Identification of Causal Plasma Proteins and Immune Mediators for Lung Cancer in Never-Smokers: A Large-Scale...

As global smoking rates decline, Lung Cancer in Never-Smokers (LCINS) has emerged as a distinct and increasingly important clinical entity. However, its underlying pathogenesis remains poorly understood, highlighting the need to identify circulating molecular determinants and candidate biomarkers.

Genetic instruments for plasma proteins were obtained from pQTL datasets (Fenland study and UK Biobank Pharma Proteomics Project), while summary statistics for LCINS were sourced from TRICL-ILCCO. Two-sample Mendelian Randomization (MR) was performed using inverse-variance weighting as the primary approach, complemented by sensitivity analyses for heterogeneity (Cochran's Q), horizontal pleiotropy (MR-Egger intercept), and outlier detection (MR-PRESSO). Tissue-level validation was conducted using CPTAC proteomic data. Further analyses, including Bayesian colocalization, Protein-Protein Interaction (PPI) network construction, Gene Ontology (GO) enrichment, and mediation analysis, were applied to investigate shared genetic architecture, functional pathways, and immune-mediated mechanisms.

Among 543 cis-acting plasma proteins evaluated in the discovery phase, 35 exhibited evidence of association with LCINS risk, including 15 positively and 20 inversely associated proteins. After False Discovery Rate (FDR) correction, only genetically proxied plasma RARRES1 remained statistically significant (OR = 1.17, adjusted P = 2.72 × 10-5), and this finding was replicated in an independent cohort. Bayesian colocalization supported a shared causal variant (PPH4 = 0.83). Functional enrichment analyses implicated RARRES1-related networks in immune regulation and extracellular processes. Two-step MR mediation analysis nominated CD16-CD56 intensity on NKT cells as a potential immune-mediated pathway, accounting for an estimated 11.8% of the genetic effect.

This study prioritizes genetically proxied plasma RARRES1 as a candidate protein associated with LCINS risk and highlights CD16-CD56 intensity on NKT cells as a potential immune-related mechanism. These findings provide biologically plausible evidence supporting a role for RARRES1 in LCINS pathogenesis and its potential utility in biomarker development.

Genetically proxied plasma RARRES1 is associated with increased LCINS risk and may contribute to disease susceptibility through immune-related alterations involving NKT-cell phenotypes. RARRES1 represents a promising candidate for future mechanistic studies and blood-based diagnostic biomarker development in LCINS.