Original Article

Celiac Disease in Osteoporotic Indians YA Gokhale*, PD Sawant**, CM Chodankar***, ND Desai+, MV Patil***, S Maroli++, MN Patil+++, NK Hase#

Abstract Objective : 1) The aim of the study was to identify the atypical celiac disease (CD) in a cohort of symptomatic osteoporotic patients, younger than 55 years of age and 2) To study associated clinical and laboratory features and outcome with gluten-free diet. Material and Methods : We studied 33 patients (F:M = 28:5), mean age 29 years (range 15-52 years) with osteoporosis (WHO diagnostic criteria, T-score less than -2.5 on DEXA scan) from January 2000 - June 2002. Serological screening for CD was done by detecting circulating IgA antibodies to tissue transglutaminase by ELISA. Patients with presence of antibodies to transglutaminase were subjected to biopsy from the 2nd part of the duodenum by upper GI endoscopy. The biopsies were reported independently by two pathologists who were blinded for the serology report. Measurement of mucosal thickness, crypts and villi were done with an ocular micrometer. Other parameters like complete hemogram, serum iron, total iron binding capacity (TIBC), calcium profile, 25-OH-D, parathyroid hormone (PTH) were evaluated. Assessment of clinical and laboratory parameters was performed within 4-12 weeks of starting gluten-free diet (GFD). Results : Thirteen patients had circulating IgA antibodies to transglutaminase. Intestinal biopsies were performed on 11 patients and were consistent with the diagnosis of CD (total villous atrophy - two, subtotal villous atrophy with crypt hyperplasia - nine). Patients with CD had significant anaemia when compared with non-CD osteoporotic patients. Other important observations in these 11 patients were low serum calcium and phosphorus, low 25-OH-D, high PTH. Significant improvement in clinical and laboratory parameters was noted in all patients within 6-12 weeks of starting GFD. Conclusion : Symptomatic osteoporotic patients (younger than 55 years of age) especially with associated anaemia should be investigated for CD. Simple measures like omission of wheat from diet (GFD) lead to significant improvement in symptoms within weeks.

INTRODUCTION

C

eliac disease (CD), gluten-sensitive enteropathy is a disease that results from permanent intolerance to wheat protein, gluten. Classically the patient manifests with diarrhoea. Such classical symptomatic patients represent the tip of the icerberg.1 Majority of patients of CD who are atypical (i.e. do not have GI symptoms) manifest as anaemia, osteoporosis, convulsions, short stature, dental enamel defects. *Associate Professor of Medicine; #Professor of Medicine; *** Asso. Professor in Pathology; ** Professor and Head of Gastroenterology; + Lecturer in Gastroenterology; ++ Associate Professor of Pharmacology; +++Technician in Medicine, Department of Medicine, Gastroenterology, Pathology and Pharmacology, Lokmanya Tilak Medical College and General Hospital, Sion, Mumbai. Received : 10.8.2002; Revised : 5.12.2002; Accepted : 10.4.2003

JAPI • VOL. 51 • JUNE 2003

Diagnosis of CD by 1970 European 2 criteria was cumbersome and entailed three jejunal mucosal biopsies, showing an initial flat mucosa, restoration of the mucosa to normal on gluten-free diet, and a deterioration of the mucosa after gluten challenge. With the availability of highly sensitive and specific serological tests [anti-endomysial antibodies (anti-EMA) and anti-tissue transglutaminase (anti-tTG)]3 screening for CD has become easy. Patients with presence of such antibodies can be selectively subjected to jejunal mucosal biopsies. In 1990, the European Society of Paediatric Gastroenterology, Hepatology and Nutrition, proposed a revised celiac sprue diagnostic protocol.3-5 Based on these criteria, if the symptoms (either typical or atypical) and serological results are suggestive, a single small intestinal biopsy followed by a favorable clinical and serological response to gluten-free diet (GFD) is now considered sufficient to definitely confirm the diagnosis provided other 579

causes of villous atrophy (e.g. milk-protein intolerance and gastrointestinal) are excluded. With the availability of serological tests, incidence of CD in some countries is reported to be 1-5%.6,7 Much work on CD is not done in India, especially in the adult population. Asymptomatic CD (i.e. without GI symptoms) is more common presentation of CD in adults than in children. Chronic malabsorption of Calcium and vitamin D leads to metabolic bone disease in these patients. We screened osteoporotic patients (WHO definition of osteoporosis) for CD.

MATERIAL AND METHODS Thirty-three consecutive symptomatic osteoporotic patients from one medical OPD and ward between January 2000 - June 2002 were included in the study. Inclusion Criteria : 1. Osteoporosis (WHO definition, Tscore 10 years; 3. Patients on steroids; 4. Diarrhoea at the time of presentation to the hospital. Methods Detail medical history, drug history, examination, complete blood count, erythrocyte sedimentation rate, serum iron, total iron binding capacity, calcium profile, liver and renal biochemistry, 25-OH-D, i PTH were performed. Other investigations like arterial blood gases and electrolytes, thyroid function tests, gonadal and gonadotropic hormones were performed wherever indicated. Sera of all patients were tested for presence of IgA tissue transglutaminase (tTG) antibody by ELISA using commercially available kits from Genesis Diagnostics, UK. Antibody titer > 10 units were considered positive. Sera of 21 asymptomatic medical students and doctors were also tested as controls. Three to four intestinal mucosal biopsies were obtained with GI endoscopy form the 2nd part of the duodenum in patients with presence of tTG antibodies. Longitudinally oriented biopsies, with muscularis mucosa seen in the section were only included. Biopsies were reported independently by two pathologists, who were blinded for the tTG antibody report. Measurements of mucosal thickness, villi and crypts were done by ocular micrometry as follows. A graticule were interposed in the 10X eyepiece. The number of squares of the graticule which could be accommodated in one small square of RBC chamber of Neubauer’s chamber (i.e. 1000 µm /5 = 200 µm) was counted under 10X objective. In our case four squares from ocular graticule could be accommodated. Thus each square from graticule measured 50 µm (when 10X objective and 10X eye piece were used). We used following measurements : 1. mucosal thickness (M) - from tip of villus to muscularis mucosa; 2. crypt dept (C) - from dilatation at the beginning of a villous to muscularis 580

mucosa, average of three crypts were taken; 3. villous height (V) = M - C; 4. villi/crypt ratio (V/C); and 5. villi to mucosal thickness (V/M). Biopsies with MARSH III lesion i.e. total or subtotal villous atrophy in addition to crypt hyperplasia and intraepithelial and lamina propria infiltration with mucosal cell, were considered to be consistent with the diagnosis of CD. All patients were started on supplements of hematinics, Calcium, vitamin D, alendronate while investigations for CD were being performed (i.e. for average three months). Patients diagnosed as CD on serology and biopsy (Marsh III lesion) were started on Gluten-Free Diet (GFD) and followed up at two weekly interval for three months and thereafter for nine months to three years. Assessment of clinical and laboratory parameters were performed on follow up. Repeat serology on GFD was performed on eight patients and a repeat DEXA scan at the end of one year in five patients. The ethical committee of our hospital approved the protocol and informed consent was obtained from all patients.

RESULTS Out of 33 (28 F : 5 M) osteoporotic patients, aged 15-52 years (mean age 29 years) who were screened for IgA tTG antibodies, 13 (12 F : 1 M) were detected to have circulating IgA antibodies to tTG. None of the 21 medical student and doctors had the circulating antibody. Out of 13 patients with circulating tTG antibodies, 11 underwent duodenal mucosal biopsies. Two patients had total villous atrophy (Fig. 1), nine had subtotal villous atrophy (Fig. 2). There was crypt hyperplasia in all biopsies (> 150 micrometer crypt depth), intra-epithelial lymphocytes and lamina propria infiltration was present in all biopsies (Fig. 3). Brief clinical details, including presenting complaints of 11 biopsy proven cases are depicted in Table 1. Average height and weight was low 140 cm and 40 kg, respectively. Two patients gave history of diarrhoea in the past three years and five other on direct inquiry gave history of vague abdominal pain, flatulence and bloating, all of which improved within 3-4 weeks of starting GFD. Significant improvement in presenting symptoms was noted in all patients within 8-12 weeks of starting GFD, whereas there was no improvement on supplements of Calcium, vitamin D, haematinics, and alendronate while reports of investigations for CD were awaited and the patients were consuming wheat (duration of supplements was ≥ 3 months). Weight gain, complete abolition of bone pain and improvement in laboratory parameters was noted in all patients. Average hemoglobin was 8.5 gm% (range 2.2-11.8 gm%). Patients with CD had significant anaemia (p < 0.05) when compared with non-CD osteoporotic patients. Other laboratory parameters were as depicted in Table 2. Most patients had osteomalacia with osteoporosis and secondary hyperparathyroidism. The means and ranges of the measurements and indices from the 11 biopsies were as shown in Table 3. JAPI • VOL. 51 • JUNE 2003

At follow up abdominal symptoms were first to improve i.e. within 3-4 weeks of starting GFD whereas bone pain and fatigue improved at 8-12 weeks completely in 10 patients and partially in one patient. Repeat serology was available in eight patients after 7-12 months of GFD. Six patients became IgA tTG negative. In two patients the antibody still persisted - in one the titers decreased and in the other patient titers remained the same (she was not strictly adherent to GFD). Result of DEXA scan in patients of CD were as follows: T-score and mean BMD at the hip were 5.77 to -2.5 and 0.564 gm/cm2 respectively and at the lumbar spine T-score -5.15 to -2.5 and mean BMD 0.496 gm/cm2. Repeat DEXA at 1 year of GFD was available in five patients and showed improvement in the mean BMD at the hip and the lumbar spine to 0.739 gm/cm2 and 0.936 gm/cm2 respectively.

DISCUSSION Despite the central importance of wheat as a dietary staple throughout the world, it is astounding that its presumptive role in precipitating celiac disease was discovered only in 1939 by the Dutch Paediatrician WK Dick.8 Later in 1954 John Paulley provided the first authentic description of villous atrophy in CD. Antigliadin antibodies were discovered in 1970.9 IgA antiendomysial antibody assay by immunofluorescence, using monkey oesophagus as substrate became available in 1983.9 Tissue transglutaminase has been reported to be the target for endomysial antibodies in CD.10 IgA class tTG can be detected by ELISA and are highly sensitive (95%) and specific (96%).11 CD results from an inappropriate T-cell mediated Fig. 1 : Flattened villi in the duodenal mucosal biopsy of a patient of CD immune response against ingested gluten in (H and E x 40). genetically predisposed people. The enzyme tTG is one of the targets of the autoimmune response in CD. The modification of gliadin by host tTG has a key role in enhancing the gliadin specific T cell response.5 Marsh12 described the spectrum of mucosal changes in response to gluten as preinfiltrative (type 0) lesion, the infiltrative (type 1) lesion - present in 40% untreated CD patients, hyperplastic (type 2) lesion - seen in 20% untreated CD patients, the destructive (type 3) lesion characterized by villous atrophy is seen in 40% patients and hypoplastic (type 4) lesion. What is certain is that the majority of individuals irrespective of type of lesion are asymptomatic. 12 The Fig. 2 : Subtotal villous atrophy in duodenal mucosa in a patient of CD (H and E x 40). JAPI • VOL. 51 • JUNE 2003

581

Table 1 : Clinical features of 11 patients of CD Sr. No.

Age /Sex

Presenting complaints**

1. 2. 3. 4. 5. 6.

25/F 18/M 15/F 32/F 49/F 19/F

7. 8.

50/F 42/F

9. 10.

18/F 52/F

11.

35/F

Bed ridden due to aches and pains Backache, fatigue, weight loss Growth retardation, aches and pains Dorsal backache Hip pain Weakness due to severe anaemia Signs of childhood rickets Dorsal backache Weakness due to severe anaemia Multiple vertebral fractures Crippled due to bodyache Backache, bodyache, carpopedal spasm Dorsal backache

Duration (months)

Past H/O hospitalization for diarrhoea

H/O vague GI symptoms - Abdominal pain, bloating, gases*

30 18 30 18 3-4 12

— yes

— Abdominal pain

— — —

Abdominal pain Abdominal pain, bloating —

12 24

— Yes

— Abdominal pain, bloating

6-8 12

— —

Abdominal pain Bloating, abdominal pain

12

—

Flatulence

*These symptoms completely revolved with in 3-4 week of starting GFD. **These symptoms revolved within 8-12 week of starting GFD.

Table 2 : Laboratory parameters of 11 patients of CD Mean Hb S Ca SP S Proteins S ALKP 25-OH-D i PTH S iron TIBC

Range

8.5* 2.2-11.8 8 (N 9-11 mg%) 5.6-9.5 3.3 (N 3-4.5 mg%) 1.4-4.5 7.1 (N 6-8 gm%) 4.8-8.9 500 (N 108-306 IU) 3.6 (N 9.9-45 ng/ml) 400 (N 12-72 pg/ml) 40 (N 60-400 mg%) 400 (N 250-400 mg%)

7 6 2 7

low, 4 normal low, 5 normal low high, 4 low

*Hb in patients with CD was significantly lower (P < 0.05) when compared with the Non-CD group.

Table 3 : Measurements and indices of biopsies

Mucosal thickness (M) Crypt depth (C) Villous height (V) V/C V/M

Fig. 3 : Duodenal mucosal biopsy in a patient of CD, with blunted villous. Note intra- epithelial lymphocytes (IEL) and lamina propria lymphocytes (H and E x 400). 582

Average

Range

502 µ

(250-800)

324 µ 178 µ 0.55 0.38

(250-550) (140-283) (0.52-0.8) (0.35-0.44)

disease remains not only clinically latent but there is mucosal latency too12 and half of the newly diagnosed CD patients do not have GI symptoms at the onset. It takes few years (upto 3 years) to develop type 3 lesions. Mucosal involvement can be even patchy.13 In the normal intestinal mucosa the villi are three times the length of the crypts13,14 and project from a simple gland layer that lies on a muscularis mucosa. Whereas in patients with CD with crypt hyperplasia (Marsh II) the crypt depth is increased, > 150 µm15 and Marsh III lesion consists of obvious villous shortening i.e. partial or subtotal villous atrophy with crypt hyperplasia.15 All our patients had Marsh III lesions (total or subtotal JAPI • VOL. 51 • JUNE 2003

villous atrophy).

REFERENCES

Adult CD is more likely than childhood CD to present without GI symptoms, with approximately 50% of adults having no history of diarrhoea.3 In many adults who present with GI symptoms the diarrhoea may be episodic or the GI symptoms may be limited to flatulence, vague abdominal discomfort and weight loss. In our cohort though none had diarrhoea at presentation, past history of short hospitalization for diarrhoea in previous three years was obtained in two patients. Another five patients on direct inquiry gave history of bloating, flatulence and vague abdominal discomfort, which completely improved within 3-4 weeks of starting GFD. In CD without GI symptoms, the extra-intestinal features often result from nutrient malabsorption and may involve virtually all organ systems. Often the only clues are routine blood test abnormalities - anaemia, hypocalcaemia. Increasing recognized extra-intestinal manifestations include osteopenic bone disease, neurological symptoms and infertility. Bone mineral density is almost invariably low in patients with CD.16 Osteoporosis occurs in over 25-35% of patients.17,18 Nut et al in Italy reported prevalence of undiagnosed CD in osteoporotic women (mean age 66.6 ± 8.5 SD) to be 9.4%.19 Our patient population is young, mean age 29 years, thus patients with senile osteoporosis are eliminated. We do not routinely screen postmenopausal women for osteoporosis and we have included patients with symptomatic osteoporosis this has given us a higher prevalence. Osteopenic bone disease develops as a result of impaired calcium absorption secondary to defective calcium transport by the diseased small intestine impaired absorption of vitamin D, and binding of intraluminal calcium and magnesium to unabsorbed dietary fatty acids, forming insoluble soaps.3 With prolonged calcium malabsorption, patient may develop secondary hyperparathyroidism (as seen in our patients), further exacerbating the osteopenia.3,20 Patients may present with bone pain (low back, rib cage, pelvis) and rarely pathological fracture. CD may be associated with dermatitis herpetiformis, several autoimmune diseases like Sjogren’s syndrome, polymyositis, SLE, IDDM, autoimmune thyroiditis. The current approach to evaluating CD has been modified by the advent of highly sensitive and specific serological tests. Positive IgA EMA or tTG followed by a positive small intestinal biopsy and improvement on gluten-free diet are necessary for making the diagnosis.

CONCLUSION Symptomatic osteoporotic patients (especially younger population) with associated anaemia should be investigated for CD. Simple measures like omission of wheat from diet (GFD) lead to significant improvement in symptoms within weeks. Acknowledgements We thank the ‘Staff Research Society’ of Lokmanya Tilak Medical College for funding this research. JAPI • VOL. 51 • JUNE 2003

1.

Catassi C, Fabiani F, Ratsch IM, et al. The celiac iceburg in Italy. A multicenter antigliadin antibodies screening for celiac disease in school-age subjects. Acta Paediatricia 1996;412:2935.

2.

Meewisse GW. Diagnostic criteria in celiac disease. Acta Paedatr Scand 1970;59:909-11.

3.

Richard J, Farrell MD, Ciaran P, et al. Diagnosis of celiac sprue. Am J Gastroenterol 200;96:3237-46.

4.

Walker SJA, Guanalini S, Schmitz K, et al. Revised criteria for the diagnosis of celiac disease. Report of a working group. Arch Dis Child 1990;60:909-11.

5.

Richard J, Farrel MD, Ciaran P. Celiac sprue. N Engl J Med 2002;346:180-8.

6.

Harold H, Grahm B, Peter F, et al. Celiac disease in primary care: case finding study. BMJ 1999;318:164-7.

7.

Cario C, Iise-Maria R, Lenora G. Why is celiac disease endemic in the people of the Sahara? Lancet 1999;354:647-8.

8.

Dicke WK, Weilers HA, Van de Kamer. Celiac disease II The presence in wheat of a factor having a deleterious effect in cases of celiac disease. Acta Paediatr 1952;42:34-42.

9.

Martin WJ, Brian BS. Endomysial antibody in the diagnosis and management of celiac disease. Postgrad Med J 2000;76:4668.

10. Satu S, Tuula H, Kaija L, et al. Transglutaminase autoantibody enzyme linked immunosorbent assay in detecting celiac disease. Gastroenterology 1999;115:1322-9. 11. Stefano G, Punit G. Do you still need a biopsy to diagnose celiac disease? Current Gastroent Reports 2002;3:32-8. 12. Marsh MN. Gluten major histocompatibility complex, and the small intestine - a molecular and immunobiologic approach to the spectrum of gluten sensitivity (celiac sprue). Gastroenterology 1992;102:330-54. 13. Scott BB, Losowsky MS. Patchiness and duodenal-jejunal variation of the mucosal abnormality in celiac disease and dermatitis herpetiformis. Gut 1976;17:984-92. 14. Whitehead R. Normal appearance in small intestine biopsy specimen. In: Whitehead R’s Mucosal Biopsy of the GI Tract. 5th Ed. Philadelphia, WB Saunders, 1997;III:175-184. 15. Katri K, Markka M, Jukka P, et al. Celiac disease without villous atrophy. Revision of criteria called for. Digestive Dis Sci 2001;46:879-87. 16. Corazza GR, Di Sario, Cecchetti L, et al. Bone mass and metabolism in patients with celiac disease. Gastroenterology 1995;109:122-8. 17. Kemppainen T, Kroger H, Janatuinen F, et al. Osteoporosis in adult patients with celiac disease. Bone 1999; 24:249-55. 18. Douglas M, Stavras S, Beverly D. Osteoporosis in a North America adult population with celiac disease. Am J Gastroenterol 2001;96:112-9. 19. Nuti R, Martini G, Valenti R, et al. Prevalence of undiagnosed celiac syndrome in osteoporotic women. J Intern Med 2001;250:361-6. 20. Selby PL, Davies M, Adams JE, et al. Bone loss in celiac disease is related to secondary hyperparathyroidism. J Bone Miner Res 1999; 14:652-7.

583