Host immune classifier to predict survival with chemoimmunotherapy in PD-L1 ≥50% metastatic NSCLC.

8533 Background: In the 1st line treatment of advanced NSCLC with PD-L ≥50%, immune checkpoint inhibitor (ICI) monotherapy and chemoimmunotherapy have demonstrated survival benefits over chemotherapy alone. However, no additional biomarkers are currently available to guide the choice between these options. Here we report an analysis from the INSIGHT study evaluating the Host Immune Classifier (HIC), a clinically validated blood-based test, to determine its potential to stratify patient survival and optimize treatment selection. Methods: Pre-treatment plasma samples were collected from 271 IIIB/IV NSCLC, 1st line ICI-treated (PD-L1 ≥50%) patients enrolled in the prospective, multicenter observational INSIGHT study (NCT03289780). Samples were analyzed by the HIC test, which categorizes patients as Hot or Cold based on the expression profile of eight proteins measured by MALDI mass spectrometry. Survival outcomes were compared between patients treated with ICI monotherapy (ICI-M) and chemoimmunotherapy (ICI-C), grouped by HIC result. To ensure comparable cohorts, baseline clinical characteristics were balanced using inverse probability weighting (IPW). Overall survival (OS) was evaluated using Kaplan-Meier estimates (95% CI) and the log-rank test. Hazard ratios were calculated with Cox proportional hazards models. Results: Among the analysis cohort, 171 subjects received ICI-M, and 100 received ICI-C. After IPW, baseline characteristics, including age, sex, smoking history, ECOG PS, and histology, were balanced. The distribution of HIC classifications was similar, with 70% Hot and 30% Cold. In HIC-Hot subjects, overall survival (OS) did not differ significantly between treatment groups (log-rank p=0.15, Table 1). In contrast, HIC-Cold subjects had significantly better OS with ICI-C than ICI-M (log-rank p=0.012, Table 1). Multivariate analysis confirmed the HIC test as an independent predictor of OS, adjusting for other prognostic factors. Conclusions: The HIC test is a robust, independent predictor of OS, unaffected by common prognostic factors. HIC-Hot patients had similar OS when treated with ICI-M or ICI-C, suggesting the potential for treatment de-escalation. Conversely, HIC-Cold patients experienced significantly poorer OS with ICI-M but showed improved OS with ICI-C, indicating the need for more aggressive treatment. These findings underscore the potential clinical utility of the HIC test in guiding 1st line ICI treatment strategies for NSCLC with PD-L1 ≥50%. Clinical trial information: NCT03289780 . HIC Test Classification Treatment N 12 Month % OS p value 24 Month % OS p value Median OS (Months) Hazard Ratio (95% CI) p value Hot ICI-M 114 58%(48%, 67%) 0.273 46%(35%, 56%) 0.448 18(11, 31) 0.7 (0.5, 1.1) 0.14 ICI-C 66 66% (52%, 77%) 52%(37%, 65%) Median Not Reached Cold ICI-M 47 26%(14%, 39%) 0.028 10% (1%, 32%) 0.005 3(2, 6) 0.5 (0.3, 0.9) 0.014 ICI-C 28 51%(32%, 67%) 36% (17%, 56%) 13(4, NR*) *Not Reached.