Performance and clinical utility of germline genetic testing criteria for predisposition to myeloid neoplasms in adults

Abstract

Abstract Introduction Hereditary hematopoietic malignancies (HHMs) are increasingly recognized as clinically relevant entities due to their implications for treatment, donor selection, and family counselling. Additionally, associated genes may cause non-malignant hematologic or extra-hematologic features, such as thrombocytopenia or pulmonary fibrosis. Despite their clinical impact, HHMs remain underdiagnosed, often missed due to phenotypic variability, incomplete family histories, and the lack of standardized testing criteria. To address this, a multidisciplinary team in Catalonia developed standardized germline testing criteria, later endorsed by the Spanish Society of Hematology. We evaluated their implementation and diagnostic yield in a multicentric real-world cohort. Methods The Catalan-Balearic Group of HHMs launched a prospective multicenter registry in October 2022 to collect clinical and genetic data from individuals undergoing germline testing based on predefined criteria. Criteria were grouped into: (1) Patient Criteria – (a) HM or aplastic anemia (AA) diagnosed <50 years; (b) therapy-related myeloid neoplasms with TP53-related tumors; (c) hypoplastic myelodysplastic neoplasm; (d) suggestive clinical features (e.g., thrombocytopenia, dysmorphisms, cytopenias); (e) family history of hematological malignancies (HM) or suggestive features; (f) suspicious somatic variants in genes known to be associated with HHM with VAF >30% for single nucleotides variants or >20% for insertion/deletion (based on ESMO Precision Medicine Working Group recommendations, Kuzbari 2023). (2) Unaffected Individual Criteria – (a) suggestive clinical features; (b) positive family history when the affected relative was unavailable for testing. Pre-test genetic counselling and informed consent were integral components of the process. Testing was performed using NGS panels or whole-exome sequencing. Germline status was confirmed using non-hematopoietic tissues. Data were managed via REDCap. Results A total of 198 individuals meeting at least one criterion were enrolled. Most were male (58%) and of European ancestry (84%). Hematological alterations were present in 190 (96%): 154 with HM, 17 with AA, and 19 with clonal cytopenia of unknown significance. Median age at diagnosis was 52.5, 27.0 and 32.0 years, respectively. Germline variants were detected in 79/198 individuals (39.9%), of which 60 (30.3%) were classified as pathogenic or likely pathogenic (P/LP). Most frequent affected genes were DDX41 (n=26, 43%), RUNX1 (n=8, 13%), and RTEL1 (n=5, 8%). Other genes mutated included ETV6, GATA2, ELANE, NAF1, ERCC6L2 (biallelic), CSF3R, TP53, TINF2, CHEK2, ATM, and TERT. The most frequently met criteria were HM/AA <50 years (n=88) and germline-suspicious VAF (n=87), followed by family history (n=32), and suggestive features in affected (n=18) or unaffected (n=16) individuals. Highest diagnostic yields were observed in HM/AA individuals with suggestive features (we found a causative germline variant in 61%, 11/18), unaffected individuals with suggestive features (44%, 7/16), and those with high VAF somatic variants (41%, 36/87). HM/AA <50 years yielded 22% (19/88), and family history showed a 25% yield (8/32). Notably, all germline variants initially identified through tumor sequencing had VAF >35%, with the lowest confirmed germline VAF at 39% (DDX41). These findings, though limited, suggest that establishing the threshold at ≥35%, instead of ≥30%, may improve prioritization of variants likely to be germline in HM, as current ESMO guidelines are largely based on solid tumor data. In total, germline findings impacted management in several cases, including HSCT donor selection (n=15), modified treatment strategies, and extra-hematologic surveillance. Cascade testing identified 67 P/LP carriers among 153 relatives (44%). Conclusions Standardized criteria enabled the identification of germline variants in nearly one-third of tested individuals, higher than the established 10–15% in solid tumors. Criteria incorporating clinical features or suspicious somatic variants showed the highest yield. Our findings suggest that raising the VAF threshold to ≥35% may improve germline variant prioritization in HM. Implementing structured criteria for HHM detection advances precision hematology by facilitating personalized care, guiding donor selection, treatment decisions, and enabling at-risk family identification.