Tumor/normal genomic profiling in patients with metastatic solid tumors identifies pathogenic germline variants of therapeutic importance.

Abstract

1501 Background: Tumor molecular profiling via next-generation sequencing (NGS) is routinely utilized to direct patients toward clinical trials of targeted therapeutics. NGS testing of paired tumor/normal samples identifies incidental pathogenic germline variants (PGVs), having potential implications for patients and their families. Methods: From 2011-2018, 1,015 patients with metastatic, refractory solid tumors underwent targeted (1700 genes) exome and transcriptome sequencing of matched tumor/normal samples through the Michigan Oncology Sequencing program. Identified PGVs that conferred increased cancer risk or were associated with certain autosomal recessive conditions were reported to the treating oncologist. Chart reviews were conducted every 3 months to assess whether PGV identification impacted treatment decision making. Results: 169 PGVs were identified in 160 unique patients (15.8% of cohort). 69 PGVs (41%) harbored a clear somatic second hit event in the tumor. PGVs associated with defects in double-strand DNA repair ( BRCA1, BRCA2, ATM, PALB2, BRIP1) or DNA mismatch repair ( MLH1, MSH2 and PMS2) were identified in 49 patients (5% of cohort, 31% of patients with PGVs), 37 of which had not previously been identified. 14 PGVs in DNA double-strand repair and 7 PGVs in DNA mismatch repair were identified in cancer types not commonly associated with hereditary breast ovarian cancer or Lynch syndromes, including cancers of unknown primary origin and sarcomas. 7 patients received a PARP inhibitor (PARPi), 3 patients received an immune checkpoint inhibitor (ICI) and 1 patient received both PARPi and ICI therapy on the basis of a PGV in DNA repair. 6 patients achieved clinical benefit, defined as time on treatment ≥ 6 months. A patient with cancer of unknown primary origin and PGV in MSH2 achieved exceptional response to ICI therapy, with complete response ongoing and lasting 23 months. Conclusions: Targeted NGS of matched tumor/normal samples identified PGVs in about 1 in every 6 patients with metastatic solid tumors. Approximately 40% of PGVs are associated with a somatic second hit in the tumor, supporting their role in tumor pathogenesis. Unexpected PGVs with therapeutic implications are identified in patients with diverse cancer types, providing opportunities to use targeted therapies with potential for significant clinical benefit. Given this finding, testing for PGVs in DNA repair genes should be considered in all patients with metastatic solid tumor malignancies.