Microscopic colitis

Abstract

Microscopic colitis (MC) is an inflammatory bowel disease (IBD) and a common cause of chronic non-bloody diarrhoea, especially in elderly women. There are two main subtypes, lymphocytic colitis (LC) and collagenous colitis (CC) which are clinically indistinguishable and can be separated only by their characteristic histopathological features. The colonoscopy is usually macroscopically normal although subtle mucosal changes have been reported. The aetiology of MC is unknown and the genetic factors are poorly investigated.

This thesis aims to describe MC in a large urban cohort and compare LC and CC regarding clinical and endoscopic features, both at diagnosis and at follow-up (F-U), and to observe the occurrence of coeliac disease, ulcerative colitis (UC) and Crohn’s disease (CD). We also reported the histological change of the MC phenotype over time and patients’ MC medication at last F-U. Further, we tested immune-related genes, known to impact several autoimmune diseases, for their potential CC-predisposing role. Finally, this thesis aims to report on chromoendoscopic findings in MC.

A retrospective study of 795 patients showed that the clinical features of LC (n=451) and CC (n=344) were similar, though watery diarrhoea occurred at a lower frequency in LC (43%) than CC (55%) as did nocturnal diarrhoea, LC (18%) and CC (28%). The mean age at diagnosis was lower in LC, at 59 years compared 63 years in CC. Subtle endoscopic mucosal findings were frequently reported at a higher rate in CC (37%) than LC (25%). Our study confirms MC’s strong association with coeliac disease, which occurred in 6% of the patients. UC and CD occurred in 2.1% of the patients.

In the MC cohort, 687 patients had a clinical F-U after a mean time of 2.89 years at which 64% were in clinical remission. The cumulative clinical remission was higher in LC. About half of the patients had received medical treatment for MC at last F-U and about a quarter of them were on steroids; both these parameters were lower in patients with LC. The mean time for the first F-U colonoscopy (n=187) was 3.33 years and the histological remission was 44% and the cumulative histologic remission was higher in LC. Histological change of phenotype over time was not uncommon and was observed in 12 % of the patients (10 CC to LC, 13 LC to CC).

Three independent CC and control cohorts were genotyped with Immunochip and 42 markers gave rise to significant genome-wide associations signals, all within the HLA region of chromosome 6. The most pronounced risk effects were observed for 8.1 haplotype alleles including DQ2.5. The HLA genotype is associated with CC and indicates immune-driven pathogenesis.

Previously, two case reports described that chromoendoscopy in CC patients showed uneven surface. We reported additional chromoendoscopic findings in 13 MC patients that showed continuous mucosal changes. Our study supports the fact that chromoendoscopy can reveal mucosal changes in MC and therefore might be diagnostically useful. LC and CC are similar but not identical, since LC has a milder clinical presentation and a better prognosis than CC. Conversions between subtypes and between MC and UC or CD exist. It is not uncommon with macroscopic changes of the colonic mucosa in patients with MC which are more manifest with chromoendoscopy. Specific HLA alleles are associated with CC, indicating an autoimmune role.