Early On-Treatment Host Fitness Dynamics Refine Prognostic Stratification in Advanced NSCLC Receiving First-Line Immunotherapy.

PURPOSE: We assessed whether baseline and early on-treatment measurements of host fitness using the Prognostic Nutritional Index (PNI) provide complementary prognostic information in advanced non-small cell lung cancer (NSCLC). MATERIALS AND METHODS: Within the prospective, multicenter OPTIMUS trial, 497 patients with advanced NSCLC receiving first-line pembrolizumab-based therapy were analyzed. PNI was calculated from serum albumin and lymphocyte count at baseline and before cycle 2 (early on-treatment). Using a cutoff of 50.5, 441 patients with paired measurements were classified by baseline/early on-treatment PNI status as high/high, high/low, low/high, or low/low. RESULTS: Low baseline PNI was associated with lower objective response and higher treatment-related mortality. Among responders, low early on-treatment PNI was associated with lower odds of 6-month response maintenance (adjusted OR, 0.40; 95% CI, 0.17-0.97). Response-maintenance rates were lowest in the low/low group at both 6 and 12 months. In multivariable analyses including both PNI time points, low early on-treatment PNI was associated with shorter PFS (HR, 1.50; 95% CI, 1.11-2.04), whereas low baseline PNI was not (HR, 1.27; 95% CI, 0.94-1.71). Both low baseline PNI (HR, 1.53; 95% CI, 1.07-2.19) and low early on-treatment PNI (HR, 1.47; 95% CI, 1.02-2.12) were associated with shorter OS. The higher mortality risk in the low/low group was generally consistent across PD-L1 expression and treatment regimen subgroups. CONCLUSION: PNI assessment at baseline and before cycle 2 provides complementary prognostic information, with early on-treatment PNI particularly related to PFS and response durability (ClinicalTrials.gov Identifier: NCT04909164).