Predictive modeling to evaluate long-term treatment effectiveness of darvadstrocel in patients with complex perianal fistulas in Crohn's disease.

Background: Current therapies for complex Crohn's perianal fistulas (CPF) have a limited ability to achieve long-term healing. Darvadstrocel (DVS) is an expanded allogeneic adipose-derived mesenchymal stem cell therapy that has demonstrated efficacy in treating complex CPF in clinical trials. There are, however, limited long-term comparative data with standard of care (SoC). The aim of this study was to combine clinical trial data and real-world evidence using statistical methodologies to predict long-term effectiveness of DVS versus SoC in patients with CPF.

Methods: Data were pooled from a clinical trial (ADMIRE-CD) and two retrospective chart review studies (INSPECT and PREFACE). Predictive statistical models extrapolated clinical outcomes beyond observed follow-up using parametric curves, which were implemented into a semi-Markov model to obtain the number of patients in remission. The setting was multinational and multicenter. ADMIRE-CD was conducted in 49 hospitals in 7 European countries and Israel. INSPECT used data from the ADMIRE study. PREFACE involved patients from Belgium, France, Germany, Italy, and Spain. The participants were patients with complex CPF treated with DVS or SoC. Times to remission and relapse (clinical, and clinical plus patient-centric remission) were analyzed. Additionally, the proportion of patients in clinical and patient-centric remission was examined.

Results: In total, 513 patients were included in the analysis (ADMIRE-CD [N = 200] and PREFACE [N = 313]). Patients in ADMIRE-CD and PREFACE were similar in age (median [interquartile range, IQR], 36 [20.0] versus 36 [22.0] years, respectively) and gender (males, 54% and 52%, respectively). The median (IQR) duration of Crohn's disease was 9.4 [11.3] years for patients in ADMIRE-CD and 6.5 [12.9] years for patients in PREFACE. The estimated time to remission was shorter for patients treated with DVS versus SoC. The estimated time to relapse was longer for patients treated with DVS versus SoC. A higher estimated proportion of patients treated with DVS versus SoC had clinical and patient-centric remission at 24 months (48% and 35%, respectively) and 48 months (49% and 32%, respectively).

Conclusion: This novel approach enabled pooled data from a clinical trial and real-world settings to predict long-term effectiveness of DVS versus SoC in patients with complex CPF.