The incidence of nasopharyngeal carcinoma (NPC) exhibits marked geographic variation globally and is intimately associated with endemic Epstein-Barr Virus (EBV) infection. Although most patients present with locoregionally-advanced disease, several population-level screening trials using blood-based EBV assays have successfully detected most cases in the asymptomatic period. Given the rarity of NPC, it is uncertain whether screening could be cost-effective, and what the optimal screening program might be. We therefore sought to define the cost-effectiveness of three blood-based screening strategies in high-incidence regions globally, and hypothesized that screening of men and women in both high-incidence and middle-income countries could be costeffective. Materials/Methods: Using incidence data from 340 cancer registries in 69 countries, we selected 42 high-incidence populations in 15 countries with a lifetime NPC incidence exceeding 0.25%. We developed a decision-analytic model comparing three previously-tested annual screening strategies (EBV DNA PCR + endoscopy + MRI, PCR + endoscopy, and ELISA serology + endoscopy) to no screening. Transition probabilities and stage distributions were based upon screening trials. Clinical outcomes were extracted from trial data and the AJCC 8 update, and microcosting was performed using WHO-CHOICE methods. One-way and probabilistic sensitivity analyses were performed to study the impact of age, gender, utilities, and costs. A willingness-to-pay (WTP) threshold was set at twice the local per-capita GDP. Results: Among the 42 selected populations, screening with PCR + endoscopy + MRI, PCR + endoscopy, and ELISA + endoscopy decreased NPC-specific mortality by 62%, 53%, and 72%, respectively. These strategies thereby improved 5-year NPC-specific survival from 74% in an unscreened population to 91%, 89%, and 94%. In the base case scenario, screening with PCR + endoscopy + MRI was cost-effective at the WTP threshold in 40% of selected populations, compared to 38% with PCR + endoscopy and 67% with ELISA + endoscopy. The addition of MRI to PCR was marginally more cost-effective than endoscopy alone, but was associated with small improvements in 5-year NPC-specific survival (2%). Among populations with a lifetime incidence exceeding 0.60%, screening was always cost-effective. Moreover, screening in many middle-income and most high-income countries was also cost-effective. Sensitivity analysis informed the optimal age at first screening in each region, as well as the screening of women. Probabilistic sensitivity analysis was in close agreement with base case analyses. Conclusion: We observed significant reductions in NPC-specific mortality with the use of blood-based screening in high-incidence populations globally. Despite economic differences and the rarity of NPC, these strategies were cost-effective in most high-incidence regions, and could be extended to both adult men and women, generally beginning at age 40.
