Prognostic utility of a personalized ctDNA minimal residual disease test in patients undergoing curative resection of HPV-negative head and neck squamous cell carcinoma.

e18113 Background: Circulating tumor DNA (ctDNA) is emerging as a powerful prognostic biomarker as a minimal residual disease (MRD) test in the postoperative(postop) setting in multiple solid tumors. This study reports an initial experience using a personalized ctDNA assay after curative surgery for head and neck squamous cell cancer (HN-SCC) and its association with disease recurrence. Methods: Patients(pts) who underwent curative surgery for HPV negative HN-SCC who had a postop tumor informed ctDNA (Signatera, Natera) drawn before any adjuvant therapy, were included in this IRB approved study. Pts with stage I-IVB were included and were treated with standard surgical resection with or without postop radiation or chemoradiation as per standard treatment guidelines. Pts with primary or locoregionally recurrent disease were included. Pts were categorized as low risk (pT1-2N0, no adverse risk factors); high risk (positive margins and/or extranodal extension); or intermediate risk (all others). Chi square analysis was used to compare baseline demographics between risk groups. Kaplan-Meier analysis was used to calculate recurrence free survival (RFS) for which any local, regional or distant recurrence was considered an event. Univariate and multivariable analysis (MVA) was performed using Cox proportional hazards regression to identify variables associated with disease recurrence. Results: Of 57 pts included in this study, 38 (67%) were male, 40 (70%) had oral cavity cancer, 44 (77%) had stage IVa/b disease and 43 (75%) presented with primary treatment naive disease. The median age was 67 (range: 44-87), median KPS was 80 and median follow up was 11.8 months (range: 1.7-23.4). Positive postop ctDNA results were observed in 23(40%) pts and negative results were observed in 34(60%). Postop ctDNA positivity correlated with baseline risk grouping and was positive in 0/5 low risk patients, 6/21 intermediate risk pts and 17/31 high risk pts (p=0.015). Of pts with positive postop ctDNA, 13(57%) experienced disease recurrence (9 distant, 2 regional, 1 local, 1 local and distant) with a median time to recurrence of 5.3 months (0.3-11.9). Nine percent of pts with a negative postop ctDNA recurred. The actuarial 1 year RFS was 38.2% (95% CI: 15.5-60.9) for ctDNA positive pts vs. 87.4% (95% CI: 74.0-100) for ctDNA negative pts (p<0.0001). On MVA, a positive postop ctDNA was significantly associated with recurrence (HR 7.52; 95% CI: 1.99-28.6; p=0.003), while high risk grouping compared to low/intermediate risk grouping was not (HR 1.26; 95% CI: 0.38-4.18; p=.70). Conclusions: A positive postoperative ctDNA in pts who undergo curative resection of HN-SCC is a powerful prognostic biomarker that is associated with substantially higher rates of disease recurrence. This finding warrants validation in prospective clinical trials.