Background: At the primary analysis of TOPAZ-1 (NCT03875235; data cut-off [DCO] Aug 11, 2021), durvalumab (D) + gemcitabine and cisplatin (GC) significantly improved overall survival (OS) vs placebo (PBO) + GC in participants (pts) with advanced biliary tract cancer (BTC), with OS curves that separated over time (Oh et al. NEJM Evid 2022), that persisted with further follow-up (DCO Feb 25, 2022; HR, 0.76 [95% CI, 0.64–0.91]; 18-month OS, 34.8% [D + GC] vs 24.1% [PBO + GC]; Oh et al. Ann Oncol 2022 Abs 56P). We characterized long-term survivors (LTS) in TOPAZ-1. Methods: Pts untreated for unresectable locally advanced, recurrent or metastatic BTC received D (1500 mg every 3 weeks [Q3W]) or PBO, + G (1000 mg/m²) and C (25 mg/m²) on days 1 and 8 Q3W, up to 8 cycles, followed by D (1500 mg Q4W) or PBO. Characteristics and outcomes of LTS (pts who survived ≥18 months after randomization; DCO Feb 25, 2022) were assessed. Results: There were more LTS with D + GC vs PBO + GC (Table). Characteristics of LTS were consistent with the full analysis set (FAS, all randomized pts) including age, sex, region, primary tumor location, disease classification (metastatic vs locally advanced) and PD-L1 expression. Recurrent disease was more common than initially unresectable disease in LTS vs FAS. Median exposure to study treatment was 11.3 months (mo; D), 9.7 mo (PBO) and 5.5 mo (GC, both arms) in LTS and 7.3 mo (D), 5.8 mo (PBO) and 5.1 mo (GC, both arms) in FAS. The objective response rate in LTS was greater with D + GC vs PBO + GC (44.3%, D + GC; 33.8%, PBO + GC), and greater for both groups vs FAS (26.7%, D + GC; 18.7%, PBO + GC). A higher proportion of LTS in the PBO + GC group received subsequent anticancer therapy, including immunotherapy, than LTS in the D + GC group. The frequency of treatment-related adverse events leading to discontinuation in LTS was consistent with FAS. Conclusions: In TOPAZ-1, characteristics of LTS were generally consistent with the FAS. Overall, LTS were more common with D + GC; LTS in the PBO + GC group appear to be associated with higher frequency of subsequent treatments, including immunotherapy. Clinical trial information: NCT03875235.

PD-L1, programmed cell death ligand-1; TAP, tumor area positivity.