Cost-Effectiveness of Screening for Nasopharyngeal Carcinoma with Plasma Epstein-Barr Virus DNA

Abstract

Screening for nasopharyngeal carcinoma (NPC) in endemic areas of Asia results in improved cancer outcomes. Screening is not routinely performed in the US, where NPC is rare. We examined the cost-effectiveness of screening for NPC with plasma Epstein-Barr Virus (EBV) DNA for Asian-American men in the US. We used a Markov cohort model to estimate discounted mean life-years, quality-adjusted life-years (QALYs), costs, and incremental cost-effectiveness ratios for screening compared to usual care without screening. Test accuracy was based on a study of 20,174 men in Hong Kong screened with plasma EBV DNA. Confirmatory testing was an MRI and nasopharyngoscopy. A false-positive could result in additional costs due to further workup and incidental findings. Estimates for the annual incidence of NPC and remaining life-years were based on data from the Surveillance, Epidemiology, and End Results (SEER) program. Costs were based on Medicare reimbursement rates. The model assumed that stage I NPC was treated with radiation, stage II-IVB with chemoradiation, and stage IVC with chemotherapy. The base case analysis considered one-time screening for 50-year-old Asian-American men. We used a $150,000/QALY gained willingness-to-pay threshold and performed one-way and probabilistic sensitivity analyses. The calibrated model produced annual incidence rates for all men aged 50-100 years old in the US of 0.18, 0.39, 0.95, and 0.22 per 100,000 people for stages I, II, III-IVB, and IVC, respectively, consistent with those reported by SEER. For Asian-American men in the base case, usual care without screening resulted in detection of cancer at stages I, II, III-IVB, and IVC for 5%, 17%, 64%, and 14%, respectively, whereas screening resulted in earlier detection with a stage distribution of 43%, 23%, 33%, and 1%, respectively. This corresponded to an additional 0.00045 life-years and 0.00050 QALYs at a cost of $58 per person. Screening cost $116,000/QALY gained. In contrast, due to lower NPC risk, screening cost $647,000/QALY gained if implemented for all 50-year-old men in the US. The cost per QALY gained was sensitive to test accuracy, NPC prevalence, and stage distribution estimates. Probabilistic sensitivity analysis demonstrated screening Asian-American men had an 86% probability of being cost-effective at $150,000/QALY. Screening for NPC with plasma EBV DNA for 50-year-old Asian-American men in the US may be cost-effective if high levels of test sensitivity and specificity can be achieved for US populations and testing can be conducted inexpensively. The cost per QALY gained would ultimately depend on how screening is implemented, test cost, and accuracy. Although NPC is a rare disease in the US, patients’ outcomes could be improved with earlier detection. Our analyses suggest that clinical studies to address accuracy, yield, and effectiveness of screening in this population would be useful.