5-Hydroxymethylcytosines in circulating cell-free DNA and overall survival in patients with multiple myeloma.

Abstract

8032 Background: The epigenetic mark 5-methylcytosines (5mC) have been associated with poor prognosis and survival in multiple myeloma (MM), but the prognostic role of 5-hydroxymethylcytosines (5hmC) as marks of tissue-specific enhancers generated from 5mC through active demethylation is unknown. We showed that 5hmC can be profiled in circulating cell-free DNA (cfDNA) and is associated with relapse/death in another lymphoproliferative disorder diffuse large B-cell lymphoma. To date, no study has investigated genome-wide 5hmC profiles in cfDNA for its prognostic significance in MM. Methods: A total of 354 newly diagnosed MM patients at the University of Chicago Medical Center were prospectively enrolled between 2010-2017. Blood samples were collected at the time of diagnosis. Patients were followed through 31 December 2020 (avg. follow-up = 77.8 mths). We collected baseline clinical, laboratory, and cytogenetic data from electronic medical records. Vital status was ascertained in 351 of the 354 patients (deaths = 73) using the National Death Index. We profiled genome-wide 5hmC in cfDNA using the 5hmC-Seal and next-generation sequencing. The 5hmC-Seal data were mapped to the human genome reference (hg19) and annotated to gene bodies. Overall survival (OS) was defined as time from diagnosis until death from any cause. We used Cox proportional hazards model and the elastic net regularization to identify genes with modified 5hmC levels that are associated with OS. Patients were randomly divided into a training set (n = 264) and testing set (n = 87). Results: The median age at diagnosis was 61.8 years and 47% (n = 165) were males. We used the differential 5hmC enrichment levels of a preliminary four-gene marker panel (i.e., YPEL1, VIPR2, PLAC8L1, and CYP2D6) to compute a weighted prognostic score (wp-score). In the training set (deaths = 55), MM patients with high wp-score had worse OS (Hazard Ratio [HR] = 2.2, 95% Confidence Interval [CI]: 1.3-3.9; p = 0.004). The same trend was observed in the testing set (deaths = 18) (HR = 3.5, 95% CI: 1.1-10.6; p = 0.02). The 5hmC-based wp-score remained significantly associated with OS after controlling for standard prognostic factors, suggesting that 5hmC-based wp-score for this 4-gene panel is an independent prognostic factor for MM. We also explored population-specific 5hmC and wp-score. We found that 5hmC profiles in cfDNA differ between Blacks (n = 117) and Whites (n = 234). In addition, 5hmC marker genes associated with OS differ between Blacks (13 genes) and Whites (20 genes). Conclusions: Our results suggest that 5hmC in cfDNA at the time of diagnosis correlate with OS in MM and the 5hmC marker genes associated with OS differ between Blacks and Whites. These findings suggest that a plasma-derived cfDNA 5hmC-modified gene panel holds promise as a noninvasive approach for predicting prognosis in MM and may be integrated in clinical practice to improve precision care.