OP07 Consistent IBD treatment approaches across South Asian and White ethnicities despite phenotypic variations: a study of 33,157 patients using the IBD BioResource

Abstract

Background The current evidence suggests ethnic distinctions in IBD phenotype, and differences in the provision of treatment have been reported. This multi-centre cohort study utilised the UK IBD BioResource dataset to evaluate phenotypic differences between South Asian (SA) and White (WH) IBD, and to explore if these were associated with differences in treatment. Methods Phenotypic and outcome data were extracted from the IBD BioResource. Chi2 (categorical data) and Mann-Whitney U (continuous data) tests were used. Propensity score matching (PSM) accounted for age at diagnosis, sex, smoking status, disease location and behaviour and perianal disease (CD). Differences in medication use (multivariable logistic regression) and surgical outcomes (Kaplan-Meier and Cox regression analysis) were assessed in propensity-matched (PM) cohorts. Results 33,157 (31,932 WH; 1225 SA) individuals were included (48.1% CD, 45.4% UC, 6.5% IBD-U). UC was the predominant disease subtype in SA (UC, SA 57.3% vs WH 44.9%, p<0.001). SA were younger at diagnosis [CD, SA 24 (IQR 17-36) vs WH 26 (IQR 19-39) years, p<0.001; UC, SA 29 (IQR 22-38) vs WH 35 (25-48) years, p<0.001]. SA CD had less ileal disease (SA 30.3% vs WH 38.4%, padj=0.008), and more perianal involvement (SA 38.5% vs WH 32.3%, p=0.009) than WH. SA CD had less stricturing disease (SA 16.9% vs WH 25.6%, padj<0.001). SA UC were more likely to have extensive disease (SA 41.7% vs WH 34.1%, padj<0.001). Initial analyses in non-PSM cohorts showed that fewer SA CD underwent surgery [SA (n=157,37.4%) vs WH (n=7532,50.4%), p<0.001], and that similar proportions of SA (n=33,5.1%) and WH (n=747,5.5%; p=0.15) UC underwent a colectomy. PSM was used to match 355 SA to 355 WH in CD, and 525 SA to 525 WH in UC. Variables were well-balanced. There were no differences in 5-ASA, corticosteroid, thiopurine, anti-TNF or Vedolizumab use (Table 1). In CD, 126 (36.5%) SA and 152 (44.7%) had surgery. Survival analysis in CD showed no difference in the time to surgery (Fig 1A, log-rank 0.28). SA ethnicity was not associated with increased risk of surgery in CD (HR 0.82, 95% CI 0.63-1.07, p=0.14). In UC, 25 (4.8%) and 37 (7.1%) WH had a colectomy. There was no significant difference in the time to colectomy (Fig 1B, log-rank 0.12) nor was SA ethnicity associated with an increased risk of having a colectomy (HR 0.65, 95% CI 0.39-1.11, p=0.12). Conclusion In the largest analysis of SA IBD to date, we have demonstrated phenotypic differences associated with ethnicity. Accounting for these variations, we have shown comparable provision of medical and surgical treatment in SA and WH. These findings indicate consistent care of IBD patients from different ethnic backgrounds in the UK.