P. E. Clark1, K. Taparra2, and J. A. Miller1; 1Cleveland Clinic, Cleveland, OH, 2Stanford Cancer Institute, Stanford, CA
Purpose/Objective(s): The disproportionate burden of morbidity and mortality from Epstein-Barr Virus (EBV)-associated nasopharyngeal carcinoma (NPC) in the United States occurs among Asian American (AA) and Native Hawaiian and other Pacific Islander (NHPI) populations. We sought to define a high-risk population in the US for EBV-based NPC screening and estimate the number needed to screen (NNS) and NPC mortality reduction. We hypothesized that a single lifetime screen among high- and intermediate-incidence populations could be cost-effective. Materials/
Methods: Age-, sex-, and ethnicity-specific WHO type II/III NPC incidence data were obtained from the SEER Asian and Pacific Islander dataset for 12 intermediate- (1-2 cases per 100,000) and high-incidence (>2 cases per 100,000) populations. Performance data from prospective trials were used to estimate NNS, NPC mortality reduction, and resource utilization for a single lifetime serologic screen using a validated NPC cohort model. Six screening strategies were evaluated consisting of combinations of EBV serology, nasopharyngeal swab EBV PCR, nasopharyngeal endoscopy, and MRI. Cost-effectiveness was evaluated with incremental cost-effectiveness ratios (ICER).
Results: Despite comprising only 3.7%, 3.8%, and 1.6% of US person-years, 21.3%, 10.5%, and 6.8% of US NPC cases developed in the Chinese, Filipino, and Vietnamese populations, respectively. Moreover, despite representing only 1.5% of US person-years, 4.3% of total cases developed in the Laotian, Cambodian, Native Hawaiian, Samoan, and Guamanian populations. Intermediate- and high-incidence populations accounted for 10.7% of US person-years and 42.7% of NPC cases. Screening at ages of peak incidence using EBV anti-BNLF2b serology and selective endoscopy was the preferred screening strategy due to highest sensitivity and lowest cost. For a hypothetical cohort of 50-year-old men and women, the median NNS to detect one case among high-incidence populations was 1,992 (range, 1,856-4,609). A median of 7.1 NPC deaths were averted per 100,000 screened. When limited to men, median NNS was 1,407 (range, 1,349-2,818; 34-85 endoscopies per detected case). Screening with anti-BNLF2b met the willingness-to-pay threshold in all five high-incidence populations (median ICER per GDP 0.82, range 0.66-1.64). The threshold was also met among intermediate-incidence Filipino, Guamanian, and Native Hawaiian men.
Conclusion: Nearly half of US NPC cases arise among the 11% with AA or NHPI ethnicity who have no access to screening. If screening were adopted, an appropriate population would be men/women age 35-65 of Chinese, Vietnamese, Laotian, Cambodian, or Samoan ethnicity, or men age 35-60 of Filipino, Guamanian, or Native Hawaiian ethnicity. Once-lifetime screening with anti-BNLF2b serology could be cost-effective with approximately 2,000 screened subjects per detected case. NPC screening might mitigate health disparities, and these data may aid the design of screening initiatives.
