INTRODUCTION
Advent of next-generation sequencing and large-scale analysis of mutations in cancer have shown that germline mutations in cancer-predisposing genes are more common than previously thought and found in a variable fraction from 5 to 20% of cancer patients, with different prevalence among cancer types. Germline variants are often scored also in sporadic cases of cancer with no familial history. To date, approximately 100 cancer-predisposing genes have been associated to hereditary syndromes. The PanCancer analysis scored 8% frequency of pathogenic and likely pathogenic germline mutations in a large cohort of 10,389 patients with 33 different cancer types. Some cancer types show a significant association with mutations in well-known cancer predisposing genes, i.e. ovarian cancer in BRCA1/2 , while new associations were revealed, such as loss of function mutations in BUB1B in lung cancer and in SDHA in melanoma. These data suggest that a relevant number of cancer-predisposing genes and their phenotypic consequences are yet to be identified.
Investigation of an increased risk for cancer development has several important clinical implications, both for carriers of the pathogenic variants and their relatives. For cancer patients, identification of a germline pathogenic variant can yield critical information about prognosis and direct therapeutic choices, both in terms of efficiency and toxicity of chemiotherapeutic/radiotherapeutic regimens and of surgical intervention strategies. Concerning other family members, identification of pathogenic germline variants in cancer-predisposing genes may enhance targeted cancer surveillance and improve cancer prevention.
One of the biggest challenges in this field is the classification of the identified germline variants, for two main reasons: i) most are classified as variants of unknown significance (VUS) and, therefore, have not been assigned a defined biological role nor have they been attributed to particular disease phenotypes. Consequently, such variants have no relevance for clinical applications. ii) each cancer-predisposing gene may confer different cancer risks, ranging from high to low penetrance (probability) of cancer development, depending both on cancer type and type of variants affecting the genes.
Acute myeloid leukemia (AML) is among the malignancies with the lowest frequency of germline variants (4.2% in PanCancer) and traditionally it has been among the less investigated cancer types from this point of view. However, germline predisposition of hematological neoplasms is increasingly assessed and, in recent years, advent of molecular testing has led to identification of specific hereditary hematological syndromes . The current guidelines from the European LeukemiaNet (ELN) for diagnosis and management of AML in adult patients recommend investigation of germline predisposition. This is particularly critical for clinical management of AML patients that often receive allogeneic stem-cell transplantation to consolidate disease remission. Considering that usually relatives are the best donors, it is critical to exclude the presence of hematopoietic stem cells harboring pathogenic variants from the donor before transplantation. The list of cancer predisposing genes in hematological neoplasms is likely to be largely incomplete and we are in need for further investigations in order to uncover other players of hematological hereditability and define the penetrance of the identified variants.
We designed a comprehensive gene panel for the assessment of germline pathogenic variants in hematological neoplasms, covering the entire coding sequence of 256 genes. The employment of large gene panels enlarges the testing potential compared to genetic screens restricted to high penetrance genes, allowing discovery of more potential predisposing variants; however, it also poses the challenge of assigning a specific functional/phenotypic consequence to newly identified VUS. Variants can also be reclassified when more information on gene function and/or data from family history become available over time, based on new studies. We studied 4 families with two siblings each affected by myeloid neoplasms in order to exploit the potential of family segregation studies in this context.