106P PLR-NIVO: A novel AI-assisted prognostic score in nivolumab-treated metastatic NSCLC

Background:The efficacy of immune checkpoint inhibitors varies considerably among patients due to clinical and biological heterogeneity.This study aimed to evaluate the prognostic value of a novel, practical scoring system, the PLR-NIVO score, developed using an artificial intelligence-assisted approach for predicting progression-free survival (PFS) in patients with non-small cell lung cancer (NSCLC) treated with nivolumab in the second-line setting.Methods: This retrospective study included 100 patients with NSCLC treated with second-line nivolumab.The PLR-NIVO score was constructed using liver metastasis status, PD-L1 expression (0-49% vs 50%), presence of actionable driver mutations, serum lactate dehydrogenase (LDH) level, and platelet-to-lymphocyte ratio (PLR).Artificial intelligence-supported analytical methods, specifically utilizing the ChatGPT-5.1 Pro LLM, were applied alongside existing literature to assist in variable selection and score development.The optimal cut-off value was determined using ROC analysis.Survival outcomes were evaluated using the Kaplan-Meier method, and prognostic factors were assessed with Cox regression analysis.Results: ROC analysis identified an optimal PLR-NIVO cut-off value of 3 for predicting 6-month PFS (AUC: 0.666, p=0.005).Median PFS was 17.4 months in patients with a PLR-NIVO score <3 and 10.2 months in those with a score 3 (p=0.03).Median overall survival (OS) was also longer in the low-score group (28.4 vs 12.6 months), although the difference did not reach statistical significance (p=0.07).In multivariable Cox regression analysis, a high PLR-NIVO score remained an independent prognostic factor for shorter PFS (HR: 2.4, 95% CI: 1.04-5.36,p=0.03).The incidence of any treatment-related adverse event was 32.9% in the low-risk group compared with 10% in the high-risk group (p=0.024). Conclusions:The PLR-NIVO score, developed using an AI-assisted analytic approach, is an independent and clinically applicable prognostic tool for predicting progressionfree survival in patients with NSCLC receiving second-line nivolumab.This scoring system may support risk stratification in patients undergoing immunotherapy.