Seven-Gene Score Incorporating Methylation and Expression Distinguishes AML Risk Groups
In a study reported in the Journal of Clinical Oncology, Marcucci et al assessed whether including epigenetic changes—ie, DNA methylation—as molecular risk factors could improve risk stratification in acute myeloid leukemia (AML). They found that a seven-gene score integrating DNA methylation and genetic information identified new risk subsets in patients with cytogenetically normal AML.
Seven-Gene Profile in Training Set
In a training set, next-generation sequencing analysis of methylated DNA identified differentially methylated regions associated with prognostic mutations in 134 previously untreated patients aged ≥ 60 years with cytogenetically normal AML who had received first-line cytarabine/daunorubicin-based therapy in Cancer and Leukemia Group B (CALGB)/Alliance first-line protocols. Seven genes (CD34, RHOC, SCRN1, F2RL1, FAM92A1, MIR155HG, and VWA8) had differentially methylated regions and expression levels associated with improved overall survival (P ≤ .001); high methylation and low expression of each gene was associated with improved survival.
A low-vs-high weighted summary score of the seven-gene expression levels was associated with improved complete response rate (86% vs 43%), 3-year disease-free survival (26% vs 4%), and 3-year overall survival (30% vs 5%; all P < .001). The score remained a significant predictor of complete response rate (P = .01), disease-free survival (P < .001), and overall survival (P < .001) on multivariate analysis adjusting for other clinical and molecular prognostic factors.
Validation of Weighted Score
The weighted seven-gene score was validated in four independent patient groups (n = 355), consisting of younger-patient (< 60 years) and older-patient sets with untreated cytologically normal AML receiving intensive cytarabine/daunorubicin-based therapy in CALGB/Alliance first-line protocols and younger-patient and older-patient sets from the German AML Cooperative Group 1999 trial.
In the CALGB sets, low score was associated with no significant difference in complete response rates (P = .12) but significantly longer disease-free survival (P = .04) and overall survival (P = .02) among older patients and with significantly better complete response rate (P = .004), disease-free survival (P < .001), and overall survival (P < .001) among younger patients.
In the German AML Cooperative Group sets, low score was associated with no significant difference in complete response rate (P = .20) but significantly longer disease-free survival (P < .001) and overall survival (P = .01) among older patients and with significantly better complete response rate (P = .007), disease-free survival (P = .003), and overall survival (P = .001) among younger patients.
Among all patients (n = 481), low scores were associated with significantly better complete response rate (85% vs 58%, P < .001), 3-year disease-free survival (40% vs 15%, P < .001), and 3-year overall survival (48% vs 15%, P < .001). On multivariate analysis, the score remained an independent predictor for complete response rate (P < .001), disease-free survival (P < .001), and overall survival (P < .001). Comparison of the weighted summary score with other molecular prognosticators and previously reported expression profiles using the Akaike information criterion indicated that the weighted score provided a better prognostic model for cytogenetically normal AML than other previously reported profiles.
Unweighted Summary Score
Since the weights in the expression-weighted summary score were nearly identical for all genes, an unweighted summary score using the number of individual genes with high expression was assessed. In the training set, patients with zero or one highly expressed genes had a complete response rate of 96%, 3-year disease-free survival of 32%, and 3-year overall survival of 39% compared with 25%, 0%, and 4% in patients with six or seven highly expressed genes.
In the two older-patient validation sets, patients with zero or one highly expressed genes had complete response rates of 89% and 69% compared with 50% for patients with six or seven highly expressed genes and had 3-year overall survival of 44% and 46% vs 10% and 12%. In the two younger-patient validation sets, patients with zero or one highly expressed genes had complete response rates of 91% and 100% compared with 71% and 53% in patients with six or seven highly expressed genes and had 3-year overall survival of 82% and 76% vs 24% and 7%. The unweighted summary score also had better Akaike information criterion scores than other previously reported gene expression profiles, indicating that it provided a better prognostic model.
The investigators concluded, “A seven-gene score encompassing epigenetic and genetic prognostic information identifies novel AML subsets that are meaningful for treatment guidance.”
Guido Marcucci, MD, of The Ohio State University, is the corresponding author for the Journal of Clinical Oncology article
The study was supported by National Cancer Institute grants, Coleman Leukemia Research Foundation, Deutsche Krebshilfe-Dr Mildred Scheel Cancer Foundation, Pelotonia Fellowship Program, and Conquer Cancer Foundation. The study authors reported no potential conflicts of interest.
The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.
