Multi-Institutional Study Evaluating the Role of Circulating Tumor DNA in the Management of Appendiceal Cancers

PURPOSE Conventional surveillance methods are poorly sensitive for monitoring appendiceal cancers (AC). This study investigated the utility of circulating tumor DNA (ctDNA) in evaluating systemic therapy response and recurrence after surgery for AC. METHODS Patients from two specialized centers who underwent tumor-informed ctDNA testing (Signatera) were evaluated to determine the association between systemic therapy and ctDNA detection. In addition, the accuracy of ctDNA detection during surveillance for the diagnosis of recurrence after complete cytoreductive surgery (CRS) for grade 2-3 ACs with peritoneal metastases (PM) was investigated. RESULTS In this cohort of 94 patients with AC, most had grade 2-3 tumors (84.0%) and PM (84.0%). Fifty patients completed the assay in the presence of identifiable disease, among which ctDNA was detected in 4 of 7 (57.1%), 10 of 16 (62.5%), and 19 of 27 (70.4%) patients with grade 1, 2, and 3 diseases, respectively. Patients who had recently received systemic chemotherapy had ctDNA detected less frequently (7 of 16 43.8% v 26 of 34 76.5%; odds ratio, 0.22 95% CI, 0.06 to 0.82; P = .02). Among 36 patients with complete CRS for grade 2-3 AC-PM, 16 (44.4%) developed recurrence (median follow-up, 19.6 months). ctDNA detection was associated with shorter recurrence-free survival (median 11.3 months v not reached; hazard ratio, 14.1 95% CI, 1.7 to 113.8; P = .01) and showed high accuracy for the detection of recurrence (sensitivity 93.8%, specificity 85.0%). ctDNA was more sensitive than carcinoembryonic antigen (62.5%), CA19-9 (25.0%), and CA125 (18.8%) and was the only elevated biomarker in four (25%) patients with recurrence. CONCLUSION This study revealed a reduced ctDNA detection frequency after systemic therapy and accurate recurrence assessment after CRS. These findings underscore the role of ctDNA as a predictive and prognostic biomarker for grade 2-3 AC-PM management.