Genomic comparison of matched primary and metastatic germ cell tumors (GCT).

Abstract

4556 Background: Tumor genomic analysis may be useful in patients with GCT as a means of identifying potentially actionable genomic alterations or mutations such as TP53 that confer resistance to chemotherapy. As GCTs often exhibit significant morphologic heterogeneity, we evaluated the level of concordance between genomic alterations in matched primary and metastatic GCT samples. Methods: GCT patients enrolled on an institutional prospective sequencing protocol with available primary and metastatic tumor tissue were eligible. Each tumor was subjected to MSK-IMPACT, an exon capture sequencing assay, which detects copy number alterations (CNAs) and mutations in 410 cancer-related genes. For each primary-metastasis pair, concordance and clonality was assessed using the FACETS algorithm. Results: Matched primary-metastasis tumor pairs were available for 36 patients (78% nonseminoma, 22% seminoma, median age 33.5 years). All patients received chemotherapy, with 25 (69%) receiving treatment prior to analysis of the metastatic samples. The frequency of genetic alterations was low with a median of 3 mutations (1-7), 7 amplifications (1-26) and 1 deletion (1-9) detected per sample, with no significant difference in mutational/CNA burden between primaries and metastases. Of 109 unique mutations across patients, only 44 (40%) were concordant between the primary and matched metastasis, including 5 of 9 hotspot mutations. For CNAs, 184 (81%) of 226 were concordant. Only 24 of 109 (22%) mutations were clonal (defined as predicted to be present in all cancer cells) in either the primary or metastatic matched samples; of these, only 4 were clonal in both the primary and metastatic samples, including 2 hotspots. However, 4 of 5 alterations in TP53/MDM2 were shared by both the primary and metastasis pairs. In a separate exploratory cohort, 4 TP53 mutations were identified in 3 primary tumors and 1 metastasis, and all 4 mutations were also detected by cell-free DNA profiling. Conclusions: Genomic concordance, particularly for mutations, is poor between primary and metastatic GCT samples. Cell-free DNA analysis may help overcome this limitation by identifying alterations in progressive tumors without need for a new biopsy.