Abstract 1681: Prospective clinical sequencing distinguishes independent primary cancers from recurrent or metastatic disease

Abstract

Abstract Background: Standard histopathology and immunophenotyping of cancer are sometimes insufficient to distinguish independent primary cancers from recurrent/metastatic disease. Accurate classification is essential when making treatment decisions. We hypothesized that broad NGS, which has increasingly been adopted to inform treatment in patients with advanced cancer, also helps discriminate between multiple primary tumors and recurrent/metastatic disease. Methods: We analyzed results from prospective NGS (‘MSK-IMPACT’) performed at Memorial Sloan Kettering between 2014-2018 involving >25,000 samples. Patients where MSK-IMPACT had been performed on more than one sample were classified based on 1) histologic determination of cancer type, and 2) the genomic similarity of the tumors. Genomic similarity was defined as having at least one shared somatic mutation or >20% shared copy number segment alterations. Tumors with purity <20% based on pathologic estimation were excluded. Cases were manually reviewed to determine whether the genomic data was useful in classifying tumors and how these results ultimately influenced patient care. Results: We identified 41 cases where tumor sequencing was utilized to help determine the clonal relationship between either synchronous or metachronous tumors. In five patients with NSCLC, NGS confirmed that multiple lesions were separate cancers as opposed to metastatic or multifocal disease. This determination changed management in two patients who were recommended surveillance rather than adjuvant chemotherapy based on this finding. In contrast, another patient did not receive NGS results prior to starting treatment and could have avoided chemotherapy. Two patients with stage I NSCLC were thought to have recurrent disease but NGS demonstrated de novo metastases from a second NSCLC. This is notable because first-line treatment of metastatic NSCLC is dictated by tumor genomics. Four patients were thought to have recurrent cancer but genomics revealed two unrelated tumors. Three patients were treated definitively with localized therapy for a presumed new primary cancer but genomics later suggested metastasis from a synchronous (N=1) or prior (N=2) tumor. In several additional cases, molecular profiling was useful in supporting what was suggested to be the same (N=19) or different (N=10) tumors histologically. Conclusion: Broad NGS can be utilized to distinguish independent malignancies from recurrent/metastatic disease, including cases where standard work-up is inconclusive. We identified multiple instances where clonal relationships were used to guide treatment, including whether to pursue potentially curative versus palliative therapy. These data demonstrate a potential additional use of NGS, which to date has largely been reserved for aiding treatment decisions in patients with advanced incurable cancer. Citation Format: Alison M. Schram, Chai Bandlamudi, Ahmet Zehir, Marc Ladanyi, Michael F. Berger, Barry S. Taylor, David M. Hyman. Prospective clinical sequencing distinguishes independent primary cancers from recurrent or metastatic disease [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1681.