Best epidemiological abstracts at ECCO’24

Balarajah et al (OP07) presented a multi-centre cohort study utilising the UK IBD Bioresource to specifically look at the phenotypic differences between South Asian (SA) and White (WH) patients and the possible differences in treatment pathways. The group used propensity score matching (PSM) as well as multivariable logistic regression, Kaplan-Meier and Cox regression analyses to generate their results. A total of 31,932 WH and 1225 SA patients were included. UC was the predominant disease subtype in SA patients. 24 vs 26 years for CD; 29 vs 35 years for UC) and had less ileal disease (30% vs 38%) and more perianal involvement (39% vs 32%); however, interestingly, SA patients had less stricturing disease (17% vs 26%). For UC, SA patients more commonly had extensive disease (42% vs 34%). In the PSM analysis (710 CD and 1050 UC patients) there was no difference between SA and WH patients with respect to 5-ASA, corticosteroid, thiopurine, anti-TNF or vedolizumab use. Survival analysis in CD showed no difference in time to surgery, and there was no increased risk of surgery in SA patients (HR 0.82, 95%CI 0.63–1.07). No significant difference in time to colectomy was observed between the patient groups, nor was SA ethnicity associated with an increased risk of colectomy (HR 0.65, 95%CI 0.39–1.11). This study highlights the phenotypic differences in SA patients but also indicates the comparable provision of medical and surgical management in the UK.

Poulsen et al (OP34) used a population-based cohort design from eastern Denmark to investigate the re-resection rates and risk of recurrence in a contemporary cohort who underwent primary intestinal resection between 2010 and 2020. Disease recurrence was defined as a colonoscopy with SES-CD ≥3, Rutgeerts score ≥i2, inflammation/stenosis on imaging, calprotectin >250 µg/g, starting steroids after resection or re-resection. Cox regression and PSM were used for analysis. 631 patients had a primary resection, with a median follow-up of 118 months. Prior to resection, 53% and 39% received immunomodulators and biologics, respectively; following resection the rates were 50% and 42%. Resection rates at 2, 5 and 10 years from diagnosis were 41%, 67% and 84%, respectively. Re-resection rates due to disease activity at 5 and 10 years after primary resection were 5% and 10%, respectively. Biological therapy had no influence on re-resection or recurrence rates within the first year following resection (HR 0.4, 95%CI 0.12–1.34). Only one in four patients remain relapse-free during long-term follow-up and despite extensive biological use 15% still require further resections.

Attauabi et al (OP38) used another population-based Danish cohort study design to address the problem of screening IBD patients for metabolic bone disease. Using a prospective design from May 2021, as part of the IBD Prognosis study, all patients had a DXA scan at diagnosis. A T-score <2.5 SD was diagnostic for osteoporosis, while a T-score between 1 and 2.5 SD defined osteopenia. 316 patients were included and had DXA scanning at a median of 90 days (IQR 51–139) from diagnosis. Of UC patients, 27% and 10% had osteoporosis and osteopenia, respectively; the corresponding rates for CD patients were 20% and 8%. Low albumin (<36 g/L) was the only factor associated with osteopenia in UC patients; no risk factors were identified in CD patients. Risk of osteoporosis was independently associated with age higher than 65 years in UC (aOR 13.4, 95%CI 3.8–51.4) and CD (aOR 6.4, 95%CI 1.1–35.5). Female sex was not relevant in either UC or CD. The risk of metabolic bone disease was not associated with IBD phenotype, medications or endoscopic severity.

Revés et al (DOP05) presented data looking at bowel damage (BD) and disability in a multicentre European cohort study of newly diagnosed CD patients (CROCO) intended to prospectively characterise BD progression using the Lémann index (LI). At 1 year following inclusion, MR enterography was performed in all patients with additional endoscopy and/or pelvic MRI based on disease location. Any bowel damage was indicated by an LI >0; disability was assessed using the IBD-disability index (IBD-DI). Of the 261 patients in the study, 135 had completed their year 1 visit and 100 had LI calculations completed. 90% of patients had ileal/ileocolonic involvement, 65% had an inflammatory phenotype and 11% had perianal disease. 53% had initiated biological therapy within the first year, with higher rates of biological use in stricturing (77%) and penetrating (79%) behaviour. In total, 61% of patients had some degree of BD but overall the median LI was low, at 0.6 (IQR 0–2). In a multivariable logistic model only older age at diagnosis appeared protective against BD (p=0.037). Of the 100 patients evaluated for LI, 84 completed the IBD-DI at year 1; these patients has a median score of 17.3 (IQR 10.7–32.6), with 30% experiencing moderate to severe disability (IBD-DI >35). No association was seen between LI and IBD-DI (OR 1.1, 95%CI 0.4–3.0), and there were no differences in the median LI across disability categories (p=0.67).