Content uploaded by Karl Ulrich Petry
Author content
All content in this area was uploaded by Karl Ulrich Petry on Dec 20, 2017
Content may be subject to copyright.
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=ierp20
Download by: [Sanofi Pasteur MSD], [Mathieu Uhart] Date: 04 July 2016, At: 00:30
Expert Review of Pharmacoeconomics & Outcomes
Research
ISSN: 1473-7167 (Print) 1744-8379 (Online) Journal homepage: http://www.tandfonline.com/loi/ierp20
An estimate of the public health impact and cost-
effectiveness of universal vaccination with a 9-
valent HPV vaccine in Germany
Nathalie Largeron, Karl Ulrich Petry, Jorge Jacob, Florence Bianic, Delphine
Anger & Mathieu Uhart
To cite this article: Nathalie Largeron, Karl Ulrich Petry, Jorge Jacob, Florence Bianic,
Delphine Anger & Mathieu Uhart (2016): An estimate of the public health impact and cost-
effectiveness of universal vaccination with a 9-valent HPV vaccine in Germany, Expert Review of
Pharmacoeconomics & Outcomes Research, DOI: 10.1080/14737167.2016.1208087
To link to this article: http://dx.doi.org/10.1080/14737167.2016.1208087
Accepted author version posted online: 01
Jul 2016.
Published online: 01 Jul 2016.
Submit your article to this journal
Article views: 1
View related articles
View Crossmark data
1
Publisher: Taylor & Francis
Journal: Expert Review of Pharmacoeconomics & Outcomes Research
DOI: 10.1080/14737167.2016.1208087
An estimate of the public health impact and cost-effectiveness of universal vaccination with a 9-valent
HPV vaccine in Germany
Nathalie Largeron [1], Karl Ulrich Petry [2], Jorge Jacob [3], Florence Bianic [4], Delphine Anger [4],
Mathieu Uhart [1]
1. Sanofi Pasteur MSD - Health Economics, 162 avenue Jean Jaurès, CS 50712, Lyon Cedex 07, Lyon
69367, France
2. Klinikum Wolfsburg – OBGYN, Sauerbruchstr. 7, Wolfsburg 38440, Germany
3. Mapi Group Ringgold - RSW&A, 3rd floor, Beaufort House, Crisket Field Road, Uxbridge, London
UB1 8QG, UK
4. Mapi Group - RSW&A, 41 rue des Trois Fontanot, Nanterre 92000, France
Corresponding author: Mathieu Uhart muhart@spmsd.com
Funding
This manuscript was funded by Sanofi Pasteur MSD (SPMSD).
Declaration of Interest
N Largeron and M Uhart are employees of SPMSD. KU Petry has an institutional grant from SPMSD and
works as an occasional advisor for SPMSD and Roche Diagn. J Jacob, F Blanc and D Anger are employees
of Mapi Group, which received funding from SPMSD to conduct this study. The authors have no other
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
2
relevant affiliations or financial involvement with any organization or entity with a financial interest in or
financial conflict with the subject matter or materials discussed in the manuscript apart from those
disclosed.
Abstract
Introduction: Since 2007, the German Standing Vaccination Committee recommends HPV vaccination
for girls aged 12-17 with a 2- (Cervarix®) or 4-valent (Gardasil®) vaccine. A 9-valent vaccine (Gardasil 9®)
recently received a European market authorization in 2015.
Methods: A dynamic transmission model was calibrated to the German setting and used to estimate
costs and QALYs associated with vaccination strategies.
Results: Compared to the current vaccination program, the 9-valent vaccine extended to boys shows
further reductions of 24% in the incidence of cervical cancer, 30% and 14% in anal cancer for males and
females, as well as over a million cases of genital warts avoided after 100 years. The new strategy is
associated with an ICER of 22,987€ per QALY gained, decreasing to 329€ when considering the vaccine
switch for girls-only.
Conclusion: Universal vaccination with the 9-valent vaccine can yield significant health benefits when
compared to the current program.
Keywords: cost-effectiveness, Germany, HPV, cervical cancer, vaccination
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
3
1. Introduction
Human papilloma virus (HPV) infections are responsible for nearly all cases of cervical cancers (4,647 per
year in Germany), 19% of vulvar cancers (605 per year), 71% of vaginal cancers (360 per year), 88% of
anal cancers (1,627 per year) and 90% genital warts (113,039 per year).[1, 2] HPV is a family of viruses
that infect epithelial tissues including skin and moist membranes.[3] For this reason, HPV is one of the
most common sexually transmitted infections.[3] Among the over 100 different types of HPV identified,
some are referred as high-risk because they increase the probability of developing anogenital cancers.
This group includes types 16, 18, 31, 33, 45, 52, and 58. High-risk types 16 and 18 alone contribute for
70% of invasive cervical cancers and high-grade cervical intraepithelial neoplasia (CIN).[4] Low-risk types
6 and 11 account for 85% of genital warts cases.[5, 6]
The economic burden that HPV infections and related diseases cause in Germany is heavy on healthcare
system and society.[7, 8] Screening and vaccination strategies against HPV represent the primary
method to prevent HPV-related pre-cancers and cancers.
Early detection of cervical cancer was introduced in Germany is 1971. Currently, all woman aged 20
years or older are entitled to an annual free-of-charge Pap smear, with no age limit for cessation. The
German screening program is a self-referring screening policy without an invitation and regulation
system. Insurance data shows that around 15 million smears are currently taken every year, which
implies a compliance with annual screening of around 50% of the target population.[9, 10] In 2015, the
national health institution Gemeinsamer Bndesausschuss (G-BA) decided to change the screening policy
to include HPV testing every five years for woman older than 30 years. However for an interim period,
these woman will be allowed to choose annual Pap smear screening instead. Since there is a lot of
uncertainty around the impact of this policy in the cost of screening in Germany, the base case analysis
is based on the current screening practice.[11]
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
4
Currently two vaccines are commercially available: Gardasil® (Sanofi Pasteur MSD) which is a
recombinant vaccine with protection against HPV types 6, 11, 16, and 18, and Cervarix®
(GlaxoSmithKline) that offers protection against the HPV types 16 and 18.[12-14] While Cervarix® is
indicated for the prevention of cervical pre-cancerous lesions and cervical cancer caused by HPV types
16 and 18, the 4-valent vaccine indication also includes vulvar, vaginal and anal cancers, and respective
precancerous dysplastic diseases (Vulvar intraepithelial neoplasia (VIN), vaginal intraepithelial neoplasia
(VaIN) and anal intraepithelial neoplasia (AIN)) as well as genital warts.[13, 14] In 2007, the German
Standing Vaccination Committee (STIKO) recommended vaccination with either Gardasil® or Cervarix® in
a three-dose schedule for girls aged 12-17. The vaccines are free of charge in the target age group,
although reimbursement is also available for women aged 18-26.[15-17] Although both vaccines are
equally reimbursed, Gardasil® holds the majority of the market share (up to 90%).[9] A recent update
shifted the target age group to 9-14, and limited the recommended number of doses to two. A third
dose is scheduled for girls 14 or older or for catch-up vaccinations.[18] Official figures for vaccination
coverage rates (VCR) in light of the new recommended age-classes are not yet available. The latest
figures from Robert Koch Institute (RKI) show compliance rate of 55.6% among the target population (12
to 17 year old girls).[19]
Gardasil 9® is a new vaccine that offers protection against HPV types 6, 11, 16, 18, 31, 33, 45, 52 and 58.
Gardasil 9® is expected to prevent infection from the majority of the HPV types with carcinogenic
properties, since types 16, 18, 31, 33, 45, 52 and 58 are amongst the most commonly detected.[20]
Hartwig et al (2015) estimate that 89% of the HPV positive cervical, vaginal vulvar and anal cancers are
attributable to the 7 oncogenic HPV types for which Gardasil 9® offers protection, whereas the high risk
types 16 and 18 included in the previous vaccines (Gardasil® and Cervarix®) account for 75% of HPV
related cancers. Therefore, the new vaccine significantly broadens the protection offered by the first
generation of HPV vaccines.[1]
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
5
In the United States, Gardasil 9® has been approved by the Food and Drug Administration in December
2014, for females aged 9-26 and males aged 9-15.[21] The Committee for Medicinal Products for Human
Use of the European Medicine Agency recommended the marketing authorisation of Gardasil 9® that
received approval from the European Commission on June 2015.[22] The indication is for individuals of
at least nine years of age and the dosing schedule is fully aligned with the national recommendations on
HPV vaccination, with 2 doses below 15 years of age and 3 doses above.
While HPV vaccination was originally focused on the prevention of cervical cancer and targeted girls
only, more and more countries recommended the HPV vaccination for girls and boys. Indeed, the
burden of HPV-related diseases is heavy on the male population. In Germany, about 800, 740 and
12,000 new cases of penile, anal and head & neck (H&N) cancers are reported each year.[2] The
economic burden is important as well. It was shown in Europe that non-cervical cancers accounted for a
substantial proportion of the economic burden of HPV-related cancers, and that this burden was mainly
driven by men (about 70%).[7] Although female vaccination can indirectly protect males, it does not
reach the homosexual population. In addition, evidence shows that a reasonable protection in the male
population is only achievable when the VCR is high on females.[9] However, to increase the coverage in
females may be challenging especially in countries like Germany where the vaccination is done by self-
referral. Furthermore, some experts argue that universal (boys and girls) vaccination can greatly
contribute to contain the virus propagation and that it is the only way to ultimately eradicate it.[23]
Accordingly, universal vaccination is being recommended and introduced in several European countries
(Austria, Norway, Switzerland and 9 regions in Italy).
Cost-effectiveness studies of the 9-valent vaccine in the United States showed that universal vaccination
is likely to be cost saving when compared to the current strategy.[24-26] In Canada, the HPV-ADVISE-
CAN model showed that the 9-valent vaccine is cost effective with respect to the 4-valent vaccine if the
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
6
price increment does not exceed $24.[27] A recently published paper concluded that universal
vaccination with the 9-valent vaccine in Austria could reduce the cervical cancer incidence by an
additional 17%, compared to vaccination with the 4-valent vaccine and be a cost-effective or cost-saving
strategy, depending on the price per dose.[28]
In Germany, cost-effectiveness studies of the new vaccine are not available in the literature; however
studies regarding HPV vaccination have been published. Hillemanns et al. (2008) used an empirically
calibrated Markov model of the natural history of HPV to assess the cost-effectiveness of the 4-valent
vaccine administered to 12-year-old girls alongside existing cervical screening programmes in
Germany.[29] The authors estimated that 2,835 cervical cancer cases and 679 deaths could be
prevented in a cohort of 400,000, at an incremental cost-effectiveness ratio (ICER) of 10,530€ per
quality-adjusted life-years (QALY) gained.[29] Schobert et al. (2012) used an HPV dynamic transmission
model in the German context.[30] They found that vaccination with the 4-valent vaccine of girls aged 12-
17 was cost-effective (ICER 5,525€/QALY) and that the ICER would increase substantially (10,293€/QALY)
when the vaccine effects on HPV6/11 diseases were excluded.[30] Horn et al. (2013) developed a
dynamic mathematical model for the natural history and transmission of HPV infections. They found
that over 100 years, a 4-valent HPV vaccination program could prevent around 37% of cervical cancer
cases assuming a 50% VCR in the 12 years old girls.[9]
Although cost-effectiveness evidence for a new technology is not mandatory in Germany, the STIKO
might consider available studies in their recommendations. In accordance, the scope of the current
study is to provide an epidemiological and cost-effectiveness analysis of the implementation of the 9-
valent vaccine in Germany in comparison of the current practice. As the 4-valent vaccine holds the vast
majority of the German HPV vaccine market, the current practice is represented in our model by a 4-
valent vaccination program targeted to girls aged from 9 to 17 years old.
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
7
2. Methods
2.1. Mathematical model
Elbasha et al. (2007) developed a deterministic SIRS (Susceptible-Infected-Recovered-Susceptible) model
to assess the cost-effectiveness of HPV vaccination.[31] The first version of the model simulated the
epidemiology of HPV 6/11/16/18 infections and related diseases (CIN, cervical cancer and genital warts).
A German adaptation of this model was performed by Schobert et al. (2012).[30] The model structure
was first updated in 2010 to take into account all HPV-related diseases, adding to the original model the
indication on vaginal, vulvar, anal, H&N and penile cancers as well as respiratory papillomatosis) and
more recent data on the natural history of the infections.[32, 33] In 2014 the effect of HPV types 31, 33,
45, 52, and 58 were added to the original four (types 6, 11, 16, and 18) in order to be suitable for the 9-
valent vaccine. However, the model uses a conservative assumption considering that the types
33/33/45/52/58 are only responsible for cervical diseases and anal cancer. The contribution of the five
additional types to the burden of other diseases (vaginal, vulvar, penile, H&N cancers, genital warts, and
recurrent respiratory papillomatosis [RRP]) is not modelled currently.
The current model features a high level of detail, involving hundreds of inputs and several thousand
ordinary differential equations (ODEs). Sub-models are evaluated successively: an initialization
demographic and epidemiologic model informs an economic model that can compare vaccination
strategies and evaluates the epidemiological impact and cost effectiveness of the implementation of a
new vaccination program.
The demographic model defines the characteristics of the population being simulated and describes
how persons enter, age, and exit various categories. The population is divided into 17 age groups,
classified into three levels of sexual activity. Persons then move between successive age groups and exit
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
8
the model upon death. Cancer patients have an additional age and stage-dependent death rate. Patients
with CIN or genital warts do not face an additional risk for death.
The epidemiologic model simulates HPV transmission and the occurrence of consequential diseases. The
acquisition of infection and progression of persons from infection to disease follow a natural history
structure that remains similar across the versions. The population was divided into distinct
epidemiologic categories, according to the person’s status with respect to infection, disease, screening,
and treatment over time. The epidemiologic module includes one HPV6-specific model (RRP, genital
warts and CIN1), one HPV11-specific model (RPP and genital warts), one separate model for each
disease related to HPV16 or HPV18 (cervical precancerous lesion and cancer, vulvar precancerous lesion
and cancer, vagina precancerous lesions and cancer, anal precancerous lesion and cancer, penile
precancerous lesions and cancer, and H&N cancer). The 5 additional types (HPV31, 33, 45, 52 and 58)
have been merged together and one separate model has been created for them: one for cervical
diseases and another one for anal diseases.[28]
Finally the economic model considers the implementation of screening and vaccination strategies that
will impact the infection transmission among the population and the development of the diseases. By
assigning costs and utilities to each health state (defined by the person’s status), the model will
generate cost-effectiveness results along with the epidemiological results.
The adaptation to the German setting was carried out by informing the model with German-specific
inputs. This study evaluates two strategies using a two-dose schedule of the 9-valent vaccine: a universal
vaccination program covering equally boys and girls in the recommended age group, and a girls-only
vaccination scenario. Both are compared to the current practice of a two-dose vaccination with the 4-
valent vaccine covering girls aged between 9 and 17 years.
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
9
2.2. Epidemiological model inputs
The parameters are divided in five distinct sections: demographics, sexual behaviour, disease and
treatment patterns, screening, and natural history of disease. The parameter groups are summarised in
Table 1 along with the references.
2.2.1.1. Demographics
Age-stratified population figures and all-cause mortality rates were retrieved from the German Federal
Statistical Office (Statistisches Bundesamt).[48, 49]
2.2.1.2. Sexual behaviour
German-specific sexual behavioural data is not available in the literature. Previous German and Austrian
cost-effectiveness models relied on the second UK National Survey of Sexual Attitudes and Lifestyle
(NATSAL-2) [9, 30, 50], as sexual behaviour patterns are similar in Germany and the UK.[51] As a new
version of this study (NATSAL-3) was published in 2014, we used the updated figures in the model. Since
results were not reported per model requirements, NATSAL authors were contacted.[35] The amount of
sexual mixing among members of different age cohorts (a value between 0 and 1 with 0 representing no
mixing) and the amount of sexual mixing among members of different sexual activity groups requested
in the model were extracted from Elbasha et al. (2010).[32, 33] The sexual behaviour parameters are
displayed in Table 2.
2.2.1.3. Natural history of disease
We assume that the natural history of the disease in Germany follows the same patterns as in the US, as
the WOLVES study shows comparable data between California and Lower Saxony.[52-54] Therefore we
relied on the extensively calibrated parameters previously described and reported by Elbasha et al
(2010).[32] For the transmission rates, calibration techniques were used to obtain the best set of
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
10
parameters for Germany. The model parameters on the natural history of disease can be found in
Appendix.
2.2.1.4. Disease and treatment patterns
Calibration was used to estimate the parameters related to the percent of treated CIN, VaIN, VIN, and
Carcinoma in Situ (CIS), as well as the percentage of females with cancer recognising their symptoms
and seeking treatment.
The hysterectomy rates were derived from a 2013 publication by the German Federal Statistical Office
(DRG related hospital statistics - Fallpauschalenbezogene Krankenhausstatistik) reporting the total
number of hysterectomies performed. We calculated the hysterectomy rates using population numbers
by age class from 2013.[36, 49]
2.2.1.5. Cancer mortality
The model requires HPV-related cancers associated mortality (i.e. the fraction of individuals with cancer
who are expected to die over the course of 1 year) stratified by age and stage (local, regional, and
distant). Since survival data were only available by age class, an extrapolation to estimate the stage-
stratified data as required by the model was performed. UK data from Cancer Research UK was used to
calculate the relative risk of each stage and applied to the German specific survival statistics (assumed to
be representative of regional stage) in order to calculate survival rates for local and distant cancers.[55]
To estimate HPV-related cancers associated mortality, we used survival data from the European Cancer
Registry (EUROCARE-5) and the German Centre for Cancer Registry Data (ZfKD).[2, 56] The ZfKD was
preferred where possible because data were more recent and more representative of the German
population.[2] EUROCARE-5 data were used for H&N and penile cancers. Assumptions were necessary to
conform to the model inputs: mortality for vaginal cancer was assumed equal to vulvar cancer, mortality
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
11
for anal cancer was assumed equal to colon and rectum pooled. The five-year mortality rates were then
converted to one-year death probabilities (Table 3).
2.2.1.6. Screening
Age-specific screening adherence was calculated from a report from the Central Institute for ambulatory
health care in the Federal Republic of Germany.[39] Those values were used to inform the percent of
females screened for cervical cancer in the model (Table 4). From the same source we retrieved the
percentage of females receiving a follow-up screening test after abnormal Pap smear diagnostics. The
percentage of woman screened at least once every three years was found in a retrospective cohort
study from Rϋckinger et al. (2008).[38]
2.2.2 Economic model inputs
Inputs of the economic model include the vaccination strategy, vaccine properties, costs, and utilities.
2.2.2.1. Vaccination strategy
The most recent report on the German Health Interview and Examination Survey for Children and
Adolescents (KiGGS Wave 1) of the Robert Koch institute was used to inform the vaccination strategy
section.[19] As the recommendation regarding HPV vaccination shifted recently, (from the age group 12-
17 to 9-14, with catch up vaccination until the age of 17) coverage rates in light of the new vaccination
schedules are not yet available. Thus we adapted the figures from the KiGGS Wave 1 to fit the new
recommended age classes as shown in the Table 5.
A compliance rate of 90% was assumed for the second dose.
2.2.2.2. Vaccine properties
Clinical trial data provide values for the prophylactic efficacy of the vaccine.[57-62] The duration of
protection was assumed to be lifelong in the base case, the relative effectiveness of the vaccine was
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
12
assumed as zero if less than the full regimen of two doses is received, and no herd immunity was
considered. Table 6 summarizes the vaccine efficacy parameters related to the protection against
transient and persistent infections.
2.2.2.3. Costs
All costs collected from the literature were inflated to 2014€, using the German Consumer Price Index
(CPI).[64] A discount rate of 3% is considered for costs as per indication of the Institute for Quality and
Efficiency in Health Care (IQWiG). All costs used in the model are listed in Table 7.
The cost of vaccination in Germany varies across federal states according to the availability of office
supply. For the purpose of this analysis, a cost of 140€ for a dose of was considered for the 4-valent
vaccine.[65] This corresponds to a tenth of the price of a 10 dose pack (1400€). A cost of 146.5€ was
used for the 9-valent vaccine corresponding to a tenth of the price of a 10 dose pack.[66] Vaccination
administration costs also differ across federal states. An average administration cost of 9€ per dose was
considered for both vaccines.
No publications reporting German specific costs per episode of care of cervical cancer were identified.
Siebert et al (2004) reported 5-year treatment costs for cervical cancer stratified by stage.[67] Schobert
et al (2012) used these figures to inform cost per episode of care parameters, therefore, we applied the
same assumption.[30] Cost of CIN 1, 2, 3, VaIN and CIS were collected from two German studies.[68, 69]
Costs per episode of care for non-cervical cancers could not be found as literature on the economic
burden of non-cervical cancers in Germany is scarce. In a cost-effectiveness study from Brisson et al
(2013) [70], the authors calculated the ratio between costs of each non-cervical cancer and cervical
cancer using published literature. They applied these ratios to Canadian cervical cancer cost data to get
estimates for vaginal, penile and oropharyngeal cancers. We retrieved the ratios calculated in this
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
13
publication and applied them to the German cervical cancer costs to estimate the costs of the remaining
HPV related cancers.
Unit costs of screening and diagnostic tests were retrieved from a cost-effectiveness model from
Hillemans et al (2008).[29]
2.2.2.4. Health-related quality of life
Hinz et al. (2005) used the EQ-5D questionnaire in a large cohort (N=2,022) between 16 and 93 years
old, which informed the utilities for the general population.[71] No German-specific utilities for health
states were found, hence the health utility values for cancer patients were derived from several sources.
In the absence of UK-specific stage-stratified data in the population with HPV-related diseases, a
combination of best available UK and US data were used to calculate the required utilities. (Table 7)
2.3. Model calibration and validation
The ZfKD provided most of the calibration targets required to validate the inputs. Annual numbers of
incident cases and deaths as well as incidence and mortality rates for all cancers were retrieved from
this source.[2] The registry did not report data for H&N cancers, therefore a cluster of sites was used as
a proxy (HPV-related oral cancers, oropharynx cancers and larynx cancers).[75] Genital warts values
were retrieved from Kraut et al. (2010).[75, 76] The proportions of diseases attributable to HPV infection
were collected from two publications.[1, 77] Regarding the incidence of CIN, we chose to use incidence
rates observed in the UK, as Germany does not have a systematic screening registry and the existing
literature shows that the epidemiology of these lesions are similar across both countries.[53, 54] These
incidence rates were calculated using the most recent statistics for the Cervical Screening Programme
and population figures in the UK.[78, 79]
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
14
The epidemiological model produces the incidence rates of the diseases related to the 4 HPV types
included in the 4-valent vaccine and to the 9 types included in the 9-valent vaccine. As a consequence,
the overall incidences collected in the literature were adjusted using the HPV-attribution reported in
Hartwig et al (2015), as shown in Table 8.[1]
In the calibration process, the model inputs were iteratively modified in order to get model outcomes
closer to the validation targets. The targets with greatest impact on overall cost-effectiveness and the
targets with best quality data were prioritized. Since the natural history parameters used in the model
followed an extensive calibration process in the original US model, they were not modified. We focused
on other parameters such as transmission rates. To fine-tune the results to match each target, local
variables such as mortality rates and proportion of individuals of seeking treatment, were adjusted.
2.4. Model analyses
With the set of inputs previously described, we estimated the total number of events, incidence and
mortality of HPV-related diseases (cervical cancer, CIN, anal cancer and genital warts) as well as costs
and QALYs per person over a time horizon of 100 years. Incremental ICERs were then calculated with the
quotient: Incremental costs / Incremental QALYs.
In Germany, HPV vaccination is delivered through the statutory health insurance (SHI) plans, and
purchase is done through several sickness funds. All analyses were performed from the SHI perspective.
A deterministic sensitivity analysis (DSA) was conducted in order to access the robustness of the results.
The parameters modified in the DSA were: vaccine price, VCR, duration of protection, utilities, and
discount rates, inclusion of cross protection and inclusion of H&N and penile indication. Sensitivity
analyses were performed deterministically, modifying the value of one base case parameter at a time
and recording the corresponding ICER.
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
15
Additional exploratory analyses were performed in order to evaluate the cost-effectiveness of the 9-
valent vaccine against the 2-valent vaccine.
3. Results
3.1. Model Calibration
The summary of the incidence targets collected from the literature and the model outcomes are
summarized in Table 9. After calibration, the model shows a good fit on the estimated incidence of
cervical, anal and penile cancers, as well as genital warts. The main obstacle of the calibration relates to
the incidence of CIN, which is significantly underestimated by our model.
In addition, the incremental proportion of CIN cases attributable to the 5 additional genotypes included
in the 9-valent vaccine is also underestimated. Hartwig et al. (2015) showed that the HPV6/11/16/18 are
responsible for about 24% of CIN1 and 45% of CIN2 cases whereas HPV6/11/16/18/31/33/45/52/58
targeted by the 9-valent vaccine account for 48% and 82% of CIN 1 and CIN 2+, respectively.[1] This
means that the 9-valent HPV infections were responsible for twice as much of CIN1 and 1.8 times more
for CIN2+ compared to the 4-valent. Our calibrated model estimates that the 9-valent HPV accounts for
1.3 times more for CIN1 and 1.25 times more for CIN2+ the 4-valent HPV.
3.2. Epidemiological results
Figure 1 and Figure 2 show the epidemiological impact of the different scenarios over a time horizon of
100 years. The curves show that the incidence and mortality rates of HPV-related diseases stabilize
before the end of the analysed time horizon (100 years).
These results show the added benefits of the 9-valent vaccine. Considering the scenarios with
vaccination coverage only for girls, 9-valent vaccine is associated with a reduction of 31,500 additional
cases and 7,408 additional deaths of cervical cancer over 100 years. Furthermore, when compared to
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
16
the current vaccination program, the new vaccine shows an additional reduction of 21% and 19% on CIN
1 and CIN 2+ incidence, respectively. This represents a reduction of 234,899 cases of CIN 1 and 399,410
cases of CIN 2+ over 100 years, as reported in Table 10.
Adopting a vaccination program with universal coverage is associated with further epidemiological
benefits. This reflects the indirect benefits, through herd immunity effects, of vaccinating boys. Over 100
years, the universal coverage scenario is associated with a reduction of 14,954 cases of cervical cancer,
90,330 cases of CIN 1 and 171,603 cases of CIN 2+ when compared to a girl-only coverage with the same
vaccine. When compared to the current vaccination program, universal coverage with the 9-valent
vaccine can avoid up to 46,454 cases of cervical cancer, 325,229 cases of CIN 1 and 571,013 cases of CIN
2+, as shown in Table 11. The universal coverage also shows added benefit in the incidence of genital
warts and anal cancer; the model estimates that anal cancer incidence can be reduced by 12% and 29%
in females and males, respectively. In addition, we observe a reduction of 20% and 22% in the incidence
of genital warts for males and females respectively, when comparing 9-valent universal vaccination with
the current practice. This corresponds to 1.5 million cases of genital warts and 8,456 cases of anal
cancer avoided over 100 years.
3.3. Cost-effectiveness results
The model estimates showed that switching for the 9-valent vaccine in Germany is highly cost-effective
with an ICER of 329€/QALY (Table 12). The ICER increased to 22,987€/QALY when universal vaccination
with the 9-valent vaccine was considered.
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
17
Both strategies show an estimated ICER below the threshold commonly used by the National Institute
for Health and Clinical excellence (NICE) in the UK (£30,000/QALY or 40,000 €/QALY).1 We use this value
as a reference since there is not a fixed threshold for the ICER in Germany.
When considering universal coverage with the 9-valent vaccine, there are considerable health benefits
and cost savings in all diseases considered in the model. CIN and anal cancer are associated with the
most significant cost savings and additional health benefits. In the instance of girls only coverage, the
benefits can only be seen in cervical diseases and anal cancer.
The effect of the new vaccine price on its cost-effectiveness is graphically represented in Figure 3. All
threshold analyses were performed considering the base case price of the 4-valent vaccine (140 €). The
results show that changing the vaccination strategy to the 9-valent vaccine when targeting girls only
remains under the NICE cost effectiveness threshold if incremental price of the new vaccine does not
exceed 126€. The same strategy can be cost-saving if the price increment per dose of the new vaccine
does not exceed 5€. As for universal vaccination the threshold analysis shows that this strategy can be
cost effective if the price increment per dose of the new vaccine does not exceed 60€.
3.4. Sensitivity analyses
The results of the DSA are summarized in the tornado diagrams represented in Figure 4. The ICER of the
switch to a 9-valent HPV vaccine remained below the NICE cost-effectiveness threshold of 40,000€/QALY
in all the sensitivity analyses.
Decreasing the duration of protection of the vaccines (assumed lifelong in the base case) to 20 years
improves the ICER in both comparisons. For the scenarios with girl-only vaccination (HPV9 Girls vs HPV4
Girls), the 9-valent vaccine becomes a cost saving strategy while the same alteration causes a reduction
1 Converted to Euros. Rate: 1.35869 Reference: http://www.xe.com/ date: 06/01/2016 17:00 UTC and rounded to
40,000
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
18
in the ICER from 22,987 €/QALY in the base case to 14,827€/QALY in the scenario with 9-valent
universal vaccination. This alteration produces higher costs and lower QALYs per person for the three
scenarios.
Assuming a lower discount rate for outcomes (1.5% in spite of 3% in the base case), causes an overall
decrease of the ICER for both comparisons. The lower discount rate translates in higher QALYs per
person while the costs remain unchanged, thus resulting in a lower ICER. A lower discount rate for
outcomes has greater impact on universal vaccination compared to the current practice, where the ICER
drops to 8,748 €/QALY.
Boosting the VCR to 70% increases the ICER in both comparisons. HPV 9 girl-only vaccination vs. current
practice increases to 707 €/QALY whereas universal vaccination with the 9-valent vaccine vs. current
practice increases to 27,986 €/QALY. The increment in the costs due to the additional number of
vaccines administered is higher than the QALY benefits, regardless of the target population of the
vaccine.
Using utilities from Elbasha et al. (2010) causes a small increase in the ICER for both comparisons. The
utilities associated with the different health states are lower than the ones used in the base case.
Therefore the QALYs per person are lower.
Including all diseases simulated by the model (base case only includes the diseases disclosed in the
vaccine label), accounts for the costs and QALYs associated with H&N and penile cancers as well as RRP.
The benefits of a universal program with the 9-valent vaccine are greatly enhanced if the analysis
includes the additional diseases, as the ICER decreases to 14,286 €/QALY. If the vaccination is targeted
only to girls, the inclusion of the additional diseases has a negligible impact on the ICER.
(Figure 4)
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
19
3.5. Scenario analysis
The additional scenario analysis results show that the 9-valent vaccine is a dominant strategy against the
2-valent vaccine when only girls are covered in the vaccination program. Similarly to the base case, the
ICER increases to 11,596 €/QALY for a universal vaccination program with the 9-valent vaccine,
compared to girls only coverage with the 2-valent vaccine.
4. Discussion
We present the first cost-effectiveness analysis of the implementation of a vaccination program with the
new 9-valent HPV vaccine extended to boys in Germany. All analyses were performed in a model
originally designed for the US. The adaptation for the German context was achieved through an
extensive data collection and calibration. After the calibration process, the model accurately estimated
indicators such as incidence and mortality of HPV-related diseases in Germany. The impact on health
outcomes and costs were estimated through various scenarios allowing testing of different vaccination
strategies and assumptions.
The results from this analysis show that, in the current German setting, replacing the current 4-valent
vaccine with the new 9-valent technology in the vaccination program is highly cost-effective with an
ICER of 329 €/QALY. The very low ICER (329 €/QALY), along with the assumption that the coverage rate
would remain the same (i.e. same number of vaccine doses administered) and the very similar price of
both vaccines, suggests that the net budget impact of the switch to the 9-valent vaccine would be low. If
the vaccination programme is extended to boys (in the same age groups recommended for girls) the
ICER remains cost-effective with a ratio equal to 22,987 €/QALY. It is noteworthy the ICER reported in
the base case considers the indicated diseases only. The inclusion of all diseases in the analysis
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
20
decreases the ICER between universal vaccination with the 9-valent vaccine and the current practice to
14,286 €/QALY.
These results are in line with the conclusions reported in three US studies presented during the ACIP
(Advisory Committee on Immunization Practices) meeting. All studies estimated that a universal
vaccination programme with the 9-valent vaccine was likely to fall within an acceptable range of cost-
effectiveness or even become cost saving compared to the current universal vaccination programme
with the 4-valent vaccine.[24-26] In Canada, the new vaccine was shown to be cost-effective for a price
increment lower than +CAN$24.[27]
Our model predicts that replacing the current vaccine recommendation with the 9-valent vaccine could
lead to further reductions of 17%, 21% and 19% in the incidence of cervical cancer, CIN 1 and CIN 2+,
respectively. Universal vaccination would allow to further reduce the incidence in 24% for cervical
cancer, 21% for CIN 1 and 24% for CIN 2+. Several studies regarding the effects of the 9-valent vaccine
report the same trend, although they do not concern the German population. [24, 25]
The effect of universal vaccination in Germany was discussed by Horn et al. (2013). Vaccination of girls
only was generally more effective than vaccination both genders. They estimated that 83,567 cases of
cervical cancer were prevented assuming a VCR of 40% among girls, while vaccinating 20% of both boys
and girls resulted in 75,152 cases avoided.[9] These results are consistent with our model, which
predicts additional health benefits in the scenario where a higher coverage rate (70%) is assumed for
girls only, compared to the base case scenario with universal vaccination (cumulative VCR of 55.6% at
the age of 17 years for each gender). However, increasing the vaccination coverage in girls may be very
difficult to achieve since the majority of vaccines in Germany are administered by private physicians,
rather than school delivery programs as it happens in other European countries. Furthermore, universal
vaccination would provide additional benefits by protecting men exposed to male partners and
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
21
unvaccinated females. On the other hand, a universal vaccination schedule against HPV would allow for
a more efficient way to stop the virus transmission and ultimately achieve the virus eradication.[23]
Lastly, universal vaccination may contribute to raise awareness to the prevention of HPV-related
diseases. Overall, universal vaccination is justified by epidemiological, equity and vaccination efficiency
factors. With this study we demonstrate that it is also economically viable.
Regarding screening strategies, the HPV-DNA test was not considered in the model as it was only
recently recommended; the Pap smear test remains the current practice and data on the
implementation of the HPV-DNA test are still scarce; moreover the model does not allow for a flexibility
in the use of mixed screening strategies.
It must be noted that the base case estimates account only for the diseases mentioned in the vaccine
SPC (Summary of Product Characteristics). As shown in the sensitivity analysis, the health and economic
benefits of a universal vaccination program with the 9-valent vaccine are substantially increased if H&N
and penile cancers are considered in the analysis. Besides, the results presented underestimate the
additional benefits of 9-valent vaccination on CIN. The estimates from the calibrated model show a
significant underestimation on the incidence of both grades of CIN. Furthermore, due to lack of German-
specific data, our model was calibrated towards CIN incidence rates observed in the UK. Since this
country has an organized screening program and higher VCR, CIN incidence rates may be higher in
Germany. In turn, our calibrated model underestimates the CIN attribution to the 5 additional types
included in the 9-valent vaccine. In addition, we do not account for neonatal morbidity and mortality
due to cervical lesions. It is widely accepted that women who undergo excisional treatments are at
increased risk of preterm delivery and low birth weight.[80, 81] A German study showed that HPV
vaccination could be cost-effective considering only the decrease in neonatal morbidity and mortality
due to the lower number of conisations.[82] Finally, the indirect costs related to productivity losses
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
22
were not considered in this study. However HPV-related cancers affect productivity. Lerner et al. (2010)
showed that women with HPV-related cervical lesions had higher absence rates and productivity loss
compared with healthy woman.[83]
In conclusion, a 9-valent vaccination program can yield significant incremental public health benefits and
was shown to be cost-effective when compared to the current 4-valent vaccination program. Inclusion
of boys in the 9-valent vaccination program would constitute an efficient and cost-effective strategy to
further reduce HPV-related cancers and diseases in the German population.
5. Key Issues
• The 9-valent vaccine yields significant incremental public health benefits and is shown to be
cost-effective when compared to the current 4-valent vaccination program.
• Inclusion of boys in the 9-valent vaccination program would constitute an efficient strategy to
further reduce HPV-related cancers and diseases in both sexes in Germany.
• Some potential benefits are concealed by the underestimation on the CIN incidence and the
attribution of HPV diseases related to the new genotypes included in the vaccine.
• All vaccination strategies evaluated remained within an acceptable range of cost-effectiveness
and the sensitivity analyses show robustness of the results across various assumptions on the
vaccine duration of protection and the VCR.
• Including all the HPV-related diseases without limiting to the ones indicated in the SPC,
improves the cost-effectiveness of the 9-valent vaccine, especially when considering the
universal vaccination strategy.
• This study accounts only for the direct medical costs. A wider societal perspective may yield
additional advantages of the new vaccine, as HPV-related diseases are associated with long-
term maternal consequences and productivity losses.
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
23
References
*of interest
**of considerable interest
1. Hartwig, S., J.J. Baldauf, and G. Dominiak-Felden, Estimation of the epidemiological burden of
HPV-related cancers, precancerous lesions, and genital warts in women and men in Europe:
potential additional benefit of a nine-valent compared to the quadrivalent HPV vaccine.
Papillomavirus Research, 2015. 1: p. 90-100.
*Provides an overview of the percentage of HPV-related diseases attributable to the
genotypes included in both vaccines
2. Robert Koch Institute. Geman Centre for Registry Cancer Data (Zentrum für Krebsregisterdaten -
ZfKD). 2014 2/17/2014; Available from:
http://www.krebsdaten.de/Krebs/DE/Content/ZfKD/zfkd_node.html.
3. NHS.uk. What is HPV? 2015.
4. Introduction of HPV vaccines in European Union countries - an update. 2015, European Centre
for Disease Prevention and Control.
5. Aubin, F., et al., Human papillomavirus genotype distribution in external acuminata
condylomata: a Large French National Study (EDiTH IV). Clin Infect.Dis., 2008. 47(5): p. 610-615.
6. Menton, J.F., et al., Molecular epidemiology of sexually transmitted human papillomavirus in a
self referred group of women in Ireland. Virol.J., 2009. 6: p. 112.
7. Preaud, E. and N. Largeron, Economic burden of non-cervical cancers attributable to human
papillomavirus: a European scoping review. J Med Econ, 2013. 16(6): p. 763-76.
8. Kotsopoulos, N., M.P. Connolly, and V. Remy, Quantifying the broader economic consequences
of quadrivalent human papillomavirus (HPV) vaccination in Germany applying a government
perspective framework. Health Econ Rev, 2015. 5(1): p. 54.
9. Horn, J., et al., Estimating the long-term effects of HPV vaccination in Germany. Vaccine, 2013.
31(19): p. 2372-2380.
**Assesses the public health benefits of a universal vaccination programm against HPV in
Germany.
10. Seidel, D., et al., Socio-demographic characteristics of participation in the opportunistic German
cervical cancer screening programme: results from the EPIC-Heidelberg cohort. J.Cancer
Res.Clin.Oncol., 2009. 135(4): p. 533-541.
11. Hillemanns, P., et al., New Screening Proposals: the Federal Joint Commission Defines the
Parameters for Cervical Cancer Screening from 2018: Statement of the Gynecology Oncology
Working Group (AGO). Geburtshilfe Frauenheilkd, 2016. 76(2): p. 145-146.
12. European Medicine Agency - Committee for Medicinal Products for Human Use, Gardasil human
papillomavirus vaccine [types 6, 11, 16, 18] (recombinant, adsorbed) - Summary of opinion.
European Medicine Agency, Committee for Medicinal Products for Human Use.
13. European Medicines Agency. Gardasil, Summary of Product Characteristics. 2015 2015;
Available from:
http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/medicines/000703/h
uman_med_000805.jsp.
14. European Medicines Agency. Cervarix, Summary of Product Characteristics. 2015 2015; Available
from:
http://www.ema.europa.eu/ema/index.jsp?curl=pages/medicines/human/medicines/000721/h
uman_med_000694.jsp&mid=WC0b01ac058001d124.
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
24
15. Rieck, T., et al., Utilization of administrative data to assess the association of an adolescent
health check-up with human papillomavirus vaccine uptake in Germany. Vaccine, 2014. 32(43):
p. 5564-5569.
16. Hense, S., et al., HPV vaccine uptake after introduction of the vaccine in Germany: an analysis of
administrative data. Hum.Vaccin.Immunother., 2014. 10(6): p. 1729-1733.
17. Delere, Y., et al., HPV vaccination coverage among women aged 18-20 years in Germany three
years after recommendation of HPV vaccination for adolescent girls: results from a cross-
sectional survey. Hum.Vaccin.Immunother., 2013. 9(8): p. 1706-1711.
18. Robert Koch Institute, Recommendations of the Standing Committee on Vaccination (STIKO) at
the Robert Koch Institute / Effective: August 2014. 2015.
19. Poethko-Muller, C., N. Buttmann-Schweiger, and G.G.S.S.G. Ki, [HPV vaccination coverage in
German girls: results of the KiGGS study: first follow-up (KiGGS Wave 1)].
Bundesgesundheitsblatt Gesundheitsforschung Gesundheitsschutz, 2014. 57(7): p. 869-77.
20. Li, N., et al., Human papillomavirus type distribution in 30,848 invasive cervical cancers
worldwide: Variation by geographical region, histological type and year of publication.
Int.J.Cancer, 2011. 128(4): p. 927-935.
21. Merck. Merck's Investigational 9-valent HPV Vaccine, V503, Prevented 97 Percent of Cervical,
Vaginal and Vulvar Pre-cancers Caused by Five Additional HPV Types, in Phase III Study. 2015
2015 [cited 2014 12/1/2014]; Available from: http://www.mercknewsroom.com/news-
release/research-and-development-news/mercks-investigational-9-valent-hpv-vaccine-v503-
prevente.
22. Summary of opinion Gardasil 9 human papillomavirus 9-valent vaccine (recombinant, adsorbed).
2015, European Medicine Agency, Committee for Medicinal Products for Human Use.
23. Mark Jit, Cost effectiveness of elimination of HPV-related cancers, in 5th Helsinki HPV workshop
on vaccination women and men against premature death. 2016: Helsinki , Finland.
24. Brisson, M., Incremental Cost-Effectiveness of the 9-valent vs. the 4-valent HPV vaccine in the
U.S., in Advisory Committee on Immunization Practices (ACIP) meeting. 2014: Atlanta.
25. Weiss, T.P., M.; Dasbach, Potential health and economic impact of the investigational 9-valent
HPV vaccine in the United States, in International Papillomavirus Conference. 2014: Seattle.
26. Chesson, H., Overview of cost-effectiveness of 9-valent HPV vaccination, in Advisory Committee
on Immunization Practices. 2015: Atlanta.
27. Drolet, M., et al., Potential cost-effectiveness of the nonavalent human papillomavirus (HPV)
vaccine. Int J Cancer, 2014. 134(9): p. 2264-8.
28. Boiron, L., et al., Estimating The Cost-Effectiveness Profile of A Universal Vaccination Programme
With A Nine-Valent Hpv Vaccine In Austria. Value Health, 2015. 18(7): p. A585.
29. Hillemanns, P., et al., Cost-effectiveness of a tetravalent human papillomavirus vaccine in
Germany. Public Health, 2008. 17: p. 77-86.
30. Schobert, D., V. Remy, and O. Schoeffski, Cost-effectiveness of vaccination with a quadrivalent
HPV vaccine in Germany using a dynamic transmission model. Health Econ.Rev., 2012. 2(1): p.
19.
*Reports the methods and results of the German adaptation performed on a previous version
of the same model used in this study
31. Elbasha, E.H., E.J. Dasbach, and R.P. Insinga, Model for assessing human papillomavirus
vaccination strategies. Emerg.Infect.Dis., 2007. 13(1): p. 28-41.
**Describes in detail the structure of the first version of the model used in this study
32. Elbasha, E.H. and E.J. Dasbach, Impact of vaccinating boys and men against HPV in the United
States. Vaccine, 2010. 28(42): p. 6858-6867.
*Provides input values for model key parameters
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
25
33. Elbasha, E.H. and E.J. Dasbach, An Integrated Economic evaluation and HPV disease transmission
models - Technical report accompanying the Manuscript "Impact of vaccinating Boys and Men
against HPV in the Inited States. Vaccine, 2010. 28(42): p. 6858-6867.
34. German Federal Statistics Office. Facts & Figures -> Society & state ->Population ->Deaths, Life
expectancy
https://www.destatis.de/EN/FactsFigures/SocietyState/Population/Deaths/Deaths.html;jsession
id=4F274DF95FF7F226BB3576989FBA7AE7.cae2 2015; Available from:
https://www.destatis.de/EN/FactsFigures/SocietyState/Population/Deaths/Deaths.html.
35. National Survey of Sexual Attitudes and Life Styles III, N.S.f.S. Research, Editor.
36. Federal Statistics Office (Statistisches Bundesamt), Fallpauschalenbezogene
Krankenhausstatistik (DRG-Statistik) Diagnosen, Prozeduren, Fallpauschalen und Case Mix der
vollstationären Patientinnen und Patienten in Krankenhäusern, avaliable at:
https://www.destatis.de/DE/Publikationen/Thematisch/Gesundheit/Krankenhaeuser/Operation
enProzeduren5231401137014.pdf?__blob=publicationFile. 2014.
37. Sroczynski, G., et al., Decision-analytic modeling to evaluate the long-term effectiveness and
cost-effectiveness of HPV-DNA testing in primary cervical cancer screening in Germany.
GMS.Health Technol.Assess., 2010. 6: p. Doc05.
38. Ruckinger, S., et al., Participation in cervical cancer screening by age and region--a cohort study
with a 3 1/2 year follow-up on 2,223,135 women in Bavaria. Gesundheitswesen, 2008. 70(6): p.
e17-e21.
39. Kerek-Bodden, H.A., L; Brenner,G, Durchführung einer versichertenbezogenen Untersuchung zur
Inanspruchnahme der Früherkennung auf Zervixkarzinom in den Jahren 2002, 2003 und 2004 auf
der Basis von Abrechnungsdaten, in http://www.ag-cpc.de/media/ZI-Abschlussbericht-
080602.pdf. 2008.
40. McCrory, D.C., et al., Evaluation of cervical cytology. Evid Rep Technol Assess (Summ), 1999(5):
p. 1-6.
41. Nanda, K., et al., Accuracy of the Papanicolaou test in screening for and follow-up of cervical
cytologic abnormalities: a systematic review. Ann.Intern.Med., 2000. 132(10): p. 810-819.
42. Mitchell, M.F., et al., Colposcopy for the diagnosis of squamous intraepithelial lesions: a meta-
analysis. Obstet.Gynecol., 1998. 91(4): p. 626-631.
43. Insinga, R.P., et al., Progression and regression of incident cervical HPV 6, 11, 16 and 18
infections in young women. Infect.Agent.Cancer, 2007. 2: p. 15.
44. Myers, E.R., et al., Mathematical model for the natural history of human papillomavirus infection
and cervical carcinogenesis. Am.J.Epidemiol., 2000. 151(12): p. 1158-1171.
45. Sanders, G.D. and A.V. Taira, Cost-effectiveness of a potential vaccine for human papillomavirus.
Emerg.Infect.Dis., 2003. 9(1): p. 37-48.
46. Ho, G.Y., et al., Natural history of human papillomavirus type 16 virus-like particle antibodies in
young women. Cancer Epidemiol.Biomarkers Prev., 2004. 13(1): p. 110-116.
47. Onda, T., et al., Characterization of IgA response among women with incident HPV 16 infection.
Virology, 2003. 312(1): p. 213-221.
48. Federal Statistics Office (Statistisches Bundesamt). Mortality Table Germany 2010-12. 2015;
Available from:
https://www.destatis.de/DE/ZahlenFakten/GesellschaftStaat/Bevoelkerung/Sterbefaelle/Tabell
en/SterbetafelDeutschland.xlsx?__blob=publicationFile.
49. German Federal Statiscs Office. Facts & Figures -> Society & state ->Population ->Current
Population. 2015 [cited 2015; Available from:
https://www.destatis.de/EN/FactsFigures/SocietyState/Population/CurrentPopulation/CurrentP
opulation.html.
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
26
50. Zechmeister, I.B., B.F.; Garnett,G.; Neilson,A.R.; Siebert,U., Cost-effectiveness analysis of human
papillomavirus-vaccination programs to prevent cervical cancer in Austria. Vaccine, 2009.
27(37): p. 5133-5141.
51. Schmitt, D.P., Sociosexuality from Argentina to Zimbabwe: a 48-nation study of sex, culture, and
strategies of human mating. Behav.Brain Sci., 2005. 28(2): p. 247-275.
52. Petry, K.U., et al., Prevalence of low-risk HPV types and genital warts in women born 1988/89 or
1983/84 -results of WOLVES, a population-based epidemiological study in Wolfsburg, Germany.
BMC Infect Dis, 2012. 12: p. 367.
53. Petry, K.U., et al., Prevalence of high-risk HPV types and associated genital diseases in women
born in 1988/89 or 1983/84--results of WOLVES, a population-based epidemiological study in
Wolfsburg, Germany. BMC Infect Dis, 2013. 13: p. 135.
54. Luyten, A., et al., Early detection of CIN3 and cervical cancer during long-term follow-up using
HPV/Pap smear co-testing and risk-adapted follow-up in a locally organised screening
programme. Int J Cancer, 2014. 135(6): p. 1408-16.
55. Cancer Research UK. Cervical cancer survival statistics.
http://www.cancerresearchuk.org/cancer-info/cancerstats/types/cervix/survival/ 2014
11/27/2014.
56. Istituto Nazionale Tumori and Istituto Superiore di Sanità, Eurocare Survival of cancer patient in
Europe. 2015.
57. Joura, E.A., et al., A 9-valent HPV vaccine against infection and intraepithelial neoplasia in
women. N Engl J Med, 2015. 372(8): p. 711-23.
58. Ault, K.A., Effect of prophylactic human papillomavirus L1 virus-like-particle vaccine on risk of
cervical intraepithelial neoplasia grade 2, grade 3, and adenocarcinoma in situ: a combined
analysis of four randomised clinical trials. Lancet, 2007. 369(9576): p. 1861-8.
59. Garland, S.M., et al., Quadrivalent vaccine against human papillomavirus to prevent anogenital
diseases. N Engl J Med, 2007. 356(19): p. 1928-43.
60. Giuliano, A.R., et al., Efficacy of quadrivalent HPV vaccine against HPV Infection and disease in
males. N Engl J Med, 2011. 364(5): p. 401-11.
61. Joura, E.A., et al., Efficacy of a quadrivalent prophylactic human papillomavirus (types 6, 11, 16,
and 18) L1 virus-like-particle vaccine against high-grade vulval and vaginal lesions: a combined
analysis of three randomised clinical trials. Lancet, 2007. 369(9574): p. 1693-702.
62. Palefsky, J.M., et al., HPV vaccine against anal HPV infection and anal intraepithelial neoplasia. N
Engl J Med, 2011. 365(17): p. 1576-85.
63. Group, F.I.S., Quadrivalent vaccine against human papillomavirus to prevent high-grade cervical
lesions. N Engl J Med, 2007. 356(19): p. 1915-27.
64. Federal Statiscs Office (Statistisches Bundesamt). German Consumer Price Index (CPI). 2015 2015
08/12/2015]; Available from:
https://www.destatis.de/EN/FactsFigures/NationalEconomyEnvironment/Prices/ConsumerPrice
Indices/Tables_/ConsumerPricesCategories.html?cms_gtp=151232_list%253D3%2526151226_sl
ot%253D2&https=1.
65. Rote Liste Service GmbH. Gardasil Price. 2016 08/02/2016]; Available from: http://online.rote-
liste.de/suche/praep/20976.
66. Lauer Fischer. Gardasil 9 Price. 2016 15/03/2016]; Available from: http://www2.lauer-
fischer.de/produkte/lauer-taxe/lauer-taxe/.
67. Siebert U, M.C., Sroczynski G, Velasco-Garrido M, Gerhardus A, Gibis B, Liquidbased preparation
and computer-assisted examination of cervical smears. Clinical effectiveness, economic
evaluation and systematic decision analysis. Health Technology Assessment https://www.mig.tu-
berlin.de/fileadmin/a38331600/2004.publications/2004.siebert_hta.35.pdf, in Series of the
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
27
German Institute for Medical Documentation and Information commissioned by the Federal
Ministry of Health and Social Security. 2004. p. 1-444.
68. Petry, K.U., et al., Cost of screening and treatment of cervical dyskaryosis in Germany.
Eur.J.Gynaecol.Oncol., 2008. 29(4): p. 345-349.
69. Hampl, M., et al., Economic burden of vulvar and vaginal intraepithelial neoplasia: retrospective
cost study at a German dysplasia centre. BMC.Infect.Dis., 2011. 11: p. 73.
70. Brisson, M., et al., Comparative cost-effectiveness of the quadrivalent and bivalent human
papillomavirus vaccines: a transmission-dynamic modeling study. Vaccine, 2013. 31(37): p. 3863-
3871.
71. Hinz, A., et al., [The Quality of Life Questionnaire EQ-5D: modelling and norm values for the
general population]. Psychother.Psychosom.Med.Psychol., 2006. 56(2): p. 42-48.
72. Sullivan, P.W., et al., Catalogue of EQ-5D scores for the United Kingdom. Med.Decis.Making,
2011. 31(6): p. 800-804.
73. Hu, D. and S. Goldie, The economic burden of noncervical human papillomavirus disease in the
United States. Am.J.Obstet.Gynecol., 2008. 198(5): p. 500-507.
74. Hillemanns, P., et al., Estimation of the incidence of genital warts and the cost of illness in
Germany: a cross-sectional study. BMC.Infect.Dis., 2008. 8: p. 76.
75. Klussmann, J.P., et al., Annual cost of hospitalization, inpatient rehabilitation, and sick leave for
head and neck cancers in Germany. Clinicoecon.Outcomes.Res., 2013. 5: p. 203-213.
76. Kraut, A.A., et al., Incidence of anogenital warts in Germany: a population-based cohort study.
BMC.Infect.Dis., 2010. 10: p. 360.
77. Hartwig, S., et al., Estimation of the epidemiological burden of human papillomavirus-related
cancers and non-malignant diseases in men in Europe: a review. BMC.Cancer, 2012. 12: p. 30.
78. Health & Social Care Information Centre, Cervical Screening Programme, England 2013-14. 2015.
79. Population Estimates for UK, England and Wales, Scotland and Northern Ireland, Mid-2013. 2014
2014 [cited 2014 1/19/2014]; Available from: http://www.ons.gov.uk/ons/publications/re-
reference-tables.html?edition=tcm%3A77-322718.
80. Kyrgiou, M., et al., Obstetric outcomes after conservative treatment for intraepithelial or early
invasive cervical lesions: systematic review and meta-analysis. Lancet, 2006. 367(9509): p. 489-
98.
81. Bruinsma, F.J. and M.A. Quinn, The risk of preterm birth following treatment for precancerous
changes in the cervix: a systematic review and meta-analysis. Bjog, 2011. 118(9): p. 1031-41.
82. Soergel, P., et al., The cost efficiency of HPV vaccines is significantly underestimated due to
omission of conisation-associated prematurity with neonatal mortality and morbidity. Hum
Vaccin Immunother, 2012. 8(2): p. 243-51.
83. Lerner, D., et al., The impact of precancerous cervical lesions on functioning at work and work
productivity. J Occup Environ Med, 2010. 52(9): p. 926-33.
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
28
6. Appendix
(Table 13)
Figure 1: Epidemiological impact of three vaccination strategies on the incidence and mortality of cervical diseases
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
29
Figure 2: Epidemiological impact of three vaccination strategies on the incidence of genital warts and anal cancer
Figure 3: Price threshold analysis of HPV9 vaccination vs HPV4 Girls vaccination with the 4-valent vaccine priced at €140
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
30
Figure 4: Tornado diagrams
Table 1: Summary table on the epidemiological input groups with some of the references used
Parameter References
Demographics
Annual all-cause mortality rate
[34] Female and male population
Female and male population >12yo
Sexual behavior
Percent of population in low/medium/high sexual activity category
[35]
Mean number of sexual partners by activity category
Mean number of sexual partners by age group
Sexual mixing among activity categories
Sexual mixing among age groups
Disease and treatment patterns
Female population receiving hysterectomy each year [36]
Age- and stage- specific mortality rates [2]
Population recognising their symptoms and seeking treatment by disease Calibration
Percent of cases treated
Screening
Percent of females receiving a follow-up screening test after abnormal Pap
smear [37]
Percent of females screened every 3 years [38]
Age-specific percent of females screened in the past year [39]
Diagnostic performance of PAP test and colposcopy [40-42]
Natural history of disease
Probability of transmitting genital, anal, penile, and head and neck HPV infection
per sexual partnership, by sex and HPV genotype [43]
Recurrence rate of treated CIN, by stage
Rate of cancer progression, by stage [44, 45]
Fraction of persistent cervical HPV infections, by type 16 or 18 [32, 33]
Clearance rate of cervical HPV infections, by type 16 or 18 [43]
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
31
Fraction of people who seroconvert following a cervical HPV infection, by type 16
or 18 [46, 47]
Degree of protection against cervical HPV infections provided by natural
immunity following seroconversion, by type 16 or 18 [32, 33]
Fraction of females transiently infected who progress to CIN over the course of
one year, by type 16 or 18 [43]
Table 2: Summary table on sexual behaviour
Definition of sexual activity categories Mean number of sexual partners per year
Low mean number of sexual partners/year: ≤ 1 Age group Male Female
Medium mean number of sexual partners/year: 2-4 13-14 0.0001 0.0001
High mean number of sexual partners/year: 5+ 15-29 1.7 1.4
Size of categories and mean number of partners 30-34 2.0 1.6
Percent of population Mean number of partners 35-44 1.7 1.3
Male Female Males Females 45-49 1.5 1.2
Low 85.10% 90.70% 0.79 0.75 50-54 1.5 1.0
Medium 11.90% 7.60% 2.54 2.52 55-59 1.1 1.5
High 3.00% 1.70% 9.8 9.66 60-64 1.1 1.0
Sexual mixing 65-69 1.1 0.9
Among members of different age cohort 70-74 1.0 0.7
Between debut and cessation 0.4 75-79 0.9 0.6
After cessation 0.1 80-84 0.8 0.5
Among members of different sexual activity groups 0.5 85+ 0.5 0.3
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
32
Table 3: Summary table on cancer mortality
Cancer type Age group Annual probability of death
(years) Local Cancer Regional Cancer Distant Cancer
Cervical cancer
15-44 0.010 0.032 0.081
45-54 0.022 0.069 0.173
55-64 0.031 0.097 0.224
65-74 0.039 0.123 0.309
75+ 0.069 0.219 0.552
Vaginal cancer
15-44 0.013 0.023 0.041
45-54 0.017 0.030 0.053
55-64 0.035 0.061 0.109
65-74 0.053 0.091 0.162
75+ 0.095 0.163 0.291
Vulvar cancer
15-44 0.011 0.023 0.050
45-54 0.014 0.030 0.064
55-64 0.028 0.061 0.132
65-74 0.042 0.091 0.197
75+ 0.075 0.163 0.353
Anal cancer
(Females)
15-44 0.030 0.066 0.114
45-54 0.032 0.072 0.123
55-64 0.032 0.072 0.123
65-74 0.041 0.091 0.157
75+ 0.077 0.172 0.295
Anal cancer
(Males)
15-44 0.032 0.072 0.123
45-54 0.038 0.085 0.147
55-64 0.040 0.088 0.152
65-74 0.049 0.109 0.188
75+ 0.081 0.180 0.310
Penile cancer
15-44 0.008 0.037 0.080
45-54 0.015 0.072 0.159
55-64 0.017 0.083 0.183
65-74 0.027 0.130 0.286
75+ 0.038 0.181 0.398
Head & Neck
cancer
(Females)
15-44 0.052 0.075 0.094
45-54 0.071 0.104 0.130
55-64 0.080 0.116 0.145
65-74 0.088 0.128 0.160
75+ 0.149 0.216 0.271
Head & Neck
cancer (Males)
15-44 0.093 0.134 0.168
45-54 0.107 0.155 0.194
55-64 0.121 0.176 0.220
65-74 0.141 0.204 0.255
75+ 0.176 0.255 0.319
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
33
Table 4: Cervical cancer screening rates
Age group Percentage of females screened in the past year (%) Reference
0-19 0.00
[39]
20-24 54.62
25-29 55.93
30-34 53.90
35-39 52.09
40-44 50.29
45-49 49.51
50-54 48.80
55-59 46.94
60-64 43.76
65-69 37.63
70-74 27.50
75-79 19.26
>80 9.02
Table 5: Vaccination Coverage Rates
Age group Vaccination coverage rate (%) Reference
9 – 10 16.3
[19]
11 – 12 37.7
13 – 14 45.6
15 – 17 55.6
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
34
Table 6: Summary table on vaccine assumptions
Vaccine assumptions HPV
16
HPV
18
HPV 31, 33, 45, 52
and 58
Cervical cancer
Vaccine efficacy for preventing cervical HPV16/18/31/33/45/52/58 infections:
- Male* 0.411 0.621 0.411
- Female** 0.760 0.963 0.760
Degree of protection of the vaccine against cervical HPV16/18 infections becoming
persistent 0.988 0.984 0.988
Degree of protection of the vaccine against HPV16/18 -related CIN 0.979 1.000 0.979
Vaginal and vulvar cancers
Vaccine efficacy for preventing vaginal/vulvar HPV16/18 infections:
- Male* 0.411 0.621
- Female** 0.760 0.963
Degree of protection of the vaccine against vaginal/vulvar HPV16/18 infections
becoming persistent 0.988 0.984
Degree of protection of the vaccine against HPV16/18-related /VaIN/VIN 1.000 1.000
Anal cancers
Vaccine efficacy for preventing anal infections
- Male* 0.411 0.621 0.621
- Female** 0.760 0.963 0.963
Degree of protection of the vaccine against anal infections becoming persistent
- Male* 0.787 0.960 0.960
- Female** 0.988 0.984 0.984
Degree of protection of the vaccine against HPV16/18 -related AIN neoplasia 0.000 0.000 0.000
Penile and H&N cancers
Vaccine efficacy for preventing penile and H&N infections
- Male* 0.411 0.621
- Female** 0.760 0.963
Degree of protection of the vaccine against penile and H&N infections becoming
persistent
- Male* 0.787 0.960
- Female** 0.988 0.984
Degree of protection of the vaccine against HPV16/18 -related PIN and H&N
neoplasia 0.000 0.000
* Preventing male genital infections through male vaccination is assumed to prevent transmission of genital infections to females
** Preventing female genital infections through vaccination is assumed to prevent transmission of genital infections to males
***The efficacy against anal, Head and Neck, Penile and RRP diseases is conferred through protection against infection only.
Source:
Females: Future II study group 2007 [63] and Joura 2007 [61] for disease endpoints, Internal data file (protocol 007 and 012 combined per
protocol) and Elbasha (2010) [32] for transient and persistent infections
Males: Giuliano et al. (2011) [60] and Elbasha (2010) [32]
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
35
Table 7: Costs and utilities
HPV-related disease Costs (€) References Utilities References
Females Males Females Males
CIN 1 363.30
[68]
0.822
[32, 72, 73]
CIN 2 0.822
CIN 3, CIS 1,619.50 0.822
Cervical cancer, local disease 8,812.70
[67]
0.822
Cervical cancer, regional disease 18,331.70 0.732
Cervical cancer, distant disease 20,092.30 0.542
Cervical cancer, cancer survivor 0.822
VaIN 2 1,117.40 [69] 0.822
VaIN 3, CIS 0.822
Vaginal cancer, local disease 7,667.00
[67, 70]
0.822
Vaginal cancer, regional disease 15,948.60 0.732
Vaginal cancer, distant disease 17,480.30 0.542
Vaginal cancer, cancer survivor 0.822
Vulvar cancer, local disease 7,667.00 0.822
Vulvar cancer, regional disease 15,948.60 0.732
Vulvar cancer, distant disease 17,480.30 0.542
Vulvar cancer, cancer survivor 0.822
Penile cancer, local disease 6,168.90 0.751
Penile cancer, regional disease 12.832.20 0.661
Penile cancer, distant disease 14,064.60 0.471
Penile cancer, cancer survivor 0.751
Anal cancer, local disease 8,988.90 0.645
Anal cancer, regional disease 18,698.30 0.555
Anal cancer, distant disease 20,494.20 0.365
Anal cancer, cancer survivor 0.645
Head & Neck cancer, local disease 10,575.20 0.756
Head & Neck cancer, regional disease 21,998.00 0.666
Head & Neck cancer, distant disease 24,110.80 0.476
Head & Neck cancer, cancer survivor 0.756
Genital warts 633.80 [74] 0.900
Table 8: HPV attribution rates
Females Males
References
HVP4 HVP9 HVP4 HVP9
Cervical cancer 72.8% 89.0% NA NA
[1]
CIN 1 24.0% 48.5% NA NA
CIN 2+ 45.5% 82.3% NA NA
Vaginal cancer 50.7% 60.6% NA NA
Vulvar cancer 14.2% 16.2% NA NA
Anal cancer 76.3% 78.7% 76.3% 78.7%
Head & Neck cancers 17.8% 17.8% 18.5% 18.5%
Penile cancer NA NA 34.4% 34.4%
Genital warts 90.0% 90.0% 90.0% 90.0%
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
36
Table 9: Overview of the calibration incidence targets
Incidence rates (per 100,000)
HPV 16,18, 6, 11 related HPV-9 related (adding 31, 33, 45, 52, and 58)
Target Calibration Target Calibration
Female
Cervical cancer 7.21 7.14 8.81 8.82
CIN 1 72.61 30.20 146.74 39.23
CIN 2+ 71.22 61.97 128.82 77.68
Vaginal 0.30 0.19 0.36 0.19
Vulvar 0.45 0.26 0.52 0.26
Anal 1.14 1.14 1.18 1.17
Genital warts 171.99 169.76 171.99 169.76
Male
Penile cancer 0.45 0.44 0.45 0.44
Anal cancer 0.84 0.85 0.87 0.86
Genital warts 132.89 133.26 132.89 133.26
Table 10: Disease events prevented with HPV9 Girls in comparison with the current strategy (HPV4
girls).
Disease event Years since start of vaccination programme
25 50 100
Females
Cervical cancer 378 5,210 31,500
CIN 1 14,894 72,335 234,899
CIN 2/3 23,364 119,048 399,410
Anal cancer 2 47 438
Males
Anal cancer 1 22 240
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
37
Table 11: Disease events prevented with HPV9 Universal in comparison with the current strategy (HPV4
girls).
Disease event Years since start of vaccination programme
25 50 100
Females
Cervical cancer 692 8,546 46,454
CIN 1 38,882 141,381 398,993
CIN 2/3 40,312 183,373 571,013
Vaginal cancer 2 45 315
Vulvar cancer 3 65 429
Genital warts 75,189 172,001 364,313
Anal cancer 16 381 3,036
Males
Genital warts 182,712 470,841 1,084,422
Anal cancer 37 804 5,420
Table 12: Cost effectiveness results in the base case analysis
Scenarios Strategies
Costs / person
(€)
QALYs /
person
Incremental
costs (€)
Incremental
QALYs
Cost per
QALY gained
(€/QALY)
Scenario 1 HPV9 Girls 336,41 28.36934 0.24 0.00073 328.77
HPV4 Girls 336,17 28.36861 - 0.00157 -
Scenario 2 HPV9 Universal 372.26 28.37018 36.09 0.00073
22,987.26
HPV4 Girls 336.17 28.36861 - 0.00157 -
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
38
Table 13. Natural history of disease related parameters (Appendix)
Probability of transmitting genital HPV infection [84]
Transmission HPV 16 HPV 18 HPV 6 HPV 11 HPV 31, 33, 45,
52 or 58
To males 0.1109 0.1109 0.415* 0.415* 0.076
To females 0.1109 0.1109 0.415* 0.415* 0.076
Stage Recurrence rate Reference Cancer progression
Reference
CIN 1 0.05 Assumption Direction Rate
CIN 2 0.05 Assumption local->regional 0.1
CIN 3 0.05 Assumption regional->distant 0.3 [44, 45]
Parameters HPV 16 HPV 18
Fraction of persistent cervical HPV infections (Elbasha[31]) 0.25 0.075
Clearance rate of cervical HPV infections (Insinga[43])
Male 0.3955 0.37755
Female 0.354 0.348
Fraction of people seroconvert following a cervical HPV infection (Ho[46] and Onda[47])
Male 0.6 0.6
Female 0.6 0.6
Degree of protection against cervical HPV infections provided by natural immunity following
seroconversion (Elbasha[31])
Male 0.5 0.5
Female 0.8 0.8
Fraction of females transiently infected with HPV16 progress to CIN over the course of one
year (Insinga[43])
CIN 1 0.105 0.068
CIN 2 0.045 0.055
CIN 3 0.024 0.009
Probability of transmitting anal HPV infection (Calibration)
To males 0.16 0.16
To females 0.173 0.173
Probability of transmitting penile HPV infection (Calibration)
To males 0.123 0.123
To females 0.123 0.123
Probability of transmitting head and neck HPV infection (Calibration)
To males 0.14118 0.13228
To females 0.14118 0.13228
*Adjusted during the calibration process
Downloaded by [Sanofi Pasteur MSD], [Mathieu Uhart] at 00:30 04 July 2016
