Cyclin-dependent kinase 6 body methylation and association with prognosis of patients with acute myeloid leukemia.

Abstract

e19000 Background: Cyclin-dependent kinase 6 (CDK6) plays critical roles in leukemogenesis and leukemia stem cells activation, and its inhibition has emerged as a novel therapeutic strategy in acute myeloid leukemia (AML) management. However, association between CDK6 methylation and prognosis in patients with AML remains unclear. Methods: Methylation profile, somatic alterations, and CDK6 gene expression data of 200 clinically annotated patients with de novo AML were retrieved from the TCGA Pan-Cancer Atlas project deposited in the UCSC Xena database. Specifically, methylation levels of all 64 CpGs across CDK6 were extracted, and then methylation abundances of promoter and gene body regions were calculated as average levels of CpGs located in the respective regions. Patients were categorized into two subgroups (hypo- and hyper-methylated) as per the median methylation level for either region or single CpG site. Log-rank tests were conducted to compare overall survival (OS) of patients stratified by methylation status. A Cox proportional hazards model was fitted to control confounding factors, including age, white blood cell count, cytogenetic risk classification (favorable, intermediate, and poor risk), and TP53 mutations, as reported previously. Associations of methylation status with cytogenetic risk and somatic alterations were examined using the Wilcoxon and Kruskal-Wallis rank sum tests, respectively. Spearman rank correlation coefficient ( R) was used to evaluate correlations between methylation and expression. Results: Of the 200 patients, 109 (54.5%) were men, and median (interquartile range [IQR]) age was 57.5 (44.75-67) years. Median (IQR) methylations of promoter and body were 5.2% (4.8%-5.7%) and 55.6% (53.7%-57.9%), respectively. Body methylation was observed to correlate with gene expression ( R, -0.22; P = .005), but not promoter methylation. In Kaplan-Meier analysis, low methylation of body rather than promoter predicted prolonged OS (median, 27.4 vs 9.2 months; hazard ratio [HR], 0.43; 95% CI, 0.29-0.63; P < .001). Notably, the OS improvement persisted with multivariable Cox model adopted (HR, 0.64; 95% CI, 0.42-0.98; P = .038). In addition, the body hypomethylation was strongly associated with early age, favorable cytogenetic risk, mutations in TP53, NPM1, and IDH1/ 2, as well as PML- RARA and MYH11- CBFB fusions (all P < .05). Further analysis identified that cg08300570, cg09653641, cg10481072, and cg13590558 in the body were all prognostic for OS (all P < .05). Intriguingly, all these CpGs resides in the enhancer. Conclusions: This is the first study to unravel clinical significance of CDK6 body methylation in AML, and it may serve as an independent prognostic marker. Given that azacitidine and decitabine are widely used, our findings may have implications for combination therapy of CDK4/6 inhibitors with hypomethylating agents in AML. Prospective studies are warranted.