Abstract 774: Methylation-derived biologic age, telomere length, and lung cancer mortality in heavy smokers from CARET

Abstract

Epigenetic alterations, including changes in DNA methylation, and telomere length (TL) are established molecular markers of aging, morbidity, and mortality. We hypothesized that older methylation-derived biologic age (meAge) measured prior to lung cancer diagnoses would be associated with increased post-diagnosis mortality in high risk heavy smokers. Our study included 372 lung cancer cases from the β-Carotene and Retinol Efficacy Trial (CARET), a unique population of heavy smokers at high risk for lung cancer. Genome-wide DNA methylation data (Illumina MethylationEPIC) and TL were assayed in blood collected on average 4.9 years prior to diagnosis. We estimated meAge using scores developed by Hannum (leukocytes; 65/71 CpGs), Horvath (pan tissue; 335/353 CpGs), and Levine (DNAm PhenoAge, whole blood; 513 CpGs). We assessed Pearson correlations between meAge, age at blood draw, and log2-transformed TL, and we defined meAge acceleration (ΔmeAge) as the difference between each meAge and age. We examined associations between ΔmeAge, with and without TL adjustment, and all-cause mortality after lung cancer diagnoses using Cox proportional hazards models adjusted for age, sex, smoking status, pack years, time between blood draw and diagnosis, and methylation-estimated blood cell counts, including a strata variable for early (I/II), late (III/IV), or unknown stage. We assessed mortality in those with both old meAge and short TL (dichotomized at the medians) compared to those with both young meAge and long TL. We modeled all cases and by histotype (adenocarcinoma N=148, squamous cell N=115, small cell N=81). In our study population, mean age (64.3) was underestimated by Hannum and Levine meAges (52.4 and 57.0, respectively) and overestimated by Horvath meAge (70.3). Age was positively correlated with meAges (Hannum r=0.55, P=1E-30; Horvath r=0.49, P=2E-24; Levine r=0.41, P=2E-16), as were meAges with each other (all r≥0.66, P≤3E-48). TL was negatively correlated with age (r=-0.13, P=0.01) and meAges (Hannum r=-0.20, P=0.0001; Horvath r=-0.17, P=0.001; Levine r=-0.15, P=0.003). While each year of age was associated with 3% higher mortality in all cases (95% confidence interval (CI): 1.01-1.06), we did not observe an independent association with ΔmeAge (Hannum 0.99, 0.97-1.01; Horvath 1.00, 0.99-1.02; Levine 1.00, 0.99-1.01) regardless of TL adjustment or histotype. Shorter TL was associated with increased mortality in those who developed small cell lung cancer, as we reported previously; here, we observed that this association was independent of ΔmeAge. Combined old meAge and short TL was not associated with mortality. Though prospectively collected ΔmeAge has been reported to predict all-cause mortality in population-based studies, we did not observe that pre-diagnosis ΔmeAge, nor combined meAge-TL, were associated with mortality in a group of heavy smokers who went on to develop lung cancer. Citation Format: Laurie Grieshober, Stefan Graw, Matt J. Barnett, Mark D. Thornquist, Gary E. Goodman, Chu Chen, Devin C. Koestler, Carmen J. Marsit, Jennifer A. Doherty. Methylation-derived biologic age, telomere length, and lung cancer mortality in heavy smokers from CARET [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 774.