Statistical Analyses

The primary objective of this study was to estimate the NPV and PPV for ctHPVDNA-based surveillance to identify patients with recurrence of HPV-associated OPSCC. We reasoned that a blood-based surveillance test would be useful if the NPV was > 90%. With a sample size of 115 patients and an estimated 2-year recurrence rate of approximately 12%,⁸ we anticipated that approximately 100 patients would test negative for the blood test. This sample size yielded 80% power to claim the test was useful (ie, NPV > 90%) at a 1-sided alpha level .05, if the true NPV was 97% or higher. Recurrence was defined as biopsy-proven disease that was diagnosed after the 3-month post-treatment PET/CT. Kaplan-Meier estimates of local control (LC), regional control (RC), local-regional control (LRC), distant metastasis–free survival (DMFS), relapse-free survival (RFS), cause-specific survival (CSS), and overall survival (OS) were calculated relative to the last day of CRT. Statistical significance was assessed using a 2-tailed log-rank test. Ninety-five percent binomial CIs were calculated using the Clopper-Pearson method.