Abstract
Chemotherapy combined with checkpoint blockade antibodies targeting programmed cell death protein (PD-1) has achieved remarkable success in non-small cell lung cancer. However, few patients benefit from long-term treatment. Therefore, biomarkers capable of guiding the optimal therapeutic selection and reducing unnecessary toxicity are of pressing importance. In our research, we gathered serial blood samples from two groups of non-small cell lung cancer patients: 49 patients received a combination of therapies, and 34 patients went under chemotherapy alone. Utilizing non-targeted metabolomic analysis, we examined different metabolites' disparity. Among the lot, L-phenylalanine emerged as a significant prognostic marker in the combination treatment of non-small cell lung cancer patients, interestingly absent in patients under sole chemotherapy. The reduced ratio of L-phenylalanine concentration (two-cycle treatment vs. pre-treatment) was associated with improved progression-free survival (hazard ratio = 1.8000, 95% confidence interval: 0.8566‒3.7820, p < 0.0001) and overall survival (hazard ratio = 1.583, 95% confidence interval: 0.7416‒3.3800, p < 0.005). We further recruited two validation cohorts (cohort 1: 40 patients and cohort 2: 30 patients) to validate the sensitivity and specificity of L-phenylalanine prediction. Our results demonstrate that a model based on L-phenylalanine variations could serve as an early risk-assessment tool for non-small cell lung cancer patients undergoing treatment, potentially facilitating strategic clinical decision-making.