Investigators pooled data from 42 phase 3 randomized trials including 25,827 patients to compare immune checkpoint inhibitors with chemotherapy in advanced non-small cell lung cancer. The pooled results indicated that ICIs significantly improved both overall survival and progression-free survival versus chemotherapy for male and female patients. The cohort showed a male-specific pooled OS hazard ratio of 0.74 and PFS hazard ratio of 0.59; for females the pooled OS HR was 0.83 and PFS HR 0.76, as reported in the source. The study team then calculated trial-level treatment-by-sex interaction hazard ratios and observed a statistically greater benefit for males than females for OS (interaction HR 0.90; P = 0.01) and for PFS (interaction HR 0.82; P = 0.0006). Heterogeneity differed by outcome and sex; the authors reported I2 values and applied random-effects models accordingly.
The authors performed multiple pre-specified subgroup analyses. Sex-based differences in OS appeared in subgroups treated with ICI monotherapy, anti‑PD‑1 monotherapy, and in patients with PD-L1 ≥50%, but those OS interaction signals did not survive False Discovery Rate adjustment. For PFS, however, sex differences favoring males persisted after FDR correction in the ICI monotherapy subgroup, the anti‑PD‑1 monotherapy subgroup, and the PD‑L1 ≥50% subgroup. In contrast, no significant sex-based differences in either OS or PFS were observed for patients treated with ICI plus chemotherapy. The study team also reported that ICI benefit versus chemotherapy was generally present across Asian and non‑Asian populations, non‑squamous histology, first-line and later-line settings, and across most treatment modalities when sexes were analyzed separately.
The authors applied GRADE and reported moderate-quality evidence for OS in both sexes and low-quality evidence for PFS in both sexes. Multivariable meta-regression identified ICI type and race as significant contributors to between-study heterogeneity for males, while in females ICI type and race associated with OS heterogeneity and PD-L1 expression associated with PFS heterogeneity. The study team detailed these associations in supplemental tables. The authors stated that twelve trials were judged low risk of bias, while thirty trials carried a high risk of bias primarily driven by open-label designs.
Data indicated robustness in sensitivity checks. Leave-one-out analyses produced no material change to pooled male and female HRs or to interaction HRs, according to the source. The authors also re-ran analyses excluding trials with very small female intervention-arm samples (≤30) and reported similar pooled estimates. Egger’s regression provided no evidence of publication bias for OS or PFS in either sex, per the published results, and corresponding funnel plots were presented in supplementary materials.
The authors discussed several mechanisms that could contribute to the observed sex differences. Tumor and immune profiling studies, cited in the source, show sex-specific differences in T-cell function, expression of inhibitory checkpoints, and the tumor microenvironment that could reduce ICI monotherapy efficacy in females. The authors also observed a lower prevalence of high PD-L1 expressors in females across multiple clinical datasets, which could reduce relative benefit versus chemotherapy. They noted clinical evidence suggesting females may derive relatively greater benefit from chemotherapy than males—an effect that would reduce the apparent incremental benefit of ICI when measured against a chemotherapy control.
The authors explicitly warned that females were underrepresented in many trials and that baseline imbalances—smoking prevalence among them—could confound sex-specific estimates. They reported that several subgroup OS signals lost statistical significance after False Discovery Rate correction, and they cautioned against overinterpretation. The source did not provide granular individual-patient dosing or covariate-level distributions beyond the trial-level summaries. The authors concluded that the observed attenuation of PFS benefit for ICI monotherapy in females requires confirmation in randomized trials specifically designed to test sex-based differences, and they recommended prospective RCTs that balance sex representation and control for key confounders.