Detectable Postoperative Circulating Tumor Human Papillomavirus DNA and Association with Recurrence in Patients With HPV-Associated Oropharyngeal Squamous Cell Carcinoma

Purpose: The purpose of our study is to determine the rate of detectability of ctHPVDNA after surgery but before adjuvant therapy in patients with human papillomavirus (HPV)-associated oropharyngeal squamous cell carcinoma (HPV[+]OPSCC) and to investigate whether detectable ctHPVDNA at this time point may be associated with risk of recurrence. Methods and Materials: We examined prospectively collected samples from patients with OPSCC in a blinded fashion using a multianalyte polymerase chain reaction assay. We collected 45 samples from patients with HPV(+)OPSCC preop (before any treatment) and 159 samples postop (before or at the start of adjuvant radiation therapy). We identified samples via the radiation oncology biobank or via participation in a clinical trial. Radiation therapy consisted of 60 Gy ± cisplatin or de-escalation (30 Gy to 36 Gy in 20 bid fractions + docetaxel). For our preliminary analysis, 32 patients had paired samples available pre- and postop. We performed additional exploratory analyses including associations of patient and tumor characteristics with recurrence using Cox proportional hazards models for all 159 postop samples. We compared detectability of ctHPVDNA across groups using logistic regression. We used Kaplan-Meier to estimate recurrence-free survival. Results: In a paired analysis of 32 pre- and postop timepoints, 94% of patients had detectable ctHPVDNA preop and 41% did postop. Recurrence-free survival at 18 months was 83% (95% confidence interval, 47%-95%) for patients with detectable postop ctHPVDNA compared with 100% for patients with undetectable postop ctHPVDNA (P = .094). In an exploratory analysis of nonpaired postop samples, ctHPVDNA was detectable in 26% of patients (41 of 159) (median of 22 days postop). Age (odds ratio,1.06, P = .025), lymphovascular space invasion (odds ratio, 3.17, P = .011) and extranodal extension (odds ratio = 5.67, P = .001) were associated with detectable ctHPVDNA after surgery. Detectable postop ctHPVDNA was significantly associated with recurrence-free survival (P < .001). Conclusion: Among patients with detectable preop ctHPVDNA, a significant proportion have detectable postop ctHPVDNA in paired postop samples collected before the initiation of adjuvant radiation therapy. Future prospective study is warranted to investigate the association of detectable postop ctHPVDNA with recurrence, including in comparison to established clinical and pathologic risk factors.