Predictors and timing of response to teduglutide in patients with short bowel syndrome dependent on parenteral support

Background and aims

This study aimed to identify predictors and estimate time to teduglutide response among adult patients with short bowel syndrome with intestinal failure (SBS-IF) dependent on parenteral support (PS).

Methods

Post-hoc analysis was performed on individual patient data from teduglutide-treated patients in the phase III teduglutide trial STEPS and the STEPS-2 extension. Response was defined as ≥20% PS volume reduction from baseline for two consecutive visits. Early responders experienced the reduction at 20 and 24 weeks during STEPS while late responders experienced the reduction during STEPS-2. Timing and predictors for response were assessed among the treated population using Cox proportional hazard model. Time to response was compared in aetiological subgroups using Kaplan-Meier analysis. Patient characteristics and time to response were compared between early vs. late responders.

Results

A total of 34 patients were included in this analysis; overall median time to response was 4.3 months. The presence of stoma predicted a positive response to teduglutide (hazard ratio [HR]: 5.6; 95% confidence interval [CI]: 1.4-21.9; p = 0.013). Vascular disease (vs. inflammatory bowel disease [IBD]) as cause of major intestinal resection (HR: 0.2; 95% CI: 0.0-0.8; p = 0.015), presence of ileocecal valve (HR: 0.1; 95% CI: 0.0-0.8; p = 0.047), and female sex (HR: 0.3; 95% CI: 0.1-1.0; p = 0.026) are negatively associated with response. In subgroup analyses, patients with IBD (vs. vascular disease), with (vs. without) a stoma, and without (vs. with) colon-in-continuity had a shorter time to response (all p < 0.05). The mean times to response were 3.6 (standard deviation (SD): 1.1) months for early responders (n = 27) and 10.0 (SD: 6.1) months for late responders (n = 7). Fewer early responders had colon-in-continuity (51.9%) and ileocecal valve (0.0%) compared to late responders (100% and 28.6%, respectively; both p < 0.05). Early responders had a lower mean percentage of colon remaining compared to late responders (24.6% vs. 57.1%, respectively; p = 0.016).

Conclusions

Time to response to teduglutide depends on bowel anatomy and SBS-IF aetiology. IBD, presence of a stoma, and absence of ileocecal valve were associated with earlier response to teduglutide. These findings may enhance management of patients with SBS-IF; however, due to sample size limitations, additional studies are needed to confirm these findings.