LBA8505 Background: The vast majority of pts with PM have unresectable disease, due to co-morbidities and/or advanced stage. In unresectable PM, standard first line treatment has been C(or carboplatin)P for 20 years but it is usually only moderately efficient, mostly on epithelioid (E)PM, with a global median overall survival (mOS) around 12 months (m), an improved quality-of-life (QoL), and mild toxicity. Trial CM743 demonstrated a significant improvement in mOS for nivolumab/ipilimumab (NI) over CP (18.1 versus 14.1 m), particularly for non-EPM, but 30% NI pts experienced > grade 3 toxicity. Thus, alternative therapies to improve mOS and tolerability are needed. Methods: Canadian Cancer Trials Group (CCTG) IND 227 is an academic, open-label, randomized P3 study of the CCTG, National Cancer Institute of Naples (NCIN) and Intergroupe Francophone de Cancérologie Thoracique (IFCT). The study was supported by grants to CCTG (Canadian Cancer Society -707213); Merck & Co. Inc provided pembrolizumab (pembro) and a grant to support study conduct. Untreated unresectable PM pts ≥ 18 years with ECOG performance status 0-1, adequate hematological, renal and hepatic function, no active autoimmune disease, brain metastases, interstitial lung disease or other active co-morbidities, were randomized 1:1 to CP with or without pembro and stratified by histology (E vs. non-E). Carboplatin was allowed if C contraindicated. Radiological review (BICR) and PD-L1 testing was centralised. mOS was the primary endpoint; median progression-free survival (PFS) and response rate (RR), QoL (EORTC QLQ-C30 + QLQ-LC13) and economic analysis were secondary. Results: A total 440 PM pts were randomized; 218 to CP and 222 to CP-pembro. Data cut-off was 16 September 2022 when the required events was observed. All pts were included in the analyses. Arms were well balanced. Seven pts opted not to receive CP. Median exposure of CP was comparable between arms. Subsequent treatment with chemotherapy was similar in both arms, while 59 pts in CP arm (vs 17 CP-pembro pts) received immunotherapeutics. Using a stratified log rank test, mOS was 17.3m vs 16.1m for CP-pembro vs CP (HR 0.79 95%CI 0.64-0.98; p=0.0324), mPFS was 7.13m vs 7.16m (HR 0.80, 95%CI 0.65-0.99, p=0.0372). BOR was significantly higher for CP-pembro (63% vs 40%, p < 0.0001). Grade 3 or higher pembro related adverse events (AEs) occurred in 19% of patients (88% were grade 3; the most common were fatigue (5%) and diarrhea, pneumonitis and irAE (2% each)), and 16% of pts discontinued pembro per protocol for related AEs (most commonly diarrhea and pneumonitis). Conclusions: With a statistically significant improvement in mOS and acceptable tolerability. CP-pembro is an option for treatment naïve, unresectable PM. Clinical trial information: NCT02784171 .