mRNA-based profile to predict platin sensitivity in NSCLC assessed by Nanostring: A translational analysis of the ETOP/EORTC SPLENDOUR trial.

Abstract

e20555 Background: Platinum-based chemotherapy is a standard treatment for non-small cell lung cancer (NSCLC), yet clinical benefit is highly variable and no validated biomarkers are available to guide treatment selection. The Platin Drug Response Predictor (Platin-DRP) is a 205-gene mRNA signature previously validated in breast cancer and NSCLC using microarray-based platforms, including an analysis performed in the SPLENDOUR trial cohort using Affymetrix gene expression profiling and reported separately (abstract submitted to ELCC 2026). Here, we evaluated the clinical association of Platin-DRP with outcomes in patients from the randomized phase 3 SPLENDOUR trial, using tumor biopsies analyzed on the Nanostring platform (UIN: researchregistry11317). Methods: SPLENDOUR was a randomized trial enrolling 514 patients with advanced NSCLC treated with platinum chemotherapy with or without denosumab. This secondary, exploratory analysis included 93 patients with available tumor biopsies and tumor cell content ≥5% (cisplatin n = 33, carboplatin n = 60). The tumor biopsies were analyzed on the Nanostring platform, and the gene expression values were used to calculate platin DRP scores. Associations between the continuous DRP score and progression-free survival (PFS) and overall survival (OS) were assessed using Cox proportional hazards models at a one-sided (1s) type I error of 5%. Additional analyses evaluated pre-specified DRP thresholds (35, 50, 70) and response rate (RR). Results: In the analyzed cohort of 93 patients, the median PFS and OS were 3.9 and 5.6 months, respectively. Higher DRP scores were significantly associated with longer OS in the carboplatin cohort (n = 60; HR (per 50-point DRP increase) = 0.65; 95% upper confidence limit (uCL) 0.930; 1s p = 0.024) and in the pooled cohort stratified by chemotherapy regimen (n = 93; HR (per 50-point DRP increase) = 0.72; 95% uCL 0.957; 1s p = 0.029). In the cisplatin cohort, the association between DRP score and OS was not statistically significant (n = 33; HR (per 50-point DRP increase) = 0.87; 95% uCL 1.399; 1s p = 0.32). Threshold-based analyses showed improved OS in carboplatin-treated patients with high DRP scores, with a median OS of 16.9 months for scores > 70 (n = 16) versus 4.6 months for scores ≤70 (n = 44; 1s p = 0.045). Moreover, DRP scores derived from Affymetrix and Nanostring platforms were strongly correlated (Spearman ρ = 0.91, p < 0.001). Conclusions: Using a Nanostring-based assay, higher Platin-DRP scores were associated with improved OS, but not PFS, in platinum-treated advanced NSCLC patients. These findings suggest cross-platform robustness of the DRP signal and potential clinical relevance for guidance in NSCLC.