Landmark analysis of overall survival (OS) by objective response in patients (pts) with previously treated, advanced HER2-positive biliary tract cancer (BTC): Post hoc analysis of the HERIZON-BTC-01 trial.

545 Background: Zanidatamab is a dual human epidermal growth factor receptor 2 (HER2)-targeted bispecific antibody that received accelerated US approval and conditional approval in the EU and China for pts with previously treated, unresectable, or metastatic HER2-positive (immunohistochemistry [IHC] 3+) BTC. In the pivotal phase 2 HERIZON-BTC-01 trial, zanidatamab showed clinically meaningful antitumor activity (confirmed objective response rate, 52%) and a promising median OS (mOS) of 18.1 months in patients with HER2 IHC 3+ tumors. Here, we assess whether tumor response to zanidatamab was prognostic of longer survival. Methods: In HERIZON-BTC-01 (NCT04466891), pts with locally advanced, unresectable, or metastatic HER2 -amplified BTC with disease progression on or following prior gemcitabine-containing therapy were prospectively separated by HER2 IHC score (cohort 1: IHC 2+ or 3+; cohort 2: IHC 0 or 1+). Pts received zanidatamab 20 mg/kg intravenously every 2 weeks and underwent tumor response assessments every 8 ± 1 weeks. In this post hoc analysis, only pts with IHC 3+ tumors were included, and landmark survival models were employed at week 9 (initial response evaluation) and week 25 (time point by which all responses occurred). Responders had ≥1 assessment of complete (CR) or partial response (PR) by the landmark. Results: Between September 15, 2020, and March 16, 2022, 62 pts with IHC 3+ tumors were enrolled. The median (range) duration of follow-up for these pts was 34.0 (28–45) months. By the week 9 landmark, there were 27 responders (3 CR, 24 PR); 20 pts had stable disease (SD), 12 pts had progressive disease (PD), and 1 pt did not have an assessment. Post-landmark mOS (95% CI) was longer for responders (24.5 [16.6, not estimable (NE)] months) and pts with SD (14.4 [6.5, 21.2] months) vs all others (8.9 [2.3, 14.3] months). Hazard ratio (95% CI) was 0.40 (0.19, 0.83) for responders vs pts with SD and 0.38 (0.17, 0.82) for pts with SD vs all others. Seven (35%) of the 20 pts with SD by week 9 achieved PR by week 25. There were 36 responders (3 CR, 33 PR), 12 pts with SD, and 6 pts with PD by week 25. Considering the additional 4 months to landmark, post-landmark mOS (95% CI) was consistent with the week 9 landmark (responders, 20.0 [13.0, 29.3] months; SD, 8.4 [0.5, 15.2] months; PD, 5.4 [1.8, NE] months). Conclusions: In HERIZON-BTC-01, responders to zanidatamab by week 9 or 25 and pts with SD by week 9 demonstrated improved survival compared with all others (PD and not assessed). These results support a prognostic association between objective response or SD with zanidatamab and longer OS in BTC. These findings will be confirmed in the first-line setting in the ongoing phase 3 HERIZON-BTC-302 trial (NCT06282575) evaluating zanidatamab with standard-of-care treatment in pts with HER2-positive (IHC 3+ or IHC 2+/amplified) BTC. Clinical trial information: NCT04466891 .