KRAS-mutant advanced NSCLC: efficacy and clinical outcomes from network meta-analysis and real-world evidence

KRAS is a common driver gene in non-small cell lung cancer (NSCLC). Owing to the absence of universally effective targeted agents and marked heterogeneity, optimal treatment selection for KRAS-mutant NSCLC remains uncertain. This study combined network meta-analysis (NMA) with real-world evidence to inform precision treatment strategies.

Thirteen randomized controlled trials were included in a Bayesian NMA to compare efficacy outcomes across regimens in KRAS-mutant and KRAS G12C-mutant NSCLC. In parallel, real-world data from advanced KRAS-mutant NSCLC patients treated at our center were analyzed.

Across KRAS and G12C NMAs, immunotherapy-based regimens generally outperformed chemotherapy-based regimens. PD-(L)1 inhibitors monotherapy ranked highest for overall survival (OS), while chemotherapy plus PD-(L)1 inhibitors and anti-angiogenic therapy (anti-VEGF) ranked highest for progression-free survival (PFS). Chemotherapy plus dual immune checkpoint blockade (PD-(L)1 and CTLA-4) yielded a greater improvement in OS than in PFS. In our real-world cohort, first-line PD-(L)1 monotherapy achieved the best outcomes (objective response rate: 88.9%; median PFS: 22.2 months), followed by chemotherapy plus PD-(L)1 ± anti-VEGF. On multivariable analysis, ECOG performance status 0-1, PD-L1 TPS ≥ 50%, TMB ≥ 10 mut/Mb, and chemotherapy plus PD-(L)1 ± anti-VEGF were independently associated with longer first-line PFS. In subsequent lines, G12C inhibitors significantly improved outcomes versus other therapies, whereas non-G12C disease derived limited benefit from available strategies.

Immunotherapy constitutes the cornerstone of first-line therapy for KRAS-mutant NSCLC. Within this framework, PD-(L)1 monotherapy may be appropriate for carefully selected patients with high PD-L1 expression, whereas chemoimmunotherapy can extend benefit to broader subgroups; escalation with anti-VEGF or anti-CTLA-4 agents appears clinically promising and merits biomarker-driven prospective evaluation. KRAS G12C inhibitors are effective in later lines, while treatment options for non-G12C disease remain limited.