Surgical Drain Fluid as a Novel Liquid Biopsy Analyte to Measure Postoperative Minimal Residual Disease in HPV+ Oropharyngeal Cancer

Abstract

<h3>Purpose/Objective(s)</h3> Postoperative risk stratification of patients undergoing surgery for HPV⁺ oropharyngeal squamous cell carcinoma (OPSCC) is controversial. While pathologic staging and extranodal extension (ENE) are established criteria for adjuvant radiotherapy+/-chemotherapy in HPV-negative head and neck cancer, their prognostic significance in HPV⁺ OPSCC is less clear. Moreover, pathologic interpretation can be variable, especially in the community setting. We propose an objective measure of postoperative minimal residual disease (MRD) using HPV DNA in surgical drain fluid (SDF). Unlike plasma, this novel liquid biopsy analyte provides a proximal, site-specific measure of locoregional MRD. We hypothesize that elevated HPV viral load in SDF predicts the presence of established pathological features crucial to adjuvant radiotherapy decision-making. <h3>Materials/Methods</h3> Thirty-two SDF specimens were collected from the surgical wound bed of 24 HPV⁺ OPSCC patients undergoing transoral resection and neck dissection. Each specimen was collected 24 hours after surgery and was treated independently in cases of bilateral neck dissection. DNA was isolated to compare HPV16 load as measured by TaqMan quantitative PCR (qPCR) to pathological diagnosis of ENE, number of positive lymph nodes, and positive lymph node size. Paired plasma samples were obtained from 21 node⁺ patients at the time of SDF collection, to compare HPV DNA detection between biofluid types. Serial dilutions of the HPV16 E6T2aE7 plasmid were used to establish the limit of detection (LOD) for our assay. Statistical analyses included Wilcoxon rank-sum test, Spearman correlation, and Fisher's exact test. AJCC 8^th edition criteria were used for nodal staging. <h3>Results</h3> TaqMan qPCR revealed a lower LOD of 20 HPV copies/µL. The median HPV copies/µL was 43-fold higher in ENE⁺, compared to ENE⁻ necks (<i>P</i> = 0.003). Moreover, 67% of ENE⁺ necks had detectable HPV in SDF, indicative of MRD, compared to only 26% of ENE⁻ necks (<i>P</i> = 0.0049). Additionally, the number of positive nodes correlated with HPV copies/µl measured in SDF (ρ = 0.45, <i>P</i> = 0.0098), with N1/N2 necks harboring 7.5-fold higher HPV copies/µl than N0 necks (<i>P</i> = 0.0003). Size of largest metastatic deposit also correlated significantly with HPV levels in SDF (ρ = 0.75, <i>P</i> < 0.0001). Only 1 paired plasma sample from an ENE⁺ patient had detectable HPV by TaqMan, illustrating the importance of proximal biofluid sampling to most sensitively detect MRD after surgery. <h3>Conclusion</h3> Postoperative TaqMan qPCR of SDF enables MRD detection of HPV⁺ OPSCC, which is highly enriched compared to plasma, and correlates significantly with aggressive pathologic features including nodal stage and ENE. If confirmed, SDF-based HPV quantitation could become a useful and automated adjunct to traditional pathology for personalized adjuvant radiotherapy decision-making.