# # Key Findings

  • Progression-free survival: Pooled hazard ratio 0.799 (95% CI 0.677–0.922) across 15 RCTs, indicating a statistically significant benefit.
  • Non-small-cell lung cancer & cervical cancer: Subgroup analyses showed improved overall survival (HR 0.544 for NSCLC; HR 0.722 for cervical cancer) and progression-free survival.
  • Adjuvant scheduling: Adjuvant ICI therapy demonstrated stronger survival benefit (OS HR 0.742; PFS HR 0.638) compared to other timing.
  • Safety: Combination therapy increased the odds of grade ≥3 immune-related adverse events (OR 2.217, 95% CI 1.743–2.821).

# # Limitations

  • High heterogeneity (I² 72–82%) across primary endpoints undermines reliability of pooled estimates.
  • Only 15 RCTs are available; evidence base is small for broad conclusions.
  • English-language restriction may introduce publication bias.
  • Overall survival confidence interval upper bound of 1.000 indicates marginal statistical significance.