Feasibility and Clinical Utility of Reporting Hereditary Cancer Predisposition Pathogenic Variants Identified in Research Germline Sequencing: A Prospective Interventional Study

Megan L Hutchcraft,Shulin Zhang,Nan Lin,Justine C Pickarski,Elizabeth A Belcher,Sainan Wei,Thèrése Bocklage,Rachel W Miller,John L Villano,Michael J Cavnar,Joseph Kim,Susanne M Arnold,Frederick R Ueland,Jill M Kolesar,Megan L. Hutchcraft,Justine C. Pickarski,Elizabeth A. Belcher,Rachel W. Miller,John L. Villano,Michael J. Cavnar,Susanne M. Arnold,Frederick R. Ueland,Jill M. Kolesar
DOI: https://doi.org/10.1200/po.23.00266
2024-01-01
JCO Precision Oncology
Abstract:PURPOSE Patients with cancer frequently undergo research-grade germline sequencing but clinically actionable results are not routinely disclosed. The objective of this study is to evaluate the feasibility of reporting clinically relevant secondary findings (SF) identified in germline research sequencing using the institutional molecular tumor board (MTB) and the treating oncology physician. METHODS This prospective, interventional cohort study enrolled Total Cancer Care participants with any cancer diagnosis at a single institution. Patients underwent research-grade germline whole-exome sequencing, with bioinformatic analysis in a Clinical Laboratory Improvement Amendments–certified laboratory to verify pathogenic/likely pathogenic germline variants (PGVs) in any American College of Medical Genomics and Genetics SF v2.0 genes. After a protocol modification in consenting patients, the MTB reported PGVs to treating oncology physicians with recommendations for referral to a licensed genetic counselor and clinical confirmatory testing. RESULTS Of the 781 enrolled participants, 32 (4.1%) harbored cancer predisposition PGVs, 24 (3.1%) were heterozygous carriers of an autosomal recessive cancer predisposition syndrome, and 14 (1.8%) had other hereditary disease PGVs. Guideline-directed testing would have missed 37.5% (12/32) of the inherited cancer predisposition PGVs, which included BRCA1, BRCA2, MSH6, SDHAF2, SDHB, and TP53 variants. Three hundred fifteen participants consented to reporting results; results for all living patients were reported to the clinical team with half referred to a licensed genetic counselor. There was concordance between all research variants identified in patients (n = 9) who underwent clinical confirmatory sequencing. CONCLUSION MTB reporting of research-grade germline sequencing to the clinical oncology team is feasible. Over a third of PGVs identified using a universal testing strategy would have been missed by guideline-based approach, suggesting a role for expanding germline testing.
oncology
What problem does this paper attempt to address?