Currently, there are two registered prophylactic HPV vaccines, a bivalent vaccine and a quadrivalent vaccine. To decide which vaccine could best be used in countries’ national immunization programs, an analytical comparison in terms of cost-effectiveness between both vaccines is required. Accordingly, the primary aim of the current study was to make a specific comparison between the bivalent and the quadrivalent HPV vaccines, taking the specific characteristics of both vaccines explicitly into account. These include not only the benefits of protection against HPV16 and 18, but also those of cross-protection against other high-risk HPV types and, for the low-risk HPV types, protection against genital warts. Note that, although differences in immunogenicity between both vaccines have been reported , the clinical relevance of this has not been demonstrated yet. Even though we conducted the study for the situation in the Netherlands, the general conclusions of our analyses are likely to apply, with minor modifications, to other European countries and the USA, because of strong similarities in HPV prevalence and implementation of prophylactic vaccination programs in these countries.
From our base-case analyses, it follows that use of the bivalent HPV vaccine gives the highest reduction in cervical cancer incidence. Consequently, the highest number of life-years gained was obtained by implementing the bivalent vaccine, resulting in a more favourable ICER compared to that obtained with the quadriavalent vaccine (€21,500/LYG vs €22,700/LYG). This implies that the bivalent vaccine might be €2.80 more expensive if only LYs are considered. Considering QALYs gained on the basis of protection against (pre-)cervical malignancies and genital warts, the quadrivalent vaccine provided the highest health gain. The ICER of the quadrivalent vaccine was estimated to be more favourable than that of the bivalent vaccine at €16,300/QALY gained vs. €17,600/QALY. This implies that the quadrivalent vaccine can be approximately €7 per dose (based at a list price of €105/dose) more expensive than the bivalent vaccine to be equally cost-effective.
The ICERs of both vaccines were found to be particularly sensitive to the vaccine price (range: €5,600 - 17,600/QALY), the duration of protection (range: €16,300 - 48,900/QALY), the discount rate for health benefits (range: €5,300 - 97,100/QALY) and, to herd-immunity benefits (range: €10,700 - 17,600/QALY). In most scenarios, the ICER (€/QALY) of the quadrivalent vaccine remained more favourable than that of the bivalent vaccine. This difference, as argued above, may justify a slightly higher vaccine price for the quadrivalent vaccine. Interestingly, without discounting, the bivalent vaccine results in a slightly higher QALY gain than the quadrivalent vaccine, despite the benefits in terms of prevention of genital warts of the latter. This illustrates that the health gains of prevention of cervical cancer prevention are highly sensitive to the discount rate applied. Furthermore, it is important to note that the acceptable price differential between both vaccines highly depends on the vaccine price applied for the reference vaccine. In particular, in the base-case (bivalent vaccination costs was €105/dose), the price differential was estimated at €7 per dose. Reducing the vaccination costs of the bivalent vaccine to, for example, €45 per dose, the price differential decreases to €4. Possible lower vaccine costs will further reduce the price differential between both vaccines. Inclusion of additional herd-immunity benefits resulted in an increased price difference between both vaccines primarily due to the benefits of providing indirect protection against genital warts in males. Interestingly, the absolute herd-immunity benefit in females will be highest if vaccine coverage is moderate, while in males the herd-immunity benefit will highest when vaccine uptake is high.
In our study, the indirect effects could not be directly considered as a static model was used. The use of static models for infectious disease modelling has been criticised by us and others [28–30]. However, in balancing complexity and transparency, static models are still of major importance and can be used for initial assessments. Here, we included the potential herd-immunity benefits based on a previous Dutch modelling study . Further research should be directed to develop a dynamic transmission model including the most important high- and low-risk HPV types. Another limitation of the current study is that potential additional health benefits of providing protection against other than cervical cancers were not included. The inclusion of these cancers will further improve the ICER for both vaccines. Consequently, price differentials between both vaccines might diminish , as the benefit of providing protection against genital warts will play a less prominent role. Finally, we did not perform any probabilistic sensitivity analysis (PSA). The use of PSA is highly recommended as it provides some insight in the certainty of the outcomes. However, comparing the PSAs of different vaccines is rather critical as potentially a highly favourable option (upper limit 95% confidence interval) of one vaccine is compared with a rather unfavourable option (lower limit 95% confidence interval) of the other vaccine. For that reason we currently didn’t embark on a formal PSA, but rather performed an extensive deterministic sensitivity analysis.
Our results are generally in line with the findings of others. In particular, several studies estimated that the bivalent HPV vaccine resulted in the highest reduction in cervical cancer incidence due to the additional benefits of cross-protection [28, 30]. However, as only the quadrivalent vaccine provides protection against genital warts, the decision which vaccine to use is not that straightforward. From a health-economic perspective, the most cost-effective vaccine should be preferred considering both the health gains of providing protection against cervical cancer and genital warts. Several studies made this type of comparison between both vaccines [25–28]. Here, we estimated that a €6.70 (7%) price differential, in favour of the quadrivalent HPV vaccine, resulted in equal ICERs. This justifiable price difference between both vaccines estimated in the present study is slightly different from that found in other studies . For example, Jit et al. estimated that the bivalent vaccine should be approximately 25% less expensive . Indeed, even larger discrepancies have been reported [31, 33]. The difference between our study and those of others can be explained by various factors. Firstly, as mentioned above, we used a static model in which herd-immunity benefits are not directly taken into account. As indicate above, inclusion of herd-immunity benefits results in an increased price difference (e.g., the price difference reaches a maximum of approximately 16% when herd-immunity is taken into account). Secondly, we included vaccine-specific cross-protection against high-risk HPVs in our analyses. As there is evidence that the bivalent vaccine is more cross-protective than the quadrivalent vaccine [5, 6], this results in a smaller price difference. Finally, according to health-economic guidelines in the Netherlands, future health benefits are discounted with a relatively low discount rate of 1.5%, and therefore, the health benefits of cervical cancer are devalued less than in most other countries. Nevertheless, in general we can conclude that in western countries a higher vaccine price for the quadrivalent vaccine is justified due to the additional benefits of providing protection against genital warts.
In contrast, Demarteau et al. estimated that for Taiwan a higher vaccine price for the bivalent vaccine is justified . The difference between this study and the studies performed in western countries can be primarily explained by the higher incidence of cervical cancer in Taiwan. Due to the potentially higher effectiveness of the bivalent vaccine in providing protection against cervical cancer a relevantly higher health gain might be obtained in settings with a high burden of cervical cancer. In most western countries the burden of cervical cancer has already been dramatically reduced since the introduction of cervical cancer screening. Consequently, in these settings the benefits of providing protection against genital warts are predominant and the ICER of the quadrivalent vaccine is even more sensitive to the inclusion of the benefits of providing protection against genital warts than found in our study. This illustrates that the decision which vaccine to use on health-economic grounds will highly depend on the disease burden of cervical cancer versus the burden of genital warts.