Utility of multi-gene panel-based germline analysis following genomic profiling and cascade testing in advanced solid tumors: An initial report of the BRANCH study.

Abstract

10615 Background: Genetic testing for pathogenic germline variants (PGVs) is performed to identify hereditary cancer syndromes (HCS) and to evaluate the application of targeted therapies in clinical oncology. Multi-gene panels (MGPs) have become popular for identifying HCS; however, their clinical utility in therapeutic applications and confirmatory germline testing for presumed PGVs and in cascade testing are not well established. Methods: BRANCH [UMIN000046085] is a multicenter clinical study to confirm PGVs in: i) patients with advanced solid tumors in whom a presumed PGV has been detected by comprehensive genomic profiling (CGP) (cohort A); ii) patients who received a germline BRCA (gBRCA) test (cohort D); and iii) first-degree relatives of patients with PGVs identified in this study (cohort C). PGVs were examined in 35 HCS genes using an MGP combined with digital multiple-ligation-dependent probe amplification. Results: MGP results were available for 75 patients in cohort A and 83 in cohort D (3 February 2023). The most common cancer in cohort A was colorectal cancer (13%) followed by ovarian (12%) and breast (11%) cancers. Of 103 tested presumed PGVs, 46 (45%) were confirmed as PGVs, with the highest germline conversion rates in BRCA1 (8/8, 100%), followed by PALB2 (5/6, 83%) and BRCA2 (15/19, 79%) among the genes tested in ≥3 patients. In tissue-based CGP, true PGVs had a higher variant allelic frequency (median 55.7% [range, 31.9-96.8]) than non-PGVs (37.0% [9.3-86.9]). In cohort D, 61 (74%) patients had breast cancer, including one with both breast and ovarian cancers. All BRCA PGVs identified by gBRCA test were identified by MGP in 28 BRCA-positive patients. Intriguingly, among 55 patients (66%) in whom no BRCA PGV was detected by gBRCA test, non-BRCA PGVs were identified in 6 (11%) patients (2 PALB2 and 1 each of ATM, CHEK2, MSH2 [large deletion], and TP53). Thirty relatives in 18 pedigrees received MGP-based cascade testing in cohort C. The most frequent target genes were BRCA2 followed by BRCA1 and PALB2. The same PGVs as the probands were confirmed in 11 (37%) individuals and novel PGVs were identified in 2 (7%) relatives, leading to double PGVs in homologous recombination genes ( PALB2+ BRCA2 and BRCA2+ ATM), respectively . Conclusions: MGP successfully determined germline variants among presumed PGVs identified by CGP in patients with advanced solid tumors, and detected non-BRCA PGVs in 10% of gBRCA-negative patients. Furthermore, MGP-based cascade testing has the potential to identify other PGVs in addition to those detected in the probands. Clinical trial information: UMIN000046085 .