Cost-effectiveness of the next generation nonavalent human papillomavirus vaccine in the context of primary human papillomavirus screening in Australia: a comparative modelling analysis

Abstract

Summary: Background: First generation bivalent and quadrivalent human papillomavirus (HPV) vaccines have been introduced in most developed countries. A next generation nonavalent vaccine (HPV9) has become available, just as many countries are considering transitioning from cytology-based to HPV-based cervical screening. A key driver for the cost-effectiveness of HPV9 will be a reduction in screen-detected abnormalities and surveillance tests. We aimed to evaluate the cost-effectiveness of HPV9 in Australia, a country with HPV vaccination of both sexes that is transitioning to 5-yearly HPV-based screening. Methods: We used Policy1-Cervix and HPV-ADVISE—two dynamic models of HPV transmission, vaccination, and cervical screening—to estimate the cost-effectiveness of HPV9 versus quadrivalent vaccine (HPV4), assuming lifelong vaccine protection, two vaccine doses, and that additional costs were incurred in girls only. Policy1-Cervix was used to estimate the lifetime risk of cervical cancer diagnosis and death. Probabilistic sensitivity analysis of the cost-effectiveness outcomes was done with both models, and results are presented as the median and 10th to 90th percentiles of simulation runs (referred to as 80% uncertainty intervals [UIs]). Findings: Compared with cytology-based screening, HPV screening is predicted to reduce lifetime risk of cervical cancer diagnosis by 18% and of death by 20%, even in unvaccinated cohorts. Under base-case assumptions (lifelong protection, full efficacy at two doses), HPV4 will provide a further reduction in diagnosis of 54% and in death of 53% and HPV9 will provide a further reduction in both diagnosis and death of 11%, compared with cytology-based screening in unvaccinated cohorts. For HPV9 to remain a cost-effective alternative to HPV4, the incremental cost per dose in girls should not exceed a median of AUS$35·99 (80% UI 28·47–41·18) with Policy1-Cervix or AUS$22·74 (15·49–34·45) with HPV-ADVISE, at a willingness-to-pay threshold of AUS$30 000 per quality-adjusted life-year. Interpretation: Differing methods and assumptions led to some differences in the estimates produced by the two models. However, on the basis of median results, HPV9 will be a cost-effective alternative to HPV4 if the additional cost per dose is AUS$23–36 (US$18–28). These results will be important when determining the optimum price of the vaccine in Australia. Funding: National Health and Medical Research Council, Australia.