Circulating HPV DNA Kinetics and Clinical Outcomes in a Large Cohort of Radiotherapy-Treated p16-Positive Oropharyngeal Cancers

Abstract

<h3>Purpose/Objective(s)</h3> Human papillomavirus (HPV) driven oropharyngeal cancer (OPC) has heterogeneous outcomes with a 5-year recurrence rate of up to 30% after definitive (chemo-)radiotherapy (CRT/RT). Circulating HPV DNA (ctDNA) is a convenient dynamic biomarker that can be repeatedly measured for risk stratification and response monitoring. However, the understanding of HPV ctDNA kinetics during treatment and its association with clinical and treatment factors is nascent. We characterized HPV ctDNA kinetics in the largest-to-date cohort of RT- or CRT-treated OPC. <h3>Materials/Methods</h3> Among 262 patients with p16 positive, stage I-III (TNM-8) OPC treated with standard of care curative-intent RT (n=107) or CRT (n=155), blood samples were collected at baseline (pre-treatment), week 4 of RT (mid-treatment), and 3-month follow-up (post-treatment). Early-treatment (end of week 1) blood draws were also performed in a subset of patients (n=92) with detectable baseline ctDNA. Copies of HPV-16 ctDNA were quantified using a previously validated multiplex droplet digital PCR assay and normalized to plasma volume. Primary outcome was recurrence-free survival (RFS) and secondary outcomes were progression-free survival (PFS) and overall survival (OS). <h3>Results</h3> With median follow up of 42.0 months (range: 3.5-77.5), 3-year RFS was 88%. Pre-treatment HPV ctDNA was detected in 220 (84%) patients and ranged from 0.32 to 2.4 × 10^5 copies/mL in plasma. Baseline ctDNA level was moderately associated with T category (median copies/mL: 350 in T1-2 and 730 in T3-4, p = 0.011) and strongly associated with N category (N0: 83, N1: 357, N2: 609, N3: 5903, p < 0.001). Among patients with detectable pre-treatment ctDNA, clearance (undetectable ctDNA) occurred in 3.5%, 14.3%, and 90.7% of early-, mid-, and post-treatment samples, respectively. Clearance at each timepoint was consistently associated with longer RFS, PFS, and OS, though the study was only powered to detect significant differences at post-treatment. 3-year RFS, PFS, and OS were 91.0%, 88.5%, and 92.1% respectively if ctDNA was cleared post-treatment, versus 62.2%, 50.6%, and 56.0% if ctDNA was not cleared. As a continuous variable, log-transformed ctDNA levels were independently prognostic of RFS (adjusted for disease stage) at early-treatment (Cox proportional hazards HR: 1.37, 95% CI: 1.01-1.86, p = 0.046) and post-treatment (HR: 1.30, 95% CI: 1.08-1.56, p = 0.0047), but not at pre- or mid-treatment. An increase (peak) in ctDNA level from pre-treatment to early-treatment was associated with inferior RFS (p=0.18), PFS (p=0.03), and OS (p=0.03). <h3>Conclusion</h3> Detectability of HPV ctDNA during and post-treatment is prognostic of recurrence and survival in RT-treated p16+ OPC. Quantitative ctDNA levels at early-treatment and post-treatment are prognostic, independent of disease stage. Early-treatment ctDNA kinetics show promise as a potential dynamic biomarker of treatment response.