Original Article . . . . .

. . . . . . . . .

Detection of Fungemia by Polymerase Chain Reaction in Critically Ill Neonates and Children Urmila H. Tirodker, MD James P. Nataro, MD, PhD Sherri Smith, MD Lauren LasCasas, MD Karen D. Fairchild, MD

OBJECTIVE: To compare polymerase chain reaction (PCR) with blood culture for the detection of fungemia in neonatal and pediatric intensive care unit patients.

STUDY DESIGN: A total of 0.2 to 0.5 ml of blood was collected simultaneously with blood culture in patients with suspected sepsis. Following DNA extraction, the gene for fungal 18S rRNA was amplified and PCR products analyzed by agarose gel electrophoresis. RESULTS: Of 70 patient samples, nine of nine with a positive blood culture for Candida were PCR positive, and one of four with a positive blood culture for Malassezia was PCR positive. In total, 13 of 57 samples with negative blood culture for fungus were PCR positive. Seven of these 13 patients had other evidence of invasive fungal disease. In contrast, none of the 44 blood culture-negative/PCR-negative patients had other evidence of fungal infection. CONCLUSIONS: PCR may be a useful adjunct to blood culture for the rapid detection of fungemia in high-risk patients. Journal of Perinatology (2003) 23, 117–122. doi:10.1038/sj.jp.7210868

Division of Pediatric Critical Care (U.H.T.), Department of Pediatrics, University of Maryland School of Medicine, Baltimore, USA; Division of Infectious Diseases (J.P.N.), Department of Pediatrics, University of Maryland School of Medicine, Baltimore, USA; Division of General Pediatrics (S.S., L.L), Department of Pediatrics, University of Maryland School of Medicine, Baltimore, USA; and Division of Neonatology (K.D.F.), Department of Pediatrics, University of Maryland School of Medicine, Baltimore, USA. Current affiliation of U.H.T.: Division of Critical Care, St. Christopher’s Hospital for Children, Philadelphia, PA, USA. This study was funded in part by a grant from the National Foundation for Infectious Diseases (K.F.). Address correspondence and reprint requests to Karen D. Fairchild, MD, Neonatology N5W68, University of Maryland School of Medicine, 22 S. Greene St., Baltimore, MD 21201, USA.

INTRODUCTION Disseminated Candida infection is an increasingly important cause of mortality and morbidity in neonatal intensive care unit (NICU) and pediatric intensive care unit (PICU) patients.1 Increased survival of very low-birth-weight infants and other immunocompromised patients, as well as increased use of indwelling intravascular catheters, broad-spectrum antibiotics, steroids, and parenteral alimentation have contributed to an increase in Candida sepsis over the last decade.2,3 A recent survey of very low birth-weight (less than 1500 g) infants revealed that fungal species, predominantly Candida albicans, were responsible for 9% of all cases of late-onset sepsis with 28% mortality.4 Another study documented a 10-fold increase in the incidence of invasive candidiasis among hospitalized neonates in the last 15 years.5 Since Candida has a propensity for deep tissue invasion, severe morbidities, including meningoencephalitis, endophthalmitis, osteomyelitis, and endocarditis, are not uncommon. Mortality from disseminated fungal infection is approximately 20 to 30% in NICU and PICU patients,6–8 but may increase to greater than 50% if treatment is delayed.2,9,10 Early detection of fungal sepsis and institution of appropriate therapy has been shown to improve patient outcome.2,10 In one study, patients with candidemia of more than 2 days’ duration prior to initiation of antifungal therapy had a two-fold higher mortality than those who were treated more promptly.11 Early removal of central venous catheters in neonates and children with candidemia is also associated with significantly decreased mortality and morbidity.12–13 Early and reliable detection of disseminated Candida infection is crucial yet problematic. In septic neonates, phlebotomy may be difficult and the small volume of blood obtained for culture may limit sensitivity. Blood culture detection of candidemia may take 48 hours or more, and some cases of systemic candidiasis are not diagnosed until autopsy.1,11,14 As a result of these impediments to detection of Candida sepsis and the high morbidity and mortality associated with delayed treatment, NICU and PICU patients are often treated empirically with antifungal agents, with attendant risks of toxicity and emergence of drug resistance. Detection of Candida DNA by polymerase chain reaction (PCR) may provide a rapid and reliable screen for candidemia in high-risk patients. PCR amplification of the highly conserved multicopy 18S ribosomal RNA (rRNA) gene present in multiple fungal species, but not in bacterial or human DNA, has been developed and tested in the clinical setting.15,17 In a study of

Journal of Perinatology 2003; 23:117–122 r 2003 Nature Publishing Group All rights reserved. 0743-8346/03 $25

www.nature.com/jp

117

Tirodker et al.

neutropenic cancer patients, Einsele reported that whole-blood PCR for the 18S rRNA gene was highly sensitive compared with culture for the detection of fungemia. Daily blood sampling revealed that in many cases fungal PCR was positive several days before Candida could be detected by blood culture.15,16 We therefore sought to compare 18S rRNA PCR with blood culture to assess its value in the detection of fungemia in NICU and PICU patients with suspected bloodstream infection.

MATERIALS AND METHODS Study Population and Blood Collection The study was carried out in a 40-bed level 3 NICU and 18-bed level 3 PICU from November 1999 through November 2000. Every patient in the NICU and PICU with suspected sepsis was eligible for inclusion. When a patient underwent blood culture for suspected infection, if excess blood remained in the sterile catheter or syringe after obtaining necessary bloodwork, it was collected in an EDTAcontaining tube and frozen at
201C. Volume of blood collected for PCR ranged from 0.2 to 0.5 ml. The Hospital Institutional Review Board determined the study to be exempt from the consent requirement because of the small volume of waste blood collected. Each patient was classified as low, medium, or high index of suspicion for septicemia based on the judgment of the most senior clinician (generally the attending or fellow) caring for the patient at the time the blood culture was obtained. Criteria for determining index of suspicion included, but were not limited to, increased apneic episodes, temperature instability, hemodynamic compromise, significant increase in ventilatory support, and results of laboratory tests including complete blood count and differential and serum C-reactive protein. Results of fungal PCR were not made available to the clinicians caring for the patients. Blood Culture Blood was inoculated for culture into Pediatric Plus bottles (Becton Dickinson Diagnostic Instrument Systems, Sparks, MD) that were monitored on a BACTEC 9240 automated system for 5 days unless a specific request was made to hold the culture longer. Positive fungal blood cultures were examined microscopically, subcultured on blood, chocolate, and Sabouraud’s agar, and inoculated on a Vitek yeast card for species identification according to standard methods. Sample DNA Extraction DNA extraction and PCR were performed using a modification of the methods described by Einsele.15–17 Patient and healthy control blood samples (0.2 to 0.5 ml) were treated with red cell lysis buffer (10 mm Tris [pH 7.6], 5 mm MgCl2, 10 mm NaCl) for 10 min at 371C and centrifuged for 10 min at 5000  g. White cell lysis buffer (10 mm Tris [pH 7.6], 10 mm EDTA, 50 mm NaCl, 1% SDS, and 150 mg/ml proteinase K [Boehringer Mannheim, Indianapolis, IN]) was added to the pellet and the samples were incubated at 118

PCR Detection of Fungemia in ICU Patients

651C for 1 hour. Following centrifugation, fungal cell walls were lysed by suspending the pellet in fungal lysis buffer (50 mm Tris [pH 7.6], 1 mm EDTA, 28 mm b-mercaptoethanol, and 3 units lyticase [Sigma, St. Louis, MO]). After incubation at 371C for 1 hour, 10% SDS was added and incubation continued at 651C for 20 minutes. Protein precipitation with potassium acetate was followed by DNA extraction with isopropanol. DNA was stored at
801C and assayed in batch by PCR. In order to verify successful DNA extraction from the small blood volumes obtained for the study, an initial PCR was performed on all samples for the constitutively expressed GAPDH gene. Candida Culture and Determination of in vitro PCR Sensitivity A single colony of C. albicans (adult patient isolate) was swirled in 1 ml of thioglycolate broth and agitated at 351C for 48 hours. The inoculum was centrifuged for 5 minutes at 12,000  g and the pellet washed twice with sterile normal saline. The fungal saline suspension was adjusted photometrically to a concentration of 107 CFU/ml .The suspension was then serially diluted 10-fold (106 to 100 CFU/ml ) and plated on Sabouraud-dextrose medium for confirmation of colony count. Serial dilutions of C. albicans were added to 1 ml of whole blood from a healthy volunteer and subjected to sample DNA extraction as described above, followed by PCR amplification. Candida DNA Extraction The washed pellet of C. albicans prepared as described above was resuspended in 500 ml of lysis buffer (50 mm Tris [pH 7.5], 10 mm EDTA, 0.5% b-mercaptoethanol, and 0.3 mg/ml Zymolase [ICN, Costa Mesa, CA]) and incubated at 371C for 1 hour. Proteinase K 50 mg/ml in 1% SDS was added and the mixture incubated for 1 hour at 561C, then heated to 951C for 5 minutes. DNA was extracted with phenol–chloroform–isoamyl alcohol (25:24:1) and precipitated with isopropanol.18 DNA quantitation was performed by spectrophotometric measurement of optical density at 260 nm. Serial dilutions of purified C. albicans DNA (4.5 ng to 0.45 fg) in sterile normal saline were prepared and subjected to PCR. PCR Extracted DNA (5 ml) was amplified in a 50-ml reaction mixture with 1  PCR buffer (10 mm Tris [pH 8.8], 1.5 mm MgCl2, 25 mm KCl, and 0.5 units Perfect Match Polymerase Enhancer [Stratagene, La Jolla, CA]), 2.5 U of AmpliTaq polymerase (Perkin Elmer Corporation, Norwalk, CT), 100 mm of deoxynucleotide triphosphates, and 30 pmol of each primer (forward primer 50 -ATT GGA GGG CAA GTC TGG TG-30 , reverse primer 50 -CCG ATC CCT AGT CGG CAT AG-30 ). These primers amplify a 482 to 503 base pair (bp) fragment of the 18S rRNA gene, depending on the fungal species.15 The amplification was performed in a Perkin-Elmer 480 Thermocycler and consisted of a hot start of 951C for 5 minutes Journal of Perinatology 2003; 23:117–122

PCR Detection of Fungemia in ICU Patients

Tirodker et al.

Study Sample PCR and Blood Culture Results Of 70 samples from 63 patients, nine blood cultures from five patients (two PICU, three NICU) yielded Candida species. Of these five patients, four were classified as high index of suspicion for septicemia and one was classified as moderate index of suspicion. One PICU patient had four samples submitted over an 11-day period (none on the same day), all of which were positive by 18S rRNA PCR and blood culture (C. albicans). Another PICU patient had two samples submitted on separate days, both of which were PCR positive and culture positive (mixed C. albicans and C. krusei). Three NICU patients had blood cultures positive for C. albicans, and in each case fungal PCR was also positive. In summary, all nine Candida culture-positive samples were positive by 18S rRNA PCR. Four positive blood cultures from four NICU patients yielded Malassezia furfur, and of these, only one was positive by 18S rRNA PCR. This PCR-positive patient was classified as high index of suspicion for septicemia. In contrast, the three patients whose blood cultures yielded M. furfur, but whose fungal PCR was negative, had a low index of suspicion for sepsis. Of the 57 blood samples negative for fungal growth by culture, 13 were PCR positive for 18S rRNA. Of the 13 PCR-positive/blood culture-negative patients, seven had other evidence of either concurrent or recent invasive fungal infection (Table 1). Four patients had positive blood cultures for Candida from 3 to 28 days prior to study sample collection, and three of these patients had been or were currently being treated with amphotericin. Two neonates had C. albicans isolated from urine obtained by suprapubic bladder aspiration. Both of these patients were classified as high index of suspicion for sepsis and had no other positive cultures. One patient had necrotizing enterocolitis with intestinal perforation and Candida peritonitis at the time of study sample collection. The remaining six patients with positive PCR and negative culture had no other evidence of fungal infection. Of the 44 samples that were both fungal PCR negative and blood culture

followed by 35 cycles of 941C for 30 seconds, 621C for 1 minute, and 721C for 2 minutes, and a final extension at 721C for 5 minutes. The PCR products were analyzed on ethidium bromidestained 2% agarose gels. Gels were analyzed for the presence of fungal PCR product by two independent observers blinded to blood culture results and patient information. The entire protocol, from sample preparation through gel interpretation, could be completed in 5 hours. To minimize the risk of contamination, sample preparation and PCR analysis were carried out in separate rooms. Sterile gloves and barrier tips were used for all steps in the protocol. Labware, solutions, and workspaces were autoclaved, UV irradiated, and/or HCl-washed, as appropriate. To monitor for contamination, negative controls (sterile water and whole blood from a healthy volunteer) were used in each extraction and amplification. Purified C. albicans DNA served as a positive control.

RESULTS Patient Characteristics Over a 12-month period, DNA was extracted from 70 blood samples collected from 63 patients (46 NICU, 17 PICU). Of these samples, 53 (including all the NICU samples) were from a peripheral blood draw, 13 from a central venous line, and four from an arterial line. Index of suspicion for septicemia was high in 29 cases (41%), medium in 17 (25%), and low in 24 (34%). A total of 47 samples (67%) were from patients on antibiotics and seven samples (10%) were from patients on antifungal agents at the time of collection. Sensitivity of 18S rRNA PCR for Detection of C. albicans The lower limit of PCR detection of purified C. albicans DNA was 45 pg, corresponding to approximately 100 organisms per sample (Figure 1). The sensitivity of PCR for detecting C. albicans organisms added to whole blood was 100 CFU/0.5 ml.

1

2

3

4

5

6

7

8

9

10

11

12

489 bp

Figure 1. Sensitivity of 18S rRNA PCR for detection of Candida albicans. Serial dilutions of DNA extracted from C. albicans organisms added to 0.5 ml whole blood from a healthy adult (lanes 2 to 7) or purified C. albicans DNA (lanes 8 to 12) were subjected to PCR for the 18S rRNA gene present in multiple species of fungi. The 489 bp C. albicans PCR products were analyzed by electrophoresis on an ethidium bromide-stained 2% agarose gel. Lane 1: Molecular weight marker. Lane 2: Whole blood negative control. Lanes 3 to 7: Whole blood with 101, 102, 103, 104, 106 CFU of C. albicans. Lanes 8 to 12: Serial dilutions of purified C. albicans DNA (4.5, 0.45 ng, 45, 4.5, 0.45 pg). Limits of sensitivity were 102 organisms added to whole blood and 45 pg of purified C. albicans DNA. This gel is representative of three experiments with similar results. Journal of Perinatology 2003; 23:117–122

119

Tirodker et al.

PCR Detection of Fungemia in ICU Patients

Table 1 Characteristics of Patients with Positive PCR and Negative Blood Culture for Fungus Index of suspicion*

Clinical factors

Antifungal therapy

Relevant cultures

1

High

NEC

Blood: C. albicans 28 days prior to sample collection

2 3

High High

Completed amphotericin course On amphotericin

4 5 6 7

High High High High

8 9 10

High High High

11 12 13

Moderate Moderate Low

NEC, intestinal perforation

On amphotericin

NEC, intestinal perforation NEC ARDS, endocarditis Wrist abscess

Blood: C. albicans 3, 5 and 8 days prior to sample collection. Blood: Enterococcus faecalis Peritoneal fluid: Klebsiella Blood: C. glabrata 28 days prior to sample collection Blood: C. albicans 14 days prior to sample collection. Urine: C. albicans Urine: C. albicans 4 days after study sample collection Peritoneal fluid: Heavy C. albicans 4 days after study sample collection Blood: Klebsiella sp. Blood: Enterobacter cloacae Blood: Pseudomonas aeruginosa Sputum: heavy C. albicans, many PMNs Abscess fluid: Pseudomonas aeruginosa Blood: Staphylococcus epidermidis None

*Index of suspicion for septicemia at the time of sample collection. NEC=necrotizing enterocolitis.

negative, no patient had other culture evidence of invasive fungal disease.

DISCUSSION In hospitalized neonates and children, Candida is a leading cause of nosocomial sepsis, with significant mortality and morbidity.1,3,5 Early detection and aggressive treatment has been shown to decrease mortality, yet blood culture is less than optimal as a single test for rapid and sensitive detection of invasive fungal disease. In this pilot study of NICU and PICU patients with suspected sepsis, we have shown that 18S rRNA PCR rapidly detected fungal DNA in all whole-blood samples from which Candida was isolated by culture. PCR was also negative in all blood cultures that did not yield Candida species. Further evidence suggested that PCR might detect low-level candidemia that is not detected by conventional blood culture methods. This is the largest clinical study to date using PCR to detect fungemia in critically ill neonates and children. In blood volumes as small as 0.2 ml, 18S rRNA PCR detected fungal DNA in all nine samples with blood culture-proven candidemia. This finding is similar to that of the only other published neonatal/pediatric study, in which Jordan19 performed PCR for the chitin synthase gene to detect Candida DNA in 26 of 27 paired blood samples from 16 patients with culture-proven candidemia. In contrast to the high 120

sensitivity of our assay for the detection of candidemia, sensitivity for detecting blood culture-proven M. furfur was only 25%. Although M. furfur DNA can be amplified with the 18S rRNA primers used in this study,15,17 clinical studies have not demonstrated the ability of PCR to detect septicemia with this relatively low-virulence yeast. Interesting, the single patient in our study in whom PCR was positive and blood culture grew M. furfur was classified as high index of suspicion for septicemia, whereas the three patients in whom culture was positive and PCR negative were classified as low index suspicion for sepsis. This leads us to speculate that our DNA extraction and PCR protocol may not be optimal for the detection of low colony count M. furfur in the blood. In vitro, the lower limit of sensitivity of our 18S rRNA PCR was 45 pg of purified C. albicans DNA, which represents approximately 100 Candida organisms. The lower limit of PCR detection of C. albicans added to whole blood was also 100 CFU/0.5 ml. These results are comparable to other similar studies,19–22 although higher sensitivity has been reported using larger volumes of blood,17 nested PCR,18,23 and DNA hybridization of PCR products.15–17 Animal studies have shown PCR to have equal or greater sensitivity than blood culture for in vivo detection of candidemia. A study in a neutropenic mouse model demonstrated that whole-blood 18S rRNA PCR followed by DNA hybridization was more sensitive than blood culture for the detection of experimental C. albicans bloodstream infection (100 versus 67% sensitivity).24 Journal of Perinatology 2003; 23:117–122

PCR Detection of Fungemia in ICU Patients

In the same study, PCR did not detect Candida DNA in the blood of animals that were colonized but not infected with Candida. Similarly, nested PCR has been reported to be more sensitive than blood culture for the detection of candidemia in an immunocompetent rabbit model.23,25 Our study consisted of a convenience sample of patients undergoing blood culture in whom extra blood was available for PCR testing. The prevalence of culture-proven candidemia was higher in our study population than in the overall NICU and PICU patient populations during the same time period, suggesting that we may have inadvertently selected a subgroup of patients with a higher likelihood of fungal bloodstream infection. Inclusion of a larger number of lower risk patients would likely increase the number of false-positive PCR results. However, in patients at high risk for fungal sepsis, such as very low-birth-weight infants or neutropenic or otherwise immunocompromised patients, PCR may aid in deciding which patients with symptoms of sepsis should receive antifungal therapy before blood culture results are available. In addition, PCR testing might reduce unnecessary use of empiric antifungal therapy, since the negative predictive accuracy of PCR for detecting culture-proven candidemia in this study was 100%. In this study, 13 patients had a positive fungal PCR result with a negative blood culture, giving a positive predictive value of PCR for detecting culture-proven fungemia of 43%. However, it has been estimated that a single blood culture, even by the lysis centrifugation method, fails to detect 25 to 50% of disseminated candidiasis.26 This leads us to speculate that some of our ‘‘falsepositive PCR’’ results may actually represent false-negative blood cultures in patients with disseminated fungal infection. Among the PCR-positive/culture-negative patients were two neonates with candiduria, strong clinical indicators of septicemia, and no other positive cultures to account for their symptomatology. Another neonate had necrotizing enterocolitis, intestinal perforation, and C. albicans peritonitis. Given that presence of Candida in the urinary and gastrointestinal tracts are strong predictors of systemic fungal infection,27,28 these patients may have had Candida bloodstream infection that was not detected by a single blood culture. In addition to detecting low colony count infection, PCR may also detect nonviable organisms, or free Candida DNA following clearance of the pathogen. Four of our PCR-positive/ culture-negative patients had previous culture-proven candidemia from 3 to 28 days prior to sample collection, and 3 were being treated with amphotericin B. Other investigators have also found persistence of positive fungal PCR in patients on appropriate antifungal therapy and have shown this to be associated with higher mortality.15 However, whether these cases represent nonviable organisms, free Candida DNA in the blood, or persistence of low-level infection is unknown. In summary, despite a high sensitivity and negative predictive value of PCR for the detection of candidemia in this pilot study, interpretation of positive PCR results in patients with negative Journal of Perinatology 2003; 23:117–122

Tirodker et al.

blood cultures remains a challenge. Large studies including septic and nonseptic patients with and without mucocutaneous fungal infection or colonization are necessary to elucidate whether a positive 18S rRNA PCR result represents disseminated fungal infection, transient fungemia that will clear without therapy, nonviable organisms, or sample contamination. Use of quantitative PCR or a combination of tests, such as PCR and a beta-glucan assay, are also under investigation for rapid screening of high-risk patients. Ultimately, these assays may facilitate early institution of antifungal therapy in patients with disseminated candidiasis as well as avoidance of empiric and potentially toxic therapy for patients without fungal infection. Acknowledgements This work was supported in part by a grant from the National Foundation for Infectious Diseases (K.F.). We also thank Penny Bamford for providing expert technical advice and Patrick Murray for helpful discussions.

References 1. Butler KM, Baker CJ. Candida: An increasingly important pathogen in the nursery. Pediatr Clin North Am 1988;35:543–63. 2. Stamos JK, Rowley AH. Candidemia in a pediatric population. Clin Infect Dis 1995;20:571–5. 3. Saiman L, Ludington E, Pfaller M, et al. Risk factors for candidemia in neonatal intensive care unit patients. The National Epidemiology of Mycosis Survey Study Group. Pediatr Infect Dis J 2000;19:319–24. 4. Stoll BJ, Gordon T, Korones SB, 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. Kossoff EH, Buescher ES, Karlowicz MG. Candidemia in a neonatal intensive care unit: Trends during fifteen years and clinical features of 111 cases. Pediatr Infect Dis J 1998;17:504–8. 6. Faix RG. Invasive neonatal candidiasis: Comparison of albicans and parapsilosis infection. Pediatr Infect Dis J 1992;11:88–93. 7. Benjamin DK, Ross K, McKinney RE, Auten R, Fisher RG. When to suspect fungal infection in neonates: A clinical comparison of Candida albicans and Candida parapsilosis fungemia with coagulase-negative staphylococcal bacteremia. Pediatrics 2000;106:712–8. 8. Yamamura DL, Rotstein C, Nicolle LE, Ioannou S. Candidemia at selected Canadian sites: Results from the Fungal Disease Registry, 1992–1994. Fungal Disease Registry of the Canadian Infectious Disease Society. Can Med J 1999;160:493–9. 9. Johnson DE, Thompson TR, Green TP, Ferrieri P. Systemic candidiasis in very low-birth-weight infants (less than 1,500 grams). Pediatrics 1984;73:138–43. 10. Pacheco-Rios A, Avila-Figueroa C, Nobigrot-Kleinman D, Santos JI. Mortality associated with systemic candidiasis in children. Arch Med Res 1997;28:229–32. 11. Fraser VJ, Jones M, Dunkel J, et al. Candidemia in a tertiary care hospital: epidemiology, risk factors, and predictors of mortality. Clin Infect Dis 1992;15:414–21. 12. Dato VM, Dajani AS. Candidemia in children with central venous catheters: Role of catheter removal and amphotericin B therapy. Pediatr Infect Dis J 1990;9:309–14. 121

Tirodker et al.

13.

14.

15.

16.

17.

18.

19. 20.

122

Eppes SC, Troutman JL, Gutman LT. Outcome of treatment of candidemia in children whose central catheters were removed or retained. Pediatr Infect Dis J 1989;8:99–104. Baley JE, Kliegman RM, Fanaroff AA. Disseminated fungal infections in very low birth weight infants: clinical manifestations and epidemiology. Pediatrics 1984;73:144–52. Einsele H, Hebart H, Roller G, et al. Detection and identification of fungal pathogens in blood by using molecular probes. J Clin Microbiol 1997;35:1353–60. Hebart H, Loffler J, Reitze H, et al. Prospective screening by a panfungal polymerase chain reaction assay in patients at risk for fungal infections: implications for the management of febrile neutropenia. Br J Haematol 2000;111:635–40. Van Burik JA, Myerson D, Schreckhise RW, Bowden RA. Panfungal PCR assay for detection of fungal infection in human blood specimens. J Clin Microbiol 1998;36:1169–75. Chryssanthou E, Andersson B, Petrini B, Lofdahl S, Tollemar J. Detection of Candida albicans DNA in serum by polymerase chain reaction. Scand J Infect Dis 1994;26:479–85. Jordan JA. PCR identification of four medically important Candida species by using a single primer pair. J Clin Microbiol 1994;32:2962–7. Buchman TG, Rossier M, Merz WG, Charache P. Detection of surgical pathogens by in vitro DNA amplification. Part I. Rapid identification of Candida albicans by in vitro amplification of a fungus-specific gene. Surgery 1990;108:338–46.

PCR Detection of Fungemia in ICU Patients

21. Holmes AR, Cannon RD, Shepherd MG, Jenkinson HF. Detection of Candida albicans and other yeasts in blood by PCR. J Clin Microbiol 1994;32:228–31. 22. Miyakawa Y, Mabuchi T, Fukazawa Y. New method for detection of Candida albicans in human blood by polymerase chain reaction. J Clin Microbiol 1993;31:3344–47. 23. Bougnoux M, Dupont C, Mateo J, et al. Serum is more suitable than whole blood for diagnosis of systemic candidiasis by nested PCR. J Clin Microbiol 1999;37:925–30. 24. van Deventer AJM, Goessens WHF, van Belkum A, van Vliet HJA, van Etten EWM, Verbrugh HA. Improved detection of Candida albicans by PCR in blood of neutropenic mice with systemic candidiasis. J Clin Microbiol 1995;33:625–8. 25. Polanco A, Mellado E, Castilla C, Rodriguez-Tudela JL. Detection of Candida albicans in blood by PCR in a rabbit animal model of disseminated candidiasis. Diagn Microbiol Infect Dis 1999;34:177–83. 26. Berenguer J, Buck M, Witebsky F, Stock F, Pizzo PA, Walsh TJ. Lysiscentrifugation blood cultures in the detection of tissue-proven invasive candidiasis. Disseminated versus single-organ infection. Diagn Microbiol Infect Dis 1993;17:103–9. 27. Phillips JR, Karlowicz MG. Prevalence of Candida species in hospitalacquired urinary tract infections in a neonatal intensive care unit. Pediatr Infect Dis J 1997;16:190–4. 28. Calandra T, Bille J, Schneider R, Mosimann F, Francioli P. Clinical significance of Candida isolated from peritoneum in surgical patients. Lancet 1989;2:1437–40.

Journal of Perinatology 2003; 23:117–122