Tarlatamab in small cell lung cancer with brain metastases: A real-world experience.

Abstract

8108 Background: Tarlatamab, a DLL3-directed bispecific T-cell engager, has demonstrated survival benefit in previously treated small cell lung cancer (SCLC) in pivotal clinical trials. However, real-world evidence remains limited, particularly in ethnically diverse populations and in patients with a high burden of central nervous system (CNS) disease who are often underrepresented in clinical trials. We evaluated the safety, radiographic response, and treatment durability of tarlatamab in a real-world cohort treated at a single academic center. Methods: We retrospectively reviewed patients with extensive-stage SCLC treated with tarlatamab at Sylvester Comprehensive Cancer Center between January 2024 and October 31, 2025. Demographic, clinical, radiographic, and toxicity data were abstracted from medical records. Radiographic response was assessed per RECIST criteria, with best overall and 6-month responses recorded. Objective Response Rate (ORR) and Disease Control Rate (DCR) were calculated. Progression-free survival (PFS) and Overall Survival (OS) were estimated using Kaplan–Meier methods with multivariable analyses performed using Cox regression. Results: Among 23 patients (median age 72 years), male sex, Hispanic ethnicity, and brain metastases were each present in 61% of the cohort at tarlatamab initiation. 43% had received ≥2 prior lines of therapy. Cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANs) occurred predominantly during cycle 1 and were limited to grade 1 or 2 events, with no grade ≥ 3 toxicities observed. Median time on treatment was 92 days, with a median of 4 cycles. Three patients required treatment discontinuation due to toxicities. Radiographic response was evaluable in 18 patients; five were not evaluable due to early death after cycle 1 day 1 (n=2), loss to follow-up (n=2), and transition to hospice after first cycle (n=1). In those 18 patients, best overall response included partial response in 5 patients (27.7%) and stable disease in 3 patients (16.7%), yielding an ORR of 27.7 % and a DCR of 44.4%. Median PFS was 139 days, and median OS was 323 days (95 % CI, 31 to 614). No variables were independently associated with outcomes on multivariable Cox proportional regression analysis. Conclusions: In a predominantly Hispanic, heavily pretreated real-world SCLC population with a high prevalence of brain metastases, tarlatamab demonstrated feasible administration, manageable immune-mediated toxicity, and clinically meaningful antitumor activity. These findings support the effectiveness of tarlatamab beyond clinical trial populations and highlight the importance of real-world evidence in informing care for underrepresented patients. Best Overall Response N (%) Complete Response 0 (0) Partial Response 5 (27.7) Stable Disease 3 (16.7) Progression of Disease 10 (55.5) ORR 5 (27.7) DCR 8 (44.4)