Article (PDF Available)

Incremental cost-effectiveness evaluation of vaccinating girls against cervical cancer pre- and post-sexual debut in Belgium

Health Economics, GlaxoSmithKline Vaccines, Rue Fleming 20, B-1300 Wavre, Belgium. Electronic address: .
Vaccine (Impact Factor: 3.62). 06/2013; 31(37). DOI: 10.1016/j.vaccine.2013.06.008
Source: PubMed

ABSTRACT

Background:
Vaccination against human papillomavirus (HPV) to prevent cervical cancer (CC) primarily targets young girls before sexual debut and is cost-effective. We assessed whether vaccination with the HPV-16/18 AS04-adjuvanted vaccine added to screening remains cost-effective in females after sexual debut compared to screening alone in Belgium. The role of protection against non-HPV-16/18 was also investigated.

Methods:
A published Markov cohort model was adapted to Belgium. The model replicated the natural history of HPV infection, the effects of screening, and vaccination. Vaccine efficacy (VE) included non-HPV-16/18 protection based on the PATRICIA clinical trial data. Pre- and post-HPV exposure VE were differentiated. Lifetime vaccine protection was assumed. Input data were obtained from literature review, national databases and a Delphi panel. Costing was from a healthcare payer perspective. Costs were discounted at 3% and effects at 1.5%. The incremental cost-effectiveness ratio (ICER) per quality-adjusted life-year (QALY) gained and the number of lesions prevented with vaccination from age 12 to 40 was evaluated. The specific effect of non-HPV-16/18 protection was investigated. Univariate sensitivity analysis was performed on key variables.

Results:
The model estimated that vaccinating a cohort of 100,000 girls at age 12 would prevent 646 CC cases over a lifetime (102 non-HPV-16/18) with an ICER of €9171/QALY. Vaccinating at age 26 would prevent 340 CC cases (40 non-HPV-16/18) with an ICER of €17,348/QALY and vaccinating at age 40 would prevent 146 CC cases (17 non-HPV-16/18) with an ICER of €42,847/QALY. The ICER remained under the highly cost-effective threshold (1×GDP/capita) until age 33 years and under the cost-effective threshold (3×GDP/capita) beyond age 40.

Conclusion:
Extending HPV vaccination to females post-sexual debut could lead to a substantial reduction in CC-related burden and would be cost-effective in Belgium.

Full-text (PDF)

Available from: Nadia Demarteau, Sep 09, 2015
Vaccine
31 (2013) 3962–
3971
Contents
lists
available
at
SciVerse
ScienceDirect
Vaccine
j
our
nal
homep
ag
e:
www.elsevier.com/locate/vaccine
Incremental
cost-effectiveness
evaluation
of
vaccinating
girls
against
cervical
cancer
pre-
and
post-sexual
debut
in
Belgium
Nadia
Demarteau
a,
,
Georges
Van
Kriekinge
a
,
Philippe
Simon
b
a
Health
Economics,
GlaxoSmithKline
Vaccines,
Rue
Fleming
20,
B-1300
Wavre,
Belgium
b
Chef
de
Clinique
Gynécologie,
ULB-Hôpital
Erasme,
808,
Route
de
Lennik,
1070
Bruxelles,
Belgium
a
r
t
i
c
l
e
i
n
f
o
Article
history:
Received
1
August
2012
Received
in
revised
form
13
May
2013
Accepted
4
June
2013
Available online 15 June 2013
Keywords:
Vaccination
Human
papillomavirus
Cost-effectiveness
Cervical
cancer
Adult
Belgium
a
b
s
t
r
a
c
t
Background:
Vaccination
against
human
papillomavirus
(HPV)
to
prevent
cervical
cancer
(CC)
primarily
targets
young
girls
before
sexual
debut
and
is
cost-effective.
We
assessed
whether
vaccination
with
the
HPV-16/18
AS04-adjuvanted
vaccine
added
to
screening
remains
cost-effective
in
females
after
sexual
debut
compared
to
screening
alone
in
Belgium.
The
role
of
protection
against
non-HPV-16/18
was
also
investigated.
Methods:
A
published
Markov
cohort
model
was
adapted
to
Belgium.
The
model
replicated
the
natu-
ral
history
of
HPV
infection,
the
effects
of
screening,
and
vaccination.
Vaccine
efficacy
(VE)
included
non-HPV-16/18
protection
based
on
the
PATRICIA
clinical
trial
data.
Pre-
and
post-HPV
exposure
VE
were
differentiated.
Lifetime
vaccine
protection
was
assumed.
Input
data
were
obtained
from
literature
review,
national
databases
and
a
Delphi
panel.
Costing
was
from
a
healthcare
payer
perspective.
Costs
were
discounted
at
3%
and
effects
at
1.5%.
The
incremental
cost-effectiveness
ratio
(ICER)
per
quality-
adjusted
life-year
(QALY)
gained
and
the
number
of
lesions
prevented
with
vaccination
from
age
12
to
40
was
evaluated.
The
specific
effect
of
non-HPV-16/18
protection
was
investigated.
Univariate
sensitivity
analysis
was
performed
on
key
variables.
Results:
The
model
estimated
that
vaccinating
a
cohort
of
100,000
girls
at
age
12
would
prevent
646
CC
cases
over
a
lifetime
(102
non-HPV-16/18)
with
an
ICER
of
D
9171/QALY.
Vaccinating
at
age
26
would
prevent
340
CC
cases
(40
non-HPV-16/18)
with
an
ICER
of
D
17,348/QALY
and
vaccinating
at
age
40
would
prevent
146
CC
cases
(17
non-HPV-16/18)
with
an
ICER
of
D
42,847/QALY.
The
ICER
remained
under
the
highly
cost-effective
threshold
(1×GDP/capita)
until
age
33
years
and
under
the
cost-effective
threshold
(3×GDP/capita)
beyond
age
40.
Conclusion:
Extending
HPV
vaccination
to
females
post-sexual
debut
could
lead
to
a
substantial
reduction
in
CC-related
burden
and
would
be
cost-effective
in
Belgium.
© 2013 The Authors. Published by Elsevier Ltd.
1.
Introduction
Worldwide,
cervical
cancer
is
the
third
most
common
cancer
in
women
[1].
Although
screening
can
reduce
the
incidence
of
and
mortality
from
cervical
cancer,
by
detecting
and
allowing
treatment
Abbreviations:
CC,
cervical
cancer;
CIN,
cervical
intraepithelial
neoplasia;
CIS,
carcinoma
in
situ;
GDP,
gross
domestic
product;
HPV,
human
papillomavirus;
ICER,
incremental
cost-effectiveness
ratio;
QALY,
quality-adjusted
life-year;
TVC,
total
vaccinated
cohort;
VE,
vaccine
efficacy.
Corresponding
author
at:
Health
Economics,
GlaxoSmithKline
Vaccines,
Rue
Fleming
20,
W23,
B-1300
Wavre,
Belgium.
Tel.:
+32
26
56
85
86;
fax:
+32
26
56
63
00.
E-mail
addresses:
nadia.x.demarteau@gskbio.com,
nadia.x.demarteau@gsk.com
(N.
Demarteau).
of
pre-cancerous
changes,
cervical
cancer
remains
an
important
public
health
problem
in
Europe,
with
an
estimated
31,000
cases
and
14,000
deaths
in
2004
[2].
In
Belgium,
about
600
cases
of
cervi-
cal
cancer
occur
annually
according
to
Belgian
Cancer
Registry
data
[3]
,
and
the
estimated
total
annual
cost
of
cervical
cancer
is
D
6.5
million
[4].
Human
papillomavirus
(HPV)
is
established
as
a
causal
factor
in
cervical
cancer,
identified
in
99.7%
of
cervical
cancers
world-
wide
[5].
Vaccination
against
high-risk
oncogenic
serotypes
of
HPV
offers
potential
for
primary
prevention
of
cervical
cancer
[6].
Two
HPV
vaccines
are
widely
available,
a
quadrivalent
vaccine
against
HPV
types
16,
18,
6
and
11
(Gardasil
TM
,
1
Merck/Sanofi-Pasteur)
and
a
bivalent
vaccine
against
HPV
types
16
and
18
with
the
AS04
1
Gardasil
is
a
trademark
of
Merck
&
Co.
Inc.
0264-410X
©
2013 The Authors. Published by Elsevier Ltd.
http://dx.doi.org/10.1016/j.vaccine.2013.06.008
Open access under CC BY license.
Open access under CC BY license.
Page 1
N.
Demarteau
et
al.
/
Vaccine
31 (2013) 3962–
3971 3963
adjuvant
(Cervarix
®
,
2
GlaxoSmithKline
Vaccines).
Clinical
data
have
demonstrated
that
both
vaccines
offer
protection
against
oncogenic
non-vaccine
HPV
types
[7–9].
The
HPV-16/18
AS04-adjuvanted
vaccine
(HPV-16/18
vaccine)
however
potentially
offers
a
better
protection
against
oncogenic
non-vaccine
HPV
types
than
the
HPV-
16/18/6/11
vaccine
[10,11].
Current
HPV
vaccination
programmes
mainly
target
adoles-
cent
girls,
before
the
onset
of
sexual
activity
(sexual
debut).
In
Belgium,
HPV
vaccination
is
recommended
at
age
12
years,
with
a
reimbursed
catch-up
programme
to
age
18
years
[12].
Other
countries
have
extended
the
catch-up
programme
to
include
young
women;
for
example,
Australia
has
extended
the
HPV
vaccination
programme
to
girls
and
young
women
aged
12–26
years
[13],
and
the
US
also
recommends
vaccination
up
to
age
26
years
[14].
Cost-effectiveness
analyses
have
been
conducted
for
HPV
vac-
cination
programmes
in
girls
aged
12
years
for
many
countries
among
which
are
France
[15]
and
Belgium
[6,16].
Few
economic
evaluations
of
HPV
vaccination
in
young
adult
and
adult
women,
assessing
catch
up
programmes,
have
been
published.
In
2010
a
review
identified
only
8
publications
on
this
topic
[17].
The
same
literature
query
run
in
2012,
identified
an
additional
22
articles
of
which
5
articles
evaluated
catch-up
vaccination
programmes
and
17
not
discussing
the
topic.
No
specific
evaluation,
adapted
to
the
Belgian
setting,
of
the
cost-effectiveness
of
vaccination
after
the
recommended
age
at
vaccination
have
been
performed
to
date.
Healthcare
decision-makers
may
need
information
on
the
pro-
jected
clinical
impact
and
cost-effectiveness
of
HPV
vaccination
in
young
adult
and
adult
women
when
deciding
whether
to
imple-
ment
or
extend
catch-up
vaccination
programmes
in
older
age
groups.
In
this
paper,
we
present
the
results
of
a
mathematical
model
evaluating
HPV
vaccination
with
a
HPV-16/18
vaccine
in
addition
to
current
screening
at
different
ages
in
Belgium
compared
with
screening
alone
from
the
perspective
of
the
Belgian
health
care
payer.
We
also
investigated
the
role
of
protection
against
non-HPV-
16/18
related
lesions
on
the
cost-effectiveness
of
the
vaccine.
2.
Materials
and
methods
2.1.
Model
design
This
analysis
used
a
previously
published
Markov
model
that
models
a
cohort
of
women
over
a
lifetime,
reproducing
the
natural
history
of
HPV
infection,
the
effect
of
screening
and
the
effect
of
HPV
vaccination
[18].
Briefly,
a
single
cohort
of
girls
(N
=
100,000)
enters
the
model
at
age
12
or
more
and
move
throughout
different
stages
of
the
natural
history
of
cervical
cancer
in
yearly
cycles
governed
by
the
transition
probabilities
in
each
stage.
The
natural
history
in
the
model
is
modified
by
the
effect
of
screening
and
vaccination
in
each
cycle.
This
model
was
adapted
to
include
Belgian-specific
settings
related
to
screening
and
cost.
This
model
does
not
take
into
account
the
economy
that
could
be
derived
from
the
reduction
in
antenatal
hospitalization,
preterm
deliveries
and
neo
natal
cares
that
can
be
awaited
from
the
reduction
in
the
number
of
conisation
resulting
in
the
primary
prevention
of
cervical
dysplasias
[19].
2.2.
Model
inputs
The
transition
probabilities
used
in
the
model
are
shown
in
Table
1.
Whenever
available,
Belgian-specific
data
were
used.
However,
for
parameters
relating
to
the
natural
history
of
the
disease
(i.e.
2
Cervarix
is
a
registered
trademark
of
the
GlaxoSmithKline
group
of
companies.
independent
of
treatment
or
screening)
we
used
published
data
from
other
countries,
as
these
parameters
were
assumed
to
be
identical
across
countries.
The
validity
of
the
model
was
assessed
by
comparing
the
age-specific
cervical
cancer
incidence
and
deaths
estimated
by
the
model
without
vaccination
to
the
cervical
cancer
incidence
reported
by
the
Belgian
cancer
registry.
2.2.1.
Vaccine
effectiveness
From
the
model
construct,
the
vaccine
effectiveness
only
applies
to
the
transition
from
no
HPV
to
HPV.
Vaccine
effectiveness
data
were
derived
from
published
clinical
trials
(Table
1).
For
women
before
the
age
of
sexual
debut
(pre-exposure
to
HPV),
vaccine
effec-
tiveness
was
based
on
vaccine
efficacy
reported
for
girls
and
women
who
were
DNA-negative
for
the
relevant
HPV
type
at
study
entry
(Total
Vaccinated
Cohort
[TVC]
naïve
cohorts)
[7,39].
This
popu-
lation
was
selected
as
representative
of
vaccine
efficacy
among
girls
and
women
pre-HPV
exposure.
For
women
post-HPV
expo-
sure,
we
used
vaccine
efficacy
reported
from
vaccine
clinical
trials
for
women
who
were
HPV
DNA-negative
regardless
of
serostatus
at
baseline
[7]
.
These
data
represent
vaccine
efficacy
against
inci-
dent
infection
among
women
with
and
without
previous
infection
but
without
current
infection.
This
population
even
though
limited
to
25
years
of
age
was
selected
as
the
most
representative
among
available
data
for
vaccination
post-sexual
debut,
in
whom
the
vac-
cine
effect
is
limited
to
prevention
of
incident
infection,
and
who
may
have
had
a
previous
infection
that
has
cleared.
The
cut-off
age
between
pre-
and
post-exposure
was
set
at
17
years,
based
on
the
reported
median
age
of
first
sexual
intercourse
in
industrialised
countries
[20].
Vaccination
before
age
17
years
was
modelled
using
the
pre-exposure
vaccine
efficacy,
while
vaccination
after
age
17
years
was
modelled
using
the
post-exposure
efficacy.
Overall
vaccine
effectiveness
took
into
account
efficacy
against
vaccine
HPV
types
(HPV-16
and
HPV-18),
and
protection
efficacy
against
10
other
HPV
types
(HPV-31/33/35/39/45/51/52/56/58/59)
[7,8,22].
Data
on
efficacy
against
each
HPV
type
were
combined
with
the
proportion
of
each
HPV
type
within
each
lesion
to
estimate
overall
expected
vaccine
effectiveness,
calculated
as
follows:
VE
=
i
%HPV
i
×
Ve
i
where
VE
is
the
vaccine
effectiveness
against
a
specific
lesion;
i
=
HPV-16/18,
HPV-31/33/35/39/45/51/52/56/58/59;
%HPV
i
is
the
frequency
with
which
HPV
type
i
occurs
in
a
specific
lesion;
Ve
i
is
the
type-specific
(i)
vaccine
efficacy.
Vaccination
was
assumed
to
confer
lifelong
protection.
In
sce-
narios
without
protection
against
non-HVP-16/18,
this
component
of
vaccine
effectiveness
was
set
at
0.
2.2.2.
Costs
The
analysis
was
performed
from
the
perspective
of
the
Belgian
healthcare
payers,
and
therefore
included
direct
medical
costs
only.
A
two-round
Delphi
panel
with
six
Belgian
healthcare
prac-
titioners
estimated
the
average
use
of
medical
resources
for
treatment
of
cervical
intraepithelial
neoplasia
(CIN)
of
different
severities
(grade
1
and
grade
2/3),
cervical
cancer
and
the
current
standard
of
Papanicolaou
smear-based
screening.
Unit
cost
data
(in
2010
D
)
obtained
from
Belgian
healthcare
tariffs
[21]
for
each
resource
were
used
to
estimate
the
costs
associated
with
the
treat-
ment
of
each
lesion
and
screening.
This
does
not
take
into
account
additional
costs
such
as
those
related
to
pre-term
deliveries
for
conisation
[22,19].
Page 2
3964 N.
Demarteau
et
al.
/
Vaccine
31 (2013) 3962–
3971
Table
1
Input
data.
Variable
Base
case
value
a
Reference
State:
HPV
Onc
HPV
infection
rate Yearly
incidence
of
HPV
Onc Due
to
lack
of
Belgian
data,
incidence
is
based
on
HPVonc
incidence
by
age
in
France
estimated
from
HPV
prevalence,
progression
to
CIN,
HPV
clearance
and
natural
mortality
[35]
HPV
regression
to
No
HPV
Age
dependent
probabilities
[36–38]
HPV
progression
to
CIN1
0.05
[39]
HPV
progression
to
CIN2/3
0
Spontaneous
progression
within
1
year,
assumed
to
be
0
(i.e.
need
at
least
2
years
to
develop
CIN2/3)
State:
CIN1
and
CIN1
detected
CIN1
detected
0.67
[40]
CIN1
regression
to
No
HPV
0.50
[41,42]
CIN1
progression
to
CIN2/3
0.12
[41]
CIN
treatment
practice
and
efficacy
28%
of
CIN1
detected
patients
undergoing
treatment;
95%
efficacy
of
CIN1
treatment
(patients
returning
to
Normal
(No
HPV)
state
after
treatment)
Expert
opinion
State:
CIN2/3,
persistent
CIN2/3
and
CIN2/3
detected
CIN2/3
regression
to
No
HPV
0.267
[41]
CIN2/3
regression
to
CIN1
0
Assumption
CIN2/3
progression
to
persistent
CIN2/3
0.128
[41]
CIN2/3
detected
0.75
[40]
Persistent
CIN2/3
progression
to
cancer 0.0025
increase
from
20
years
to
40,
0.0025/8
increase
from
40
to
60;
0.0025/16
increase
thereafter
Assumptions
(no
data
available)
CIN2/3
treatment
practice
and
efficacy
96%
of
CIN2/3
detected
patients
undergoing
treatment;
90%
efficacy
of
CIN2/3
treatment
(patients
returning
to
Normal
(No
HPV)
state
after
treatment)
Expert
opinion
State:
Cervical
cancer
Cancer
death
with
ICC
Age-dependent
Cancer
cured 19.0%
[43]
State:
Natural
mortality
Death
rates
Overall
death
rate
(age
specific)
[44]
Utilities
CIN1
detected
0.009376
(0.00750;
0.01125)
[45–49]
CIN2/3
detected 0.009376
(0.00750;
0.01125) [45–49]
Cancer
0.273000
(0.21840;
0.32760)
[45–49]
Cancer
cured
0.062000
(0.04960;
0.07440)
[45–49]
Vaccine
effectiveness
against
HPV-related
lesions
in
Belgium:
Cervarix
®
Pre-exposure,
TVC-naive
CIN1
HPV-16/18
distribution
20.8%
[50]
Vaccine
effectiveness
HPV-16/18
b
98%
(88.4%;
100.0%)
[51,7]
Non-HPV-16/18
c
42.8%
[50]
Vaccine
effectiveness
non-HPV-16/18
48%
(28.9%;
61.9%)
[28,7]
Total
vaccine
effectiveness
CIN1
40.9%
(30.8%;
47.3%)
CIN2/3
HPV-16/18
distribution
51.5%
[50]
Vaccine
effectiveness
HPV-16/18
98%
(88.4%;
100.0%)
[51,7]
Non-HPV-16/18
c
40.3%
[50]
Vaccine
effectiveness
non-HPV-16/18
68%
(45.7%;
82.4%)
[7,8]
Total
vaccine
effectiveness
CIN2/3
77.8%
(63.9%;
84.7%)
Cervical
cancer
HPV-16/18
distribution
77.5%
[50]
Vaccine
effectiveness
HPV-16/18
b
98%
(88.4%;
100.0%)
[51,7]
Non-HPV-16/18
c
17.8%
[50]
Vaccine
effectiveness
non-HPV-16/18
68%
(45.7%;
82.4%)
[7,8]
Total
vaccine
effectiveness
cervical
cancer
88.1%
(76.7%;
92.2%)
Cervarix
®
Post-exposure,
DNA-negative,
regardless
of
serostatus
CIN1
HPV-16/18
distribution
20.8%
[50]
Vaccine
effectiveness
HPV-16/18
89%
(81.6%;
94.0%)
[27]
Non-HPV-16/18
c
42.8%
[50]
Vaccine
effectiveness
non-HPV-16/18
34%
(19.0%;
45.6%)
[27,7]
Total
vaccine
effectiveness
CIN1
33.1%
(25.1%;
39.1%)
CIN2/3
HPV-16/18
distribution
51.5%
[50]
Vaccine
effectiveness
HPV-16/18
d
89%
(81.6%;
94.0%)
[27]
Page 3
N.
Demarteau
et
al.
/
Vaccine
31 (2013) 3962–
3971 3965
Table
1
(Continued)
Variable
Base
case
value
a
Reference
Non-HPV-16/18
c
40.3%
[50]
Vaccine
effectiveness
non-HPV-16/18
47%
(28.2%;
61.6%)
[27,7]
Total
vaccine
effectiveness
CIN2/3 64.7%
(53.4%;
73.2%)
Cervical
cancer
HPV-16/18
distribution
77.5%
[50]
Vaccine
effectiveness
HPV-16/18
d
89%
(81.6%;
94.0%)
[27]
Non-HPV-16/18
c
17.8%
[50]
Vaccine
effectiveness
non-HPV-16/18
47%
(28.2%;
61.6%)
[27,7]
Total
vaccine
effectiveness
cervical
cancer 77.4%
(68.3%;
73.2%)
Cost
data
(D
,
2010)
Regular
screening
negative
pap
34
(27;
41)
[21]
Regular
screening
+
false
positive
37
(29;
44)
Treatment
CIN1
detected
657
(525;
788)
Treatment
CIN2/3
detected
1616
(1293;
1939)
Cancer
(Stage
I–IV)
8190
(6552;
9828)
Vaccine
cost
(full
course)
431
(345;
518)
Discount
cost/outcomes
3.0%/1.5%
(0%/0%;
5%/5%)
CIN,
cervical
intraepithelial
neoplasia;
CIS,
carcinoma
in
situ;
DNA,
deoxyribonucleic
acid;
HPV,
human
papillomavirus;
HPVonc,
oncogenic
human
papillomavirus;
ICC,
invasive
cervical
cancer;
TVC,
total
vaccinated
cohort.
a
Lower
value
upper
value
for
univariate
sensitivity
analysis
when
appropriate.
b
Pre-exposure
vaccine
efficacy
against
HPV-16/18
assumed
98%
against
any
lesion
type,
as
observed
for
CIN2+
[7].
c
Non-vaccine
oncogenic
HPV
types
(HPV-31/33/35/39/45/51/52/56/58/59).
d
Post-exposure
vaccine
efficacy
against
HPV-16/18
conservatively
assumed
89%
against
any
lesion
type,
as
observed
for
CIN1.
2.3.
Utility
Utility
values
for
detected
HPV
lesions
(CIN1,
CIN2/3,
cervical
cancer)
were
extracted
from
published
literature
(Table
1).
We
assumed
that
undergoing
a
pap
test
was
associated
with
no
utility
loss
even
though
it
may
be
the
case
no
study
reported
such
utility
decrement.
2.4.
Discounting
Based
on
the
Belgian
guidelines,
costs
were
discounted
at
an
annual
rate
of
3%
and
health
benefits
at
1.5%
[23,24]).
2.5.
Model
outcomes
All
analyses
were
run
with
a
cohort
of
100,000
girls
aged
12
years,
followed
over
a
lifetime
(95
years).
For
each
scenario,
the
incremental
lifetime
costs,
QALY,
cancer
cases
between
a
cohort
undergoing
vaccination
and
screening,
and
a
cohort
undergoing
screening
alone
were
computed.
The
incremental
cost
and
QALY
were
discounted
at
3%
and
1.5%,
respectively
while
cervical
cancer
cases
were
reported
undiscounted.
The
ICER
represents
the
ratio
between
the
incremental
costs
and
the
incremental
QALY.
2.5.1.
Base
case
The
base
case
compared
two
cohorts,
one
receiving
HPV
vacci-
nation
and
screening,
and
one
receiving
screening
alone.
Screening
alone
(every
3
years
from
age
25
to
65
years
for
59%
of
the
pop-
ulation
[25])
was
compared
with
the
same
screening
plus
HPV
vaccination
with
80%
vaccination
coverage
assumed.
The
vaccina-
tion
scenarios
were
run
with
age
at
vaccination
ranging
from
12
to
40
years
in
increments
of
2
years.
For
each
scenario,
the
model
estimated
the
number
of
cervi-
cal
cancer
cases,
the
number
of
quality-adjusted
life-years
(QALYs)
and
total
cost.
Based
on
these
results,
the
incremental
costs,
QALY,
cervical
cancer
cases
avoided
and
incremental
cost-effectiveness
ratios
(ICER)
were
estimated
for
each
vaccination
scenario
com-
pared
with
no
vaccination.
Two
thresholds
for
cost-effectiveness
were
considered:
very
cost-effective
(<1×
Belgian
gross
domes-
tic
product
[GDP]
per
capita
per
QALY
gained,
i.e.
D
32,200/QALY),
and
cost-effective
(<3×
Belgian
GDP
per
capita
per
QALY
gained,
i.e.
D
96,600/QALY).
GDP
per
capita
was
taken
from
International
Monetary
Fund
data
[26]
,
mid-point
of
2010
values
at
current
and
constant
prices
rounded
to
2
significant
figures.
All
analyses
were
run
with
and
without
protection
against
non-
HPV-16/18.
2.5.2.
Catch-up
scenario
The
number
of
incident
cervical
cancer
and
CIN2+
cases
under
different
catch-up
scenarios
was
estimated
over
40
years.
Two
catch-up
scenarios
were
assessed:
vaccination
from
12
to
18
and
from
12
to
25
years
of
age.
For
each
the
number
of
cervical
cancer
cases
was
summed
over
the
14
cohorts
(12–25
years
of
age)
either
with
or
without
vaccination.
The
analysis
was
conducted
both
with
and
without
protection
against
non-HPV-16/18.
2.6.
Sensitivity
analyses
One-way
and
probabilistic
sensitivity
analyses
were
conducted
to
explore
the
effect
on
the
results
of
uncertainty
in
the
input
val-
ues.
Vaccine
cost,
treatment
cost,
utilities,
and
HPV
incidence
were
varied
from
20%
below
to
20%
above
the
base-case
value.
The
dura-
tion
of
protection
was
reduced
to
25
years
or
10
years
without
the
administration
of
a
booster.
Input
parameters
related
to
vaccine
effectiveness
were
varied
by
reported
confidence
intervals
[7,27],
capped
at
100%.
Discount
rates
of
0%
and
5%
were
applied
for
both
costs
and
health
outcomes,
reflecting
the
range
observed
across
other
countries’
health
economic
guidelines
[23].
Probabilistic
sensitivity
analyses
using
@RISK
software
(Palisade
Corporation,
US)
were
conducted
to
estimate
confidence
intervals
around
the
ICER
for
vaccination
at
15,
20,
25,
30,
35
or
40
years
of
age.
In
these
analyses,
distributions
were
assigned
to
each
variable
(
Table
2).
When
range
data
were
available,
normal
distribution
was
selected,
deriving
the
standard
deviation
from
the
observed
range
and
mean
as
the
base-case
value.
When
no
range
data
were
avail-
able,
a
uniform
distribution
from
80%
to
120%
of
base-case
value
limited
to
100%
was
applied.
For
transition
probabilities,
utilities
and
vaccine
efficacy
data,
the
distributions
were
limited
between
0
and
1.
Ten
thousand
simulations
were
run
with
each
vari-
able
sampled
from
the
distribution
for
each
age
at
vaccination.
A
non-parametric
80%
confidence
interval
was
derived
from
these
simulations.
Page 4
3966 N.
Demarteau
et
al.
/
Vaccine
31 (2013) 3962–
3971
Table
2
Inputs
for
multivariate
probabilistic
sensitivity
analysis.
Variable
Distribution
Distribution
parameters
Reference
Transition
probabilities
Age-dependent
mortality
data Uniform
distribution Age-dependent
Assumption
Age-dependent
HPV
incidence
Uniform
distribution
Age-dependent
Assumption
HPV
Onc
to
CIN1
progression
Normal
distribution
0.05
(SD
0.009)
[39]
Regression
from
CIN1
to
No
HPV
Normal
distribution
0.5
(SD
0.145)
[41,42]
CIN1
Onc
to
CIN2/3
progression
Normal
distribution
0.12
(SD
0.021)
[41]
Regression
from
CIN2/3
to
No
HPV
Normal
distribution
0.267
(SD0.058)
[41]
Proportion
CIN1
Onc
treated Uniform
distribution 0.224–0.336 Assumption
CIN1
treatment
success Uniform
distribution 0.9–1 Assumption
CIN2/3
progress
to
persistent
CIN2/3
Uniform
distribution
0.102–0.154
Assumption
Proportion
CIN2/3
treated
Uniform
distribution
0.768–1
Assumption
CIN2/3
treatment
success
Uniform
distribution
0.72–1
Assumption
Cervical
cancer
to
death
Uniform
distribution
0.066–0.098
Assumption
Cervical
cancer
to
cured
Uniform
distribution
0.152–0.228
Assumption
Utility
data
No
HPV
Fix
(1)
1
Assumption
HPV
Fix
(1)
1
Assumption
CIN1
Fix
(1) 1
Assumption
CIN2/3
Fix
(1)
1
Assumption
Death
Fix
(0)
0
Assumption
Disutility
of
CIN1
detected
Uniform
distribution
0.010–0.015
Assumption
Disutility
of
CIN2/3
detected
Uniform
distribution
0.008–0.011
Assumption
Disutility
of
Cancer Uniform
distribution 0.218–0.328 Assumption
Disutility
of
cancer
cured
0.050–0.074
Assumption
Screening
effectiveness
CIN1
detected
Normal
distribution
0.67
(SD
0.045)
[40]
CIN2/3
detected
0.75
(SD
0.045)
[40,7]
Percentage
estimated
pos
Pap
smear
Uniform
distribution
0.044–0.066
Assumption
Vaccine
effectiveness
Pre-exposure
Vaccine
effectiveness
against
HPV-16/18
Normal
distribution
0.98
(SD
0.045)
[51]
Non-HPV-16/18
CIN2/3
and
CC
a
Normal
distribution
0.68
(SD
0.092)
[8]
Non-HPV-16/18
CIN1
a
Normal
distribution
0.48
(SD
0.084)
[28]
Post-exposure
Vaccine
effectiveness
against
HPV-16/18 Normal
distribution
0.89
(SD
0.045)
[27]
Non-HPV-16/18
CIN2/3
and
CC
a
Normal
distribution
0.47
(SD
0.084)
[27]
Non-HPV-16/18
CIN1
a
Normal
distribution
0.34
(SD
0.067)
[27]
HPV
type
and
CIN
distributions
Proportion
of
HPV-16/18
in
CC
Normal
distribution
0.775
(SD
0.1735)
[50]
Proportion
of
HPV-16
and
-18
among
CIN1
Uniform
distribution
0.166–0.250
Assumption
Proportion
of
HPV-16
and
-18
among
CIN2/3
Uniform
distribution
0.412–0.618
Assumption
Cost
data
HPV-16/18
vaccine
Uniform
distribution
D
324–D
539
Assumption
HPV-16/18
vaccine
booster
cost
Uniform
distribution
Fix
at
one-third
of
the
vaccine
price
Assumption
CC,
cervical
cancer;
CIN,
cervical
intraepithelial
neoplasia;
HPV,
human
papillomavirus;
SD,
standard
deviation.
a
Non-vaccine
oncogenic
HPV
types
(HPV-31/33/35/39/45/51/52/56/58/59).
3.
Results
3.1.
Model
validation
The
cervical
cancer
incidence
modelled
without
vaccination
fit-
ted
the
reported
age-dependent
cervical
cancer
incidence
observed
in
Belgium
between
2001
and
2003.
Similar
results
were
obtained
for
cervical
cancer
deaths.
3.2.
Base
case
Table
3
shows
the
base-case
incremental
costs,
QALY,
cervical
cancers
avoided
and
ICER
for
vaccination
with
the
HPV-16/18
vac-
cine
compared
with
no
vaccination
in
Belgium
for
different
ages
at
vaccination
with
and
without
protection
against
non-HPV-16/18.
At
age
12
years,
with
protection
against
non-HPV-16/18,
vac-
cination
was
highly
cost-effective
(ICER
D
9171/QALY),
and
was
projected
to
save
646
cases
of
cervical
cancer
per
100,000
girls
vaccinated.
As
age
at
vaccination
increased
the
number
of
cases
of
cervical
cancer
avoided
and
the
cost-effectiveness
of
vaccination
decreased
(ICER
increased).
However,
vaccination
at
age
40
years
was
still
projected
to
save
146
cases
of
cervical
cancer
per
100,000
females
vaccinated.
As
expected,
the
cost-effectiveness
and
number
of
cases
of
cervi-
cal
cancer
prevented
were
both
improved
when
protection
against
non-HPV-16/18
was
included.
Protection
against
non-HPV-16/18
resulted
in
an
additional
102
cervical
cancers
prevented
for
vacci-
nation
at
12
years
of
age
to
17
cervical
cancer
cases
for
vaccination
at
40
years
of
age
and
reduced
the
discounted
ICER
by
D
2456
or
D
6181,
respectively.
Page 5
N.
Demarteau
et
al.
/
Vaccine
31 (2013) 3962–
3971 3967
Table
3
Modelled
effect
of
vaccination
for
a
single
age-cohort
(n
=
100,000)
of
12–40-year-old
females
in
Belgium
with
lifetime
vaccination
protection
compared
with
screening
alone
(with
and
without
accounting
for
cross
protection).
Age
(y)
Incremental
cost
vs.
screening
alone
(discounted
3%)
Incremental
QALY
vs.
screening
alone
(discounted
1.5%)
Total
cervical
cancer
cases
avoided
vs.
screening
alone
ICER
(D
/QALY)
Impact
on
HPV-16/18-related
lesions
and
beyond
(with
protection
against
non-HPV-16/18)
12
27,118,169D
2957
646
9171
14
26,963,497D
2942
628
9164
16
27,042,960D
2849
597
9492
18
28,437,184D
2339
482
12,156
20
28,709,073D 2222
450 12,923
22
29,109,872D 2096
418
13,892
24
29,707,625D
1926
380
15,428
26
30,209,689D
1741
340
17,348
28
30,661,509D
1567
303
19,571
30
31,365,112D
1368
264
22,920
32
31,870,671D
1185
228
26,903
34
32,218,222D
1040
200
30,975
36
32,599,077D
926
178
35,201
38
32,754,699D
837
160
39,126
40
32,859,034D 767
146 42,847
Impact
on
HPV-16/18-related
lesions
alone
(without
protection
against
non-HPV-16/18)
12
28,916,356D
2487
544
11,627
14
28,825,167D
2474
529
11,649
16
28,931,404D
2396
502
12,075
18
29,725,188D
2051
423
14,490
20
29,993,099D 1949
395
15,389
22
30,374,664D
1840
367
16,511
24
30,929,007D
1693
334
18,272
26
31,273,112D
1533
300
20,400
28
31,588,949D
1381
268
22,868
30
32,231,791D
1209
233
26,670
32
32,600,352D
1048
202
31,121
34
32,834,653D
920
177
35,671
36
33,192,372D
820
158
40,483
38
33,259,490D
741
142
44,866
40
33,295,522D
679
129
49,028
ICER,
incremental
cost-effectiveness
ratio;
QALY,
quality-adjusted
life-year;
y,
years.
Fig.
1
shows
the
projected
ICER
with
and
without
cross
protec-
tion,
together
with
cost-effectiveness
thresholds.
When
cross
protection
was
included,
the
ICER
remained
under
the
‘very
cost-effective’
threshold
(1×GDP
per
capita,
i.e.
D
32,200)
until
age
35
years,
and
under
the
‘cost-effective’
threshold
(3×GDP
per
capita,
i.e.
D
96,600)
beyond
age
40
years
(the
highest
age
modelled).
Without
protection
against
non-HPV-16/18,
the
ICER
remained
below
the
‘very
cost-effective’
threshold
until
age
33
years,
and
was
still
below
the
‘cost-effective’
threshold
beyond
age
40
years.
3.3.
Catch-up
Fig.
2
shows
the
incident
cases
predicted
by
the
model
for
cohorts
from
12
to
25
years
of
age
over
40
years
post-vaccination.
Extending
catch-up
from
18
to
25
years
resulted
in
additional
cases
prevented
and
a
faster
decrease
in
cervical
cancer
post-
vaccination.
3.4.
Sensitivity
analyses
The
results
of
the
one-way
sensitivity
analyses
are
shown
in
Fig.
3A–F.
The
discount
rate
was
the
most
influential
factor.
With
a
dis-
count
rate
of
5%
the
ICER
was
always
above
the
threshold
of
1×GDP/capita
(D
32,200),
whereas
with
a
discount
rate
of
0%
the
age
at
which
the
ICER
crossed
the
threshold
of
1×GDP/capita
was
above
40
years.
The
other
factors
tested
in
sensitivity
analyses
(vac-
cine
price,
vaccine
effectiveness,
treatment
costs,
utility
values
HPV
incidence
and
vaccine
duration
of
protection)
had
relatively
small
effects
(almost
no
effect
for
treatment
cost
and
utility
value)
on
the
age
at
which
the
ICER
crossed
the
cost-effectiveness
threshold.
The
age
at
which
the
ICER
crossed
the
threshold
of
1×GDP/capita
(D
32,200)
was
26–35
years
for
the
lower
bound
values
and
35–38
years
for
the
upper
bound
values.
Fig.
3H
shows
the
results
of
the
probabilistic
sensitivity
analy-
sis.
These
findings
indicate
that
HPV
vaccination
would
be
expected
to
remain
cost-effective
for
vaccination
up
to
25–30
years
of
age,
accounting
for
the
overall
uncertainty
around
the
model
input
parameters.
The
upper
bound
of
the
ICER
confidence
interval
was
below
the
very
cost-effective
threshold
for
vaccination
at
age
25
years
and
slightly
above
the
threshold
for
vaccination
at
age
30
years.
4.
Discussion
The
modelling
results
presented
here
show
that
screening
plus
vaccination
against
HPV
using
the
HPV-16/18
vaccine
at
age
12
years,
in
Belgium,
would
save
646
cases
of
cervical
cancer
over
the
lifetime
of
a
100,000
girl
cohort,
compared
with
screening
alone,
and
would
be
highly
cost-effective
(estimated
ICER
D
9171/QALY).
These
results
are
consistent
with
other
cost-effectiveness
evalua-
tions
for
Belgium
for
a
vaccination
of
12
years
old
with
base
case
estimates
of
D
10,546
[16]
and
D
32,665
[25]
for
a
quadrivalent
vac-
cine
for
a
vaccine
protecting
against
HPV-16/18.
The
higher
value
from
Thiry
et
al.
is
driven
by
an
assumed
duration
of
protection
of
15
years
instead
of
lifetime
and
lower
vaccine
efficacy.
If
pro-
tection
against
non-vaccine
HPV
types
was
not
included
in
our
analysis,
the
projected
results
were
slightly
less
favourable,
with
vaccination
projected
to
save
544
cases
of
cervical
cancer
at
an
ICER
of
D
11,627/QALY.
Both
with
and
without
protection
against
non-vaccine
types,
the
projected
number
of
cases
of
cervical
cancer
Page 6
3968 N.
Demarteau
et
al.
/
Vaccine
31 (2013) 3962–
3971
Fig.
1.
Impact
of
age
at
vaccination
on
ICER
and
number
of
cervical
cancer
cases
avoided
(with
and
without
protection
against
non-HPV-16/18).
CC,
cervical
cancer;
GDP,
gross
domestic
product
(D
32,200
for
Belgium);
ICER,
incremental
cost-effectiveness
ratio;
QALY,
quality-adjusted
life-year.
prevented
and
the
cost-effectiveness
of
vaccination
progressively
declined
as
age
at
vaccination
increased.
However,
vaccination
was
still
highly
cost-effective
in
young
adult
and
adult
women.
At
a
threshold
for
cost-effectiveness
of
1×
Belgian
GDP/capita
per
QALY
(i.e.
D
32,200),
vaccination
remained
cost-effective
up
to
age
35
years
with
protection
against
non-vaccine
types,
or
up
to
age
33
years
without
protection
against
non-vaccine
types.
At
a
thresh-
old
for
cost-effectiveness
of
3×
Belgian
GDP/capita
(i.e.
D
96,600),
vaccination
was
cost-effective
beyond
the
age
of
40
years
with
or
without
protection
against
non-vaccine
types.
The
projected
number
of
cervical
cancer
cases
prevented
by
protection
against
non-vaccine
types
was
larger
for
vaccination
at
age
12
years
com-
pared
with
vaccination
at
age
40
years
(102
and
17
additional
cases
prevented,
respectively).
However,
the
impact
on
the
ICER
was
larger
for
vaccination
at
older
ages.
This
is
due
to
the
costs
associated
with
cases
prevented
at
different
ages.
At
ages
above
40
years
proportionally
more
of
the
costs
prevented
relate
to
cancer,
while
in
younger
girls
a
greater
proportion
of
the
costs
prevented
relate
to
screening
costs,
which
are
lower
than
cancer
costs.
Thus,
the
cost
averted
for
each
case
prevented
will
tend
to
be
higher
in
older
age
cohorts,
resulting
in
a
larger
impact
of
protection
against
non-vaccine
types
on
ICER
for
vaccination
at
older
ages.
Sensitiv-
ity
analyses
indicated
that
the
results
were
robust
to
variations
in
the
input
parameters.
To
our
knowledge,
this
study
is
the
first
to
evaluate
HPV
vaccination
in
young
adult
and
adult
women
in
Belgium.
Recent
clinical
trial
data
have
reported
that
the
HPV-16/18
vac-
cine
has
protection
efficacy
against
10
non-vaccine
HPV
types
and
that
important
vaccine
efficacy
was
observed
irrespective
of
HPV
DNA
status
and
serostatus
[8,27,28].
The
present
study
is
the
first
to
investigate
the
effect
of
this
protection
against
non-HPV-16/18
data
on
the
cost-effectiveness
of
HPV
vaccination
in
young
adult
and
adult
women.
Our
results
suggest
that
including
protection
against
non-vaccine
HPV
types
increased
the
projected
number
of
cervi-
cal
cancer
cases
prevented
and
improved
the
cost-effectiveness
of
adding
vaccination
to
screening.
Our
findings
indicate
that
extending
HPV
vaccination
pro-
grammes
beyond
age
12
years
could
produce
substantial
additional
health
benefits,
and
that
HPV
vaccination
with
the
HPV-16/18
vac-
cine
would
remain
cost-effective
in
adult
women
up
to
the
age
of
33
years
to
over
40
years
(depending
on
the
cost-effectiveness
thresh-
old
and
whether
protection
against
non-HPV-16/18
is
included)
in
Belgium.
This
is
consistent
with
other
reports
even
though
adapted
to
other
country
specific
settings,
that
extending
HPV
vaccina-
tion
beyond
young
girls
may
be
effective
and
cost-effective.
In
the
Netherlands,
an
evaluation
using
a
Markov
model
found
that
HPV
vaccination
remained
generally
cost-effective
in
women
up
to
age
25
years
[29],
and
some
countries
such
as
Australia
[13]
and
the
US
[14]
already
recommend
HPV
vaccination
up
to
age
26
years.
Adult
women
remain
at
risk
of
HPV
infection
from
new
sexual
relation-
ships,
and
the
protective
effect
of
HPV
vaccination
against
cancer
may
be
seen
more
quickly
in
adult
women
than
after
vaccination
of
adolescent
girls
[30].
Our
analysis
showed
that
the
extension
of
catch-up
cohorts
would
be
expected
to
produce
a
faster
reduction
in
cervical
cancer.
Rapid
clinically
measurable
reductions
in
lesion
incidence
have
been
observed
soon
after
HPV
vaccine
introduction
0
50
100
150
200
250
300
350
4035302520151050
incident CC cases
Cohorts follow-up time (years)
No vaccination
12 to 18 yoa vaccinated + CP
12 to 25 yoa vaccinated + CP
12 to 18 yoa vaccinated
12 to 25 yoa vaccinated
Fig.
2.
Yearly
incident
cervical
cancer
cases
in
the
cohorts
over
time
for
three
vaccination
scenarios
(no
vaccination,
80%
catch
up
from
12
to
18
years
of
age
and
80%
vaccination
from
12
to
25
years
of
age),
both
with
and
without
cross
protection.
CC,
cervical
cancer;
CP,
protection
against
non-HPV-16/18;
yoa,
years
of
age.
Page 7
N.
Demarteau
et
al.
/
Vaccine
31 (2013) 3962–
3971 3969
Fig.
3.
(A–G)
One-way
sensitivity
analyses;
(H)
probabilistic
sensitivity
analyses
(mean
value
of
ICER
accounting
for
protection
against
non-HPV-16/18
and
80%
confidence
intervals).
GDP,
gross
domestic
product
(D
32,200
for
Belgium);
HPV,
human
papillomavirus;
ICER,
incremental
cost-effectiveness
ratio;
QALY,
quality-adjusted
life-year.
in
Australia,
where
recent
data
showed
a
significant
decrease
in
high-grade
cervical
abnormalities
within
3
years
of
the
introduc-
tion
of
HPV
vaccination
for
girls
and
women
aged
12–26
years
[13].
The
strengths
of
our
model
include
its
simplicity,
transparency
and
adaptability
to
reflect
country-specific
epidemiology
and
dis-
ease
management
patterns,
in
this
case
for
Belgium.
It
has
already
been
applied
to
economic
evaluations
of
HPV
vaccination
in
a
range
of
countries
including
Chile,
Finland,
Ireland,
Poland,
Taiwan
and
France
[18,31,15].
The
present
results
apply
to
Belgium,
but
could
also
be
relevant
to
other
countries
with
similar
cervical
cancer
man-
agement
and
HPV
epidemiology.
Our
model
has
limitations.
It
is
not
a
transmission
dynamic
model
and
therefore
cannot
capture
indirect
benefits
such
as
herd
protection,
which
may
underestimate
the
potential
benefits
of
pro-
tection
against
non-HPV-16/18.
On
the
other
hand,
our
model
cannot
account
for
the
existing
herd
effect
on
the
girls
included
in
the
catch-up
and
thus
may
overestimate
cost-effectiveness
for
the
additional
catch-up
cohort.
The
herd
effect
across
age
cohorts
may,
however,
be
more
limited
for
HPV
than
for
other
viruses
such
as
pertussis
or
influenza
[32].
In
addition,
the
model
does
not
account
for
any
expected
effects
of
HPV
vaccination
on
other
HPV-associated
cancers
[33]
or
neonatal
morbidity
and
mortality,
which
would
underestimate
vaccine
cost-effectiveness
[34].
Also
the
input
data
are
mainly
based
on
a
pre-vaccination
situation.
The
implementation
of
the
vaccine
worldwide
may
have
an
impact
on,
e.g.
disease
management
or
a
change
in
HPV
incidence
from
herd
effect
which
may
also
impact
the
cost-effectiveness
of
vaccinating
the
new
cohorts.
We
assumed
vaccine
coverage
of
80%.
As
we
used
a
static
model,
the
vaccine
coverage
rate
does
not
affect
the
ICER
estimate,
although
it
could
affect
the
estimate
of
the
absolute
numbers
of
lesions
prevented.
The
results
presented
here
remain
a
modelling
exercise,
and
are
yet
to
be
confirmed
by
real-world
data
from
long-term
follow-up
studies.
However,
such
data
may
be
several
years
away,
whereas
decisions
about
vaccination
programme
implementation
must
be
made
in
the
near
term.
Our
model
offers
a
valuable
method
of
esti-
mating
the
potential
clinical
and
economic
benefits
of
extending
HPV
vaccination
to
young
adult
and
adult
women.
5.
Conclusions
The
results
presented
here
indicate
that
extending
HPV
vacci-
nation
with
the
HPV-16/18
vaccine,
in
combination
with
screening,
to
adult
and
young
adult
women
post-sexual
debut
could
provide
a
substantial
reduction
in
cervical
cancer
disease
burden
in
Belgium,
compared
with
screening
alone.
HPV
vaccination
with
the
HPV-
16/18
vaccine
is
projected
to
be
cost-effective
in
women
up
to
age
33–40
years.
Acknowledgements
The
authors
thank
Maud
Boyer
(Business
and
Decision
Life
Sci-
ences)
for
publication
co-ordination
and
Carole
Nadin
for
medical
writing
services,
on
behalf
of
GlaxoSmithKline
Biologicals
SA,
Rix-
ensart,
Belgium.
Page 8
3970 N.
Demarteau
et
al.
/
Vaccine
31 (2013) 3962–
3971
Disclosure
statement
ND
and
GVK
are
employees
of
the
GlaxoSmithKline
group
of
companies
and
ND
owns
stock
in
the
GlaxoSmithKline
group
of
companies.
PS
has
performed
consultancy
work
for
GlaxoSmithKline
group
of
companies.
He
received
funding
for
board
membership
and
lec-
tures
from
GlaxoSmithKline
group
of
companies.
None
of
these
activities
was
directly
related
to
the
current
study.
Author
contributions
PS,
GVK
and
ND
conceived
and
designed
the
study;
ND
devel-
oped
the
model;
ND
and
GVK
acquired
the
data;
all
authors
reviewed
and
commented
on
drafts,
and
approved
the
final
manuscript.
Role
of
the
funding
source
This
study,
including
preparation
of
the
manuscript,
was
funded
by
GlaxoSmithKline
Biologicals
SA,
which
was
involved
in:
design
and
conduct
of
the
study;
data
collection,
analysis
and
interpre-
tation;
manuscript
preparation,
review
and
decision
to
submit
for
publication.
References
[1]
Arbyn
M,
Castellsague
X,
de
SS,
Bruni
L,
Saraiya
M,
Bray
F,
et
al.
Worldwide
bur-
den
of
cervical
cancer
in
2008.
Ann
Oncol
2011;22(December
(12)):2675–86.
[2]
Arbyn
M,
Raifu
AO,
Autier
P,
Ferlay
J.
Burden
of
cervical
cancer
in
Europe:
estimates
for
2004.
Ann
Oncol
2007;18(October
(10)):1708–15.
[3]
Belgian
Cancer
Registry.
Annual
new
diagnoses;
2008.
Available
from
URL:
http://kankerregister.nettools.be/default.aspx?url=p
130.htm
[cited
27.12.2011].
[4]
Annemans
L,
Remy
V,
Lamure
E,
Spaepen
E,
Lamotte
M,
Muchada
JP,
et
al.
Economic
burden
associated
with
the
management
of
cervical
cancer,
cervical
dysplasia
and
genital
warts
in
Belgium.
J
Med
Econ
2008;11(1):135–50.
[5]
Walboomers
JM,
Jacobs
MV,
Manos
MM,
Bosch
FX,
Kummer
JA,
Shah
KV,
et
al.
Human
papillomavirus
is
a
necessary
cause
of
invasive
cervical
cancer
world-
wide.
J
Pathol
1999;189(September
(1)):12–9.
[6]
Thiry
N,
De
Laet
C,
Hulstaert
F,
Neyt
M,
Huybrechts
M,
Cleemput
I.
Cost-
effectiveness
of
human
papillomavirus
vaccination
in
Belgium:
do
not
forget
about
cervical
cancer
screening.
Int
J
Technol
Assess
Health
Care
2009;25(April
(2)):161–70.
[7]
Paavonen
J,
Naud
P,
Salmeron
J,
Wheeler
C,
Chow
SN,
Apter
D,
et
al.
Efficacy
of
human
papillomavirus
(HPV)-16/18
AS04-adjuvanted
vaccine
against
cervical
infection
and
precancer
caused
by
oncogenic
HPV
types
(PATRICIA):
final
analy-
sis
of
a
double-blind,
randomised
study
in
young
women.
Lancet
2009;374(July
(9686)):301–14.
[8]
Skinner
R,
Apter
D,
Chow
SN,
Wheeler
C,
Dubin
G.
Cross-protective
efficacy
of
Cervarix
against
oncogenic
HPV
types
beyond
HPV-16/18:
final
analysis
of
cross-protection
PATRICIA
study.
In:
Presented
at
the
25th
international
papillomavirus
conference.
2009.
[9]
Brown
DR,
Kjaer
SK,
Sigurdsson
K,
Iversen
OE,
Hernandez-Avila
M,
Wheeler
CM,
et
al.
The
impact
of
quadrivalent
human
papillomavirus
(HPV;
types
6,
11,
16,
and
18)
L1
virus-like
particle
vaccine
on
infection
and
disease
due
to
oncogenic
nonvaccine
HPV
types
in
generally
HPV-naive
women
aged
16–26
years.
J
Infect
Dis
2009;199(April
(7)):926–35.
[10]
Szarewski
A.
Cervarix(R):
a
bivalent
vaccine
against
HPV