Cost-Effectiveness Analysis of Chemoradiation Therapy Versus Transoral Robotic Surgery for Human Papillomavirus–Associated, Clinical N2 Oropharyngeal Cancer

Abstract

To perform a cost-effectiveness analysis of primary chemoradiation therapy (CRT) versus transoral robotic surgery (TORS) for clinical N2, human papillomavirus (HPV)-positive oropharyngeal carcinoma. We developed a Markov model to describe the health states after treatment with CRT or TORS, followed by adjuvant radiation therapy or CRT in the presence of high-risk pathology (positive margins or extracapsular extension). Outcomes, toxicities, and costs were extracted from the literature. One-way sensitivity analyses (SA) were performed over a wide range of parameters, as were 2-way SA between the key variables. Probabilistic SA and value of information studies were performed over key parameters. The expected quality-adjusted life years (QALYs)/total costs for CRT and TORS were 7.31/$50,100 and 7.29/$62,200, respectively, so that CRT dominated TORS. In SA, primary CRT was almost always cost-effective up to a societal willingness-to-pay of $200,000/QALY, unless the locoregional recurrence risk after TORS was 30% to 50% lower, at which point it became cost effective at a willingness-to-pay of $50-100,000/QALY. Probabilistic SA confirmed the importance of locoregional recurrence risk, and the value of information in precisely knowing this parameter was more than $7M per year. If the long-term utility after TORS was 0.03 lower than CRT, CRT was cost-effective over nearly any assumption. Under nearly all assumptions, primary CRT was the cost-effective therapy for HPV-associated, clinical N2 OPC. However, in the hypothetical event of a large relative improvement in LRR with surgery and equivalent long-term utilities, primary TORS would become the higher-value treatment, arguing for prospective, comparative study of the 2 paradigms.