Local Cost-Effectiveness of Nasopharyngeal Carcinoma Screening Strategies in Southern China: Secondary Analysis of the PRO-NPC-001 Cluster-Randomized Trial

Abstract

Population-based screening for endemic nasopharyngeal carcinoma (NPC) detects most cases at an early stage. In a cluster-randomized trial conducted in Guangdong, a combination of Epstein-Barr Virus (EBV) anti-VCA/EBNA1 IgA serology and endoscopy reduced NPC mortality. We conducted a secondary analysis of this trial in conjunction with local incidence and cost data, hypothesizing that screening would be cost-effective in this region. We estimated population-level NPC mortality reduction, resource utilization, and cost-effectiveness of 12 unique screening strategies in six populations in Guangdong/Guangxi using a previously-validated time-inhomogeneous decision-analytic cohort model. These 12 strategies evaluated combinations of serology, nasopharyngeal swab PCR (NP PCR), endoscopy, and head/neck MRI. Incidence data, screening costs, and healthcare costs were obtained from local cancer registries, laboratories conducting ELISA/PCR, and the Guangdong provincial healthcare system. We evaluated variable screening ages, sexes, intervals, and durations to identify optimal screening approaches from the perspective of the healthcare system in southern China. An incremental cost-effectiveness ratio (ICER) willingness-to-pay threshold of 1.50 times the per-capita GDP was considered cost-effective in southern China. For the base strategy screening 50-year-old men and women using only serology and endoscopy, the average cost per screened subject for a single round of screening over a five-year cycle was ¥175.69. The addition of MRI improved sensitivity (76% vs. 62%) and approximately doubled screening costs. Triage with NP PCR was cost-neutral when used in conjunction with MRI and reduced endoscopy/MRI utilization by 37% with a 3-4% reduction in screening sensitivity. Among 50-year-old men and women, screening was cost-effective in all populations provided that medium-risk subjects were not referred for endoscopy/MRI (ICER/GDP 0.62-0.83). The use of NP PCR without MRI (ICER/GDP 0.83) was dominated by the base strategy (ICER/GDP 0.62) due to higher costs and NPC mortality. After a single five-year screening cycle, screening reduced population NPC mortality by 14% with serology + endoscopy and 21% with serology + endoscopy + MRI. Introduction of MRI with or without NP PCR could be cost-effective in all populations. For MRI-based strategies, the most efficient use of resources was deferral of endoscopy unless MRI was abnormal (ICER/GDP 0.67). Overall, the best-performing strategies balanced NPC mortality, screening costs, and MRI utilization. EBV serology-based screening for endemic NPC is likely to be cost-effective among adult men and women in Guangdong and Guangxi. Referring medium-risk subjects for endoscopy/MRI should be avoided, and NP PCR should be used to triage individuals for MRI rather than endoscopy. These data may aid the design of population-based screening programs in this region.