Tumor Purity as a Prognostic and Predictive Biomarker of Postoperative Radiation Therapy Outcomes in Stage IIIA-N2 Non-Small Cell Lung Cancer: A Transcriptomic Analysis From the Lung ART Trial.

Abstract

Tumor purity (TP), the proportion of malignant cells within a tumor sample, is an important feature of the tumor microenvironment. Using transcriptomic data from the Lung ART-IFCT 0503 trial, we investigated the relevance of TP and its potential to predict benefit from postoperative radiation therapy (PORT). RNA sequencing was successfully performed on 285 samples. TP was inferred using the ESTIMATE algorithm. Associations with overall survival (OS) and disease-free survival (DFS) were assessed using Kaplan-Meier and multivariable Cox models. Among 285 patients with resected stage IIIA-N2 non-small cell lung cancer, 144 received PORT. The median age was 61 years, 31% were women, and 80% had nonsquamous histology. Baseline characteristics were well balanced between arms. The median TP was 0.64 (range, 0.41-0.92) and was slightly higher in the PORT arm (0.65 vs 0.63; P = .006). TP correlated with H&E pathologist-estimated cellularity (r = 0.23, P < .001), was higher in squamous tumors (0.68 vs 0.63, P < .001), and increased with necrosis (r = 0.31, P < 10^-6). Transcriptomic analysis confirmed associations with proliferation-related pathways and reduced hypoxia signatures (false discovery rate < 0.001). TP inversely correlated with T-cell immune infiltration by immunohistochemistry (CD3⁺ r = -0.52; CD8⁺ r = -0.45). High-TP was associated with worse OS (48.5 vs 106.5 months, P < .001) and DFS (18.4 vs 48.0 months, P = .017). A TP × PORT interaction was observed for OS (P = .049) and a trend for DFS (P = .07). TP reflects proliferative and tumor microenvironment features; a lower TP is independently associated with improved prognosis in resected stage IIIA-N2 non-small cell lung cancer. PORT may be more effective in tumors with lower TP; however, this finding is exploratory and requires independent validation.