4q Copy Neutral Loss of Heterozygosity and KMT2A Partial Tandem Duplication Identified By Chromosomal Genomic Array Testing (CGAT) Correlated with Decreased Overall Survival in Acute Myeloid Leukemia (AML)

Background:

CGAT can find copy neutral loss of heterozygosity (cnLOH) and genomic copy number alterations (CNAs) that are undetectable by conventional fluorescence in situ hybridization (FISH) or karyotype. This study investigates whether specific genetic aberrations identified by CGAT are associated with worse event free survival (EFS) or overall survival (OS) in patients with AML.

Methods:

A discovery cohort was formed of adult patients with newly diagnosed AML or high-grade myeloid neoplasm (≥10% myeloid blasts in bone marrow or peripheral blood) who had CGAT performed at Fred Hutchinson Cancer Center from 2012-2022 within 60 days of treatment start. CGAT was performed using CytoScan HD Array, targeting genome-wide region with 2.4 million markers for copy number and 750,000 genotype-able SNPs. Based on prior assay validation in our CLIA-certified diagnostic lab, the cutoff was 100 Kb for CNAs and 10 Mb for cnLOH.

Survival predictive power analysis (SPP), a log-rank based statistical tool within the Nexus Copy Number software, was utilized to identify aberrations in the discovery cohort that were significantly associated with EFS (events: failure to achieve remission, relapse, or death) and OS. SPP considers all aberrations in an unbiased way. Initial SPP calls were manually curated to (1) remove those representing constitutional absence of heterozygosity or overlapping with benign copy number variance, (2) connect adjacent SPP into contiguous regions, and (3) remove those seen in <5 patients. Multivariable analysis was performed to determine the hazard ratio (HR) for EFS and OS, adjusted for patient age, secondary/treatment related AML, ELN 2017 risk, and allogeneic hematopoietic cell transplantation (HCT) as a time-varying covariate.

A validation cohort was formed of patients with de novo AML treated on 7 SWOG trials from 1986-2009. Pretreatment samples were available for patients treated with intensive chemotherapy who survived >28 days after initiation of therapy. CGAT was performed on samples where no clonal abnormalities were detected in at least 20 metaphase cells. Genetic abnormalities identified in the discovery cohort were tested in univariate analysis of EFS and OS (log rank test).

Results:

The discovery cohort consisted of 187 patients, the majority of which had normal karyotype (n=119; 64%). Median age was 63 years (range 20-87). 97 (52%) went on to receive HCT. By ELN 2017 criteria, 169 had intermediate risk, 3 had favorable risk disease, and 15 had adverse risk by karyotype and FISH.

Initial SPP analysis identified 132 CGAT lesions for EFS and 73 for OS. After curation, 5 genomic regions were significantly associated with decreased EFS by univariable analysis: 4q cnLOH (n=5, p=0.002), 5q (EGR1) deletion (n=6, p<0.001), 7q deletion (n=4, p<0.001), 9p cnLOH (n=7, p<0.001), and KMT2A-PTD (n=10, p<0.001). Further, 2 abnormal CGAT findings were significantly associated with decreased OS: 9p cnLOH (p=0.002) and KMT2A-PTD (p=0.04). Multivariable analysis for EFS demonstrated a HR of 3.5 for 4q cnLOH (95%CI 1.4-9.0, p=0.008), 2.3 for 5q (EGR1) deletion (95%CI 0.7-7.3, p=0.16), 4.5 for 7q deletion (95%CI 1.5-13.4, p=0.007), 5.7 for 9p cnLOH (95%CI 2.6-12.7, p<0.001), and 2.9 for KMT2A-PTD (95%CI 1.5-5.6, p=0.002). Multivariable analysis of OS demonstrated a significantly increased risk of death in patients with 9p cnLOH (HR 4.9, 95%CI 1.9-12.5, p=0.001) and KMT2A-PTD (HR 2.3, 95%CI 1.1-5.2, p=0.04); the HRs for the other recurrent abnormalities were not significant.

In the validation cohort (n=202), 5 patients had 4q cnLOH, none had 5q deletion, 2 had 7q deletion, 2 had 9p cnLOH, and 3 had KMT2A-PTD. Given the low number of events, survival analysis was only performed for 4q cnLOH and KMT2A-PTD. 4q cnLOH and KMT2A-PTD were associated with worse OS (p=0.01 and p=0.04, respectively).

Discussion:

This study presents the novel finding that 4q cnLOH and KMT2A-PTD detected by CGAT are associated with worse outcomes in AML patients. One of the notable genes on 4q is TET2, an epigenetic modifier implicated in the pathogenesis of AML. KMT2A-PTD portends more aggressive disease and worse prognosis in AML patients. CGAT is a powerful adjunct to karyotype and FISH. Larger studies are needed to evaluate the effect of recurrent abnormalities by CGAT to better inform the prognosis and management of patients with AML.