The differential risk of cervical cancer in HPV vaccinated and unvaccinated women: a mathematical modelling study

Abstract

Background: With increased uptake of vaccination against HPV, protection against cervical cancer will also increase for unvaccinated women, due to herd immunity. Still, the differential risk between vaccinated and unvaccinated women might warrant a vaccination-status screening approach. To understand the potential value of stratified screening protocols, we estimated the risk differentials in HPV and cervical cancer between vaccinated and unvaccinated women. Methods: We used STDSIM, an individual-based model of HPV transmission and control, to estimate the HPV prevalence reduction over time, after introduction of HPV vaccination. We simulated scenarios of bivalent or nonavalent vaccination in females-only or females and males, at 20% coverage increments. We estimated relative HPV type-specific prevalence reduction compared to a no-vaccination counterfactual and then estimated the age-specific cervical cancer risk by vaccination status. Results: The relative cervical cancer risk for unvaccinated compared to vaccinated women ranged from 1.7 (bivalent vaccine for females and males; 80% coverage) to 10.8 (nonavalent vaccine for females-only; 20% coverage). Under 60% vaccination coverage, which is a representative coverage for several western countries, including the United States, the relative risk varies between 2.2 (bivalent vaccine for females and males) and 9.2 (nonavalent vaccine for females). Conclusions: We found large cervical cancer risk differences between vaccinated and unvaccinated women. In general, our model shows that the relative risk is higher in lower vaccine coverages, using the nonavalent vaccine, and when vaccinating females only. Impact: To avoid a disbalance in harms and benefits between vaccinated and unvaccinated women, vaccination-based screening needs serious consideration.