). However, a significant decrease in the annual rate of decline of FVC% predicted was seen in the ibuprofen group (⫺1.62 ⫾ 0.52 for placebo vs ⫺0.07 ⫾ 0.51 for ibuprofen; P ⫽ .03) (Figure 2). There were no significant differences in changes in the predicted maximum midexpiratory flow (FEF25%-75%) (data not shown).

High-Dose Ibuprofen in Cystic Fibrosis: Canadian Safety and Effectiveness Trial

251

Figure 1. Flowchart of patient recruitment and participation. Figure 2. Annualized rate of change in lung function in the ibuprofen and placebo groups.

Table I. Subject baseline characteristics Placebo group (n ⴝ 72) Age Height z score Weight z score Weight, % of ideal Body mass index z score FEV1% predicted FVC% predicted

11.1 ⫾ 3.1 ⫺0.50 ⫾ 0.97 ⫺0.48 ⫾ 0.96 101 ⫾ 13.0 ⫺0.28 ⫾ 0.91 91.0 ⫾ 17.4 96.2 ⫾ 18.0

Ibuprofen group (n ⴝ 70) 12.0 ⫾ 3.3 ⫺0.52 ⫾ 0.91 ⫺0.44 ⫾ 0.97 102 ⫾ 14.6 ⫺0.21 ⫾ 0.95 92.3 ⫾ 18.4 95.2 ⫾ 14.0

All values are mean ⫾ standard deviation.

Chest radiograph scores were available for 62 participants receiving placebo and 60 participants receiving ibuprofen. The mean scores at baseline demonstrated moderate abnormality compared with a perfect score of 25 and did not differ between the groups (placebo, 19.5 ⫾ 2.23; ibuprofen, 19.7 ⫾ 2.11). The mean changes in radiograph scores over the course of the study were not significantly different between the 2 groups (placebo, ⫺0.61 ⫾ 2.33; ibuprofen, ⫺0.54 ⫾ 2.20; P ⫽ .9). Nutritional status (weight and body mass index z scores) did not change significantly in either group over the course of the study. The frequency of hospitalization was quite low and did not differ significantly between the 2 groups (1 or more hospitalizations over the 2-year period, 36% in the placebo group vs 27% in the ibuprofen group [P ⫽ .3]; 1 or more respiratory hospitalizations, 26% vs 20% [P ⫽ .4]; 1 or more gastrointestinal hospitalizations, 6% vs 4% [P ⫽ .7]). Total days spent in the hospital during the randomized follow-up period were 561 for the placebo group and 248 for the ibuprofen group. Poisson regression analysis of days per year of follow-up, with a Pearson adjustment for overdispersion, gave hospitalization rates of 4.1 days per year in the placebo group and 1.8 days per year in the ibuprofen group (P ⫽ .07). 252

Lands et al

Table II. Adverse events leading to study withdrawal Event

1 2

Conjunctivitis Abdominal cramps, nausea, diarrhea Abdominal pain, gastritis

Ibuprofen 4 months Ibuprofen 2.5 months

Epigastric pain, nausea, diarrhea Abdominal pain, reflux esophagitis Abdominal pain, admitted with hepatitis Nausea, vomiting, tinnitus Reactive arthritis Elevated liver enzyme levels

Placebo

14 months on and off 3 weeks

Placebo

12 months

3 4 5 6 7 8 9 10 11

Gastrointestinal bleeding Abdominal pain

Group

Time in study

Patient

Placebo

Placebo

9 months on and off Ibuprofen 18 months Placebo 8 months Placebo 18 months on and off Ibuprofen 23.5 months Placebo 1.5 months

Post hoc analysis of days in the hospital revealed a significant age factor (P ⫽ .026); that is, older patients spent more days in the hospital than younger patients. Including age in the Poisson regression model explained more of the variance, and the treatment group difference was significant (P ⫽ .024). This analysis suggests that the doubling of the days spent in the hospital for patients in the placebo group represents a true treatment effect. A total of 18 patients (9 in each group) did not complete the full 2 years of follow-up. Eleven withdrew due to adverse events (4 in the ibuprofen group and 7 in the placebo group); Table II presents results in the order in which the patients withdrew. As far as we know, all centers discontinued the study drug during treatment with intravenous aminoglycosides. There were no changes in renal, The Journal of Pediatrics • September 2007

hepatic, or hematologic values that warranted discontinuation of the study drug. There were no differences in the percentage of patients taking concomitant therapy (anti-inflammatory, inhaled corticosteroids, inhaled or oral antibiotics, bronchodilators, pancreatic enzymes, vitamins) at the beginning of the trial or any time during the study. There was no discernible center effect for any of the outcomes.

DISCUSSION We found that high-dose ibuprofen significantly slowed the decline in FVC in a group of children and adolescents with CF who initially had well-preserved lung function. The ibuprofen group had fewer days spent in the hospital. Ibuprofen generally was well tolerated, with only 4 subjects withdrawing due to adverse events. Our findings are consistent with those of Konstan et al,6 who found that children with mild lung disease benefited the most from this therapeutic strategy. Compared with their patients, the patients in the current study were younger and had better initial lung function and nutritional status. However, their placebo group, whether taken as the entire group or limited to those under age 13 years, had greater annual losses in lung function (FEV1, 3.6%/year for all vs 4.2%/year for those under age 13) compared with our group (⫺2.7%/year). When it became clear after 1 year of enrollment that we could not achieve our projected sample size, the investigative team, after consulting with the Safety Monitoring Committee, decided to continue the underpowered study, with the main focus on safety outcomes. We encountered few adverse events leading to withdrawal. The main impediment to recruitment was similar to that reported for previous studies,7 concerns about the risk of adverse events, especially gastrointestinal events. One patient experienced significant gastrointestinal bleeding. At the time of the event, we were not advising any prophylactic therapy, such as H2 antagonists; but after the event, on the advice of the Safety Monitoring Committee, all centers were advised that they could prescribe H2 antagonists for gastrointestinal protection. We do not know the extent of adherence to this advisory. A recent single-center study of high-dose ibuprofen did not find any improvement in lung function, but did report significant adverse events, primarily gastrointestinal.14 In that study, the sample size was not powered to detect significant declines in lung function, and only 4 of 9 patients with significant gastrointestinal complaints received H2 antagonists or other gastrointestinal protection. More recently, concern has been expressed about the long-term use of nonsteroidal anti-inflammatory drugs, including ibuprofen, and the risk of cardiovascular events.15 This may be related to the relatively low dosage of these medications prescribed to treat conditions other than CF. Konstan et al16 have demonstrated that low-dose ibuprofen can actually increase neutrophil recruitment; however, when peak serum levels of 50 to 100 ␮g/mL are achieved, neutrophil recruitment is diminished in both CF and healthy sub-

jects. Therefore, low-dose ibuprofen actually may be proinflammatory. One other impediment to using high-dose ibuprofen is the need for therapeutic monitoring, including pharmacokinetics every 2 to 3 years and monitoring for adverse renal, hepatic, and hematologic effects.17 The costs of such monitoring and of the ibuprofen itself are relatively minor compared with those for most current therapies, however. We did not achieve our recruitment goals, and our control group did better than we had predicted when calculating our required sample size. Nonetheless, we did achieve a virtual stoppage (⬍0.1%) in the rate of decline in FVC, along with fewer days in the hospital after adjusting for age. Due to a wide variance in the rate of decline in FEV1, our 45% decrease in the rate of decline in the ibuprofen group was not statistically significant. To detect a significant change in the rate of decline of FEV1 requires a study lasting at least 1 year with a relatively large sample size.4 As general results for patients with CF continue to improve, the sensitivity of spirometry as an outcome measure diminishes, yet no new markers have been firmly established. However, our results demonstrate that adverse events occurred less often with ibuprofen than perceived by the treating community in CF. In summary, high-dose ibuprofen in patients with CF was found to be safe and to have a positive impact on the rate of decline in lung function and on the duration of hospitalization. This study supports and extends the original observations of Konstan et al.6,16 Slowing the rate of progression will result in enhanced longevity and quality of life for patients with CF.

REFERENCES 1. Dupuis A, Hamilton D, Cole DEC, Corey M. Cystic fibrosis birth rates in Canada: a decreasing trend since the onset of genetic testing. J Pediatr 2005;147:312-5. 2. Eigen H, Rosenstein B, FitzSimmons S, Schidlow D. A multicenter study of alternate-day prednisone therapy in patients with cystic fibrosis. J Pediatr 1995;126: 515-23. 3. Equi A, Balfour-Lynn IM, Bush A, Rosenthal M. Long-term azithromycin in children with cystic fibrosis: a randomised, placebo-controlled crossover trial. Lancet 2002;360:978-84. 4. Davis PB, Byard PJ, Konstan MW. Identifying treatments that halt progression of pulmonary disease in cystic fibrosis. Pediatr Res 1997;41:161-5. 5. Elkins MR. A controlled trial of long-term inhaled hypertonic saline in patients with cystic fibrosis. N Engl J Med 2006;354:229-40. 6. Konstan MW, Byard PJ, Hoppel CL, Davis PB. Effect of high-dose ibuprofen in patients with cystic fibrosis. N Eng J Med 1995;332:848-54. 7. Oermann CM, Sockrider MM, Konstan MW. The use of anti-inflammatory medications in cystic fibrosis: trends and physician attitudes. Chest 1999;115:1053-8. 8. Wang X, Dockery DW, Wypij D, Fay ME, Ferris BG Jr. Pulmonary function between 6 and 18 years of age. Pediatr Pulmonol 1993;15:75-88. 9. Diggle PJ, Liang KY, Zager SL. Analysis of Longitudinal Data. Oxford, UK: Clarendon Press; 1994. 10. Kovesi TA, Swartz R, MacDonald N. Transient renal failure due to simultaneous ibuprofen and aminoglycoside therapy in children with cystic fibrosis. N Engl J Med 1998;338:65-6. 11. Ramsey BW, Farrell PM, Pencharz P. Nutritional assessment and management in cystic fibrosis: a consensus report. Am J Clin Nutr 1992;55:108-16. 12. American Thoracic Society. Standardization of spirometry, 1994 update. Am J Respir Crit Care Med 1995;152:1107-36. 13. Brasfield D, Hicks G, Soong S, Tiller RE. The chest roentgenogram in cystic fibrosis: a new scoring system. Pediatrics 1979;63:24-9. 14. Fennell PB, Quante J, Wilson K, Boyle M, Strunk R, Ferkol T. Use of high-dose ibuprofen in a pediatric cystic fibrosis center. J Cystic Fibrosis 2007;6:153-8.

High-Dose Ibuprofen in Cystic Fibrosis: Canadian Safety and Effectiveness Trial

253

15. Hippisley-Cox J, Coupland C. Risk of myocardial infarction in patients taking cyclo-oxygenase-2 inhibitors or conventional non-steroidal anti-inflammatory drugs: population-based nested case-control analysis. BMJ 2005;330:1366. 16. Konstan MW, Krenicky JE, Finney MR, Kirchner HL, Hilliard KA, Hilliard JB, et al. Effect of ibuprofen on neutrophil migration in vivo in cystic fibrosis and healthy subjects. J Pharmacol Exp Ther 2003;306:1086-91. 17. Prescott W, Johnson C. Anti-inflammatory therapies for cystic fibrosis: past, present, and future. Pharmacotherapy 2005;25:555-73.

APPENDIX Investigators Recruiting Patients in Their Center Danny Vaze, Janeway Hospital Dan Hughes, IWK Health Centre

254

Lands et al

Georges Rivard, Centre Hospitalier de l’Université Laval Normand Petit, Centre Hospitalier Rouyn-Noranda Larry Lands, Montreal Children’s Hospital Jacques-Edouard Marcotte, Hôpital Ste Justine Ian Maclusky, Toronto Hospital for Sick Children Linda Pedder, McMaster University Medical Centre Bryan Lyttle, Children’s Hospital of Western Ontario Vijay Kumar, Sudbury Regional Hospital Kumar Ramlall, Royal University Hospital Mark Montgomery, Alberta Children’s Hospital

The Journal of Pediatrics • September 2007

Primary Sclerosing Cholangitis in Childhood is Associated with Abnormalities in Cystic Fibrosis–Mediated Chloride Channel Function HARPREET PALL, MD, JULIAN ZIELENSKI, PHD, MAUREEN M. JONAS, MD, DEBORAH A. DASILVA, RN, KIMBERLY M. POTVIN, XIAO-WEI YUAN, MSC, QIUJU HUANG, MD, AND STEVEN D. FREEDMAN, MD, PHD

Objective To determine whether primary sclerosing cholangitis (PSC) in childhood is associated with abnormalities in cystic fibrosis transmembrane conductance regulator (CFTR). Study design Subjects with PSC diagnosed in childhood (n ⴝ 20) were recruited from Children’s Hospital. Subjects had testing with sweat chloride concentration, nasal transmembrane potential difference, and extensive genetic analysis of the CFTR gene. Disease control subjects consisted of 14 patients with inflammatory bowel disease alone and no liver disease. t Tests were performed to determine statistical significance. Results In the PSC group, CFTR chloride channel function (⌬Chloride free ⴙ isoproterenol) was markedly diminished at ⴚ8.6 ⴞ 8.2 mV (reference range: ⴚ24.6 ⴞ 10.4 mV). In contrast, disease control subjects had normal function, at ⴚ17.8 ⴞ 9.7 mV (P ⴝ .008). Sweat chloride concentration in subjects with PSC was greater than in disease control subjects (20.8 ⴞ 3.4 mmol/L vs 12.0 ⴞ 1.6 mmol/L, P ⴝ .045). Comprehensive CFTR genotyping revealed that 5 of 19 (26.3%) subjects with PSC had a CFTR mutation or variant, compared with 6 of 14 (42.9%) disease control subjects. Conclusions There is a high prevalence of CFTR-mediated ion transport dysfunction in subjects with childhood PSC. (J Pediatr 2007;151:255-9) ystic Fibrosis (CF) is characterized by abnormal secretion of fluid, electrolytes, and macromolecules by exocrine glands. To date, close to 1200 CF-causing mutations have been reported.1 In addition to the formation of inspissated secretions, cystic fibrosis transmembrane conductance regulator (CFTR) dysfunction results in an excessive host inflammatory response.2,3 Primary sclerosing cholangitis (PSC) is a slowly progressive cholestatic disease of unknown cause, characterized by fibro-obliterative inflammation of the biliary tract. Although the underlying cause and pathogenesis of PSC is not known, it is associated with inflammatory bowel disease (IBD) in approximately 80% of cases.4 Conversely, only about 4% of patients with IBD have or will have PSC, and it is impossible to predict who will do so. PSC and CF have features in common. Chronic low-grade inflammation and damage to bile ducts characterize both conditions. Due to their similarities, studies have examined the prevalence of CFTR mutations in adults with PSC. In one study, only 1 of 19 subjects with PSC had neither a CFTR mutation/variant nor the M470V genotype.5 CFTR function in these patients was decreased as measured by nasal transmembrane potential difference (NTPD). Another study failed to demonstrate an association of common CF disease– causing mutations with PSC.6 However, exhaustive genotyping as well as functional analyses were not performed in that study. In the current study, we hypothesized that a defect in CFTR mediated chloride channel function is present in subjects with PSC diagnosed in childhood (⬍18 years) as

C

CF CFTR ERCP IBD

Cystic fibrosis Cystic fibrosis transmembrane conductance regulator Endoscopic retrograde cholangiopancreatography Inflammatory bowel disease

MRCP NTPD PPAR PSC

Magnetic resonance cholangiopancreatography Nasal transmembrane potential difference Peroxisome proliferator activated receptor Primary sclerosing cholangitis

See editorial, p 230 From the Division of Pediatric Gastroenterology (H.P., M.J.), Children’s Hospital Boston, Harvard Medical School, Boston, MA; Program in Genetics and Genomic Biology (J.Z., X.-W.Y., Q.H.), The Hospital for Sick Children, Toronto, Ontario, Canada; and the Division of Gastroenterology (D.D., K.P., S.F.), Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA. Supported by the Thrasher Research Fund (grant 02820-4) and The General Clinical Research Centers at Children’s Hospital Boston (NIH M01 RR02172) and The Beth Israel Deaconess Medical Center (NIH M01 RR01032). Dr Pall was supported in part by the Harvard/MIT Clinical Investigator Training Program. This project was also partially funded by Genome Canada through the Ontario Genomics Institute as per research agreement 2004-OGI-3-05. The study sponsors did not have a role in the study design, data analysis, or preparation of the manuscript. Submitted for publication Nov 5, 2006; last revision received Feb 12, 2007; accepted Mar 30, 2007. Reprint requests: Dr Harpreet Pall, Children’s Hospital Boston, Division of Gastroenterology, 300 Longwood Avenue, Boston, MA 02115. E-mail: harpreet.pall@ childrens.harvard.edu. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.062

255

compared with disease control subjects with IBD and no liver disease. In contrast to the previous adult study by our group in a separate set of patients,5 subjects with PSC diagnosed in childhood may have no CFTR dysfunction. Alternatively, genetic diseases may be more likely to manifest in children than in adults if there are more severe mutations with a further decrease in CFTR function.

METHODS Patient Selection and Study Design This study was conducted at Children’s Hospital Boston and Beth Israel Deaconess Medical Center in Boston. Institutional review board approval was obtained at both hospitals. Written informed consent was obtained from the parents or guardians of the children who served as subjects of the investigation and, when appropriate, from the subjects themselves. Children with PSC diagnosed before the age of 18 years were recruited. The criteria for the diagnosis of PSC included presence of typical cholangiographic abnormalities of PSC involving bile ducts segmentally or extensively by endoscopic retrograde cholangiopancreatography (ERCP) or magnetic resonance cholangiopancreatography (MRCP), and/or liver biopsy demonstrating typical features of PSC. Patients were excluded if they had other liver disease, such as viral hepatitis, drug-induced liver disease, and metabolic or hereditary liver disease by conventional clinical, laboratory, and histologic criteria. PSC with features of autoimmune hepatitis (overlap syndrome) was not used as an exclusion criterion, as this is thought to be a subset of PSC. For entry into this study, subjects with PSC had to be at least 12 years of age, as this was the lower age limit to reliably tolerate NTPD. Disease control subjects were age-matched subjects with IBD and no liver disease. Disease control subjects had normal alanine aminotransferase and ␥-glutamyl transferase levels and did not undergo routine biliary imaging or liver biopsies. Diagnostic criteria for this group included clinical, radiologic, and/or histologic evidence of either ulcerative colitis or Crohn disease. There was no exclusion based on sex, race, and ethnic background. Subjects had functional testing for CFTR consisting of NTPD and sweat chloride measurements and underwent complete CFTR gene sequencing for mutations. Sweat Chloride All patients underwent sweat chloride testing by the quantitative pilocarpine iontophoresis method, described by Gibson and Cooke, at Children’s Hospital Boston.7 Results were classified as abnormal (⬎60 mmol/L), borderline (40 to 60 mmol/L), or normal (⬍40 mmol/L). These reference values are based on those recommended by the consensus statement from the United States Cystic Fibrosis Foundation, in which the diagnostic criteria have been revised to include the “atypical” CF phenotype.8 256

Pall et al

Nasal Transmembrane Potential Difference Nasal transmembrane potential difference was performed as described by Knowles et al.9 Basal potential difference (infusion PD) and the response to inhibition by perfusion of amiloride (⌬Amil) was recorded to reflect activity of inwardly directed sodium transport. A chloride-free solution was perfused with amiloride to measure basal chloride secretion, and a maximal chloride secretory response (⌬Cl-free ⫹ Iso) was then elicited by perfusing with isoproterenol. ⌬Clfree ⫹ Iso was considered to be abnormal if it was outside the 99% probability limits (7.65 to 22.6 mV) for healthy control subjects. Reference ranges were established from healthy individuals analyzed in a previous study from our group.10 NTPD testing was performed in the same time period by the same operator and was not blinded to diagnosis. Analysis of the CFTR Gene for DNA Alterations Patients underwent an exhaustive search for DNA alterations in the CFTR gene with multiplex, heteroduplex gel (mHET) analysis as previously described.11 PCR-amplified DNA fragments corresponding to all the CFTR exons, their flanking intron sequences, and the promoter region (up to1 kb upstream of exon 1) were examined by mHET and sequencing analyses. The estimated detection rate for this protocol is 95% of known CFTR gene mutations.11 In addition, variants in the polythymidine tract (5T, 7T and 9T) of intron 812 and the M470V (1540A ¡ G) polymorphism in exon 10 were studied.13 Abnormalities in these loci produce either less amounts of correctly spliced CFTR (T-tract) or subfunctional CFTR (M470V). Other genetic liver disease modifiers such as ␣-1-antitrypsin were not assessed. Statistical Analysis All phenotype measurements by NTPD were described as mean ⫾ SD. Sweat chloride concentration was the average of both arms and expressed as mean ⫾ SEM. t Tests were performed to determine statistical significance between groups.

RESULTS Demographics Twenty subjects with PSC were recruited, 19 of whom completed NTPD. There was a variable degree of portal tract fibrosis in the patients with PSC, with some having mild fibrosis and others more advanced disease and portal-toportal tract bridging fibrosis. Seven of 20 subjects with PSC had recent autoimmune markers. The diagnosis of overlap syndrome with autoimmune hepatitis in 4 patients was based on characteristic biopsy features accompanied by positive autoimmune markers. Fourteen disease control subjects were recruited, 12 of whom completed NTPD (Table I). Phenotype Testing The sweat chloride concentration in subjects with PSC was greater than in disease control subjects (20.8 ⫾ 3.4 The Journal of Pediatrics • September 2007

Table I. Demographics of PSC subjects and disease control subjects Characteristics

PSC

IBD and no liver disease

Age (y) Sex Male Female Ethnicity Caucasian Black Other IBD Ulcerative colitis Crohn disease None Overlap syndrome Sinusitis/asthma Family history of CF Family history of PSC

17.9 ⫾ 0.8

17.5 ⫾ 0.4

15 5

6 8

19 1 0

12 0 2

13 5 2 4 5 1 0

3 11 0 0 5 0 0

Age at the time of enrollment is expressed as the mean ⫾ SEM. Overlap syndrome is sclerosing cholangitis with features of autoimmune hepatitis based on liver biopsy.

mmol/L vs 12.0 ⫾ 1.6 mmol/L, P ⫽ .045), although still within normal range. With regard to NTPD, the maximum baseline potential difference was normal in both groups. Baseline potential difference in subjects with PSC was ⫺26.5 ⫾ 7.0 mV and in disease controls ⫺22.3 ⫾ 6.1 mV (normal reference range: ⫺23.0 ⫾ 8.1 mV).10 The response to inhibition of sodium uptake (⌬Amiloride) was also normal in both groups, with subjects with PSC at 15.0 ⫾ 6.9 mV and disease control subjects at 11.8 ⫾ 4.8 mV (normal reference range: 13.7 ⫾ 5.2 mV). CFTR chloride channel function (⌬Chloride free ⫹ isoproterenol, which reflects maximum CFTR-mediated chloride conductance) was markedly diminished in subjects with PSC, at ⫺8.6 ⫾ 8.2 mV compared with disease control subjects, in whom values were normal (⫺17.8 ⫾ 9.7 mV, P ⫽ .008) (normal reference range: ⫺24.6 ⫾ 10.4 mV) (Figure, A). The individual values are shown in the Figure (B). There was no correlation between severity of liver disease by biopsy and degree of CFTR dysfunction, although the numbers were small.

Genetic Testing For purposes of definition, mutations are defined as CF-causing changes; variants are associated with decreased CFTR function and/or present in other CFTR-associated diseases; polymorphisms are changes not linked to specific diseases. Comprehensive CFTR genotyping revealed that 5 of 19 (26.3%) subjects with PSC had a CFTR mutation or variant, compared with 6 of 14 (42.9%) disease control subjects (Table II). The R75Q variant was common in both groups of subjects and is currently not listed as a CF causing mutation, although its contribution to CF-like phenotypes cannot be excluded. Homozygosity for 470V (GG) at the 1540 locus produces a less functional CFTR protein variant

Figure. A, Isoproterenol response in subjects with PSC compared with disease control subjects. Overall mean and standard error bars (in mV) for isoproterenol response are shown for subjects with PSC (PSC) and disease control subjects (IBD). Mean for subjects with PSC was ⫺8.6 ⫾ 8.2 mV compared with disease control subjects, in whom values were normal (⫺17.8 ⫾ 9.7 mV, P ⫽ .008). Normal Ref Range: ⫺24.6 ⫾ 10.4 (SD).10 B, Individual isoproterenol responses. Individual mean isoproterenol response is shown for subjects with PSC (PSC) and disease control subjects (IBD Control). Note that 11 of the 19 (57.9%) subjects with PSC had a diminished response of ⬍7 mV, with no overlap with disease control subjects. These values are ⬎2 SD below the mean for the normal reference range. Testing was performed by the same operator.

and was identified in 4 of 19 subjects with PSC compared with 2 of 14 disease control subjects. Eight of 19 subjects with PSC had a polymorphism, as did 6 of 14 disease control subjects. Three disease control subjects had the T-tract 5/7. There is little data regarding the frequency of CFTR mutations/variants/polymorphisms in normal populations. The data indicate the following frequencies: IVS8T5, 3% to 5%, depending on population; R75Q, 3% to 6%, depending on population; 1716G ¡ A, 1% to 2% (personal communication from Julian Zielenski, Hospital for Sick Children, Toronto, Canada).

Primary Sclerosing Cholangitis in Childhood is Associated with Abnormalities in Cystic Fibrosis–Mediated Chloride Channel Function 257

Table II. Genotype-phenotype correlation Subject no. PSC 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 IBD 21 22 23 24 25 26 27 28 29 30 31 32 33 34

Mutations/Variants

Polymorphism

R75Q 4521G/A(8T/9T) 2694T/G

Sweat chloride (mmol/L)

⌬Cl ⴙ Iso (mV)*

7/7 7/7 7/7 7/9 7/7 7/7 7/9 7/7 7/7 7/9 7/9 7/7 7/7

20.2 19.7 14.8 19.4 NP 5.1 5.8 9.9 56.3 30.3 32.1 8.7 4.8 45.3 18.6 7.6 21.6 45 11 18.9

⫺1.2 ⫺17.3 ⫺3.9 ⫺13.3 NP ⫺12.5 ⫺17.8 ⫺12.8 ⫺5.1 ⫺5.5 ⫺29 ⫺2.4 ⫺3.8 ⫺2.2 ⫺0.45 ⫺4.5 ⫺1.8 1.2 ⫺11.5 ⫺20.1

7/9 7/7 7/7 5/7 7/7 7/7 5/7 7/7 7/7 7/7 7/7 7/7 7/7 5/7

8.9 15 17.8 4.8 10.8 22.9 15.1 20.1 10.1 14 3.5 11.8 9.4 4

⫺13 ⫺15 ⫺20 ⫺10.4 ⫺30 ⫺10.5 NP ⫺8 ⫺17 NP ⫺35 ⫺8 ⫺33 ⫺14

1540 locus

T Tract

GG GG AA AA AG GG

7/7 7/7 7/7 7/7 7/7 7/7

AA AG AA AA AG AG AG GG AA AG AG AA AG AG AG AG AA GG AG AG AG GG AA AG AA AG AG

NP E725K R75Q 4006 - 200G/A, 1233A/T 2694T/G 4521G/A, 4700T8/9 1525 - 61A/G 3030G/A R75Q 1001 ⫹ 11C/T 1716G ¡ A

S1235R/2752 - 26A ¡ G

1001 ⫹ 11C/T 185 ⫹ 324C/T

IVS8T5 875 ⫹ 40A/G, 125G/C IVS8T5 R75Q Q1352H

4521A(hom), 4700T8/8

R75Q/IVS8T5

125G/C 1898 ⫹ 152T/A

Subjects with IBD were disease control subjects. CFTR mutations shown in bold are E725K, S1235R, and 2752 - 26A ¡ G. The 1540 locus is presented as genotype AA, AG (M470V), or GG (470V). *(⌬Chloride free ⫹ isoproterenol). NP (not performed). Subjects 4 and 8 had no IBD. Subjects 7, 8, 12, and 16 had overlap syndrome.

DISCUSSION In this prospective study, we have demonstrated an increased prevalence of CFTR functional abnormalities in subjects with PSC diagnosed in childhood compared with disease control subjects. The mean values for the impairment in the chloride secretory response observed in subjects with childhood PSC was intermediate to that expected in patients with classic CF and healthy control subjects. It is noteworthy that 11 of 19 subjects with PSC had very low isoproterenol stimulated chloride secretory responses. The mean value of ⫺8.6 mV was lower than that seen in our previous study of adults with PSC (⫺14 mV median [⫺9, ⫺20, interquartile range]).5 NTPD values are not known to change as a function of age in older children and adults (personal communication 258

Pall et al

from Peter Durie and Lynda Ellis, Hospital for Sick Children, Toronto, Canada). A reduced chloride response has been reported in other single organ disorders associated with CFTR dysfunction, such as congenital bilateral absence of the vas deferens,14 idiopathic pancreatitis,10 and chronic rhinosinusitis.15 NTPD is more sensitive than sweat testing for the assessment of chloride ion secretion in disease states characterized by single organ involvement associated with mild CFTR dysfunction. Hence, it is not surprising that most subjects with PSC had negative sweat tests. However, the mean sweat chloride value in subjects with PSC was greater than in disease control subjects. Although the sample size was too small for a valid correlation, exhaustive CFTR genotyping was not a reliable The Journal of Pediatrics • September 2007

predictor of phenotype in this study. There were several variants detected that are not true CF-causing mutations. It is important to point out that what may be regarded as a variant, that is, not proven to be a disease-causing mutation leading to the classic form of CF, may be a causative mutation leading to the expression of disease in other milder CF related disorders, or “CFTR-opathies,” such as idiopathic pancreatitis.10 The polymorphisms identified in our study are also important because they could affect CFTR function, in part through potentiating mutations or variants. Some of the exonic and intronic polymorphisms may actually disrupt yet unknown sequence elements playing a role in regulating transcription or splicing of the gene. It is interesting to note that CFTR variants and mutations were detected in disease control subjects as well. Because the onset of PSC in patients with IBD is variable, it is possible that disease control subjects may develop this condition in the future. None of the patients in this study met the CF Consensus criteria for the diagnosis of classic CF. A conceptual model of cholangiopathies was recently described.16 PSC probably represents a common phenotypic end point arising from the interaction of multiple environmental and genetic factors. Its strong association with IBD (90% in this cohort) suggests that chronic portal bacteremia may provide an initial insult to cholangiocytes. Bacterial overgrowth in the CF intestine, as recently shown in CF mice,17 may also play a causative role as an instigating factor. The fact that only 4% of patients with IBD will have PSC may be related to CFTR dysfunction resulting in an excessive host inflammatory response attributable to increased levels of proinflammatory cytokines and neutrophils.2,3 Further support for the concept that CFTR dysfunction in the setting of colitis predisposes to bile duct injury comes from experiments with exon 10 cftr⫺/⫺ mice, in which induction of colitis with dextran sodium sulfate results in a mononuclear cell infiltrate in the portal tracts in conjunction with bile duct proliferation.18 This was not observed in wild-type control mice. Peroxisome proliferator activated receptor (PPAR)-␣ abnormalities may play a role in contributing to this excessive inflammatory response, as shown in this CF mouse model of bile duct injury.19 Thus, abnormal qualitative or quantitative innate immune responses to an insult in the liver may predispose to the development of PSC. We conclude that NTPD has identified a subgroup of subjects with PSC diagnosed in childhood who have CFTRmediated ion transport dysfunction as compared with disease control subjects. PSC is a genetically complex disease with no simple mendelian pattern of inheritance. CFTR may play a role as a modifier in the development of PSC, but this may

not be the only mechanism and does not exclude other independent or coexistent environmental or genetic factors or disorders of immune regulation. Genotyping did not identify an increase in disease proven CFTR mutations, which is not unexpected, given the fact that these individuals do not have the classic form of CF.

REFERENCES 1. Cystic fibrosis mutation database (2004). Cystic fibrosis genetic analysis consortium. Available from: http://www.genet.sickkids.on.ca/cftr/. 2. Muhlebach MS, Stewart PW, Leigh MW, Noah TL. Quantitation of inflammatory responses to bacteria in young cystic fibrosis and control patients. Am J Respir Crit Care Med 1999;160:186-91. 3. Khan TZ, Wagener JS, Bost T, Martinez J, Accurso FJ, Riches DW. Early pulmonary inflammation in infants with cystic fibrosis. Am J Respir Crit Care Med 1995;151:1075-82. 4. Broome U, Olsson R, Loof L, Bodemar G, Hultcrantz R, Danielsson A, et al. Natural history and prognostic factors in 305 Swedish patients with primary sclerosing cholangitis. Gut 1996;38:610-5. 5. Sheth S, Shea JC, Bishop MD, Chopra S, Regan MR, Malmberg E, et al. Increased prevalence of CFTR mutations and variants and decreased chloride secretion in primary sclerosing cholangitis. Hum Genet 2003;113:286-92. 6. Gallegos-Orozco JF, E Yurk C, Wang N, Rakela J, Charlton MR, Cutting GR, et al. Lack of association of common cystic fibrosis transmembrane conductance regulator gene mutations with primary sclerosing cholangitis. Am J Gastroenterol 2005;100:874-8. 7. Gibson LE, Cooke RE. A test for concentration of electrolytes in sweat in cystic fibrosis of the pancreas utilizing pilocarpine by iontophoresis. Pediatrics 1959;23:545-9. 8. Rosenstein BJ, Cutting GR. The diagnosis of cystic fibrosis: a consensus statement: Cystic Fibrosis Foundation Consensus Panel. J Pediatr 1998;132:589-95. 9. Knowles MR, Carson JL, Collier AM, Gatzy JT, Boucher RC. Measurements of nasal transepithelial electric potential differences in normal human subjects in vivo. Am Rev Respir Dis 1981;124:484-90. 10. Bishop M, Freedman SD, Zielenski J, Ahmed N, Dupuis A, Martin S, et al. The cystic fibrosis transmembrane conductance regulator gene and ion channel function in patients with idiopathic pancreatitis. Hum Genet 2005;118:372-81. 11. Zielenski J, Aznarez I, Onay T, Tzountzouris J, Markiewicz D, Tsui L. CFTR mutation detection by multiplex heteroduplex (mHET) analysis on MDE gel. In: Methods Mol Med. Vol 70, 2002; 3-19. 12. Chu CS, Trapnell BC, Curristin S, Cutting GR, Crystal RG. Genetic basis of variable exon 9 skipping in cystic fibrosis transmembrane conductance regulator mRNA. Nat Genet 1993;3:151-6. 13. Cuppens H, Lin W, Jaspers M, Costes B, Teng H. Polyvariant mutant cystic fibrosis transmembrane conductance regulator genes. J Clin Invest 1998;101:487-96. 14. Chillon M, Casals T, Mercier B, Bassas L, Lissens W, Silber S, et al. Mutations in the cystic fibrosis gene in patients with congenital absence of the vas deferens. N Engl J Med 1995;332:1475-80. 15. Wang X, Moylan B, Leopold DA, Kim J, Rubenstein RC, Togias A, et al. Mutation in the gene responsible for cystic fibrosis and predisposition to chronic rhinosinusitis in the general population. JAMA 2000;284:1814-9. 16. Lazaridis KN, Strazzabosco M, LaRusso NF. The cholangiopathies: disorders of biliary epithelia. Gastroenterology 2004;127:1565-77. 17. Norkina O, Burnett TG, De Lisle RC. Bacterial overgrowth in the cystic fibrosis transmembrane conductance regulator null mouse small intestine. Infect Immunol 2004;72:6040-9. 18. Blanco PG, Zaman MM, Junaidi O, Sheth S, Yantiss RK, Nasser IA, et al. Induction of colitis in cftr⫺/⫺ mice results in bile duct injury. Am J Physiol Liver Physiol 2004;287:G491-6. 19. Pall H, Zaman M, Andersson C, Freedman S. Decreased peroxisome proliferator activated receptor alpha is associated with bile duct injury in cystic fibrosis transmembrane conductance regulator⫺/⫺ mice. J Pediatr Gastroenterol Nutr 2006;42:275-81.

Primary Sclerosing Cholangitis in Childhood is Associated with Abnormalities in Cystic Fibrosis–Mediated Chloride Channel Function 259

Clinical Trial of Safety and Efficacy of IHN-A21 for the Prevention of Nosocomial Staphylococcal Bloodstream Infection in Premature Infants MITCHELL DEJONGE, MD, DAVID BURCHFIELD, MD, BARRY BLOOM, MD, MARIA DUENAS, MD, WHIT WALKER, MD, MARK POLAK, MD, ELIZABETH JUNG, MD, DIETRA MILLARD, MD, ROBERT SCHELONKA, MD, FABIEN EYAL, MD, AMY MORRIS, MBA, BARRY KAPIK, MS, DESTREY ROBERSON, RN, KAREN KESLER, PHD, JOE PATTI, PHD, AND SETH HETHERINGTON, MD

Objective To determine if INH-A21, an intravenous immune globulin (IGIV) derived from donors with high titers of antibody to surface adhesins of Staphylococcus epidermidis and S aureus prevents late-onset sepsis (LOS) in very low birth weight (VLBW) infants. Study design In this double-blind, placebo-controlled study, infants with birth weights 500 to 1250 g were randomized to receive up to four doses of INH-A21 (Veronate®) or placebo. The primary objective was to determine the safety and efficacy of INH-A21 versus placebo for prevention of S aureus LOS in VLBW infants. Results A total of 1983 infants from 95 neonatal intensive care units were randomized, and received at least one dose of study drug. S aureus LOS developed in 50 of 989 (5%) and 60 of 994 (6%) infants who received placebo or INH-A21, respectively (P ⴝ .34). No differences were found in the frequencies of LOS caused by coagulase-negative staphylococci (CoNS), Candida spp, or overall mortality. No adverse events were statistically significantly associated with INH-A21 infusions compared with placebo. Conclusion INH-A21 failed to reduce the incidence of staphylococcal LOS or candiSee editorial, p 232 demia in premature infants. (J Pediatr 2007;151:260-5) lthough advances in medical care provided by neonatal intensive care units have dramatically improved survival among premature infants,1,2 one of the costs has been an increased frequency of complications, especially nosocomial (hospital acquired) infections or late-onset sepsis (LOS). The overall rate of LOS among very low birth weight (VLBW) infants ⱕ1500 g birth weight ranges from 16% to 25%, with rates of 40% among the smallest infants (500-600 g).3-5 Infants who develop LOS have increased mortality, longer hospital stays, more frequent complications of prematurity, and are more likely to have adverse neurodevelopmental outcomes at follow-up compared with uninfected infants.3,6 Infants born before 32 weeks gestation are relatively deficient in IgG. Low level of IgG at birth is an identified risk factor for LOS in LBW infants.3 However, prior studies administering immune globulin for intravenous administration to premature infants have failed to show a clinically significant reduction in overall LOS rates.7-9 Intravenous immune globulin (IGIV) targeted against specific pathogens in neonates, however, has not been tested. The most common pathogens of LOS in premature infants include coagulasenegative staphylococcus (CoNS), Staphylococcus aureus, Enterococcus spp, and Candida spp.5,10 S aureus has been considered an infrequent pathogen in the past, but in a recent review it accounted for approximately 7% of LOS.11 A similar result was noted in the prior Phase II clinical trial of INH-A21.12 INH-A21 is an experimental donor-selected anti-staphylococcal human IGIV. It contains elevated levels of antibodies against the staphylococcal fibrinogen-binding pro-

A

BSI CoNS IGIV LIS

260

Bloodstream infection Coagulase-negative staphylococci Intravenous immune globulin Late-onset sepsis

NEC SAEs VLBW

Necrotizing enterocolitis Serious adverse events Very low birth weight

From the DeVos Children’s Hospital (M. DeJonge), Grand Rapids, Michigan; Shand’s Children’s Hospital, University of Florida (D.B.), Gainesville, Florida; Wesley Medical Center (B.B.), Wichita, Kansas; St. John Hospital and Medical Center (M. Duenas), Detroit, Michigan; Greenville Memorial Hospital (W.W.), Greenville, South Carolina; West Virginia University Hospital (M.P.), Morgantown, West Virginia; St. John’s Mercy Medical Center (E.J.), St. Louis, Missouri; Evanston Hospital (D.M.), Evanston, Illinois; University of Alabama at Birmingham Hospital (R.S.), Birmingham, Alabama; University of South Alabama Children’s & Women’s Hospital (F.E.), Mobile, Alabama; Inhibitex, Inc. (A.M., B.K., D.R., J.P., S.H.), Alpharetta, Georgia; and Rho, Inc. (K.K.), Chapel Hill, North Carolina. This study is registered at www.clinicaltrials. gov; registration number ⫽ NCT00113191. The study was funded in its entirety by Inhibitex, Inc. A. Morris, B. Kapik, D. Roberson, J. Patti, and S. Hetherington were all employees of Inhibitex during the course of the study. Submitted for publication Nov 20, 2006; last revision received Mar 30, 2007; accepted Apr 25, 2007. Reprint requests: Seth Hetherington, MD, 4222 Emperor Blvd., Suite 335, Durham, NC 27703. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.04.060

teins, clumping Factor A and Ser-Asp dipeptide repeat G, that are found in ⬎95% of strains of S aureus and most strains of S epidermidis, respectively.13-15 These antigens play an important role in the adherence of bacteria, which is the initiating step in establishing infection. Donors for INH-A21 represent approximately 2% of the normal blood donor population with the highest levels of antibodies to either antigen. Compared with seven lots of commercially available IGIV from five manufacturers, INH-A21 contains 2- to 5-fold higher levels of anti-clumping Factor A and 1.75- to 6-fold higher titers of anti-Ser-Asp dipeptide repeat G (data on file). We report results from a randomized, double-blind, multicenter, placebo-controlled study designed to determine the safety and efficacy of INH-A21 versus placebo for prevention of LOS as a result of S aureus in VLBW infants.

METHODS Patients and Eligibility This clinical trial was conducted at 95 study centers in the United States and Canada between May 26, 2004 and January 21, 2006. Premature infants between postnatal days 3 to 5 (beginning at hour 49 and through hour 120 after delivery) were eligible for enrollment if they met the following criteria: birth weight ⱖ500 and ⱕ1250 g and expected to survive at least 4 weeks and to require intravenous access for 10 to 14 days. Infants were excluded if there was evidence of active sepsis (defined by one of the following: culture-proven early-onset sepsis and not clinically stable, or clinical signs of sepsis and blood cultures pending), severe congenital anomaly, congenital immunodeficiency, evidence of significant fluid overload or volume depletion, or serum creatinine ⬎1.6 mg/dL. Infants were excluded who had received or were likely to receive another IGIV product or immune globulin before first infusion of study drug or were receiving antibiotics for prevention of catheter-related or nosocomial infections. Study Design, Study Groups, and Randomization Following informed consent, infants meeting entry criteria were randomized (1:1) to receive 1.5 mL/kg of study drug INH-A21 (750 mg/kg) or placebo (0.45% NaCl). Infants were randomized using a standard block randomization, stratified within site and birth weight group (500-900 g and 901-1250 g). Infants received up to four infusions of study drug on study days 1, 3, 8, and 15, provided intravenous access was present for general medical care. Infusions were administered by a rate-escalation protocol as described previously.12 The dose selected and infusion schedule were based on results from a previous dose-escalation study and population pharmacokinetic modeling of anti-staphylococcal antibodies.12,16 Infants were followed for up to 70 days at the enrolling institution, or up to the time of discharge home, permanent transfer to another hospital, or death. The protocol, study design, and parental consent forms were approved by the Institutional Review Board at each participating institution. An independent Data and Safety

Monitoring Board reviewed available safety data and infection rates at predefined intervals. The study was conducted according to the guidelines of Good Clinical Practice as established by the International Conference on Harmonization (http:// www.fda.gov/cder/guidance/959fnl.pdf).

Outcome Measures The primary outcome was the proportion of infants with LOS caused by S aureus. Sepsis for known bacterial or fungal pathogens was defined as the presence of clinical signs and one positive blood culture or culture from an otherwise sterile site (cerebrospinal fluid; peritoneal, pleural, or joint fluid; but not urine). For CoNS, the diagnosis of sepsis was considered “definite” when clinical signs of sepsis were present and accompanied by two documented cultures for CoNS obtained within a 24-hour period. Cultures could be two separate blood samples or one blood culture plus a culture from an otherwise sterile site (excluding urine, superficial soft tissue, or upper respiratory tract). The diagnosis was considered “probable” if clinical signs were present with one positive blood culture and antibiotics were administered on four or more consecutive days. Clinical signs of infection considered indicative of infection included: hyperglycemia (⬎140 mg/dL), increased apnea, leukocytosis (white blood cell count ⬎20,000 cells/ mm3), neutropenia (absolute neutrophil count ⬍1,500/mm3), temperature instability, hypotension, increased respiratory support, lethargy, unexplained metabolic acidosis, increased band-to-mature neutrophil ratio (⬎0.2), pulmonary infiltrates on chest roentgenogram, inflammation at a vascular line site, or gastrointestinal symptoms.3 Secondary outcomes of the trial included the proportions of infants with Candida bloodstream infection (BSI), all CoNS sepsis (definite and probable), all staphylococcal sepsis, and mortality. Vital signs were monitored throughout the infusion, and concomitant medications and adverse events were monitored during the study. Specific diagnoses related to prematurity (morbidities associated with prematurity) were recorded and included anemia, hyperbilirubinemia, patent ductus arteriosus, apnea, bradycardia, periventricular or intraventricular hemorrhage, retinopathy of prematurity, air leak syndrome, feeding intolerance, gastroesophageal reflux, necrotizing enterocolitis (NEC), bronchopulmonary dysplasia, respiratory distress of prematurity, cystic periventricular leukomalacia, progressive hydrocephalus, and focal gastrointestinal perforation (not associated with NEC). Adverse events were defined as serious if they resulted in death, were immediately lifethreatening, or required intervention by procedure or surgery. Statistical Analysis The primary analysis compared the proportion of infants with S aureus sepsis in each treatment group. Only the first episode of sepsis for each infant was analyzed. The null hypothesis that the proportions in the two treatment groups were the same was tested using a Cochran-Mantel-Haenszel ␹2 statistic controlling for birth weight group, with a two-

Clinical Trial of Safety and Efficacy of IHN-A21 for the Prevention of Nosocomial Staphylococcal Bloodstream Infection in Premature Infants

261

Table I. Summary of selected characteristics of the treatment groups Characteristic

Figure 1. Disposition of patients randomized and enrolled into the clinical trial.

sided ␣ of 0.05. The study had 90% power to detect a 50% reduction in S aureus LOS, assuming a 6% incidence rate in the placebo group. The analysis was performed using the intent-to-treat population defined as all infants who were randomized and received at least one infusion of study drug. The Cochran-Mantel-Haenszel ␹2 test statistic controlling for birth weight group was used to compare secondary outcomes, adverse events, and morbidities associated with prematurity by treatment group. Analyses within each birth weight stratum were conducted using a Pearson’s ␹2 test. Two-sided significance tests were used throughout, with a P value ⱕ .05 being declared statistically significant for the primary hypothesis. Other P values are presented as descriptive statistics. No attempts were made to adjust for multiple testing. The analysis was conducted using the SAS software version 8.02 (SAS Institute Inc., Cary, NC).

RESULTS Subjects From 95 neonatal intensive care units in the US and Canada, 2017 infants were enrolled. Thirty-four infants did not receive any infusion of study drug (INH-A21 or placebo), usually because they became clinically unstable, and were not included in the analysis (Figure 1). Among the 1983 infants enrolled who received at least one infusion of study drug, 989 and 994 received placebo and INH-A21, respectively. The mean and median ages at the time of the first infusion were 4.2 and 4.0 days, for placebo and INH-A21, respectively. There were no significant differences by birth weight, gestational age, sex, or maternal race between the two groups 262

DeJonge et al

Placebo INH-A21 P (N ⴝ 989) (N ⴝ 994) value*

Estimated gestational age 26.8 (2.2) 26.8 (2.2) (wk; mean ⫾ SD) Male sex, n (%) 504 (51%) 487 (49%) Birth weight (g; mean ⫾ SD) 896 (197) 891 (204) Maternal use of antibiotics,† 572 (58%) 583 (59%) n (%) Cesarean delivery, n (%) 715 (72%) 700 (70%) Apgar score at 5 min mean (⫾ SD) 7.3 (1.7) 7.3 (1.7) median 8 8 Maternal race/ethnicity, n (%) Caucasian, non-Hispanic 535 (54%) 524 (53%) Black, non-Hispanic 279 (28%) 285 (29%) Hispanic 115 (12%) 134 (14%) Asian 22 (2%) 24 (2%) Hawaiian or Pacific Islander 8 (⬍1%) 7 (⬍1%) Native North American/ 13 (1%) 8 (⬍1%) Native Alaskan Other 16 (2%) 12 (1%) Missing 1 (⬍1%) 0 (0%)

.76 .38 .36 .75 .36 .72

.74

*P values corresponds to a test of no difference between treatments using either a Cochran-Mantel-Haenszel test (general association) stratified by birth weight group for categorical variables, or an analysis of variance model with treatment and birth weight group as fixed effects for continuous variables. †Use of antibiotics within 24 hours before delivery by the mother.

(Table I). Overall, infants received a mean of 3.4 complete study-drug infusions, with a mean of 3.6 infusions per infant among those with birth weights 500 to 900 g and 3.3 infusions for those infants with birth weights 901 to 1250 g. There was no difference in mean number of infusions between placebo and INH-A21 groups.

Primary Outcomes Overall, 110 (6%) infants developed at least one episode of S aureus LOS: 50 (5%) among the placebo and 60 (6%) among the INH-A21 recipients (P ⫽ .34; Table II). Oxacillin resistance was present in 27 (23%) of 116 tested S aureus isolates; two strains had mixed sensitivities, and two isolates were not tested. The time of onset of S aureus LOS was similar between the two treatment groups (Figure 2; available at www.jpeds.com). The median (and ranges) time to S aureus LOS was 18 days (range 1-70 days) and 19 days (range 2-70 days) for placebo and INH-A21 recipients, respectively. No difference was seen between treatment groups in frequency of S aureus infections by birth weight stratum: 33 infections occurred in the placebo group and 39 infections in the INHA21 group for infants with a birth weight of 500 to 900 g (P ⫽ .48). There was no difference in infection rate analyzed by the number of infusions received for the overall population or by treatment group (data not shown). The Journal of Pediatrics • September 2007

Table II. Number of infants with primary and secondary outcomes

Outcome Bloodstream Infection S aureus† Any CoNS Definite CoNS‡ Probable CoNS§ All staphylococcal infection S aureus and definite CoNS S aureus and any CoNS Candida spp† Mortality

Table V. Serious adverse events* other than death occurring in >0.5% of study infants

Placebo INH-A21 (N ⴝ 989) (N ⴝ 994) n (%) n (%) P value* 50 (5%) 227 (23%) 92 (9%) 154 (16%)

60 (6%) 247 (25%) 107 (11%) 148 (15%)

.34 .32 .28 .67

139 (14%)

160 (16%)

.20

264 (27%) 30 (3%) 73 (7%)

285 (29%) 33 (3%) 57 (6%)

.32 .72 .13

*P value corresponds to a test of no difference between treatments using a CochranMantel-Haenszel test (general association) stratified by weight group. †All infections other than CoNS are defined as clinical signs of infection plus at least one positive culture from blood or an otherwise sterile site. ‡A definite CoNS infection is defined as clinical signs of infection plus at least two positive blood cultures or at least one positive blood and one otherwise sterile site culture. The two positive cultures must be drawn within 24 hours of one another. §A probable CoNS infection is defined as documented evaluation for infection, plus one positive blood culture, plus clinical signs of infection, plus antibiotics for ⱖ4 days.

Secondary Outcomes No difference was found between the two groups for any secondary outcome. Among the placebo recipients, 92 infants developed definitive CoNS LOS compared with 107 infants receiving INH-A21 (P ⫽ .28). Sixty-three infants developed Candida BSI. Antifungal prophylaxis had been administered to 65 placebo and 68 INH-A21 recipients (7% each). Among 1850 infants who did not receive antifungal prophylaxis, 30 of 924 (3%) placebo and 32 of 926 (3%) INH-A21 recipients developed Candida BSI. There were 130 deaths in the study population through study day 70, with no significant differences between the treatment groups (P ⫽ .13). The number of infants with sepsis caused by other organisms included Escherichia coli (n ⫽ 63, 3%), Klebsiella sp (n ⫽ 38, 2%), Enterobacter sp (n ⫽ 36, 2%), Pseudomonas sp (n ⫽ 35, 2%), and Serratia (n ⫽ 18, 1%), with no significant differences by treatment group (Table III; available at www. jpeds.com). Adverse Events There were no differences between treatment arms in the total numbers of adverse events, serious adverse events (SAEs), adverse events considered related to treatment, or adverse events leading to interruption of infusion or permanent discontinuation of study drug (Table IV; available at www.jpeds.com). The most common SAEs were NEC, gastrointestinal perforation (not NEC), retinopathy of prematurity, pneumothorax, sepsis, hydrocephalus, and bradycardia, with no differences between the two groups (Table V). Four SAEs among placebo recipients were considered possibly related to study drug by the investigator (two NEC, one

Event

Placebo (n ⴝ 989) n (%)

INH-A21 (n ⴝ 994) n (%)

‡P value

NEC GI Perforation (not NEC) ROP Pneumothorax Sepsis† Hydrocephalus Bradycardia

51 (5%) 18 (2%) 38 (4%) 4 (⬍1%) 18 (2%) 14 (1%) 3 (⬍1%)

40 (4%) 27 (3%) 39 (4%) 6 (⬍1%) 17 (2%) 11 (1%) 7 (⬍1%)

.24 .23 1.00 .75 .87 .55 .34

*Serious adverse events were defined, per protocol, as an event leading to death, an immediately life-threatening event, or an event requiring a surgical intervention. †Sepsis in this instance was reported as a clinical diagnosis, with or without an identified pathogen and was only recorded as a serious adverse event if it led to death. ‡P value corresponds to a test of no difference between treatments using a CochranMantel-Haenszel test (general association) stratified by weight group.

intestinal perforation, and one air embolism) compared with two such events among INH-A21 recipients (one NEC and one pulmonary hemorrhage). Frequencies of the 16 prespecified common morbidities of prematurity were not statistically significantly different between the two treatment groups (Table VI; available at www.jpeds.com).

DISCUSSION The potential to prevent infection in a high-risk population of premature infants through administration of immune globulin has been attractive for several reasons. The neonate born before 32 weeks gestation is deficient in IgG. Specific antibody is believed to be a critical component of host defense against staphylococci, which represent the most common pathogens for LOS.17-20 Multiple trials have attempted to reduce LOS by administration of IGIV.7,10,21-23 The meta-analysis by Ohlsson and Lacy9 concluded that IGIV could decrease nosocomial infections in this population, but it would not affect mortality. An approach using IGIV derived from donors with high levels of pathogen-specific antibody, however, had not been studied. The adhesion proteins, microbial surface components recognizing adhesive matrix molecules (MSCRAMMs), are surface-expressed virulence factors, and animal models have demonstrated the ability of antibody to prevent infection.24 In addition, a Phase II trial with INH-A21, although not powered for statistical significance, showed trends for reduction of S aureus LOS and Candida BSI in premature infants.12 In the current large, adequately powered, randomized clinical trial, we were unable to demonstrate clinical benefit from infusion of IGIV with elevated levels of antibody against surface antigens of S epidermidis and S aureus. The changes in the infusion schedule between the Phase II trial and the current trial do not explain the apparent difference in efficacy between the two trials. The lots of INH-A21 used in the current trial had 50% to 75% higher levels of anti-MSCRAMM antibodies than the single lot

Clinical Trial of Safety and Efficacy of IHN-A21 for the Prevention of Nosocomial Staphylococcal Bloodstream Infection in Premature Infants

263

used in the previous Phase II trial. Based on population pharmacokinetic modeling, the levels of anti-staphylococcal antibodies likely were higher at all time points compared with those observed with the previous schedule.16 Because the mean time to onset of S aureus LOS was approximately 3 weeks, we also do not believe that the time of initiation of infusions explains the lack of effect. Reduction of infections in premature infants through administration of IGIV may not be as simple as supposed previously. Risk factors for LOS are complex, involving the environment, immune status of the infant, and risk factors for hospital acquired infections; and they may not be addressed by administration of IgG. Although opsonic activity generally is accepted as being critical, there is no evidence to support a quantitative link between measured opsonic activity and risk of infection for premature infants. Levels of IgG antibodies and opsonic activity to staphylococci either at birth or during the first 2 postnatal weeks do not differ between infants who develop CoNS sepsis and those who do not.17 In term infants, antibody can be protective for certain pathogens, but this has not been shown for staphylococci or premature infants.25 In the end, the lack of efficacy for INH-A21 may rest with the selected target. Antibodies directed against capsular polysaccharide and protein targets are both opsonic, but clinical efficacy only has been achieved to date with anti-capsular polysaccharide antibodies. However, anti-capsular polysaccharide antibodies to S aureus have failed to protect patients in recent clinical trials.26 The safety profile of INH-A21 is similar to that reported in other studies of IGIV in neonates. Acute renal failure and aseptic meningitis, reported among adults but not neonates, were not observed.27-36 Neutropenia, anecdotally reported to be associated with IGIV use in neonates, was reported in 4% of infants in each treatment group.37 There was no significant difference in the frequencies of NEC or of surgical intervention for NEC between the two groups. By contrast, Baker et al found less NEC among infants treated with prophylactic IGIV compared with control infants who received placebo.10 LOS continues to be a major health risk for premature infants. Our data are consistent with studies showing the predominance of gram-positive organisms as causative agents.3,5,10,38 The percentage of infants affected by S aureus was similar to that seen in our prior Phase II study and is higher than that reported by older studies, but it is consistent with that recently reported by Healy et al.11,12 We may be observing the inflection point of the return of S aureus as a major pathogen among neonates. Clearly, the need for new strategies to reduce staphylococcal LOS persists. The authors specifically acknowledge Amy Burdan for medical writing assistance in preparing the manuscript.

REFERENCES 1. Martin JA, Hamilton BE, Ventura SJ, Menacker F, Park MM. Births: final data for 2000. Natl Vital Stat Rep 2002;50:1-101. 2. Fanaroff AA, Hack M, Walsh MC. The NICHD neonatal research network: changes in practice and outcomes during the first 15 years. Semin Perinatol 2003;27:281-7.

264

DeJonge et al

3. Fanaroff AA, Korones SB, Wright LL, Verter J, Poland RL, Bauer CR, et al. Incidence, presenting features, risk factors and significance of late onset septicemia in very low birth weight infants. The National Institute of Child Health and Human Development Neonatal Research Network. Pediatr Infect Dis J 1998;17:593-8. 4. Stoll BJ, Gordon T, Korones SB, Shankaran S, Tyson JE, Bauer CR, et al. Late-onset sepsis in very low birth weight neonates: a report from the National Institute of Child Health and Human Development Neonatal Research Network. J Pediatr 1996;129:63-71. 5. Stoll BJ, Hansen N, Fanaroff AA, Wright LL, Carlo WA, Ehrenkranz RA, et al. Late-onset sepsis in very low birth weight neonates: the experience of the NICHD Neonatal Research Network. Pediatrics 2002;110:285-91. 6. Stoll BJ, Hansen NI, Adams-Chapman I, Fanaroff AA, Hintz SR, Vohr B, et al. Neurodevelopmental and growth impairment among extremely low-birth-weight infants with neonatal infection. JAMA 2004;292:2357-65. 7. Fanaroff AA, Korones SB, Wright LL, Wright EC, Poland RL, Bauer CB, et al. A controlled trial of intravenous immune globulin to reduce nosocomial infections in very-low-birth-weight infants. National Institute of Child Health and Human Development Neonatal Research Network. N Engl J Med 1994;330:1107-13. 8. Hill HR. Additional confirmation of the lack of effect of intravenous immunoglobulin in the prevention of neonatal infection. J Pediatr 2000;137:595-7. 9. Ohlsson A, Lacy JB. Intravenous immunoglobulin for preventing infection in preterm and/or low-birth-weight infants. Cochrane Database Syst Rev 2004: CD000361. 10. Baker CJ, Melish ME, Hall RT, Casto DT, Vasan U, Givner LB. Intravenous immune globulin for the prevention of nosocomial infection in low-birth-weight neonates. The Multicenter Group for the Study of Immune Globulin in Neonates. N Engl J Med 1992;327:213-9. 11. Healy CM, Palazzi DL, Edwards MS, Campbell JR, Baker CJ. Features of invasive staphylococcal disease in neonates. Pediatrics 2004;114:953-61. 12. Bloom B, Schelonka R, Kueser T, Walker W, Jung E, Kaufman D, et al. Multicenter study to assess safety and efficacy of INH-A21, a donor-selected human staphylococcal immunoglobulin, for prevention of nosocomial infections in very low birth weight infants. Pediatr Infect Dis J 2005;24:858-66. 13. Vernachio J, Bayer AS, Le T, Chai YL, Prater B, Schneider A, et al. Anticlumping factor A immunoglobulin reduces the duration of methicillin-resistant Staphylococcus aureus bacteremia in an experimental model of infective endocarditis. Antimicrob Agents Chemother 2003;47:3400-6. 14. Patti JM, Allen BL, McGavin MJ, Hook M. MSCRAMM-mediated adherence of microorganisms to host tissues. Annu Rev Microbiol 1994;48:585-617. 15. Höök M, Patti JM, McGavin M, Gurusiddappa S, Lindgren PE, Jönsson K, et al. MSCRAMMs-microbial recognition systems for extracellular matrix molecules. In: Wadström T, Holder, I, and Kronvall, G, eds Molecular Pathogenesis of Surgical Infections. Stuttgart: Gustav Fischer Verlag; 1994137-44. 16. Capparelli EV, Bloom BT, Kueser TJ, Oelberg DG, Bifano EM, White RD, et al. Multicenter study to determine antibody concentrations and assess the safety of administration of INH-A21, a donor-selected human staphylococcal immune globulin, in low-birth-weight infants. Antimicrob Agents Chemother 2005;49:4121-7. 17. Krediet TG, Beurskens FJ, van Dijk H, Gerards LJ, Fleer A. Antibody responses and opsonic activity in sera of preterm neonates with coagulase-negative staphylococcal septicemia and the effect of the administration of fresh frozen plasma. Pediatr Res 1998;43:645-51. 18. Fischer GW, Cieslak TJ, Wilson SR, Weisman LE, Hemming VG. Opsonic antibodies to Staphylococcus epidermidis: in vitro and in vivo studies using human intravenous immune globulin. J Infect Dis 1994;169:324-9. 19. Etzioni A, Obedeanu N, Blazer S, Benderly A, Merzbach D. Effect of an intravenous gammaglobulin preparation on the opsonophagocytic activity of preterm serum against coagulase-negative staphylococci. Acta Paediatr Scand 1990;79:156-61. 20. Clark LA, Easmon CS. Opsonic requirements of Staphylococcus epidermidis. J Med Microbiol 1986;22:1-7. 21. Christensen RD, Hardman T, Thornton J, Hill HR. A randomized, double-blind, placebo-controlled investigation of the safety of intravenous immune globulin administration to preterm neonates. J Perinatol 1989;9:126-30. 22. Fischer GW. Use of intravenous immune globulin in newborn infants. Clin Exp Immunol 1994;97(suppl 1)73-7. 23. Jenson HB, Pollock BH. Meta-analyses of the effectiveness of intravenous immune globulin for prevention and treatment of neonatal sepsis. Pediatrics 1997;99:E2. http://www.pediatrics.org/cgi/content/full/99/z/e2. 24. Vernachio JH, Bayer AS, Ames B, Bryant D, Prater BD, Syribeys PJ, et al. Human immunoglobulin G recognizing fibrinogen-binding surface proteins is protective against both Staphylococcus aureus and Staphylococcus epidermidis infections in vivo. Antimicrob Agents Chemother 2006;50:511-8. 25. Lin FY, Weisman LE, Azimi PH, Philips JB 3rd, Clark P, Regan J, et al. Level of maternal IgG anti-group B streptococcus type III antibody correlated with protection

The Journal of Pediatrics • September 2007

of neonates against early-onset disease caused by this pathogen. J Infect Dis 2004;190:928-34. 26. Shinefield H, Black S, Fattom A, Horwith G, Rasgon S, Ordonez J, et al. Use of a Staphylococcus aureus conjugate vaccine in patients receiving hemodialysis. N Engl J Med 2002;346:491-6. 27. Ahsan N, Wiegand LA, Abendroth CS, Manning EC. Acute renal failure following immunoglobulin therapy. Am J Nephrol 1996;16:532-6. 28. Brannagan TH 3rd, Nagle KJ, Lange DJ, Rowland LP. Complications of intravenous immune globulin treatment in neurologic disease. Neurology 1996; 47:674-7. 29. Cantu TG, Hoehn-Saric EW, Burgess KM, Racusen L, Scheel PJ. Acute renal failure associated with immunoglobulin therapy. Am J Kidney Dis 1995;25:228-34. 30. Michail S, Nakopoulou L, Stavrianopoulos I, Stamatiadis D, Avdikou K, Vaiopoulos G, et al. Acute renal failure associated with immunoglobulin administration. Nephrol Dial Transplant 1997;12:1497-9. 31. Sati HI, Ahya R, Watson HG. Incidence and associations of acute renal failure complicating high-dose intravenous immunoglobulin therapy. Br J Haematol 2001;113:556-7.

32. Cayco AV, Perazella MA, Hayslett JP. Renal insufficiency after intravenous immune globulin therapy: a report of two cases and an analysis of the literature. J Am Soc Nephrol 1997;8:1788-94. 33. Hansen-Schmidt S, Silomon J, Keller F. Osmotic nephrosis due to high-dose immunoglobulin therapy containing sucrose (but not with glycine) in a patient with immunoglobulin A nephritis. Am J Kidney Dis 1996;28:451-3. 34. Casteels-Van Daele M, Wijndaele L, Hanninck K, Gillis P. Intravenous immune globulin and acute aseptic meningitis. N Engl J Med 1990;323:614-5. 35. Scribner CL, Kapit RM, Phillips ET, Rickles NM. Aseptic meningitis and intravenous immunoglobulin therapy. Ann Intern Med 1994;121:305-6. 36. Sekul EA, Cupler EJ, Dalakas MC. Aseptic meningitis associated with high-dose intravenous immunoglobulin therapy: frequency and risk factors. Ann Intern Med 1994;121:259-62. 37. Lassiter HA, Bibb KW, Bertolone SJ, Patel CC, Stroncek DF. Neonatal immune neutropenia following the administration of intravenous immune globulin. Am J Pediatr Hematol Oncol 1993;15:120-3. 38. Gladstone IM, Ehrenkranz RA, Edberg SC, Baltimore RS. A ten-year review of neonatal sepsis and comparison with the previous fifty-year experience. Pediatr Infect Dis J 1990;9:819-25.

Clinical Trial of Safety and Efficacy of IHN-A21 for the Prevention of Nosocomial Staphylococcal Bloodstream Infection in Premature Infants

265

Table IV. Summary of adverse events

Outcome variable

Figure 2. Time of onset of S aureus infections. Number of infections occurring among patients randomized to placebo (hatched bars) or INHA21 treatment (open bars) during each time interval.

Table III. Number of infants with nonstaphylococcal and non-candidal bloodstream infections. A count of patients

Organism Enterobacter sp Enterococcus sp Eschericha coli Group B streptococcus Klebsiella sp Pseudomonas sp Serratia sp Viridans streptococcus Other Alpha hemolytic streptococcus Anaerobes Gram-negative coccus Gram-positive rods Other gram-negative bacillus Other gram-positive coccus Viral pathogen Not specified

265.e1

DeJonge et al

Placebo (n ⴝ 989) n (%)

INH-A21 (n ⴝ 994) n (%)

Total (n ⴝ 1983) n (%)

15 (2) 29 (3) 31 (3) 9 (⬍1) 23(2) 23 (2) 9 (⬍1) 7 (⬍1) 20 (2) 1 (⬍1)

21 (2) 25 (3) 32 (3) 2 (⬍1) 15 (2) 12 (1) 9 (⬍1) 4 (⬍1) 21 (2) 3 (⬍1)

36 (2) 54 (3) 63 (3) 11 (⬍1) 38 (2) 35 (2) 18 (⬍1) 11 (⬍1) 41 (2) 4 (⬍1)

0 (0) 0 (0) 4 (⬍1) 8 (⬍1)

2 (⬍1) 1 (⬍1) 4 (⬍1) 5 (⬍1)

2 (⬍1) 1 (⬍1) 8 (⬍1) 13 (⬍1)

2 (⬍1)

1 (⬍1)

3 (⬍1)

1 (⬍1) 4 (⬍1)

0 (0) 5 (⬍1)

1 (⬍1) 9 (⬍1)

Any Adverse Event Within 48 h of infusion Considered drug-related Leading to interruption or discontinuation of infusion Leading to permanent discontinuation of infusion Any SAE Within 48 h of infusion Considered drug-related

Placebo (n ⴝ 989) n (%)

INH-A21 (n ⴝ 994) n (%)

P value*

883 (89%) 629 (64%) 60 (6%) 47 (5%)

890 (90%) 662 (67%) 68 (7%) 59 (6%)

.85 .15 .48 .24

65 (7%)

54 (5%)

.28

182 (18%) 43 (4%) 3 (⬍1%)

164 (16%) 48 (5%) 2 (⬍1%)

.24 .61 .65

*P value corresponds to a test of no difference between treatments using a CochranMantel-Haenszel test (general association) stratified by weight group.

Table VI. Common morbidities of prematurity identified in study patients* Placebo INH-A21 (n ⴝ 989) (n ⴝ 994) n (%) n (%) P value†

Condition Air leak syndrome Anemia Apnea Bradycardia Chronic lung disease‡ Cystic periventricular leukomalacia Feeding intolerance Gastroesophageal reflux Gastrointestinal perforation Hyperbilirubinemia Patent ductus rrteriosus Progressive hydrocephalus Respiratory distress Necrotizing enterocolitis (Stage ⱖ2) Intraventricular hemorrhage (Grade I-IV) Retinopathy of prematurity (Stage 1-5)

119 (12%) 857 (87%) 908 (92%) 960 (97%) 740 (75%) 41 (4%)

131 (13%) 854 (86% 909 (91%) 954 (96%) 746 (75%) 39 (4%)

.43 .63 .77 .18 .90 .80

823 (83%) 343 (35%) 34 (3%) 861 (87%) 446 (45%) 48 (5%) 947 (96%) 84 (8%)

828 (83%) 320 (32%) 35 (4%) 869 (87%) 456 (46%) 47 (5%) 952 (96%) 71 (7%)

.96 .24 .92 .81 .73 .89 .98 .10

347 (35%)

317 (32%)

.13

406 (41%)

424 (41%)

.45

*Patients with more that one morbidity for a particular condition are counted only once for that condition. †P value corresponds to a test of no difference between treatments using a CochranMantel-Haenszel test (general association) stratified by weight group. ‡Defined as use of oxygen on postnatal day 28.

The Journal of Pediatrics • September 2007

Nebulized Hypertonic Saline in the Treatment of Viral Bronchiolitis in Infants BRIAN A. KUZIK, MD, MSC, FRCP(C), SAMIM A. AL QADHI, MD, MBCHB, STEVEN KENT, BSC(MED), MD, FRCP(C), MICHAEL P. FLAVIN, MB, MRCP(UK), FRCP(C), WILMA HOPMAN, MA, SIMON HOTTE, MD, AND SARAH GANDER, MD

Objective To investigate the use of nebulized 3% hypertonic saline (HS) for treating viral bronchiolitis in moderately ill hospitalized infants by a prospective, randomized, double-blinded, controlled, multicenter trial. Study design A total of 96 infants (mean age, 4.7 months; range, 0.3 to 18 months) admitted to the hospital for treatment of viral bronchiolitis were recruited from 3 regional pediatric centers over 3 bronchiolitis seasons (December 2003 to May 2006). Patients were randomized to receive, in a double-blind fashion, repeated doses of nebulized 3% HS (treatment group) or 0.9% normal saline (NS; control group), in addition to routine therapy ordered by the attending physician. The principal outcome measure was hospital length of stay (LOS). Results On an intention-to-treat basis, the infants in the HS group had a clinically relevant 26% reduction in LOS to 2.6 ⴞ 1.9 days, compared with 3.5 ⴞ 2.9 days in the NS group (P ⴝ .05). The treatment was well tolerated, with no adverse effects attributable to the use of HS. Conclusions The use of nebulized 3% HS is a safe, inexpensive, and effective treatment for infants hospitalized with moderately severe viral bronchiolitis. (J Pediatr 2007;151:266-70) espiratory syncytial virus (RSV) accounts for the majority of viral bronchiolitis cases, although other viruses, including human metapneumovirus, adenovirus, parainfluenza, rhinovirus, and influenza, also play important roles.1-3 Given that virtually all children become infected with RSV by age 2 years and that at least 1% of these children will develop bronchiolitis See editorial, p 235 sufficient to require hospitalization,4 the burden of this disease is high, accounting for up to 17% of all infant hospitalizations,5 at an annual cost of more than $500 million in the From the Department of Paediatrics, United States alone.6 Sheikh Khalifa Medical City, Abu Dhabi, Despite the high prevalence and morbidity of bronchiolitis, therapy remains conUnited Arab Emirates (B.K., S.Q.); Department of Paediatrics, University of British troversial and without widely accepted therapeutic guidelines other than supportive Columbia, Victoria General Hospital, Victo7,8 care. Bronchiolitis is characterized by airway plugging with sloughed epithelium, ria, British Columbia, Canada (S.K.); Department of Paediatrics, Queen’s University, mucus, and edema rather than bronchospasm.9,10 Nevertheless, the use of nebulized Kingston General Hospital, Kingston, Onbronchodilators continues to be common,11,12 despite extensive evidence supported by 3 tario, Canada (M.F., S.H., S.G.); Clinical Re13-15 search Unit, Kingston General Hospital, meta-analyses that the benefits are limited, short term, and do not justify routine use. Kingston, Ontario, Canada (W.H.). Similarly, although steroids might reasonably be expected to decrease the inflammatory Supported by the Queen Alexandra Founresponse in bronchiolitis, published data are conflicting, with equally well-designed dation for Children, British Columbia, Can16-18 19-21 ada; Vancouver Island Health Authority, or ineffective. The studies concluding that steroids may be either effective Youth and Maternal Programme, British primary treatment, therefore, remains largely supportive, with administration of fluids and Columbia, Canada; and an Ontario Tho8,22 supplemental oxygen, observation, and mechanical ventilatory support as needed. racic Society block term grant. No reprint requests are available from the Several reports over the last decade have demonstrated that inhalation of nebulized authors. 6% to 10% hypertonic saline (HS) improves both immediate and long-term clearance of Submitted for publication Aug 17, 2006; 23-26 small airways in patients with cystic fibrosis. The exact mechanism is unknown but last revision received Mar 7, 2007; accepted Apr 9, 2007. is thought to facilitate removal of inspissated mucus through osmotic hydration, disrupReprint requests: Brian A. Kuzik, MD, MSc, tion of mucus strand cross-linking, and reduction of mucosal edema.27,28 In otherwise

R

ANOVA HS KGH LOS NS

266

Analysis of variance Hypertonic saline Kingston General Hospital Length of stay Normal saline

RDAI RSV SaO2 SKMC VGH

Respiratory Distress Assessment Instrument Respiratory syncytial virus Oxygen saturation Sheikh Khalifa Medical City Victoria General Hospital

FRCP(C), Department of Paediatrics, Royal Victoria Hospital of Barrie, 208-1 Quarry Ridge Road, Barrie, Ontario, Canada L4M 6M2. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.04.010

healthy infants hospitalized with viral bronchiolitis, the regular administration of nebulized 3% HS combined with epinephrine decreased length of stay (LOS) by approximately 22% compared with infants receiving the same dose of epinephrine mixed in 0.9% normal saline (NS).29 Similarly, in ambulatory infants with mild bronchiolitis, inhalation of nebulized 3% HS (with terbutaline) improved clinical scores but did not produce a decrease in hospital admission rate.30 Both of the aforementioned studies used 3 times per day dosing, which is significantly less than the 3 to 6 times per hour regimens often used to deliver nebulized medication to children in respiratory distress.31-33 The purpose of the present study was to investigate the addition of frequently nebulized 3% HS to standard therapy of moderately ill infants hospitalized with typical viral bronchiolitis in a prospective, randomized, double-blind, controlled fashion. The primary objective was to compare the LOS of these infants with that of a control group of infants receiving standard therapy plus frequently nebulized NS.

METHODS Patients Infants up to age 18 months who were admitted to the hospital for the treatment of moderately severe viral bronchiolitis were eligible for study. The diagnosis of moderately severe bronchiolitis required a history of a preceding viral upper respiratory infection, the presence of wheezing or crackles on chest auscultation, plus either an oxygen saturation (SaO2) of ⬍94% in room air or significant respiratory distress as measured by a Respiratory Distress Assessment Instrument (RDAI)34 score of ⱖ4. In brief, 6 separate assessments of retractions and auscultatory findings are made and assigned a numerical score; the sum of these scores provides the RDAI score ranging from 0 to 17, with increasing scores indicating increasing respiratory distress. Exclusion criteria included a history of any of the following: previous episode of wheezing, chronic cardiopulmonary disease or immunodeficiency; critical illness at presentation requiring admission to intensive care; the use of nebulized HS within the previous 12 hours; or premature birth (gestational age ⱕ 34 weeks). Setting The study was conducted at 3 regional tertiary care hospitals: Sheikh Khalifa Medical City (SKMC), Abu Dhabi, United Arab Emirates; Victoria General Hospital (VGH), Victoria, British Columbia, Canada, and Kingston General Hospital (KGH), Kingston, Ontario, Canada. VGH and KGH serve multiethnic populations in the west coast and central regions of Canada, respectively. Data were collected during the winter bronchiolitis seasons between December 2003 and April 2006.

Study Design Patients admitted to hospital with bronchiolitis were assessed within 12 hours for entry into the study. If inclusion/ exclusion criteria were satisfied, then informed consent was obtained, and the patient was randomized to receive treatment with 4 mL of nebulized study solution containing either 3% HS (study group) or NS (control group). The study solution was administered in a double-blind fashion every 2 hours for 3 doses, followed by every 4 hours for 5 doses, followed by every 6 hours until discharge. After study enrollment, any additional (nonprotocol) treatments were at the sole discretion of the attending physician, who was blinded to the study treatment. If additional treatments included nebulized medication, the medication was nebulized in 4 mL of the assigned study solution (ie, HS or NS). All inhaled therapies were delivered to a settled infant from a standard oxygen-driven hospital nebulizer through a tight-fitting facemask, or head box, whichever was better tolerated by the infant. Patients were randomized independently at each study site to receive either HS or NS using a computer-based randomization program. Study solutions were prepared by a research pharmacist and were identical in appearance and odor. The identity of the study solutions was blinded to all participants, care providers, and investigators. Clinical response was determined by the designated study physician using RDAI scores and SaO2 readings at study entry and then at least once daily. Determination of LOS LOS was defined as the time between study entry (within 12 hours of admission to the hospital) and the time at which the infant either reached protocol-defined discharge criteria as measured by the study physician or was discharged from the hospital on independent clinical grounds by the attending physician, whichever came first. Protocol-defined discharge criteria required both an RDAI score ⬍4 and an SaO2 of at least 95% in room air for 4 hours. Ethics The study was approved by the ethics and human research committees of the 3 participating hospitals. Informed written consent was obtained from at least 1 parent of each infant before enrollment. Statistical Strategy A reduction in LOS of 1 day was previously proposed as being clinically significant32 and was adopted in this study. It was anticipated that this would require a sample size of approximately 46 patients per trial arm, for 80% power, to show a P value ⱕ .05. This number is based on a prestudy mean LOS at the largest study hospital (SKMC) of 4.1 ⫾ 1.7 days (unpublished data). Data were entered into an Excel spreadsheet (Microsoft Corp, Redmond, WA) and imported into SPSS version 12.0.1 software (SPSS Inc, Chicago, IL)

Nebulized Hypertonic Saline in the Treatment of Viral Bronchiolitis in Infants

267

Table I. Patient demographics and illness status at baseline

% male Age (months) Duration of illness before admission (days) Respiratory distress clinical score % SaO2 in room air Infants treated with bronchodilator before study entry (%) Infants treated with systemic steroids before study entry (%) Infants treated with antibiotics before study entry (%) Infants tested for RSV RSV positive (%)

Treatment

HS (n ⴝ 47)*

NS (n ⴝ 49)*

P

57% 4.4 ⫾ 3.7 4.5 ⫾ 2.3

61% 4.6 ⫾ 4.7 4.0 ⫾ 2.4

.84 .54 .30

7.8 ⫾ 2.5

8.1 ⫾ 3.3

.69

94.9 ⫾ 3.9 37 (86%)

95.2 ⫾ 3.4 41 (91%)

.71 .52

1 (2.5%)

1 (2.4%)

6 (15%)

4 (9.8%)

.52

40 25 (62%)

40 30 (75%)

1.0 .39

1.0

*Sample sizes vary slightly for the individual comparisons due to missing data.

for analysis on an intention-to-treat basis. Descriptive analyses were completed overall and also for the control and study groups separately. The ␹2 test (Fisher’s exact) was used to examine the association between categorical variables and group, and independent sample t tests and Levene’s test for equality of variance were used to assess the association between numeric variables and group. One-way analysis of variance (ANOVA) was used to compare data from the 3 study sites. To test for the potential effect of age on the results, the patients were divided into 3 age groups (0 to 6 months, 7 to 12 months, and 13 to 18 months), and the effects of age and treatment were tested in a 2-way ANOVA.

RESULTS Study Population A total of 96 previously well infants (mean age, 4.7 ⫾ 4.2 months; range, 10 days to 18 months) with viral bronchiolitis were enrolled from 3 centers during the bronchiolitis seasons from December 2003 to May 2006. Thirty-two infants were enrolled from the 2 Canadian sites (VGH and KGH), and 64 infants were enrolled from SKMC. Fortyseven infants were randomized to the HS treatment group, and 49 were randomized to the NS control group. Five infants (2 from the HS group and 3 from the NS group) were withdrawn at parental request before study completion but were included in the final intention-to-treat analysis. The HS and NS groups were comparable at baseline and typically presented on the fifth day of illness (range, 1 to 14 days) with borderline hypoxia (mean SaO2, 95%; range, 85% to 100%) and moderate respiratory distress (mean clinical score, 8 out of 17; range, 4 to 17) (Table I). Some 69% of all 268

Kuzik et al

Table III. Treatments received during the study

Study solution alone (nebulizations/day) Albuterol ⫹ study solution (nebulizations/day) Racemic epinephrine ⫹ study solution (nebulizations/day) Steroids ⫹ study solution (nebulizations/day) Total nebulizations/day Number of patients given any systemic steroid (%) Number of patients given any antibiotic (%)

HS (n ⴝ 47)*

NS (n ⴝ 49)*

P

3.2 ⫾ 3.0

3.8 ⫾ 4.1

.46

3.1 ⫾ 3.5

3.6 ⫾ 3.6

.49

2.7 ⫾ 3.7

1.6 ⫾ 2.4

.13

0.39 ⫾ 0.83

0.26 ⫾ 0.60

.42

9.1 ⫾ 3.0 8 (17%)

9.2 ⫾ 4.5 7 (14%)

.93 .78

5 (11%)

10 (20%)

.26

*Sample sizes vary slightly for the individual comparisons due to missing data.

tested infants were positive for RSV. Subset comparison of the SKMC and Canadian sites revealed minimal differences at baseline (Table II; available at www.jpeds.com). Although the Arab infants tended to be sicker (RDAI 8.9 ⫾ 2.9 vs 6.2 ⫾ 1.9; P ⬍ .001) and more likely to receive previous treatment with a bronchodilator (98% vs 70%; P ⬍ .001), all other measurements were comparable.

Treatment Received After enrollment, all treatments (protocol and add-on) received by infants in the HS and NS groups were comparable (Table III). The infants received a mean of 9 nebulizations of study solution per day delivered alone (38% of treatments) or co-administered with albuterol (salbutamol; 37%), racemic epinephrine (racepinephrine; 23%), or inhaled steroid (3%). Subset comparison of the SKMC and Canadian sites revealed minimal differences in the treatments received (Table IV; available at www.jpeds.com). Treatment at SKMC was more likely to include antibiotics (P ⫽ .002) as well as the addition of racemic epinephrine to the inhaled study solution (P ⫽ .003). Adverse Effects of HS All participants tolerated therapy without apparent adverse effects and were eventually discharged after achieving full recovery. No infants were withdrawn by the medical staff due to clinical deterioration or the need for intensive care support. Although 5 infants were withdrawn at parents’ request because of perceived adverse effects of therapy, only 2 of these infants were receiving HS. One of these infants (a 2-month-old male) cried very vigorously during his third inhalation (HS alone) and again with his fifth inhalation (HS with racemic epinephrine) and was withdrawn at that time. This was not associated with any significant acute change in his clinical condition, and he was eventually discharged on day 6. The second infant (a 3-month-old female) was withdrawn because of agitation after her second inhalation (HS The Journal of Pediatrics • September 2007

Figure. Percentage of patients in each group remaining in the hospital.

with albuterol), which was not associated with a significant change in her respiratory status. She was eventually discharged on day 2.

Response to Therapy The endpoint of LOS was identified by the attending physician using clinical grounds alone (45% of patients) or by reaching protocol-established discharge criteria as measured by the study physician (55% of patients), whichever came first. One-way ANOVA confirmed that the LOS did not differ significantly between study sites for either the NS group (P ⫽ .12) or the HS group (P ⫽ .44). Infants in the control group had a mean LOS of 3.5 ⫾ 2.9 days, whereas infants treated with nebulized 3% HS were discharged on average 1 day sooner, with a 26% reduction in LOS to 2.6 ⫾ 1.9 days (P ⫽ .05). There was a trend toward greater improvement in infants under age 6 months, but this difference did not attain statistical significance (P ⫽ .17). The percentage of patients from each group remaining in hospital each day is shown in the Figure.

DISCUSSION This study demonstrates that inhaled 3% HS is an effective treatment for infants up to age 18 months hospitalized with viral bronchiolitis. Repeated inhalations of nebulized HS reduced the LOS by approximately 1 day, from 3.5 ⫾ 2.9 to 2.6 ⫾ 1.9 days. This is a clinically relevant benefit with the potential for widespread impact on the treatment of bronchiolitis. The infants that we studied came from a population that was geographically and ethnically very diverse. Nevertheless, these infants were very similar to those described in other

bronchiolitis studies, with a slight male predominance (62%), primary infection with RSV (69%), mean age of 4.7 months, and LOS in the control group of 3.5 to 4 days.7,35,36 Strict inclusion and discharge criteria were used to minimize possible confounding effects of uncharacterized and evolving wheezing phenotypes and to minimize between-site variability. The clinical scoring system chosen has been widely used in other studies on bronchiolitis16,32,37 and has been proposed to be the scoring system of choice for further studies.15 Therefore, our findings should be universally applicable to other previously healthy infants hospitalized with moderately severe viral bronchiolitis. The majority of our patients received bronchodilators before study entry. In addition, although our study protocol did not require or encourage the co-administration of bronchodilator with the study solution, blinded attending physicians prescribed bronchodilators approximately 5 times per day. This finding was not unexpected, because the use of bronchodilators in bronchiolitis remains widespread, with some reporting it in more than 80% of patients.11,12 It is also possible that attending physicians prescribed bronchodilators to prevent possible adverse effects of HS. Although inhalation of HS may cause bronchoconstriction in asthmatics,38 and co-administration with a bronchodilator is often recommended,24,39 others have reported that inhalation of 4.5% to 7% HS (without a bronchodilator) can be performed safely in healthy nonasthmatic children40,41 or in children with moderately severe small airway obstruction secondary to cystic fibrosis.42 In our study, there were no apparent adverse effects attributable to the use of HS without a bronchodilator, although the numbers were insufficient to allow further exploration of this issue. However, there was no increase in add-on bronchodilator therapy in the treatment group, suggesting that the use of HS in this setting was not associated with a clinically significant increase in lower airway obstruction. The use of inhaled HS in the treatment of viral bronchiolitis in hospitalized infants is a novel therapy that was first reported in 200339 and recently strengthened with the publication of a 2-year extension of the original study.29 These authors demonstrated that 3 times a day dosing with 4 mL of 3% HS containing 1.5 mg of epinephrine compared with the same dose of epinephrine in NS reduced the LOS from 3.6 ⫾ 1.6 days to 2.8 ⫾ 1.3 days, a 22% improvement (P ⬍ .05). They included epinephrine to prevent possible adverse effects of HS and attributed the beneficial effects in the treatment group to the presence of HS. Our study was very similar but differed primarily in the inclusion of slightly older infants (up to age 18 months), plus the much more frequent dosing of HS (9.1 ⫾ 3.0 inhalations/day). In our hands, increasing the frequency of inhaled HS produced a further reduction in the LOS to 26%, but this reduction was not significant compared with 3 times a day dosing. The routine use of 3% HS in the treatment of infants hospitalized with bronchiolitis has the potential for enormous economic benefit. A 26% reduction in LOS not only will return infants to home and their parents to work a day sooner,

Nebulized Hypertonic Saline in the Treatment of Viral Bronchiolitis in Infants

269

but also will also substantially reduce hospital costs. The estimated hospital costs for bronchiolitis in the US, which includes the widespread use of bronchodilators nebulized with NS,11,12 exceed $580 million per year.6,43 Therefore, the substitution of NS with the comparably priced 3% HS, with the subsequent reduction in LOS, has the potential to save the US healthcare system more than $150 million annually. In summary, inhaled 3% HS is a safe, inexpensive, and effective treatment for previously well infants admitted to the hospital with moderately severe viral bronchiolitis. Further research is needed to determine the optimum dosing and to identify whether there is any benefit from co-administered bronchodilator. We thank Jaishen Rajah, Senior Consultant in Paediatrics, SKMC, for his initial statistical guidance.

REFERENCES 1. Gleazen WP, Denny FW. Epidemiology of acute lower respiratory disease in children. N Engl J Med 1973;288:498-505. 2. Ray CG, Minnich LL, Holberg CJ, Shehad ZM, Wright AL, Barton LL, et al. Respiratory syncytial virus–associated lower respiratory illnesses: possible influence of other agents. Pediatr Infect Dis 1993;12:15-9. 3. Smyth RL, Openshaw PJM. Bronchiolitis. Lancet 2006;368:312-22. 4. Hall CB. RSV and parainfluenza virus. N Engl J Med 2001;344:1917-28. 5. McConnochie KM, Roghmann KJ, Liptak GS. Hospitalization for lower respiratory tract illness in infants: variation in rates among counties in New York State and areas within Monroe Counties. J Pediatr 1995;126:220-9. 6. Pelletier AJ, Mansbach JM, Camargo CA. Direct medical costs of bronchiolitis hospitalizations in the United States. Pediatrics 2006;118:2418-23. 7. Wright RB, Pomerantz WJ, Luria JW. New approaches to respiratory infections in children: bronchiolitis and croup. Emer Med Clin North Am 2002;20:93-114. 8. Subcommittee on Diagnosis and Management of Bronchiolitis, American Academy of Pediatrics. Diagnosis and management of bronchiolitis. Pediatrics 2006; 118:1774-93. 9. Welliver JR, Welliver RC. Bronchiolitis. Pediatr Rev 1993;14:134-9. 10. Aherne W, Bird T, Court SDM, Gardner PS, McQuillin J. Pathological changes in viral infections of the lower respiratory tract in children. J Clin Pathol 1970;23:7-18. 11. Kostagal UR, Robbins JM, Kini NM, Schoettker PJ, Atherton HD, Kirschbaum MS. Impact of a bronchiolitis guideline: a multi-site demonstration project. Chest 2002;121:1789-97. 12. Christakis DA, Cowan CA, Garrison MM, Molteni R, Marcuse E, Zerr DM. Variation in inpatient diagnostic testing and management of bronchiolitis. Pediatrics 2005;115:878-84. 13. Kellner JD, Ohlsson A, Gadomski AM, Wang EEL. Efficacy of bronchodilator therapy in bronchiolitis: a meta-analysis. Arch Pediatr Adolesc Med 1996;150:1166-72. 14. Kellner JD, Ohlsson A, Gadomski AM, Wang EEL. Bronchodilators for bronchiolitis. Cochrane Database Syst Rev 2000;CD001266. 15. Flores G, Horwitz RI. Efficacy of beta-2 agonists in bronchiolitis: a reappraisal and meta-analysis. Pediatrics 1997;100:233-9. 16. Schuh S, Coates AL, Binnie R, Allin T, Goia C, Corey M, et al. Efficacy of oral dexamethasone in outpatients with acute bronchiolitis. J Pediatr 2002;140:27-32. 17. Csonka P, Kaila M, Laippala P, Iso-Mustajarvi M, Vesikari T, Ashorn P. Oral prednisolone in the acute management of children age 6 to 35 months with viral respiratory infection-induced lower airway disease: a randomized, placebo-controlled trial. J Pediatr 2003;143:725-30. 18. Weinberger M. Oral prednisolone in the acute management of children age 6 to 35 months with viral respiratory infection–induced lower airway disease: a randomized, placebo-controlled trial. J Pediatr 2004;145:137-8.

270

Kuzik et al

19. Patel H, Platt R, Lozano JM, Wang EEL. Glucocorticoids for acute viral bronchiolitis in infants and young children. Cochrane Database Syst Rev 2004:CD004878. 20. Bulow SM, Nir M, Levin E, Friis B, Thomsen LL, Nielsen JE, et al. Prednisolone treatment of respiratory syncytial virus infection: a randomized controlled trial of 147 infants. Pediatrics 1999;104:e77. 21. van Woensel JB. Long-term effects of prednisolone in the acute phase of bronchiolitis caused by respiratory syncytial virus. Pediatr Pulmonol 2000;30:92-6. 22. King VJ, Viswanathan M, Bordley C, Jackman AM, Sutton SF, Lohr KN, et al. Pharmacologic treatment of bronchiolitis in infants and children: a systematic review. Arch Pediatr Adolesc Med 2004;158:127-37. 23. Elkins MR, Robinson M, Rose BR, Harbour C, Moriarty CP, Marks GB, et al. A controlled trial of long-term inhaled hypertonic saline in patients with cystic fibrosis. N Engl J Med 2006;354:229-40. 24. Eng PA, Morton J, Douglass JA, Riedler J, Wilson J, Robertson CF. Short-term efficacy of ultrasonically nebulized hypertonic saline in cystic fibrosis. Pediatr Pulmonol 1996;21:77-83. 25. Riedler J, Reade T, Button B, Robertson CF. Inhaled hypertonic saline increases sputum expectoration in cystic fibrosis. J Paediatr Child Health 1996;32:48-50. 26. Suri R, Grieve R, Normand C, Metcalfe C, Thompson S, Wallis C, et al. Effects of hypertonic saline, alternate day and daily rhDNase on healthcare use, costs and outcome in children with cystic fibrosis. Thorax 2002;57:841-6. 27. Robinson M, Hemming AL, Regnis JA, Wong AG, Bailey DL, Bautovich GJ, et al. Effect of increasing doses of hypertonic saline on mucociliary clearance in patients with cystic fibrosis. Thorax 1997;52:900-3. 28. Tomooka LT, Murphy C, Davidson TM. Clinical study and literature review of nasal irrigation. Laryngoscope 2000;110:1189-93. 29. Tal G, Cesar K, Oron A, Houri S, Ballin A, Mandelberg A. Hypertonic saline/ epinephrine treatment in hospitalized infants with viral bronchiolitis reduces hospitalizations stay: 2 years experience. IMAJ 2006;8:169-73. 30. Sarrell E, Tal G, Witzling M, Someck E, Houri S, Cohen HA, et al. Nebulized 3% hypertonic saline solution treatment in ambulatory children with viral bronchiolitis decreases symptoms. Chest 2002;122:2015-20. 31. Wainright C, Altamirano L, Cheney M, Cheney J, Barber S, Price D, et al. A multicentre, randomized, double-blind, controlled trial of nebulized epinephrine in infants with acute bronchiolitis. N Engl J Med 2003;349:27-35. 32. Patel H, Platt RW, Pekeles GS, Ducharme FM. A randomized controlled trial of the effectiveness of nebulized therapy with epinephrine compared with albuteral and saline in infants hospitalized for acute viral bronchiolitis. J Pediatr 2002;141:818-24. 33. Ray MS. Comparison of nebulized adrenaline versus salbutamol in wheeze associated with respiratory tract infection in infants. Indian Pediatr 2002;39:12-22. 34. Lowell DI, Lister G, Von Koss H, McCarthy P. Wheezing in infants: the response to epinephrine. Pediatrics 1987;79:939-45. 35. Cheney J, Barber S, Altamirano L, Cheney M, Williams C, Jackson M, et al. A clinical pathway for bronchiolitis is effective in reducing readmission rates. J Pediatr 2005;147:622-6. 36. Todd J, Bertock D, Dolan S. Use of a large national database for comparative evaluation of the effect of a bronchiolitis/viral pneumonia clinical care guidelines on patient outcome and resource utilization. Arch Pediatr Adolesc Med 2002;156:1086-90. 37. Menon K, Sutcliffe T, Klassen TP. A randomized trial comparing the efficacy of epinephrine with salbutamol in the treatment of acute bronchiolitis. J Pediatr 1995;126:1004-7. 38. Cataldo D, Foidant JM, Lau L, Bartsch P, Djukanovic R, Louis R. Induced sputum: comparison between isotonic and hypertonic saline solution in patients with asthma. Chest 2001;120:1815-21. 39. Mandelberg A, Tal G, Witzling M, Someck E, Houri S, Balin A, et al. Nebulized 3% hypertonic saline solution treatment in hospitalized infants with bronchiolitis. Chest 2003;123:481-7. 40. Araki H. Inhalation of hypertonic saline as a bronchial challenge in children with mild asthma and normal children. J Allergy Clin Immunol 1989;84:99-107. 41. Williams PV. Inhalation bronchoprovocation in children. Immunol Clin North Am 1998;18:149-64. 42. Suri R, Marchall LJ, Wallis C, Mecalfe C, Shute JK, Bush A. Safety and use of sputum induction in children with cystic fibrosis. Pediatr Pulmonol 2003;35:309-13. 43. Friedman SM. The inflation calculator. Available at http://www.westegg.com/ inflation/.

The Journal of Pediatrics • September 2007

Table II. Site-specific patient demographics and illness status at baseline

% male Age (months) Duration of illness before admission (days) Respiratory distress clinical score % oxygen saturation in room air Previous treatment with bronchodilator (%) Previous treatment with systemic steroids (%) Previous treatment with antibiotics (%) Tested for RSV (%) RSV positive (%)

Table IV. Site-specific treatments received

SKMC (n ⴝ 64)*

VGH ⴙ KGH (n ⴝ 32)*

P

59 4.4 ⫾ 4.3 4.2 ⫾ 2.6

59 5.3 ⫾ 4.1 4.2 ⫾ 1.9

1.0 .34 .87

8.9 ⫾ 2.9

6.2 ⫾ 1.9

⬍.001

94.7 ⫾ 3.8

95.8 ⫾ 3.3

.17

98

70

Treatment

⬍.001

3.4

0.0

.91

15.5

4.3

.27

89.7 62.1

87.5 61.3

.74 .94

Study solution alone (nebulizations/day) Albuterol ⫹ study solution (nebulizations/ day) Racemic epinephrine ⫹ study solution (nebulizations/day) Steroids ⫹ study solution (nebulizations/ day) Total nebulizations/day Number of patients given any systemic steroid (%) Number of patients given any antibiotic (%)

SKMC (n ⴝ 64)*

KGH ⴙ VGH (n ⴝ 32)*

2.4 ⫾ 2.4

6.0 ⫾ 4.1

.03

3.8 ⫾ 3.8

2.4 ⫾ 2.4

.12

2.9 ⫾ 3.6

0.48 ⫾ 1.0

⬍.01

0.24 ⫾ 0.72

0.48 ⫾ 0.72

.77

9.4 ⫾ 3.8 10 (16%)

8.9 ⫾ 4.1 5 (16%)

.71 1.0

15 (23%)

0 (0%)

⬍.01

P

*Sample sizes vary slightly for the individual comparisons due to missing data.

*Sample sizes vary slightly for the individual comparisons due to missing data.

Nebulized Hypertonic Saline in the Treatment of Viral Bronchiolitis in Infants

270.e1

Deaths and Injuries Attributed to Infant Crib Bumper Pads BRADLEY T. THACH, MD, GEORGE W. RUTHERFORD, JR, MS,

AND

KATHLEEN HARRIS

Objective To document deaths attributed to bumper pads and injuries from their use that are potentially preventable. Study design The US Consumer Product Safety Commission maintains files on cases voluntarily reported to them of deaths and injury related to commercial products. These cases represent an unknown fraction of total occurrences. We searched this database for deaths related to crib bumpers for the years 1985 to 2005. We also searched other Consumer Product Safety Commission databases for crib-related injuries that potentially might have been prevented by bumpers. Additionally, we examined 22 retail crib bumpers and described features that could be hazardous. Results Twenty-seven accidental deaths reported by medical examiners or coroners were attributed to bumper pads. The mechanism of death included suffocation and strangulation by bumper ties. Twenty-five nonfatal injuries were identified, and most consisted of minor contusions. All retail bumpers had hazardous properties. Conclusions These findings suggest that crib and bassinet bumpers are dangerous. Their use prevents only minor injuries. Because bumpers can cause death, we conclude that they should not be used. (J Pediatr 2007;151:271-4)

ost infant cribs sold in the United States are used with bumper pads. Whether crib bumper pads pose a risk to infants for accidental suffocation is controversial. Recently, the Juvenile Product Manufacturing Association (JPMA) asked the US Consumer Products Safety Commission (CPSC) to review crib deaths involving suffocation or strangulation. On the basis of their own analysis of an unpublished CPSC review, representatives of the JPMA independently concluded, “there were no deaths directly related to the traditional use of crib bumper pads.”1 However, several organizations, including the CPSC and the American Academy of Pediatrics, have stated that crib bumpers are a potential risk when they are “pillow like.”2,3 In addition, the First Candle Sudden Infant Death Syndrome Alliance cautions that bumper pads should be “thin, firm but not pillow like.”4 These are subjective assessments and open to interpretation; thus caregivers may have difficulty in applying these criteria to their purchases of bumper pads. Because there are no detailed and systematically gathered data on hazards of crib bumper pads, we searched for cases of accidental death attributed to crib bumpers in CPSC databases. Also, because crib bumpers are intended to reduce the risk of injury, we searched CPSC’s injury database for non-fatal crib injuries that conceivably might have been prevented by crib bumpers. Finally, we have examined crib bumpers currently on the market for features that might be construed as pillow-like or otherwise potentially dangerous.

M

METHODS Bumper-related suffocation deaths were identified through a search of CPSC databases from Jan 1, 1985, through Dec 31, 2005, made available to the public. Three CPSC databases were searched. These include the Death Certificate, Injury and Potential Injury Incidents, and In-Depth Investigations databases. The CPSC receives death certificates from all 50 states, the District of Columbia, and New York City; these include deaths from all suffocation codes, with the exception of the suffocation code for “falling earth” that was in use with the International Classification of Diseases, Ninth Revision coding system. This information is stored in the Death Certificate database. The CPSC also collects information on deaths from medical examiners, coroners, and other sources such as police and fire departments and media articles that are stored in the Injury and Potential Injury Incidents database or stored in the In-Depth Investigations database. The information in the 3 databases contains unique information about deaths and duplicates

CPSC JPMA

US Consumer Products Safety Commission Juvenile Product Manufacturing Association

NEISS

National Electronic Injury Surveillance System

See editorial, p 237 From Washington University Department of Pediatrics, St. Louis, Missouri. Submitted for publication Sep 7, 2006; last revision received Feb 28, 2007; accepted Apr 16, 2007. Reprint requests: Bradley T. Thach, MD, Washington University Department of Pediatrics, 660 S Euclid, Campus Box 8208, St. Louis, MO 63110. E-mail: Thach@ kids.wustl.edu. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.04.028

271

Figure 2. Death scene reconstruction of case #15. Mannequin placed in position in which the infant was found dead. Figure 1. Death scene reconstruction of case #1. Infant’s neck was actually extended with his face pressed into the bumper. This is not shown in photo because of inability to extend mannequin’s neck.

reports that may provide additional information about deaths. Because the CPSC does not receive all deaths reported in the United States, the deaths in the study should be considered a minimum number. The databases were searched for the keywords “bumper,” “pad,” and “padding” for deaths involving infants aged from 1 month through 2 years. The search was not restricted in sleeping location, external cause of death code, or other identifier. Deaths identified in all of the databases were combined and sorted by state, age, and sex to identify duplicate cases, and deaths were removed that were duplicates or out-of-scope (eg, mattress pad, heating pad), yielding a final dataset of 27 deaths. Crib-related injury cases were identified though CPSC’s National Electronic Injury Surveillance System (NEISS). NEISS is a probability sample of US hospital emergency departments stratified by emergency department size and geographic location. This database was searched from Jan 1, 2000, through Dec 31, 2004, by using product codes for cribs, portable cribs, crib extender rails or youth bed rails, and cribs not specified for infants aged ⱕ6 months. This age range was selected because after 6 months it is doubtful that bumpers would prevent head injury because most infants can raise their heads above the bumper pad. Although it is possible to determine national estimates using the NEISS, we made no attempt to do so because of the small number of cases identified. Files on these deaths and injuries were obtained and reviewed. Cases with evidence of non-traditional use of bumper pads were excluded. The authors assessed infant bumpers for sale at a St. Louis, Missouri, retail store; 22 different bumpers were examined and graded for softness, potential space between bottom of bumper and mattress, bumper width, and length of 272

Thach, Rutherford, and Harris

fabric fasteners that attach the bumper to the crib. Softness was graded on a scale of 1 to 3, with 3 being the consistency of a comforter or soft pillow and 1 being that of a typical couch cushion. We considered a typical cushion to be firm enough to provide comfort when a person otherwise would be sitting on or against a hard surface. It was obvious that softness varied a great deal from bumper to bumper. However, the site of the investigation necessitated a subjective assessment of this property.

RESULTS In this search, we found 27 cases of infant death involving bumper pads or similarly padded bassinets (4 of the 27 cases). In 26 cases, a death scene investigation was conducted. In 1 case, it was uncertain whether a formal investigation was made. Additionally, CPSC personnel conducted an additional scene investigation in 18 of the 27 cases. In all cases except 1 (#14), an autopsy was performed. Three types of infant death involving crib bumpers pads were found: 1) face against bumper (Figure 1); 2) infant wedged between bumper and other object (Figure 2), and 3): bumper tie around infant’s neck. There were 11 deaths in type 1 cases; 13 deaths in type 2 cases, and 3 deaths in type 3 cases (Table I; available at www.jpeds.com). There were 25 non-fatal crib injuries in the database (Table II; available at www.jpeds.com). It was unclear in most reports whether bumpers were present or not. Summaries in Tables I and II are those of the medical examiner or other health care workers (Table II). Twenty-two different crib bumper pads were evaluated for relevant properties at a retail outlet store in St. Louis (Table III; available at www.jpeds.com).

DISCUSSION Recently, the Canadian Healthy Environment and Consumer Safety Bureau in a brief report cited 23 “incidences” involving bumper pads, including 1 suffocation and 1 The Journal of Pediatrics • September 2007

strangulation death.5 The present report provides details of multiple infant deaths in which crib or bassinet bumper pads were thought to play a causal role. Also, it is a report of nonfatal injuries that might have been prevented had crib bumper pads been used. It must be emphasized that our search of the crib database reveals only an undetermined fraction of the actual incidents occurring in the United States in the period studied, because incidents are inconsistently reported to the CPSC and may or may not be published in media sources. Data on accidental deaths from US Vital Statistics are not coded by product. Thus CPSC data is the only resource at the national level with codes allowing for the identification of bumper-related deaths. The degree of underreporting is indicated by cases coming from only 17 states, with some states with large populations (New York, Texas) contributing only 1 case each and other less-populated states (Wisconsin, Missouri) reporting 3 cases each. It is important to consider limitations of our study. Underreporting of cases is one obvious limitation. In addition, scene investigations and autopsies were performed by different individuals, so there was no consistent protocol for these procedures. We have divided the bumper- and padded bassinetrelated deaths into 3 categories. The first are those in which the infant’s face was in close contact with the bumper surface, and death was either judged or could be assumed to be caused by asphyxia possibly resulting from re-breathing expired air or by nasal and oral compression.6,10 From past studies, the softest of the retail bumpers examined that had the characteristics of comforters or soft pillows would pose the greatest risk for this type of death.6-10 Case #6 in Table I is of particular interest because the bumper had a plastic covering, and it was suggested in the death scene report that moisture on the plastic caused the face to adhere to the bumper surface. This indicates that applying a nonporous covering over a bumper might not make it safer. Half the cases were in category 2. Here the infant’s head was determined to be wedged between a bumper and another surface. Death caused by wedging is a traditional diagnosis, and cases continue to be reported.11-13 An important contributing factor in wedging deaths is that many infants lack the motor development needed to extricate themselves.14 Death presumably results from asphyxia caused by re-breathing, nose and mouth compression, or a combination of these. Wedging occurs when the baby pushes his/her head into a narrow space between 2 surfaces. An important feature of the surface is that it is elastic and can spring back to its original shape after deformation. This characteristic provides the force pressing against the infant’s head, which causes the entrapment. Couch cushions are elastic and are universally recognized as a common cause of wedging deaths.12,13 Because the firmer and thicker retail bumpers we evaluated were elastic, like couch cushions, we deemed them to be more hazardous for wedging than the softer thinner bumpers. Considering this, it would not seem to be helpful to suggest that crib bumper pads be firm.4 The last category of death was strangulation. Infant deaths involving neck compression by cords, ribbons, or bands Deaths and Injuries Attributed to Infant Crib Bumper Pads

of various kinds is well-recognized, and frequent warnings to eliminate this hazard have been issued in past years. Current manufacturing standards state that “ribbons, strings, and ties on bumper guard should not exceed 9 inches.”15 It is relevant that in our own survey of commercially available bumpers there were 2 with fabric fasteners longer than 9 inches (case #5 and #10). Therefore, a strangulation hazard may still exist for some bumpers on the market. In theory, bumpers prevent injury from a baby’s head hitting crib bars or from extremities projecting through the bars. We cannot tell from the reports of crib injuries how effective bumpers are in protecting infants, because we do not know whether a bumper was present. The exception is the 1 case in which, ironically, the infant’s knee was reportedly contused when it struck a crib bumper pad (Table II, case #14). In the remaining cases, contusions and abrasions to the face and head conceivably could have been prevented had a bumper been in place. However, it is unclear whether a bumper would have prevented an arm or leg from passing through the crib rails, because we found an open space between the bumper and the crib mattress in all the bumper pads we examined. It is conceivable that a bumper might have contributed to the arm and leg injuries because it could provide a mechanism for limb entrapment. This could amplify the force on the limb exerted by an infant struggling to free itself. The seven reported cases of limb fractures or closed head injury were likely not caused by accidents. It is difficult to imagine an infant exerting a force sufficient to cause a limb fracture or hitting its head against a wooden slat with force enough to cause closed head injury. Currently, such cases would immediately raise a pediatrician’s suspicion of intentional injury. In summary, we report a number of fatal accidental infant deaths directly attributable to crib bumper pads. In direct contradiction to the JPMA interpretation of a CPSC staff data review that there were no incidents directly related to normal bumper use, we found 27 cases of death reported in the same CPSC databases. Moreover, an examination of commercial bumper pads indicates that these products continue to have characteristics that appear to be dangerous. Furthermore, a review of cases of non-fatal injuries in cribs indicates that these are not serious and might or might not have been prevented by bumper pads. This case series provides evidence that the risks from crib bumper pads or padded bassinettes (death) outweigh the possible benefits provided by such padding (minor bruises and contusions). Furthermore, our data does not suggest any way in which changes in bumper design can reduce risk of death. We conclude that bumpers should not be placed in cribs or bassinets.

REFERENCES 1. Crib safety Trade group departs from government recommendations. Consumer Reports. March 2005. Available at www.ConsumerReports.org. Accessed May 18, 2007. 2. Consumer Products Safety Commission. Crib safety tips— use your crib safely. Document #5030. Available at www.cpsc.gov. Accessed May 18, 2007. 3. Task Force on Sudden Infant Death Syndrome. The changing concept of sudden infant death syndrome: diagnostic coding shifts, controversies regarding the sleeping

273

environment, and new variables to consider in reducing risk. Pediatrics 2005; 116:1245-55. 4. First Candle/SIDS Alliance, First Candle/SIDS Alliance Recommendations for Parents and Caregivers. Baltimore, MD, October 2005. Available at www.Firstcandle.org. Accessed May 18, 2007. 5. Canadian Consumer Product Safety Bureau. Policy statement for bumper pads. August 17, 2005. Doucument # 05-100287-569. Available at www.hc.gc.ca/cps-spc/ legislation/pol/bumper-bordure_e.html. Accessed May 18, 2007. 6. Kemp JS, Thach BT. Sudden death in infants sleeping on polystyrene-filled cushions. N Engl J Med 1991;324:1858-64. 7. Chiodini B, Thach BT. Impaired ventilation in infants sleeping face down: potential significance for sudden infant death syndrome. J Pediatr 1993;123:686-92. 8. Bolton DPG, Taylor BJ, Campbell AG, Galland BC, Cresswell CA. A potential danger for prone sleeping babies: rebreathing of expired gases when face down in soft bedding. Arch Dis Child 1993;69:187-90. 9. Carleton JN, Donoghue AM, Porter WK. Mechanical model testing of rebreathing potential in infant bedding materials. Arch Dis Child 1998;78:323-8. 10. Patel A, Harris K, Thach BT. Inspired C02 and 02 in sleeping infants rebreathing from bedding: relevance for sudden infant death syndrome. J Appl Physiol 2001;91:2537-45.

274

Thach, Rutherford, and Harris

11. Drago DA, Dannenberg AL. Infant mechanical suffocation deaths in the United States, 1980-1997. Pediatrics 1999;103:e59. Available at http://www.pediatrics.org/cgi/ content/full/103/5/e59. Accessed May 18, 2007. 12. Scheers NJ, Dayton CM, Kemp JS. Sudden infant death with external airways covered: case comparison study of 206 deaths in the United States. Arch Pediatr Adolesc Med 1998;152:540-7. 13. Kemp JS, Unger B, Wilkins D, Psara R, Ledbetter T, Graham M, et al. Unsafe sleep practices and an analysis of bedsharing among infants dying suddenly and unexpectedly: results of a four-year, population-based, death scene investigation study of sudden infant death syndrome and related deaths. Pediatrics 2001;106: e41. 14. Paluszynska D, Harris K, Thach BT. Influence sleep position experience on ability of prone sleeping infants to escape from asphyxiating microenvironments by changing head position. Pediatrics 2004;114:1634-9. 15. American Society for testing and Materials. Standard consumer safety performance specification for infant bedding and related accessories, 2000, voluntary safety standard for bumper pads. www.astm.org. Annual book of ASTM standards, volume 15.07. Code of Federal Regulations. Washington, DC: US Government Printing Office; 2000.

The Journal of Pediatrics • September 2007

Table I. Medical examiners’ summaries of deaths 1. “Face obstructed by crib bumper pad- positional asphyxia. A male infant, age 2 months, died after he was found with his face against a bumper pad in his crib at home by his mother.” 2. “Died of asphyxiation caused by pressure against an overstuffed crib bumper during sleep. A 7-month old female was found unresponsive in her crib by her mother. The victim was placed on her back in the crib.” 3. “A coroner determined a 7-month-old male infant died in a crib due to positional asphyxiation—face in corner of crib against bumper pad. Victim was on his back with head turned to right, and his face was up into the corner of the bumper pad.” 4. “This incident involved the death of a 4-month-old infant due to positional asphyxia. The infant was found unresponsive by his mother. He had crawled face first into the corner of his crib with his nose and mouth pressed against the protective bumpers.” 5. “A 14-month-old baby boy died sleeping in a crib with his face pressed firmly against a bumper pad.” 6. “Baby got face into plastic bumper pad of cradle. Crib pad was much too large for this size of bed. Night was very hot, and it was felt that the crib pad adhered to the victim due to the heat. Baby got face into plastic bumper pad. Anoxia consistent with accidental suffocation.” 7. “A 13-month-old male was found dead in his crib while he and his mother were visiting at his grandmother’s house. The infants face was resting against a properly installed plastic bumper pad.” 8. “A 3-month-old male died of SIDS in his crib with his face against the bumper pad.” 9. “A 2-month-old female was found dead in her wicker infant basket for a nap after being fed at noon. She was found on her stomach, head turning to the left with face pressed slightly against the padded basket liner. The medical examiner found no anatomic cause and attributed the death to probable suffocation.” 10. “A 2-month-old male died of anoxia when he was sleeping and his face was pressed against the bumper of the ‘bassinet/carrier’ (cradle). The victim was dead on arrival. Note: Mother stated that the baby died due to the tilt of the bassinet/carrier.” 11. “Baby suffocated at home in the corner of the crib against the crib bumper. Suffocation—accidental.” 12. “Baby found face down in crib, pinned between bumper pad and sibling sister. A male infant, age 4 months, placed for a nap in a crib with a twin sister was found wedged between the bumper pad and his sister. Cause of death asphyxia due to positional crib accident.” 13. “A 4-month-old male was found dead in his crib at home. Reports indicated that the victim became wedged between the mattress and the bumper pad of his crib. The death was declared an accident; cause of death was listed as asphyxia by suffocation.” 14. “A 10-month-old male died of positional asphyxia, wedged between his crib railing and a dresser 6 inches away. He apparently stood on the crib bumper pads and climbed over the crib railing.” Author’s note: This case indicates yet another hazard of bumper use. The bumper allowed the infant to climb from a relatively safe environment into a hazardous one. 15. “Found unresponsive wedged between pillow and bumper pad. Positional asphyxia. Note: Mother reported the baby’s head had slipped off the edge of the pillow. His head was wedged between the pillow and the bumper pads inside the bed.” 16. “Seven-month-old girl was placed in her crib for a nap after being fed by her mother. Child was found later in her crib with her head wedged between the mattress and the bumper pad attached to side slats. Child was pronounced dead on arrival at hospital.” 17. “Found by mother with face wedged between crib mattress and bumper pads. COD: asphyxia.” 18. “An 11-month-old female slid off a day bed mattress. The crib bumper pad is believed to have become caught around the victim’s neck, and as she slid forward and she was unable to breathe and suffocated. The cause of death is mechanical asphyxia, the manner of death is considered accidental.” 19. “A 2-1/2-month-old male died due to probable suffocation. According to an investigator with the sheriff’s department, the infant’s mother found him face down in his crib. The investigator stated the baby’s head got caught between a baby blanket and the bumper pads in his crib. He was pronounced dead at the scene.” 20. “Face wedged in crib between pillow, mattress, and bumpers, external facial compression (suffocation).” 21. “An 8-month-old female died after being trapped tight against a side rail padding and mattress in her crib.” 22. “A 6-day-old female was found not responsive in her infant basket. She was on her stomach with her head turned to one side. Her face was pressed into the crevice between the basket mattress and padded sideliner. After an autopsy was preformed, the medical examiner ruled that death was caused by probable suffocation due to an external airway obstruction.” 23. “The baby was found wedged between adult pillows and crib bumper. The baby had originally been placed on her side and was found on her stomach.” 24. “A 2-month-old male was found dead in his crib. Autopsy examination revealed no cause of death, but findings frequently seen in sudden infant death syndrome. Based on circumstances surrounding the death as currently known, this death meets the criteria for sudden infant death syndrome.” Author’s note: The original death scene investigation makes no mention of infant’s head position at death, and so the medical examiner lacked this important information. A subsequent CPSC death scene investigation (Figure 2) indicated that the baby’s face was covered by a comforter, and his head was wedged between the mattress and the bumper pads. 25. “A 6-month-old female was strangled by the strings of her bumper pads while sleeping in her full size crib. She had placed her head through a loop formed by the tied fabric attachment strings of the bumper pad.” 26. “Asphyxiation by string-ligature. Father noted the string around baby’s neck. He pulled baby from crib, pulling the string from the bumper pad in the process. Police surmise that the baby had grasped the loosened tie in his hand then rolled over pulling the tie across the front of his neck. A mark was made.” 27. “Tie of bumper pad became tangled around neck. Cerebral anoxemia and anoxia; ligature compression of vessels.”

Deaths and Injuries Attributed to Infant Crib Bumper Pads

274.e1

Table II. Consumer Product Safety Commission file summaries of crib accidents 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

“Patient struck face on side of a crib at home, contusion on face.” “Child has a dent in side of head after pushing against bars of crib at home.” “Hit head on crib Dx. Head abrasion.” “Patient struck left knee against side of a crib, knee contusion.” “Patient fell forward in crib, bumping head on crib at home 7 days ago; head injury, head contusion.” “Four-month-old male, contusion to head, hit head on crib.” “Patient was in crib; mom came home, and patient had a bump on her forehead. Dx: mild head injury.” “Patient sustained head injury hit head on crib.” “Patient hit head against metal bassinet at home 2 days ago, has abrasion in forehead, crying, minor head injury, abrasion.” “Contusion to head when struck on crib.” “Patient’s legs were sticking out of crib bars this AM. Now his hip is making a popping sound. DX: sprain right leg.” “Mother states child hit face on side of crib. Dx: nasal contusion.” “Patient hit mouth on crib and sustained cut injury to inner mouth.” “Knee contusion— hitting bumper pads in baby bed-home.” “Left arm caught between bars in crib, contusion left arm.” “Trauma (R) forearm; patient got forearm stuck in the baby crib rail, crying and pain. Patient got arm struck in crib, was alone in bedroom, strain elbow.” “Contused head on bassinet.” “Patient caught arm in crib at home, not using arm; nursemaids elbow.” “Fx (Left Forearm), patient got her arm caught in the rails of the crib, cried a lot of pain.” “Patient got leg caught in crib, twisted thigh, arrives with swollen thigh, Lt femur fracture.” “Patient accidentally hit head against crib side. Dx: closed head trauma.” “Patient’s arm got stuck between crib and wall, and father states he heard a crack. Dx: Lt humerus fracture.” “Patient pushed against crib, dad heard snap. Femur fractured.” “Patient hit head on crib; closed head injury.” “Five-month-old female with fractured femur. Patient got leg caught in baby bed rails at home. Patient admitted.”

274.e2

Thach, Rutherford, and Harris

The Journal of Pediatrics • September 2007

Table III. Features of 22 retail crib bumper pads

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

Softness scale

Thickness (inches)

Length of bands attaching bumper to crib bars (inches)

3 2 1 2 1 2 1 1 2 2 3 3 2 1 2 1 3 1 1 2 3 3

1-1/16 1-1/4 1-3/4 1-3/4 1-1/4 1-5/8 1-1/2 2-3/4 2-1/8 1-3/4 1-3/4 1-5/8 2-3/4 2-1/4 3-3/4 2-3/4 1-3/4 1-7/8 2 1-1/2 1-5/8 1-3/4

6-1/2 8 8-1/4 6-3/4 9-1/4 7-1/2 8 7 7 9-1/8 8-3/4 8-1/4 8-1/4 7 8 7-1/2 6-1/4 7-1/2 6 8-1/2 8-1/2 8-1/2

Potential for head wedging

high high high high high

high high high high high high

In the assessment of softness, 1 is the hardest and 3 the softest; 2 is intermediate. The hardest and thickest (⬎2 inches) bumpers were deemed to have the highest potential for wedging.

Deaths and Injuries Attributed to Infant Crib Bumper Pads

274.e3

Insulin Resistance in Adolescents ANN M. RODDEN, DO, VANESSA A. DIAZ, MD, MS, ARCH G. MAINOUS III, PHD, RICHELLE J. KOOPMAN, MD, MS, AND MARK E. GEESEY, MS

Objectives

To investigate the relationship of other body mass index (BMI) ranges with Homeostasis Model Assessment– Insulin Resistance (HOMA-IR), a surrogate marker for insulin resistance in adolescents. Study design Cross-sectional analysis of a nationally representative sample of 1837 nondiabetic, nonpregnant 12 to 19 year old persons from the National Health and Nutrition Examination Survey, 1999-2002. The main outcome measurement of insulin resistance was calculated as HOMA-IR >3.16. Results Having a BMI >75th percentile is associated with a high HOMA-IR levels. As the BMI percentile increases, the odds of high HOMA-IR levels increase (BMI percentile 75-84.9, OR 4.277, 95% CI 2.090-8.752; BMI percentile 85-94.9, OR 4.299, 95% CI 2.158-8.563; BMI >95th percentile, OR 17.907, 95% CI 11.360-28.228). Conclusion Adolescents with BMI percentile of 75 to 84.9, which represents approximately 1.2 million US adolescents, have not previously been identified as having higher HOMA-IR levels. (J Pediatr 2007;151:275-9)

he proportion of adolescents who are overweight has risen from 6% in the 1970s to 17% in 2004.1-3 The classification of obesity for children and adolescents differs from that of adults. Instead of having set body mass index (BMI) levels (obese is a BMI ⱖ30, overweight a BMI of 25-29.9 for adults), children and adolescents are characterized by BMI percentiles. Being overweight is defined as a BMI of greater than or equal to the 95th percentile of Centers for Disease Control and Prevention (CDC) growth charts, and risk of being overweight is defined as a BMI percentile between 85 and 94.9. Normal weight is characterized as being between the 5th and 85th percentile.4 Being overweight is an established risk factor for insulin resistance, a precursor of type II diabetes mellitus, as well as hypertension, coronary heart disease, stroke, and cancer.5-9 It is unclear however, whether contemporary classifications of levels of BMI as overweight or risk of overweight are the only levels of BMI to have independent risk for insulin resistance because these are arbitrary classifications based on percentiles. The purpose of this study is to evaluate the association of BMI with an insulin resistance surrogate marker among adolescents while controlling for factors that are related to insulin resistance. In particular, we examine the prevalence of higher Homeostasis Model Assessment–Insulin Resistance (HOMA-IR) levels at specific BMI levels that consist of overweight, risk for overweight, and several groupings in the “normal” BMI category in a nationally representative sample of adolescents.

T

METHODS Study Population Data from the 1999-2002 National Health and Nutrition Examination Survey (NHANES) were analyzed. NHANES included participants from a nationally representative sample of non-institutionalized residents of the United States. The survey was conducted by the National Center for Health Statistics (NCHS) and included laboratory and interview information. Samples were weighted to be representative of the US population so population estimates could be made. The NHANES sampling weights account for unequal probaBMI CDC CV fitness HOMA-IR

Body mass index Centers for Disease Control and Prevention Cardiovascular fitness Homeostasis Model Assessment–Insulin Resistance

NCHS NHANES PIR

National Center for Health Statistics National Health and Nutrition Examination Survey Poverty income ratio

From the Medical University of South Carolina, Department of Family Medicine, Charleston, South Carolina. Submitted for publication Sep 22, 2006; last revision received Feb 13, 2007; accepted Mar 16, 2007. Reprint requests: Dr Ann M. Rodden, Medical University of South Carolina, Department of Family Medicine, 295 Calhoun Street, Charleston, South Carolina. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.023

275

Table I. Demographics and physical fitness of 12 to 19 year old nondiabetic adolescents by BMI

Sample size US population estimate % of population Age* Sex Male Female PIR ⬍1.0 ⱖ1.0 Ethnicity Non-Hispanic White Non-Hispanic Black Hispanic Activity Little Regular Heavy ⬍1 h/wk Heavy ⱖ1 h/wk CV fitness Low Moderate High

Total

95%

1837 31,285,208 100% 15.4 (0.1)

583 11,495,615 36.7 (1.5) 15.5 (0.1)

393 6,728,754 21.5 (1.1) 15.6 (0.2)

217 3,677,716 11.8 (1.1) 15.1 (0.2)

288 4,151,849 13.3 (1.2) 15.2 (0.2)

356 5,231,274 16.7 (1.2) 15.5 (0.2)

51.8 (1.7) 48.2 (1.7)

54.7 (2.9) 45.3 (2.9)

47.6 (4.0) 52.4 (4.0)

40.8 (4.0) 59.2 (4.0)

57.3 (4.4) 42.7 (4.4)

54.2 (4.9) 45.8 (4.9)

20.6 (1.9) 79.4 (1.9)

19.8 (2.8) 80.2 (2.8)

15.0 (2.5) 85.0 (2.5)

20.1 (4.0) 79.9 (4.0)

26.2 (3.7) 73.8 (3.7)

26.0 (3.0) 74.0 (3.0)

67.0 (2.0) 15.2 (2.1) 17.8 (2.2)

71.7 (2.7) 11.5 (1.8) 16.8 (2.9)

69.8 (2.5) 14.5 (2.7) 15.7 (2.1)

68.1 (2.5) 14.1 (2.6) 17.9 (3.3)

58.5 (4.4) 21.4 (3.3) 20.1 (3.6)

59.2 (4.1) 20.2 (3.3) 20.6 (3.2)

15.9 (1.2) 51.4 (2.0) 4.6 (0.8) 28.2 (2.2)

12.5 (2.1) 54.5 (4.1) 4.8 (1.6) 28.2 (3.8)

11.6 (1.9) 53.2 (3.5) 5.6 (1.6) 29.7 (3.4)

17.2 (4.1) 47.8 (5.8) 3.6 (1.7) 31.4 (5.2)

22.6 (3.2) 46.2 (5.2) 4.6 (1.9) 26.6 (4.7)

22.8 (4.1) 48.5 (5.1) 4.4 (1.1) 25.5 (3.4)

31.4 (2.1) 43.6 (1.7) 25.0 (1.5)

21.1 (2.3) 49.6 (2.8) 29.3 (3.1)

26.3 (3.9) 42.9 (4.3) 30.8 (3.7)

31.7 (5.2) 45.6 (5.4) 22.7 (4.1)

46.6 (3.9) 39.4 (3.3) 14.0 (3.1)

50.5 (4.6) 32.2 (4.0) 17.3 (3.6)

P value

.003

.006 ⬍.001

.110

⬍.001

Otherwise % (SE). *Mean (SE).

bilities of selection as a result of planned oversampling, sample design, and nonresponse, and they were matched to known population control totals to be representative of the US population. Sample weights also account for missing data. A fasting sample of adolescents between 12 and 19 years of age had laboratory tests obtained for further evaluation. Adolescents were excluded from this sample if they were currently pregnant or had previously been informed by a physician that they had diabetes.

Demographic Data Persons were categorized based on self report as nonHispanic White, non-Hispanic Black, or Hispanic. The poverty income ratio (PIR) was available from the NHANES data. This ratio takes into consideration the annual income before taxes of the family, adjusting for family size. A value ⬍1.00 was considered below the official poverty threshold. Body Mass Index BMI was calculated from the measured weight and height (kg/m2) collected by protocol. The CDC 2000 growth chart guidelines for grouping BMI percentiles by age in adolescents were used.4 Adolescents were grouped into the following categories by BMI percentile: ⬍50, 50.0-74.9, 7584.9, 85-94.9 (risk for overweight), and ⱖ95 (overweight). Dietary Variables Dietary history for the 24 hours before the interview was gathered in person by trained dietary interviewers using a 276

Rodden et al

multiple pass method. A standard set of measuring guides were used to help the respondent report the volume and dimensions of the food items consumed to simplify portion size estimation. The dietary recalls were further characterized as reliable and meeting the minimum criteria by the NCHS if ⬍25% of foods were missing descriptive information, ⬍15% were missing amounts, and the respondent remembered at least one food item per meal. Persons with unreliable data as determined by the NCHS were excluded. The total daily carbohydrate intake in grams was quantified from this 24-hour dietary recall.

Physical Activity Each participant was asked a series of questions pertaining to physical activity. Typical physical activity levels were determined from these questions. These categories consist of little or no regular recreation, sport, or physical activity; regular recreation or work-related physical activity of any time period during the week; regular, heavy physical activity ⬍1 hour a week; and regular, heavy physical activity 1 hour or more a week. Cardiovascular Fitness The cardiovascular (CV) fitness test consisted of a submaximal exercise test performed by trained health technicians. Each participant performed a treadmill protocol chosen based on age, sex, BMI, and self-reported physical activity. An estimated maximal oxygen uptake was calculated based on The Journal of Pediatrics • September 2007

heart rate measurements during the submaximal testing. Each participant’s CV fitness was categorized as low, moderate, or high, based on the estimated maximal oxygen uptake for age and sex.10

Homeostasis Model Assessment–Insulin Resistance (HOMA-IR) The gold standard for measuring insulin sensitivity is the hyperinsulinemic-euglycemic clamp technique, which is invasive, expensive, and time-consuming to perform. Because the NHANES collected fasting laboratory data, fasting glucose and insulin were obtained on participants. Although HOMA-IR is not the gold standard, HOMA-IR is able to be used in large populations of adolescents as a screening tool for insulin resistance.11 In the adolescent population, the homeostatic model assessment has been identified as being one of the best non-invasive techniques with sensitivity and specificity of 76% and 66%, respectively, with a cutoff value of 3.16. This differs from the value found to be useful in adults of ⬎2.50.11 Consequently, we will consider the HOMA-IR cutoff value of 3.16 to be high HOMA-IR. HOMA-IR was calculated as fasting insulin (␮U/mL) ⫻ fasting glucose (mg/ dL) divided by 22.5. Data Analysis Because of the complex survey design of the NHANES 1999-2002, univariate analysis and descriptive statistics were performed using SUDAAN software (Research Triangle Institute, Research Triangle Park, NC). Associations between variables and BMI percentiles were assessed using ␹2 tests. Associations between high HOMA-IR and variables also were assessed using ␹2 tests and t tests when appropriate. A P value ⬍ .05 was considered significant. We assessed predictors of high HOMA-IR in these nondiabetic 12 to 19 year olds using a logistic regression. An unadjusted model and a model adjusting for the demographics only were initially performed. The final adjusted model assessed the relationship of high HOMA-IR levels with BMI controlling for age, sex, ethnicity, PIR, and carbohydrates, along with self-reported physical activity and CV fitness. The total daily carbohydrate intake was included in the final adjusted model because it has been found to be statistically significant in previous studies evaluating insulin resistance.12

RESULTS Table I shows the characteristics of the unweighted study population of 1837 persons representing a weighted population of 31,285,208 adolescents between 12 and 19 years of age. The average age was 15.4 years, which was similar across all BMI categories. Overweight and risk for overweight adolescents reported less physical activity than normal weight adolescents. Overall, the majority of normalweight adolescents were classified as being of moderate fitness on the CV fitness test. More than half of those considered overweight achieved low CV fitness test results. Insulin Resistance in Adolescents

Table II. Demographics and physical fitness of 12 to 19 year old nondiabetic adolescents by BMI percentile

Sample size US population estimate Age–mean (SE) Sex Male Female PIR ⬍1.0 ⱖ1.0 Ethnicity Non-Hispanic White Non-Hispanic Black Hispanic BMI percentile ⬍50 50-74.9 75-84.9 85-94.9 ⱖ95 Activity Little Regular Heavy ⬍1 h/wk Heavy ⱖ1 h/wk CV fitness Low Moderate High Carbohydrates (g)

HOMA-IR 3.16

1225 22,341,944 15.5 (0.1)

581 8,513,896 15.3 (0.1)

73.0 (2.0) 71.8 (2.6)

27.0 (2.0) 28.2 (2.6)

66.2 (3.5) 75.0 (2.1)

33.8 (3.5) 25.0 (2.1)

75.7 (2.4) 65.1 (1.9) 65.5 (2.9)

24.3 (2.4) 34.9 (1.9) 34.5 (2.9)

90.8 (1.1) 86.3 (3.2) 66.2 (5.3) 62.2 (4.3) 27.2 (3.2)

9.2 (1.1) 13.3 (3.2) 33.8 (5.3) 37.8 (4.3) 72.8 (3.2)

61.7 (3.0) 72.6 (2.3) 65.5 (8.9) 80.8 (2.7)

38.3 (3.0) 27.4 (2.3) 34.5 (8.9) 19.2 (2.7)

59.8 (3.2) 78.9 (2.5) 80.3 (2.8) 338.6 (6.4)

40.2 (3.2) 21.1 (2.5) 19.7 (2.8) 303.5 (8.6)

P value

.180* .696

.029

.003

⬍.001

.005

⬍.001

.001*

Otherwise % (SE). *t test comparison of means.

Table II compares the same population by HOMA-IR levels. More than a third of Hispanic and Non-Hispanic Black adolescents had HOMA-IR ⬎3.16. Adolescents with high HOMA levels tended to report less activity (P ⫽ .005) and performed worse on the CV fitness test (P ⬍ .001). The prevalence of high HOMA-IR in the adolescent population with BMI ⱖ95th percentile was 73%. This decreased to 37.8% of adolescents in the BMI group with a percentile of 85 to 94.9. In the BMI percentile weight category with a BMI percentile of 75 to 84.9, which is characterized as normal weight, almost 34% of adolescents in this group also had high HOMA-IR levels. The amount of high HOMA-IR levels decreased below the 75th percentile BMI to about 10%. In all three models noted in Table III, a BMI ⱖ75th percentile was associated with high HOMA-IR levels. The odds of having high HOMA-IR in the final adjusted model was similar in adolescents with a BMI percentile of 75 to 84.9 (OR 4.277, 95% CI 2.090-8.752) and a BMI percentile of 85 to 94.9 (OR 4.299, 95% CI 2.158-8.563). 277

Table III. Association between HOMA-IR and BMI in nondiabetic adolescents age 12 to 19 years* Odds ratio (95%CI) BMI percentile

Unadjusted

Adjusted for demographics*

Adjusted†

⬍50 50-74.9 75-84.9 85-94.9 ⱖ95 Age Sex Male Female Ethnicity Non-Hispanic White Non-Hispanic Black Hispanic PIR ⬍1.0 ⱖ1.0 Activity Little Regular Heavy ⬍1 h/wk Heavy ⱖ1 h/wk CV fitness Low Moderate High Carbohydrates

1 1.577 (0.902-2.757) 5.066 (2.872-8.937) 5.959 (3.748-9.473) 26.521 (17.557-40.063)

1 1.539 (0.863-2.746) 4.328 (2.412-7.766) 5.633 (3.288-9.649) 25.283 (16.942-37.730) 0.952 (0.877-1.094)

1 1.742 (0.960-3.163) 4.277 (2.090-8.752) 4.299 (2.158-8.563) 17.907 (11.360-28.228) 0.949 (0.862-1.043)

1 1.156 (0.868-1.540)

1 1.044 (0.678-1.610)

1

1

1.267 (0.895-1.792)

1.243 (0.808-1.914)

1.457 (1.002-2.120)

1.509 (0.981-2.320) 1.045 (0.635-1.721) 1 2.468 (1.495-4.074) 1.888 (1.233-2.890) 4.378 (1.357-14.120) 1 1.572 (0.853-2.898) 1.032 (0.535-1.991) 1 1.000 (0.998-1.002)

*Adjusted for age, sex, and ethnicity. †Adjusted for age, sex, ethnicity, PIR, carbohydrates, physical activity, and CV fitness.

DISCUSSION Adolescents with BMI percentile of 75 to 84.9, which represents approximately 1.2 million US adolescents, have high HOMA-IR levels. Although several previous studies have identified an increase in insulin resistance in the overweight and risk for overweight adolescent populations,9,13,14 this is the first study to show high HOMA-IR levels in the BMI percentile of 75 to 84.9, a group considered to be “normal” weight. Adolescents with insulin resistance tend to have higher total cholesterol with lower high-density lipoprotein, higher fasting triglycerides, and higher systolic blood pressure.9 This study demonstrated that more than 8.5 million US adolescents have high HOMA-IR levels. Many of these persons may not be identified based on our current BMI categorizations because more than 3.2 million adolescents who are considered to be of normal weight have high HOMA-IR levels. Of these adolescents, more than 1.2 million are in the BMI percentile group of 75 to 84.9. In fact, the proportion of adolescents with high HOMA-IR levels in this BMI group is similar to the proportion in the risk for overweight group. Adolescents in the BMI percentile range of 75 to 84.9 have conventionally been considered “normal weight.” Our find278

Rodden et al

ings suggest that the prevalence of high HOMA-IR levels in this “normal weight” group is similar to that of those considered at risk for overweight. Thus, a BMI ⱖ75th percentile is associated with high HOMA-IR levels, and these adolescents may benefit from counseling on diet and exercise. Other factors such as diet, exercise, and physical activity were controlled for in the analysis because they are known to be related to insulin resistance. The percentage of energy intake from carbohydrates has been noted to be related to increased insulin resistance in an adult population.12 Cardiovascular (CV) fitness, a measure of a person’s maximal oxygen consumption during exercise and physical activity may be related to insulin resistance.15 A third of adolescents have low CV fitness levels.16 This is associated with an increased prevalence of cardiovascular disease risk factors.16,17 Similarly, a sedentary lifestyle, a factor distinct from CV fitness, is associated with obesity, diabetes, and cardiovascular disease, and it is very prevalent in adolescents.18-20 A BMI ⬎75th percentile is associated with higher HOMA-IR. Identifying persons with insulin resistance at an early age may allow for preventive measures to be discussed and implemented. Lifestyle interventions have been attempted in adults with insulin resistance and have reduced the The Journal of Pediatrics • September 2007

incidence of diabetes compared with placebo treatment.21,22 In the Diabetes Prevention Program Trial comparing placebo, metformin, and lifestyle changes, the people in the lifestyle changes arm had a reduced incidence of diabetes.21 Recently a randomized, controlled trial of metformin demonstrated improvement in BMI and fasting insulin in overweight adolescents.23 If people at risk for disease could be identified earlier in life, these lifestyle and medical interventions might prevent diabetes, heart disease, strokes, and even cancer. There are limitations to this study. First of all, there is still no universally accepted definition for insulin resistance in adolescents. However, this study uses HOMA-IR, which has been identified in previous studies as more reliable than other non-invasive tests in the adolescent population.11 Several laboratory modalities and values have been noted that use non-invasive measurement techniques to identify this intermediary state known as insulin resistance.11,24,25 Compared with other tests such as the quantitative insulin sensitivity check index (QUICKI) and fasting glucose/insulin ratio (FGIR), HOMA-IR has been found to be more reliable in adolescents.11 Second, physical activity was assessed based on self-report. Because it might be expected that respondents overreport physical activity, lower levels of physical activity may still be beneficial. Third, insulin resistance may be a transient process during puberty because of hormonal changes. Tanner stages were not collected in the 1999-2002 NHANES, so this could not be controlled for in the analysis. HOMA-IR values tend to peak at 13 years of age in girls and 14 years in boys.26 To adjust for these peaks, the analysis was performed with age categorized as 12 to 14 and 15 to 19 years old, and it did not affect the results (analysis not included). This study suggests that adolescents with a BMI percentile of 75 to 84.9 should be considered a group that includes persons who may need counseling about diet and exercise similar to those with a BMI ⱖ85th percentile. Further studies are needed to develop effective lifestyle or medical interventions for this group.

REFERENCES 1. Ogden CL, Carroll MD, Curtin LR, McDowell MA, Tabak CJ, Flegal KM. Prevalence of Overweight and Obesity in the United States, 1999-2004. JAMA 2006;295:1549-55. 2. Jolliffe D. Extent of overweight among US children and adolescents from 1971 to 2000. Int J Obes 2004;28:4-9. 3. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM. Prevalence of overweight and obesity among us children, adolescents, and adults, 1999-2002. JAMA 2004;291:2847-50. 4. Kuczmarski RJ, Ogden CL, Guo SS, Grummer-Strawn LM, Flegal KM, Mei Z, et al. 2000 CDC Growth Charts for the United States: methods and development. Vital & Health Statistics - Series 11: Data from the National Health Survey. 2002;246:1-190.

Insulin Resistance in Adolescents

5. Cruz ML, Shaibi GQ, Weignesberg MJ, Spruijt-Metz D, Ball GDC, Goran MI. Pediatric obesity and insulin resistance: chronic disease risk and implications for treatment and prevention beyond body weight modification. Annu Rev Nutr 2005; 25:435-68. 6. Facchini FS, Hua N, Abbasi F, Reaven GM. Insulin resistance as a predictor of age-related diseases. J Clin Endocrinol Metab 2001;86:3574-8. 7. The STOPP-T2D Prevention Study Group. Presence of diabetes risk factors in a large U.S. eighth-grade cohort. Diabetes Care 2006;29:212-7. 8. Duncan GE. Prevalence of diabetes and impaired fasting glucose levels among US adolescents. Arch Pediatr Adolesc Med 2006;160:523-8. 9. Williams DE, Cadwell BL, Cheng YJ, Cowie CC, Gregg EW, Geiss LS, et al. Prevalence of impaired fasting glucose and its relationship with cardiovascular disease risk factors in US adolescents, 1999-2000. J Pediatr 2005;116:1122-6. 10. National Center for Health Statistics. NHANES Cardiovascular Fitness Procedure Manual. Available at: http://www.cdc.gov/nchs/data/nhanes/cv.pdf. 11. Keskin M, Kurtoglu S, Kendirci M, Atabek ME, Yazici C. Homeostasis model assessment is more reliable than the fasting glucose/insulin ratio and quantitative insulin sensitivity check index for assessing insulin resistant among obese children and adolescents. Pediatrics 2005;115:e500-e503. 12. Diaz VA, Mainous AG, Koopman RJ, Geesey ME. Are ethnic differences in insulin sensitivity explained by variation in carbohydrate intake? Diabetologia 2005; 48:1264-8. 13. Duncan GE, Li SM, Zhou X. Prevalence and trends of a metabolic syndrome phenotype among US adolescents, 1999-2000. Diabetes Care 2004;27:2438-43. 14. Sinha R, Fisch G, Teague B, Tamborlane WV, Banyas B, Allen K, et al. Prevalence of impaired glucose tolerance among children and adolescents with marked obesity. New Eng J Med 2002;346:802-10. 15. Imperatore G, Cheng YJ, Williams DE, Fulton J, Gregg EW. Physical activity, cardiovascular fitness, and insulin sensitivity among U.S adolescents. Diabetes Care 2006;29:1567-71. 16. Carnethon MR, Gulati M, Greenland P. Prevalence and cardiovascular disease correlates of low cardiorespiratory fitness in adolescents and adults. JAMA 2005; 295:2981-8. 17. Kasa-Vubu JZ, Lee CC, Rosenthal A, Singer K, Halter JB. Cardiovascular fitness and exercise as determinants of insulin resistance in postpubertal adolescent females. J Clin Endocrinol Metab 2005;90:849-54. 18. Winer N, Sowers JR. Epidemiology of diabetes. J Clin Pharm 2004;44:397-405. 19. Reilly JJ, Jackson DM, Montgomery C, Kelly LA, Slater C, Paton JY. Total energy expenditure and physical activity in young scottish children: mixed longitudinal study. Lancet 2004;363:211-2. 20. Lowry R, Howell W, Galuska DA, Fulton JE, Kann L. Television viewing and its association with overweight, sedentary lifestyle, and insufficient consumption of fruits and vegetables among US high school students: differences by race, ethnicity, and gender. J Sch Health 2002;72:413-21. 21. Diabetes Prevention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med 2002;346:393-403. 22. Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen H, IlanneParikka P, et al. Finnish Diabetes Prevention Study Group. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med 2001;344:1343-50. 23. Srinivasan S, Ambler GR, Baur LA, Garnett SP, Tepsa M, Yap F, et al. Randomized, controlled trial of metformin for obesity and insulin resistance in children and adolescents: improvement in body composition and fasting insulin. J Clin Endocrinol Metab 2006;91:2074-80. 24. Gungor N, Saad R, Janosky J, Arslanian S. Validation of surrogate estimated of insulin sensitivity and insulin secretion in children and adolescents. J Pediatr 2004; 144:47-55. 25. Conwell LS, Trost SG, Brown WJ, Batch JA. Indexes of insulin resistance and secretion in obese children and adolescents. Diabetes Care 27:2004;314-9. 26. Lee JM, Okumura MJ, Davis MM, Herman WH, Gurney JG. Prevalence and determinants of insulin resistance among U.S. adolescents. Diabetes Care 2006; 29:2427-32.

279

A Cognitive Behavioral Therapy Program for Overweight Children ERICA L. T.

VAN DEN

AKKER, MD,* PATRYCJA J. PUIMAN, MD,* MIEKE GROEN, MSC, REINIER TIMMAN, PHD, MIEKE T.M. JONGEJAN, MD, PHD, AND WIM TRIJSBURG, PHD†

Objective To assess the 1-year results of a multidisciplinary, cognitive behavioral therapy treatment program for overweight and obese children. Study design Children (n ⴝ 73; 8 to 15 years old) participated in a prospective study aimed at reduction of the body mass index–standard deviation score (BMI-SDS), adapting a healthy lifestyle and creating a positive self-image and higher selfesteem, by use of a group approach and parental involvement. Reduction in BMI-SDS and percent overweight were measured and analyzed by use of MIXED modeling. Results The participants achieved a 0.6 BMI-SDS reduction, comparable to a weight loss of 18.7% after 1 year (P < .0001). The proportion of dropouts was 33%. Compared with the follow-up group, dropouts were older, increased in BMI-SDS before start of treatment, and were less successful in BMI-SDS reduction during treatment. Conclusions This treatment program had a positive effect on BMI-SDS in overweight and obese children at 1-year follow-up. Differences between the characteristics of the dropout and follow-up group may reflect predictor variables for treatment outcome. (J Pediatr 2007;151:280-3) orldwide the growing incidence of obesity and overweight in children is a reason for concern.1-3 Obese children are likely to become overweight or obese adults, with an increased risk of diabetes mellitus type 2, cardiovascular disease and death.4-7 Even more alarming is the rising prevalence of abnormal glucose tolerance, dyslipidemia, and hypertension among obese children.8,9 One of the key elements in the battle against obesity is early treatment to prevent comorbidity and risk of becoming obese adults. Obesity treatment aims at lifestyle changes and has an overall low success rate.10 From previous studies we have learned that treatment programs for overweight and for From the Department of Pediatrics, Erasobese adults are less successful than those designed for children.11 This is not surprising, mus University Medical Center-Sophia taking into account that eating habits and exercise patterns are developed at a young age. Children’s Hospital, Rotterdam, the Netherlands (E.L.T.A., P.J.P.); Department of PeTherefore lifestyle changes are easier to achieve when sedentary behavior and unhealthy diatrics, Sint Franciscus Hospital, Rotterdietary habits have not yet fully developed. Cognitive behavioral therapy in children is dam, the Netherlands (E.L.T.A., P.J.P., 10,12 reported as one of the most successful methods for reducing overweight and obesity. M.T.M.J.); Department of Medical Psychology, Sint Franciscus Hospital, Rotterdam, Nonetheless, when treating obese children, family engagement and group therapy appear the Netherlands (M.G.); Department of to positively influence success rates.13-15 Medical Psychology and Psychotherapy, Erasmus University Medical Center, RotterSince 1995, the pediatric department of the Sint Franciscus Hospital in Rotterdam, dam, the Netherlands (R.T., W.T.). the Netherlands, has had a cognitive behavioral outpatient treatment program for overSubmitted for publication Sep 27, 2006; last weight and obese children, in which parental involvement and a group approach are revision received Feb 1, 2007; accepted Mar 19, 2007. imbedded. This program is called the “Dikke Vrienden Club” (DVC), which translates Reprint requests: Erica L. T. van den Akinto “Big Friends Club,” with a pun on “close” friends intended. Both overweight and ker, MD, ErasmusMC–Sophia Children’s obese children between 8 and 15 years of age can take part in this program. The DVC Hospital, Pediatrics–Endocrinology, Dr. Molenwaterplein 60, PO Box 2060, 3000 aims at reduction of body mass index (BMI), adapting a healthy lifestyle and creating a CBRotterdam, Netherlands. E-mail: e.l.t. positive self-image. In this article we present 1-year follow-up results of this program. [email protected].

W

METHODS Study Population All children who participate in the DVC program are between 8 to 15 years old and are overweight or obese, defined as a BMI standard deviation score (BMI-SDS) of more BMI BMI-SDS

280

Body mass index Body mass index–standard deviation score

DVC

Dikke Vrienden Club (Big Friends Club)

*The first two authors contributed equally to the article. † This article is dedicated to Professor Doctor Wim Trijsburg, for his outstanding work in the clinical and scientific field of psychology. Professor Trijsburg deceased April 8th 2007 at the age of 58. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.042

than 1.1 and 2.3, respectively. The exclusion criteria were overweight caused by a somatic treatable disorder, mental retardation, behavioral problems defined as a score exceeding 70 on the “child behavioral checklist” (CBCL),16,17 being insufficiently fluent in the Dutch language, and insufficient motivation of parents or the child to actively participate in the program.

Design The DVC focuses on 4 elements: (1) Teaching children about a healthy diet and physical exercise. (2) Coping with psychosocial consequences of overweight and obesity. (3) Creating a positive self-image and higher self esteem. (4) Reducing BMI-SDS, preferably by maintenance of weight during growth. Participants are referred by a physician or may contact the pediatric department themselves. During the first outpatient clinic visit, a pediatrician will conduct an interview and physical examination to exclude somatic disorders that might cause overweight or obesity. In addition, the parents and the child are evaluated for motivation to participate in the treatment. When this is found to be sufficient, they are asked to sign a consent form and declaration of intent. The children are then placed on the 3-month waiting list. DVC groups comprise 8 to 10 children. The program provides for an intake session, 8 children-sessions and 2 parent-sessions during the first 12 weeks. The meetings are led by a team consisting of a psychologist, a dietician, and a physiotherapist. Both at the start and the end of this intensive 12-week period, the children are seen by the DVC-team and a pediatrician. The program uses different behavioral therapy techniques such as operant and cognitive therapy strategies. The first 90 minutes of the children-sessions are dedicated to normal eating and exercise behavior and strategies to deal with difficulties concerning eating or physical inactivity. By providing group therapy, the children can recognize and share their concerns and learn from each other. Gradually their eating and exercise behaviors will change, first by extrinsic reinforcement and later on by intrinsic reinforcement. Furthermore, attention is paid to the psychosocial aspects of obesity like being picked on by peers. The last hour of each session is led by the physiotherapist. By creating positive exercise experiences through games and sports, the children improve their physical condition and exercise behavior. Implementing more exercise in daily life and therefore decreasing sedentary behavior is another goal of this part of the session. The role parents play in their child’s lifestyle is emphasized. Hence, 2 parent-sessions are scheduled. The first takes place before the children’s first session. In this session the DVC team explains the DVC-key elements and the content of the program. During the session the parents learn about a healthy diet, normal exercise behavior, psychosocial aspects of obesity, and the fact that obesity increases the risk of physical and psychological morbidity. Part of the session is devoted to A Cognitive Behavioral Therapy Program for Overweight Children

changing interaction patterns between the parents and their children by teaching them how to support their child instead of controlling them, how to give positive feedback, and how to apply positive reinforcement. The second parent-session takes place 4 weeks later. During this session the parents are invited to ask questions and to share their problems. The DVC team stimulates the parents to search for answers within the group to increase the parent’s problem-solving capacity. After the intensive 12-week period, the children are awarded an A-diploma if they managed to decrease their BMI-SDS. To support each other after the intensive period, the children are paired into age-matched buddy-teams. They find emotional support and help by contacting each other when necessary. The buddy-team–system is a tool to reach the set eating and exercise goals. It enhances self-control and shifts the extrinsic reinforcement provided by the DVC team and the parents to intrinsic reinforcement. At 6 months and 12 months after start of the program, children-parent–sessions aimed at prevention of relapse are organized. The goals are evaluated and the children receive feedback on their diet and exercise pattern. All together, the children, parents and the DVC team try to solve the problems that arose during the past months. The main purpose of these 2 sessions is reinforcement of the key elements of the DVC program. Additionally, B- and C-diplomas are awarded during these sessions for children who retained or further reduced BMI-SDS. In between, the children are seen by a pediatrician 6, 9, and 12 months after the start of the program, and height and weight are measured. At these occasions the children receive individual feedback about their weight change. Children who did not return at the end of the 12-week period, or who did not show up for the follow-up visits were classified as dropouts.

Measurements Data were collected at 4 time points: first outpatient clinic visit (t ⫽ ⫺3 months); intake at start of the DVC program (t ⫽ 0); end of the 12-week intensive program (t ⫽ 3 months); and 1 year after start of the program (t ⫽ 12 months) (Figure). Height was measured in the upright position and defined in centimeters. Weight was measured in kilograms by a digital scale (SECA, Hamburg, Germany). BMI was calculated as bodyweight (kg) divided by height in meters squared (m2). BMI standard deviation score (BMISDS) represents an age- and sex-specific standard deviation. The BMI-SDS was calculated with the software program Growth Analyser version 3.5 (Dutch Growth Foundation, Rotterdam, the Netherlands; www.growthanalyser.org). Percent overweight was calculated using the following equation: 100 ⫻ (actual weight ⫺ 50th percentile weight-for-height)/ 50th percentile for weight for height. The weight-for-height growth charts are based on nationally collected data.18 Statistical Analysis The course of BMI-SDS was analyzed with the PROC MIXED procedure in SAS 8.2 (SAS Institute, Cary, 281

Table I. Characteristics study population

Girls t ⫽ ⫺3 months Age (years) Weight (kg) BMI BMI-SDS t⫽0 Weight (kg) BMI BMI-SDS t ⫽ 3 months Weight (kg) BMI BMI-SDS t ⫽ 12 months Weight (kg) BMI BMI-SDS Percentage overweight (t ⫽ 12 months) ⫺ (t ⫽ 0)

Table II. Final MIXED model parameters of BMI-SDS decrease

Follow-up (SD) n ⴝ 49

Drop-outs (SD) n ⴝ 24

P value

n ⫽ 36 (74%)

n ⫽ 17 (71%)

.65*

10.0 (8-13) 11.4 (8-14) 61.4 (42.4-107.4) 72.4 (54.4-113.8) 26.6 (22.4-36.2) 28.7 (24.1-40.0) 2.6 (1.8-3.4) 2.7 (2.0-3.6)

.001† .002† .01† .25†

63.0 (42.6-111.8) 75.1 (56.4-117.2) 26.8 (21.8-36.8) 29.3 (24.2-41.4) 2.6 (1.7-3.4) 2.7 (2.0-3.7)

.001† .003† .08†

59.4 (36.1-103.9) 73.6 (54.9-113.5) ⬍.001† 24.8 (20.6-34.0) 28.6 (23.7-40.1) ⬍.001† 2.2 (1.3-3.3) 2.6 (1.8-3.6)‡ .002† 61.4 (35.0-111.1) 24.4 (18.7-36.3) 2.0 (0.8-3.3) ⫺18.7%

*Fisher’s exact test, 2-sided. †Student’s t test. ‡n ⫽ 21.

NC).19 This analysis method allows the use of incomplete cases. First, a saturated MIXED model was postulated, including the BMI-SDS as dependent variable, and the main effects time, squared time, age, sex, and the interaction effects between these main effects. Then the covariance structure and random part of the model were determined, by use of the restricted maximum likelihood function. Finally, insignificant interaction effects were removed step by step from the fixed part of the model. The ordinary maximum likelihood function was used to determine the difference between the saturated and the final model. SPSS 12.0.1 (SPSS, Chicago, IL) was used for

Figure. Dropout group versus follow-up group.

282

van den Akker et al

Effect Intercept Sex Age Dropout Time linear Time quadratic Time cubic Dropout䡠time linear Dropout䡠time quadratic

Estimate

P value

3.68 ⫺0.24 ⫺0.067 0.19 0.69 ⫺0.56 0.088 ⫺0.10 0.083

⬍.001 .013 .018 .047 ⬍.0001 ⬍.0001 ⬍.0001 .20 .012

relations between dropout and BMI-SDS increase (Fisher’s exact test). Relation between age and dropout was analyzed with an independent sample t test. P values ⬍ .05 were considered statistically significant.

RESULTS Seventy-three children (53 girls) participated in the DVC from 1999 to 2003 (Table I). Mean age was 10.5 years (range 8.0-14.0 years). Mean BMI at inclusion was 27.3 kg/m2 (range 22.4-40.0) and mean BMI-SDS was 2.6 (range 1.8-3.6). Seventy participants (96%) completed the 12-week program. One year later, 49 (67%) children came back for follow-up; the other 24 (33%) were considered dropouts. Mean BMI-SDS showed a significant reduction of 0.3 BMISDS after the 12-week program (P ⬍ .0001) (Table I, Figure). In the analysis of changes in BMI-SDS, the saturated fixed model (⫺2 log-likelihood ⫽ ⫺7.7) was not significantly better than the final model (⫺2 log-likelihood ⫽ ⫺3.4; ␹2 (5) ⫽ 4.3; P ⫽ .51). The variables of the final MIXED model are presented in Table II. For the follow-up group, those children who attended the 1-year follow-up visit, the treatment resulted in a significant decrease of 0.6 mean BMI-SDS (range ⫺2.0-0.16) (P ⬍ .0001) (Table I, Figure). This decrease is comparable to a mean weight loss of 18.7%. Forty-five children (92%) managed to reduce their BMI-SDS, 33 of whom achieved this reduction by losing weight. The other 12 children gained weight, but the BMI-SDS decrease was accomplished by height gain. Four children in the total follow-up group (8%) had a BMI-SDS elevation with a mean of 0.1 SDS in comparison to the end of the intensive 12-week program. The children who did not attend the 1-year follow-up visit were already less successful during the first 3 months of the DVC program. The mean BMI-SDS for this dropout group increased between the first outpatient clinic visit (t ⫽ ⫺3 months) and the intake session at start of the study (t ⫽ 0) (P ⬍ .001). During the 12-week program, the dropout group decreased less in BMI-SDS, than non-dropouts (Figure). Dropout was also related to age at start, with a higher mean age for dropouts compared to the follow-up group (P ⬍ .01). Analyses showed that 62% of children aged 12 years or The Journal of Pediatrics • September 2007

older dropped out, compared with 21% of younger children. No different effect of the treatment was found for boys and girls. Characteristics of the follow-up and dropout groups are summarized in Table I.

DISCUSSION

1-year follow-up. Further follow-up studies are needed to see what perspective the DVC program offers on long-term weight control. In addition, further identification of predictor variables could lead to better and more individually based recommendations on different treatment programs, and therefore increase the chances of success.

Overall, the overweight and obese children participating in the DVC program achieved a mean BMI-SDS reduction of 0.6 at 1-year follow-up. When converted to percent overweight, the BMI-SDS reduction results in an 18.7% loss of percent overweight. This means that a shift from a mainly obese group to a predominantly overweight group had occurred. The 18.7% reduction is higher than that found in most other studies. Three other studies using an approach similar to ours reported 12-month success rates ranging from 5.8% to 13.1% weight reduction.13,20 One of these also used the BMI-SDS and found a decrease of 0.38 BMI-SDS.21 In this study, after the intensive 12-week program, the decline in overweight persisted until 1-year follow-up. Other studies describing long-term follow-up all show BMI-SDS reduction during the intensive treatment program, yet this is not always maintained during follow-up.10,14,15,20,21 Only 3 participants in our study dropped out during the intensive 12-week program. However, the 33% dropout rate after 1 year shows that the full program might not suit all children. Only 1 other study with a 1-year follow-up reported a dropout rate of 22%.21 In our study, the dropouts were older than children in the follow-up group. Therefore the DVC program might be more effective in younger children. Another study also reported success rates that were higher for children under 14 years of age compared with older children.13 In contrast, a 2-year follow-up study reported better results for older and more overweight children.22 We identified 2 other predictors for dropout: mean BMI-SDS increasing during the waiting list period and achieving less reduction in mean BMI-SDS during the intensive 12-week treatment period. On the basis of these findings, we suggest that the characteristics of the dropout group in the DVC program may reflect predictor variables for treatment outcome. Further identification of predictor variables could be helpful in designing treatment programs focused on children’s individual needs. Some subgroups might benefit more from an individual or more intensive program. Comparison of our findings to those from other studies is complicated by several factors. First, definitions of overweight and obesity lack uniformity. Use of the BMI-SDS as a standard international definition for childhood overweight and obesity, as proposed by Cole et al,23 will enable comparison of study results. Second, study results do not always permit conclusions about dropout rates or rates of successfully treated children. This study was neither randomized nor controlled. However, from our waiting list data and from patient-controlled studies performed by others, we know that BMI-SDS reduction is not achieved without treatment.13,21,24 In conclusion, children who participated in this cognitive behavioral therapy program declined in overweight at

1. Obesity: Preventing and Managing the Global Epidemic. Geneva, Switzerland: World Health Organization; 2000. 2. James PT, Leach R, Kalamara E, Shayeghi M. The worldwide obesity epidemic. Obes Res 2001;9(Suppl 4):228S-33S. 3. Hedley AA, Ogden CL, Johnson CL, Carroll MD, Curtin LR, Flegal KM. Prevalence of overweight and obesity among US children, adolescents, and adults, 1999-2002. JAMA 2004;291:2847-50. 4. Dietz WH. Childhood weight affects adult morbidity and mortality. J Nutr 1998;128(Suppl):411S-4S. 5. Must A, Strauss RS. Risks and consequences of childhood and adolescent obesity. Int J Obes Relat Metab Disord 1999;23(Suppl 2):S2-11. 6. Togashi K, Masuda H, Rankinen T, Tanaka S, Bouchard C, Kamiya H. A 12-year follow-up study of treated obese children in Japan. Int J Obes Relat Metab Disord 2002;26:770-7. 7. Mossberg HO. 40-year follow-up of overweight children. Lancet 1989; 2(8661):491-3. 8. Weiss R, Dziura J, Burgert TS, Tamborlane WV, Taksali SE, Yeckel CW, et al. Obesity and the metabolic syndrome in children and adolescents. N Engl J Med 2004;350:2362-74. 9. Freedman DS, Dietz WH, Srinivasan SR, Berenson GS. The relation of overweight to cardiovascular risk factors among children and adolescents: the Bogalusa Heart Study. Pediatrics 1999;103(Pt 1):1175-82. 10. Summerbell CD, Ashton V, Campbell KJ, Edmunds L, Kelly S, Waters E. Interventions for treating obesity in children. Cochrane Database Syst Rev 2003;(3):CD001872. 11. Epstein LH, Valoski AM, Kalarchian MA, McCurley J. Do children lose and maintain weight easier than adults: a comparison of child and parent weight changes from six months to ten years. Obes Res 1995;3:411-7. 12. Whitlock EP, Williams SB, Gold R, Smith PR, Shipman SA. Screening and interventions for childhood overweight: a summary of evidence for the US Preventive Services Task Force. Pediatrics 2005;116:e125-44. 13. Braet C, Van Winckel M, Van Leeuwen K. Follow-up results of different treatment programs for obese children. Acta Paediatr 1997;86:397-402. 14. Epstein LH, McCurley J, Wing RR, Valoski A. Five-year follow-up of family-based behavioral treatments for childhood obesity. J Consult Clin Psychol 1990;58:661-4. 15. Epstein LH, Valoski A, Wing RR, McCurley J. Ten-year follow-up of behavioral, family-based treatment for obese children. JAMA 1990;264:2519-23. 16. Ferdinand RF, Visser JH, Hoogerheide KN, van der Ende J, Kasius MC, Koot HM, et al. Improving estimation of the prognosis of childhood psychopathology; combination of DSM-III-R/DISC diagnoses and CBCL scores. J Child Psychol Psychiatry 2004;45:599-608. 17. Krol NP, De Bruyn EE, Coolen JC, van Aarle EJ. From CBCL to DSM: a comparison of two methods to screen for DSM-IV diagnoses using CBCL data. J Clin Child Adolesc Psychol 2006;35:127-35. 18. Fredriks AM, van Buuren S, Burgmeijer RJ, Meulmeester JF, Beuker RJ, Brugman E, et al. Continuing positive secular growth change in The Netherlands 1955-1997. Pediatr Res 2000;47:316-23. 19. Verbeke G, Molenberghs G, eds. Linear mixed models in practice—A SAS oriented approach. New York: Springer; 1997. 20. Israel AC, Guile CA, Baker JE, Silverman WK. An evaluation of enhanced self-regulation training in the treatment of childhood obesity. J Pediatr Psychol 1994;19:737-49. 21. Reinehr T, Kersting M, Alexy U, Andler W. Long-term follow-up of overweight children: after training, after a single consultation session, and without treatment. J Pediatr Gastroenterol Nutr 2003;37:72-4. 22. Nuutinen O, Knip M. Long-term weight control in obese children: persistence of treatment outcome and metabolic changes. Int J Obes Relat Metab Disord 1992;16:279-87. 23. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000;320(7244):1240-3. 24. Johnson WG, Hinkle LK, Carr RE, Anderson DA, Lemmon CR, Engler LB, et al. Dietary and exercise interventions for juvenile obesity: long-term effect of behavioral and public health models. Obes Res 1997;5:257-61.

A Cognitive Behavioral Therapy Program for Overweight Children

283

REFERENCES

Socioeconomic Position, Maternal IQ, Home Environment, and Cognitive Development SHILU TONG, PHD, PETER BAGHURST, PHD, GRAHAM VIMPANI, PHD,

AND

ANTHONY MCMICHAEL, PHD

Objective To assess whether socioeconomic position, maternal intelligence (IQ), and the home environment are interrelated to cognitive development in childhood. Study design Prospective cohort study (n ⴝ 723) with cognitive tests at ages 2, 4, 7, and 11 to 13 years. Results There were statistically significant positive associations of father’s occupational prestige, Home Observation for Measurement of Environment (HOME) score, and maternal IQ with cognitive performance in childhood. After adjustment for confounding factors, there was an increase in cognitive development by 0.8 to 2.0, 2.9 to 4.8, and 4.2 to 9.0 points for a 10-unit increment in father’s occupational prestige, maternal IQ, and HOME score, respectively. Conclusions These results demonstrate that socioeconomic position, maternal IQ, and the home environment are independently and positively predictive of children’s cognitive development. These findings provide additional rationale for implementing social policies that reduce socioeconomic inequalities. (J Pediatr 2007;151:284-8) hildhood intelligence (IQ) is a predictor of various health outcomes in adulthood (eg, cardiovascular disease, some cancers, diabetes, suicide, motor vehicle injuries, and premature deaths).1-6 Hence, a better understanding of the determinants of child IQ should assist in the reduction of lifetime risks to health. Several explanations for the association between IQ and adult health have been advanced,7 including that IQ is 1) a predictor of health-favoring social circumstances in later life (eg, high educational attainment and high job status); 2) a cumulative index of psychological and physiological insults (eg, birth complications, suboptimal postnatal care, and illness); 3) an intrinsic indicator of general body integrity (as measured via the brain’s capacity to process information rapidly, correctly, and reliably); 4) a proxy for stress management skills—people with higher intelligence scores may be less likely to place themselves in stressful environments or cope better if they do; or 5) an asset for optimal interpretation From the School of Public Health Queensof health prevention messages— children who scored highly in intelligence tests were land University of Technology, Kelvin Grove, Australia (S.T.); the Public Health more likely to give up smoking in adult life.8 Research Unit, Children, Youth, and WomThe relative roles of socioenvironmental and genetic factors in cognitive developen’s Health Service, and Disciplines of Paediatrics and Public Health, University of Adment remain unresolved. Although a proportion of the variation in human intelligence elaide, Adelaide, Australia (P.B.); The may be attributable to genetic factors,9 socioenvironmental factors are important deterDiscipline of Paediatrics and Child Health, minants of childhood IQ and are modifiable through a range of early-years intervention University of Newcastle, Newcastle, Australia (G.V.); and the National Centre for strategies such as early childhood education and care.10,11 A number of studies have Epidemiology and Population Health, Ausexamined the link between socioeconomic characteristics and cognitive development.12-14 tralian National University, Canberra, Australia (A.M.). In general, children in disadvantaged families manifest poorer cognitive performance than Supported by a series of grants from the do those in better-off families. However, few studies have adjusted satisfactorily for National Health and Medical Research parental and environmental factors that may confound the relation between socioecoCouncil and the Channel 7 Children’s Research Foundation; Dr Shilu Tong is supnomic backgrounds and cognitive development. For example, children in disadvantaged ported by an NHMRC Research Fellowfamilies may be more exposed to environmental neurotoxins such as lead and other ship. 15-21 chemicals and may have poorer quality of home environment. Submitted for publication Sep 16, 2006; last revision received Jan 16, 2007; accepted In this study, we examined the association of socioeconomic position, maternal IQ, Mar 15, 2007. and the home environment with cognitive development during childhood, using data Reprint requests: Dr Shilu Tong, School of from a prospective cohort study design that enabled adjustment for a wide range of Public Health, Queensland University of Technology, Kelvin Grove, Qld. 4059, Ausconfounding factors.

C

HOME

284

Home Observation for Measurement of Environment

IQ MDI

Intelligence Quotient Mental Development Index

tralia. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.020

METHODS The primary objective of the Port Pirie Cohort Study was to examine the relation between exposure to environmental lead and child development. The children living in and around the lead-smelting town of Port Pirie, South Australia, were followed from birth to age 11 to 13 years. Details of the research design have been reported elsewhere.17-21

Sample Of the 723 live births who were originally recruited in the Port Pirie Cohort Study, 601, 548, 494, and 375 were followed and assessed, respectively, at ages 2, 4, 7, and 11 to 13 years. The children who were evaluated at each age differed little from those lost to follow-up on most characteristics, including sociodemographic, environmental, and biomedical factors, except that they lived in families with slightly higher social class than those lost to followup.17-21 For example, possible loss to follow-up bias was evaluated at age 11 to 13 years. The results show that 55 children lost to follow-up were in families that either left the Port Pirie district or could not be contacted despite intensive efforts (ie, 11.7% of the base population), although a small number of families (7.8%) simply discontinued their participation.19 Measurements of Socioeconomic Characteristics A crude measure of children’s socioeconomic position often used in Australia is the Daniel Scale,22 which is based on the hierarchy of the father’s occupational prestige and is used as an approximate indicator of family income. The Daniel score is inversely related to occupational prestige, that is, the higher the Daniel score, the lower the prestige. For example, a manufacturing worker was scored much higher than a professional occupation (eg, manager or engineer). The Daniel score of each family at birth was used in this study, as some evidence suggests that the father’s social class at birth is an important predictor of child’s IQ.13,23 Two important predictors of children’s abilities are the Home Observation for Measurement of the Environment (HOME) inventory24 and maternal IQ. The HOME inventory was used to assess each child’s caregiving environment, when each child was 3 years of age. Maternal IQ was measured with the Wechsler Adult Intelligence Scale-Revised,25 whereas the children were in the age range of 3 to 5 years. Maternal IQ rather than maternal education was used in this study because the former combines both inherited and acquired cognitive capacity and is a better integrated predictor of children’s IQ than is the latter. Measurement of Cognitive Function Each child’s abilities were assessed by using the Bayley Scales of Infant Development at age 2 years,26 the McCarthy Scales of Children’s Abilities at age 4 years,27 and the revised version of the Wechsler Intelligence Scale for Children

(WISC-R) at ages 7 and 11 to 13 years.28 The Bayley Scales of Infant Development, which comprises two standardized scores, the Mental Development Index (MDI) and Psychomotor Development Index, is suitable for children ages 30 months and younger. Only MDI scores were used in this study because MDI score primarily reflects cognitive abilities. The McCarthy Scales of Children’s Abilities, which are applicable to children 3 to 7 years old, comprises five scales: verbal, perceptual performance, quantitative, memory, and motor. The first three of those scales combined form the general cognitive index. The WISC-R was used to assess the cognitive function of each child at ages 7 and 11 to 13 years. The WISC-R is a test of general intelligence developed for use with children ages 6 to 16 years. All children were assessed at each age by a research psychologist who was unaware of children’s socioeconomic characteristics.

Confounders Information was also collected on the following potential confounding variables: child’s sex, birth weight, head circumference, birth length, Apgar score at 5 minutes after birth, presence of neonatal jaundice, maternal age, duration of gestation, maternal smoking and drinking habits, parental marital status, and lifetime average blood lead concentration up to age 2 years. Birth weight was measured in the clinical setting and recorded in grams. Duration of gestation was calculated from the date of the last menstrual period reported by the mother and checked against medical records. The child’s head circumference and birth length were also measured immediately after delivery. To assess early-life exposure to environmental lead, sequential blood samples were collected from the pregnant women, the umbilical cord, at 6, 12, and 24 months, and then periodically from the young child. The lifetime average blood lead concentration up to 2 years was used in this study because we had previously found that lead exposure in the first 2 years of life is most critical to cognitive development.17-21 Analysis We first examined univariate relations between socioeconomic characteristics and cognitive development. All available data were retrieved for the 601 children whose cognitive function was initially assessed at age 2 years. Analysis of variance was used to test for linear trend across groups defined on the basis of the quintiles, separately for each of three indices: the Daniel Scale, the HOME score, and maternal IQ. Covariates that might confound the association between socioeconomic characteristics and cognitive function were explored. Multiple linear regression model was then used to examine the association of occupational prestige, maternal IQ, and home environment, with cognitive development as measured at ages 2, 4, 7, and 11 to 13 years, after adjustment for a range of potential confounding factors.

Socioeconomic Position, Maternal IQ, Home Environment, and Cognitive Development

285

HOME score with children’s cognitive development across childhood.

Figure 1. Socioeconomic characteristics and children’s IQ at ages 2, 4, 7, and 11 to 13 years (quintile 1 represents the lowest father’s occupational prestige, HOME scores, and maternal IQ; quintile 5 represents the highest father’s occupational prestige, HOME scores, and maternal IQ). Quintile 1; Quintile 2; Quintile 3; Quintile 4; Quintile 5.

RESULTS Overview of the Predictors of Cognitive Development The mean value of cognitive scores at ages 2, 4, 7, and 11 to 13 years was 109.2, 107.1, 104.7, and 100.0, respectively. The father’s occupational prestige, HOME score, and maternal IQ were clearly associated with cognitive function at all ages examined (P ⱕ .01), and there appeared to be a consistent dose-response relation (Figure 1). Daniel Scale and Cognitive Development Table I (available at www.jpeds.com) presents a statistically significant positive association between father’s occupational prestige and children’s cognitive function at ages 2, 4, 7, and 11 to 13 years in simple (unadjusted) regression analyses. Children whose father’s occupational prestige was low had poorer cognitive function than those whose father’s occupational prestige was higher. The strength of the association decreased slightly after adjustment for a range of putative confounders (model I). The association between father’s occupational prestige and children’s cognitive function was attenuated by adjustment for maternal IQ (model II) and quality of home environment (model III). However, the relation remained statistically significant or marginally significant. Maternal IQ, Quality of Home Environment, and Cognitive Development Similar patterns were observed for maternal IQ (Table II; available at www.jpeds.com) and quality of home environment (Table III; available at www.jpeds.com). Statistically significant and positive associations were observed for those two indices across different ages in simple (unadjusted) regression analyses. Children whose mother’s IQ or quality of home environment (HOME score) was low had poorer cognitive function than those whose mother’s IQ or HOME score was higher. The magnitude of the association decreased slightly after adjustment for a set of confounders (models I through III). However, after that adjustment there remained a statistically significant association of mother’s IQ and 286

Tong et al

Socioeconomic Position, Maternal IQ, Quality of Home Environment, and Cognitive Development: A Comparison Further analyses display a consistent relation between these predictors and cognitive development after adjustment for confounding factors in the final model. For example, for every 10-unit increase in father’s occupational prestige, children’s cognitive performance increased by 0.8 (95% CI: ⫺0.7 to 2.3), 1.5 (95% CI: 0.1 to 2.9), 2.0 (95% CI: 0.5 to 3.4), and 1.1 (95% CI: ⫺0.4 to 2.6) points at ages 2, 4, 7, and 11 to 13 years, respectively. There was a similar pattern for the effect of maternal IQ on cognitive performance by age. For every 10-unit increase in HOME scores, children’s cognitive performance improved by 9.0 (95% CI: 5.4 to 12.7), 7.7 (95% CI: 4.2 to 11.3), 4.2 (0.7 to 7.6), and 6.2 (95% CI: 2.5 to 9.9) points at ages 2, 4, 7, and 11 to 13 years, respectively (Figure 2).

DISCUSSION In this cohort of children, socioeconomic position, maternal IQ, and the quality of home environment were consistently associated with cognitive development, even after adjustment for a wide range of confounders. Our results indicate that the three measures of socioeconomic characteristics have an independent impact on childhood cognitive development. In general, the higher the occupational prestige and maternal IQ and the better the home environment, the higher the children’s cognitive function. We also found that the association between socioeconomic position and cognitive development was markedly attenuated by adjustment for maternal IQ and quality of home environment. It suggests that it is important to take these factors into account in the assessment of the association between socioeconomic position and childhood intelligence. The results of this study also indicate that other parental factors (eg, parental smoking behavior) and environmental variables (eg, lead exposure) are unlikely to explain the association between socioeconomic characteristics and cognitive development. Further, there were different patterns for the effects of HOME score versus maternal IQ or father’s occupational prestige on cognitive performance (Figure 2). This suggests that the quality of home environment may influence cognitive development through different mechanisms and/or pathways to maternal IQ or father’s occupational prestige. In particular, it appears that the quality of home environment has maximal impact in early childhood, and the other two variables have maximal impact in later childhood. Several studies have examined the association between socioeconomic status and cognitive function. Kaplan et al12 undertook a population-based study of 496 Finnish men ages 58 and 64 for whom there were data on parent’s socioeconomic position, their own education level, and performance on neuropsychological tests. They reported childhood socioThe Journal of Pediatrics • September 2007

Figure 2. Adjusted relations of socioeconomic factors to cognitive development by age.

economic position to be positively associated with cognitive function in adulthood. Jefferis et al13 examined the combined effect of social class and birth weight on cognitive development and found that the postnatal socioeconomic environment has a substantial impact on cognitive function through to early adulthood. More recently, Lawlor et al23 reported that father’s social class at the time of birth was an important predictor of childhood intelligence, even after adjustment for maternal characteristics and perinatal and childhood factors. Our findings are essentially consistent with the findings of these studies. Although both maternal IQ and quality of home environment are important determinants of children’s cognitive function,29-31 there are few empirical data on how these factors inter-relate to each other. The results of this study demonstrate that maternal IQ and quality of home environment tend to have independent impacts on children’s cognitive development. There are three major strengths in this study. We systematically examined the inter-relations between socioeconomic characteristics and cognitive development after adjustment for a wide range of confounding factors (including parental smoking and lead exposure). Second, a relatively homogenous community-based sample was used. For example, all children involved in the study were Caucasian, and therefore the assessment is unlikely to be confounded by cultural factors. Finally, internationally standardized tests of cognitive function were used at various ages, and stringent quality control procedures were implemented in this study.17-21 This study has three possible limitations. First, follow-up bias could have arisen if the children who left the study differ—in the relation that they display between their socioeconomic characteristics and cognitive development—from those who remained in the study. However, any such bias is unlikely to be

substantial because our analyses showed that the children who were evaluated at ages 11 to 13 years did not differ significantly from those lost to follow-up on most characteristics, including sociodemographic, environmental, and biomedical factors, except that fathers of the children remaining in the cohort had slightly higher occupational prestige than those lost to followup.17-21 Therefore, to maximize available information, we used the total dataset available at each age (2, 4, 7, and 11 to 13 years); the longitudinal analysis of only those subjects who contributed data at all four ages revealed a similar pattern. Second, there is no recognized gold standard for measuring socioeconomic characteristics. We used the Daniel scale, the HOME score, and maternal IQ to measure child’s socioeconomic characteristics. Nevertheless, these indices may not fully reflect all potentially relevant childhood socioeconomic background, and residual confounding (eg, by paternal IQ) may exist. Finally, this study is unavoidably unable to distinguish between a genetic and an environmental effect of maternal IQ on cognitive development. The findings of this study have two important pubic health implications. First, socioeconomic attributes appear to be independently and positively predictive of children’s cognitive function. The association between socioeconomic characteristics and cognitive development is unlikely to be explained by other parental factors (eg, parental smoking behaviour) and environmental variables (eg, lead exposure). Second, socioeconomic characteristics may directly or indirectly contribute to the documented relation between childhood intelligence and adult morbidity and mortality as they are major determinants of cognitive development. The authors particularly wish to thank the families who participated in this study.

Socioeconomic Position, Maternal IQ, Home Environment, and Cognitive Development

287

REFERENCES 1. Gunnell D, Magnusson PK, Rasmussen F. Low intelligence test scores in 18 year old men and risk of suicide: cohort study. BMJ 2005;330:167. 2. Batty GD, Mortensen EL, Nybo Andersen AM, Osler M. Childhood intelligence in relation to adult coronary heart disease and stroke risk: evidence from a Danish birth cohort study. Paediatr Perinat Epidemiol 2005;19:452-9. 3. Hart CL, Taylor MD, Smith GD, Whalley LJ, Starr JM, Hole DJ, et al. Childhood IQ and all-cause mortality before and after age 65: prospective observational study linking the Scottish Mental Survey 1932 and the Midspan studies. Br J Health Psychol 2005;10:153-65. 4. Hart CL, Taylor MD, Smith GD, Whalley LJ, Starr JM, Hole DJ, et al. Childhood IQ and cardiovascular disease in adulthood: prospective observational study linking the Scottish Mental Survey 1932 and the Midspan studies. Soc Sci Med 2004;59:2131-8. 5. Martin LT, Fitzmaurice GM, Kindlon DJ, Buka SL. Cognitive performance in childhood and early adult illness: a prospective cohort study. J Epidemiol Community Health 2004;58:674-9. 6. Hart CL, Taylor MD, Davey Smith G, Whalley LJ, Starr JM, Hole DJ, et al. Childhood IQ, social class, deprivation, and their relationships with mortality and morbidity risk in later life: prospective observational study linking the Scottish Mental Survey 1932 and the Midspan studies. Psychosom Med 2003;65:877-83. 7. Whalley LJ, Deary IJ. Longitudinal cohort study of childhood IQ and survival up to age 76. BMJ 2001;322:819-21. 8. Taylor MD, Hart CL, Davey Smith G, Starr JM, Hole DJ, Whalley LJ, et al. Childhood mental ability and smoking cessation in adulthood: prospective observational study linking the Scottish mental survey 1932 and the Midspan studies. J Epidemiol Community Health 2003;57:464-5. 9. Posthuma D, Luciano M, Geus EJ, Wright MJ, Slagboom PE, et al. A genomewide scan for intelligence identifies quantitative trait loci on 2q and 6p. Am J Hum Genet 2005;77:318-26. 10. Hatcher PJ, Goetz K, Snowling MJ, Hulme C, Gibbs S, Smith G. Evidence for the effectiveness of the Early Literacy Support programme. Br J Educ Psychol 2006;76:351-67. 11. Tong S, Baghurst P, McMichael A. Birthweight and cognitive development during childhood. J Paediatr Child Health 2006;42:98-103. 12. Kaplan GA, Turrell G, Lynch JW, Everson SA, Helkala EL, Salonen JT. Childhood socioeconomic position and cognitive function in adulthood. Int J Epidemiol 2001;30:256-63. 13. Jefferis BJ, Power C, Hertzman C. Birth weight, childhood socioeconomic environment, and cognitive development in the 1958 British birth cohort study. BMJ 2002;325:305-14.

288

Tong et al

14. Lee S, Kawachi I, Berkman LF, Grodstein F. Education, other socioeconomic indicators, and cognitive function. Am J Epidemiol 2003;157:712-20. 15. Ascioglu M, Dolu N, Golgeli A, Suer C, Ozesmi C. Effects of cigarette smoking on cognitive processing. Int J Neurosci 2004;114:381-90. 16. Tong S, McMichael AJ. Maternal smoking and neuropsychological development in childhood: a review of evidence. Dev Med Child Neurol 1992;34:191-7. 17. McMichael AJ, Baghurst PA, Wigg NR, Vimpany GV, Robertson EF, Roberts RJ. Port Pirie Cohort Study: environmental exposure to lead and children’s abilities at the age of four years. N Engl J Med 1988;319:468-75. 18. Baghurst PA, McMichael AJ, Wigg NR, Vimpany GV, Robertson EF, Roberts RJ, Tong SL. Environmental exposure to lead and children’s intelligence at the age of seven years: the Port Pirie Cohort Study. N Engl J Med 1992;327:1279-84. 19. Tong SL, Baghurst PA, McMichael AJ, Sawyer M, Burns JM. Lifetime exposure to environmental lead and children’s intelligence at ages 11-13 years. BMJ 1996;312:1569-75. 20. Tong SL, Baghurst P, Sawyer M, Burns JM, McMichael AJ. Declining blood lead levels and changes in cognitive function during childhood: the Port Pirie Cohort Study. JAMA 1998;280:1915-9. 21. Burns JM, Baghurst P, Sawyer M, McMichael AJ, Tong SL. Lifetime low-level exposure to environmental lead and children’s emotional and behavioural development at ages 11-13 years: the Port Pirie Cohort Study. Am J Epidemiol 1999;149:740-9. 22. Daniel A. The measurement of social class. Comm Health Studies 1984;3:218-22. 23. Lawlor DA, Batty GD, Morton SMB, Deary IJ, Macintyre S, Ronalds G, et al. Early life predictors of childhood intelligence: evidence from the Aberdeen children of the 1950s study. J Epidemiol Community Health 2005;59:656-63. 24. Caldwell B, Bradley R. Home observation for measurement of the environment. New York: Dorsey, 1985. 25. Wechsler D. Wechsler adult intelligence scale – revised. New York: Psychological Corporation USA, 1985. 26. Bayley N. Bayley Scales of Infant Development. New York: The Psychologic Corporation, 1969. 27. McCarthy D. Manual for the McCarthy Scales of Children’s Abilities. New York: The Psychologic Corporation, 1972. 28. Wechsler D. Manual for the Wechsler Intelligence Scale for Children – Revised. New York: Psychological Corporation, 1974. 29. Bradley RH, Whiteside-Mansell L. Home Environment and Children’s Development. In: Families, risk, and competence. Editors: Lewis M, Feiring C. Mahwah, NJ: Lawrence Erlbaum, 1998, pp 133-60. 30. Sternberg RJ, Grigorenko EL. Environmental effects on cognitive abilities. Mahwah, NJ: L. Erlbaum Associates, 2001. 31. Grigorenko E. Heritability and Intelligence. In: Handbook of intelligence. Editor: Sternberg RJ. Cambridge/New York: Cambridge University Press, 2000, pp 53-91.

The Journal of Pediatrics • September 2007

Table I. Regression coefficients (95% CI) of father’s occupational prestige as a predictor of cognitive development Adjusted* Age (y)

Unadjusted

Model I

Model II

Model III

2 4 7 11–13

0.28 0.34 0.33 0.30

0.26 (0.14, 0.37) 0.27 (0.15, 0.39) 0.31 (0.19, 0.43) 0.23 (0.11, 0.36)

0.16 (0.02, 0.31) 0.22 (0.08, 0.37) 0.24 (0.09, 0.38) 0.15 (⫺0.01, 0.30)

0.08 (⫺0.07, 0.23) 0.15 (0.01, 0.29) 0.20 (0.05, 0.34) 0.11 (⫺0.04, 0.26)

*Model I: Variables that were adjusted for included child’s sex, birth weight, head circumference, birth length, Apgar score at 5 minutes, neonatal jaundice, maternal age, duration of gestation, maternal smoking and drinking habits, parental marital status, and lifetime average blood lead concentration up to age 2 years; model II: all the variables in model I plus maternal IQ; model III: all the variables in model II plus HOME.

Table II. Regression coefficients (95% CI) of maternal IQ as a predictor of cognitive development Adjusted* Age (y)

Unadjusted

Model I

Model II

Model III

2 4 7 11–13

0.44 0.56 0.57 0.46

0.43 (0.27, 0.58) 0.50 (0.35, 0.64) 0.57 (0.43, 0.71) 0.41 (0.26, 0.56)

0.40 (0.25, 0.56) 0.45 (0.30, 0.60) 0.53 (0.39, 0.47) 0.41 (0.25, 0.56)

0.29 (0.13, 0.45) 0.34 (0.18, 0.50) 0.48 (0.33, 0.63) 0.32 (0.16, 0.48)

*Model I: Variables that were adjusted for included child’s sex, birth weight, head circumference, birth length, Apgar score at 5 minutes, neonatal jaundice, maternal age, duration of gestation, maternal smoking and drinking habits, parental marital status, and lifetime average blood lead concentration up to age 2 years; model II: all the variables in model I plus Daniel scores; model III: all the variables in model II plus HOME.

Table III. Regression coefficients (95% CI) of quality of home environment as a predictor of cognitive development Adjusted* Age (y)

Unadjusted

Model I

Model II

Model III

2 4 7 11–13

1.21 1.38 1.02 1.06

1.13 (0.85, 1.41) 1.16 (0.87, 1.45) 0.82 (0.53, 1.11) 0.95 (0.66, 1.24)

1.03 (0.74, 1.33) 1.05 (0.75, 1.35) 0.65 (0.35, 0.95) 0.86 (0.55, 1.16)

0.90 (0.54, 1.27) 0.77 (0.42, 1.13) 0.42 (0.07, 0.76) 0.62 (0.25, 0.99)

*Model I: Variables that were adjusted for included child’s sex, birth weight, head circumference, birth length, Apgar score at 5 minutes, neonatal jaundice, maternal age, duration of gestation, maternal smoking and drinking habits, parental marital status, and lifetime average blood lead concentration up to age 2 years; model II: all the variables in model I plus Daniel scores; model III: all the variables in model II plus maternal IQ.

Socioeconomic Position, Maternal IQ, Home Environment, and Cognitive Development

288.e1

A Randomized, Controlled Trial of Tonsillectomy in Periodic Fever, Aphthous Stomatitis, Pharyngitis, and Adenitis Syndrome M. RENKO, MD, PHD, E. SALO, MD, PHD, A. PUTTO-LAURILA, MD, PHD, H. SAXEN, MD, PHD, P. S. MATTILA, MD, PHD, J. LUOTONEN, MD, PHD, O. RUUSKANEN, MD, PHD, AND M. UHARI, MD, PHD

Objective We carried out a prospective, randomized, controlled trial to clarify the effect of tonsillectomy on the clinical course of periodic fever, aphthous stomatitis, pharyngitis, and adenitis (PFAPA) syndrome. Study design Twenty-six consecutive children (mean age 4.1 years) with at least 5 PFAPA attacks were recruited from 3 tertiary care pediatric hospitals during 1999-2003 and randomly allocated to tonsillectomy or follow-up alone. They were all followed up with symptom diaries for 12 months. Tonsillectomy was allowed after 6 months in the control group if the attacks recurred. Results Six months after randomization all 14 children in the tonsillectomy group and 6/12 children in the control group (50%) were free of symptoms (difference 50%, 95% confidence interval 23% to 75%, P < .001). Tonsillectomy was performed on 5/6 of the patients in the control group who still had symptoms after 6 months. The remaining unoperated child in the control group had recurrences of the fever episodes throughout the follow-up, but the symptoms became less severe, and the parents did not choose tonsillectomy. Conclusion Tonsillectomy appeared to be effective for treating PFAPA syndrome. The fever episodes ceased without any intervention in half of the control subjects. We conclude that although the mechanisms behind this syndrome are unknown, tonsillectomy can be offered as an effective intervention for children with PFAPA. (J Pediatr 2007;151:289-92) eriodic fever syndrome refers to recurrent bouts of fever at regular intervals without any definitive infection. The most usual type in children is periodic fever, aphthous stomatitis, pharyngitis, and adenitis (PFAPA) syndrome, characterized by periodic episodes of high fever lasting 3 to 6 days and recurring regularly every 3 to 8 weeks, often associated with aphthous stomatitis, pharyngitis, and adenitis. PFAPA was first described by Marshall et al1 in 1987, and larger series of patients have been described since then.2-4 PFAPA is sporadic, but 3 familial types of periodic fever syndrome with underlying genetic abnormalities have been characterized, that is, Mediterranean fever (MEFV), hyperimmunoglobulinemia D syndrome (HIDS) and tumor necrosis factor receptor–associated periodic syndrome (TRAPS).5-7 The cause of PFAPA syndrome is unknown. Affected children are healthy between episodes and the prognosis is excellent even when the recurrent bouts of fever last for years.4 No amyloidosis or other complications have been reported, even though they are possible in familial types of periodic fever syndrome, such as in MEFV.8 Treatment with antipyretics is of limited value and antimicrobial agents are ineffective,3 and although a single dose of a corticosteroid (1 to 2 mg/kg prednisone) leads to rapid resolution of the fever in most patients,3,4 it does not prevent subsequent episodes. The only therapies found to induce remission in at least some patients are continuous cimetidine therapy9 and tonsillectomy,2-4 but no randomFrom the Departments of Pediatrics (M.R., M.U.) and Otorhinolaryngology (J.L.), Uniized, controlled studies have been published on their efficacy. We carried out a prospecversity of Oulu, the Departments of Peditive, randomized, controlled trial to clarify the effect size of tonsillectomy on PFAPA. atrics (E.S., H.S.) and Otorhinolaryngology

P

METHODS Patients Consecutive children with at least 5 PFAPA attacks were recruited from 3 tertiary care pediatric hospitals from 1999 to 2003. The criteria for an attack were high fever HIDS MEFV PFAPA

Hyperimmunoglobulinemia D syndrome Mediterranean fever Periodic fever, aphthous stomatitis, pharyngitis and adenitis

TRAPS

Tumor necrosis factor receptor–associated periodic syndrome

(P.M.), University of Helsinki and the Department of Pediatrics (A.P-L., O.R.), University of Turku, Finland. Submitted for publication Oct 17, 2006; last revision received Dec 29, 2006; accepted Mar 6, 2007. Reprint requests: Marjo Renko, MD, PhD, Department of Pediatrics, University of Oulu, Box 5000, FIN-90014 University of Oulu, Finland. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.015

289

Table. Background characteristics of the 26 children with PFAPA syndrome allocated randomly to either tonsillectomy (TE) or follow-up alone Characteristic

Tonsillectomy Control group (n ⴝ 14) (n ⴝ 12)

Number of boys 8 (57%) Age, years 4.2 (1.5-14.0) Number of episodes per child 9.0 (5-20) Duration of episodes, days 3.4 (2-4) Interval between episodes, 25.9 (21-28) days Number of children without 7 (50%) any other symptoms than fever Fevermax, °C 39.7 (39-40) CRPmax, mg/L 159.3 (17-342) Leucocytesmax, ⫻ 103/mm3 21.0 (9.1-45.9)

8 (67%) 4.0 (1.5-7.2) 9.5 (4-20) 3.8 (2.5-6) 25.0 (18-28) 4 (33%)

39.9 (39-41) 121.0 (18-210) 16.1 (7.6-22.3)

Numbers are means (range) unless otherwise indicated. Fevermax, CRPmax, and leucocytesmax refer to the mean (range) of the highest measured values during episodes.

Figure. Study protocol.

(ⱖ38.5°C) of unknown origin recurring with a typical, regular pattern and asymptomatic intervals of 2 to 5 weeks. Accompanying signs of aphthous stomatitis, pharyngitis, and adenitis were recorded. Thirty-five children were evaluated, and the parents of 28 of these gave written consent for participation in the study (Figure). The protocol was acceptable to the Ethical Committee of the Northern Ostrobothnia Hospital District. The 28 children had had an average of 9 febrile episodes (range 4 to 20) before recruitment, with fever lasting for a mean of 3.6 days (range 2 to 6) with a mean interval of 25.6 days (range 18 to 28). In 41% of cases the fever was the only symptom during the episodes (Table), 29% had exudative tonsillitis during at least 1 of the episodes and 21% had either cervical lymphadenopathy, aphthous stomatitis or pain in the mouth or throat. One child had a sibling with PFAPA. Two children whose parents were not willing to participate also had a sibling who had had PFAPA that had been cured after tonsillectomy. The parents of these children wanted tonsillectomy to be performed as soon as possible. Because 1 child was found to have acute lymphoblastic leukemia during the follow-up and 1 was lost from follow-up, the final series consisted of 26 white children (Figure). The child with subsequent diagnosis of leukemia was a 6-year-old boy who was randomized to the control group after 8 fever episodes. After entering the study he had only 2 episodes of fever. At the 6-month control visit, he had a history of back pain, was slightly pale, and had enlarged submandibular lymph nodes. Laboratory tests revealed a leukocyte count of 66,000 cells/mm3 with 17% blasts. Acute lymphoblastic leukemia was diagnosed, and he was excluded from the study. 290

Renko et al

Serum levels of immunoglobulins G, A, M, E, and D were measured in 20 of 26 children and were within the age-adjusted normal values in all cases. Either the sedimentation rate or the C reactive protein (CRP) and blood leukocyte count during an episode was available for 24 of 26 children (Table).

Intervention and Follow-up The participants were randomly allocated to either the treatment group, for tonsillectomy to be performed within 1 month, or the control group (Figure). They were then all followed up with symptom diaries filled in by the parents for 1 year. Tonsillectomy was allowed after 6 months in the control group if the child’s symptoms still persisted. Outcome Measures and Statistical Methods The main outcome measure was disappearance of fever episodes according to the symptom diaries. For the sample size calculations we estimated that the fever episodes would disappear in 90% of cases in the tonsillectomy group and in 30% of cases in the control group during the first 6 months of the follow-up. We chose the type I error to be 5% and wanted to be 90% sure when claiming the efficacy to be less than a difference of 60%. With these assumptions, the sample size was calculated to be 12 children per group. To make sure that there would be enough children for analysis in both groups, we chose to enroll 28 children. A balanced randomization was used at each center to minimize bias. The proportion of children with disappearance of symptoms was compared between the treatment groups and the absolute difference and 95% confidence interval (CI) for the difference were calculated. The statistical significance of the difference in proportions was tested with the SND test. The mean number of PFAPA episodes per person-month at risk was determined The Journal of Pediatrics • September 2007

for both treatment groups and the mean difference with CI was calculated. The statistical significance of the difference was tested with the Mann-Whitney U-test.

RESULTS All 14 children in the tonsillectomy group and 6 of 12 (50%) in the control group were free of symptoms 6 months after randomization (difference 50%, CI 23%-75%; P ⬍ .001). Tonsillectomy was performed on 5 of 12 children (42%) in the control group after the follow-up period of at least 6 months because symptoms persisted, after which the symptoms disappeared in all cases. There were no complications after tonsillectomy. One child had persistent fever episodes throughout the follow-up, but the symptoms became less severe, and the parents did not want tonsillectomy to be performed. Four of the 14 children in the tonsillectomy group had 1 fever episode compatible with periodic fever in 6 months after tonsillectomy (0.05 episodes per person-month at risk), and the 12 children in the control group had altogether 34 such episodes in the same time interval (0.44 episodes per personmonth at risk, difference 0.40, 95% CI 0.17 to 0.62; P ⫽ .007). There were no significant differences in the outcome between the children who had periodic fever with or without aphtous stomatitis, pharyngitis, tonsillitis or lymphadenitis. In the tonsillectomy group 7 of 14 (50%) had fever as the only symptom during episodes (Table) and all of them had a successful outcome. In the control group, there were 4 of 12 (33.3%) children with fever as the only symptom during the episodes. In 2 of them, the fevers disappeared spontaneously during the 6-month follow-up, in 1 child tonsillectomy was performed because of persisting symptoms, and 1 child’s parents did not choose tonsillectomy. The 6 children in the control group who were cured without any intervention did not differ in age, sex, or features of the PFAPA attacks from the 6 in whom the attacks recurred.

DISCUSSION Tonsillectomy appeared to be an effective treatment for PFAPA syndrome, although half of the children in the control group had self-resolution during the follow-up of 6 months. The mean number of PFAPA episodes per personmonth at risk during the 6 months of the follow-up was reduced from 0.44 to 0.05 as a result of tonsillectomy. This striking effect has been found previously in some retrospective patient series.2,3,10,11 Galanakis et al,2 in a retrospective evaluation of the preoperative symptoms of 40 children who underwent tonsillectomy for recurrent pharyngitis, found that 15 (37.5%) of them had a history compatible with PFAPA syndrome, and all of these were symptom-free after tonsillectomy. The cause of PFAPA syndrome is unknown. Some clinical features resemble 3 familial types of periodic fever syndrome, that is, MEFV, HIDS, and TRAPS,5-7 but the episodes do not appear as regularly in these as they do in PFAPA. Cyclic neutropenia is a rare syndrome with profound neutropenic episodes causing fever, pharyngitis, lymphade-

nopathy, and pyogenic infections at intervals of 21 to 24 days. The occurrence of cyclic neutropenia is most often familial, however, and it can be differentiated easily from PFAPA syndrome by means of serial leukocyte counts just before the fever. The diagnosis is confirmed by bone marrow examination and genetic testing.5,7 The diagnosis of PFAPA syndrome is based on the clinical picture and the exclusion of other diseases causing recurrent episodes of fever. The children in our series had aphthous stomatitis, cervical lymphadenopathy, and pharyngitis reported less often than the patients described in the literature, and exudative tonsillitis reported more frequently. Signs of mild pharyngitis and lymphadenitis are not very easy to discover in small children and we think it is possible that some of those localizing signs have been missed by physicians. There were 2 children whose symptoms had begun at an older age than is stated in the diagnostic criteria published by Thomas et al.3 Both these patients had a typical clinical picture, however, and were cured by tonsillectomy, as was the one with a sibling with a history of PFAPA. A limitation of the study is inability to confirm cases other than by compatible symptom complex. Additionally, parents and investigators, obviously, were not blinded to treatment group. Even though epidemiologic data are lacking, PFAPA cannot be said to be a rare syndrome in children. Patients are otherwise healthy, and they have no tendency to suffer from recurrent infections in general. Their growth and psychomotor development are normal. On the other hand, the periodic fevers associated with PFAPA may persist for years when untreated.3 Recognition of the syndrome brings significant relief to the parents and eliminates unnecessary courses of antimicrobial therapy. Single-dose prednisone therapy (1 to 2 mg/kg) has been recommended as first-line medication,3,5 but repeated doses of prednisone have led to a decrease in the interval between episodes5,7 and may cause other well-known systemic side-effects in small children. The fear of these side effect results in a poor compliance.11 Tonsillectomy proved to be very effective for treating PFAPA syndrome and was readily accepted by most of the parents. In this series the parents of 1 child preferred the symptoms of PFAPA over tonsillectomy. The parents of 1 child had earlier experience of tonsillectomy and PFAPA syndrome, and they chose tonsillectomy for their child as soon as possible and were not willing to participate in this study because of the possibility of being randomized into the control group. We conclude that tonsillectomy can be offered as an effective treatment when discussing treatment for children with this peculiar syndrome.

REFERENCES 1. Marshall GS, Edwards KM, Butler J, Lawton AR. Syndrome of periodic fever, pharyngitis, and aphthous stomatitis. J Pediatr 1987;110:43-6. 2. Galanakis E, Papadakis CE, Giannoussi E, Karatzanis AD, Bitsori M, Helidonis ES. PFAPA syndrome in children evaluated for tonsillectomy. Arch Dis Child 2002;86:434-5. 3. Thomas KT, Feder HM Jr, Lawton AR, Edwards KM. Periodic fever syndrome in children. J Pediatr 1999;135:15-21. 4. Padeh S, Brezniak N, Zemer D, Pras E, Livneh A, Langevitz P, et al. Periodic fever, aphthous stomatitis, pharyngitis, and adenopathy syndrome: clinical characteristics and outcome. J Pediatr 1999;135:98-101.

A Randomized, Controlled Trial of Tonsillectomy in Periodic Fever, Aphthous Stomatitis, Pharyngitis, and Adenitis Syndrome

291

5. John CC, Gilsdorf JR. Recurrent fever in children. Pediatr Infect Dis J 2002;21:1071-7. 6. Drenth JP, van der Meer JW. Hereditary periodic fever. N Engl J Med 2001;345:1748-57. 7. Long SS. Distinguishing among prolonged, recurrent, and periodic fever syndromes: approach of a pediatric infectious diseases subspecialist. Pediatr Clin North Am 2005;52:811-35. 8. Bakkaloglu A. Familial Mediterranean fever. Pediatr Nephrol 2003;18:853-9.

9. Feder HM Jr. Cimetidine treatment for periodic fever associated with aphthous stomatitis, pharyngitis and cervical adenitis. Pediatr Infect Dis J 1992; 11:318-21. 10. Dahn KA, Glode MP, Chan KH. Periodic fever and pharyngitis in young children: a new disease for the otolaryngologist? Arch Otolaryngol Head Neck Surg 2000;126:1146-9. 11. Tasher D, Somekh E, Dalal I. PFAPA syndrome: new clinical aspects disclosed. Arch Dis Child 2006;91:981-4.

50 Years Ago in The Journal of Pediatrics STEROID

THERAPY FOR RHEUMATIC FEVER

McCue, CM. J Pediatr 1957;50:255-61

Although the incidence of rheumatic fever had been declining in the United States for more than 30 years, in the 1950s it remained a feared illness. Aspirin was available to treat systemic and articular manifestations of the disease, but its efficacy in carditis was uncertain. Many affected children were left with disability, and some died of their illness. The report by Dr. McCue describes her experience with glucocorticoids to treat 94 children with acute rheumatic fever at the Medical College of Virginia during the years 1950 to 1957. Not all the children had carditis, but of those who did, some had pericarditis or severe congestive heart failure. Rheumatic fever was diagnosed using contemporary Modified Jones Criteria and treated using a protocol that evolved during the course of the study. Most patients received between 200 to 300 mg of cortisone or 60 mg prednisone orally for 21 days, followed by a 60-day taper. Outcome was judged according to grade of heart murmur, radiographic size of the cardiac silhouette, and survival. This was an uncontrolled study. However, of 36 patients with carditis who began treatment during the first 28 days of illness, 74% were judged to be either free of heart disease or greatly improved, a significantly better outcome when compared with a group of patients treated 10 years earlier. Furthermore, dramatic responses were observed in individual cases, some of which attending physicians judged to be life-saving. Dr. McCue’s report was one of several studies from the same era, which included larger controlled, multicenter trials. Taken together, these studies failed to establish the superiority of glucocorticoids over aspirin for treatment of acute rheumatic carditis. Since their publication, the incidence of rheumatic fever has continued to decline in most industrialized countries. In contrast, it remains endemic in much of Africa, Asia, and South America,1 where it far exceeds Kawasaki disease as the most common cause of acquired heart disease in childhood. Because of this, it is rather disturbing that, despite advances in cardiac imaging and development of newer immunosuppressive medications, no therapy has ever been convincingly shown to reduce the risk of heart valve lesions in acute rheumatic carditis.2 Efforts are underway to develop a vaccine against group-A streptococcus, the microorganism that elicits the immunologic response that causes rheumatic fever. However, until an efficacious, non-rheumatigenic vaccine becomes available, the need for an effective and affordable treatment for rheumatic carditis will remain as urgent as it was in 1957. Eli M. Eisenstein, MD Department of Pediatrics Hadassah-Hebrew University Medical Center Mount Scopus, Jerusalem, Israel 10.1016/j.jpeds.2007.04.001

REFERENCE 1. Rheumatic fever and rheumatic heart disease. Report of a WHO expert consultation Geneva 29 October 1-November 2001. WHO technical report series 923; Geneva, Switzerland, 2004. Available at: http://www.who.int/cardiovascular_diseases/resources/en/cvd_trs923.pdf. Accessed April 30, 2007. 2. Cilliers AM, Manyemba J, Saloojee H. Anti-inflammatory treatment for carditis in acute rheumatic fever. Cochrane Database Syst Rev 2003;2:CD003176.

292

Renko et al

The Journal of Pediatrics • September 2007

Effect of Prebiotic Supplementation and Calcium Intake on Body Mass Index STEVEN A. ABRAMS, MD, IAN J. GRIFFIN, MBCHB, KELI M. HAWTHORNE, MS,

AND

KENNETH J. ELLIS, PHD

Objective To assess the effects of a prebiotic supplement and usual calcium intake on body composition changes during pubertal growth. Study design We measured anthropometry and body fat with dual-energy X-ray absorptiometry in 97 young adolescents who were randomized to receive either a daily prebiotic supplement or maltodextrin (control) for 1 year. Results Subjects who received the prebiotic supplement had a smaller increase in body mass index (BMI) compared with the control group (BMI difference 0.52 ⴞ 0.16 kg/m2, P ⴝ .016), BMI Z-score (difference 0.13 ⴞ 0.06, P ⴝ .048) and total fat mass (difference 0.84 ⴞ 0.36 kg, P ⴝ .022). The prebiotic group maintained their baseline BMI Z-score (0.03 ⴞ 0.01, paired t test, P ⴝ .30), although BMI Z-score increased significantly in the control group (0.13 ⴞ 0.03, P < .001). In considering subjects whose usual calcium intake was >700 mg/d, those who received the prebiotic supplement had a relative change in BMI that was 0.82 kg/m2 less than control subjects (P < .01), and BMI Z-score that was 0.20 less than control subjects (P ⴝ .003). Differences tended to be maintained 1 year after supplementation was stopped. Conclusion Prebiotic supplementation and avoidance of a low calcium intake can have significant effects in modulating BMI and other body composition changes during puberty. (J Pediatr 2007;151:293-8) lthough the optimal calcium intake is uncertain, few challenge the concept that very low calcium intakes during puberty have long-term detrimental health effects on bone.1,2 We have shown that the use of a prebiotic (a food substance that promotes the growth of potentially beneficial bacteria in the intestines), specifically an inulin-type fructan (ITF), was associated with increased calcium absorption and bone mineralization in nonobese pubertal children.3 It has been hypothesized that increased dietary calcium leads to decreased weight gain and lower body mass index (BMI) in adults via an effect on fat cells,4,5 although a recent systematic review did not confirm this relationship.6 The relatively few data available in children suggest a similar benefit in both young children and adolescents.7-11 Prebiotics may also play a role in weight maintenance.12-14 Although few data exist regarding the mechanism of this effect, it appears to involve both fiber and hormonal effects, which increase the sense of satiety and lead to long-term decreases in undesired weight gain.14,15 An interaction between prebiotic supplementation and calcium intake on body composition has not been previously considered. As part of our blinded, randomized controlled trial of prebiotic supplementation and bone mineralization, we obtained longitudinal anthropometry, dietary, and body composition data in those supplemented and the control group.3 That trial was not From the USDA/ARS Children’s Nutrition specifically designed to evaluate body composition or BMI as related to either diet or Research Center, and Texas Children’s Hospital, Department of Pediatrics, Baylor prebiotic supplementation. Thus subjects were unlikely to have altered their dietary College of Medicine, Houston, Texas. pattern on the basis of an expectation that enrollment in the study would lead to body fat Supported by federal funds from the changes. USDA/ARS under Cooperative Agreement number 58-6250-6-001, the NIH, NCRR In this report, we have used the data from that trial, including a post-supplemenGeneral Clinical Research for Children tation follow-up, to examine the effects of prebiotic supplementation and usual diet on Grant number RR00188, NIH AR43708 weight and body composition changes during puberty. We hypothesized that this analysis and NIDDK, P30 DK56338. Submitted for publication Nov 1, 2006; last would demonstrate that both calcium intake and prebiotic supplementation would be revision received Jan 23, 2007; accepted related to study changes in total body fat, BMI and BMI Z-score and that these would Mar 19, 2007. be related to each other such that higher calcium intakes were associated with an enhanced Reprint requests: Steven A. Abrams, MD, 1100 Bates St, Houston TX 77030. E-mail: prebiotic effect on BMI.

A

BMI DXA

Body mass index Dual-energy X-ray absorptiometry

ITF

Inulin-type fructan

[email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.043

293

METHODS By public advertising, we identified 50 girls and 50 boys for this study. Of these 97 subjects (49 boys and 48 girls) completed the 1-year study intervention study. All subjects were between 9.0 to 13.0 years of age and were selected to approximately match the ethnic distribution of the greater Houston area. All subjects received a screening physical examination including Tanner staging before inclusion in the study. To be enrolled, subjects had to be healthy, nonobese, Tanner stage 2 or 3, and girls had to be premenarchal. Written informed consent was obtained from a parent or legal guardian for each subject; written assent was obtained from all of the study subjects. The Institutional Review Board of Baylor College of Medicine and Affiliated Hospitals approved this protocol. Within 8 weeks of the screening visit described above, subjects were admitted for 24 hours to the General Clinical Research Center of Texas Children’s Hospital in Houston, TX. During this stay, measurements of calcium absorption and dual-energy X-ray absorptiometry (DXA) measurement of bone mineralization,3 as well as body fat determination were carried out. At the end of this baseline study, subjects were randomized, in a double-blinded fashion, and stratified by sex to 1 of 2 carbohydrate supplement groups; either 8 g/d of a prebiotic, ITF (Beneo Synergy1, Orafti, Tienen Belgium), or 8 g/d of a maltodextrin control. The ITF prebiotic was a co-spray dried 1:1 mixture of oligofructose (average degree of polymerization, DPav ⫽ 4) and long-chain inulin (DPav ⫽ 25). Subjects were instructed to mix the carbohydrate supplement with calcium-fortified orange juice and to drink it with breakfast daily for 12 months. The supplemented juice (180-240 mL/d) provided approximately 80 to 110 kcal/d and the prebiotic 12 kcal/d (1.5 kcal/g).16 The maltodextrin control provided approximately 32 kcal/d (4 kcal/g). The energy contribution of the juice, but not the supplements, was included in the energy intake calculations. To provide some dietary variation, subjects were also allowed to use milk to mix the carbohydrate supplement. Dietary recalls and discussions with families demonstrated that all subjects primarily used orange juice, and this accounted for more than 95% of total study days. Twelve months after the initial baseline study, subjects returned for a follow-up visit in which anthropometry and DXA measurements were performed. Subjects then discontinued the supplement, and follow-up anthropometry and DXA measurements were performed 12 months later (2 years after enrollment). At the screening visit, a dietary history was obtained to determine what subjects usually ate on a normal day. Food preferences were also obtained. Inpatient menus for the overnight study visit were based on this reported typical calcium intake. All foods and beverages during the inpatient and outpatient visits were pre-weighed and post-weighed to accurately determine intake. Subjects were instructed to keep weighed food records for 2 days after the first overnight visit 294

Abrams et al

and for 2 days after the 1-year visit. Subjects were called at home every 2 months during the 1-year period to obtain a 24-hour dietary recall of the previous day’s intake and to ensure that the subject maintained a relatively consistent calcium intake. Energy intake was not regulated during any time of the study. Another 24-hour dietary recall was obtained at the 2-year follow-up visit. Dietary intake data were collected with Nutrition Data System for Research software versions 4.03 and 4.05, developed by the Nutrition Coordinating Center, University of Minnesota, Minneapolis, MN. Because of the possibility that the study intervention lead to a change in dietary intake, we utilized the intake values at the one-year visit (both dietary recall and in-patient and outpatient weighed diet averages) as the primary study dietary variable for energy and calcium intake. Data from the 1-day weighed inpatient food record at the 1-year visit along with the 2-day weighed home food records at this visit were pooled. This average value was used to represent the dietary intake at the end of 1 year of supplementation. Whole body fat was determined by use of a Hologic QDR-4500A DXA absorptiometer (Hologic, Inc, Waltham, MA) scanning in the fan-beam mode. Total body fat was calculated as the product of body weight and body fat percentage (DXA). BMI Z-scores were calculated from the online database of our institution.17 Statistical comparisons between prebiotic and control groups were made by use of the generalized linear model (analysis of variance) function of SPSS 13.0 for Windows (SPSS, Inc., Chicago, IL). For each analysis the final value was used as the dependent variable, and the initial value used as a covariate in analyses. The principle model used included the initial height Z-score, the Tanner breast stage after 1 year, and the calcium and energy intake after 1 year. The initial height Z-score adjusted for size difference at the start of the study but was not a significant factor in any outcomes. The Tanner stage at 1 year was significantly related to changes in BMI and BMI Z-score (P ⬍ .05) in most of the analyses. A significant interaction was found between energy intake and prebiotic use, but not between energy use and calcium intake or calcium intake and prebiotic use for BMI and BMI Z-score outcomes. Therefore only the energy intake and prebiotic use interaction were included in the final model. The effect of calcium intake on changes in body composition was assessed by including this as a continuous independent variable or a dichotomous independent variable (with cutoffs of either 700 mg/d, 800 mg/d, or 900 mg/d). We also evaluated whether a significant change in body composition had occurred during the year of the study in either of these groups. As BMI normally increases during puberty, we evaluated the age and sex normalized value, the BMI Z-score, to determine whether a potentially undesirable increase in BMI had occurred in either group. BMI z-score at the start and end of the study were compared in each of the 2 groups individually with a paired t test. Two subjects (both in the control group) had a baseline BMI Z-score ⬍1.65. Analyses were carried out with and The Journal of Pediatrics • September 2007

Table I. Baseline anthropometry before supplementation with prebiotic or control* Prebiotic (n ⴝ 48)

Non-prebiotic P (control) (n ⴝ 49) value

Age (y) 11.8 ⫾ 0.2 Weight (kg) 42.6 ⫾ 1.3 Height (cm) 149.1 ⫾ 1.3 Weight Z-score 0.2 ⫾ 0.1 Height Z-score 0.0 ⫾ 0.2 Tanner stage (2/3) 34/14 M/F 25/23 BMI (kg/m2) 18.99 ⫾ 0.37 BMI Z-score 0.26 ⫾ 0.18 Body fat (%) 24.7 ⫾ 0.9 Total fat mass (kg) 10.7 ⫾ 0.6 Calcium intake (mg/d) 904 ⫾ 45 Energy intake (kcal/d) 2287 ⫾ 67

11.4 ⫾ 0.2 41.3 ⫾ 1.3 148.3 ⫾ 1.3 0.3 ⫾ 0.1 0.3 ⫾ 0.1 37/12 24/25 18.62 ⫾ 0.37 0.20 ⫾ 0.18 24.5 ⫾ 0.9 10.4 ⫾ 0.6 882 ⫾ 45 2195 ⫾ 66

.02 .46 .67 .56 .14 .61 .76 .49 .73 .89 .69 .73 .33

*One missing case (control) for body fat and total fat mass. Data are mean ⫾ SEM. Tanner stage refers to breast/penile as stages 2 or 3.

without these subjects included in the analysis because of the possibility of the higher BMI Z-score subjects affecting the results. Omission of these subjects did not substantially change the results, and the results are presented with all subjects included. All data are presented as mean ⫾ SEM. Values of P ⬍ .05 were considered significant.

RESULTS A total of 97 subjects, 48 who received the prebiotic supplementation and 49 who received the control completed the 1-year study. Of these, follow-up data were available at 2 years for 89 subjects, 44 of whom had received the prebiotic supplementation. Sex and ethnicity were considered in the analysis and were not found to be significantly related to any of the outcomes of the study and thus dropped from the analysis. Anthropometric characteristics of the study subjects are shown in Table I. By study design, there were no subjects enrolled with a BMI Z-score ⬎2.0 (maximum was 1.86; 2 subjects had a BMI Z-score ⱖ1.65 representing the 95 percentile). One subject with a BMI Z-score of ⫺2.13 was the only subject with a BMI Z-score ⬍⫺2.0. Results for the effect of prebiotic supplementation on the basis of this model are shown in Table II. The increment in BMI over the study year was 0.73 kg/m2 for the prebiotic group and 1.24 kg/m2 for the control group (Difference 0.52 ⫾ 0.16 kg/m2, P ⫽ .016). Omitting calcium intake from the analysis negligibly changed the relative prebiotic effect on BMI by 0.01 kg/m2. The same negligible effect of including calcium intake in the model on the other body composition outcomes was found. In the prebiotic group no significant increase in BMI Z-score occurred during the study (0.03 ⫾ 0.01; P ⫽ .30); in contrast BMI Z-score increased significantly in the control group (0.13 ⫾ 0.04; P ⫽ .001).

We did not find any significant relationship between the baseline BMI Z-score (P ⫽ .66) or whether the subject tended to be overweight at the start of the study (BMI Z-score ⬎1.0, P ⫽ .84) on the effect of the prebiotic on changes in BMI Z-score. Omitting the 2 subjects (both control group) with a baseline BMI Z-score ⱖ1.65 or the 1 subject (control group) with a BMI Z-score ⬍⫺2.0 similarly had no substantial effect on the findings or their statistical significance. To examine the effect of the prebiotic on body composition at different energy intakes, we modeled the data identically as before but omitted energy intake and the energy intake-prebiotic interaction as covariates. For BMI, the between-group differences became 0.49 ⫾ 0.022 kg/m2, P ⫽ .026 and for BMI Z-score, 0.123 ⫾ 0.06, P ⫽ .062, values virtually identical to the analysis that had included energy intake and energy intake-prebiotic interaction (Table II). Thus, although there was a statistically significant interaction of energy intake with prebiotic use, the effect of energy intake and the interaction with prebiotic supplementation on the overall differences were small. We found a P ⫽ .08 for the effect of calcium intake on BMI and BMI Z-score when calcium intake was considered as a continuous linear covariate. However, when calcium intake was considered as a nonlinear dichotomous variable and subjects were categorized as having higher or lower relative calcium intakes using cutoff points of 700 mg/d, 800 mg/d or 900 mg/d, the results depended on the cutoff point selected. A highly significant (P ⬍ .01) effect of calcium intake on BMI and BMI Z-score was seen (Table III) in a model with 700 mg/d intake or 800 mg/d intake as cutoff points. In both cases, the change in BMI, BMI Z-score, fat mass, and weight over the study year were significantly less for subjects above the calcium intake cutoff point than for those below it. Use of a calcium intake cutoff point of 900 mg/d showed no significant effect of calcium intake on BMI or BMI Z-score. Omitting the 2 subjects with a BMI Z-score ⱖ1.65 or the subject with a BMI ⬍⫺2.0 had a very small effect on the results. Use of the prebiotic had no significant effect on body composition outcomes for calcium intakes ⬍700 mg/d (Figure). This value was chosen based on the findings in Table III that there was a nonlinear effect of calcium intake on body composition with an effect seen at intakes as low as 700 mg/d as a cutoff point. The rationale for the use of cutoff points was the likelihood that the nonlinear effects would not allow for identification of the calcium effect. Furthermore, cutoff values are often used in establishing dietary requirement and nutritional guidelines. In subjects with calcium intakes ⱖ700 mg/d, the prebiotic supplementation was associated with a BMI difference of 0.82 kg/m2, BMI Z-score difference of 0.2, fat mass difference of 1.3 kg, and a body weight of 2.0 kg compared with those not receiving the prebiotic (values lower for prebiotic group, all P ⬍ .01, except BMI, P ⬍ .001).

Effect of Prebiotic Supplementation and Calcium Intake on Body Mass Index

295

Table II. Effect of supplementation on body composition after 1 year*

BMI Z-score BMI (kg/m2) Weight (kg) Height (cm) Body fat (%) Total fat mass (kg)

Prebiotic (n ⴝ 48)

Control (n ⴝ 49)

Difference

P value

0.25 ⫾ 0.045 19.52 ⫾ 0.15 47.7 ⫾ 0.4 155.7 ⫾ 0.3 23.3 ⫾ 0.4 11.24 ⫾ 0.25

0.38 ⫾ 0.044 20.03 ⫾ 0.15 49.0 ⫾ 0.4 155.7 ⫾ 0.3 24.2 ⫾ 0.4 12.07 ⫾ 0.25

0.13 ⫾ 0.06 0.52 ⫾ 0.21 1.3 ⫾ 0.6 0.0 ⫾ 0.5 ⫺0.8 ⫾ 0.6 0.84 ⫾ 0.36

.048 .016 .048 .99 .14 .022

*One missing case (control). Evaluation of each outcome was done with a model in which the covariates were the baseline height Z-score, the energy and calcium intakes at 1 year, the Tanner stage at 1 year, and the interaction of the prebiotic and energy intake. Each model individually included the baseline value for the dependent variable. The interaction of energy intake and prebiotic supplementation was significant, P ⬍ .01 for each body composition variable.

Table III. Study year differences in change in BMI Z-score, BMI, total fat mass, and weight in subjects consuming a calcium intake below the cutoff point compared with those consuming calcium intakes above the cutoff point 700 mg/d‡

800 mg/d§

900 mg/d储

Calcium intake cut-off*†

Difference

P value

Difference

P value

Difference

P value

BMI Z-score BMI (kg/m2) Total fat mass (kg) Weight (kg)

0.20 ⫾ 0.07 0.67 ⫾ 0.25 1.02 ⫾ 0.42 1.3 ⫾ 0.7

.008 .006 .016 .076

0.17 ⫾ 0.07 0.54 ⫾ 0.22 0.58 ⫾ 0.36 1.2 ⫾ 0.6

.014 .015 .13 .06

0.08 ⫾ 0.07 0.25 ⫾ 0.23 0.13 ⫾ 0.40 0.6 ⫾ 0.7

.28 .28 .74 .37

*Values shown are the difference between changes during the study year in the body composition parameters for subjects with calcium intakes lower than the indicated cut-off compared to those with higher intakes. No significant effects were seen using any cutoffs ⬎900 mg/d. All values are positive, indicating a larger change (increase) in those with intakes below the cut-off compared to those above the cut-off. †Total sample was 97 subjects (one missing for total fat mass). Covariate model included prebiotic supplementation, height Z-score, Tanner stage, energy intake and interaction of energy intake and prebiotic supplementation. ‡n ⫽ 23 below cut-off, n ⫽ 74 above. §n ⫽ 40 below cut-off, n ⫽ 57 above. 储n ⫽ 54 below cut-off, n ⫽ 43 above.

Follow-up data were available for 89 of the 97 (43 prebiotic) subjects at 2 years. The prebiotic intervention had been stopped after 1 year and DXA and anthropometry were performed 12 months later. We found a BMI difference of 0.68 ⫾ 0.36 kg/m2, P ⫽ .061 for the prebiotic effect, and 0.91 ⫾ 0.41, P ⫽ .03 for the calcium effect on BMI (lower values in those who received prebiotics and those who had a calcium intake ⱖ700 mg/d). Therefore, although the variability in BMI increased, the magnitude of each effect was maintained or slightly increased during the year after stopping the prebiotic supplement.

DISCUSSION We found that supplementation with a prebiotic, in addition to its benefit to bone mineralization, had a significant benefit in the maintenance of an appropriate BMI increase during pubertal growth in primarily nonobese young adolescents. This effect was significantly modified in a nonlinear fashion by the dietary intake of calcium such that the maximum benefit to the prebiotic occurred when low calcium intakes were avoided. BMI normally increases during puberty at a yearly rate of about 0.6 to 0.8 kg/m2.18 We found that the prebiotic group had an increase in BMI of about 0.7 kg/m2 during the supplementation year, consistent with expected increases dur296

Abrams et al

ing puberty and that the control group had an increase of 1.2 kg/m2. We further considered the effect on the age- and sex-normalized BMI, the BMI Z-score. The changes in BMI Z-score demonstrated no significant change in the BMI Z-score in the prebiotic group compared with a significant increase in the Z-score of the control group during the study year. Thus the overall greater increase in BMI during the year in the control group was likely not ideal. It is not clear why the control group had an increase in BMI Z-score, although this may be related to overall trends toward increased BMI currently. It is possible that it was related to the placebo. Regardless, further data are needed to determine whether an actual decrease in BMI could be achieved with prebiotic supplementation or only a limitation in an undesirable increase in BMI. The calcium effects were more difficult to quantify because this was not a calcium intervention trial. There was a significant interaction of calcium intake with prebiotic supplementation, but only when subjects with low intakes of calcium were considered as a group. Because the calcium intake assessments used were at a single time point, although robust and reflective of dietary recall data, weighed diets, and their inpatient dietary records, the specific value at which a calcium effect could be seen and the magnitude of the effect should be interpreted cautiously. Nonetheless our data demThe Journal of Pediatrics • September 2007

Figure. Effects of prebiotic supplementation versus control on body composition outcomes in 97 subjects (96 for fat mass). Covariate model included prebiotic supplementation, height Z-score, Tanner stage, energy intake and interaction of energy intake and prebiotic supplementation. Differences not significant for all results in ⬍700 mg/d calcium intake comparisons. *P ⬍ .01. **P ⬍ .001.

onstrate that calcium intake values below about 800 mg/d were associated with a greater increase in BMI than greater intakes and that the benefit to prebiotic supplementation was greater when calcium intake is not low. The mechanism of a potential calcium effect on weight, however, is likely related to a direct effect on fat cell metabolism.4,5 These findings of a beneficial effect of avoidance of a low calcium intake on maintenance of BMI are consistent with our hypothesis and with a recent study in which adolescents in Brazil in the lowest quartile of calcium intake had a significantly greater BMI than those in the highest quartile.11 Lappe et al8 found no effect of calcium intake on weight gain in girls during puberty but the non-supplemented group had an average intake above the level found to have an effect in our data. Dixon et al9 found an inverse relationship between BMI and calcium intake, but only in 7- to 10-year-old children without hypercholesterolemia. Our data are consistent with these results, but we did not evaluate lipid status in our study subjects and thus cannot identify a specific relationship between lipid status and prebiotic supplementation.

A single study of preschool children reported that a 300-mg increase in calcium intake was associated with 1 kg less body fat.5,7 Our data found a 1.0-kg difference in body fat mass in those with calcium intakes ⬎700 mg/d compared with those with lower intakes and a 0.8-kg difference in those who received the prebiotic. Combination of a calcium intake ⬎700 mg/d and prebiotics led to a 1.3 kg difference in fat mass. These data are consistent with the earlier study in preschool children. In contrast, Phillips11 did not find a relationship between body fat percentage or BMI and dairy food intake during pubertal development in nonobese girls. A limitation in this study was the lack of control of energy intake during the study. We did not regulate or investigate the consequences to the whole diet of the supplemental orange juice provided with the study. The increment in calories from the maltodextrin (control) was ⬍2% of the mean caloric intake and much less than the intake from the orange juice. It is unlikely to have contributed substantially to the increase in BMI Z-score that occurred in that group. We did not specifically recruit overweight or obese subjects because of the potentially confounding effect of these on bone mineral metabolism changes; however, we found no effect of baseline BMI Z-score on the prebiotic effect. Energy expenditure was not assessed, nor was physical activity estimated. We used multiple methods to assess dietary intakes. However, all such methods are limited in their ability to assess long-term intakes and therefore may not represent exact measures of usual intake. We did not have dietary data available at 2 years, so we cannot determine whether the trend toward a persistent benefit was associated with dietary changes. The mechanism of the prebiotic effect on BMI and body fat has only minimally been evaluated to date. In a rat model, ITF regulates appetite via increases in gastrointestinal peptides that modulate food intake such as glucagon-like peptide-1.12 A recent pilot study involving 10 young adults suggested that prebiotics reduce hunger and food consumption.13 The lack of a significant increase in BMI Z-score in the prebiotic group despite the supplementation with both the prebiotic and with juice implies an overall regulatory effect on energy intake associated with the diet or with the prebiotic. In our study, although energy intake was assessed at the beginning and end of the study, the tools used and the assessment methods would not be able to identify small changes in intake over a long period of time. Our findings demonstrate the potential for dietary interventions such as calcium and prebiotic supplementation to assist in maintaining appropriate rates of increase in BMI during puberty in nonobese young adolescents, as well as their known benefits for bone and gastrointestinal health. Controlled trials evaluating other population groups, including those who are more overweight and combining these interventions with exercise and behavior modification should be conducted to evaluate the overall potential optimal strategy for decreasing the risk of excessive increase in BMI during early adolescence.

Effect of Prebiotic Supplementation and Calcium Intake on Body Mass Index

297

The authors would like to acknowledge the assistance of Leslie Cruz, Penni Hicks, PhD, and Adrianne Morse in data analysis and E. O’Brian Smith, PhD for statistical advice. Orange juice used in the study was provided by The Coca-Cola Company, Houston TX and the inulin-type fructan by Orafti, Tienen, Belgium.

REFERENCES 1. Dietary reference intakes for Calcium, Magnesium, Phosphorus, Vitamin D, and Fluoride, Standing Committee on the Scientific Evaluation of Dietary Reference Intakes, Food and Nutrition Board, Institute of Medicine. Washington, DC: National Academy Press, 1997. 2. Greer FR, Krebs NF. Optimizing bone health and calcium intakes of infants, children and adolescents. Pediatrics 2006;117:578-85. 3. Abrams SA, Griffin IJ, Hawthorne KM, Liang L, Gunn SK, Darlington G, et al. A combination of prebiotic short-and long-chain inulin-type fructans enhances calcium absorption and bone mineralization in young adolescents. Am J Clin Nutr 2005; 82:471-6. 4. Zemel MB. Calcium modulation of hypertension and obesity: mechanisms and implications. J Am Coll Nutr 2001;20:428-35. 5. Heaney RP, Davies KM, Barger-Lux J. Calcium and weight: clinical studies. J Am Coll Nutr 2002;21:152S-5S. 6. Trowman R, Dumville JC, Hahn S, Torgerson DJ. A systematic review of the effects of calcium supplementation on body weight. Br J Nutr 2006;95:1033-8. 7. Carruth BR, Skinner JD. The role of dietary calcium and other nutrients in moderating body fat in preschool children. Int J Obes 2001;25:559-66.

298

Abrams et al

8. Lappe JM, Rafferty KA, Davies KM, Lypaczewski G. Girls on a high-calcium diet gain weight at the same rate as girls on a normal diet: a pilot study. J Am Diet Assoc 2004;104:1361-7. 9. Dixon LB, Pellizzon MA, Jawad AF, Tershakovec AM. Calcium and dairy intake and measures of obesity in hyper-and normocholesterolemic children. Obes Res 2005;13:1727-38. 10. dos Santos LC, Martini LA, Padua Cintra I, Fisberg M. Relationship between calcium intake and body mass index in adolescents. Arch Latinoam Nutr 2005;55:345-9. 11. Phillips SM, Bandini LG, Cyr H, Colclough-Douglas S, Naunova E, Must A. Dairy food consumption and body weight and fatness studied longitudinally over the adolescent period. Int J Obes 2003;27:1106-13. 12. Cani PD, Dewever C, Delzenne NM. Inulin-type fructans modulate gastrointestinal peptides involved in appetite regulation (glucagons-like peptide-1 and ghrelin) in rats. Br J Nutr 2004;92:521-6. 13. Cani PD, Joly E, Horsmans Y, Delzenne NM. Oligofructose promotes satiety in healthy human: a pilot study. Eur J Clin Nutr 2006;60:567-72. 14. Delzenne NM, Cani PD, Dabuioul C, Neyrinck AM. Impact of inulin and oligofructose on gastrointestinal peptides. Br J Nutr 2005;93(Suppl 1):S157-61. 15. Archer BJ, Johnson SK, Devereux HM, Baxter AL. Effect of fat replacement by inulin or lupin-kernel fibre on sausage patty acceptability, post-meal perceptions of satiety and food intake in men. Br J Nutr 2004;91:591-9. 16. Roberfroid MB. Caloric value of inulin and oligofructose. J Nutr 1999; 129:1436S-7S. 17. Shypailo RJ, Carter J. Children’s BMI-percentile-for-age Calculator. 2001. Retrieved 12/12/2006 from the Baylor College of Medicine, Children’s Nutrition Research Center Web Site: (http://www.bcm.edu/cnrc/bodycomp/bmiz2.html). 18. Maynard LM, Wisemandle W, Roche AF, Chumlea WC, Guo SS, Siervogel RM. Childhood body composition in relation to body mass index. Pediatrics 2001; 107:344-50.

The Journal of Pediatrics • September 2007

Long-Term Follow-Up in 12 Children with Pulmonary Arteriovenous Malformations: Confirmation of Hereditary Hemorrhagic Telangiectasia in all Cases AURORE CURIE, MD, GAËTAN LESCA, MD, VINCENT COTTIN, MD, PHD, PATRICK EDERY, MD, PHD, GABRIEL BELLON, MD, PHD, MARIE E. FAUGHNAN, MD, MSC, AND HENRI PLAUCHU, MD, PHD

Objective To assess whether pulmonary arteriovenous malformation (PAVM) is associated with hereditary hemorrhagic telangiectasia (HHT). Study design This study was a review of 12 children (sex ratio ⴝ 1) including family history, mutation analysis, and long-term follow-up. Results Five children were under age 3 years when PAVM was diagnosed. Presentations included pulmonary symptoms (n ⴝ 8), cerebral abscess (n ⴝ 2), and transient ischemic attack (TIA) (n ⴝ 1); 1 patient was asymptomatic. Nine of the 12 children (75%) had a family history of PAVM. The diagnosis of HHT was confirmed in all cases. A mutation in ENG was found in 9 of the 10 children available for testing. No mutation in ACVRL1 was found. During long-term From the Department of Clinical Genetics follow-up (mean, 16 years), the following complications occurred: TIA (n ⴝ 2), hemopand National Reference Centre of Rendutysis (n ⴝ 2), and cerebral abscess (n ⴝ 2). Nine children experienced recurrence of Osler Disease, Hoˆtel-Dieu Hospital, Civil Hospices of Lyon, University of ClaudePAVM. The children with no recurrence were those without a family history of PAVM. Bernard Lyon 1, France (A.C., G.L., H.P.); Conclusions The diagnosis of HHT should be considered in a child with an apparently the Department of Molecular Genetics, Edouard Herriot Hospital, Civil Hospices of isolated PAVM. Because serious complications may occur at any age, we recommend Lyon, France (G.L.); the Department of screening for PAVM and long-term follow-up in children from families with HHT, Pneumology and Reference Centre of Orespecially those with an ENG mutation. (J Pediatr 2007;151:299-306) phaned Pulmonary Disease, Louis Pradel ulmonary arteriovenous malformations (PAVMs) are abnormal vascular connections between pulmonary arteries and veins that provide a direct, capillary-free communication between the pulmonary arterial and pulmonary venous circulations, leading to hypoxemia, dyspnea, exercise intolerance, and cyanosis. These pulmonary symptoms seem to correlate with the degree of right-to-left shunting present.1 Furthermore, large, subpleural, thin-walled PAVMs may rupture into the pleura, leading to spontaneous hemothorax, or central PAVMs may rupture into the bronchi, leading to massive hemoptysis. These hemorrhagic complications occur in ⬃15% of untreated patients.1-6 Paradoxical embolism to the systemic circulation also can occur due to the direct connection between the pulmonary artery and vein, leading to such neurologic complications as cerebral abscess (septic emboli), transient ischemic attack (TIA), or stroke (thromboemboli) in about 40% of adults with untreated PAVMs.1-6 Studies based mainly on adults have shown that up to 90% of patients with PAVMs actually have hereditary hemorrhagic telangiectasia (HHT), also known as Rendu-OslerWeber disease.4,7-11 Nevertheless, HHT may have remain undiagnosed in a given family at the time of the initial diagnosis of PAVM in a family member.4 HHT is an autosomal dominant hereditary disorder caused by mutation in 1 of 2 genes: ENG, encoding endoglin, or ACVRL1, encoding activin receptor-like kinase 1. Both of these genes are involved in the transforming growth factor-␤ signaling pathway. Clinical features of

P

CEMRA CT HHT

Contrast-enhanced pulmonary magnetic resonance angiography Computed tomography Hereditary hemorrhagic telangiectasia

PAVM TCE TIA

Pulmonary arteriovenous malformation Transcatheter embolotherapy Transient ischemic attack

Hospital, UMR754 University of ClaudeBernard Lyon 1, France (V.C.); the Department of Paediatric Clinical Genetics, Debrousse Hospital, Civil Hospices of Lyon, University of Claude-Bernard Lyon 1, France (P.E.); the Department of Paediatric Pulmonology, Debrousse Hospital, Civil Hospices of Lyon, University of ClaudeBernard Lyon 1, France (G.B.); the Department of Medicine, Division of Respiratory Medicine, St Michaels’s Hospital, University of Toronto, Toronto, Canada (M.F.). Supported by the French Ministry of Medical Research, the French Rendu-Osler Disease Network financed by INSERM, the Association Française contre les Myopathies, and the French Ministry of Health (Reference Center of Rendu-Osler Disease, Hôtel-Dieu, Lyon). M.F. also received support from the Nelson Arthur Hyland Foundation, St Michael’s Hospital Research Institute. Submitted for publication Jul 20, 2006; last revision received Dec 21, 2006; accepted Mar 14, 2007. Reprint requests: Dr Aurore Curie, Service de Génétique Clinique, Hôpital HôtelDieu, 1 Place de l’Hôpital, 69288 Lyon Cedex 02, France. E-mail: aurorecurie@ yahoo.fr. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.021

299

HHT include recurrent spontaneous epistaxis often leading to anemia, mucocutaneous telangiectasia, and visceral arteriovenous malformations in the pulmonary, cerebral, or hepatic vascular bed.12 A family history of HHT is almost always found but may be neglected. Clinical diagnosis of HHT is considered definite if 3 of the criteria are present, possible when 2 criteria are present, and uncertain when only 1 criterion is present.13 The prevalence of PAVMs in HHT patients ranges from approximately 15% to 36% in predominantly adult series.2,3,5,14,15 PAVM is most often diagnosed in the second or third decade of life.1,12 Although 92 children with PAVM have been reported since 1965, data on the outcome of PAVMs have been less well described in children compared with adults.1,5,8-10,16-27 Furthermore, family history (including HHT but also PAVM), molecular data, and long-term follow-up are often lacking. In the present study, we report detailed data on 12 children presenting with apparently isolated PAVMs in whom a diagnosis of HHT could be definitely established.

METHODS Between 1975 and 2000, 12 children were referred to the Department of Medical Genetics of Hoˆtel-Dieu Hospital in Lyon after diagnosis and initial treatment of PAVMs, to determine whether these apparently isolated lesions could be the lead manifestation of an undetected HHT family disease. Only 1 child (patient 11) belonged to a previously known HHT family; the other children’s families had no previous diagnosis or knowledge of Rendu-Osler-Weber disease.

Pulmonary Investigations Chest radiograph, room air arterial blood gas or pulse oximetry, hemoglobin level, and pulmonary angiography were performed in all patients. Nine children underwent thoracic computed tomography (CT) or contrast-enhanced pulmonary magnetic resonance angiography (CEMRA). Treatment of PAVMs in Lyon was limited to surgical resection or ligation until 1992, after which most patients have been treated with transcatheter embolotherapy (TCE). Diagnosis of HHT A comprehensive family history for epistaxis and other clinical symptoms related to HHT was elicited. The patients and parents underwent physical examination, including an extensive search for telangiectasia (especially at characteristic sites, such as the lips, oral cavity, fingers, and nasal mucosa). We applied the HHT consensus clinical diagnostic criteria13 to evaluate the certainty of the HHT diagnosis in each patient. When a blood sample was available, mutation analysis (in ENG and ACVRL1) was performed in Lyon.28,29 Written informed consent was obtained from the patients, or from their parents if they were under age 18, in accordance with the French bioethics law. 300

Curie et al

Follow-Up Follow-up data were collected for all patients, including assessment for HHT symptoms, assessment of pulmonary symptoms and complications of PAVM, as well as pulse oximetry or arterial blood gas, hemoglobin level, and, more recently, contrast echocardiography. If either test was positive, then thoracic CT was performed. The follow-up period started at the time of the PAVM diagnosis.

RESULTS Clinical Presentation and Family History The study group comprised 6 females and 6 males, ranging in age from 1 day to 18 years (Table I). Five children were under age 3 years at the time of PAVM diagnosis. Patients 10 and 12, the aunt and the mother of patient 1, also presented with PAVM before age 18 years. The most frequent presenting complaints were pulmonary symptoms (n ⫽ 8; 67%), including dyspnea on exertion, respiratory distress at rest (sometimes occurring during an infectious episode), cyanosis, and hemoptysis. Three children presented with inaugural neurologic complications (25%), 2 with a cerebral abscess (patients 8 and 12) and 1 with a TIA (patient 11). In the latter case, TIA indirectly led to the diagnosis of PAVM, which was eventually considered because the child’s hemoglobin level was increased despite recurrent epistaxis. In the remaining case (patient 9), PAVM was discovered on routine chest radiograph performed for minor thoracic trauma. On clinical examination, 10 children were cyanotic (85%), 5 had clubbing (42%), and 9 had a pulmonary bruit (75%), but examination was unremarkable for the 2 remaining children. Although only 6 patients had telangiectasia or epistaxis at the time of diagnosis, the others developed these clinical features during the follow-up period (Table I). Long-term follow-up and family history allowed us to confirm the clinical diagnosis of HHT in all children (Table I). Although 11 of the 12 children and their families had never heard of HHT at the time of their first admission to the hospital, the family history revealed relatives with epistaxis or telangiectasia in all families and relatives with PAVM in 8 families (67%) (Table I). Pulmonary Investigations Laboratory and imaging findings are shown in Table II. Arterial blood gas measurements with the patient on room air revealed hypoxemia; arterial oxygen tension (PaO2) ranged from 31.5 to 88 mm Hg (mean, 56 mm Hg). Polycythemia, secondary to chronic hypoxemia, was present in 9 of the 12 children (75%). Chest radiographs were abnormal in all cases, with the most common image a round or oval mass of uniform density. In all patients, the diagnosis of PAVM was confirmed by conventional pulmonary angiography, which revealed large single lesions (15 to 50 mm diameter) in 8 patients (67%), multiple lesions in 3 patients (25%), and diffuse PAVM in only 1 patient (8%). A total of 26 discrete PAVMs were documented, excluding the diffuse PAVMs. The Journal of Pediatrics • September 2007

Long-Term Follow-Up in 12 Children with Pulmonary Arteriovenous Malformations: Confirmation of Hereditary Hemorrhagic Telangiectasia in all Cases

Table I. Clinical features of the 12 children Family members with HHT and PAVM Patient case no.

Age at PAVM diagnosis

Sex

1

F

1 day

2

M

3 4

Epistaxis*

Family member with HHT

Number

Family relationship

Clinical manifestations

Age of PAVM diagnosis (years)

Her mother (case 12) Her maternal aunt (case 10) Her maternal great-grandfather His mother His maternal great-aunt The son of the great-aunt

Bacterial meningitis Pulmonary symptoms Pulmonary symptoms Pulmonary symptoms Cerebral abscess Cerebral abscess

18 15 54 31 25 31

Her mother Her maternal grand-mother His father His grand-mother His aunt

TIA Pulmonary symptoms Cerebral abscess Pulmonary symptoms Cerebral abscess

47 45 48 50 41

Her sister C Her sister D Her father His mother

Cerebral abscess Pulmonary symptoms Cerebral abscess Pulmonary symptoms

23 25 54 33

His His His His His

Pulmonary symptoms Pulmonary symptoms Pulmonary symptoms Pulmonary symptoms Hemoptysis

65 30 31 23 18

Clinical manifestations of PAVMs Presenting symptom

Cyanosis

Dyspnea

Clubbing

Pulmonary bruit

Pulmonary symptoms

⫹

⫹

⫺

⫹

⫹ 10 years

⫹ 10 years

7

3

11 months

Pulmonary symptoms

⫹

⫺

⫺

⫹

⫹ 11 years

⫹ 12 years

8

3

F F

13 months 26 months

⫹ ⫹

⫹ ⫹

⫹ ⫹

⫹ ⫹

⫹ 13 months ⫹ 18 years

⫺ 15 years ⫹ 9 years

4 3

0 2

5

M

2.5 years

Hemoptysis Pulmonary symptoms Pulmonary symptoms

⫹

⫹ on exertion

⫺

⫹

⫹ 19 years

⫹ 10 years

4

3

6

M

6.5 years

⫹

⫺

⫺

⫹

⫹ 10 years

6

0

7

M

⫹

⫹

⫹

⫹

⫹ 10 years

⫹ 15 years

3

0

8

F

10 years and 2 months 12 years

Pulmonary symptoms Pulmonary symptoms Cerebral abscess

⫹

⫹ on exertion

⫺

⫺

⫹ 18 years

⫹ 32 years

5

3

9 10

M F

12.5 years 15 years

⫺ ⫹

⫺ ⫹

⫺ ⫹

⫺ ⫹

⫹ 12.5 years ⫹ 23 years

⫹ 9 years ⫹ 15 years

3

1

11

M

16.5 years

⫺ Pulmonary symptoms TIA

⫺

⫺

⫺

⫺

⫹ 30 years

⫹ 15 years

7

5

12

F

18 years

Cerebral abscess

⫹

⫺

⫹

⫹

⫹ 18 years

⫹ 7 years

⫹, present. *Age of occurrence.

Telangiectasia*

⫹ 7.5 years

maternal grandfather maternal aunt C maternal aunt P first cousin J first cousin F

301

302 Curie et al

Table II. Laboratory, imaging findings and molecular data on the 12 children Laboratory findings Patient case no 1 2 3 4 5 6 7

SaO2 (%)

31.5 (4.2 kPa) 70 59 78 80 60

8

The Journal of Pediatrics • September 2007

9 10 11 12

PaO2, mmHg (kPa)

50 (6.65) 40 (5.3) 32.2 (4.3)

Hb (g/dL)

Gene mutation DNA level

Protein level

Reference

Chest radiography

Chest CT scan

Pulmonary angiography

Number

Location

⫹ ⫹ ⫹ ⫹ ⫹ ⫹

* ⫹ * * * ⫹

⫹ ⫹ ⫹ ⫹ ⫹ ⫹

1 1 1 2 1 2

LLL LLL RUL RUL and LLL LLL LLL LLL

c.461delG p.Gly154fsX162 c.689⫹2T⬎C Exon skipping Not available for mutation screening c.1A⬎G p.M1 ? c.277C⬎T p.Arg93X Not available for mutation screening

19

No mutation found in ENG or ACVRL1 c.1346_1347delCT p.Ser449fxX499

Lesca et al. 2006

⫹

*

⫹

1

Lesca et al. 2004

⫹

⫹

⫹

12

c.1134_1977del c.461delG c.1522C⬎T c.461delG

Lesca Lesca Lesca Lesca

⫹ ⫹ ⫹ ⫹

⫹ ⫹ ⫹ ⫹

⫹ ⫹ ⫹ ⫹

2 1 1 1

16

83 (11.07) 88

16.2 15.7 17,6 15

LUL, left upper lobe; RUL, right upper lobe; LLL, left lower lobe; Hb, hemoglobin. ⫹, abnormal; ⫺, negative. *Not performed.

Unknown p.Gly154fsX162 p.Gln508X p.Gly154fsX162

Lesca et al. 2004 Lesca et al. 2004

PAVM

178 15 11 17 17,8 17,5

67 89 80

Imaging findings

unpublished data Lesca et al. 2006

et et et et

al. al. al. al.

2006 2004 2004 2004

LUL and right lung LLL and RUL LLL LLL LLL

Size (mm) 35 ⫻ 20 30 30 25 30 50 20 50

8 15 20 15

Table III. Treatment and long-term follow-up Follow-up Patient case no

Surgery (year)

Embolization (year)

Follow-up (years)

Recurrence of PAVM (number)

Time of recurrence after first PAVM (years)

1 2 3 4 5 6 7 8 9 10 11 12

⫹(1988) ⫹(1992) ⫹(1967) ⫹(1989) ⫹(1983) ⫹(1994) ⫹(1980) ⫺ ⫺ ⫹(1983) ⫹(1991) ⫹(1983)

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫹(1992) ⫹(2000) ⫺ ⫺ ⫺

17 14 14 16 22 4 26 13 5 22 15 22

⫹(1) ⫹(1) ⫺ ⫹(4) ⫹(5) ⫺ ⫺ ⫹(2) ⫹(3) ⫹ ⫹(2) ⫹(5)

14.5 7 ⫺ 7 16 ⫺ ⫺ 10 2 17 6 16.5

Initial treatment

Embolization (number of sessions)

Complications

⫹(1) ⫹(1) ⫺ ⫹(4) ⫹(1) ⫺ ⫺ ⫹(2) ⫹(2) ⫺ ⫹(2) ⫹(4)

⫺ ⫺ ⫺ ⫺ ⫺ ⫺ ⫺ TIA and cerebral abscess Cerebral abscess ⫺ TIA and hemoptysis Hemoptysis

Figure. Anatomic view. A, Voluminous PAVM (star) involving the right upper lobe of patient 3, with appearance of medusa head with large feeding arteries (arrows). B, Left lower lobectomy (star) of patient 7 with an obvious voluminous PAVM (arrow).

Most of the PAVMs were located in the lower lobes (20/26; 77%). PAVMs were detectable on thoracic CT scan performed in 7 cases and on CEMRA in 2 cases (patients 2 and 9).

Molecular Testing A mutation in the ENG gene was found in 9 of the 10 children available for screening. Detailed data are presented in Table II. No mutation in either ENG or ACVRL1 was found in the remaining child. Two patients (3 and 6) had been lost to follow-up when molecular screening was established in 2001. Treatment and Follow-Up Initial treatment and follow-up are described in Table III. Surgical resection was carried out in 10 patients (83%), and 2 patients underwent TCE. Surgical procedures included wedge resection and partial or complete lobectomy (Figure). Patients

8 and 9 were treated by TCE and required 4 and 3 TCE sessions, respectively. The mean follow-up period was 16 years (range, 4 to 26 years), with a total of 190 patient-years. Only 2 patients were lost to follow-up (patients 3 and 6), after 14 and 4 years, respectively. During the follow-up period, 2 patients developed TIAs (3 and 4 years after the initial treatment); patient 8 had a right arm weakness for 24 hours, and patient 11 presented with abnormal gait for 24 hours. Patient 8 had reperfused PAVMs, and patient 11 had a new PAVM; both patients underwent TCE. In patient 9, brain abscess led to the diagnosis of recurrent PAVM 2 years after initial treatment. Two patients (11 and 12) suffered from minor hemoptysis. In all, 9 patients (75%) experienced recurrence of PAVM, consisting of new feeding vessels appearing after surgery or recanalization of embolized vessels and occurring 2

Long-Term Follow-Up in 12 Children with Pulmonary Arteriovenous Malformations: Confirmation of Hereditary Hemorrhagic Telangiectasia in all Cases

303

to 17 years after the initial PAVM treatment. All of these patients were treated by 1 to 4 TCE sessions, except patient 10, who declined treatment. HHT complications unrelated to PAVM occurred in 4 patients, including multiple liver vascular malformations (patient 5) and cerebral arteriovenous fistulas (patient 6 at 8 years and patient 9 at 13 years).

DISCUSSION Comprehensive family history, careful clinical examination of relatives, and follow-up allowed us to confirm the diagnosis of HHT in all families. Clinical diagnosis of HHT is more difficult in children because of the age-related development of clinical manifestations. All of our 12 patients were carefully examined for subtle manifestations, but most did not display additional features on initial examination. This is in keeping with the classical reports on HHT, showing that only 50% to 60% of HHT patients develop epistaxis before age 20 and about 50% develop telangiectasia before age 30.12,30 Penetrance is estimated as up to 98% after 50 years.12 PAVM may occur very early in life and present with serious, sometimes life-threatening, complications. Our series includes a high proportion of early-onset forms; about half of the children were under age 3 years at the time of PAVM diagnosis. Patients were usually older in other recent series.26,27,30 Patient 1 presented with a large PAVM at birth. PAVM has occasionally been reported in newborns.19-21,31 This suggests that, at least in some patients, PAVM develops during intrauterine life. In the present series, the high proportion of symptomatic PAVM (all except 1) likely reflects the fact that HHT was not known, except in 1 family, and that pulmonary screening could not be proposed. Several recent data from genotype–phenotype correlation studies has strongly established the significantly higher frequency of PAVM as well as complications (including cerebral abscess) in patients with an ENG mutation compared with those with an ACVRL1 mutation.32-36 In our series, the fact that all mutations were found in ENG is unlikely due to a selection bias, because we and others previously showed that ACVRL1 mutations are almost twice as common as ENG mutations.29,37,38 Taken together, and in accordance with the literature, our results suggest that the predominance of PAVM in families with ENG mutations may be even higher for early-onset forms.27 Although there is no specific ENG mutation associated with PAVM, we recently suggested that truncated mutations may be associated with a greater risk of epistaxis and telangiectases in HHT patients.32 Although such a trend was also observed for PAVM, it did not achieve significance. The fact that we did not find any ENG missense mutation in the present series may provide evidence in favor of this hypothesis, but these mutations are less frequent than truncating mutations for this gene.28,29 Eight of the 12 children (67%) had a family history of PAVM. According to some authors, the prevalence of PAVM is about 15% to 20% in unselected patients with HHT, but may reach approximately 35% in HHT families in which at least 1 member has PAVM.3,6,39 This is in accor304

Curie et al

dance with recent data from our group showing that PAVM is often family-clustered, especially in families with an ENG mutation.32 Interestingly, in the present series, the children with no recurrence of PAVM also had no family history of pulmonary involvement. The treatment of choice (safe, well-tolerated, and effective) for PAVM in adults is TCE using coils or other intravascular devices.1,40 – 42 TCE is performed for all PAVMs with feeding arteries of at least 3 mm in diameter and has dramatically limited the need for surgery. The first large series of children treated with TCE has recently been published,26 and the improvement of medical survey in children from HHT family members should lead to the increased use of this technique. Antibiotic prophylaxis also should be recommended for all bacteremic procedures, such as dental work, once PAVM has been diagnosed,6,26 to prevent cerebral abscess. The follow-up period in our series (mean, 16 years) is the longest reported so far, and allowed us to detect 5 recurrences of PAVM occurring more than 10 years after initial treatment. The high recurrence rate (75%) might be related in part to the improved imaging techniques for detection of PAVMs during the follow-up period (some of the PAVMs may have been initially missed), but it raises concerns about the risk of developing novel PAVMs during childhood and adolescence and emphasizes the need for long-term follow-up. Because PAVM may present with life-threatening complications early in life, and because effective treatment has been described, we advocate screening for PAVM in children of HHT families. Pulmonary screening protocols have been proposed for adults.1,6,15,43,44 We have recently reported that 100% sensitivity and negative predictive value could be obtained in adults when combining anteroposterior chest radiographs with contrast echocardiography.39 A screening algorithm based on the combined use of both tests, followed by chest CT if either test is positive, was suggested. An alternative is screening directly by chest CT. But the algorithm may obviate the need for chest CT in patients without PAVM (the majority of HHT patients). The need to avoid unnecessary chest CT in children is obvious because of radiation-induced cancer.45 Contrast echocardiography is often suggested as a screening tool in children,26 and a preliminary study on a small group of children was reported recently.46 Further studies are needed to determine the most appropriate screening strategy for children. The fact that chest radiography allowed detection of large PAVMs in our patients as well as in those reported earlier suggests that it should be performed at least once during childhood, even in asymptomatic children of HHT families, to detect large PAVMs.1,6,8,9,17 Physical examination and eventually pulse oximetry may be useful to disclose cyanosis, which is often noted in published cases of PAVM in children. Contrast echocardiography seems to be the most appropriate screening test in children. As in adults, definite diagnosis of PAVM relies on chest CT or pulmonary angiography when TCE is indicated. Despite the fact that most mutations in ENG and ACVRL1 are private (ie, different from one family to another), the recent development of effective molecular techThe Journal of Pediatrics • September 2007

niques will allow physicians to use genetic data for clinical screening of patients from HHT families. Once a pathogenic mutation has been identified in the index case, genetic testing of the other family members is straightforward. Keeping in mind that genetic testing in children should be combined with appropriate counseling regarding medical, ethical, and psychological aspects, this approach may reduce the cost of conventional clinical screening and long-term follow-up, which will be focused on mutation carrier family members.47 In conclusion, PAVM may occur very early in life, and serious, sometimes life-threatening, complications may result if they are left untreated. The diagnosis of HHT should be considered in a child with apparently isolated PAVM. Children of HHT families should benefit from PAVM screening and long-term follow-up, especially in families with an ENG mutation or a history of PAVM. We thank Jean-Pierre Pracros for his helpful comments and Daniel Floret, Benoît Guibert, and Gérard Champsaur for referring some of the families.

REFERENCES 1. Swanson KL, Prakash UB, Stanson AW. Pulmonary arteriovenous fistulas: Mayo clinic experience, 1982-1997. Mayo Clin Proc 1999;74:671-80. 2. Ference BA, Shannon TM, White RI Jr, Zawin M, Burdge C. Life-threatening pulmonary hemorrhage with pulmonary arteriovenous malformations and hereditary hemorrhagic telangiectasia. Chest 1994;106:1387-90. 3. Gossage JR, Kanj G. Pulmonary arteriovenous malformations: a state-of-the-art review. Am J Respir Crit Care Med 1998;158:643-61. 4. White RI Jr, Lynch-Nyhan A, Terry P, Buescher PC, Farmlett EJ, Charnas L, et al. Pulmonary arteriovenous malformations: techniques and long-term outcome of embolotherapy. Radiology 1988;169:663-9. 5. Kjeldsen AD, Oxhoj H, Andersen PE, Green A, Vase P. Prevalence of pulmonary arteriovenous malformations (PAVMs) and occurrence of neurological symptoms in patients with hereditary haemorrhagic telangiectasia (HHT). J Intern Med 2000;248:255-62. 6. White RI Jr. Pulmonary arteriovenous malformations: how do we diagnose them and why is it important to do so? Radiology 1992;182:633-5. 7. White RI Jr, Pollak JS. Pulmonary arteriovenous malformations: diagnosis with three-dimensional helical CT. A breakthrough without contrast media. Radiology 1994;191:613-4. 8. Dines DE, Arms RA, Bernatz PE, Gomes MR. Pulmonary arteriovenous fistulas. Mayo Clin Proc 1974;49:460-5. 9. Dines DE, Seward JB, Bernatz PE. Pulmonary arteriovenous fistulas. Mayo Clin Proc 1983;58:176-81. 10. Mansour KA, Hatcher CR Jr, Logan WD Jr, Abbott OA. Pulmonary arteriovenous fistula. Am Surg 1971;37:203-8. 11. Pick A, Deschamps C, Stanson AW. Pulmonary arteriovenous fistula: presentation, diagnosis and treatment. World J Surg 1999;23:1118-22. 12. Plauchu H, de Chadarevian J-P, Bideau A, Robert J-M. Age-related clinical profile of hereditary hemorrhagic telangiectasia in an epidemiologically recruited population. Am J Med Genet 1989;32:291-7. 13. Shovlin CL, Guttmacher AE, Buscarini E, Faughnan ME, Hyland RH, Westermann CJJ, et al. Diagnostic criteria for hereditary hemorrhagic telangiectasia (RenduOsler-Weber syndrome). Am J Med Genet 2000;91:66-7. 14. Nanthakumar K, Graham AT, Robinson TI, Grande P, Pugash RA, Clarke JA, et al. Contrast echocardiography for detection of pulmonary arteriovenous malformations. Am Heart J 2001;141:243-6. 15. Haitjema T, Disch F, Overtoom TT, Westermann CJ, Lammers JW. Screening family members of patients with hereditary hemorrhagic telangiectasia. Am J Med 1995;99:519-24. 16. Higgins CB, Wexler L. Clinical and angiographic features of pulmonary arteriovenous fistulas of children. Radiology 1976;119:171-5. 17. Mitchell RO, Austin EH III. Pulmonary arteriovenous malformation in the neonate. J Pediatr Surg 1993;28:1536-8. 18. Bennhagen RG, Holje G, Laurin S, Pesonen E. Coil embolization of a neonatal pulmonary arteriovenous malformation. Pediatr Cardiol 2002;23:235-8.

19. Grady RM, Sharkey AM, Bridges ND. Transcatheter coil embolisation of a pulmonary arteriovenous malformation in a neonate. Br Heart J 1994;71:370-1. 20. Taylor GA. Pulmonary arteriovenous malformation: an uncommon cause for cyanosis in the newborn. Pediatr Radiol 1983;13:339-41. 21. Clarke CP, Goh TH, Blackwood A, Venables AW. Massive pulmonary arteriovenous fistula in the newborn. Br Heart J 1976;38:1092-5. 22. Sommer B, Kaufmann HJ, Kumm M. Pulmonary arteriovenous fistula: ultrasonographic approach. Pediatr Radiol 1990;20:353-4. 23. Audenaert SM, Wood BP. Radiological case of the month: pulmonary arteriovenous fistula associated with Rendu-Osler-Weber syndrome. Am J Dis Child 1990; 144:575-6. 24. Oliveira GH, Seward JB, Cortese DA, Dines DE. Contrast transesophageal echocardiography in the diagnosis and localization of diffuse pulmonary telangiectasias. Chest 2000;118:557-9. 25. Meredith JT. Cyanosis due to pulmonary arteriovenous malformation. Am Fam Physician 1988;38:187-91. 26. Faughnan ME, Thabet A, Mei-Zahav M, Colombo M, Maclusky I, Hyland RH, et al. Pulmonary arteriovenous malformations in children: outcomes of transcatheter embolotherapy. J Pediatr 2004;145:fbib28 826-31. 27. Mei-Zahav M, Letarte M, Faughnan ME, Abdalla SA, Cymerman U, MacLusky IB. Symptomatic children with hereditary hemorrhagic telangiectasia. Arch Pediatr Adolesc Med 2006;160:596-601. 28. Lesca G, Plauchu H, Coulet F, Lefebvre S, Plessis G, Odent S, et al. Molecular screening of ALK1/ACVRL1 and ENG genes in hereditary hemorrhagic telangiectasia in France. Hum Mutat 2004;23:289-99. 29. Lesca G, Burnichon N, Raux G, Tosi M, Pinson S, Marion MJ, et al. Distribution of mutations in ENG and ACVRL1 genes in HHT French patients. Hum Mutat 2006;27:598. 30. Folz BJ, Zoll B, Alfke H, Toussaint A, Maier RF, Werner JA. Manifestations of hereditary hemorrhagic telangiectasia in children and adolescents. Eur Arch Otorhinolaryngol 2006;263:53-61. 31. Koppen SKC, Dalinghaus M, Westermann CJ. Neonatal pulmonary arteriovenous malformation in hereditary haemorrhagic telangiectasia. Arch Dis Child Fetal Neonatal Ed 2002;87:F226-7. 32. Lesca G, Olivieri C, Burnichon N, Pagella F, Carette MF, Gilbert-Dussardier B, et al. Genotype-phenotype correlations in hereditary hemorrhagic telangiectasia: data from the French-Italian HHT network. Genet Med 2007;9:14-22. 33. Berg J, Porteous M, Reinhardt D, Gallione C, Holloway S, Umasunthar T, et al. Hereditary haemorrhagic telangiectasia: a questionnaire-based study to delineate the different phenotypes caused by endoglin and ALK1 mutations. J Med Genet 2003;40:585-90. 34. Bayrak-Toydemir P, McDonald J, Markewitz B, Lewin S, Miller F, Chou LS, et al. Genotype–phenotype correlation in hereditary hemorrhagic telangiectasia: mutations and manifestations. Am J Med Genet 2006;140:463-70. 35. Letteboer TGW, Mager JJ, Snijder RJ, Koeleman BPC, Lindhout D, Ploos van Amstel JK, et al. Genotype–phenotype relationship in hereditary haemorrhagic telangiectasia. J Med Genet 2006;43:371-7. 36. Kjeldsen AD, Moller TR, Brusgaard K, Vase P, Andersen PE. Clinical symptoms according to genotype amongst patients with hereditary haemorrhagic telangiectasia. J Intern Med 2005;258:349-55. 37. Wehner LE, Folz BJ, Argyriou L, Twelkemeyer S, Teske U, Geisthoff UW, et al. Mutation analysis in hereditary haemorrhagic telangiectasia in Germany reveals 11 novel ENG and 12 novel ACVRL1/ALK1 mutations. Clin Genet 2006;69:239-45. 38. Olivieri C, Mira E, Delu G, Pagella F, Zambelli A, Malvezzi L, et al. Identification of 13 new mutations in the ACVRL1 gene in a group of 52 unselected Italian patients affected by hereditary haemorrhagic telangiectasia. J Med Genet 2002;39:E39. 39. Cottin V, Plauchu H, Bayle JY, Barthelet M, Revel D, Cordier JF. Pulmonary arteriovenous malformations in patients with hereditary hemorrhagic telangiectasia. Am J Respir Crit Care Med 2004;169:994-1000. 40. Haitjema TJ, Overtoom TTC, Westermann CJJ, Lammers J-WJ. Embolisation of pulmonary arteriovenous malformations: results and follow up in 32 patients. Thorax 1995;50:719-23. 41. Andersen PE, Kjeldsen AD, Oxhoj H, Vase P, White RI Jr. Embolotherapy for pulmonary arteriovenous malformations in patients with hereditary hemorrhagic telangiectasia (Rendu-Osler-Weber syndrome). Acta Radiol 1998;39:723-6. 42. Dutton JAE, Jackson JE, Huhes JMB, Whyte MKB, Peters AM, Ussov W, et al. Pulmonary arteriovenous malformations: results of treatment with coil embolization in 53 patients. Am J Roentgenol 1995;165:1119-25. 43. Kjeldsen AD, Oxhoj H, Andersen PE, Elle B, Jacobsen JP, Vase P. Screening procedures and pulmonary angiography in patients with hereditary hemorrhagic telangiectasia. Chest 1999;116:432-9.

Long-Term Follow-Up in 12 Children with Pulmonary Arteriovenous Malformations: Confirmation of Hereditary Hemorrhagic Telangiectasia in all Cases

305

44. Westermann CJ, Rosina AF, de Vries V, de Coteau PA. The prevalence and manifestations of heriditary hemorrhagic telangiectasia in the Afro-Caribbean population of the Netherlands Antilles: a family screening. Am J Med Genet 2003; 116A:324-8. 45. Brenner D, Elliston C, Hall E, Berdon W. Estimated risks of radiation-induced fatal cancer from pediatric CT. AJR Am J Roentgenol 2001;176:289-96.

46. Giordano P, Nigro A, Lenato GM, Guanti G, Suppressa P, Lastella P, et al. Screening for children from families with Rendu-Osler-Weber disease: from geneticist to clinician. J Thromb Haemost 2006;4:1237-45. 47. Cohen JH, Faughnan ME, Letarte M, Vandezande K, Kennedy SJ, Krahn MD. Cost comparison of genetic and clinical screening in families with hereditary hemorrhagic telangiectasia. Am J Med Genet 2005;137A:153-60.

50 Years Ago in The Journal of Pediatrics NIEMANN-PICK

DISEASE IN A BOY OF 16 MONTHS

Pansky B, Lee R. J Pediatr 1957;51:290-9

In an age in which protein purification and demonstration of human enzymatic activities were still among the most challenging scientific goals, Pansky and Lee described the effect on the electrophoretic mobility of serum proteins of “therapy” in a patient diagnosed with Niemann-Pick disease. The therapies offered were a series of biologicals including vitamins A and B12, adrenocorticotrophic hormone, and various antibiotics. The serum proteins were identified solely by their electrophoretic mobilities, and the clinical diagnosis was based on undefined “chemical, histological, physical and marrow findings.” The physician reader 50 years later may derive comfort from the progress that medical science has made not only in accurately defining the diagnostic enzymatic deficiencies of many inherited disorders, among them several forms of Niemann-Pick disease (A, B, and C), but also in having available an enzyme replacement therapy, albeit in human clinical trials, for sphingomyelinase deficiency, the classic form of Niemann-Pick disease. A generation of clinical biochemists carefully elaborated the lysosomal enzyme deficiencies that cause more than 40 lysosomal storage disorders (LSD). A succeeding generation of molecular biologists sequenced the genes that encode these lysosomal enzymes and produced animal models that led to the current understanding of lysosomal storage disease pathophysiology. Although bone marrow transplantation demonstrated long-term efficacy in some patients with LSD, morbidity and mortality in these patients encouraged the search for alternative treatments. From early experiments treating patients with nephropathic cystinosis using cysteamine, a thiol analogue used previously in high-altitude pilots to ameliorate radiation exposure, numerous therapeutic approaches have followed to offer success in decreasing lysosomal storage and conferring clinical benefits. Following the FDA approval in 1992 of glucocerebrosidase purified from human placentas (ceredase), recombinant technology has allowed high-throughput synthesis of an enzyme analogue (cerezyme) in a bioreactor using Chinese hamster ovary cells. These therapies offered the first effective treatments for an intralysosomal enzyme deficiency causing an LSD. Enzyme replacement therapies for Fabry disease, mucopolysaccharidoses types I (Hurler), II (Hunter), VI (Maroteaux-Lamy syndrome), and Pompe’s disease have been FDA approved, and several other disorders, among them Niemann-Pick A/B, are currently in clinical trials. Alternative therapeutic approaches (miglustat) inhibit the biosynthesis of the glycosphingolipid, which accumulates in Gaucher disease, and offer clinical benefits for patients in whom enzyme replacement therapy is not an option. The field of lysosomal storage diseases has been reinvigorated by treatment options born of decades of scientific and clinical investigation. The pediatrician’s recognition of these treatable disorders at an early age is critical to long-term outcome. Hans C. Andersson, MD, FACMG Director, Hayward Genetics Center Tulane University Medical School New Orleans, Lousiana 10.1016/j.jpeds.2007.04.025

306

Curie et al

The Journal of Pediatrics • September 2007

Duodenogastro-Esophageal Reflux in Children with Refractory Gastro-Esophageal Reflux Disease ILSE HOFFMAN, MD, ALEXANDER TERTYCHNYY, MD, NADINE ECTORS, PHD, TOON DE GREEF, NANCY HAESENDONCK, AND JAN TACK, PHD

Objective To determine the role of duodenogastro-esophageal reflux (DGER) in the pathogenesis of refractory gastroesophageal reflux disease (GERD) in children. Study design Twenty-two patients (12 boys, mean age, 13.2 years) with GERD symptoms that persisted on omeprazole (1 mg/kg) underwent upper gastrointestinal endoscopy and barium x-ray, 24-hour pH and DGER (Bilitec) monitoring, and a 13C octanoic acid gastric emptying breath test. Results Patients presented mainly with epigastric pain, regurgitation, and nausea. Endoscopy revealed persistent esophagitis in 15 patients (68%). Pathologic acid and DGER exposure were present in 12 (55%) and 15 (68%) children, respectively, with combined pathologic reflux in 10 (45%). Acid exposure did not differ according to the presence of esophagitis, but patients with grade II esophagitis had significantly higher DGER exposure than those without esophagitis (9.1 ⴞ 5.3% vs 26.7 ⴞ 10.9% of the time, P < .05). Gastric emptying rate was not associated to acid or DGER exposure or persisting esophagitis. Symptoms improved after adding a prokinetic drug to the proton pump inhibitor therapy or referral for surgery (n ⴝ 5). Conclusions DGER may play a role in the pathophysiology of proton pump inhibitor–refractory GERD and esophagitis in children. (J Pediatr 2007;151:307-11) astro-esophageal reflux disease (GERD) is defined by the presence of symptoms and lesions that can be attributed to reflux of gastric contents to the esophagus.1 The pressure of the lower esophageal sphincter, the motility of the esophageal body and the stomach, the composition of the reflux material, and the sensitivity or resistance of the esophageal mucosa to the reflux material are important factors involved in the pathogenesis of GERD-related symptoms and lesions.2 The refluxate is not only composed of gastric acid and pepsin but may also contain food and regurgitated duodenal contents.3,4 In adults, reflux of duodenal contents into the stomach is a physiological event, both postprandially and at night.5-7 Hence, regurgitation of duodenal contents through the pylorus into the stomach, with following reflux into the esophagus, which is called duodenogastroesophageal reflux (DGER), is not unusual. The role of DGER has initially been evaluated by means of endoscopy with biopsies,8,9 scintigraphy,10 aspiration studies,11 and esophageal pH monitoring. The terms bile reflux and alkaline reflux have been used to describe DGER,4 but it is now well established that alkaline reflux that is found on pH monitoring is not equivalent with DGER but rather seems to reflect swallowed saliva and esophageal mucosal bicarbonate secretion.12,13 The Bilitec 2000 (Synectics Medical, Stockholm Sweden) device is a fiberoptic From the Division of Pediatrics (I.H.), Despectrophotometric probe developed to quantify DGER in an ambulatory setting. In vitro partment of Gastroenterology, the Department of Pathology (N.E.), and the Division validation studies confirmed a good correlation between the total bilirubin concentration of Internal Medicine (T.D.G., N.H., J.T.), Deof aspirated samples and the fiberoptic reading of bilirubin concentration.14 Moreover, a partment of Gastroenterology, University Hospitals Leuven, Belgium; and the Departgood correlation was found between total bilirubin content and the concentrations of ment of Pediatric Pathology (A.T.), Russian pancreatic enzymes in aspirated refluxate.14 Based on these observations, bilirubin seems State Medical University, Moscow, Russia. to be an accurate tracer for DGER monitoring in patients who have normal serum Submitted for publication Oct 4, 2006; last revision received Jan 22, 2007; accepted bilirubin levels. Mar 16, 2007. In adults, esophageal exposure to acid and to DGER has been extensively studied Reprint requests: Dr Ilse Hoffman, Univer15 by using combined pH and Bilitec monitoring in normal subjects and patients with sity Hospital Leuven-Gasthuisberg, Pediat16-18 rics, Herestraat 49, 3000 Leuven, Belgium. GERD. Both acid reflux and DGER show a graded increase in severity across the

G

DGER GERD

Duodenogastro-esophageal reflux Gastro-esophageal reflux disease

PPI

Proton pump inhibitor

E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.024

307

GERD spectrum, and they occur simultaneously in the majority of the reflux episodes. DGER has been identified as a factor that contributes to GERD that is refractory to acid suppressive therapy.16,19-21

METHODS Study Population We evaluated consecutive children with refractory GERD symptoms despite omeprazole treatment (1 mg/kg). All children and their parents were asked to stop any medication known to affect gastrointestinal motility or secretion for at least 2 weeks before the study. Inclusion criteria were the presence of GERD symptoms despite the intake of omeprazole, at appropriate doses for at least 2 months, started after an abnormal endoscopy and/or pH-monitoring that established GERD. Exclusion criteria were organic, systemic, metabolic, or neurologic disease and an underlying psychiatric illness. Therefore, all patients underwent a careful history taking, clinical examination, routine biochemistry, and an abdominal ultrasound. A barium swallow was performed to rule out a hiatal hernia, malrotation, or any other anatomic abnormality. Questionnaire Before the pH and DGER monitoring study, each patient and his or her parents completed a questionnaire, which evaluated the presence of different typical and atypical reflux symptoms (heartburn, regurgitation, epigastric pain, chest pain, night-time pain, nausea, vomiting, belching, bloating, anorexia, aspiration, throat ache, hoarseness, coughing, and hypersalivation). Esophagogastroduodenoscopy with Biopsies A fiberoptic esophagogastroduodenoscopy with biopsies was performed in all patients, after sedation with 0.1 mg/kg midazolam (with a maximum of 5 mg) and 1 mg/kg pethidine (with a maximum of 10 mg) intravenously. The type and location of the mucosal injury were documented and the severity of the damage was graded according to the Savary Miller classification,22 which was still the basis for proton pump inhibitor (PPI) reimbursement in Belgium at the time of the study. During the esophagogastroduodenoscopy, biopsies were obtained from the duodenum, the stomach (antrum), and the distal esophagus.23,24 Ambulatory pH and DGER Monitoring A Digitrapper MK III (Synectics Medical) with unipolar antimony electrodes calibrated in test solutions (pH 7 and 1) was used for the pH monitoring. The Bilitec 2000 (Synectics Medical), a fiberoptic probe, was used for DGER monitoring. The fiberoptic probe and the pH probe were taped together at the tip. This assembly was introduced through the nose and positioned under fluoroscopic control two vertebrae above the diaphragm.25 During the study, pa308

Hoffman et al

tients received no solid food, fruit juice, or coffee but a fluid food Nutridrink with vanilla taste (Nutridrink, Nutricia, Bornem, Belgium).26 They were admitted to the hospital for the duration of the study but were encouraged to carry out daily activities. Patients and their parents were asked to mark the sleeping, eating, and symptom periods by pressing a button on the data recorder. The ambulatory recording data were downloaded onto a personal computer and analyzed with the aid of commercially available software (Gastrosoft Inc, Synetics Medical, Irvine, TX). Acid reflux was quantified with the following variables obtained from computerized analysis: number of reflux episodes, number of reflux episodes lasting longer than 5 minutes per hour, and percentage of time with pH ⱕ4. DGER was quantified with the following variables obtained from computerized analysis: number of reflux episodes, number of reflux episodes lasting longer than 5 minutes per hour, and percentage of time with bilirubin absorbance ⱖ0.14. The recording was divided into meal, postprandial (2 hours after a meal), interdigestive, upright, and supine periods. A patient is considered to have pathologic DGER if the fraction of time that the esophageal mucosa is exposed to a refluxate with a bilirubin absorbance of ⬎0.14 exceeds 4.2% of the total study time.27 Pathologic acid reflux is considered to be present if the fraction of time that the esophageal mucosa is exposed to a refluxate with pH ⬍4 exceeds 4% of the recording time.28

Gastric Emptying Breath Test Gastric emptying rate was studied by means of the noninvasive 13C octanoic breath test. Octanoic acid, a medium-chain fatty acid marked with a stable 13C isotope, is rapidly absorbed from the duodenum. Because gastric emptying is the rate-limiting step for the absorption of medium-chain fatty acids, the fraction of 13C expired in the breath is a measure for the rate of gastric emptying. The 13C octanoic acid was mixed in a pancake and presented for breakfast after obtaining two basal breath test samples. After feeding, breath samples were collected every 15 minutes during 4 hours. Analysis of the expired 13C fraction in the breath samples was performed by using isotope-ratio mass spectrometry (Gilson ABCA 20-20 stable isotope analyzer and autosampler; Europa Scientific, Crewe, UK), and the gastric emptying time was calculated as previously reported.29 Statistical Analysis Values are expressed as mean and standard error (SEM) or median and interquartile ranges. Results were compared by using the Student t test, Mann-Whitney U test, or ␹2 test wherever appropriate. Probability values were considered to be significant at ⬍.05. The Journal of Pediatrics • September 2007

Table I. Distribution of symptoms

Table II. Results of pH and DGER monitoring

Symptom

No. (%) of patients

Epigastric pain Regurgitation Nausea Retrosternal pain Nightly pain Anorexia Hoarseness Vomiting Throat ache Coughing Bloating Belching Aspiration Hypersalivation

19 (86%) 16 (73%) 14 (64%) 8 (36%) 7 (32%) 6 (27%) 5 (23%) 5 (23%) 4 (18%) 4 (18%) 2 (9%) 2 (9%) 0 (0%) 0 (0%)

Total

Upright

Supine

% Time pH ⬍4 5.9 ⫾ 1.4 6.5 ⫾ 1.3 5.3 ⫾ 2.1 No. of acid reflux episodes 77 ⫾ 14 64 ⫾ 12 12 ⫾ 3 No. of long-lasting acid reflux 2 ⫾ 0.6 1 ⫾ 0.3 1 ⫾ 0.4 episodes Longest acid reflux episode 22 ⫾ 6 8⫾3 20 ⫾ 7 (min) % Time bilirubin absorbance 12.6 ⫾ 3.1 12.1 ⫾ 2.6 13.6 ⫾ 5.3 ⬎0.14 No. of DGER episodes 31 ⫾ 7 25 ⫾ 6 6⫾2 No. of long-lasting DGER 4 ⫾ 0.9 3 ⫾ 0.5 2 ⫾ 0.6 episodes Longest DGER episode (min) 82 ⫾ 29 38 ⫾ 13 50 ⫾ 24

Data are represented as number/row percentage.

RESULTS Study Population Twenty-two children (12 girls; mean age, 13.2 ⫾ 2.2 years) with symptoms suggestive of refractory GERD despite at least 2 months of omeprazole therapy (1 mg/kg) were included. Twenty-one children had an abnormal endoscopy before the PPI treatment (grade 1 esophagitis in 7 patients, grade 2 esophagitis in 10 patients, gastritis in 12 patients, and antral erosions in 5 patients). One patient had a normal endoscopy with biopsies but an abnormal pH monitoring (8.2% of time pH ⬍4). The mean time between the start of the PPI treatment and the DGER monitoring because of symptoms suggestive of PPI refractory GERD was 3.5 months. Questionnaire The most prevalent symptoms were epigastric pain, regurgitation, and nausea (Table I). Heartburn is not typically reported by children with GERD, but night-time pain and retrosternal pain were reported in 32% and 36%, respectively, of the patients. Barium Swallow All patients underwent a barium x-ray examination of the esophagus, stomach, and duodenum. The barium study did not reveal any anatomic abnormalities except for one child with a small intermittent sliding hernia. In one child with a history of infantile pyloric hypertrophy (at the age of 2 months), the barium examination did not reveal a recurrent pyloric stenosis, and barium emptied normally. Esophagogastroduodenoscopy and Biopsies All patients underwent a new esophagogastroduodenoscopy. Eight patients had a grade I esophagitis and 4 had a grade II esophagitis. Gastritis was noted in 8 children, 3 children had gastric erosions, and duodenal mucosal hy-

Figure 1. Distribution of findings on combined reflux monitoring in 22 children with PPI-refractory GERD symptoms. Graph shows the proportion of patients with pathologic acid exposure or pathologic DGER exposure or both.

peremia was noted in one child. During the endoscopy, biopsies were obtained from the duodenum, stomach, and distal esophagus. Biopsies showed esophagitis in 3 endoscopynegative patients, Helicobacter pylori–negative gastritis in 9 patients (5 chronic focal gastritis, 4 reactive gastritis), and duodenitis in 1 patient. No other inflammatory or allergic disorders were noted. In total, endoscopic or histologic esophagitis was demonstrated in 15 patients (68.1%).

Acid and Bile Reflux Testing The results of esophageal pH and DGER monitoring are listed in Table II. The mean esophageal acid exposure was slightly elevated, and pathologic esophageal acid exposure was present in 12 children (55%). Average esophageal DGER exposure was highly elevated, and 15 patients (68%) had pathologic DGER exposure. Combined pathologic acid and DGER exposure was present in 45% of the children (n ⫽ 10), although 2 (9%) had pathologic acid exposure alone and 5 (23%) had pathologic DGER alone. No pathologic reflux was demonstrated in 5 (23%) children (Figure 1). Acid exposure and DGER exposure were not significantly correlated (r ⫽ 0.05, NS). Gastric emptying rate did

Duodenogastro-Esophageal Reflux in Children with Refractory Gastro-Esophageal Reflux Disease

309

pathologic DGER, and the PPI and prokinetic treatment could be stopped.

DISCUSSION

Figure 2. Acid exposure and DGER exposure in 22 children with reflux symptoms poorly responsive to PPI therapy, according to the absence or presence of grade 1 or grade 2 esophagitis. Acid exposure did not differ significantly between the groups. Compared with patients without esophagitis, patients with grade 2 esophagitis had significantly higher DGER exposure (*P ⬍ .05).

not differ between patients with or without pathologic acid exposure (84 ⫾ 24 vs 86 ⫾ 26 minutes, NS) or with or without pathologic DGER exposure (105 ⫾ 47 vs 76 ⫾ 24 minutes, NS).

Relation Between Endoscopic Esophagitis and Acid or DGER Exposure Esophageal acid exposure and esophageal DGER did not differ between patients with or without esophagitis (Figure 2). However, patients with grade II esophagitis had significantly higher DGER exposure than patients without esophagitis (9.1% ⫾ 5.3% vs 26.7% ⫾ 10.9% of the time, P ⬍ .05). Similarly, the duration of the longest DGER episode was significantly longer in patients with grade II esophagitis compared with patients with grade I esophagitis and patients without esophagitis (242 ⫾ 132 vs 28 ⫾ 18 and 61 ⫾ 16 minutes, respectively, P ⬍ .05). Relation Between Symptoms and Acid or DGER Exposure There was no statistical difference in symptom pattern between acid, DGER, or combined reflux patients. In all groups, the most prevalent symptoms were regurgitation, epigastric pain, and nausea. Clinical Outcomes After the investigation, initial therapy consisted of prokinetic drug therapy (cisapride or domperidone) in all children. Most children with DGER had better symptom control on a combination of high-dose PPI and prokinetic drug therapy. Eventually, 4 of the children with pathologic DGER were referred for antireflux surgery. One of the children without pathologic DGER was also referred for antireflux surgery because of pathologic acid reflux that persisted during adequate PPI and prokinetic drug therapy. The Nissen fundoplication resolved all symptoms in the group with the 310

Hoffman et al

The term bile reflux is often used synonymously with DGER because bile acids or bilirubin are the constituents used most often as markers. Studies in the adult population have demonstrated a progressive increase in esophageal exposure to acid and duodenal contents across the spectrum of GERD, with a particularly high prevalence in patients with Barrett esophagus.30,31 Although DGER is also suppressed by PPIs, available literature suggests that PPIs are less efficacious to normalize DGER, compared with their effect on acid reflux.32,33 In adults, DGER has been associated with less complete responsiveness to PPI therapy34 and with persistence of symptoms and lesions during PPI therapy.20 In the past, it has been suggested that an alkaline shift on esophageal pH monitoring might reflect exposure to duodenal content in pediatric GERD.35 However, combined aspiration and pH monitoring studies in adults have shown that this is not the case,36 and the Bilitec 2000 fiberoptic spectrophotometric probe is now considered the gold standard for the evaluation of DGER.37 One previous study evaluated the use of combined pH and Bilitec monitoring in children with GERD symptoms and found that both bile and acid reflux increased with the severity of esophagitis in children, too.38 In the present study, we assessed esophageal acid and DGER exposure in consecutive pediatric patients with GERD that was refractory to standard PPI therapy. The refractoriness is confirmed by the finding of ongoing or recurrent endoscopic esophagitis after a brief interruption of drug intake, despite at least 2 months of PPI therapy in the majority of the children. For ethical reasons, performing DGER monitoring in healthy children is not feasible, and hence normal values for the pediatric population are lacking. We therefore used the upper limit of normal values previously obtained in healthy adults to define pathologic DGER exposure. Using this cutoff, we found pathologic esophageal DGER exposure in 68% of the children, and one third did not have associated pathologic acid exposure. More severe grades of esophagitis were associated with significantly higher DGER exposure, also suggesting the pathophysiologic relevance of DGER in GERD refractoriness in these patients. The present study has a number of limitations. First of all, due to the setting of this study in a tertiary referral center, this is a highly selected refractory patient population, and the true prevalence of pediatric refractory GERD remains unclear. In the adult population, refractory heartburn and regurgitation are cardinal clinical symptoms suggestive of refractory GERD. In the present study of pediatric refractory GERD, the most frequently reported symptoms were epigastric pain, regurgitation, and nausea. Studies in adults have suggested that symptoms of heartburn and regurgitation occur in relation to both acid or bile reflux, and less typical symptoms such as nausea and vomiting were related to bile reflux alone.39 The Journal of Pediatrics • September 2007

Second, as the use of solid food during Bilitec monitoring is associated with a high prevalence of meal-impaction artifacts,40 patients were put on a liquid diet during the study. This is a deviation from normal feeding patterns, and the lack of solid foods may have led to an underestimation of some of the acid and DGER measures. Third, as indicated above, normal values derived in this study were derived from an adult and not a pediatric control population. Finally, and most importantly, as this group of children with refractory GERD was studied after a brief interruption of PPI therapy and in the absence of control DGER data in children whose symptoms responded well to PPIs, we can only speculate on the contribution of DGER to PPI refractoriness in pediatric patients. However, the association of more severe reflux esophagitis with higher DGER exposure is supportive of a pathophysiologic role for DGER in refractoriness. Furthermore, the clinical improvement after addition of prokinetic therapy, or referral for surgery, would also support a role for DGER in ongoing symptoms and lesions despite PPI therapy. Indeed, previous studies in adults have confirmed the value of prokinetic therapy41,42 or surgery34 in the treatment of DGER. Given its high prevalence in the patient group we studied, nonacid reflux or DGER should be considered in children with refractory symptoms despite adequate PPI therapy. Future studies seem warranted to investigate acid and DGER exposure on PPI therapy and to investigate the response to prokinetic or surgical therapy in children with established pathologic DGER.

REFERENCES 1. Kahrilas PJ, Quigley FM. Clinical esophageal pH recording: a technical review for practice guideline development. Gastroenterology 1996;110:1982-96. 2. Okholm M, Sorensen H, Wallin L, Boeby S. Bile reflux into the esophagus. Scand J Gastroenterol 1999;34:653-7. 3. Vaezi MF, Singh S, Richter JE. Role of acid and duodenogastric reflux in esophageal injury: a review of animal and human studies. Gastroenterology 1995; 108:1897-907. 4. Vazi MF. Ballière’s Clinical Gastroenterology 2000 Vol. 14, Nov 4: pp 719-29. 5. Cuomo R, Koek G, Sifrim D, Janssens J, Tack J. Analysis of ambulatory duodenogastroesophgeal reflux monitoring. Dig Dis Sci 2000;45:2463-9. 6. Koek G, Vos R, Sifrim D, Cuomo R, Janssens J, Tack J. Mechanisms underlying duodeno-gastric reflux in man. Neurogastroenterol Motil 2005;17:191-9. 7. Mearin F, Azpiroz F, Malagelada JR, Zinsmeister AR. Antroduodenal resistance to flow in the control of duodenogastric bile reflux during fasting. Gastroenterology 1987;93:1026-33. 8. Nasrallah SM, Johnston GS, Gadacz TR, Kim KM. The significance of gastric bile reflux at endoscopy. J Clin Gastroenterol 1987;9:514-7. 9. Stein HJ, Smyrck TC, DeMeester TR, et al. Clinical value of endoscopy and histology in the diagnosis of duodenogastric reflux disease. Surgery 1992;112:796-804. 10. Drane WE, Karvelis K, Johnson DA, et al. Scintigraphic evaluation of duodenogastric reflux: problems, pitfalls and technical review. Clin Nucl Med 1987;12:377-84. 11. Johnsson F, Joelsson B, Floren CH, et al. Bile salts in the esophagus of patients with esophagitis. Scand J Gastroenterol 1988;23:712-6. 12. Singh S, Bradley LA, Richter JE. Determinants of esophageal alkaline pH environment in controls and patients with gastroesophageal reflux disease. Gut 1993;34: 309-16. 13. Devault KR, Georgson S, Castell DO. Salivary stimulation mimics esophageal exposure to refluxed duodenal contents. Am J Gastroenterol 1993;88:1040-3. 14. Stipa F, Stein HJ, Feussner H, Kraemer S, Siewert JR. Assessment of non-acid esophageal reflux:comparison between long-term reflux aspiration test and fiberoptic bilirubin monitoring. Dis Esophagus 1997;10:24-8. 15. Byre J, Romagnoli R, Bechi P, Attwood S, Fuchs K, Collard JM. Duodenogastric reflux of bile in health: the normal range. Physiol Meas 1999;20:149-58.

16. Koek G, Tack J, Sifrim D, Lerut T, Janssen J. The role of acid and duodenal gastroesophageal reflux in symptomatic GERD. Am J Gastroenterol 2001;96:2033-40. 17. Wilmer A, Tack J, Dits H, Vanderschueren S, Gevers A, Bobbaers H. Duodenogastroesophageal reflux and esophageal mucosal injury in mechanically ventilated patients. Gastroenterology 1999;116:1293-9. 18. Vaezi M, Richter J. Role of acid and duodenogastroesophageal reflux in gastroesophageal reflux disease. Gastroenterology 1996;11:1192-9. 19. Koek G, Vos R, Flamen P, Sifrim D, Lammert F, Vanbilloen B, et al. Oesophageal clearance of acid and bile: a combined radionuclide, pH, and Bilitec study. Gut 2004;53:21-6. 20. Tack J, Koek G, Demedts I, Sifrim D, Janssens J. Gastroesophageal reflux disease poorly responsive to single dose proton pump inhibitors in patients without Barrett’s esophagus: acid reflux, bile reflux or both. Am J Gastroenterol 2004;99:981-93. 21. Poelmans J, Feenstra L, Tack J. The role of duodenogastroesophagopharyngeal reflux in unexplained excessive throat phlegm. Dig Dis Sci 2005;50:824-32. 22. Lundell LR, Deut J, Bennett JR, Blum AL, Armstrong D, Galmiche JP, et al. Endoscopic assessment of esophagitis: clinical and functional correlates and further validation of the Los Angeles classification. Gut 1999;45:172-80. 23. Armstrong D, Monnier PH, Nicolet M, Blum AL, Savary M. The ‘MUSE’ system. In: Guilli R, Tytgat GNJ, DeMeester TR, Galmiche JP, eds. The esophageal mucosa. Amsterdam: Elsevier Science; 1994, pp 313-21. 24. Furuta GT. Pediatric gastrointestinal endoscopy in gastrointestinal endoscopy clinics of North America. 2001;11:683-715. 25. Vandenplas Y. Working group of the European Society of Pediatric Gastroenterology and Nutrition. A standardized protocol for the methodology of esophageal pH monitoring and interpretation of the data for the diagnosis of gastroesophageal reflux. JPGN 1992;14:467-71. 26. Zacharioudakis G, Chrysos E, Athanasakis E, Tsiaoussis J, Karmoiris K, Xynos E. Is there any Mediterranean diet not affecting Bilitec assessment of bile reflux? Digestion 2004;70:84-92. 27. Vaezi MF, LaCamra RG, Richter JE. Bilitec 2000 ambulatory duodenogastric reflux monitoring system: studies on its validation and limitations. Am J Physiol 1994;30:1050-6. 28. Richter JE, Bradley LA, DeMeester TR, Wu WC. Normal 24 hr ambulatory esophageal pH values: influences of study center, pH electrode, age and gender. Dis Dig Sci 1992;37:849-56. 29. Van Den Driessche M, Veereman-Wauters G. Gastric emptying in infants and children. Acta Gastroenterol Belg 2003;66:274-82. 30. Champion G, Richter JE, Vaezi MF, Singh S, Alexander R. Duodenogastroesophageal reflux relationship to pH and importance of Barrett’s. Gastroenterology 1994;107:747-54. 31. Stein HJ, Kauer WK, Feussner H, Siewert JR. Bile reflux in benign and malignant Barrett’s esophagus: effect of medical acid suppression and Nissen fundoplication. J Gastrointest Surg 1998;2:333-41. 32. Marshall RE, Anggiansah A, Manifold DK, Owen WA, Owen WJ. Effect of omeprazole 20 mg twice daily on duodenogastric and gastro-esophageal bile reflux in Barrett’s esophagus. Gut 1998;43:603-7. 33. Menges M, Muller M, Zeitz M. Increased acid and bile reflux in Barrett’s esophagus compared to reflux esophagitis and effect of proton pump inhibitor therapy. Am J Gastroenterol 2001;96:231-7. 34. Stein HJ, Kauer WK, Feussner H, Siewert JR. Bile reflux in benign and malignant Barrett’s esophagus: effect of medical acid suppression and Nissen fundoplication. J Gastrointest Surg 1998;24:333-41. 35. Weilin W, Shijun J, Huizhen W, Wei W. 24-hour gastroesophageal double pH monitoring acid and alkaline gastroesophageal and duodenogastric refluxes in pediatric patients. Chin Med J 1998;111:881-4. 36. Iftikhar SY, Ledingham S, Evans DF, Yusuf SW, Steele RJ, Atkinson M, Hardcastle JD. Alkaline gastro-esophageal reflux: dual probe pH monitoring. Gut 1995;37:465-70. 37. Tack J. Review article: role of pepsine and bile in gastro-esophageal reflux disease. Aliment Pharmacol Ther 2005;22(suppl 1):48-54. 38. Orel R, Markovic S. Bile in the esophagus: a factor in the pathogenesis of reflux esophagitis in children. JPGN 2003;36:266-73. 39. Vaezi M, Richter J. Contribution of acid and duodenogastroesophageal reflux to esophageal mucosal injury and symptoms in partial gastrectomy patients. Gut 1997;1: 297-302. 40. Tack J, Bisschops R, Koek G, Sifrim D, Lerut T, Janssens J. Dietary restrictions during ambulatory monitoring of duodenogastroesophageal reflux. Dig Dis Sci 2003; 48:1213-20. 41. Vaezi MF, Sears R, Richter JE. Placebo-controlled trial of cisapride in postgastrectomy patients with duodenogastroesophageal reflux. Dig Dis Sci 1996;41:754-63. 42. Koek G, Sifrim D, Lerut T, Janssens J, Tack J. Effect of the GABA against baclofen in patients with symptoms and duodenogastroesophageal reflux refractory to proton pump inhibitors. Gut 2003;52:1397-402.

Duodenogastro-Esophageal Reflux in Children with Refractory Gastro-Esophageal Reflux Disease

311

Vocal Cord Dysfunction and Feeding Difficulties after Pediatric Cardiovascular Surgery RITU SACHDEVA, MD, ELORA HUSSAIN, MD, M. MICHELE MOSS, MD, MICHAEL L. SCHMITZ, MD, RICHARD M. RAY, MD, MICHIAKI IMAMURA, MD, PHD, AND ROBERT D. B. JAQUISS, MD

Objective To evaluate the impact of vocal cord dysfunction on feeding in children after cardiovascular surgery. Study design Of the 2255 children who had cardiovascular surgery between January 2000 to January 2006, 38 (1.7%) had postoperative vocal cord dysfunction confirmed at laryngoscopy. The following data were obtained retrospectively: type of surgery, laryngoscopic examination results, swallowing studies, upper gastrointestinal (UGI) studies, and feeding route: oral, nasogastric tube (NG), and gastrostomy. Results Surgeries included aortic arch reconstruction (n ⴝ 20), patent ductus arteriosus ligation (n ⴝ 8), arterial switch (n ⴝ 3), cervical cannulation for extracorporeal membrane oxygenation (n ⴝ 2), and others (n ⴝ 5). A swallowing study confirmed dysfunction in 27 of 29 patients. Gastrostomy was placed in 18/38 patients. At discharge, 18 patients were fed by gastrostomy, 13 orally, 3 by NG, and 4 by combination oral/NG. At a median follow-up of 12 months, 20 were fed orally, 1 by NG, 7 by gastrostomy, 7 by combination gastrostomy/orally, 1 was lost to follow-up, 2 died. Conclusion Vocal cord dysfunction after pediatric cardiovascular surgery is associated with significant feeding problems and may require prolonged gastrostomy feeding. These findings support aggressive surveillance for vocal cord dysfunction, especially in patients undergoing aortic arch surgery. (J Pediatr 2007;151:312-5)

ostoperative vocal cord dysfunction is a clinically important complication in children undergoing cardiovascular surgery because it may predispose them to aspiration by impairing their ability to protect their airway. Such aspiration may in turn result in significant pulmonary morbidity and in severe cases even prove fatal. There are several possible mechanisms responsible for vocal cord dysfunction, such as operative injury to the recurrent laryngeal nerve, association of the cardiac defect with a congenital laryngotracheal anomaly, prolonged intubation with direct vocal cord trauma, neurodevelopmental delay, poor sucking and swallowing coordination, or even injury from a transesophageal echocardiography probe.1,2 Some children with vocal cord injury undergo placement of a gastrostomy tube to avoid the risk of aspiration; others remain orally fed, but with altered consistency of feeds. Unfortunately, there is relatively little published information pertaining to feeding patterns after diagnosis of postoperative vocal cord dysfunction. In this study we sought to determine the impact of vocal cord dysfunction on feeding in children after cardiovascular surgery to provide some guidance for parents and physicians caring for these children.

P

METHODS This study was approved by the Institutional Review Board of the University of Arkansas for Medical Sciences. Medical records of pediatric patients (⬍18 years) who had heart surgery from January 2000 to January 2006 and had otorhinolaryngology consultation were reviewed. All those who had postoperative vocal cord dysfunction confirmed at laryngoscopy formed the study cohort. The following data were obtained retrospectively: age and weight at surgery, cardiac diagnosis and type of surgery, whether intraoperative transesophageal echocardiography was performed, duration of endotracheal intubation, presence of neurologic injury, laryngoscopic examination results, esophageal swallowing studies, upper gastrointestinal (UGI) studies, duration of hospital stay, discharge and follow-up feeding route: oral, nasogastric tube (NG), gastrostomy. Laryngoscopic evaluations were performed by a pediatric otorhinolaryngologist by flexible, fiberoptic laryngoscopy. Vocal cord paresis was defined as incomplete abduction ECMO MRI

312

Extracorporeal membrane oxygenation Magnetic resonance imaging

NG UGI

Nasogastric tube Upper gastrointestinal

From the Department of Pediatrics, Division of Pediatric Cardiology (R.S., E.H., M.M.); the Department of Anesthesiology (M.S.), Otorhinolaryngology (R.R.), and Pediatric Cardiothoracic Surgery (M.I., R.J.), Arkansas Children’s Hospital and University of Arkansas for Medical Sciences, Little Rock, Arkansas. Submitted for publication Sep 25, 2006; last revision received Dec 19, 2006; accepted Mar 6, 2007. Reprint requests: Ritu Sachdeva, MD, 1900 Maryland, Slot 512-3, Little Rock, AR 72202. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.03.014

or adduction of the cord. Vocal cord paralysis was defined as complete immobility of the vocal cord. Esophageal swallow studies were performed by a speech pathologist in conjunction with a supervising radiologist. Patients were given barium with 3 different consistencies: thin, nectar, and honey. The swallowing was viewed in a lateral plane with continuous fluoroscopy. Laryngeal penetration was present if the barium bolus entered the glottic aperture but did not pass the level of vocal cords. If the bolus passed across the vocal cords and entered the subglottic area and trachea, it was considered aspiration. Sucking and swallowing coordination were also observed during this study. The study was repeated at an interval recommended by the speech pathologist. A barium contrast UGI study or technetium scanning was performed to identify gastroesophageal reflux on the basis of clinical suspicion. Recommendations were made by the speech pathologist to alter the feeding consistency if there was aspiration or deep penetration noted with a certain consistency of barium. If the patient had aspiration with all consistencies or had poor suck and swallow coordination in spite of speech therapy, gastrostomy tube placement was considered. In addition, a Nissen fundoplication was performed if there was evidence of significant reflux on UGI or technetium scan. Charts were reviewed to obtain information about the feeding pattern at each follow-up visit with cardiologist, otorhinolaryngologist, or speech pathologist.

RESULTS From January 2000 to January 2006, 2255 children underwent cardiovascular surgery at our institution. Vocal cord dysfunction was diagnosed in 38 (1.7%). Primary cardiovascular surgeries included coarctation repair in 10, modified Norwood procedure in 10, patent ductus arteriosus (PDA) ligation in 8, arterial switch in 3, cannulation for cervical extracorporeal membrane oxygenation (ECMO) in 2, systemic-pulmonary shunt in 1, left aortopulmonary collateral ligation in 1, left pulmonary arterioplasty in 1, truncus arteriosus repair in 1, and Fontan completion in 1. Many patients had more than 1 cardiac lesion addressed during surgery. The median age at surgery was 13.5 days (1 to 1604 days); median weight was 3.45 kg (0.9 to 14.5 kg). Transesophageal echocardiography was performed during surgery in 11/38 (29%) patients. The median duration of intubation was 10 days (2 to 71 days). The median length of hospital stay was 32.5 days (9 to 107 days). There were 2 deaths: patient 11 died at 3 months of age because of infection of ventriculoperitoneal shunt, and patient 38 died 3 days after Nissen fundoplication and gastrostomy placement. Cause of death of patient 38 was unknown and autopsy was denied by the family (Table I; available at www.jpeds.com). Five patients had significant neurologic injury diagnosed during surgery. Patient 4 had multiple seizures in the immediate postoperative period and a magnetic resonance imaging (MRI) of the brain showed multiple hemorrhagic infarcts in the white matter. Patient 11 had severe commu-

nicating hydrocephalus diagnosed 3 weeks after modified Norwood palliation and underwent a ventriculoperitoneal shunt. Patient 21 who had been on ECMO support for severe pulmonary hypertension was noted to have extensive encephalomalacia involving both cerebral hemispheres on an MRI performed 23 days after being placed on ECMO. This patient had a normal head ultrasound scanning result before being placed on ECMO. Cerebral palsy later developed as a consequence of the patient’s neurologic damage. Patient 23 had extensive encephalomalacia of both cerebral hemispheres with focal encephalomalacia of the right parietal cortex observed on MRI performed a year after initial heart surgery. This patient had heterotaxy syndrome and a very complicated postoperative course including support with ECMO. Also, patient 23 already had a gastrostomy after his initial heart surgery, before the diagnosis of neurologic injury and has continued to be fed by gastrostomy. Patient 27 had a computed tomographic scan of the head performed 1 week before heart surgery that showed focal areas of edema in both cerebral hemispheres suggesting infarction. A follow-up study performed 2 weeks after cardiac surgery showed progressive atrophy and ischemia compatible with hypoxic encephalopathy. This patient has continued to be fed by gastrostomy. Patient 17 had Down syndrome, and patient 33 had Turner syndrome. The initial laryngoscopic examination showed unilateral vocal cord paresis in 4, bilateral paresis in 1, unilateral paralysis in 30, bilateral paralysis in 2, and left vocal cord paralysis with right vocal cord paresis in 1. Of the 30 patients with unilateral paralysis, 2 patients who had undergone cannulation for cervical ECMO had right vocal cord paralysis and the remaining 28 had left vocal cord paralysis. Positioning of a paralyzed cord was in the paramedian position in all cases. None of the patients had a documented congenital laryngotracheal anomaly, except patient 23 who had laryngomalacia. A follow-up evaluation was available in 9 of 38 patients and showed normal vocal cord mobility in 3 patients, improved mobility in 4, and unchanged in 2. None of our patients underwent a medialization of the paralyzed vocal cord. Tracheostomy was done in 3 patients: patient 16 who had bilateral vocal cord paralysis and subglottic stenosis; patient 21 returned to hospital 6 months after initial discharge with chronic lung disease and severe laryngotracheomalacia and failed several attempts at extubation, thus requiring tracheostomy; patient 31 returned to the hospital 2 weeks after initial discharge with respiratory failure related to prematurity and chronic lung disease and required tracheostomy after prolonged mechanical ventilation and failure to wean from ventilatory support. Results of esophageal swallow studies, UGI and technetium scanning, gastrointestinal surgery, and the feeding pattern before heart surgery and at the time of discharge are shown in Table I. Before surgery 17 patients had never been fed orally, 16 were being fed orally, 4 by a transpyloric tube, and 1 by a nasogastric tube. Clinical suspicion of swallowing dysfunction was noted in 34 of 38 (89%) patients with vocal cord dysfunction diagnosed at laryngoscopy. A swallowing

Vocal Cord Dysfunction and Feeding Difficulties after Pediatric Cardiovascular Surgery

313

Table II. Percentage of patients on various feeding regimens at the time of discharge from the hospital and at last follow-up Feeding management Normal oral Modified oral Oral/GT or NG combination All GT or NG

Hospital discharge status (n ⴝ 38)

Last follow-up status (n ⴝ 35)*

13.2% 18.4% 13.2%

37.2% 20% 17.1%

55.2%

25.7%

GT, Gastostomy; NG, nasogastric tube. *One lost to follow-up; 2 died.

study was done in 29 patients within 30 days after surgery; all were abnormal in terms of aspiration, penetration, or discordant swallowing. Aspiration of varying consistencies of formula was noted in 23; 5 had penetration with thin or nectar consistency, and 1 had delayed swallowing with thin and nectar consistency. Results of a follow-up swallow study were available in 17 patients at a mean follow-up of 6.2 ⫾ 4.1 months from the initial study. It was normal in 8, showed improved swallowing in 4, unchanged in 3, and worsening in 2. Gastroesophageal reflux was evaluated by UGI in 20 patients and a technetium scan in 3, and both in 2. Of these 25 patients, reflux was identified in 16. Gastrostomy was placed in 18/38 (47%) patients, 16 of whom also underwent Nissen fundoplication. Two patients (31 and 32) with aspiration of all consistencies and no evidence for reflux also had gastrostomy placed. At discharge, 5 patients were sent home on normal oral feeds, 7 with thickened oral feeds, 5 with combination of nasogastric and thickened oral feeds, 3 were on nasogastric feeds, and 18 patients were fed by gastrostomy (Table II). At a mean follow-up duration of 10 months (3 to 60 months), 1 was lost to follow-up; 13 were on normal oral feeds; 7 were on thickened oral feeds; of the 18 patients who underwent gastrostomy placement, 9 were still fed by gastrostomy; 6 were on combination of gastrostomy and orally fed; 1 had normal swallow function and gastrostomy removal; and 2 died (Table II).

DISCUSSION The phenomenon of vocal cord dysfunction after pediatric heart surgery is familiar in pediatric cardiac units. Although it is sometimes viewed as relatively trivial in comparison to other possible complications of cardiovascular surgery, it is well known that the probability of aspiration is increased in patients with vocal cord dysfunction because of impaired airway protection.3,4 This potential for pulmonary injury in the setting of vocal cord dysfunction in turn has major implications for feeding strategies in these children. It must be emphasized that every single patient in our study with vocal cord dysfunction had an abnormal swallowing study result. We have demonstrated that only 13% of such children may be normally orally fed by the time of hospital discharge, and that 314

Sachdeva et al

the remaining patients will receive some or their entire nutritional intake as modified oral feeds with or without some sort of feeding tube when they leave the hospital. A large portion of our study cohort (34/38) had undergone surgery requiring dissection around the aortic arch, ductus arteriosus, or left pulmonary artery. Because the left recurrent laryngeal nerve courses around the ductus arteriosus, surgery involving the ductus, the descending aorta, or the left pulmonary artery may require mobilization of this nerve and put it at risk for injury. It is worth noting that the 2 patients who were placed on cervical ECMO via right cervical incision developed right vocal cord paralysis, the presumable mechanism being injury to the vagus nerve in the carotid sheath. The vagus nerve provides innervation to the right vocal cord via the recurrent laryngeal nerve and could be injured during manipulation of the common carotid artery and internal jugular vein for cervical ECMO. Schumacher et al5 reported 5 similar cases of right vocal cord paralysis after cervical ECMO without any other risk factors and suggested that a laryngoscopic examination should be considered for patients after ECMO. Similar to our study cohort, previous studies have also demonstrated a high risk for injury to the left recurrent laryngeal nerve during PDA ligation and arch reconstruction.6,7 Skinner at al8 evaluated the incidence and significance of recurrent laryngeal nerve injury and swallowing dysfunction after the Norwood procedure and compared Norwood patients to a group of patients undergoing biventricular aortic arch reconstruction. In this study swallowing dysfunction occurred in 48% of patients, with aspiration in 24%. Unlike other reports, left recurrent laryngeal injury was believed to be an uncommon cause of swallowing dysfunction and was seen in only 9% of their patients. They did not find any difference in incidence of left true vocal cord paralysis and aspiration between those who had Norwood procedure versus those who underwent biventricular repair with aortic arch reconstruction. Kohr et al2 reported dysphagia in 18% of children undergoing cardiac surgery. The risk factors for dysphagia identified in their study included preoperative intubation, age less than 3 years, and use of transesophageal echocardiography in children weighing less than 5.5 kg. Transesophageal echocardiography was used in only 29% of our patients. We do not have a control group to compare whether vocal cord dysfunction was associated with the duration of intubation or intraoperative use of transesophageal echocardiogram. Besides vocal cord dysfunction, other potential risk factors that could contribute to feeding difficulties include abnormal swallowing, gastroesophageal reflux, and splanchnic hypoperfusion related to surgery or underlying heart disease. In addition, delayed introduction of oral feedings as seen in 45% of our study cohort can result in significant oromotor and swallowing dysfunction. Even though we analyzed all patients ⬍18 years old who underwent cardiothoracic surgery, it is important to note that all but 1 of the patients with vocal cord dysfunction were younger than 1 year old. Vocal cord dysfunction related to The Journal of Pediatrics • September 2007

congenital heart surgery appears to be extremely uncommon in children over 1 year of age. Zbar et al7 studied 17 cases of vocal cord paralysis in infants under 12 months of age. Eight of these children with left vocal fold paralysis had a history of prior thoracic surgery. This study reported a 7.4% postoperative incidence of vocal fold paralysis after ligation of PDA. They concluded that iatrogenic injury of left recurrent laryngeal nerve during heart surgery is the most common cause of vocal cord paralysis in infants and that it persists at an average follow-up of 6 months. However, idiopathic vocal cord paralysis resolved within an average of 6 weeks. In another study Zbar at al6 found that the incidence of iatrogenic left vocal cord paralysis was 8.8% in infants undergoing PDA ligation, and the single major risk factor for this was birth weight less than 1 kg. In our study cohort, 3 of the 8 patients who had PDA ligations were premature. These 3 premature infants represent a tiny fraction of the premature neonates undergoing ductal ligation during the period of the study. On the other hand, 20 patients who had complex arch manipulation either as a part of Norwood procedure or coarctation repair developed vocal cord dysfunction. This reflects a much higher incidence of vocal cord dysfunction in these children undergoing more complex repairs. The question of whether the recurrent laryngeal nerve is at less risk during ductal ligation as opposed to arch reconstruction, or whether the consequences of nerve injury are less transient and, hence, magnified in the setting of more complex congenital heart disease remains unanswered. In a clinical review, Hartle et al9 provided an update on management of adult patients with unilateral recurrent laryngeal nerve paralysis as a result of thyroid surgery. They stated that this lesion is frequently well tolerated but may be life threatening because of the possibility of aspiration pneumonia. They recommended surgical treatment with medialization of the paralyzed vocal cord to close the glottic gap on phonation, so that the normal vocal fold can make contact with the paralyzed one. This technique reportedly is simple, has low complication rate and is highly efficient in eliminating aspiration and improving voice quality.10,11 In contrast, Bhattacharyya et al12 did not find significant improvement in aspiration after vocal cord medialization. There are no conclusive studies regarding use of medialization for unilateral vocal cord paralysis in infants. A swallowing study is a useful tool for diagnosing aspiration in patients who have vocal cord dysfunction. Aspiration of varying consistencies was noted in 80% of our patients who had a swallowing study. Thus allowing us to appropriately alter their feeding regimen or place a GT. Even though a follow-up swallow study was not available in all patients, 70% showed normal or improved swallowing on a follow-up study at 6 months, emphasizing the need for a continued follow-up evaluation by a speech pathologist. Most

of the patients who received a gastrostomy continued to be fed by gastrostomy at the time of last follow-up. Several limitations of our study are worthy of note, the most important of which is its retrospective nature. Because we did not prospectively evaluate all patients undergoing cardiovascular surgery, we do not know the true incidence of vocal cord dysfunction and have probably underestimated it. Lesser degrees of vocal cord dysfunction or temporary cord dysfunction may have gone undetected in this cohort. Not all patients who had vocal cord dysfunction had evaluation by a swallow study. Also a follow-up swallow study was not done in all those who initially had an abnormal swallow study result. Most patients do not have a follow-up laryngoscopic evaluation to determine the resolution or persistence of vocal cord dysfunction. We are currently prospectively monitoring all patients who have suspicion of vocal cord dysfunction especially after heart surgeries involving great vessels or PDA. These patients are undergoing laryngoscopic evaluations, swallowing study and UGI or technetium scan. A systematic approach like this will allow us to better treat these patients in terms of their feeding and might possibly reduce the duration of hospitalization after heart surgery. This will also enable us to better predict the time course of the cord paresis and the need for gastrostomy feeds. We thank Carl W. Chipman Jr, RN, for his help with the database.

REFERENCES 1. Khariwala SS, Lee WT, Koltai PJ. Laryngotracheal consequences of pediatric cardiac surgery. Arch Otolaryngol Head Neck Surg 2005;131:336-9. 2. Kohr LM, Dargan M, Hague A, Nelson SP, Duffy E, Backer CL, et al. The incidence of dysphagia in pediatric patients after open heart procedures with transesophageal echocardiography. Ann Thorac Surg 2003;76:1450-6. 3. Kohda E, Hisazumi H, Hiramatsu K. Swallowing dysfunction and aspiration in neonates and infants. Acta Otolaryngol Suppl 1994;517:11-6. 4. Newman LA, Keckley C, Petersen MC, Hamner A. Swallowing function and medical diagnoses in infants suspected of Dysphagia. Pediatrics 2001;108:E106. 5. Schumacher RE, Weinfeld IJ, Bartlett RH. Neonatal vocal cord paralysis following extracorporeal membrane oxygenation. Pediatrics 1989;84:793-6. 6. Zbar RI, Chen AH, Behrendt DM, Bell EF, Smith RJ. Incidence of vocal fold paralysis in infants undergoing ligation of patent ductus arteriosus. Ann Thorac Surg 1996;61:814-6. 7. Zbar RI, Smith RJ. Vocal fold paralysis in infants twelve months of age and younger. Otolaryngol Head Neck Surg 1996;114:18-21. 8. Skinner ML, Halstead LA, Rubinstein CS, Atz AM, Andrews D, Bradley SM. Laryngopharyngeal dysfunction after the Norwood procedure. J Thorac Cardiovasc Surg 2005;130:1293-301. 9. Hartl DM, Travagli JP, Leboulleux S, Baudin E, Brasnu DF, Schlumberger M. Clinical review: Current concepts in the management of unilateral recurrent laryngeal nerve paralysis after thyroid surgery. J Clin Endocrinol Metab 2005;90:3084-8. 10. Laccourreye O, Paczona R, Ageel M, Hans S, Brasnu D, Crevier-Buchman L. Intracordal autologous fat injection for aspiration after recurrent laryngeal nerve paralysis. Eur Arch Otorhinolaryngol 1999;256:458-61. 11. Hartl DM, Hans S, Vaissiere J, Riquet M, Laccourreye O, Brasnu DF. Objective voice analysis after autologous fat injection for unilateral vocal fold paralysis. Ann Otol Rhinol Laryngol 2001;110:229-35. 12. Bhattacharyya N, Kotz T, Shapiro J. Dysphagia and aspiration with unilateral vocal cord immobility: incidence, characterization, and response to surgical treatment. Ann Otol Rhinol Laryngol 2002;111:672-9.

Vocal Cord Dysfunction and Feeding Difficulties after Pediatric Cardiovascular Surgery

315

315.e1

Table I. Feeding pattern before and after cardiovascular surgery

Sachdeva et al

Follow-up swallow study (months since initial)

Upper GI/technetium scan

Penetration with thin

Normal (12)

Malrotation

Aspiration with thin and nectar Aspiration with thin and nectar Delayed swallowing with thin and nectar Penetration with thin — — Aspiration with thin and nectar Penetration with thin Aspiration with thin and nectar — — Aspiration with thin, nectar and honey Aspiration with thin and nectar Aspiration with thin

Normal (12)

Normal

Ladd’s procedure —

Normal (3)

Normal

—

The Journal of Pediatrics • September 2007

Patient no.

Preoperative feeding

Swallow study

1

TPT

2

PO

3

NPO

4

NPO

5 6 7 8

NPO PO NPO NPO

9 10

NPO NPO

11 12 13

NPO PO PO

14

PO

15

PO

16

TPT

17

PO

18

PO

19

PO

20

PO

21

NPO

22 23

NPO NPO

Aspiration with thin and nectar Aspiration with thin, nectar and honey Aspiration with thin and nectar Aspiration with thin, nectar and honey Aspiration with thin and nectar Aspiration with thin, nectar and honey — —

—

—

— — —

GI surgery

—

Feeding at discharge

Feeding at last follow-up

Follow-up (mos)

NG/honey thickened PO Honey thickened PO NG

PO

76

PO

70

PO

61

NG/honey thickened PO

Lost

9

NG PO PO GT

PO PO PO GT/thickened PO PO PO

66 32 31 16

3 12 13

Normal (2) — — Unchanged (12)

GER

— — — Nissen, GT

Normal (3) Normal (3)

GER Normal

Nissen, GT —

— — Improved swallow with nectar (6) Improved swallow with nectar (8) Aspiration with nectar (6) —

GER GER

Nissen, GT — Nissen, GT

GT NG/honey thickened PO GT PO GT

GER

Nissen, GT

GT

Died PO GT/nectar thickened PO GT/PO

Nissen, GT

Honey thickened PO GT

11

GER

Nectar thickened PO GT

Unchanged (12)

GER

Nissen, GT

GT

GT

24

Honey thickened PO NG

PO

18

PO

31 14

67 19

—

—

—

Normal (12)

—

—

Normal (3)

—

—

Aspiration with honey (3) — — —

GER

Nissen, GT

GT

GER

Nissen, GT

GT

GT/honey thickened PO GT

— GER, Malrotation

— Nissen, GT, Ladd’s

PO GT

PO GT

64 22

12

14

13

Vocal Cord Dysfunction and Feeding Difficulties after Pediatric Cardiovascular Surgery

Table I. Continued Follow-up swallow study (months since initial)

Upper GI/technetium scan

GI surgery

Aspiration with thin

—

—

—

— — —

PO

— — Aspiration with thin, nectar and honey Penetration with thin and nectar —

30

PO

Aspiration with thin

31

NPO

32

PO

33

NPO

34

NPO

35

PO

Aspiration with thin, nectar and honey Aspiration with thin, nectar and honey Aspiration with thin and nectar Aspiration with thin, nectar and honey Aspiration with thin and nectar

36

TPT

37 38

NPO NPO

Patient no.

Preoperative feeding

24

NG

25 26 27

TPT PO PO

28

NPO

29

Swallow study

Aspiration with thin and nectar Penetration with thin Aspiration with thin and nectar

Unchanged (3)

GER — GER Normal

Improved; aspiration with thin (3) — —

Follow-up (mos)

NG/nectar thickened PO GT PO GT

Nectar thickened PO GT PO GT

6

Honey thickened PO Nectar thickened PO Nectar thickened PO GT

5

GT/nectar thickened PO Honey thickened PO GT

7

Honey thickened liquids/PO solids GT/honey thickened PO GT Died

3

Normal

GT

Normal

GT

GT

—

—

—

—

— GER

—

—

Feeding at last follow-up

Honey thickened PO Nectar thickened PO Nectar thickened PO GT

— Improved; penetration with thin (3) —

Nissen, GT — Nissen, GT

Feeding at discharge

— Nissen, GT

—

—

NG/honey thickened PO GT

—

GER

Nissen, GT

Honey thickened liquids/PO solids GT

— —

GER GER

Nissen, GT Nissen, GT

GT GT

GER, Gastroesophageal reflux; GI, gastrointestinal; GT, gastrostomy; NG, nasogastric tube; NPO, nil per os; PO, per os (oral); TPT, transpyloric tube.

5 6 7

8 8 6

4 3

6 5 2

315.e2

GRAND ROUNDS

Growing Skull Fracture after Minor Closed-Head Injury JEAN-RODOLPHE VIGNES, MD, PHD, N. U. OWASE JEELANI, MRCS, MBA, MPHIL, ASHFAQ JEELANI, MD, MSC, MRCPCH, MICHEL DAUTHERIBES, MD, AND DOMINIQUE LIGUORO, MD, PHD

ead injury is a major public health issue due to its prevalence and the associated socioeconomic costs. Its incidence is increasing in urban areas.1 Minor closed-head injuries (MCHIs) constitute more than 80% of all head injuries.2 A conservative approach to diagnostic evaluation is generally recommended in infants with MCHI.3 Skull radiograph (SR) plays a limited role in these cases, except when nonaccidental injury is considered.4 Growing skull fracture (GSF), a known intracranial complication of head injury, can occur after an MCHI.5 Unless this diagnosis is considered at initial presentation, a significant delay in detection can occur, resulting in suboptimal management.6 To highlight this issue, we present 3 cases of children with a typical MCHI, as defined by previous studies,3,4 who subsequently developed GSF. This report emphasizes the importance of initial evaluation in identifying those patients with MCHI at risk of developing intracranial injury. A timely and efficient management plan is recommended.

H

CASE 1 A 5-day-old term male infant fell from a baby-changing table and cried immediately. The next day, the infant was seen by the family doctor, who noted a subcutaneous parietal hematoma but no other signs of injury or neurologic deficit. No follow-up assessment or investigations were instigated. Five months later, the infant’s mother, concerned about a persistent hematoma (Figure 1A) and progressive lethargy, brought the child for further medical attention. Examination at this point revealed hypotonia with a right incomplete motor deficit. SR revealed a large skull fracture (Figure 1B). Magnetic resonance imaging (MRI) confirmed the presence of a large left temporoparietal meningoencephalocele through the dura, as well as the bone defect, which produced a significant skull deformity (Figure 1C). The assessment protocol for child abuse was followed and yielded negative findings. Electroencephalography (EEG) revealed focal seizure activity requiring medical treatment (sodium valproate). Neurosurgical management comprised duroplasty and cranioplasty. Postoperative outcome was good, and the infant was discharged to home 7 days after the surgery. The infant exhibited early improvement in neurologic status. At 6-month follow-up, persistent mild residual hemiparesis was noted, and EEG showed continuing epileptiform discharges.

CASE 2 A 3-month-old male infant fell from his pram onto the strut of a windowpane. He cried after the fall, and his mother brought him to the emergency department. Examination revealed a right parietal subcutaneous hematoma. The infant had no history of lethargy, irritability, or vomiting, and clinical examination revealed no neurologic deficits. The infant remained well and was discharged to home after 6 hours of observation. Three months later, the parents noted that the infant was intermittently hypotonic and had a pulsatile collection on his head, which became tense with coughing and crying. An SR showed a large right parietal skull fracture (Figure 2A), and a computed tomography (CT) scan confirmed the diagnosis of a GSF (Figure 2B). A durocranioplasty was successfully performed. The child remained asymptomatic at the 1-year follow up.

CASE 3 A 3-month-old male infant fell from his mother’s arms onto a tiled floor. The infant cried immediately and was brought to the emergency department. Examination revealed a 5-cm-diameter right parietal cephalhematoma. The infant remained alert with no CT EEG GSF

316

Computed tomography Electroencephalography Growing skull fracture

MCHI MRI SR

Minor closed-head injury Magnetic resonance imaging Skull radiograph

From the Department of Neurosurgery A, CHU of Bordeaux, University of Bordeaux 2, Bordeaux, France (J.R.V., D.L.); Department of Pediatric Neurosurgery, Great Ormond Street Hospital, London, UK (O.W.); Wickford Health Centre for Children, Wickford, Essex, UK (A.J.); and Department of Neurosurgery and Pediatric Neurosurgery, CHU of Bordeaux, Bordeaux, France (M.D.). Submitted for publication Oct 20, 2006; last revision received Feb 26, 2007; accepted Apr 19, 2007. Reprint requests: Jean-Rodolphe Vignes, MD, PhD, Department of Neurosurgery A, Hôpital Pellegrin, 1 Place Amélie Raba-Léon, 33076 Bordeaux Cedex, France. E-mail: [email protected]. J Pediatr 2007;151:316-8 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.04.041

Figure 1. Case 1, preoperative radiologic images. A, A large left-sided parieto-occipital subcutaneous swelling is evident. B, An anteroposterior skull radiograph showing a large calvarial defect (GSF). C, A T2-weighted axial MRI demonstrating a large left parieto-occipital leptomeningeal cyst extruding through the calvarial defect.

Figure 2. Case 2, preoperative radiologic images. A, Lateral skull radiograph demonstrating a GSF. B, Axial CT showing extrusion of the brain parenchyma through a right parieto-occipital skull defect.

neurologic deficit. Fundoscopic examination was negative. An SR identified a parietal linear skull fracture. A CT scan revealed a cephalhematoma with an underlying cortical contusion and some traumatic subarachnoid blood overlying the contused cortex (Figure 3A). The infant was observed as an inpatient for 7 days, during which time he remained well. A repeat CT scan showed resolution of the contusions, and he was discharged home. Two weeks later, clinical examination revealed a persistent right parietal pulsatile subcutaneous mass. MRI confirmed the presence of a meningocele (Figure 3B) through the fracture. This was repaired surgically, and the infant was discharged 5 days later. One-year follow up was entirely satisfactory.

DISCUSSION GSF is a rare complication of head trauma.5,7 The prevalence ranges from 0.05% to 1.6% of all skull fractures,8 with 1% typically cited. This figure is an underestimate for the concerned pediatric population, however, because although skull fractures occur in the entire pediatric population, GSFs occur primarily in the first few years of life. Some 90% of GSFs occur before age 3 years, and more than 50% occur before age 12 months.9 Falling is the most frequent cause of the injury, followed by motor vehicle accidents and birth injury.8 Nonaccidental injury is another cause than should be considered and excluded.10 GSFs are generally associated with severe head trauma but can occur with minor head injury, with an associated risk of complications including Growing Skull Fracture after Minor Closed-Head Injury

Figure 3. Case 3, preoperative radiologic images. A, An axial CT slice. B, A coronal T2-weighted MRI showing a leptomeningeal cyst through a right parietotemporal calvarial defect.

meningoencephalocele, porencephalic cysts, hydrocephalus, brain atrophy, and functional alterations such as epileptic discharges, neurologic deficit, and psychomotor development delay.5,7 As a general rule in cases of brain injury, early diagnosis and prompt management can help minimize complications and provide the best prognosis.11 Whereas the primary injury sustained by the brain parenchyma is irreversible, efforts should be geared toward reducing secondary insults. There has been no study specifically comparing outcomes in cases of GSF with early intervention versus delayed intervention, but an understanding of the pathophysiologic mechanisms suggests a more favorable outcome in the former category. Smaller skull and dural defects are easier to repair, and prolonged progressive parenchymal herniation is likely to result in gliotic changes. Our case 3 demonstrated early detection and management with an excellent outcome. In case 1, on the other hand, the delay may be attributed to the residual hemiparesis. In case 2, despite the delay, there was no residual deficit after treatment. Nevertheless, the interval between head injury and diagnosis of GSF continues to vary from the time of initial consultation to a few years.6,9,11,12 Age is a significant risk factor for intracranial injury after head trauma, with reports of age cohorts of under 3 months, under 1 year, and under 2 years composing specific risk categories.3,13,14 There is consensus that age under 3 months is almost certainly a risk factor for brain injury after MCHI. Head injury in the younger age group is distinct from that in older children and adults because of differences in mechanisms and injury thresholds.15 A normal neurologic examination and maintenance of consciousness do not preclude the presence of significant intracranial injury in pediatric trauma patients.16 Moreover, 48% of intracranial abnormalities on CT scan are associated with normal initial clinical examination.3 Some authors feel that the Glasgow coma scale is not sensitive under age 2 years17 and recommend using the pediatric version of this scale.18 Abnormalities of scalp examination may be a marker for intracranial injury in asymptomatic infants.3,19 Cephalhematoma, a subperiostic skull hematoma, typically does not cross suture lines, is painful, and decreases slowly, unlike a strict subcutaneous hematoma. Cephalhematoma is considered an 317

indirect sign of bone fracture that is a necessary prerequisite to developing a GSF.14 Previous studies have shown that subcutaneous scalp hematomas, independent of their location, are also strongly associated with skull fracture in infants with head injury, and that skull fractures in turn are closely associated with intracranial pathology.3,15,19,20 All of our cases presented with scalp hematoma, and this tendency has been corroborated in other studies involving more patients.8 Despite the frequent occurrence of MCHI in children, management strategies differ among individuals and institutions. A CT scan is not sensitive enough to detect dural tears at the initial phase,11 and although MRI has greater sensitivity,21 its use is not routine in cases of head injury, especially in developing countries.11 Some authors have proposed using B-mode ultrasound for early detection of the dural defect in cases of diastatic skull fractures of posttraumatic collection overlying the skull, but the clinical efficacy of this approach remains to be determined.11 SR offers a limited role in the evaluation of children with MCHI.4 It is a radiation source22 and, when applied indiscriminately, affords a low yield.13 In our series, cases 1 and 2 underwent no radiologic evaluation at initial presentation and went on to develop GSF. In case 3, early detection of a fracture led to prompt management. In the literature, larger series of patients with GSF who underwent SR at the time of initial presentation exhibited skull fracture,8,23 and generally, bone diastasis of 4 mm or more was noted.8 Based on our experience and a review of the available literature, we propose that infants under age 3 months with head injury, MCHI or otherwise, and scalp hematoma are a high-risk subgroup for developing GSF. In the absence of correlative studies documenting the incidence of skull fractures in this subgroup, we recommend that all such patients undergo SR at initial presentation. Demonstration of a fracture mandates close clinical follow-up and a specialist’s evaluation. The need for further imaging, such as a CT or MRI, depends on the characteristics of each individual case. Clinical follow-up over the ensuing weeks should focus on identifying any developing neurologic deficits and on evaluating the scalp hematoma for settling. In cases 1 and 2, the scalp swelling remained persistent and in fact enlarged. In case 1, the neurologic deficit was detected on subsequent examination. In both cases, the diagnosis likely would have been made earlier had adequate follow-up been undertaken. Finally, the role of parent education should not be underestimated or overlooked. As appropriate, the parents should be informed of the possibility of GSF and instructed to watch for any persistent or progressive scalp swelling and the onset of any neurologic signs and symptoms. GSF remains a rare but serious complication of pediatric head trauma. Diagnosis and management is typically delayed by the lack of awareness by front-line health care personnel who care for these patients. Although this article highlights a particular high-risk subgroup of infants with MCHI (under age 3 months with cephalhematoma), this

318

Vignes et al

complication can occur outside this cohort as well. More work is needed to more accurately define the incidence of GSF, especially in cases of MCHI associated with cephalhematoma. Improved awareness by health care staff and better parent education is essential for prompt diagnosis and treatment to help minimize the detrimental sequelae of this complication.

REFERENCES 1. Yates PJ, Williams WH, Harris A, Round A, Jenkins R. An epidemiological study of head injuries in a UK population attending an emergency department. J Neurol Neurosurg Psychiatry 2006;77:699-701. 2. Brookes M, MacMillan R, Cully S, Anderson E, Murray S, Mendelow AD, et al. Head injuries in accident and emergency departments. How different are children from adults? J Epidemiol Community Health 1990;44:147-51. 3. Greenes DS, Schutzman SA. Clinical indicators of intracranial injury in headinjured infants. Pediatrics 1999;104:861-7. 4. Committee on Quality Improvement, American Academy of Pediatrics and Commission on Clinical Policies and Research, American Academy of Family Physicians. The management of minor closed head injury in children. Pediatrics 1999;104:1407-15. 5. Naim UR, Jamjoom Z, Jamjoom A, Murshid WR. Growing skull fractures: classification and management. Br J Neurosurg 1994;8:667-79. 6. Ziyal IM, Aydin Y, Turkmen CS, Salas E, Kaya AR, Ozveren F. The natural history of late diagnosed or untreated growing skull fractures: report on two cases. Acta Neurochir (Wien) 1998;140:651-4. 7. Muhonen MG, Piper JG, Menezes AH. Pathogenesis and treatment of growing skull fractures. Surg Neurol 1995;43:367-73. 8. Ersahin Y, Gulmen V, Palali I, Mutluer S. Growing skull fractures (craniocerebral erosion). Neurosurg Rev 2000;23:139-44. 9. Lende RA, Erickson TC. Growing skull fractures of childhood. J Neurosurg 1961;18:479-89. 10. Laskey AL, Holsti M, Runyan DK, Socolar RR. Occult head trauma in young suspected victims of physical abuse. J Pediatr 2004;144:719-22. 11. de Djientcheu VP, Njamnshi AK, Ongolo-Zogo P, Kobela M, Rilliet B, Essomba A, et al. Growing skull fractures. Child Nerv Syst 2006:1-5. 12. Zegers B, Jira P, Willemsen M, Grotenhuis J. The growing skull fracture, a rare complication of paediatric head injury. Eur J Pediatr 2003;162:556-7. 13. Browning JG, Reed MJ, Wilkinson AG, Beattie T. Imaging infants with head injury: effect of a change in policy. Emerg Med J 2005;22:33-6. 14. Palchak MJ, Holmes JF, Vance CW, Gelber RE, Schauer BA, Harrison MJ, et al. A decision rule for identifying children at low risk for brain injuries after blunt head trauma. Ann Emerg Med 2003;42:492-506. 15. Duhaime AC, Alario AJ, Lewander WJ, Schut L, Sutton LN, Seidl TS, et al. Head injury in very young children: mechanisms, injury types, and ophthalmologic findings in 100 hospitalized patients younger than 2 years of age. Pediatrics 1992;90:179-85. 16. Simon B, Letourneau P, Vitorino E, McCall J. Pediatric minor head trauma: indications for computed tomographic scanning revisited. J Trauma 2001;51:231-8. 17. Chung CY, Chen CL, Cheng PT, See LC, Tang SF, Wong AM. Critical score of Glasgow coma scale for pediatric traumatic brain injury. Pediatr Neurol 2006; 34:379-87. 18. Macgregor DM, McKie L. CT or not CT--that is the question. Whether ‘tis better to evaluate clinically and x-ray than to undertake a CT head scan! Emerg Med J 2005;22:541-3. 19. Quayle KS, Jaffe DM, Kuppermann N, Kaufman BA, Lee BC, Park TS, et al. Diagnostic testing for acute head injury in children: when are head computed tomography and skull radiographs indicated? Pediatrics 1997;99:E11. 20. Lloyd DA, Carty H, Patterson M, Butcher CK, Roe D. Predictive value of skull radiography for intracranial injury in children with blunt head injury. Lancet 1997;349:821-4. 21. Husson B, Pariente D, Tammam S, Zerah M. The value of MRI in the early diagnosis of growing skull fracture. Pediatr Radiol 1996;26:744-7. 22. Masters SJ, McClean PM, Arcarese JS, Brown RF, Campbell JA, Freed HA, et al. Skull x-ray examinations after head trauma: recommendations by a multidisciplinary panel and validation study. N Engl J Med 1987;316:84-91. 23. Mierez R, Guillen A, Brell M, Cardona E, Claramunt E, Costa JM. Growing skull fracture in childhood: presentation of 12 cases. Neurocirugia (Astur) 2003; 14:228-34.

The Journal of Pediatrics • September 2007

CLINICAL AND LABORATORY OBSERVATIONS

In Vivo Proton Magnetic Resonance Spectroscopy Assessment for Muscle Metabolism in Neuromuscular Diseases TSYH-JYI HSIEH, MD, CHIEN-KUO WANG, MD, HUNG-YI CHUANG, MD, SCD, YUH-JYH JONG, MD, MMS, CHUN-WEI LI, PHD, AND GIN-CHUNG LIU, MD

Muscle metabolites were obtained by in vivo proton magnetic resonance spectroscopy of 3 patients with Duchenne muscular dystrophy (DMD), 6 patients with spinal muscular atrophy (SMA), and 10 normal volunteers. Patients with DMD and SMA had lower trimethyl amide (TMA)/water and TMA/total creatine (tCr) ratios but normal tCr/water ratios. (J Pediatr 2007;151:319-21)

ecent reports using in vivo proton magnetic resonance spectroscopy (MRS) have opened up the possibility of monitoring muscle metabolism.1-4 MRS has not yet been used to study muscles with fatty degeneration, however. Muscle atrophy and fatty degeneration are common findings in spinal muscular atrophy (SMA) and Duchenne muscular dystrophy (DMD).5-7 The objectives of this study were to assess the ability of MRS to evaluate metabolic spectra in muscles with fatty degeneration and to compare metabolites in patients with DMD, patients with SMA, and normal controls.

R

METHODS Patients We evaluated 19 MRS studies of 6 patients with SMA type III (4 female and 2 male; mean age, 11 years; range, 4 to 18 years), 3 patients with DMD (3 male; mean age, 9 years; range, 7 to 14 years), and 10 normal volunteers (5 female and 5 male; mean age, 12 years; range, 4 to 18 years) at our hospital between August 2003 and July 2005. We evaluated the functional abilities of lower limbs according to the protocol of Brooke. Informed consent was obtained from all patients and normal volunteers and their parents. From the Departments of Medical Imaging Magnetic Resonance Imaging and MRS All patients were examined using a 3.0 Tesla whole-body MR system (General Electric Medical Systems, Milwaukee, WI). A built-in body coil for conventional magnetic resonance imaging (MRI) and a knee coil for localization and detection of MRS were used. Both the SMA and DMD groups underwent conventional MRI protocols, including fast spin-echo T2-weighted axial images and spin-echo T1-weighted axial images of upper and lower extremities, before undergoing proton MRS. In the normal volunteers, only T2-weighted axial images and MRS sequences were performed. Data were collected with standard proton MRS acquisition software provided by the manufacturer. A volume of interest (2 cm ⫻ 2 cm ⫻ 2 cm) was positioned at the soleus (Figure 1). For localization, spectra were obtained using a point-resolved spin-echo sequence. Water suppression was accomplished using 3 preceding chemical-shift-selective saturation pulses (bandwidth, 60 Hz). The following acquisition measures were used: repetition time (TR), 1.5 seconds; 128 acquisitions; 4 dummy scans; spectral width, 2500 Hz; and 2048 data points. Three echo times (30, 90, and 144 msec) were used on the same volume of interest for each patient.

(T.H., C.W., G.L.), Clinical Research (H.C.), and Pediatrics (Y.J.), Kaohsiung Medical University Hospital; Department of Radiology, Kaohsiung Municipal Hsiao-Kang Hospital (T.H.); Departments of Radiology (C.W., G.L.) and Pediatrics (Y.J.), Faculty of Medicine, College of Medicine; and Faculties of Public Health (H.C.) and Medical Radiation Technology (C.L.), College of Health Sciences, Kaohsiung Medical University, Kaohsiung, Taiwan. Supported by the National Science Council of Taiwan (grant 92-2314-B-037-053). The sponsor had no involvement in the study design; the collection, analysis, and interpretation of data; the writing of the report; or the decision to submit the manuscript for publication. Submitted for publication Dec 12, 2006; last revision received Mar 12, 2007; accepted May 11, 2007. Reprint requests: Chun-Wei Li, PhD, 100 Tzyou 1st Road, San Ming District, Kaohsiung 807, Taiwan. E-mail: [email protected]. edu.tw.

DMD MRI MRS

0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.05.026

Duchenne muscular dystrophy Magnetic resonance imaging Magnetic resonance spectroscopy

SMA TCr TMA

Spinal muscular atrophy Total creatine Trimethyl amide

319

Figure 1. The regions of interest of MRS in the soleus. T1-weighted spin-echo image acquired from a normal volunteer A, and a patient with SMA B.

The MRS analysis package provided by the manufacturer (SAGE 7.1) was used. The raw data was zero-filled once, apodized with a 3-Hz Gaussian filter, Fourier-transformed, and phase- and baseline-corrected. Marquardt curve fitting was performed using a Gaussian line shape to calculate the area under the peak. Total creatine (tCr) and trimethyl amide (TMA) concentrations were measured, and the mean ratios of tCr/water, TMA/water, and TMA/tCr were compared for the different groups.

RESULTS Spectra of analyzable quality were obtained for the 9 patients and the 10 normal volunteers at 3 echo times separately. All of the spectra of normal volunteers and 7 patients showed peaks of TMA and creatine (Figure 2). TMA peaks could not be detected in 2 other patients (1 DMD, 1 SMA) at the 30-msec echo time, because the peaks could not be differentiated from baseline noise. The TMA/water ratios in the patients with DMD, those with SMA, and normal volunteers exhibited statistically significant differences at echo times of 30 msec (mean ⫾ standard deviation, 0.00082 ⫾ 0.00045, 0.00135 ⫾ 0.00032, and 0.00273 ⫾ 0.00025; P ⫽ .0016), 90 msec (0.00333 ⫾ 0.00221, 0.00547 ⫾ 0.00156, and 0.01091 ⫾ 0.00121; P ⫽ .0089), and 144 msec (0.00383 ⫾ 0.00967, 0.01853 ⫾ 0.00684, and 0.03489 ⫾ 0.00530; P ⫽ .027). The TMA/tCr ratios also showed statistically significant differences at echo times of 30 msec (0.325 ⫾ 0.154, 0.456 ⫾ 0.109, and 0.817 ⫾ 0.084; P ⫽ .014), 90 msec (0.649 ⫾ 0.173, 0.629 ⫾ 0.122, and 1.071 ⫾ 0.095; P ⫽ .021), and 144 msec (0.524 ⫾ 0.173, 0.678 ⫾ 0.122, and 1.226 ⫾ 0.095; P ⫽ .0017). However, the tCr/water ratios showed no statistically significant differences at echo times of 30 msec (0.00333 ⫾ 0.00129, 0.00467 ⫾ 0.00091, and 0.00336 ⫾ 0.00071; P ⫽ .510), 90 msec (0.00457 ⫾ 0.00333, 0.01017 ⫾ 0.00229, and 0.01048 ⫾ 0.00178; P ⫽ .287), and 144 msec (0.00970 ⫾ 0.00779, 320

Hsieh et al

0.02243 ⫾ 0.00551, and 0.02931 ⫾ 0.00426; P ⫽ .113). The Brooke scores of the lower limbs demonstrated significant statistical differences among the patients with DMD, those with SMA, and normal volunteers (3.30 ⫾ 0.58, 2.83 ⫾ 0.41, 1.00 ⫾ 0.00; P ⬍ .001).

DISCUSSION The patients with DMD and those with SMA exhibited significantly lower TMA/tCr and TMA/water ratios compared with normal volunteers, a result compatible with an earlier in vitro proton nuclear MRS study in human patients with DMD.8 The TMA peaks are constituents of phospholipid metabolism and cell membranes, and decreased TMA is considered to be associated with a lower number of cells, reduced rate of membrane synthesis, and decreased cell turnover.9 This decrease in TMA may reflect degenerative changes in the muscles of patients with DMD and SMA. Compared with normal muscles, the muscles of patients with DMD and SMA were found to have more adipose tissue, which may have rendered the other metabolic peaks undetectable. This made the TMA peaks undetectable in 2 patients at an echo time of 30 msec, but not at longer echo times. This suggests that long echo times are preferable when performing proton MRS studies of muscles with fatty degeneration. One limitation of our study is the small number of patients with neuromuscular disease. An earlier nuclear MRS study found significantly lower creatine concentrations in patients with DMD.8 However, the difference in tCr/water ratio among the normal volunteers and the patients with DMD and SMA did not reach statistical significance in our study. Another limitation is that fiber orientation, for which we had no information, significantly influences proton MRS studies of muscles.10,11 A multivoxel MRS that can include a larger field of muscle may provide information on variant fiber orientation. The Journal of Pediatrics • September 2007

Figure 2. Proton MRS of a normal volunteer A, and a patient with SMA B, at 3 echo times showing peaks of lipid, creatine, and choline.

In conclusion, we have found that in vivo proton MRS has potential in diagnosing or monitoring neuromuscular disease. Proton MRS at longer echo times should allow us to better detect metabolite spectra. Special thanks to Yuan-Yu Chiau and Feng-O Shu for technical assistance with the MR studies.

REFERENCES 1. Stueckle CA, Claeys L, Haegele K, Zimmermann S, Mruck S, Adams S, et al. Diagnostic value of proton MR spectroscopy in peripheral arterial occlusive disease: a prospective evaluation. AJR Am J Roentgenol 2006;187:1322-6. 2. Boesch C, Machann J, Vermathen P, Schick F. Role of proton MR for the study of muscle lipid metabolism. NMR Biomed 2006;19:968-88. 3. Hu J, Xu Y, Jiang Q, Sehgal V, Shen Y, Xuan Y, et al. Spectral pattern of total creatine and trimethyl ammonium in multiple sclerosis. Magn Reson Imaging 2004;22:427-9. 4. Trump ME, Hanstock CC, Allen PS, Gheorghiu D, Hochachka PW. An (1)

H-MRS evaluation of the phosphocreatine/creatine pool (tCr) in human muscle. Am J Physiol Regul Integr Comp Physiol 2001;280:R889-96. 5. Chan WP, Liu GC. MR imaging of primary skeletal muscle diseases in children. AJR Am J Roentgenol 2002;179:989-97. 6. Liu GC, Jong YJ, Chiang CH, Jaw TS. Duchenne muscular dystrophy: MR grading system with functional correlation. Radiology 1993;186:475-80. 7. Liu GC, Jong YJ, Chiang CH, Yang CW. Spinal muscular atrophy: MR evaluation. Pediatr Radiol 1992;22:584-6. 8. Sharma U, Atri S, Sharma MC, Sarkar C, Jagannathan NR. Skeletal muscle metabolism in Duchenne muscular dystrophy (DMD): an in vitro proton NMR spectroscopy study. Magn Reson Imaging 2003;21:145-53. 9. Hu J, Xia Y, Shen Y, et al. Significant differences in proton trimethyl ammonium signals between human gastrocnemius and soleus muscle. J Magn Reson Imaging 2004;19:617-22. 10. Gao F, Bottomley PA, Arnold C, Weiss RG. The effect of orientation on quantification of muscle creatine by 1H MR spectroscopy. Magn Reson Imaging 2003;21:561-6. 11. Vermathen P, Boesch C, Kreis R. Mapping fiber orientation in human muscle by proton MR spectroscopic imaging. Magn Reson Med 2003;49:424-32.

In Vivo Proton Magnetic Resonance Spectroscopy Assessment for Muscle Metabolism in Neuromuscular Diseases

321

ABCA3 Deficiency Presenting as Persistent Pulmonary Hypertension of the Newborn ANETTE M. KUNIG, MD, THOMAS A. PARKER, MD, LAWRENCE M. NOGEE, MD, STEVEN H. ABMAN, MD,

AND JOHN

P. KINSELLA, MD

A newborn with persistent pulmonary hypertension (PH) unresponsive to conventional therapies was found to be homozygous for a mutation in the gene encoding adenosine triphosphate binding cassette protein, member A3 (ABCA3). Most causes of PH respond to lung recruitment, inhaled nitric oxide, and hemodynamic support. When PH is prolonged and does not respond to standard therapies, genetic causes of surfactant abnormalities should be considered in the differential diagnosis. (J Pediatr 2007;151:322-4)

4.1-kg full-term male infant was delivered at an outlying hospital after an uncomplicated pregnancy and labor. He developed respiratory distress in the delivery room and was initially managed with continuous positive airway pressure. During the first 12 hours of life, his gas exchange worsened, necessitating intubation and mechanical ventilation. A chest radiograph obtained early in the patient’s course showed mild parenchymal lung disease, which did not fully account for the patient’s degree of severe hypoxemia (Figure). Mean airway pressure was 20 cm H2O at the time of chest radiograph. He was not treated with exogenous surfactant due to his unstable clinical condition, tenuous oxygenation, and chest radiograph inconsistent with typical surfactant deficiency. The patient was transferred to the Children’s Hospital at age 30 hours for evaluation and management of hypoxemic respiratory failure with progressive deterioration in gas exchange and hemodynamic instability. An echocardiogram on admission revealed a structurally normal heart with bidirectional shunting at the ductus arteriosus and right-to-left shunting at the atrial level consistent with systemic levels of pulmonary artery pressure. His gas exchange and hemodynamics stabilized during treatment with high-frequency oscillatory ventilation (HFOV), inhaled nitric oxide (iNO), and high-dose dopamine infusion. The infant was treated with antibiotics, and infectious causes for pulmonary hypertension (PH) were ruled out with negative blood cultures. A complete family history was obtained, which was negative for respiratory disorders. There was no history of consanguinuity.

A

Hospital Course Hypoxemic respiratory failure persisted over the first 5 days of life, with little improvement in ventilatory support requirements. Serial echocardiography showed near-systemic levels of pulmonary artery pressure despite iNO treatment. There was no echocardiographic evidence of structural heart disease, severe left ventricular systolic or diastolic dysfunction, or pulmonary venous stenosis to account for the patient’s persistent PH. Due to the severity and persistence of the PH, a lung biopsy was performed at age 6 days to rule out alveolar-capillary dysplasia (ACD); however, the lung histology was not diagnostic for ACD, but rather showed partial desquamation of the respiratory epithelium from memFrom the Pediatric Heart Lung Center, Debranous and terminal bronchioles. In addition, there was marked alveolar epithelial partment of Pediatrics, University of Colorado Health Sciences Center, The Children’s hyperplasia and moderate widening of alveolar walls without prominent cellular proliferHospital, Denver, CO (A.K., T.P., S.A., J.K.) ation. This histology suggested a surfactant dysfunction mutation. and Department of Pediatrics, Johns Hopkins Children’s Center, Baltimore, MD (L.N.). Because of a persistent left pneumothorax after the biopsy procedure that contribSubmitted for publication Oct 31, 2006; uted to worsening hypoxemia, and in the absence of a diagnosis of severe persistent last revision received Apr 27, 2007; acpulmonary hypertension of the newborn (PPHN), the infant was cannulated for extracepted May 31, 2007. corporeal membrane oxygenation (ECMO). Despite ECMO therapy, however, he conReprint requests: Anette M. Kunig, MD, Pediatric Heart Lung Center, University of tinued to exhibit evidence of systemic levels of pulmonary artery pressure over a 10-day Colorado School of Medicine and The ABCA3 ACD ECMO HFOV iNO

322

Adenosine triphosphate binding cassette protein, member A3 Alveolar-capillary dysplasia Extracorporeal membrane oxygenation High-frequency oscillatory ventilation Inhaled nitric oxide

PH PPHN RDS SP

Pulmonary hypertension Persistent pulmonary hypertension of the newborn Respiratory distress syndrome Surfactant protein

Children’s Hospital, Pediatrics, Mail Stop 8317, PO Box 6511, Aurora, CO 80045. E-mail: [email protected]. 0022-3476/$ - see front matter Copyright © 2007 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2007.05.054

Figure. Chest radiograph of the patient early in the hospital course. The lungs are well expanded with minimal parenchymal infiltrate, suggesting that hypoxemia cannot be explained by the degree of lung disease alone.

course. Brief trials off iNO caused pulmonary artery pressure to rapidly become suprasystemic. Cardiac catheterization to evaluate for discrete pulmonary vein stenosis was declined by the infant’s family. After discussion with the family regarding the futility of continued ECMO support, therapy was withdrawn, and the patient died. Significant findings at autopsy included hyperplasia and adventitial fibrosis of the pulmonary arteries, as well as right ventricular hypertrophy. Studies obtained before death did not reveal the most common gene mutation for surfactant protein (SP)-B deficiency. However, electron microscopy of the lung specimen revealed absence of lamellar bodies, a finding consistent with ABCA3 deficiency. In the most severely affected infants with this disorder, normally formed, normally sized lamellar bodies are often absent in the cytoplasm of type II cells. The cytoplasm of type II cells in ABCA3-deficient infants instead may contain small dense bodies that likely reflect abnormally formed lamellar bodies, but the precise nature of these bodies has not yet been determined.1-3 Permission was obtained to enroll the child in a study to identify genetic mechanisms of lung disease. DNA was prepared from peripheral blood leukocytes as described previously,4 and the child was subsequently found to be homozygous for a missense mutation (L326R) in the gene encoding ABCA3.

Clinical Discussion PPHN is a complex disorder associated with a wide array of cardiopulmonary diseases characterized by marked pulmonary hypertension and altered vasoreactivity, leading to right-to-left shunting of blood across the patent ductus arte-

riosus and foramen ovale.5,6 For near-term and term newborns with hypoxemic respiratory failure and PPHN, therapies including iNO, HFOV, and exogenous surfactant treatment have decreased the use of ECMO over the last decade. Some patients fail to respond to standard treatment with iNO, HFOV, and ECMO, however. Up to 40% mortality has been reported in patients with refractory PH,7 and mortality remains very high in those patients who ultimately require ECMO. Failure to respond to iNO therapy has been associated with alveolar capillary dysplasia, discrete pulmonary vein stenosis, and severe lung hypoplasia, as seen with congenital diaphragmatic hernia. More recently, genetic abnormalities of surfactant function, also known as surfactant dysfunction mutations,8 have been recognized in patients who remain hypoxemic despite a prolonged course of iNO and ECMO.9 The last few years has brought increased recognition of surfactant dysfunction mutations and a better appreciation of their clinical presentation. Often these patients are full-term newborns with hypoxemic respiratory failure. SP-B deficiency is characterized by progressive respiratory distress soon after birth, which ultimately leads to respiratory failure and death.10 Chest radiographs typically demonstrate diffuse ground-glass opacities. Patients with SP-C mutations have varied courses ranging from mild respiratory symptoms to severe respiratory failure.10 Radiographically, these patients show diffuse interstitial infiltrates. These surfactant dysfunction mutations are characterized by parenchymal lung disease rather than pulmonary vascular disease, and persistent PH is not commonly reported. Recently, mutations in the gene encoding the transporter ABCA3 have been reported as a cause of severe neonatal lung disease.11,12 ABCA3 is localized to the limiting membrane of lamellar bodies, organelles containing concentric, onion-like layers of surfactant. Infants with ABCA3 mutations lack typical lamellar bodies, indicating that ABCA3 is critical to their formation.12,13 ABCA3 deficiency is inherited in an autosomal-recessive manner, and the clinical features have not been described in detail. In the limited number of cases of ABCA3 deficiency reported to date, newborns presented with initial signs of respiratory distress syndrome (RDS) and rapidly progressive respiratory failure that was refractory to ventilation and ECMO. Radiographic findings have not been specifically reported but have been described as showing diffuse pulmonary opacification, reticular-granular infiltrates, and air bronchograms consistent with RDS.10 In general severe, prolonged PH has not been reported as a prominent feature in children with ABCA3 mutations. Milder lung disease associated with prolonged survival due to ABCA3 mutations also has been reported recently.4 Here we report a case of ABCA3 deficiency presenting with severe hypoxemic respiratory failure and refractory PH. Our patient differed from those previously reported because his primary presentation was severe PH that was unresponsive to traditional therapies, including iNO and HFOV. He did

ABCA3 Deficiency Presenting as Persistent Pulmonary Hypertension of the Newborn

323

not present with the typical clinical signs of RDS, and for the first several days of life he had relatively clear lung fields on chest radiographs. He did not develop evidence of parenchymal lung disease until much later in his hospital course, and his initial presentation was severe PH that appeared to be out of proportion to his degree of lung disease. Information is emerging on the clinical spectrum of patients with surfactant dysfunction mutations. As more patients with these genetic disorders are identified, we may be able to begin to recognize patterns of clinical presentation for these diseases. Surfactant dysfunction mutations are associated with marked changes in the airspaces, alveolar epithelium, and interstitium. Whether or not these diseases are also associated with abnormalities of pulmonary vascular development remains unknown. Our patient failed to respond to iNO despite adequate lung recruitment and cardiac support. The other etiologies of PH that typically fail to respond to iNO were excluded after appropriate evaluation. This case suggests that perhaps iNO nonresponders and/or patients with prolonged, refractory PH should be evaluated for surfactant dysfunction mutations as well. Experience with these rare genetic surfactant abnormalities is limited at this time. Refractory PH may turn out to be common among patients with this disease. Until we understand more about the presentation and clinical course of this disease, it is important to keep a high index of suspicion in patients with severe PH. We thank the physicians, staff, and consultants involved in the care of this patient. We are grateful to the investigators who

324

Kunig et al

provided their time and helpful comments in the preparation of this manuscript.

REFERENCES 1. Tryka AF, Wert SE, Mazursky JE, Arrington RW, Nogee LM. Absence of lamellar bodies with accumulation of dense bodies characterizes a novel form of congenital surfactant defect. Pediatr Dev Pathol 2000;3:335-45. 2. Cutz E, Wert SE, Nogee LM, Moore AM. Deficiency of lamellar bodies in alveolar type II cells associated with fatal respiratory disease in a full-term infant. Am J Respir Crit Care Med 2000;161(2 Pt 1):608-14. 3. Bullard JE, Wert SE, Nogee LM. ABCA3 deficiency: neonatal respiratory failure and interstitial lung disease. Semin Perinatol 2006;30:327-34. 4. Bullard JE, Wert SE, Whitsett JA, Dean M, Nogee LM. ABCA3 mutations associated with pediatric interstitial lung disease. Am J Respir Crit Care Med 2005; 172:1026-31. 5. Kinsella JP, Abman SH. Clinical approach to inhaled nitric oxide therapy in the newborn with hypoxemia. J Pediatr 2000;136:717-26. 6. Kinsella JP, Abman SH. Recent developments in the pathophysiology and treatment of persistent pulmonary hypertension of the newborn. J Pediatr 1995;126:853-64. 7. Walsh-Sukys MC, Tyson JE, Wright LL, Bauer CR, Korones SB, Stevenson DK, et al. Persistent pulmonary hypertension of the newborn in the era before nitric oxide: practice variation and outcomes. Pediatrics 2000;105(1 Pt 1):14-20. 8. Fan LL, Deterding RR, Langston C. Pediatric interstitial lung disease revisited. Pediatr Pulmonol 2004;38:369-78. 9. Nogee LM. Genetic mechanisms of surfactant deficiency. Biol Neonate 2004;85: 314-8. 10. Whitsett JA, Wert SE, Trapnell BC. Genetic disorders influencing lung formation and function at birth. Hum Mol Genet 2004;13(Spec 2):R207-15. 11. Nagata K, Yamamoto A, Ban N, Tanaka AR, Matsuo M, Kioka N, et al. Human ABCA3, a product of a responsible gene for abca3 for fatal surfactant deficiency in newborns, exhibits unique ATP hydrolysis activity and generates intracellular multilamellar vesicles. Biochem Biophys Res Commun 2004;324:262-8. 12. Shulenin S, Nogee LM, Annilo T, Wert SE, Whitsett JA, Dean M. ABCA3 gene mutations in newborns with fatal surfactant deficiency. N Engl J Med 2004;350: 1296-303. 13. Hallman M. Lung surfactant, respiratory failure, and genes. N Engl J Med 2004; 350:1278-80.

The Journal of Pediatrics • September 2007

INSIGHTS

Severe Cerebellar Hypoplasia Associated with Osteogenesis Imperfecta Type III teoporotic bone with metaphyseal flaring. With this picture, the diagnosis of osteogenesis imperfecta type III was made. MRI of the brain showed severe cerebellar hypoplasia, associated with angulation of the brainstem, but no basilar invagination (Figure). The recognized neurological complications of osteogenesis imperfecta include cranio-vertebral junction anomalies such as basilar invagination, syringohydromyelia, hydrocephalus, and brainstem compression; there has also been an isolated report of Dandy-Walker malformation.1 There are only two reported cases of cerebellar hypoplasia associated with osteogenesis imperfecta type IV and V.2,3 The proposed mechanism for cerebellar hypoplasia is in utero vascular compromise with hypoplastic posterior circulation structures due to associated cranio-vertebral junction anomalies. Neurologic evaluation should be part of a team approach in the management of patients with severe osteogenesis imperfecta types. B. Tabarki, MD S. Al-Malki, MD H. Al-Ghamdi, MD Department of Pediatrics Al Hada Military Hospital Taif, Kingdom of Saudi Arabia

REFERENCES Figure. Sagittal MRI of the brain (T1 sequence) showing severe cerebellar atrophy with brain stem angulation.

four-year-old girl was admitted with recurrent pneumonias. She was found to have triangular facies, generalized, non-paralytic hypotonia, severe scoliosis, pectal deformity, extreme short stature, and blue scleral hue. Cerebellar signs were absent. Radiographs showed os-

A

1. Charnas LR, Marini JC. Communicating hydrocephalus, basilar invagination and other neurologic features in osteogenesis imperfecta. Neurology 1993;43:2603-8. 2. Zhou LJ, Khong PL, Wong KY, Ooi GC. A case of cerebelReprint requests: Brahim Tabarki, Departlar hypoplasia in a Chinese infant ment of Pediatrics, Al Hada Military Hospiwith osteogenesis imperfecta. tal, Taif, Kingdom of Saudi Arabia. E-mail: Hong Kong Med J 2004;10: [email protected]. 211-3. J Pediatr 2007;151:325 3. Syamlal S, Shine S, Kunju 0022-3476/$ - see front matter M. Brainstem and cerebellar hyCopyright © 2007 Mosby Inc. All rights poplasia associated with osteoreserved. genesis imperfacta type-5. J Postgrad Med 2006;52:152-3. 10.1016/j.jpeds.2007.05.027

325

CURRENT BEST EVIDENCE

Clinical Research Abstracts for Pediatricians EDITOR’S NOTE: Journals reviewed for this issue: Archives of Disease in Childhood, Archives of Pediatrics and Adolescent Medicine, British Medical Journal, Journal of the American Medical Association, Journal of Pediatrics, The Lancet, New England Journal of Medicine, Pediatric Infectious Diseases Journal, and Pediatrics. Gurpreet K. Rana, BSc, MLIS, Taubman Medical Library, University of Michigan, contributed to the review and selection of this month’s abstracts. —John G. Frohna, MD, MPH

Adenotonsillectomy less beneficial for sleep apnea in older and obese children Tauman R, Gulliver TE, Krishna J, Montgomery-Downs HE, O’Brien LM, Ivanenko A, et al. Persistence of obstructive sleep apnea syndrome in children after adenotonsillectomy. J Pediatr 2006;149:803-8. In children with obstructive sleep apnea (OSA), are there factors that predict whether symptoms will resolve after adenotonsillectomy (T&A)?

Question

Design

Cohort study.

Setting

Children’s Hospital, Louisville, Kentucky.

Participants 110 children, age 1 to 16 years (mean age, 6.4

years), along with 22 control children. Intervention The intervention group underwent 2 polysomnographic evaluations before and after T&A. History of allergy and family history of sleep-disordered breathing was obtained before each polysomnographic evaluation.

Persistence of sleep-disordered breathing, as defined by an apnea/hypopnea index (AHI) ⬎ 5/hour of total sleep time (TST).

Outcome

Significant changes in sleep stage percentages and sleep fragmentation were found in the postsurgery study compared with the presurgery study; 25% of the children had an AHI ⬍ 1, 46% had an AHI of 1 to 5, and 29% had an AHI ⬎ 5 in the postsurgery study. The frequency of children with an AHI ⬎ 5/h of TST after surgery was higher in the obese subjects than in the nonobese subjects (36.4% vs 17.6%, P ⬍ .05).

Commentary OSA is common and potentially harmful. T&A is often used to treat OSA in children. Previous studies have shown that abnormal polysomnographic findings persist after T&A in up to 40% of cases. In the present study, 29% of the patients continued to have more than 5 apneic or hypopneic episodes per hour of sleep time after T&A. Older patients and patients with high body mass index were even more likely to continue to experience OSA after surgery. However, no data on patient symptoms were obtained. It is possible that even patients with an AHI ⬎ 5 after T&A experienced improvement in symptoms after the procedure. Notably, children with genetic disorders, cerebral palsy, neuromuscular disease, or any underlying systemic disease were excluded from the study, so it is unclear how these children might benefit from T&A. It also is unclear whether the initial group of 110 patients was chosen prospectively or retrospectively, which could affect the validity of the data. When recommending T&A to parents as a treatment for OSA, physicians need to provide appropriate education so that parents will have realistic expectations. Families should be told that T&A may not cure OSA, particularly in older and obese patients. In addition, this study suggests that postoperative polysomnography should be considered in patients undergoing T&A for OSA.

Main Results

Conclusions T&A yields improvements in respiratory abnor-

malities in children with OSA syndrome, although complete normalization occurs in only 25% of patients. Obesity and AHI at diagnosis are the major determinants of surgical outcome. When normalization of respiratory measures occurs after surgery, normalization of sleep architecture will also ensue. 326

Sarah Lacy, MD University of Michigan Ann Arbor, Michigan

Predictive value of rapid influenza tests varies with prevalence Grijalva CG, Poehling KA, Edwards KM, Weinberg GA, Staat MA, Iwane MK, et al. Accuracy and interpretation of rapid influenza tests in children. Pediatrics 2007;119:e6-11. Among young children, how does prevalence of influenza alter the predictive ability of rapid influenza tests?

Question Design

Longitudinal cohort. The Journal of Pediatrics • September 2007

Setting

Three counties in the Unites States.

Participants Children enrolled in the New Vaccine Surveillance Network, age 5 years and younger, who were hospitalized with respiratory symptoms or fever between October 2000 and September 2004. Outpatients from 2002 to 2004 also were included.

Nasal and throat swabs were obtained, and influenza virus was detected by culture and reverse-transcription polymerase chain reaction (RT-PCR). Provider-ordered rapid influenza tests were compared with the criterion standard (culture and RT-PCR) to determine their sensitivity and specificity. Intervention

Trends in weekly predictive values of the rapid tests estimated during the influenza seasons.

Outcome

Main Results Rapid influenza tests had an overall sensitivity

of 63% and specificity of 97%. In 2002-2003, the prevalence of influenza in symptomatic outpatient children peaked at 21% and remained above 10% for ⬃4 weeks. In contrast, in 2003-2004, the prevalence peaked at 60% and remained above 20% for ⬃6 weeks. The positive predictive value (PPV) of the rapid tests approached 80% when the influenza prevalence was ⱖ15% but decreased to ⬍70% when the prevalence was ⬍10%. Conclusions The prevalence of influenza varies between and within seasons. On the basis of these estimates, rapid tests are of limited use when the prevalence is ⬍10%. The appropriate interpretation of rapid influenza tests requires local influenza surveillance and timely communication of this information to practitioners.

Influenza is a very common childhood disease, responsible for many pediatric ambulatory visits and hospitalizations during the winter months in the northern hemisphere. This study conveys an important message about interpreting diagnostic tests and reiterates the importance of disease prevalence in test interpretation. That is, when the disease prevalence is low, the PPV of a test also will be low, and when the disease prevalence is high, the PPV will be high. Several important points should be considered when examining the results of the present study. First, although all the tests were compared using a gold standard procedure (viral culture or RT-PCR), several antigen tests with differing performance characteristics, were lumped together. This affects interpretation of the results of a single rapid test used in an institution. Second, to assess whether a test is valid, we need to ask whether the diagnostic test was evaluated in an appropriate spectrum of patients (eg, those in whom it would be used in practice). Rapid tests to either rule in or rule out disease are useful primarily in outpatient settings for such reasons as to reduce the need for further testing and/or antibiotic therapy and to reassure parents. In this study, the rapid tests were performed only on inpatients, and the results were translated to the outpatient population. This limits the

Commentary

Clinical Research Abstracts for Pediatricians

generalizability of the results to the outpatient and emergency department settings, where these tests are most often used. Vidya Sharma, MBBS, MPH The Children’s Mercy Hospital Kansas City, Missouri

Failure to respond to name is indicator of possible autism spectrum disorder Nadig AS, Ozonoff S, Young GS, Rozga A, Sigman M, Rogers SJ. A prospective study of response to name in infants at risk for autism. Arch Pediatr Adolesc Med 2007; 161:378-83. Question Among children at high risk for autism, does failure to respond to name accurately predict a subsequent diagnosis of autism spectrum disorder (ASD)? Design

Prospective longitudinal study of infants at risk for

autism. Setting

University medical center.

Participants Infants at risk for autism (55 age 6 months and

101 age 12 months) and a control group at no known risk (43 age 6 months and 46 age 12 months). To date, 46 at-risk infants and 25 control infants have been followed up to 24 months. Intervention

Experimental task eliciting response-to-name

behavior. Autism Diagnostic Observation Schedule, Mullen Scales of Early Learning.

Outcome

At age 6 months, there was a nonsignificant trend for control infants to require a fewer number of calls to respond to name compared with the infants at risk for autism. At age 12 months, 100% of infants in the control group “passed,” responding on the first or second name call, compared with 86% in the at-risk group. Three-fourths of the children who failed the task were identified with developmental problems at age 24 months. Specificity of failure to respond to name was 0.89 for ASD and 0.94 for any developmental delay; sensitivity was 0.50 for ASD and 0.39 for any developmental delay. For a diagnosis of ASD, the likelihood ratios are 4.55 for a positive test and 0.56 for a negative test.

Main Results

Conclusions Failure to respond to name by age 12 months is highly suggestive of developmental abnormality but does not identify all children at risk for developmental problems. Lack of response to name is not universal among infants later diagnosed with ASD and/or other developmental delays. Poor response to name may be a trait of the broader autism phenotype in infancy.

Developmental pediatricians have advocated for earlier diagnosis of autism because of growing evidence that early intervention may improve long-term outcomes. However, no single diagnostic test has been found to be reliable and valid. Retrospective studies using videotapes of

Commentary

327

children subsequently diagnosed with autism have suggested a consistently decreased response to name. This also has been seen in some prospective studies. The current study compared a group of high-risk children (infant siblings of a child diagnosed with autism) with a control group representative of the general population. By age 12 months, children who did not respond to their name were much more likely to be diagnosed with developmental delay or an ASD. Although the sensitivity was low, the specificity was decent, giving a positive likelihood ratio of 4.55 for the diagnosis of ASD. This was a small study, and the authors are still planning to follow all of the children to age 36 months. Larger numbers of children also will produce more stable estimates of sensitivity and specificity. In addition, this test was applied in a group of high-risk children, resulting in spectrum bias, and care needs to be taken when applying this to the general population of patients seen in practice. Nonetheless, response-to-name behavior is an easy and inexpensive test. A high-risk child who fails to respond to his or her name at 12 months should be referred for further testing. John G. Frohna, MD, MPH University of Wisconsin Children’s Hospital Madison, Wisconsin

ALSO NOTED Hutton EK, Hassan ES. Late versus early clamping of the umbilical cord in full-term neonates: Systematic review and meta-analysis of controlled trials. JAMA 2007;297:1241-52. Delayed clamping of the umbilical cord after delivery results in placental transfusion. In children born vaginally, a 2- to 3-minute delay in clamping can increase the neonate’s blood volume by 20 to 30 mL/kg. The potential benefits and drawbacks of late cord clamping were assessed using meta-analysis. This validation study examined 15 controlled trials involving 1912 newborns. At age 2 to 6 months, those with delayed clamping (⬎2 minutes) demonstrated increased hematocrit,

328

ferritin, and stored iron levels. Importantly, these children also had a clinically significant decrease in the risk of anemia (relative risk ⫽ 0.53; 95% confidence interval ⫽ 0.4 to 0.7). Newborns with late clamping were at increased risk for asymptomatic polycythemia, but no other adverse effects were noted. Of note, this study did not address the effect of the current practice of administering oxytocin at the end of labor, and an accompanying editorial does not recommend altering this practice. Clark E, Plint AC, Correll R, Gaboury I, Passi B. A randomized, controlled trial of acetaminophen, ibuprofen, and codeine for acute pain relief in children with musculoskeletal trauma. Pediatrics 2007;119:460-7. Which pain medication (acetaminophen, ibuprofen, or codeine) provides the best analgesia for children with musculoskeletal injuries? This question was investigated by a randomized, controlled trial of children age 6 to 17 years who presented to the emergency department with pain from a musculoskeletal injury sustained in the previous 48 hours. Children were randomized to receive 1 oral dose of acetaminophen 15 mg/kg, ibuprofen 10 mg/kg, or codeine 1 mg/kg. Using a visual analog scale, children rated their own pain. At 30 minutes, there was no statistical difference between pain scores in the 3 groups. However, at 60 or more minutes, patients who had received ibuprofen exhibited significantly greater improvement in pain scores compared with those in the acetaminophen or codeine groups. In addition, more patients in the ibuprofen group had achieved adequate analgesia compared with the other 2 groups (52% vs 36% and 40%; P ⬍ .001; NNT ⫽ 7 and 9, respectively). Ibuprofen appears to be more effective than acetaminophen or codeine in the treatment of acute musculoskeletal pain, particularly if a fracture is present. However, ibuprofen alone often provides inadequate pain control.

The Journal of Pediatrics • September 2007

LETTERS

Transfusion threshold in anemic premature infants To the Editor: The Iowa1 and Premature Infants in Need of Transfu2 sion trials did not resolve an important question: “How far can we push the anemic preterm infant before transfusion?” In these trials, the low hematocrit threshold values for restrictive transfusion practices in preterm infants with no respiratory support were 26% ⫾ 5% and ⱕ23% (hemoglobin of ⱕ75 g/L), respectively. We evaluated cardiac function in stable, very-low-birth weight infants with hematocrit ranges of 14% to 21%3 and found that these infants have much higher end-diastolic and end systolic diameters and stroke volumes in comparison to infants with higher hematocrit levels. On the basis of echocardiographic measures, we suggested that many stable infants with hematocrit levels ⱕ 21% are in a high cardiac output state. The cutoff values for abnormal end-diastolic and end systolic diameters and stroke volumes were approximately 15 mm, 10 mm, and 2.6 mL/kg, respectively. Because the current “traditional” criteria for packed red blood cell transfusion are not sensitive, new variables are needed to guide the clinician when to transfuse the anemic preterm infant. Arie L. Alkalay, MD Charles F. Simmons, MD Division of Neonatology Department of Pediatrics Ahmanson Pediatric Center Cedars-Sinai Medical Center David Geffen School of Medicine at UCLA Los Angeles, California 10.1016/j.jpeds.2007.05.016

REFERENCES 1. Bell EF, Strauss RG, Widness JA, Mahoney LT, Mock DM, Seward VJ, et al. Randomized trial of liberal versus restrictive guidelines for red blood cell transfusions in preterm infants. Pediatrics 2005;115:1685-91. 2. Kirpalani H, Whyte RK, Andersen C, Asztalos EV, Heddle N, Blajchman MA, et al. The Premature Infants in Need of Transfusion (PINT) study: A randomized, controlled trial of a restrictive (LOW) versus liberal (HIGH) transfusion threshold for extremely low birth weight infants. J Pediatr 2006;149:301-7. 3. Alkalay AL, Galvis S, Ferry DA, Simmons CF, Krueger RC. Hemodynamic changes in anemic premature infants: are we allowing the hematocrits to fall too low? Pediatrics 2003;112:838-45.

Reply To the Editor: Drs Alkalay and Simmons point out that the lowest safe limits for transfusion are not yet known for preterm newborns. We agree. This reiterates our own cautious conclusion that “transfusion thresholds in ELBW infants can be moved downwards by at least 10 g/L.”1 Pushing the lower limits of transfusion further remains to be tested in randomized cone10 Letters

trolled trials with clinically meaningful primary outcomes. Although this approach is open to the charge of being a tedious and iterative stepwise procedure, it remains the prudent way forward. In our view, extrapolations to lower guidelines than were tested by the Premature Infants in Need of Transfusion Study and the Iowa study are premature. The situation may be compared with the historical yo-yo-ing back and forth of “appropriate oxygen levels” in the absence of specific targeted trials. It is only now that a series of coordinated international trials is addressing this dilemma in the appropriate manner. They all use clinically relevant primary outcomes. Alkalay and Simmons also make an interesting proposal that the criteria for establishing transfusion thresholds for preterm newborns should be physiologically based, presumably by use of echocardiography. We are aware of their work cautioning that lower limits of hemoglobin are associated with echocardiographic findings considered abnormal.1 Others have also tried to use physiological measures, including fractional oxygen extraction.2 However, use of such surrogate physiological measures have their own issues. For example, is the relative risk of a transfusion better or worse than an adaptation to lower hemoglobin with a raised cardiac output? If surrogate physiological measures are used, they need to be robust and highly predictive of meaningful (preferably longterm) clinical outcomes.3 Ultimately, their use would still require evaluation via randomized controlled trials. H. Kirpalani, MSc, FRCP(UK) R. Whyte, MB, FRCP(C) R. Roberts, M.TECH Department of Pediatrics and Clinical Epidemiology McMaster University Hamilton, Ontario, Canada 10.1016/j.jpeds.2007.05.015

REFERENCES 1. Kirpalani H, Whyte RK, Andersen C, Asztalos EV, Heddle N, Blajchman MA, et al. The Premature Infants in Need of Transfusion (PINT) study: a randomized, controlled trial of a restrictive (low) versus liberal (high) transfusion threshold for extremely low birth weight infants. J Pediatr 2006;149:301-7. 2. Wardle SP, Yoxall CW, Crawley E, Weindling AM. Peripheral oxygenation and anemia in preterm babies. Pediatr Res 1998;44:125-31. 3. Zhang B, Schmidt B. Do we measure the right end points? A systematic review of primary outcomes in recent neonatal randomized clinical trials. J Pediatr 2001; 138:76-80.

The natural history of thyroid autoimmunity and thyroid function in children with type 1 diabetes To the Editor: We read with interest the report by Radetti et al1 regarding the natural history of euthyroid Hashimoto thyThe Journal of Pediatrics • September 2007

roiditis in children. We have some questions about their data. The criteria for selection of patients and their inclusion in the study have not been made clear. According to the authors, “the presence of Hashimoto thyroiditis was diagnosed on the basis of typical ultrasound imaging findings and presence of thyroid peroxidase antibodies and thyroglobulin antibodies.” Because clinical presence of goiter was not one of the criteria for patient selection, one must surmise that the children at risk were screened by the measurement of thyroid antibodies before they were subjected to a more expensive and involved procedure such as ultrasound imaging. However, in Figures 1, 3, and 4, up to 50% of patients are shown to have normal (grade 0) TPO and TG antibodies. How were these patients identified as having Hashimoto thyroiditis? Also, 2 measurements of any markers of a chronic condition, separated by a few weeks to 32 years, hardly provide information about the “natural history” of the condition especially when the study is conducted in a retrospective manner. We measured thyroid antibodies (thyrAB) every 3 to 6 months in 236 white, black, and Hispanic (Puerto Rican) children (x ⫽ 10.4 years) with type 1 diabetes (T1D) in a 5.5-year prospective study. Initial thyrAB were measured at diagnosis of T1D in 99 children (Group I) and sometime subsequent to diagnosis in 137 children (Group II) (Table). Follow-up was from 1 year to 5.5 years (x ⫽ 2.9 years) in 160 children (62 from Group I; 98 from Group II). The prevalence of positive thyroid antibodies at diagnosis of T1D was 17.4%, similar to that reported by others.2-4 There was no significant difference in age between those children with negative and positive thyrAb. The natural history of thyrAb in our population showed that 10% of children demonstrated conversion of their initial thyrAb status. In 7 children, thyrAb were detected intermittently; in 4 initially positive, thyrAb became negative; in 5 initially negative, thyrAb became positive. Fifty children (21%) had positive antibodies at some point during the study. The prevalence of positive thyrAb obtained subsequent to diagnosis of T1D was significantly greater than those obtained at time of diagnosis (P ⬍ .05). The highest prevalence of positive thyrAb was in our population of Hispanic children. This is congruent with our data from the Philadelphia Pediatric Type 1 Diabetes Registry that has been maintained since 1985. We demonstrated that Hispanic children in Philadelphia have the highest incidence of T1D of any racial group in the United States.5-7 In children with positive thyrAb, T4 and TSH were measured every 3 to 6 months. Ten children (20%) had normal T4 levels, and their serum TSH levels were 1.5 to 5

times the upper limit of normal; in 70% of these patients, serum levels of TSH normalized without treatment. One of these children had a peak TSH level of 25 mU/L. Although height deceleration has been reported in children with subclinical hypothyroidism,8 this was not demonstrated in our population, nor were there any other signs or symptoms of hypothyroidism. The prevalence of elevated TSH in our population is lower than that reported by Radetti et al,1 yet the prevalence of overt hypothyroidism in our group of patients was similar to the 5.6% in the population described by Radetti et al.1 Of 50 of our subjects who had positive thyrAb at any point in the study, only 3 (6%) developed overt hypothyroidism. Age, duration of diabetes, and level of thyrAb were not associated with the development of overt hypothyroidism. Our data demonstrate that, as with many autoimmune disorders, there are racial disparities in the prevalence of thyroid autoimmunity in children with T1D. In our population, thyrAb at the time of diagnosis did not always predict subsequent presence or absence of antibodies. Therefore a single measurement of thyrAB at diagnosis of diabetes is not sufficient to identify thyroid autoimmunity. In some children, positive thyrAB occur later, in others positive antibodies may spontaneously remit. We concur with Radetti et al1 that most children with thyroid autoimmunity and TSH levels above normal do not develop overt hypothyroidism. Our data, and those of Moore,9 show that even patients with elevated TSH may have a benign course, and in many subjects (70% of our population) the TSH spontaneously normalizes. Follow-up, rather than empirical treatment of those children may be indicated. Terri H. Lipman, PhD, CRNP University of Pennsylvania School of Nursing Division of Endocrinology, Children’s Hospital of Philadelphia Philadelphia, Pennsylvania Iraj Rezvani, MD Department of Pediatrics Temple University School of Medicine Section of Pediatric Endocrinology Temple University Children’s Medical Center Philadelphia, Pennsylvania Angelo M. DiGeorge, MD Department of Pediatrics Temple University School of Medicine Philadelphia, Pennsylvania 10.1016/j.jpeds.2007.05.013

REFERENCES Table. Thyroid antibodies at initial measurement Male Female White Black Hispanic Total Total No. 124 112 145 63 Ab positive 23 18 28 7 % positive 18.5% 16.1% 19.3% 11.1%

Letters

28 6 21.4%

236 41 17.4%

1. Radetti G. Gottardi E. Bona G. Corrias A. Salardi S. Loche S. Study Group for Thyroid Diseases of the Italian Society for Pediatric Endocrinology and Diabetes (SIEDP/ISPED). The natural history of euthyroid Hashimoto’s thyroiditis in children. J Pediatr 2006;149:827-32. 2. Franzese A. Buono P. Mascolo M. Leo AL. Valerio G. Thyroid autoimmunity starting during the course of type 1 diabetes denotes a subgroup of children with more severe diabetes. [Letter] Diabetes Care 2000;23:1201-2. 3. McKenna MJ. Herskowitz R. Wolfsdorf JI. Screening for thyroid disease in children with IDDM. Diabetes Care 1990;13:801-3.

e11

4. Riley WJ. Maclaren NK. Lezotte DC. Spillar RP. Rosenbloom AL. Thyroid autoimmunity in insulin-dependent diabetes mellitus: the case for routine screening. J Pediatr 1981;99:350-4. 5. Lipman TH. The epidemiology of type I diabetes in children 0-14 yr of age in Philadelphia. Diabetes Care 1993;16:922-5. 6. Lipman TH. Chang Y. Murphy KM. The epidemiology of type 1 diabetes in children in Philadelphia 1990-1994: Evidence of an epidemic. Diabetes Care 2002;25:1969-75. 7. Lipman TH. Jawad AF. Murphy KM. Tuttle A. Thompson RL. Ratcliffe SJ. Levitt Katz LE. Incidence of type 1 diabetes in Philadelphia is higher in black than white children from 1995 to 1999: Epidemic or misclassification? Diabetes Care 2006; 29:2391-5. 8. Chase HP. Garg SK. Cockerham RS. Wilcox WD. Walravens PA. Thyroid hormone replacement and growth of children with subclinical hypothyroidism and diabetes. Diabetic Med 1990;7:299-303. 9. Moore DC. Natural course of “subclinical” hypothyroidism in childhood and adolescence. Arch Pediatr Adolesc Med 1996;150:293-7.

Regional Hospital of Bolzano Bolzano, Italy Gianni Bona, MD Pediatric Clinic University of Novara Novara, Italy Andrea Corrias, MD Ospedale Regina Margherita Torino Torino, Italy Silvana Salardi, MD Pediatric Clinic University of Bologna Bologna, Italy

Reply To the Editor: We thank Dr. Lipman et al for their interest in our recently published article.1 In the first part of the letter, they wonder how the diagnosis has been made, in particular because goiter was not present in all children. Obviously the presence of goiter suggested the diagnosis of Hashimoto thyroiditis in most of the children, and, in the remaining, it has been made during routine evaluation of children with other known autoimmune disease, such as diabetes and vitiligo. The diagnosis was then confirmed by the typical ultrasound findings together with the presence of thyroid-peroxidase and thyroglobulin antibodies. The presence of at least one antibody was considered necessary, which explains the presence in Figures 1, 3, and 4 of children without one antibody. The authors also describe the presence of anti-thyroid antibodies and thyroid function in a group of diabetic children followed-up for about 5 years. In agreement with our findings, antithyroid antibodies serum levels, as well as thyroid function largely fluctuated over time with no negative effect on growth. Unfortunately, they never performed a thyroid ultrasound in their patients during follow-up, because, reportedly, the presence of anti-thyroid antibodies alone is not a sufficient criterion for the diagnosis of Hashimoto thyroiditis. Moreover, the prevalence of celiac disease in their diabetic patients is not known. It has been reported, in fact, that, by following a gluten-free diet, anti-thyroid antibodies may disappear.2 We are pleased that they do agree with our conclusion that treatment with thyroxine should not be started just when TSH levels are slightly elevated but rather that a watchful observation is recommended. Giorgio Radetti, MD Department of Paediatrics Regional Hospital of Bolzano Bolzano, Italy Elena Gottardi, MD Department of Paediatrics e12 Letters

Sandro Loche, MD Ospedale Regionale per le Microcitemie Cagliari Cagliari, Italy 10.1016/j.jpeds.2007.05.014

REFERENCES 1. Radetti G, Gottardi E, Bona G, Corrias A, Salardi S, Loche S, Study Group for Thyroid Diseases of the Italian Society for Pediatric Endocrinology and Diabetes (SIEDP/ISPED). The natural history of euthyroid Hashimoto’s thyroiditis in children. J Pediatr 2006;149:827-32. 2. Ventura A, Neri E, Ughi C, Leopaldi A, Citta A, Not T. Gluten-dependent diabetes-related and thyroid-related autoantibodies in patients with celiac disease. J Pediatr 2000;137:263-5.

Extreme obesity among children in Mexico To the Editor: Freedman et al,1 using data from the Bogalusa Heart Study, concluded that the 99th percentile-for-age may be appropriate for identifying children who are at very high risk for biochemical abnormalities and severe adult obesity. A previous study from Freedman et al,2 based on the NHANES 1999-2002, determined the prevalence of extreme obesity among 6- to 11-year-old children was 4% among whites and 5% among blacks and Mexican-Americans. However, we have not found a report of extreme obesity among Mexican children in Mexico. In response to this void, we conducted a survey of 2690 boys and girls between 6 to 12 years of age in Tijuana and Ensenada. The children’s weight, height, and waist circumference (WC) were measured according to standardized protocols. Body mass index (BMI) and WC values were compared by use of age/sex BMI percentiles from growth charts from the Centers for Disease Control and Prevention. The ⱖ99th percentile of BMI was used as the cutpoint for extreme obesity. The 90th percentile for WC was used as the cutpoint for abdominal obesity because children over that cutpoint were more likely to have multiple risk factors.3 An overall prevalence of extreme obesity was 5.2% (Tijuana 5.3%, Ensenada 5.0%), with 6.3% among boys and 4.1% in girls. The higher prevalence among boys is consistent with that observed among Mexican-American boys and girls.2 The Journal of Pediatrics • September 2007

Additionally, 54 (39%) children with BMI ⱖ99th percentile had a BMI ⱖ 30 kg/m2, 13 (9.3%) had a BMI ⱖ 35 kg/m2, 4 had a BMI ⱖ 40 kg/m2, and 127 (91%) had a WC ⱖ90th percentile. Because children at these levels of extreme obesity are at risk for biochemical abnormalities and severe adult obesity,1 intensive interventions are needed in Mexico to stop this epidemic with its current and potential for future burden to the health care system. Arturo Jimenez Cruz, MD, PhD Montserrat Bacardi-Gascon, MD, EdD Elizabeth Jones, RD, EdD Hospital Infantil de las Californias

Letters

Graduate Program of Nutrition Medical School Universidad Autonoma de Baja California Mexico 10.1016/j.jpeds.2007.05.017

REFERENCES 1. Freedman DS, Mei Z, Srinivasan SR, Berenson GS. Cardiovascular risk factors and excess adiposity overweight children and adolescents: The Bogalusa Heart Study. J Pediatr 2007;150:12-7. 2. Freedman DS, Khan LK, Serdula MK, Ogden CL, Dietz WH. Racial and ethnic differences in secular trends for childhood BMI, weight, and height. Obesity 2006;14:301-8. 3. Mafeiz C, Pietrobelli A, Grezzani A, Provera S, Tato L. Waist circumference and cardiovascular risk factors in prepubertal children. Obes Res 2001;9:179-87.

e13