Routine Screening for Germline Variants in Patients With AML

Nearly 14% of adult patients with acute myeloid leukemia (AML) have pathogenic or likely pathogenic germline variants that make them susceptible to myeloid malignancies, according to a study in Blood. The researchers suggest these findings support the need for routine screening of patients with AML for pathogenic germline variants.

Corresponding study author Anupriya Agarwal, PhD, of Oregon Health & Science University, explained that patients with AML who harbor pathogenic or likely pathogenic germline variants are largely representative of those seen in a typical adult leukemia clinic and not preselected for their genetic and family history. “Thus, our study demands the routine screening for germline variants in these patients across all age spectrums as well as comprehensive and careful evaluation of family history as part of routine clinical practice,” she said.

Dr. Agarwal and colleagues collected 391 tumor samples from patients with AML and matched skin biopsies. The researchers also assessed patient electronic medical records to obtain data on clinical, prognostic, pathologic, cytogenetic, and molecular genetics. Additionally, the researchers reviewed past medical and family histories of a total of 324 patients with AML for evidence of familial clustering of hematologic malignancies in first-, second-, and third-degree relatives.

At diagnosis, the median age of the 391 patients with AML was 62 years. Approximately 59% of the patients were at least 60 years of age, but 13% were younger than 40 years.

Using the 2015 American College of Medical Genetics and Genomics (ACMG) guidelines for variant interpretation, the investigators curated up to 1,547 unique variants from 228 genes. Based on the available data, the researchers identified pathogenic or likely pathogenic germline variants in 13.6% of patients with AML (n=53) in a total of 34 genes. Approximately 41% of these variants were found in DNA damage response genes, with the most frequently mutated genes identified in CHEK2 (n=8) and DDX41 (n=7).

Around 44% of the identified pathogenic germline variants were found in genes deemed “clinically actionable,” defined by current guidelines. In addition, the investigators reported that they found pathogenic germline variants in new candidate genes, including DNAH5, DNAH9, DNMT3A, and SUZ12.

There was no strong association between the rate of germline mutations and the age of AML onset. In patients who reported a history of at least one family member with a hematologic malignancy (n=49), six were found to harbor a known pathogenic or likely pathogenic germline variant, while the other patients had one or more variants of uncertain significance. According to the researchers, these findings suggest a need for further functional validation studies.

Several variants of unknown significance (VUS) were observed in most patients who had reported family history of hematological malignancies. “This underscores the constant need for

performing functional studies in relevant models of those variants to support their pathogenicity,” said Dr. Agarwal. “Family segregation analysis can yield powerful data to re-classify a VUS, but obtaining samples from family members for variant analysis has been a challenge for a retrospective study like this. This again suggests we need to change the clinical practice for these patients going forward.”

Using CHEK2 as an example, the investigators demonstrated that three-dimensional protein modeling may be an effective modality to identify variants of unknown significance that warrant further work up with functional studies. The researchers added that they “evaluated an in-silico approach that applies ACMG/AMP curation in an automated manner using the tool for assessment and prioritization in exome studies, which can minimize manual curation time for variants.”

The researchers noted that this study was limited by intragenic insertions/deletions and copy number variations, non-assessed epigenetic events, as well as alterations in the non-coding regions.

Study authors report no relevant conflicts of interest.

Reference

Yang F, Long N, Anekpuritanang T, et al. Identification and prioritization of myeloid malignancy germline variants in a large cohort of adult AML patients [published online ahead of print, 2021 Sep 5]. Blood. doi: 10.1182/blood.2021011354.

To answer the perennial question posed by cancer patients – “Why me?” – this study surmised the germline status of deleterious variants in 228 cancer genes among 391 adults diagnosed with AML, without requiring syndromic features or positive family history, identifying 53 (~14%) with germline predisposition alleles. However, their calculation included 19 heterozygous variants from conditions considered autosomal recessive, which, by definition, requires bi-allelic mutations. Therefore, re-calculating a more conservative estimate of germline inheritance for genes known to confer susceptibility to hematopoietic malignancies,1 including autosomal dominant alleles, heterozygous alleles that confer risk, and individuals with syndromic features, one estimates a germline frequency of at least 5% (21/391): DDX41 (7), GATA2 (1), TERT (1), PTPN11 (1), ATM (2), POT1 (1), and CHEK2 (8). Importantly, these germline mutations were present across the entire age spectrum, challenging the widely held view that germline predisposition causes cancer exclusively in young people.

These findings should prompt quick change to our clinical practice, given the increasing use of allogeneic stem cell transplantation for consolidation therapy in older individuals and the preference for familial donors. The timing of germline predisposition testing becomes critical. Dr. Yang and colleagues compared paired AML and skin biopsy samples to determine variants’ germline status, taking DNA directly from the skin sample after washing out most hematopoietic cells.

Future studies comparing this method to the standard of culturing skin fibroblasts and performing family segregation and functional testing will refine the estimate of germline predisposition frequency and determine if culturing is necessary to avoid confusion from contaminating leukemia cells, which are often at high levels at diagnosis.

Currently, determination of deleterious germline alleles is recommended for young individuals diagnosed with myelodysplastic syndromes; aplastic anemia; two or more cancers, one of which is a hematopoietic malignancy; and positive family history.2,3,4 However, clinical germline testing is reasonable for any condition with a pre-test probability of >5%. Even using a conservative estimate of 5% as the frequency of germline alleles, incorporating germline predisposition testing into the standard AML diagnostic workup is necessary.5

Lucy Godley, MD, PhD
University of Chicago

References

  1. Trottier AM, Godley LA. Inherited predisposition to haematopoietic malignancies: overcoming barriers and exploring opportunities. Br J Haematol. 2021;194(4):663-676.
  2. Feurstein S, Churpek JE, Walsh T, et al. Germline variants drive myelodysplastic syndrome in young adults. Leukemia. 2021;35(8):2439-2444.
  3. Fox LC, Wood EM, Ritchie DS, Blombery P. Diagnostic evaluation and considerations in hypocellular bone marrow failure-A focus on genomics. Int J Lab Hematol. 2020;42 Suppl 1:82-89.
  4. Singhal D, Hahn CN, Feurstein S, et al. Targeted gene panels identify a high frequency of pathogenic germline variants in patients diagnosed with a hematological malignancy and at least one other independent cancer [published online ahead of print, 2021 Apr 13]. Leukemia. doi: 10.1038/s41375-021-01246-w.
  5. Liu YL, Stadler ZK. The Future of Parallel Tumor and Germline Genetic Testing: Is There a Role for All Patients With Cancer? J Natl Compr Canc Netw. 2021;19(7):871-878.