Efficacy and safety of durvalumab with or without tremelimumab in metastatic NSCLC: A meta-analysis of randomized controlled trials.

e20568 Background: Lung cancer remains the leading cause of cancer-related mortality worldwide. Non–small cell lung cancer (NSCLC) comprises 80–85% of cases, and despite therapeutic advances, stage IV disease carries a poor 5-year survival rate below 10%. While immune checkpoint inhibitors have improved outcomes, the benefit of combining PD-L1 and CTLA-4 blockade remains uncertain. This meta-analysis aimed to evaluate the safety and efficacy of combined therapy in metastatic NSCLC. Methods: A comprehensive search of PubMed, Embase, and Cochrane Central was performed through May 2025 for randomized controlled studies (RCTs) assessing durvalumab alone or in combination with tremelimumab in advanced NSCLC. Outcomes of interest included overall survival (OS), analyzed using hazard ratios (HRs), and immune-related adverse events (irAEs), analyzed using odds ratios (ORs). Effect estimates were pooled using random-effects models and heterogeneity was assessed using Cochrane Q and I². When no direct comparisons were available, individual patient data were reconstructed from the original study Kaplan–Meier survival curves using the IPDfromKM method, with log-rank testing and Cox regression applied. Statistical analyses were performed in R version 4.5.0 with the meta package. Results: Three RCTs enrolling 2,726 patients met inclusion criteria. Of these, 891 (32%) were allocated to the durvalumab group and 884 (34%) to the durvalumab + tremelimumab group, whereas 891 (32%) were allocated to chemotherapy. Combined analysis showed no significant OS benefit with durvalumab plus tremelimumab compared to durvalumab alone (HR 0.97; 95% CI, 0.74–1.27). Pooled analysis of irAEs demonstrated significantly higher risks in the tremelimumab-containing arms: enterocolitis (OR 3.95; 95% CI, 1.71–9.11), pneumonitis (OR 2.25; 95% CI, 1.02–4.98), rash (OR 2.31; 95% CI, 1.28–4.16), thyroid disorders (OR 1.48; 95% CI, 1.02–2.16), and adrenal insufficiency (OR 2.97; 95% CI, 1.10–8.02). Conclusions: In metastatic NSCLC, the addition of tremelimumab does not confer a statistically significant OS advantage in unselected populations and is associated with a toxicity increase. Benefits may be restricted to biomarker-defined subgroups, but these findings remain exploratory. Further prospective studies are needed to clarify the optimal role of CTLA-4 blockade in combination immunotherapy. Characteristics of included studies. Trial Patients, No. Females, No. (%) Age ≥ 65, No. (%) ECOG, % 0/1 SCC, No. (%) ARCTIC a 126 36 (28.57) 56 (44.44) 40.5/59.5 32 (25.4) ARCTIC b 469 161 (34.32) 221 (47.12) 32.3/67.6 114 (24.31) MYSTIC 1118 346 (30.95) 248 (50.82)d 40.9/58.8 320 (28.62) POSEIDON 1013 243 (23.99) 477 (47.09) 33.4/66.5 374 (36.92) ECOG=Eastern Cooperative Oncology Group; SCC=Squamous cell carcinoma. a = PD-L1 ≥ 25%. b = PD-L1 <25%.