Answers are grounded in this paper and cite their sources.
Purpose Immune checkpoint inhibitors (ICIs) have reshaped the treatment of non-small cell lung cancer (NSCLC), yet the prognostic value of many reported factors remains uncertain. We conducted an umbrella review to evaluate the strength and credibility of evidence linking prognostic factors to overall survival in ICI-treated NSCLC. Methods PubMed, Embase, Web of Science, and the Cochrane Library were searched through July 1, 2026. Overlapping meta-analyses were managed using prespecified criteria. Pooled hazard ratios were recalculated, and heterogeneity, prediction intervals, small-study effects, excess significance, and a 10% credibility ceiling were assessed. Evidence was graded using an adapted Ioannidis framework, and methodological quality was evaluated with AMSTAR 2. Results Seventy-six meta-analyses yielded 113 associations across eight prognostic domains, of which 112 were gradable. Seven associations achieved Class I evidence: pretreatment Advanced Lung Cancer Inflammation Index, tissue tumor mutational burden, probiotic exposure, two pretreatment imaging/artificial intelligence-based risk models, ICI-associated thyroid dysfunction, and setting-specific cutaneous immune-related adverse events. Seventeen associations were Class II, 16 Class III, 31 Class IV, and 41 Class V. Several post-baseline and medication-related associations remained susceptible to time-dependent bias and residual confounding. Only 34 reviews were rated as high or moderate quality by AMSTAR 2. Conclusions Only a small proportion of proposed prognostic factors were supported by highly credible evidence. Statistical credibility does not establish causality, predictive utility, or clinical applicability. Prospective validation using standardized definitions, appropriate time-dependent analyses, and external validation of prognostic models is warranted.
No takes yet. Share an insight, caveat, or question.
Lu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: