Crohn's disease in rheumatology clinic-An Indian experienceDebashish Danda1, G Kurian2, A Chacko2, G Chandy2, S Patra2, A Mathew1, K Thomas1
1 Department of Medicine II and Clinical Immunology & Rheumatology Unit, Christian Medical College and Hospital, Vellore, India
2 Department of Gastroenterointestinal Sciences and Wellcome Research Laboratory, Christian Medical College and Hospital, Vellore, India
Correspondence Address: Source of Support: None, Conflict of Interest: None PMID: 16388168
Source of Support: None, Conflict of Interest: None
Background: Prevalence of Crohn's disease (CD) among patients with rheumatic illnesses in India is grossly under estimated, especially when it has overtaken that of Ulcerative Colitis in the West. Aim: To study the frequency of histologically unequivocal CD amongst clinically suspected patients with enteropathic arthropathy and to ascertain if the arthritics with CD have any independent clinical predictor. Settings and designs: Retrospective datasheet analysis from a Rheumatology clinic of a large tertiary care centre. Materials and Methods: Patients of suspected enteropathic arthropathy were studied by ileocolonoscopy and segmental colonic biopsy for histological evidence of Crohn's disease and followed up. Statistical Analysis: Logistic regression analysis was done to find out any clinical predictor of histologically proven CD. Results: Fourteen of the twenty-nine patients studied had histologically confirmed CD. Those with CD were younger than those without (34.7 yr vs 41.6 yrs, p =0.057). The CD group also had significantly higher number of people with loss of weight (12 vs 1), fever (11 vs 0), perianal fistula (4 vs 0), abdominal pain (8 vs 2), history of dysentery (4 vs 0) and uveitis (6 vs 1) (p=0.00002, 0.00001, 0.026, 0.013, 0.026 & 0.01 respectively). However logistic regression analysis of the most relevant ones among these, namely, loss of weight, fever, and perianal fistula showed loss of weight as only independent predictor of CD in this subset of patients ( p = 0.03 with odds ratio of 28). Conclusion: Presence of significant loss of weight in an Indian patient with clinically suspected enteropathic arthropathy is an independent predictor of CD.
Keywords: Arthritis, Crohn′s disease, inflammatory bowel disease, weight loss
Crohn's disease (CD) is being recognized in the West more often over the last two decades., Up to 25% of patients with CD can have extra intestinal symptoms as its manifestation, arthropathy being the commonest among them., CD associated with enteric arthropathy can be of spondyloarthropathy (SpA) type, Rheumatoid (RA) type, a combination of SpA and RA type, oligoarticular, monoarticular, or Jaccoud type arthritis and other atypical arthritis. Often the arthropathy of CD can accompany, precede, or follow bowel symptoms. Symptoms of arthritis can be the dominant presentation and the bowel symptoms can be very mild. In such a situation gastrointestinal manifestation can be considered as a separate entity, both by the patient and the physician.
There is also a lack of awareness regarding CD, as it is considered to be a rare entity in our country and classical cases remain undiagnosed. Although CD was reported in Asian Indians overseas, there was no Indian data from India at the time of the study. We thus looked for histolgical evidence of CD and its clinical correlates in suspected patients of enteropathic arthritis from our Rheumatology Clinic.
In this retrospective study, patients were selected from the outpatient and inpatient departments of Medicine, Gastroenterology and Rheumatology between September 1996 and August 1998 and thereafter the follow up was recorded from the case sheets till March 2005. Patients included in this study were those of chronic inflammatory arthropathy with chronic abdominal pain, diarrhea, dysentery, perianal abscess and perianal fistula. Those with obvious irritable bowel syndrome or hemorrhoids were excluded after clinical evaluation.
Chronic inflammatory arthritis was defined as arthritis of large, small or axial joints of more than 6 months duration with daily morning stiffness of more than 1 hour or elevated inflammatory parameters like raised ESR or C-reactive protein, in the absence of any obvious infective process.
A detailed rheumatological and gastroenterological examination was done for these patients. Ileocolonoscopy and segmental colonic biopsy were done in those patients who gave informed consent. Histological evaluation of biopsy specimens was done to look for evidence of CD. Endoscopic appearance of CD namely aphthous ulcers, deep or longitudinal fissures, cobble stone appearance, edema, strictures, pseudo polyps, skip areas and rectal sparing of the ulcerative lesions were noted. But histological appearance was taken as the gold standard for diagnosis of CD. Classical histological description of discrete, non-confluent, noncaseating microgranuloma was adopted to define CD. Confluent granulomas with caseation of tuberculosis were differentiated from CD by dedicated gastrointestinal pathologist. The stains used were H&E
All patients were treated with sulphasalazine and or Methotrexate and followed up in our hospital at regular intervals.
The demographic parameters, clinical symptoms and complications were compared between those arthritics with CD and those without. This was done by univariate analysis using Chi square and Fischer's exact test. Subsequently, logistic regression analysis was done among the most relevant parameters to look for any independent clinical predictor for the diagnosis of CD. The software package SPSS was used for regression analysis.
Of the 32 patients fulfilling the study criteria, ileocolonoscopy and segmental colonic biopsy was done in twenty nine. Fourteen of these 29 patients had histological evidence of CD [Figure - 1]. Of the remaining, five patients had non-specific chronic colitis without any definite evidence of inflammatory bowel disease, five other had unclassifiable minimal inflammatory infiltrate, three had possible infective colitis without any identifiable organism and two others were totally normal.
Baseline demographic characteristics of the patients with CD and those without CD are shown in [Table - 1]. Majority of the patients with CD belonged to the eastern and southern part of India. There was however no significant difference in the geographical distribution of patients with or without CD. The CD group patients were of younger age group when compared with those without CD ( P = 0.057). The mean duration of joint symptoms was 4.6 years in the Crohn's group and 5.9 years in the other. Similarly the mean duration of bowel symptoms at presentation was 4.2 years in the Crohn's group and 7.2 years in the other group. All the other parameters in [Table - 1] were similar in those with and without CD. Inflammatory markers like ESR and CRP also did not differ between the two groups (data not shown).
The clinical parameters that varied significantly by univariate analysis between those two groups were significant weight loss, fever, perianal fistula, abdominal pain, dysentery and uveitis [Table - 2].
Logistic regression analysis of the most relevant clinical parameters that were significant on univariate analysis showed loss of weight as the only significant predictor for diagnosis of CD ( P = 0.03, odds ratio = 28, [Table - 3]).
Other interesting observations in this study were:
1. Two patients with CD and arthropathy had concomitant pulmonary tuberculosis (no evidence of intestinal tuberculosis).
2. Two patients with CD had psoriasis.
3. One patient with CD had Systemic Lupus Erythematosus (SLE) with very high anti cardiolipin antibody but without any obvious thromboembolism.
All patients with CD were treated with Sulfasalazine and Methotrexate. None of them could afford Biological agents. During the 5.54 (+/-1.71) years follow up period, dysentery, abdominal pain and fever completely resolved with treatment in all patients. Patients with perianal fistula had surgical intervention in addition to the medical therapy and there has been no recurrence of fistula till date. All patients gained weight following treatment. Although joint symptoms improved in all, four of them had recurrent arthritis requiring intra articular steroid injections. One patient with severe diffuse polyarthritis has been given a single pulse of IV Methyl Prednisolone. One of the patients had to undergo a Total Hip Replacement for advanced hip joint disease. Two patients had recurrent uveitis requiring topical and systemic steroids. The CD patient with SLE who has been on steroids developed pulmonary tuberculosis for which she was treated. Subsequently she also developed pulmonary Aspergilloma.
The non Crohn's group was also treated with Sulfasalazine with or without Methotrexate, in view of chronic inflammatory arthropathy. During the 3.27 (+/- 2.13) years follow up period of these patients, two of them had persistent bowel symptoms including recurrent bleeding per rectum, requiring blood transfusion in one. Repeat colonoscopy in both these patients revealed Ulcerative Colitis. They were given steroid pulses and enema, in addition to the existing therapy, and are currently doing well.
There was no death during the follow up period.
This first ever report on Crohn's arthropathy from India revealed histologically unequivocal CD in half the suspected cases (14 out of 29) and significant loss of weight was the single most independent clinical predictor with an Odds ratio of twenty eight. This clinical clue is a novel finding which would enable detection of more cases of CD in India, where the disease is under recognized.
One of the difficult areas in differentiation may be that from tuberculosis of the intestine., The absence of diffuse, confluent macro granuloma with caseation can rule out intestinal tuberculosis with reasonable precision.
Most of the common features of CD seen in our patients like young male in thirties, perianal fistula, recurrent abdominal pain, dysentery, uveitis, fever and loss of weight are described in literature.,
Significant loss of weight, an independent predictor of diagnosis in CD, as found in this study may be due to high level of tumour necrosis factor (TNF) alpha which can well explain the very basis for highly successful response to treatment with TNF antagonists in CD.
Limitations of this study include small number of patients, retrospective analysis of datasheets and a skewed population in a referral centre. However, this study can act as a sensitizer to create awareness for this condition in our country.
A large multicentric, longitudinal population based study involving subjects from all the regions of India and the Indian subcontinent with HLA delineation may bring out more information about CD associated arthropathy. Interestingly CD has been reported from southern India in contrast to infective colitis, microscopic colitis and ulcerative colitis from northern India.,
[Figure - 1][Table - 1], [Table - 2], [Table - 3]