Abstract Lung cancer’s high mortality is driven by advanced-stage diagnoses and high recurrence rates. Existing genomic markers lack sensitivity for early detection and predicting recurrence, presenting ongoing clinical challenges. We introduce the Boston Lung Trans-omics (BOLT) Initiative, an early-stage sub-cohort of the Boston Lung Cancer Study with 1446 paired tumor/adjacent-normal bulk RNA-sequenced tissues from 723 NSCLC patients. We performed paired differential gene expression analysis between tumor and adjacent normal tissues using limma-voom, adjusting for batch effects and patient ID to account for inter-individual confounding. Statistically significant differentially expressed genes were defined as genes with FDR-adjusted p0.05 and log2 change |LFC|2. To emphasize biological relevance and minimize false positives, we performed pathway enrichment analysis with the Gene-set Enrichment Analysis software using curated pathway data retrieved from the Molecular Signature Database, Broad Institute. We then applied weighted gene co-expression network analysis (WGCNA) to the top 5000 most variable genes across all samples to identify gene modules and assessed module-trait associations. We identified 738 significant DEGs in the paired tumor/normal comparison. Among these, 387 (52%) were upregulated in tumor compared to normal lung tissues. AGER of the receptor for advanced glycation end-products (RAGE) pathway was the top downregulated gene in tumor compared to adjacent normal tissues (FDR=3.3*10-156; LFC=-4.4) while SPP1 of the PI3K/mTORC1 signaling and glycolysis pathways was the top upregulated gene in tumor compare to normal lung tissues (FDR=3.6*10-134; LFC=4.8). Pathway enrichment analysis revealed that genes upregulated in tumors compared to normal tissues were most strongly associated with MYC target activation, MTORC1 signaling, glycolysis, and inflammatory responses. Modules defined by WGCNA recapitulated tumor/normal distinctions, suggesting system-level gene programs underlying early tumorigenesis.These preliminary insights from BOLT reveal consistent early transcriptional differences between NSCLC tumors and adjacent normal tissues characterized by upregulation of both novel and known oncogenic pathways and immune/inflammatory responses. Co-expression networks support the tumor/normal status as the principal driver of gene modules at early stages of NSCLC. Modeling time to recurrence remains a critical yet understudied area in lung cancer research. We will develop and validate predictive gene expression signatures and co-expression networks in BOLT to stratify early-stage NSCLC patients by recurrence risk and anticipated time to recurrence following primary treatment or neoadjuvant therapies. Citation Format: Yu Chen Zhao, Li Su, Lorelei A. Mucci, Timothy R. Rebbeck, Yi Li, David C. Christiani. Uncovering early transcriptional alterations in lung cancer: Preliminary insights from the Boston Lung Trans-omics (BOLT) Initiative abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1102.
Zhao et al. (Fri,) studied this question.