AUTHORS
Pankaj Pradeep, MS, Olisaemeka Ogbue, MBBS2, Terra Lasho, PhD3, Jenna Fernandez, Ph.D.4, Alejandro Ferrer, PhD5, Nickolas Steinauer, M.D., Ph.D.6, Moritz Binder, MD, MPH7, Christy Finke, BS8, Mrinal M. Patnaik, MD, MBBS9
AFFILIATIONS
1 Mayo Clinic, Rochester, Minnesota, United States
2 Mayo Clinic, Rochester, Minnesota, United States
3 Mayo Clinic, Rochester, Minnesota, United States
4 Mayo Clinic, Rochester, Minnesota, United States
5 Mayo Clinic, Rochester, Minnesota, United States
6 Mayo Clinic, Rochester, Minnesota, United States
7 Mayo Clinic, Rochester, Minnesota, United States
8 Mayo Clinic, Rochester, Minnesota, United States
9 Mayo Clinic, Rochester, Minnesota, United States
ABSTRACT
Introduction
Chronic myelomonocytic leukemia (CMML) is a clonal myeloid neoplasm characterized by persistent monocytosis and overlapping myelodysplastic and myeloproliferative features. Although germline predisposition is increasingly recognized across myeloid neoplasms and has important implications for donor selection, family counseling, and surveillance, its contribution to CMML remains poorly defined. We performed whole-genome sequencing (WGS) in an unselected CMML cohort to define the frequency and spectrum of candidate germline (GL) predisposition variants and explore their associations with clinical features and overall survival.
Methods
Following IRB approval, WGS (DRAGEN v4.3; 30× coverage) was performed on diagnostic bone marrow specimens from 50 CMML patients. Germline variants were identified using GATK HaplotypeCaller with cohort-based joint genotyping, while somatic variants were detected using an ensemble approach incorporating Mutect2, Strelka2 (PMID: 30013048), and DeepVariant (PMID: 30247488) with a custom panel of normals from gnomAD. In the absence of matched constitutional DNA, candidate variants underwent an integrated germline-somatic classification framework incorporating population allele frequencies (gnomAD), COSMIC, ClinVar annotations, variant allele fraction, recurrent myeloid mutational hotspots, ACMG classification, and expert manual review. Clinical records were reviewed for antecedent cytopenias, family history of hematologic malignancy, AML transformation, allogeneic transplantation, and overall survival (OS). Associations between germline predisposition variant status and clinical, cytogenetic, molecular, and outcome variables were evaluated.
Results
Fifty patients with CMML underwent WGS. The median age at diagnosis was 71 years (IQR, 65–78), 70% were male, and median hemoglobin, white blood cell count, platelet count, and absolute monocyte count were 11.6 g/dL, 8.7 × 10?/L, 94 × 10?/L, and 1.7 × 10?/L, respectively. 43 patients (86%) had CMML-1, while abnormal cytogenetics were identified in 9 patients (18%).
Candidate germline variants (median VAF=50.8%) were identified in 14 patients (28%), comprising 17 variants across 11 predisposition genes. Seven patients harbored pathogenic variants in predisposition genes, namely RUNX1 (n=4), GATA2, BRCA2, and CHEK2 (n=1 each). Among all GL variants, including variants of unknown significance, RUNX1 was the most frequently altered gene (n=5), followed by GATA2 and RTEL1 (n=2 each); additional variants involved BRCA2, CHEK2, CTC1, ETV6, MECOM, SAMD9, SAMD9L, and SRP72 (one each).
The most frequently mutated somatic genes in the entire cohort were TET2 (56%), SRSF2 (44%), ASXL1 (25%), KRAS (20%), IDH2 (10%), NRAS, DNMT3A, EZH2 (8% each), followed by CBL, STAG2, PHF6, IDH1, and CEBPA (4% each). RAS pathway mutations were numerically enriched (NRAS 21.4% vs 2.8% P=0.061, KRAS 28.6% 8.3%, P=0.085) in individuals with GL variants. Individuals harboring GL variants were diagnosed at a younger age than noncarriers (median 68.5 vs. 71 years, p=0.087). Antecedent cytopenias and family history of hematologic malignancy did not differ between groups.
AML transformation occurred in 28.6% (4/14) of individuals with GL variants compared with 22.2% (8/36) of noncarriers. Among patients who transformed, carriers developed AML at a younger median age (68.1 vs 72.4 years, P=0.44) and after a shorter interval from CMML diagnosis (6.1 vs 14.3 months, P=0.45), although differences were limited by small sample size. Exploratory gene-specific analyses identified trends toward higher AML transformation among carriers of RTEL1 (2/2, P=0.053), RUNX1 (2/5, P=0.58), GATA2 (1/2, P=0.42), and SAMD9 (1/1, P=0.08) variants. Exploratory gene-specific OS analyses demonstrated inferior survival among carriers of SAMD9 variants (HR=23.82, P=.0096), while BRCA2 carriers showed a similar trend (HR=3.53, P=.23). The frequency of allogeneic hematopoietic cell transplantation was similar between carriers and noncarriers (21.4% vs. 16.7%).
Conclusions
To our knowledge, this represents one of the largest whole-genome sequencing studies evaluating germline predisposition in CMML. Candidate germline predisposition variants were identified in more than one quarter of patients, supporting a potential contribution of inherited susceptibility to CMML biology. These findings support further evaluation of germline testing in CMML. Orthogonal validation using matched constitutional DNA is ongoing.
DISCLOSURES
Pankaj Pradeep, MS: No relevant conflicts of interest to declare.