Identification of High-Incidence Populations in the United States for Anti-Epstein-Barr Virus Serologic Screening for Nasopharyngeal Carcinoma.

Abstract

Background: In the United States, Epstein-Barr virus (EBV)-associated nasopharyngeal carcinoma (NPC) disproportionately impacts Asian Americans (AA) and Native Hawaiians and other Pacific Islanders (NHPI) who have no access to screening. EBV-based screening trials in Asia have detected most cases at early stages. We sought to identify a US target population for NPC screening and hypothesized that once-lifetime screening could be cost-effective.

Methods: We obtained NPC incidence data from the Surveillance, Epidemiology, and End Results Asian and Pacific Islander datasets. We estimated the number needed to screen (NNS), mortality reduction, and resource utilization using a validated model and performance data from trials. Six evaluated strategies incorporated serology, nasopharyngeal swab PCR, and endoscopy or MRI.

Results: Intermediate-incidence and high-incidence populations accounted for 10.7% of US person-years yet 42.7% of cases. Anti-BNLF2b screening with selective endoscopy was the preferred strategy. In high-incidence populations, the median NNS to detect one case was 1,992, with a median of 7.12 NPC deaths averted per 100,000 screened. Screening met the willingness-to-pay threshold in all five high-incidence populations (median incremental cost-effectiveness ratio/gross domestic product, 0.82) and among men in intermediate-incidence populations.

Conclusions: Nearly half of NPC in the United States arises among the 10% with AA or NHPI ethnicity. A suitable target population for US screening trials would be men and women aged 35 to 65 years of Chinese, Sāmoan, or Southeast Asian ethnicity, or men aged 35 to 60 years of Guamanian/Chamorro, Filipino, or Native Hawaiian ethnicity. Once-lifetime anti-BNLF2b screening could be cost-effective.

Impact: These data may aid the design of US screening trials. Targeted NPC screening might mitigate health disparities.