The cost effectiveness of a quadrivalent human papillomavirus vaccine (6/11/16/18) in Hungary

Abstract

Objective: A transmission dynamic model was used to assess the epidemiological and economic impact of a quadrivalent human papillomavirus (HPV) (6/11/16/18) vaccine in preventing cervical cancer, cervical intraepithelial neoplasia grades 2 and 3 (CIN 2/3), CIN 1 and genital warts in Hungary.

Methods: The routine vaccination of 12-year-old girls and the routine vaccination of 12-year-old girls plus a temporary catch-up programme for girls and women aged 12–24 years was evaluated.

Results: The model projected that at year 100, both strategies could reduce the incidence of HPV 6/11/16/18-related cervical cancer, CIN 2/3, CIN 1 and genital warts cases among Hungarian women by 90%, 90%, 85% and 93%, respectively. Twenty-five years after the introduction of HPV vaccination in the population, routine vaccination of girls by the age of 12 reduced the cumulative number of cases of cervical cancer, CIN 2/3, CIN 1 and genital warts by 685, 13,473, 3,423 and 163,987, respectively. The incremental cost-effectiveness ratios of the two vaccination strategies were €9,577 and €10,646 per quality-adjusted life-year (QALY) gained over a time horizon of 100 years.

Key limitations: The model did not account for the health and economic impact of other HPV diseases which may result from HPV 16, 18, 6, and 11 infections such as vaginal, vulvar, penile, anal and head-neck cancers, and recurrent respiratory papillomatosis. Epidemiological data from Hungary on these other HPV diseases as well genital warts are needed.

Conclusion: A quadrivalent HPV vaccination programme can reduce the incidence of cervical cancer, CIN and genital warts in Hungary at a cost-per-QALY ratio within the range defined as cost effective.

Keywords::

• cervical intraepithelial neoplasia
• cost-effectiveness analysis
• epidemiology
• human papillomavirus
• uterine cervical neoplasms
• vaccines

Introduction

Globally, the second most common cancer in women is cervical cancer. Each year, approximately half a million cases are diagnosed, half of which result in death^1,2. Cervical intraepithelial neoplasia (CIN) marks the pre-cancerous stages of cervical cancer, starting with the mild dysplasia (CIN 1), progressing to the more serious forms of CIN 2 and CIN 3, and finally progressing to carcinoma. Cervical cancer is known to be caused by infection with human papillomavirus^3,4 (HPV). Although more than 100 types of the virus have been identified⁵, types 16 and 18 in particular are known to be oncogenic and associated with cervical cancer. Infection with the low-risk types of HPV, 6 and 11, causes 90% of cases of condylomata acuminatum (genital warts)⁶.

It is estimated that 80% of cervical cancer cases occur in less-developed countries that have access to less than 5% of the global cancer treatment resources⁷. The ten newest members of the European Union (EU), which include most of the Eastern European nations, (EU10: Cyprus, Czech Republic, Estonia, Hungary, Latvia, Lithuania, Malta, Poland, Slovakia and Slovenia), have a 1.8-fold higher incidence rate of cervical cancer and a 2.6-fold higher cervical cancer mortality rate than their counterparts in the original 15 member nations of the EU (EU15: Austria, Belgium, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, Portugal, Spain, Sweden, The Netherlands and the United Kingdom)⁸.

The nation of Hungary, in particular, has some of the highest cervical cancer rates in the EU. Of the 25 member nations, Hungary has the sixth highest incidence rate and the fifth highest mortality rate of cervical cancer. Beginning in 2003, the Hungarian National Cancer Control Programme began an organised national cervical cancer screening programme for cervical cancer^9,10. Although screening appears to be effective in decreasing the incidence of cervical cancer⁹, uptake has been poor¹¹. In addition, Hungarian women have been shown to have high rates of smoking and the use of contraceptive pills, both of which have been associated with increased risks of cervical cancer^12–14. As a result, cervical cancer incidence is higher in Hungary than most other EU nations with more organised and accessed screening.

Although approximately 10% of HPV infections in Hungary are caused by the high-risk genotypes associated with cervical cancer, 4% of infected women have the low-risk genotypes that cause genital warts (e.g. HPV 6 and 11)¹⁵. Approximately 75% of these women with genital warts will seek treatment from a physician¹⁶, at an estimated cost of €41.24 per case¹⁷.

Prevention of HPV infection and disease with a vaccine can play an important role in complementing current cervical cancer screening initiatives in Hungary. A prophylactic quadrivalent vaccine has been approved for use in the EU for the prevention of CIN 2/3, cervical carcinoma, high-grade vulvar dysplastic lesions (VIN 2/3), and external genital warts causally related to HPV types 6, 11, 16 and 18¹⁸. In order for policy makers in Hungary to formulate HPV vaccination guidelines and to support reimbursement decisions, they will seek information on the epidemiological and economic consequences of HPV immunisation programmes using this quadrivalent HPV vaccine¹⁹. Currently, information on the long-term epidemiological and economic consequences of HPV immunisation programmes using this quadrivalent HPV vaccine in Hungary is not available. In this paper, the authors review the results from a previously developed economic model²⁰ adapted to explore the epidemiological and economic consequences of introducing a quadrivalent HPV vaccine in Hungary. Specifically, this study will answer the following questions:

What is the potential impact of a quadrivalent HPV vaccine on CIN, cervical cancer, cervical cancer mortality and genital warts in the female population in Hungary?

What is the cost effectiveness of a quadrivalent HPV vaccine programme when added to the current standard of care (i.e. cervical cancer screening and clinical management of CIN, cervical cancer and genital warts) from the perspective of the national health insurer in Hungary?

Methods

A previously developed mathematical model for evaluating the impact of quadrivalent HPV vaccination in the US was adapted to explore the potential epidemiological and economic impact of a quadrivalent HPV vaccine in Hungary. Details of this model and its structure have been previously described²⁰. Components of the model that were modified for Hungary include screening, treatment and vaccination strategies, as well as epidemiological (e.g. mortality) and economic inputs. The following describe the strategies evaluated, model parameters and output, the simulation method, and the sensitivity and validation analyses.

Screening and vaccination strategies

It was assumed that each vaccination strategy would be combined with current cervical cancer screening and HPV disease treatment practices in Hungary. The base-case vaccination strategy was the routine HPV vaccination of girls by 12 years of age. The routine vaccination of girls by 12 years in conjunction with a catch-up vaccination programme that targeted girls and women 12–24 years of age was also examined. The catch-up programme was assumed to be temporary and to last for a period of 5 years.

Model parameters and sources

Baseline assumptions and estimates were determined by a comprehensive search of the literature, input from experts and analysis of clinical trial data²⁰. Baseline epidemiological and economic parameters and sources are shown in Supplementary Tables S1, S2, S3 and S4.

Screening zand vaccination programme strategy parameters

The duration of protection for the HPV vaccine was assumed to be from 10 years to lifetime. It was also assumed that vaccination would not have any effect on the natural course of any HPV infection that may have been present at the time of vaccination and that routine vaccination would be permanent. Efficacy assumptions were 90% against HPV 6/11/16/18, 95.2% against all CIN caused by HPV 6/11/16/18 and 98.9% against genital warts caused by HPV 6/11. It was assumed that up to 85% of 12-year-old Hungarian girls would receive the full series of three doses²¹. It was also assumed that the catch-up programme for 12–24-year-old Hungarian girls and women would provide 10% coverage by year 5²¹.

Economic parameters

The economic perspective of the National Health Insurance Fund Administration (NHIFA)¹⁷ in Hungary was adopted for the analysis. This perspective only considers direct medical costs; hence, the costs associated with work and productivity losses were not included. Direct medical costs included the costs of vaccination, screening, diagnosis and treatment of detected cervical cancer, CIN and genital warts.

Direct medical costs of interventions were estimated using the average national tariffs of the NHIFA announced for the first semester of 2006 and were reported in EUR. It was assumed the cost of the three-dose vaccine series was €279. In Hungary a school-based vaccination programme for hepatitis B exists for 12 year-olds with a high coverage rate. Since HPV vaccines can be co-administered with hepatitis B vaccine, the proposed vaccination strategy does not generate any additional administration cost. Costs of conventional cytology screening examinations, colposcopy and biopsy were taken from the official code list of the NHIFA. Treatment costs of genital warts were calculated on the basis of expert opinion based on the available treatment options and utilisation. Data from the database of the NHIFA were used to estimate the average per-patient costs of the diagnosis and treatment of precancerous states and cervical cancer. Results of the costing analysis are shown in Table S3. Quality-of-life weights were measured using health utilities²², which were used to estimate quality-adjusted life-years (QALYs). The planning horizon was 100 years, and it was assumed that the population at any given time over the 100-year horizon would be 100,000. All costs and effects were discounted at an annual rate of 5% in the base-case analysis.

Model output

A number of measures were used to assess the epidemiological impact and cost effectiveness of both vaccination strategies. The epidemiological output included invasive cervical cancer, CIN 2/3, CIN 1 and female genital warts cases, as well as cervical cancer deaths. Economic output included total costs, quality-adjusted survival and cost per QALY. The incremental cost-effectiveness ratio (ICER) was measured as the incremental cost difference between two strategies divided by the incremental QALY difference between the two strategies.

Simulation method

The MSD WebModel for the Quadrivalent HPV types 6, 11, 16, 18 vaccine was used. This model was developed as series of differential equations in Mathematica (Wolfram Research, Champaign, IL, USA), and used the NDSolve subroutine in Mathematica version 7.0 to generate numerical solutions for the differential equations of the model. The baseline parameter estimates were used to solve the model for the prevaccination steady-state values of the variables. The prevaccination data were used as initial values for the vaccination model. The entire time path of the variables was then solved for until the system approached a steady state at approximately 100 years. This solution was used to generate the output described previously for each of the screening and vaccination strategies.

Validation analyses

Validation of the natural history component of the model has been previously described in detail^20,23. For the Hungarian adaptation, the validity of the model was assessed by comparing model predictions to the epidemiological data found in the literature on the incidence of cervical cancer in Hungary.

Sensitivity analysis

Extensive sensitivity analyses from prior models^20,23,24 were used to identify parameters that would be most influential to the results. The parameters examined in the sensitivity analysis included duration of vaccine protection, vaccine coverage, degree of vaccine efficacy against HPV 6/11-related disease, vaccine costs, diagnosis and treatment costs of HPV diseases, the discount rate and quality of life (i.e. health utilities). In addition, a pessimistic-scenario sensitivity analysis was conducted where the above parameters were all set to values that would be biased against vaccination.

Results

Model validation

The model predicted that in the absence of vaccination and with current screening strategies, the incidence of HPV 16/18-related cervical cancer would be 13 per 100,000 among girls and women ≥12 years of age. Arbyn et al estimated a crude rate of 19.0 for cervical cancer incidence due to all HPV types in Hungary for year 2004²⁵. Evidence suggests that approximately 70% of all these cases (i.e. 13.3 per 100,000) are caused by HPV 16/18²⁶.

Epidemiologic impact of the vaccination strategies (base case)

When compared with no vaccination, both vaccination strategies significantly reduced the incidence of HPV 6/11/16/18-related disease. The incidence of cervical cancer, CIN 2/3, CIN 1 and female genital warts were reduced by 90%, 90%, 85% and 93%, respectively, over a 100-year time horizon (Figures 1A–1E). However, the catch-up vaccination provided earlier and greater reductions in HPV-related diseases than the routine vaccination strategy. Table 1 shows the cumulative HPV 6/11/16/18-related disease events prevented by vaccination at 12 years relative to no vaccination up to 25 years after the introduction of vaccination into the population. For example, under the 25-year column, rows 1 and 4 show that 163,987 cases of genital warts (female and male) and 685 cases of cervical cancer could be prevented over 25 years if routine vaccination of 12-year-old girls were implemented in Hungary.

Figure 1. Impact of vaccination strategies on HPV-related disease in girls and women aged 12+ years. (A) HPV 16/18–related cervical cancer incidence; (B) HPV 16/18-related cervical cancer deaths. (C) HPV 6/11/16/18-related CIN 2/3 incidence; (D) HPV 6/11/16/18-related CIN 1 incidence (E) HPV 6/11-related genital warts incidence.

Table 1. Cumulative HPV (6/11/16/18) disease events prevented with vaccination at age 12 years relative to no vaccination in the Hungarian population ≥12 years of age.

HPV disease event	Years since start of vaccination programme
	5	10	15	20	25
Genital warts*	1,212	17,731	56,481	107,769	163,987
CIN 1	6	198	883	2,020	3,423
CIN 2/3	14	545	2,829	7,280	13,473
Cervical cancer	0	3	43	227	685
Cervical cancer deaths	0	0	5	38	146
Total	1,233	18,476	60,242	117,334	181,714

*Number of avoided genital wart cases in male and female population.

Economic impact of vaccination strategies (base case)

Figure 2 shows the annual discounted costs avoided by the routine vaccination of 12-year-old girls relative to no vaccination. Prior to year 23, most of the costs avoided are attributable to the prevention of genital warts caused by HPV 6/11. After year 23, the costs avoided are primarily attributable to the prevention of CIN and cervical cancer caused by HPV 16/18. For example, in year 10, approximately €145,304 and €20,933 could be avoided by the prevention of HPV 6/11-related disease and HPV 16/18-related disease, respectively. However by year 30, approximately €108,153 and €178,724 could be avoided by the prevention of HPV 6/11-related disease and HPV 16/18-related disease, respectively. The total discounted costs and QALYs over 100 years for each strategy are shown in Table 2. The cost-effectiveness ratios of the routine vaccination of 12-year-old girls and routine vaccination of 12-year-old girls plus 12–24-year-old catch-up were €9,577 /QALY and €10,646/QALY, respectively.

Figure 2. Annual discounted HPV disease treatment costs avoided in the total population by routine vaccination of 12-year-old girls relative to no vaccination, stratified by HPV type.

Table 2. Cost-effectiveness analysis of HPV vaccination strategies in the base case.

Strategy	Discounted total			Incremental
	Costs (€)	QALYs	Costs (€)	QALYs	€/QALYs*
No vaccination	1,631,534	1,705,341	–	–	–
12-year-old girls	4,988,770	1,705,692	3,357,236	351	9,577
+12–24-year-old catch-up	5,401,992	1,705,731	413,222	39	10,646

€1 ∼ HUF254.59.

*Compared with the preceding non-dominated strategy.

Sensitivity analyses

Results of the sensitivity analyses are summarised in Table 3. The most influential analysis was the pessimistic scenario biased against vaccination. In this scenario, the duration of protection was limited to 10 years, the cost of the vaccine series was 20% higher, and there was lower quality-of-life impact.

Table 3. Sensitivity analyses: cost-effectiveness ratios.

Scenario	Routine vaccination* (€/QALY)	Routine vaccination + 12–24 year-old catch-up^† (€/QALY)
Baseline analysis	9,577	10,646
Duration of vaccine protection=10 years	Weakly dominated^‡	23,291
Vaccine does not prevent HPV 6/11 infection and disease	16,880	16,957
Cost of vaccination series: +20%	11,591	12,872
Coverage with vaccination=50%	Weakly dominated^‡	9,826
Coverage with vaccination=70%	9,295	10,249
Costs of diagnosis and treatment: +20%	9,478	10,549
Costs of diagnosis and treatment: -20%	9,675	10,743
Lower QALY weights – HPV disease has higher impact on quality of life	7,356	8,291
Higher QALY weights– HPV disease has lower impact on quality of life	13,716	14,868
Discount rate=0%	Weakly dominated^c	2,435
Pessimistic scenario	Weakly dominated^c	40,083

€1 ∼ HUF254.59.

*Routine vaccination compared to the no vaccination strategy.

^†Routine vaccination + 12–24 year-old catch-up compared to Routine vaccination.

^‡Strategy A (routine vaccination) is dominated if there is another strategy, B (routine vaccination + catch-up programme), that is both more effective and more efficient than strategy A.

Discussion

In this study, a mathematical model was used to project the potential epidemiological and economic impact of two HPV 6/11/16/18 vaccination programmes in Hungary. The two vaccination strategies evaluated included one that targets 12-year-old girls and another that also includes a temporary 5-year catch-up programme for girls and women aged 12–24 years. Generally, the results from the model demonstrated that a quadrivalent HPV vaccine programme that includes the vaccination of 12-year-old girls or a routine vaccination with a catch-up programme can be cost effective (€9,577 per QALY and €10,646 per QALY, respectively). These findings are consistent with other cost-effectiveness analyses that have generally shown that vaccination of 12-year-old girls can be cost effective^27,28 and may be useful for HPV vaccine policy decisions in Hungary as well as neighbouring countries.

Since no explicit inclusion cost-effectiveness threshold exists in Hungary either for drugs or for vaccination programmes²⁹, the cost-effectiveness criteria established by the World Health Organization (WHO) were used. WHO defines cost effectiveness as follows: if the ICER is less than gross domestic product (GDP) per capita, the intervention is considered very cost effective; if it is one to three times GDP per capita, it is considered cost effective; and if it is more than three times the GDP per capita, it is considered not cost effective³⁰. In 2006, the GDP per capita in Hungary was approximately €12,098³¹. Therefore, both vaccination strategies would be considered cost effective based on these criteria. In addition, both incremental cost-effectiveness ratios are within the range of cost-effectiveness ratios for other commonly accepted medical technologies³².

A key result from this analysis was that both vaccination strategies can significantly reduce the burden of HPV disease in the short term. Figure 2 showed that the majority of the early cost reductions can be attributed to protection against infection with HPV types 6 and 11.

Sensitivity analyses were also important for generating insights for developing an HPV vaccination policy. One of the key variables identified in the sensitivity analysis was the importance of duration of vaccine protection. It was found that as the duration of protection decreased, the cost-effectiveness ratios increased. Another insight from the sensitivity analysis was the importance that protection against genital warts (HPV 6 and 11) played in reducing the burden of HPV diseases. For example, the cost-effectiveness ratio for routine vaccination increased approximately 75% to €16,880 per QALY gained in the absence of any protection against HPV 6/11 infection. Finally, the cost-effectiveness ratio of the catch-up strategy increased significantly to €40,083 per QALY gained for the pessimistic scenario in the sensitivity analysis (the routine vaccination strategy was weakly dominated in this scenario).

Although the limitations of this model have been previously described in detail²⁰, there are four key limitations: first, although the model predictions generally fell within the range of epidemiological values observed for cervical cancer, epidemiological data on genital warts were not available for calibration. The projected rates are similar to what has been reported for the UK³³. Second, only four HPV disease types (i.e. 6, 11, 16 and 18) were modelled. However, this study did not account for the additional benefits that might be realised through protection against HPV diseases associated with infection with other HPV types not directly targeted by the vaccine (i.e. cross-protection) providing a conservative estimation of both efficacy and cost effectiveness³⁴. Third, this study also did not account for other potential benefits of vaccination that would have improved the cost-effectiveness ratio, such as protection against vulvar and vaginal pre-cancers and cancers³⁵, protection against head and neck cancers³⁶, protection against recurrent respiratory papillomatosis³⁷ and mortality and productivity costs (i.e., indirect costs). Fourth, it was assumed screening would not change over time with HPV vaccination (see Table S2).

Conclusion

The results from this model suggest that in a setting of cervical cancer screening in Hungary, a prophylactic quadrivalent HPV (6/11/16/18) vaccine programme can substantially reduce the incidence of cervical cancer, CIN and genital warts, improve quality of life and survival, and be cost effective when implemented as a strategy that routinely vaccinates 12-year-old girls.

Transparency

Declaration of funding: This study was funded by Merck, North Wales, Pennsylvania, USA.

Declaration of financial/ other relationships:

E.J.D. and E.H.E. are employees of Merck and own Merck stock. LN is an employee of MSD, Budapest, Hungary and owns MSD stock. AB works as a paid consultant for MSD, Budapest, Hungary.

Some peer reviewers receive honoraria from JME for their review work. Peer reviewers 1 and 2 have disclosed that they have no relevant financial relationships.

Acknowledgement

The authors wish to thank Ralph P. Insinga from Merck for his helpful comments on the manuscript, and Karen Collins, BS, of JK Associates, Inc., for assistance with manuscript preparation.