486 Cisplatin vs carboplatin-based chemotherapy combined with immune checkpoint inhibitor (ICI)- meta-analysis of efficacy and toxicities in randomized clinical trials

<h3>Background</h3> Platinum-based chemotherapy has cytotoxic and immunomodulatory effects with preclinical studies indicating cisplatin (CIS) may be more immunogenic than carboplatin (CAR), potentially leading to better synergy with ICIs. However, clinical data is limited. Our study aimed to compare the efficacy and safety outcomes of CIS vs CAR combined with ICIs across solid tumors. <h3>Methods</h3> A literature search was conducted using PubMed, clinicaltrials.gov, and conference abstracts to identify phase 3 trials of ICIs combined with CIS or CAR-based chemotherapy. Median overall survival (OS) and progression-free survival (PFS), hazard ratios (HRs), and adverse events were evaluated using a meta-regression random effect model. Analyses were adjusted for age, sex, region (global vs Asian), and control arm median survival (&lt;1 year vs. ≥1 year) to account for differences in tumor biology. <h3>Results</h3> We identified 40 trials totaling 29,276 patients (pts): 13 used CIS-based chemotherapy (n=9669 pts), 16 CAR-based (n=11,687), and 11 either(n=7920). The ICI combined included PD1 inhibitor (n=13,567), PD-L1 inhibitor (n=12,569), CTLA4 inhibitor (n=2421), or PD1 plus CTLA4 inhibitor (n=719). Trials involved pts with urothelial (n=2831), lung (n=14,859), esophageal (n=), and other cancers (n=8757). No significant difference in OS was found between CIS-ICI and CAR-ICI cohorts (Δ 0.29 months, p=0.86) or their controls (Δ 0.08 months, p=0.94). OS trends favored Asia over global regions for both CIS-ICI vs. CAR-ICI (Δ 3.89 months, p=0.08) and CIS vs. CAR cohorts (Δ 2.24 months, p-value=0.17). No significant difference between OS HRs of CIS-ICI vs CIS and CAR-ICI vs CAR was found (p=0.92). No significant difference in PFS was found for CIS-ICI vs. CAR-ICI overall (Δ 0.34 months, p-value=0.59). In the control cohort, an increase in PFS was observed for CIS vs. CAR (Δ 1.12 months, p-value=0.01). There was a decrement in PFS in men regardless of CIS or CAR (p &lt;0.001 in the experimental arm; p=0.15 in the control arm). No significant difference in PFS HRs between CIS-ICI vs CIS and CAR-ICI vs CAR was found (p=0.13). No significant differences were observed in adverse events (AEs) except for a 7% higher rate of Grade ≥3 immune-related AEs noted for CAR-ICI vs. CIS-ICI (p &lt;0.001). <h3>Conclusions</h3> While the efficacy of CIS-ICI and CAR-ICI were similar, the lower severe irAE rate with CIS may provide a better therapeutic index. The lower PFS in men regardless of therapy and the trend for better OS with CIS-ICI in Asian regions warrant further study. The study is limited by its trial-level analysis and cancer-type heterogeneity.