Germline DNA Repair Gene Mutations and Clonal Hematopoiesis (CH) in 24,849 Patients with BRCA-Associated Cancers
by
, , , , ,
Catherine H. Marshall
1
,
Ali T. Arafa
2,
Ellen Jaeger
3,
Stamatina Fragkogianni
3,
Anne Sonnenschein
3,
Elizabeth Mauer
3,
Lukasz P. Gondek
1,
Calvin Chao
3,
Jun Luo
4 and
Emmanuel S. Antonarakis
5,*
1
Sidney Kimmel Comprehensive Cancer Center, Department of Oncology, Johns Hopkins School of Medicine, Baltimore, MD 21287, USA
2
Department of Pharmacology, University of Minnesota, Minneapolis, MN 55455, USA
3
Tempus AI Inc., Chicago, IL 60654, USA
4
Department of Urology, James Buchanan Brady Urological Institute, Johns Hopkins School of Medicine, Baltimore, MD 21287, USA
5
Department of Medicine, University of Minnesota Masonic Cancer Center, Minneapolis, MN 55455, USA
*
Author to whom correspondence should be addressed.
Cancers 2025, 17(9), 1432; https://doi.org/10.3390/cancers17091432
Submission received: 25 February 2025
/
Revised: 4 April 2025
/
Accepted: 7 April 2025
/
Published: 25 April 2025
(This article belongs to the Collection Oncology: State-of-the-Art Research in the USA)
Simple Summary
Clonal hematopoiesis (CH) is the expansion of white blood cells with mutations in genes associated with hematologic malignancies found in patients without evidence of any hematologic malignancy. CH is associated with increasing age and exposure to certain cancer therapies. We hypothesized that having a germline cancer predisposition gene mutation might be associated with an increased risk of CH. We used the deidentified Tempus dataset to answer this question. Among patients with breast cancer, we did see an association with increased risk of CH; this was not seen among those with prostate, ovarian, and pancreatic cancer.
Abstract
Purpose: To determine if the risk of clonal hematopoiesis (CH) would be higher among those with germline alterations in homologous recombination repair genes (gHRR) in the four BRCA-associated cancers (breast, ovarian, prostate, pancreas) compared to those without inherited predisposition (the sporadic group). Methods: We retrospectively analyzed deidentified data from 24,849 patient samples from the Tempus database with a primary diagnosis of breast, ovarian, prostate, and pancreatic cancers. Germline pathogenic or likely pathogenic variants in BRCA1, BRCA2, ATM, PALB2, and CHEK2 were identified across all four cancer types. CH was determined based on the presence of pathogenic or likely pathogenic alterations in any one of 52 CH-associated genes with a variant allele fraction of at least 2% found in the normal match. Age-adjusted odds ratios were calculated for risk of CH across cancer types. Results: CH was identified in 14% of patients with BRCA-associated cancers. DNMT3A, PPM1D, and TET2 were the most common CH gene alterations. After adjusting for age at time of biopsy, having any germline alteration in the breast cancer cohort was associated with a 41% increased likelihood of CH (OR 1.41; 95% CI 1.07–1.84, p = 0.014). An increase in CH prevalence was not seen in the three other cancer types. Conclusions: When accounting for age at time of testing, pathogenic germline alterations in DNA repair genes were associated with an increased risk of CH only among patients with breast cancer, but not in those with ovarian, pancreatic, or prostate cancers.
