Cabozantinib Plus Atezolizumab or Cabozantinib Alone in Patients With Advanced NSCLC Previously Treated With an Immune Checkpoint Inhibitor: Results From the Phase 1b COSMIC-021 Study

• Published in: JTO clinical and research reports 2024-10, Vol.5 (10), p.100666, Article 100666
• Main Authors: Neal, Joel W., Santoro, Armando, Gonzalez-Cao, Maria, Lim, Farah Louise, Fang, Bruno, Gentzler, Ryan D., Goldschmidt, Jerome, Khrizman, Polina, Proto, Claudia, Patel, Shiven, Puri, Sonam, Liu, Stephen V., Massarelli, Erminia, Williamson, Denise, Schwickart, Martin, Scheffold, Christian, Andrianova, Svetlana, Felip, Enriqueta
• Format: Article
• Language: English
• Subjects: Atezolizumab; Cabozantinib; Immunotherapy; Non–small cell lung cancer; Original
• ISSN: 2666-3643; 2666-3643
• DOI: 10.1016/j.jtocrr.2024.100666

We evaluated efficacy and safety of cabozantinib plus atezolizumab or cabozantinib alone in advanced NSCLC previously treated with an immune checkpoint inhibitor (ICI). COSMIC-021 (NCT03170960) is a phase 1b, multicenter study in advanced solid tumors. This analysis included patients with stage IV non-squamous NSCLC without actionable genomic aberrations in EGFR, ALK, ROS1, or BRAF-V600E who progressed on one prior ICI and less than or equal to two prior lines of systemic anticancer therapy. Patients received cabozantinib 40 mg orally/day plus atezolizumab 1200 mg intravenously every three weeks (combination cohort) or cabozantinib 60 mg orally/day (single-agent cabozantinib cohort). Primary end point of the combination cohort was objective response rate per Response Evaluation Criteria in Solid Tumors v1.1 by investigator. Outcomes in the single-agent cabozantinib cohort were exploratory. Eighty-one patients assigned to combination therapy and 31 assigned to single-agent cabozantinib received greater than or equal to one dose of study treatment. Median (range) follow-up was 26.1 months (12.1–44.2) and 22.4 months (1.5–29.0), respectively. Objective response rate was 20% (95% confidence interval: 11.7%–30.1%) in combination cohort and 6% (95% confidence interval: 0.8%–21.4%) in single-agent cabozantinib cohort. Treatment-related adverse events (TRAEs) occurred in 86% of patients in the combination cohort and 90% in the single-agent cabozantinib cohort; grade 3/4 TRAEs were 44% and 48%, respectively. There were two grade 5 TRAEs: pneumonitis (n = 1, combination) and gastric ulcer hemorrhage (n = 1, single-agent). Neither PD-L1 expression in tumor cells nor tumor mutation burden correlated with outcomes. Cabozantinib plus atezolizumab demonstrated modest clinical activity and manageable toxicity in advanced NSCLC after progression on prior ICI.