Genetic factors and gene expression patterns are recognized predictors of susceptibility to chronic obstructive pulmonary disease (COPD). Nevertheless, the extent to which these elements shape the diverse clinical presentations of COPD is not well established. This investigation aimed to identify high-risk COPD subtypes by combining genetic risk assessment via polygenic risk score (PRS) with blood-based transcriptional risk score (TRS), and to examine variations in their associated clinical and molecular profiles. High-risk categories were constructed using quantiles derived from PRS and TRS distributions, selecting thresholds that produced the greatest separation in protein biomarker levels within the COPDGene training dataset. These definitions were then applied to independent COPDGene and ECLIPSE validation cohorts. Associations between subgroup membership and clinical endpoints were analyzed using multivariable models. Protein–protein interaction networks and potential drug repurposing strategies were also contrasted across the high-risk categories. Analyses focused on two omics-derived high-risk groups within separate test cohorts (n = 1133 non-Hispanic White participants from COPDGene, n = 299 African American participants from COPDGene, and n = 468 from ECLIPSE). The subgroups were labeled “high activity” (low PRS combined with high TRS) and “severe risk” (high PRS combined with high TRS). Relative to a low-risk reference group (low PRS and low TRS), participants in both high-risk subgroups had reduced body-mass index (BMI), impaired lung function, and dysregulated pathways involved in metabolism, growth, and immune responses. The “high activity” subgroup alone exhibited more rapid longitudinal decline in FEV1 (COPDGene: −51 mL/year; ECLIPSE: −40 mL/year). Proteomic signatures in this subgroup were notably enriched for gene sets influenced by 5-lipoxygenase inhibitors and angiotensin-converting enzyme (ACE) inhibitors. Considering polygenic and transcriptional risk scores together delineated clinically and molecularly distinct subsets within the high-risk COPD population. Proteomic characterization and drug-repurposing modeling highlighted therapy enrichments that differed by subtype, providing a plausible explanation for why earlier drug-repurposing initiatives may have failed due to insufficient consideration of patient heterogeneity.
