Durable Clinical Benefit and Progression-Free Survival in Patients with PD-L1 ≥50% NSCLC Receiving Pembrolizumab: Impact of Nutritional and Inflammatory Markers

• We evaluated nutritional and inflammatory markers in advanced NSCLC with PD-L1 ≥50% treated with pembrolizumab. • Poor performance status was consistently associated with worse outcomes, and higher NLR was associated with lower DCB and shorter PFS in unadjusted analyses. • A composite nutritional–inflammatory score differentiated DCB status at baseline, but should be interpreted as an exploratory host-summary rather than a definitive prognostic index. • Interstitial lung disease did not independently predict pembrolizumab efficacy in this cohort. • Systemic host factors may guide patient selection and optimize immunotherapy outcomes. Immune checkpoint inhibitors (ICIs) are effective in advanced non-small cell lung cancer (NSCLC) with PD-L1 expression ≥50%; however, responses remain heterogeneous. We investigated the predictive value of nutritional and inflammatory indices (Prognostic Nutritional Index PNI, Geriatric Nutritional Risk Index GNRI, and neutrophil-to-lymphocyte ratio NLR), for durable clinical benefit (DCB) and progression-free survival (PFS) in this population. We retrospectively analyzed 125 patients (ECOG PS 0–2) with advanced NSCLC and PD-L1 ≥50% treated with pembrolizumab. A composite score (0–3) was defined by PNI <45, GNRI <92, and NLR ≥5 (1 point each). DCB predictors were evaluated by prespecified multivariable logistic regression (PS, smoking, stage, ILD as forced covariates), and PFS/OS by Kaplan–Meier and multivariable Cox models with landmark OS. DCB was non-progressive disease sustained for ≥6 months (monotherapy) or ≥10 months (chemo-immunotherapy). Among 125 patients, 65.6% achieved DCB. In multivariable logistic regression, PS was independently associated with DCB; the composite score showed a borderline association (p = 0.088), whereas NLR ≥5 was independently associated with lower odds of DCB. PFS was longer with NLR <5, but in adjusted Cox models neither the composite score nor NLR independently predicted PFS/OS; PS remained significant. ILD was not associated with DCB or PFS. In PD-L1–high advanced NSCLC treated with pembrolizumab, PS was the most consistent determinant. Higher inflammation (NLR ≥5) was linked to lower odds of DCB, whereas the composite nutritional–inflammatory score should be regarded as an exploratory host-summary rather than a definitive prognostic index.