Long-term overall survival with dual CTLA-4 and PD-L1 or PD-1 blockade and biomarker-based subgroup analyses in patients with advanced non-small-cell lung cancer: a systematic review and reconstructed individual patient data meta-analysis

Background Immune checkpoint inhibitors targeting PD-L1 or PD-1 as monotherapy or combined with CTLA-4 inhibitors or chemotherapy (or both) are the standard of care for patients with advanced non-small-cell lung cancer (NSCLC). However, it remains unclear which patients benefit from the addition of CTLA-4 inhibitors. We aimed to evaluate whether dual checkpoint blockade with CTLA-4 and PD-L1 or PD-1 inhibitors provides similar efficacy to PD-L1 or PD-1 inhibitor monotherapy, or whether these strategies produce distinct outcomes across NSCLC subpopulations. Methods We conducted a search of PubMed, MEDLINE, and Embase for randomised phase 3 trials published from database inception to Nov 21, 2024, that investigated PD-L1 or PD-1 inhibitors, with or without CTLA-4 inhibitors, in patients with advanced NSCLC. We focused on studies reporting Kaplan–Meier survival data at 5 years or biomarker analyses based on PD-L1, KRAS , and STK11 mutational status. Individual patient data were extracted from Kaplan–Meier curves with WebPlotDigitizer version 5 and reconstructed with the IPDfromKM method. The primary endpoint of the study was 5-year overall survival in the overall population and in subpopulations based on PD-L1 tumour proportion score (TPS), tumour histology, and mutational status (mutant vs wild-type) of KRAS and STK11 . This study was registered with PROSPERO, CRD420251081707. Findings The initial search yielded 1026 results, and six randomised clinical trials met the eligibility criteria and were included. Among the 2881 patients eligible for analysis (838 [29·1%] female and 2043 [70·9%] male), 1282 received dual CTLA-4 and PD-L1 or PD-1 blockade and 1599 received single PD-L1 or PD-1 blockade. Patients treated with dual CTLA-4 and PD-L1 or PD-1 blockade had similar median overall survival compared with those treated with single PD-L1 or PD-1 inhibition (16·1 months [95% CI 15·0–17·8] vs 16·9 months [15·5–18·3]; HR 0·95 [95% CI 0·87–1·03], p=0·19). Median overall survival was significantly longer with dual CTLA-4 and PD-L1 or PD-1 blockade among patients with PD-L1 TPS less than 1% versus those treated with single PD-L1 or PD-1 inhibition (15·5 months [95% CI 13·6–18·5] vs 14·5 months [13·4–15·9]; HR 0·85 [95% CI 0·74–0·98], p=0·021), with 5-year overall survival rates of 16·6% (95% CI 13·4–20·6) versus 9·3% (7·0–12·3), respectively. Median overall survival in patients with tumours harbouring STK11 mutations was also significantly longer with dual CTLA-4 and PD-L1 or PD-1 blockade compared with single PD-L1 or PD-1 inhibition (13·9 months [95% CI 9·8–20·8] vs 7·8 months [6·4–12·9]; HR 0·67 [95% CI 0·49–0·91], p=0·012). However, no significant differences in overall survival were found between treatment groups by tumour histology (squamous vs non-squamous NSCLC) or by KRAS mutational status. Interpretation Compared with single PD-L1 or PD-1 inhibition, dual immune checkpoint blockade with CTLA-4 and PD-L1 or PD-1 inhibitors was associated with improved overall survival in patients with advanced NSCLC and PD-L1 TPS less than 1% and in those with STK11 mutations, but not in the overall population. Prospective validation of these results in clinical trials is warranted. Funding NextGenerationUE.