Real-world analysis of CRS and ICANS following tarlatamab and its association with survival in patients with SCLC.

e23440 Background: Tarlatamab, a DLL3-targeted bispecific T-cell engager (BITE), is approved for small cell lung cancer (SCLC) in second line and later settings. We aimed to evaluate real-world data regarding cytokine release syndrome (CRS) and immune effector cell-associated neurotoxicity syndrome (ICANS), their association with overall survival (OS), and post-progression management. Methods: A retrospective cohort study was conducted using the TriNetX Global Network (May 2024–December 2025). Patients with SCLC treated with tarlatamab were identified. CRS, ICANS and presence of baseline central nervous system (CNS) metastasis were identified via ICD-10-CM codes. Survival outcomes were compared using log-rank tests and Cox proportional hazards models. Subsequent treatment patterns and median time to next treatment (TTNT) were also analyzed. Results: We identified 643 patients with SCLC treated with tarlatamab (Mean age 66 ± 10 years; range 22–88; 47.9% female; 76.3% White, 7.1% Black). Overall, 314 patients (48.8%) comprised the toxicity cohort, having developed CRS and/or ICANS of any grade. Amongst those who experienced toxicity, 159 (50.6%) had CRS only, 50 (15.9%) had ICANS only, and 96 (30.6%) experienced both toxicities. Tocilizumab was utilized in 118 (37.6%) of the toxicity cohort (median time to administration: 1 day). Baseline CNS disease was balanced between the no-toxicity (n = 126; 38%) and toxicity (n = 124; 39%) cohorts. Survival analysis demonstrated significantly superior outcomes in patients without CRS or ICANS (Log-rank p < 0.0001). The no-toxicity cohort (n = 329) demonstrated a significantly lower risk of death (HR: 0.498; 95% CI: 0.368–0.675; p < 0.0001) with a 9-month survival probability of 69.1% for the no-toxicity cohort compared to 49.7% for the toxicity cohort. Following tarlatamab discontinuation, 38.1% of the total cohort initiated subsequent therapy with the most frequent agents including lurbinectedin (15.4%) and topotecan (7.1%), with a median TTNT of 98 days (IQR: 62–145). Conclusions: In this real-world analysis of SCLC patients treated with tarlatamab, CRS or ICANS was documented in around half of the study population with a higher incidence of ICANS compared to clinical trials. Occurrence of this toxicity was associated with worse overall survival and this observation warrants further study. Subsequent therapies after tarlatamab discontinuation reflected contemporary treatment patterns in this setting. Safety and survival tarlatamab in SCLC. Category Metric Value Safety Incidence of CRS or ICANS CRS alone ICANS alone CRS + ICANS 314 (48.8%) 159 (50.6%) 50 (15.9%) 96 (30.6%) Tocilizumab Utilization Rate 118/314 (37.6%) Survival 9-Month 0.487 (95% CI: 0.362-0.655) Log-Rank P-value < 0.0001