Different efficacy of pembrolizumab in overall survival between Asians and non-Asians with advanced cancers: A systematic review and meta-regression analysis.

Abstract

e13597 Background: Immune checkpoint inhibitors have revolutionized the treatment landscape of malignancy. However, disproportionate enrollment among different races/ethnicities in global trials placed the generalizability in doubt. This meta-analysis aimed to estimate the relative efficacy of pembrolizumab between Asians and non-Asians with advanced cancers. Methods: A systematic review was conducted in PubMed, Cochrane Library, Embase, and Web of Science. The meta-analysis included phase 3 randomized controlled trials investigating pembrolizumab in advanced cancers with available subgroup analysis for Asians and non-Asians. The experimental strategies involved monotherapy versus standard-of-care (SOC) or best supportive care, an add-on to SOC, and combined with multikinase inhibitor versus standard chemotherapy, regardless of the setting. The primary and secondary effect measures were the mean difference (MD) in the natural logarithm (i.e., log) of the hazard ratio (HR) for overall survival (OS) and progression-free survival (PFS) between the two subgroups [MD=log(HRAsians)−log(HRNon-Asians)=log(HRAsians/HRNon-Asians)]. We used the random-effects meta-analysis to calculate the pooled ratio of HRs (i.e., exp(MD)) and implemented a meta-regression analysis to identify significant covariates. The heterogeneity among the selected studies was evaluated using the Chi-squared Q test and the I2 statistic. Results: A total of 17 and 11 trials were selected in the meta-analysis for OS and PFS, respectively. These trials reported available subgroup results on Asians and non-Asians (patient numbers 2,732 [25.49%] and 7,000 [65.32%] for OS; 1,438 [22.50%] and 4,129 [64.61%] for PFS). The pooled ratio of HRs for OS was 0.87 (95% confidence interval [CI], 0.76−0.99; p = 0.0391; Chi-squared Q test Χ2 = 20.84, p = 0.29; I2 = 23.9%), favoring Asians, and no significant difference was found in PFS (the pooled ratio of HRs 0.93; 95% CI, 0.82−1.07; p = 0.2391; Chi-squared Q test Χ2 = 7.37, p = 0.77; I2 ≈ 0%). Both linear meta-regression analyses revealed an open-label design as an important covariate accounting for the observed difference, favoring non-Asians (estimated ratios of HR, 1.6318 for OS [95% CI, 1.1442-2.3273]; 1.3429 for PFS [95% CI, 1.0007−1.8020]). The sensitivity analysis showed that the percentage of subsequent treatment or post-progression programmed cell death 1 (PD-1)/ligand 1 (PD-L1) inhibitors had no significant impact. Conclusions: Asians receiving pembrolizumab for advanced cancers may have a superior OS advantage than non-Asians. Although the results warranted further exploration, this meta-analysis provided insight into clinical research design.