Abstract 6157: Population-specific 5-hydroxymethylcytosine signatures in circulating cell-free DNA and overall survival among Blacks and Whites with multiple myeloma

Bei Wang,Zhou Zhang,Benjamin Derman,Krissana Kowitwanich,Jason Cheung,Andrezej Jakubowiak,Chuan He,Wei Zhang,Brian Chiu

DOI: https://doi.org/10.1158/1538-7445.am2024-6157

IF: 11.2

2024-03-28

Cancer Research

Abstract:Background: Epigenetic modifications play important roles in the prognosis of multiple myeloma (MM). To date, DNA methylation (i.e., 5-methylcytosine) is the most extensively studied cytosine modification. Emerging evidence suggests that 5-hydroxymethylcytosine (5hmC), a product of 5mC through demethylation, in circulating cell-free DNA (cfDNA) is associated with the overall survival (OS) of MM. We also showed that African Americans (AAs) and European Americans (EAs) with MM have distinct genome-wide 5hmC profiles. However, no study has evaluated whether the association between 5hmC profiles and OS in MM differs between EAs and AAs. Methods: A total of 586 newly diagnosed patients with MM at the University of Chicago Medical Center were prospectively enrolled between 2010-2019. The current study profiled genome-wide 5hmC in plasma-derived cfDNA for 303 patients using the 5hmC-Seal and next-generation sequencing. Data cutoff was February 28, 2022. Vital status was ascertained using the National Death Index. OS was defined as the time from diagnosis until death from any cause. The 5hmC-Seal data were mapped to the human genome reference (hg19) and annotated to gene bodies. In each population, patients were randomly divided into a training set (EAs, n=147; AAs, n=64) and a testing set (EAs, n=64; AAs, n=28). After selecting the top 10% candidate genes associated with OS in age-adjusted univariate Cox proportional hazards model, we applied multivariable Cox models with elastic net regularization for feature selection through leave-one-out cross-validation. The optimal "alpha" and "lambda" were evaluated to maximize model performance measured by the Harrell's Concordance Index. The differential 5hmC enrichment levels of selected genes were used to compute a weighted prognostic score (wp-score). The associations between the wp-scores and OS were evaluated in the testing sets, adjusting for established prognostic factors (e.g., age, stage). Results: Of the 303 MM patients, 94 (30.92%) were AA and 210 (69.08%) were EA. The median follow-up was 108 months. Our preliminary analyses identified 26 and 10 5hmC marker genes associated with OS in EAs and AAs, respectively. Importantly, none of these genes overlap between the two groups. We used these genes to compute wp-scores in EAs and AAs. In the training sets, patients with a higher wp-score had worse OS. The same trend was observed in the testing sets: hazard ratios (HR) in EAs: 1.31, 95% CI: (0.83, 2.06) and HR in AA: 11.94, 95% CI: (1.88, 75.68) after controlling for established prognostic factors. Conclusions: Our findings suggest that different epigenetic modifications in cfDNA might be associated with OS between EAs and AAs with MM and the 5hmC marker gene-based wp-scores are independent prognostic factors for MM. Citation Format: Bei Wang, Zhou Zhang, Benjamin Derman, Krissana Kowitwanich, Jason Cheung, Andrezej Jakubowiak, Chuan He, Wei Zhang, Brian Chiu. Population-specific 5-hydroxymethylcytosine signatures in circulating cell-free DNA and overall survival among Blacks and Whites with multiple myeloma [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular s); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl) nr 6157.

oncology