Vivek Verma1, Charles B Simone2, Chi Lin1. 1. Department of Radiation Oncology, University of Nebraska Medical Center, Omaha, Nebraska. 2. Department of Radiation Oncology, University of Maryland Medical Center, Baltimore, Maryland.
Abstract
BACKGROUND: There are no existing high-volume studies characterizing human papillomavirus (HPV)-associated nasopharyngeal cancer (NPC). METHODS: The National Cancer Data Base (NCDB) was queried for NPC with known HPV (2004-2013). Logistic regression ascertained factors associated with HPV-positivity. Kaplan-Meier overall survival (OS) was evaluated between HPV-positive and HPV-negative cohorts; Cox proportional hazards modeling assessed factors associated with OS. Patients with nonmetastatic disease receiving definitive chemoradiotherapy underwent propensity-matched OS analysis. RESULTS: Altogether, 956 patients were analyzed (32% HPV-positive and 68% HPV-negative). Median follow-up was 23 months (range 0-67 months). The patients with HPV-positive disease were younger, less likely to be uninsured, lived in more educated areas, and presented with more advanced T (but not N/overall) classification. Median OS for HPV-positive and HPV-negative groups were 50 and 43 months, respectively (P = .171). The HPV status did not independently predict for OS (P = .183). No OS differences were observed after propensity matching (P = .734). CONCLUSION: In what we believe as the only large study of HPV-associated NPC, HPV neither correlates with nor predicts survival in NPC. Owing to the difficulty of addressing causality in database studies, further work must corroborate the findings herein.
BACKGROUND: There are no existing high-volume studies characterizing human papillomavirus (HPV)-associated nasopharyngeal cancer (NPC). METHODS: The National Cancer Data Base (NCDB) was queried for NPC with known HPV (2004-2013). Logistic regression ascertained factors associated with HPV-positivity. Kaplan-Meier overall survival (OS) was evaluated between HPV-positive and HPV-negative cohorts; Cox proportional hazards modeling assessed factors associated with OS. Patients with nonmetastatic disease receiving definitive chemoradiotherapy underwent propensity-matched OS analysis. RESULTS: Altogether, 956 patients were analyzed (32% HPV-positive and 68% HPV-negative). Median follow-up was 23 months (range 0-67 months). The patients with HPV-positive disease were younger, less likely to be uninsured, lived in more educated areas, and presented with more advanced T (but not N/overall) classification. Median OS for HPV-positive and HPV-negative groups were 50 and 43 months, respectively (P = .171). The HPV status did not independently predict for OS (P = .183). No OS differences were observed after propensity matching (P = .734). CONCLUSION: In what we believe as the only large study of HPV-associated NPC, HPV neither correlates with nor predicts survival in NPC. Owing to the difficulty of addressing causality in database studies, further work must corroborate the findings herein.