VALIDATE-PERIANAL: AN INTERNATIONAL REAL-WORLD MULTI-CENTRE VALIDATION OF THE TOPCLASS DEFINITION OF A RADIOLOGICALLY HEALED FISTULA

Abstract BACKGROUND Radiological healing on MRI is the ultimate therapeutic goal in Class 2a perianal fistulising Crohn’s disease (pfCD). The TOpClass consortium recently defined a radiologically healed fistula (TOpClass-RH). This study aimed to validate this definition in real-world practice. METHODS VALIDATE-PERIANAL was a retrospective, multi-centre, international study across three expert centres. pfCD patients with MRI evidence of a radiologically healed fistula between 2021–2023 were identified. Paired scans (baseline active and follow-up radiologically healed) were independently reviewed by three gastrointestinal radiologists using TOpClass-RH criteria, followed by consensus adjudication. At least 12 months of clinical follow-up was required. The primary outcome was sustained clinical remission; secondary outcomes included proctectomy, hospitalization, stoma formation. Reliability was assessed using Cohen’s kappa. RESULTS Of 977 patients screened, 40 met the inclusion criteria of a radiologically healed fistula in pfCD with a baseline MR scan showing active disease. Of these, 14/40 (35%) fulfilled TOpClass-RH criteria. Clinical recurrence occurred in 1/14 (7%) of TOpClass-RH patients versus 8/26 (30.8%) in the non-RH group (relative risk 4.3 [0.60–31.0], p = 0.12). Sustained clinical remission was achieved in 93% of TOpClass-RH cases, with a trend towards lower rates of proctectomy, hospitalization, and stoma formation compared to non-RH patients. Inter-rater agreement for TOpClass-RH was excellent (κ 0.89 –0.95). CONCLUSION TOpClass-RH predicted sustained clinical remission in 93% of patients and showed excellent reproducibility. These findings support TOpClass-RH as a robust and clinically meaningful radiological endpoint, providing a reliable benchmark for international trials and diagnostic stratification in pfCD. Patients not meeting TOpClass-RH criteria trended towards a fourfold rate of recurrence compared to existing standards, although limited sample size restricted statistical power.