RESUMEN
PURPOSE: Human papillomavirus (HPV) is the primary driver of cervical cancer. Although studies in other malignancies correlated peripheral blood DNA clearance with favorable outcomes, research on the prognostic value of HPV clearance in gynecologic cancers using intratumoral HPV is scarce. We aimed to quantify the intratumoral HPV virome in patients undergoing chemoradiation therapy (CRT) and associate this with clinical characteristics and outcomes. METHODS AND MATERIALS: This prospective study enrolled 79 patients with stage IB-IVB cervical cancer undergoing definitive CRT. Cervical tumor swabs collected at baseline and week 5 (end of intensity modulated radiation therapy) were sent for shotgun metagenome sequencing and processed via VirMAP, a viral genome sequencing and identification tool for all known HPV types. The data were categorized into HPV groups (16, 18, high risk [HR], and low risk [LR]). We used independent t tests and Wilcoxon signed-rank to compare continuous variables and χ2 and Fisher exact tests to compare categorical variables. Kaplan-Meier survival modeling was performed with log-rank testing. HPV genotyping was verified using quantitative polymerase chain reaction to validate VirMAP results using receiver operating characteristic curve and Cohen's kappa. RESULTS: At baseline, 42%, 12%, 25%, and 16% of patients were positive for HPV 16, HPV 18, HPV HR, and HPV LR, respectively, and 8% were HPV negative. HPV type was associated with insurance status and CRT response. Patients with HPV 16+ and other HPV HR+ tumors were significantly more likely to have a complete response to CRT versus patients with HPV 18 and HPV LR/HPV-negative tumors. Overall HPV viral loads predominantly decreased throughout CRT, except for HPV LR viral load. CONCLUSIONS: Rarer, less well-studied HPV types in cervical tumors are clinically significant. HPV 18 and HPV LR/negative tumors are associated with poor CRT response. This feasibility study provides a framework for a larger study of intratumoral HPV profiling to predict outcomes in patients with cervical cancer.
