We investigated genomic differences in sequenced biopsies of patients with metastatic cancer to identify associations between somatic alterations and anatomic site of metastasis.

We analyzed data from 20,000 patients treated at Memorial Sloan Kettering Cancer Center, covering >50 cancer types and >20 metastatic sites. Tumors were sequenced using MSK-IMPACT, a targeted sequencing assay that identifies somatic mutations, copy number changes and structural rearrangements in 468 cancer genes using tumor and matched normal DNA. We identified tumor types such as prostate cancer, lung adenocarcinoma, uterine endometrioid carcinoma, head and neck cancer and gastrointestinal stromal tumors, for which the primary biopsies differed significantly from the metastatic biopsies at the genomic level, suggesting a need for sample stratification in clinical studies. We also identified site-specific enrichment of individual somatic alterations for tumors from the same cancer type, including therapeutically relevant targets. For example, when evaluating patients with esophagogastric cancer, we found an enrichment of HER2 amplified tumors, which are candidates for trastuzumab therapy, in sequenced lung metastases as opposed to biopsies sequenced at other metastatic sites (8/16, 50% vs. 20/145, 14%; p<0.001). Looking at sequenced metastases of lung adenocarcinoma (n=468), we found that brain metastases (n=34) exhibit increased aneuploidy (median fraction of genome altered 0.55 vs. 0.44; p<0.001, purity-controlled results) and adrenal gland metastases (n=21) exhibit increased mutational burden (13.3 mut/Mb vs. 6.1 mut/Mb; p<0.001) when compared to metastases sequenced at other anatomic sites. Other examples of novel findings include enrichment of amplifications in the MITF-ETV1 pathway in melanoma patients with brain metastatic disease (13/63, 21% vs. 10/234, 4%; p<0.001) and enrichment of 20q chromosomal amplifications in patients with microsatellite stable colorectal tumors that exhibit oligo-metastatic (as opposed to poly-metastatic) patterns at first time of metastasis diagnosis (158/223, 71% vs. 28/58, 48%; p=0.002).

We provide a detailed analysis of associations between genomic features and metastatic patterns within a clinical sequencing setting. Our work has the potential to directly influence the management of patients with metastatic cancer by enabling clinicians to (a) estimate the risk of future metastasis based on the genomic profile of early stage tumors, (b) choose the appropriate site for biopsy in cases where mutation identification will influence the selection of targeted therapy, (c) optimize the use and frequency of imaging modalities for metastasis surveillance and (d) investigate and functionally characterize novel genomic alterations that have a role in metastasis and can lead to improved therapies.

Citation Format: Francisco Sanchez-Vega, Walid K. Chatila, Karuna Ganesh, Alexander Penson, Niedzica Camacho, Philip Jonsson, Ed Reznik, Jianjiong Gao, Rona Yaeger, Michael F. Berger, Barry S. Taylor, Nikolaus Schultz. Genomic characterization of organ-specific metastasis from prospective clinical sequencing of 20,000 cancer patients [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3024.