(M-118) Past as Prologue: A Novel Anti-Fucosyl-GM1 Monoclonal Antibody in Extended Stage Small Cell Lung Cancer (ES-SCLC) Phase 3 Dose Selection through Exposure/Dose-Response Analyses of Efficacy and Safety

Yue Zhao, MS – Director, Bristol Myers Squibb; Sean Cheng, PhD – Senior Research Investigator, Bristol Myers Squibb; Laureen Ojalvo, MD – Clinical Trial Physician, Bristol Myers Squibb; Yu Liu, MD, PhD – Vice President, Bristol Myers Squibb; Jiaju Wu, PhD – Senior Manager, Bristol Myers Squibb; Rafael Sarmiento, PhD – Clinical Scientist, Bristol Myers Squibb; Sarah Tannenbaum-Dvir, MD – Clinical Trial Physician, Arvina; Zariana Nikolova, MD – Vice President, NA; Satyendra Suryawanshi, PhD – Senior Director, Bristol Myers Squibb

Objectives: BMS-986012, a fully human IgG1 monoclonal antibody, targets the unique surface antigen fucosyl-GM1 highly expressed in SCLC cells but limited in healthy tissues [1]. In Phase 1 and 2 studies, BMS-986012, when administered with nivolumab (nivo) or nivo and carboplatin/etoposide (chemo), showed promising antitumor activity and manageable safety in recurrent/relapsed [2] and first-line (1L) ES-SCLC. The E-R analyses of tumor growth dynamics (TGD), overall survival (OS), and safety were conducted to determine the appropriate dose for a Phase 3 trial evaluating BMS-986012 in 1L ES-SCLC.

Methods: The E-R analyses of TGD and safety (Grade3+AEs and pruritus) utilized pooled datasets from 169 patients of all dose groups in a Phase 1 study, CA001-030 (70 to 1000 mg BMS-986012 dosed alone or combined with nivo), and an ongoing Phase 2 study, CA001-050 (420 mg BMS-986012 + nivo + chemo). Using a TGD model with exponential tumor shrinkage (TS) and linear tumor growth function (TG) [3 ], longitudinal tumor size profiles were analyzed. Regimen effects on TS/TG were evaluated, alongside the correlation between model-predicted individual TS/TG and exposure. E-R analysis for OS focused on 64 ES-SCLC patients who received 420 mg BMS-986012 + nivo + chemo. Cox proportional hazards model assessed the impact of BMS-986012 exposure on OS, with exploratory D-R analysis of Grade3+AEs and E-R of pruritus severity using ordinal logistic regression. Exposure metric (time-averaged concentration over the first cycle [Cavg1]) was chosen based on pharmacological plausibility and modeling approach.

Results: Significant regimen effects on TS/TG were noted (reference: 420 mg BMS-986012 + nivo + chemo). Higher BMS-986012 exposure correlated with increased TS and decreased TG on the triplet regimen although moderately improved tumor shrinkage did not translate into enhanced OS. The E-R analysis revealed that BMS-986012 exposure was a significant covariate, however, the hazards ratio at the 5th and 95th percentiles relative to the median was similar, indicating a flat E-R relationship across the tested exposure range. Additional E-R analysis confirmed this flat effect, showing consistency between Cavg1 quartiles and OS, with or without adjusting for baseline characteristics. The E-R analysis of safety revealed comparable risk of Grade3+AEs across treatment arms. Finally, for pruritus, adding BMS-986012 increased the likelihood of all grades compared to no pruritus, with probabilities unchanged across varying Cmax ranges tested.

Conclusions: Preliminary E-R analyses based on interim data suggest a flat exposure vs OS relationship with a tolerable safety profile. Interim data, including safety, efficacy, PK, and E-R analyses, support further development of BMS-986012 in a Phase 3 trial, using the same dose regimen evaluated in the ongoing Phase 2 trial.

Citations: [1] Ponath, P., et al (2018). Cancer. Clin Cancer Res, 24(20), 5178-5189.
[2] Chu, Q., et al (2022). JTO Clin Res Rep, 3(11), 100400.
[3] Wang, Y, et al (2009). Clin Pharmacol Ther. 2009;86(2):167-74.