denotes an abstract that is clinically relevant.
denotes that this is a recommended PHD Trainee Session.
denotes that this is a ticketed session.Session: 803. Emerging Tools, Techniques and Artificial Intelligence in Hematology: Poster II
Hematology Disease Topics & Pathways:
Research, assays, Clinical Research, Diseases, real-world evidence, Lymphoid Malignancies, Myeloid Malignancies, Technology and Procedures, molecular testing
Approximately 7-10% of cancer patients have pathogenic germline variants (PGVs) in cancer predisposition genes. However, most studies to date have focused on PGVs in patients with solid tumors. Less is known about the prevalence of PGVs in patients with hematologic malignancies (HM). We investigated the prevalence of PGVs in 72 cancer predisposition genes in a cohort of HM cases from our tumor-normal whole exome sequencing (TN-WES) database.
Methods
We identified 363 unique HM patients with TN-WES performed at our lab (Sema4, Branford, CT) since 2020. Our analysis pipeline calls germline pathogenic/likely pathogenic variants (PGVs) in 72 cancer-predisposing genes (PGV-72) from normal tissue, predominantly buccal samples. These are reported clinically as secondary findings, with the patient’s consent.
Results
The cohort (N=363) was comprised predominantly of patients with lymphoid malignancies (N=299, 82.4%), most commonly myeloma (N=222) and CLL/SLL (N=27). The most common myeloid malignancy was myelodysplastic syndrome (N=25).
Of 363 unique HM patients with TN-WES, 33 (9.1%) were found to have PGVs in one of the PGV-72 genes. There was no significant difference between those with lymphoid and myeloid malignancies (8.0% vs. 14.1%, Fisher’s exact p=0.15). Among patients with germline findings (N=33), the median age was 67 years (range 23-84). The most commonly mutated genes were APC (N=7), CHEK2 (N=7), MUTYH (N=6), and LZTR1 (N=4). LZTR1 and ATM PGVs were more common in patients with myeloid versus lymphoid malignancies (LZTR1: 4.7% vs 0.3%, p=0.02; ATM: 3.1% vs. 0%, p=0.03).
Several founder mutations were identified: all APC variants were p.I1307K; four CHEK2 variants were p.T367Mfs*15 (c.100delC); five MUTYH variants were p.G396D. Two patients had more than one PGV: one patient with myeloma (APC p.I1307K and CHEK2 p.S428F) and one patient with CLL/SLL (CHEK2 p.T367Mfs*15 and HOXB13 p.G84E). One patient with myeloma had a BRCA2 p.S1982Rfs*22 germline variant. All PGVs were heterozygous, and none were associated with a somatic second hit.
Conclusions
The prevalence of PGVs in 72 cancer predisposition genes was approximately 9% in a set of patients with HM undergoing TN-WES. Most PGVs identified were associated with moderately increased risk for solid tumors. Many of the PGVs identified were founder mutations in APC, CHEK2, and MUTYH, likely reflecting the racial/ethnic composition of our cohort.
Disclosures: Al-Kateb: Sema4: Current Employment. Jun: Sema4: Current Employment, Current equity holder in publicly-traded company. Hayes: Sema4: Current Employment. Rossi: Sema4: Current Employment. Hantash: Sema4: Current Employment. Oh: GSK, Janssen, Merck, Pfizer: Consultancy; Sema4: Current Employment, Current equity holder in publicly-traded company. Onel: Sema4: Current Employment, Current equity holder in publicly-traded company.