The role of serial analysis of hepatocellular carcinoma via circulating tumor DNA in identification of new actionable alterations

478 Background: Molecular profiling using circulating tumor DNA (ctDNA) is advantageous in cancers where tissue biopsies are not routinely obtained such as in hepatocellular carcinoma (HCC). We hypothesize that due to the expected evolution of cancer clones during treatment, serial analysis of ctDNA at the time of progression may identify additional actionable alterations. This study analyzed serial testing patterns and results for patients with HCC in a large laboratory database. Methods: Analysis of genomic results from blood samples prospectively collected between January 1, 2016-March 31, 2021 for clinical Guardant360 testing and a diagnosis of HCC as documented by ordering providers on the test requisition form (TRF) was performed. Serial tests were defined as having Guardant360 analysis performed on more than one sample collected at more than one timepoint. Chi-squared analysis was completed to assess associations between maximum variant allele fraction (maxVAF) changes and identification of new genes altered on subsequent tests. Results: 106 patients were tested serially with Guardant360 with a total of 272 tests. The median age of patients tested was 65 years (range: 16-91); 75% were male. Approximately 73% of patients had 2 serial tests, 95% had 2-4 serial tests; 5% had 5 or more. 105/106 (99%) of patients had serial tests with genomic alterations detected across multiple tests, of which 68 (64.8%) had new pathogenic alterations compared to their previous test(s) (Table). Increases in maxVAF at the time of re-testing were associated with findings of new molecular alterations (p=0.03). Potentially actionable molecular alterations accounted for 19% (13/68) of new alterations and included mutations in KIT and BRAF and amplifications in MET, BRAF, RAF1, PDGFRA, KIT, and FGFR2. Conclusions: Serial ctDNA testing in patients with HCC identified numerous potentially actionable alterations in nearly 1 in 5 patients for which a clinical trial or matched treatment option may be available. Non-invasive longitudinal genomic assessment via ctDNA provides an opportunity to examine these trends in a larger cohorts and assess impact on clinical outcomes for patients with HCC who undergo serial testing.[Table: see text]