DOP90 Efficacy of the Treat-to-target Approach in Modifying Disease Course with Ustekinumab in Patients with Moderate-to-severe Crohn’s Disease: Results from the STARDUST Trial

Abstract Background STARDUST is a phase 3b randomized trial comparing two therapeutic strategies with ustekinumab (UST) in patients (pts) with Crohn’s disease (CD): treat-to-target (T2T) using early endoscopic assessment vs standard of care (SoC). Results from the Week (W)48 and long-term extension (LTE) (W48 to W104) were published previously.1,2 Here we explore the efficacy of the T2T approach in modifying disease course with UST by analyzing time-to-disease modifying (DM) event up to W104. Methods Adult pts with moderate–severe active CD, received iv, weight-based UST ~6mg/kg at W0; then sc UST 90mg at W8. At W16, pts with ≥70-point reduction in CD Activity Index were randomized 1:1 to either T2T or SoC and received sc UST 90mg, every 12W/8W (up to 4W in the T2T arm), as per the protocol. From W48, eligible pts could continue to receive sc UST up to W104 with further protocol-guided dose adjustment. DM events were recorded through W48 and W104 starting at randomization, and time-to-event analysis was performed for pts in both arms. DM events are represented by combined bowel damage events (defined as the development of new strictures/fistulae or the occurrence of an intra-abdominal abscess) or CD-related hospitalizations or surgeries based on the adverse event analysis. Time-to-first bowel damage event, CD-related surgery, hospitalization, hospitalization or surgery was analyzed separately; Kaplan–Meier (K–M) curves with corresponding hazard ratios (HRs) and 95% confidence intervals (CIs) are shown here. Results Of 440 pts randomized to either T2T or SoC at W16, 74 discontinued before W48. Of the remaining 366 pts completing W48, 43 discontinued and did not enter LTE. At W48, 323 pts entered LTE; 20.1% of these pts discontinued before completing W104. HRs (95% CI) for time-to-first DM event from randomization through W48 and W104 are shown in Table 1, suggesting a numeric benefit in favour of the T2T arm. The separation of the K–M curves for time-to-first DM event between the two arms was apparent at W48 and continued up to W104, without significant difference (Figure 1a and b). Conclusion In STARDUST, with UST, starting at W48, numerically less DM events were observed in the T2T arm, mainly driven by the lower number of CD-related hospitalizations and bowel damage events, potentially linked to numerically higher proportion of endoscopic responders at W48.1 These results advocate for an optimized T2T approach using strictly defined targets like endoscopic response, on top of clinical and biomarkers, to facilitate decision making in clinical practice. References 1. Danese S, et al. United European Gastroenterol J. 2020;8:1264–1265 (Abstract LB11). 2. Peyrin-Biroulet L, et al. United European Gastroenterol J. 2021;9 (Suppl 1).