ORLANDO – Older AML patients are not typically offered germline genetic testing to assess their inherited risk for myeloid malignancies. However, such testing may provide an opportunity to manage cancer risk in families, as well as evaluate relatives' suitability for donating stem cells, the results of a study presented at the American Society of Hematology annual meeting suggest.
The study also found a particularly high prevalence of variants of unknown significance, illustrating the need for more efforts to definitively classify the pathogenicity of variants that can be important for treating AML patients.
The analysis was conducted within the Beat AML umbrella trial, where the primary goals are to assess the feasibility of rapid genomic assessments and match older, previously untreated AML patients to investigational therapeutic strategies, and to identify efficacy signals in the different biomarker-informed treatment arms. At the ASH meeting, investigators reported the first data demonstrating the feasibility of a rapid precision medicine approach from Beat AML. In a separate session, researchers led by Uma Borate from Oregon Health & Science University, presented data from a related project within the trial to characterize participants' inherited predisposition to myeloid malignancies.
Inherited risk for hematologic malignancies was historically thought to be relatively rare, however, the increased use of next-generation sequencing has identified more patients with germline genetic mutations that are known to increase cancer risk. Recently published data suggests that between 15 percent and 20 percent of patients with hematologic malignancies have cancer risk predisposition alleles, higher than previously thought.
When cancer risk mutations are identified in patients with hematologic malignancies, this knowledge can inform treatment, and help determine whether family members, who may or may not have inherited the same mutation, can be donors for stem cell transplantation. However, it is not well established in the field when adult AML patients should be tested for inherited gene mutations. Jo-Anne Vergilio, senior director of pathology at Foundation Medicine –– one of the labs that provided testing for this analysis –– noted that patients' physician examinations, medical history, blood counts, and family history can all impact testing decisions.
"Since genomic profiling has become part of standard-of-care testing in these diseases, algorithms are needed to identify probable familial predisposition syndromes in the interpretation of targeted sequencing panels that can then guide further clinical management," she said. "We hope this study can lay the groundwork for establishment of those guidelines.”
Importantly, with their analysis, Borate and colleagues make a case for germline assessment in older adults with AML, even though this has been more of a consideration in pediatrics. "We all think of inherited hematologic malignancies as something that is seen in a younger patient population," said Borate. "But even with [patients] 60 and older this is something we need to take into account clinically," since it could have implications for family members and their ability to be stem cell transplant donors.
The analysis involved 179 Beat AML participants who provided tumor samples from bone marrow or blood, and normal samples from saliva or skin for genetic analysis. Tumor and normal samples were analyzed by Foundation Medicine using its FoundationOne Heme next-generation sequencing test, and the normal samples were analyzed by OHSU on its GeneTrails hematological malignancies NGS panel.
Testing conducted independently by the two labs yielded a 14 percent prevalence of germline pathogenic variants, or 27 mutations in 24 AML patients. The labs classified 181 variants of unknown significance in 102 patients, a prevalence of 57 percent. There were seven variant calls that were discordant between the two labs and were excluded because a consensus determination couldn't be reached.
DDX41 mutations, which have been previously described in older AML patients, were identified in five patients and was the most commonly identified pathogenic germline risk marker in the study. Mutations were also identified in CHEK2, SBDS, MPL, BRCA2, and the Fanconia anemia-associated genes FANCA and FANCL, among others. "There wasn't really a specific pattern that was identified in any of these mutations or the way these mutations presented in patients," Borate said.
However, the vast majority of mutations identified in the analysis were VUS, and the most common were VUS in DOCK8 and CREBBP with a prevalence of more than 5 percent. Although both of these genes have been implicated in leukemias, they aren't yet established as definitely cancer linked. As more studies are done and NGS continues to be used, more VUS will make the switch to pathogenic classification, but for this analysis the authors followed the American College of Medical Genetics and Genomics criteria, and were conservative in their classifications, Borate said.
"Our 14 percent prevalence [for pathogenic variants] is the minimum estimate for this study," she added.
Jane Churpek, a hematologist oncologist and a cancer geneticist at the University of Wisconsin School of Medicine and Public Health, commented following Borate's presentation that the field should proceed cautiously when counseling patients based on germline findings. "All of us are seeing these heterozygous mutations in Fanconi [anemia] genes and others," she said. "When we state the 14 percent prevalence, it's hard because as a cancer geneticist, I wouldn't counsel a patient with a heterozygous FANCL mutation as having a strong leukemia predisposition."
Fanconi anemia is a condition characterized by bone marrow failure and other physical abnormalities, and patients are at greater risk for cancer. Most cases are caused by mutations in the FANC family of genes and patients have to have two mutated copies to get the disease.
As such, Churpek noted that while there might be an association between a heterozygous mutation and cancer, it may not be appropriate to tell carriers that this genomic abnormality is known to cause cancer or that they have a strong predisposition for developing cancer. "We just have to be careful about how we're communicating that level of risk and really proving that those variants are pathogenic and substantially increase chemogenic risk," she said.
In order to better guide the management of AML patients, the VUS rate also needs to be reduced. Borate said she and her colleagues want to further analyze the VUS identified in this preliminary effort, and more definitively classify them through functional analysis and a more extensive family cancer history. There are also opportunities to improve lab techniques for detecting germline pathogenic mutations with minimal presence in tumor and normal samples, she said.
The findings in the study, like others in hematologic malignancies, was limited by the presence of tumor mutations in normal samples. "Every time you take a so-called normal [sample] it's contaminated by tumor cells," Borate said. The investigators compared the median variant allele frequency of somatic mutations in skin and saliva samples from 37 patients, and found that the variant allele frequency was significantly lower in the former (8 percent versus 20 percent). They determined, as other groups have, that saliva may not be a good option when collecting "matched" normal samples from patients at the time of AML diagnosis.
Within the Beat AML study, enrolled patients had a limited family history taken as part of the consent process. This allowed researchers to assess if the finding of a pathogenic germline variant correlated with their self-reported family history of leukemia. Out of 179 patients, 72 percent reported a family history of leukemia, with 12 patients reporting at least one relative with AML. Out of these dozen patients, two had pathogenic mutations identified in this study.
The data from this preliminary analysis, Borate said, indicates that much more research is needed to lower the rate of VUS. Additionally, she noted that the biology of pathogenic alleles need to be defined, for example, when they are inherited in a dominant versus recessive pattern.
During her presentation, Borate noted that this analysis raises the provocative question facing may practitioners in the field: should all putative family members who are considered as stem cell donors be tested for one or more of these inherited risk mutations?
In AML, where patients may receive an infusion of blood-forming stem cells to replace damaged ones, the identification of pathogenic cancer risk mutations can be particularly important for deciding which relatives can be stem cell donors. Siblings are often the top choice in this regard, but other relatives may also be considered. "If we suspect an inherited predisposition to AML, we really want to be careful [about] the donors we're using, particularly if they're family members," Borate said in an interview. "The risk is [the donors] could have the same mutation and pass that on to the patient in these stem cells … [and] that those stem cells may go on to develop the same malignancy."