Feed

Environmental exposures are upstream determinants of molecular variation, yet exposure-to-biomarker gradients remain insufficiently quantified in harmonized multiomics cancer cohorts. Using TCGA lung adenocarcinoma as a model, we evaluated cumulative tobacco exposure and smoking history as determinants of variant-derived biomarkers and tested whether integrating clinical, genomic, transcriptomic, and proteomic data improves identification of high tumor mutational burden (TMB). This secondary analysis used public cBioPortal-linked TCGA data. Among 522 patients with clinical and molecular annotations, 302 ever-smokers with nonmissing pack-years, TMB, and covariates comprised the primary adjusted logistic-regression cohort, and 250 cases had matched multiomics data for prediction modeling. Higher cumulative tobacco exposure was associated with greater odds of high TMB: Compared with pack-year tertile 1, tertile 3 had an adjusted odds ratio of 2.28 (95% CI: 1.18, 4.41; p trend = 0.013), and each 10 pack-year increment was associated with an odds ratio of 1.12 (95% CI: 1.02, 1.23). Smoking categories also showed strong gradients for TMB, total nonsynonymous mutation counts, and C > A substitution fraction. Current smoking was positively associated with TP53 mutation and inversely associated with EGFR mutation relative to never-smoking. High-TMB tumors showed 201 differentially expressed transcripts and 15 differentially abundant proteins. Driver-augmented models discriminated high TMB better than broader multiomics models, although integrated scores retained prognostic relevance. These findings support exposure-aware biomarker development in lung adenocarcinoma. High TMB was defined as a cohort-specific top-quartile analytic endpoint rather than a universal clinical threshold. The findings support biomarker interpretation and hypothesis generation, not direct immunotherapy-response prediction or a clinically deployable model.