Association of a Serum Proteomic Signature With Survival and Immune-Related Adverse Events in Patients With NSCLC Treated With Immune Checkpoint Inhibitors

Background: Serum proteomic signatures may reflect tumor- and host-related biology and serve as prognostic biomarkers in patients receiving immune checkpoint inhibitors (ICIs). We evaluated the association of the VeriStrat serum proteomic classification with survival outcomes and immune-related adverse events (irAEs) in patients with non-small cell lung cancer (NSCLC) treated with ICIs. Methods: We retrospectively reviewed patients with NSCLC who received ICI-containing therapy and underwent VeriStrat testing at Northwestern Memorial Hospital from October 2015 through June 2023. Patients were classified as proteomic signature Good (PS-Good) or Poor (PS-Poor). Progression-free survival (PFS) and overall survival (OS) were assessed among patients receiving palliative-intent ICI therapy. First any-grade and grade 3 or higher irAEs were evaluated in all ICI-treated patients using cumulative incidence functions and Fine-Gray competing-risk regression. Results: Among 162 ICI-treated patients included in the toxicity analysis, 129 received palliative-intent therapy and were included in the survival analysis; 91 (71%) were PS-Good and 38 (29%) were PS-Poor. PS-Good status was associated with longer PFS (median, 6 vs 3 months; hazard ratio [HR], 0.50; 95% CI, 0.33-0.77; P<0.01) and OS (median, 20 vs 8 months; HR, 0.59; 95% CI, 0.39-0.91; P=0.02). These associations remained significant after multivariable adjustment for PFS (adjusted HR, 0.46; 95% CI, 0.26-0.82; P<0.01) and OS (adjusted HR, 0.50; 95% CI, 0.28-0.87; P=0.01). Any-grade irAEs showed a nonsignificant trend toward a higher cumulative incidence in PS-Good patients. At 12 months, the cumulative incidence was 32.6% for PS-Good versus 22.5% for PS-Poor (subdistribution HR, 1.53; 95% CI, 0.77-3.02; P=0.23). The cumulative incidence of grade 3 or higher irAEs was similar between groups (16.3% vs 15.0%; subdistribution HR, 1.12; 95% CI, 0.48-2.62; P=0.79). Conclusions: PS-Good classification was independently associated with improved survival in patients with NSCLC receiving ICI-containing therapy. Although any-grade irAEs were numerically more frequent among PS-Good patients, proteomic classification was not significantly associated with any-grade or high-grade irAE risk. Prospective validation is warranted.