Vaccine 27 (2009) A46–A53
Contents lists available at ScienceDirect
Vaccine
journal homepage: www.elsevier.com/locate/vaccine
Efficacy, duration of immunity and cross protection after HPV vaccination:
A review of the evidence
Paolo Bonanni ∗ , Sara Boccalini, Angela Bechini
Department of Public Health, University of Florence, Viale G.B. Morgagni 48, 50134 Florence, Italy
a r t i c l e
i n f o
Article history:
Received 3 September 2008
Received in revised form 24 October 2008
Accepted 27 October 2008
Keywords:
HPV vaccine
Immunity
Long-term efficacy
Cross protection
a b s t r a c t
The efficacy and immunogenicity of HPV vaccines has proven excellent in several phase 2 and phase 3
trials involving tens of thousand women. A decrease in antibody titres was observed in follow-up studies
of vaccinees, with initial sharp decline reaching a plateau in the longer term. Only few subjects lost their
antibodies during the 5–6 years after vaccination. However, no breakthrough disease occurred even in
those subjects. The administration of a challenge dose of quadrivalent vaccine at month 60 of follow-up
resulted in a strong anamnestic response. The mechanism by which vaccination confers protection and
the reasons for continuing vaccine efficacy remain to be elucidated. The same applies to the possibility of
inducing an anamnestic response following viral challenge via genital mucosa.
Data strongly suggest that both vaccines can have a variable level of cross protection against HPV types
genetically and antigenically-closely related to vaccine types. Demonstration of cross protection against
combined endpoints (CIN2/3 and AIS) for combined HPV types, and, as a single type, for HPV-31, has been
reached for the quadrivalent vaccine, and there is evidence of cross protection against HPV 31 and 45
persistent infections (as single types) for the bivalent vaccine.
Assays used for antibody detection were different for the two vaccines, and standardisation of methods
for anti-HPV L1 protein detection is presently underway. The possibility to use universally accepted tests
for antibody measurement would make comparison between vaccines and among different studies much
easier.
© 2008 Elsevier Ltd. All rights reserved.
1. Introduction
Human Papillomavirus (HPV) is the causative agent of several
skin and mucosal diseases, including virtually all cases of the second most common world female malignancy, cervical cancer, (in
its forms of squamous cancer and adenocarcinoma), and of genital
warts, the most common genital diseases occurring in the sexually
active population [1].
Cancers of other sites are also causally linked to HPV, albeit not
in all cases: vulva (30–85%), vagina (about 60%), penis (about 40%),
anus (70%), larynx (10–20%) and tonsils (about 50%) [2].
More than 100 HPV types have been identified. AlphaPapillomaviruses have a tropism for mucosal surfaces, and at least
13 types have been recently confirmed to be potentially oncogenic
(plus 5 suspect types) by the International Agency for the Research
on Cancer (IARC). Among them, types 16 and 18 are by far the main
responsible for cervical cancer (>70% in all geographical areas), followed by 31, 33, 35, 45, 52 and 58 [3]. In the group of low-risk HPV,
types 6 and 11 are responsible for more than 90% of genital warts.
∗ Corresponding author. Tel.: +39 055 4598511; fax: +39 055 4598935.
E-mail address: paolo.bonanni@unifi.it (P. Bonanni).
0264-410X/$ – see front matter © 2008 Elsevier Ltd. All rights reserved.
doi:10.1016/j.vaccine.2008.10.085
The natural history of HPV infection is rather short in the
majority of subjects, since clearance of the virus or of low-grade
cytological and histological lesions occurs in the majority of cases
in few months to 1–3 years. However, cervical cancer development
is a long process (usually lasting 15–20 years). It implies the persistence of infection with a high-risk HPV type in a minority of infected
women, leading to pre-cancerous lesions in the middle term (3–5
years), and eventually to the development of invasive cancer in the
long term (>10 years) [4,5].
The course and progression of pre-cancerous lesions can be
modified through adequate early treatment when cytological
abnormalities (Low-grade Squamous Intraepithelial Lesions or LSIL,
and High-grade Squamous Intraepithelial Lesions or HSIL) are
detected following the periodical collection of cervical smears (Paptest). The consequent sharp decrease of cervical cancer incidence
in countries where screening has been widely applied represents
one of the most important successes of public health interventions.
Unfortunately, many economic and organisational barriers hamper the implementation of secondary prevention programmes in
developing countries, where the disease burden is highest.
The development of HPV vaccine is a landmark in the history
of immunisation, since this is the first vaccine primarily directed
and perceived as an anti-cancer vaccine (although, in reality, the
P. Bonanni et al. / Vaccine 27 (2009) A46–A53
hepatitis B vaccine was the first anti-cancer vaccine). The Human
Papillomavirus vaccine has the potential to complement secondary
prevention in developed countries, and to control cervical cancer
morbidity and mortality worldwide if solutions will be found to
implement its use in girls living in developing countries.
The two presently available vaccines (quadrivalent and bivalent)
have proven very efficacious in the prevention of pre-cancerous
lesions (Cervical Intraepithelial Neoplasia – CIN of grade 2+), and
can exert their maximum efficiency if used at pre-adolescent age,
when sexual activity has not yet started [6]. However, they need to
extend their protective effect for many years if a substantial impact
on HPV-related diseases has to be achieved. Furthermore, the possibility to demonstrate a certain degree of cross-protection towards
HPV types genetically and antigenically related to those included in
the vaccines would add further value to their preventive potential.
The aim of this review is to present the data available to date
on long-term duration of immunity and on cross-protection, and
to highlight what is still unknown, in need of further confirmation
or difficult to demonstrate.
2. Defence mechanisms to naturally occurring HPV
infection
HPV has developed an extremely efficient strategy to evade the
attack of the host defence systems (innate and adaptive). As a matter of fact, the strict tissue tropism for squamous epithelial cells has
key implications for the usually poor response to the presence of
the infecting virus. HPV enters the basal layer of squamous epithelia (keratinocytes), where it expresses the so-called ‘early’ genes
implicated in the regulation of the viral cycle. Viral replication follows the evolution of the keratinocyte up to the stratum spinosum
and granulosum, where viral proteins expression and assembly of
new virions occur at maximum level. HPV is shed through epithelial
desquamation taking place approximately three weeks after infection. Therefore, no major cell damage occurs as a consequence of
HPV replication. The lack of cell death together with the absence
of inflammation (usually acting as activation signals) are the main
reasons for the poor response of the innate and adaptive defence
systems, together with the fact that no viraemic phase occurs, and
therefore the systemic immune system is not or is poorly activated.
Furthermore, the production of complete virions and their release
occurs far from the basal epithelium, where the contiguity to the
immune system would help a stronger response. Finally, the fact
that HPV is a double-stranded DNA virus without any RNA intermediate further explains the lack of activation signals for innate
immunity [7].
Notwithstanding all the just enumerated down-regulatory
mechanisms, humoral and cellular responses to HPV infection do
exist. Antibody production against L1 proteins appears to be linked
to the presentation of conformational, type specific epitopes, that
induce slowly increasing (taking up to 18 months) and low-level
IgG titres only in a part of those infected (59.5%, 54.1% and 68.8%
for HPV-16, -18 and -6, respectively, in a study of incident infection in 588 women attending a college) [8]. Such antibodies have
a clear neutralizing effect, as shown by the evidence that seropositive animals do not get infected after viral challenge [9], and by the
possibility to protect dogs immunised with Canine Oral Papillomavirus (COPV), Virus-Like Particles (VLPs) against viral challenge,
and to transfer such protection to susceptible dogs [10].
Cellular immunity is responsible for viral clearance from
infected cells and for the resolution of HPV-related lesions. As
a matter of fact, studies on spontaneously regressing genital
warts show large infiltrates of both CD4+ and CD8+ T cells and
macrophages in the wart stroma and epithelium, with increase in
A47
the level of pro-inflammatory cytokines [11,12]. These findings are
further confirmed by studies on the evolution of the wart cycle in
animal models, and are indicative of a Th1-type immune response.
The production of neutralizing antibodies follows wart regression,
and, although the usually low levels reached, such Ig are able to
prevent re-infection with the same HPV type.
The central role of cellular immunity for viral clearance is
confirmed by studies on patients under immunosuppressive treatments or infected by HIV. Typically, in renal transplant recipients,
the incidence of squamous intraepithelial lesions and of genital warts is increased in comparison to the immune-competent
population [7,13,14], while HIV-infected subjects showed higher
prevalence and longer duration of anogenital lesions [7,15,16].
3. Efficacy of HPV vaccines in clinical trials
Two vaccines based on the production of L1 recombinant proteins were developed and tested in large-scale clinical trials. The
L1 protein, produced in the form of pentamers, is able to spontaneously re-assemble to form Virus Like Particles (VLPs), sort of
‘empty shells’ reproducing the conformation of HPV but containing
no infectious DNA. This property of reconstituting VLPs is extremely
favourable for the induction of a systemic immune response following the administration of the vaccine.
The first vaccine to be produced and approved by the US Food
and Drugs Administration (FDA) and by the European Medicines
Agency (EMEA) was the quadrivalent vaccine, that includes VLPs of
the HPV types 16, 18, 6 and 11, and therefore aims at the prevention
of about 70% of cervical cancer cases, about 90% of genital warts, and
of vulvar and vaginal lesions. The vaccine is produced in yeast cells
and is adjuvanted with amorphous aluminium. The immunisation
schedule includes three injections at month 0, 2, 6.
The efficacy of the quadrivalent vaccine was assessed in four
clinical trials (phase 2 and 3) involving about 20,800 women overall,
with an age range 16–26 years.
The results were evaluated in three different populations [17]:
(1) Per protocol susceptible population, defined as subjects who
received all three doses of vaccine or placebo within 12 months;
were seronegative and HPV–DNA negative on PCR analysis for
HPV-6, -11, -16 or -18 at day one; remained negative on PCR
analysis for the same HPV type (to which they were negative
at day one) through one month after the third dose; had no
major protocol violations; but were included even if the result
on cervical cytologic examination at day one were abnormal.
(2) Unrestricted susceptible population, defined as subjects who
were seronegative and HPV–DNA negative on PCR analysis for
HPV-6, HPV-11, HPV-16, or HPV-18 at day one; were included
even if protocol violations were present; were included even
if results on cervical cytologic examination at day one were
abnormal.
(3) Intention-to-treat general study population, defined as subjects who were included even if they had infection or disease
associated with HPV-6, HPV-11, HPV-16, or HPV-18 (i.e.,
cervical intraepithelial neoplasia, vulvar intraepithelial neoplasia, or vaginal intraepithelial neoplasia) before vaccination;
were included even if protocol violations were present; were
included even if results on cervical cytologic examination at
day one were abnormal.
The per protocol and unrestricted susceptible populations
reflect the possible vaccine effectiveness in cohorts of preadolescent girls that are the object of routine immunisation
campaigns, while the intention-to treat population could indicate
A48
P. Bonanni et al. / Vaccine 27 (2009) A46–A53
Table 1
Primary efficacy results with combined database of efficacy studies of quadrivalent human papillomavirus (HPV) types 6, 11, 16, and 18 (HPV-6/11/16/18) vaccine (per-protocol
population). Median duration of follow-up: three years after vaccination dose one. Adapted from Barr and Tamms [18].
End point
HPV vaccine
No. of subjects
No. of cases
No. of subjects
No. of cases
HPV-16- or HPV-18-related CIN2 or 3 or AIS
HPV-16- or HPV-18-related VIN2 or 3
HPV-16- or HPV-18-related VaIN2 or 3
HPV-6-, HPV-11-, HPV-16-, or HPV-18–related CIN (CIN1, CIN2/3) or AIS
HPV-6-, HPV-11-, HPV-16-, or HPV-18-related genital warts
8492
7771
7771
7863
7899
1
0
0
6
2
8462
7742
7742
7863
7900
85
8
7
148
160
the impact of vaccine implementation in a ‘real world’ situation,
i.e. a population including both susceptible and already infected
women, and even subjects with lesions due to the relevant HPV
types.
The results reported for the per protocol population and the
unrestricted susceptible population show the extremely high efficacy in the prevention of CIN2+, VIN2+, VaIN2+ lesions and genital
warts related to vaccine types, with values always >95% (Tables 1–2)
[18]. In a combined analysis of several randomized clinical trials,
the per-protocol analysis demonstrated that after a mean of three
years, vaccine efficacy for the primary endpoint of the combined
incidence of HPV-16 and -18-related CIN2/3, or adenocarcinoma in
situ was 99% [19].
The bivalent vaccine is presently approved by the EMEA for
use in European countries. It contains VLPs from the two main
oncogenic types, 16 and 18, and is intended to prevent about 70%
of cervical cancer cases. The vaccine is produced in insect cells
(Trichoplusia ni derived cells) through Baculovirus vector, and is
adjuvanted with the proprietary adjuvant named AS04, containing
aluminium hydroxide plus Mono-Phosphoryl Lipid A (MPL). The
immunisation schedule foresees three doses at month 0, 1 and 6.
The bivalent vaccine was evaluated in phase 2 and phase 3 trials.
The initial study had the primary objective to assess vaccine efficacy in the prevention of infection with HPV-16, HPV-18, or both
(HPV-16/18), between month 6 and 18 in participants who were
initially shown to be seronegative for HPV-16/18 by ELISA and negative for HPV-16/18 DNA by PCR. Secondary objectives included:
evaluation of vaccine efficacy in the prevention of persistent infection with HPV-16/18, and the evaluation of vaccine efficacy in the
prevention of histologically confirmed LSIL (CIN1), HSIL (CIN2 or
3) squamous cell cancer, or adenocarcinoma associated with HPV16/18 infection between months 6 and 18, and months 6 and 27.
Placebo
%Efficacy
98.8
100.0
100.0
96.0
98.8
In the per protocol analysis, efficacy in the prevention of both incident and persistent infection against HPV types 16–18 was virtually
100%, and also results in the intention to treat population showed
the high efficacy profile of the vaccine (84–100%, according to the
measured outcome) [20].
The phase 3 study was conducted on about 18,000 subjects
treated with either HPV or HAV vaccine (placebo group). The study
population was defined as Total Vaccinated Cohort for Efficacy
(TVC-E), which included women who had prevalent oncogenic HPV
infections, often with several HPV types, as well as low-grade cytological abnormalities at study entry and who received at least one
vaccine dose.
Cervical cytology was assessed and subsequent biopsy for 14
oncogenic HPV types by PCR was performed. The primary endpoint,
vaccine efficacy against cervical intraepithelial neoplasia (CIN) 2+
associated with HPV-16 or HPV-18, was assessed in women who
were seronegative and DNA negative for the corresponding vaccine
type at baseline (month 0) and allowed inclusion of lesions with
several oncogenic HPV types. This interim event defined analysis
was triggered when at least 23 cases of CIN2+ with HPV-16 or HPV18 DNA in the lesion were detected in the total vaccinated cohort for
efficacy. During a mean follow-up time of 14.8 months, the 23 cases
needed for interim analysis occurred. When results were analysed
according to the pre-specified definition of vaccine type related
lesion in the vaccine and the placebo group, two cases of CIN2+
associated with HPV-16 or HPV-18 DNA were seen in the HPV-16/18
vaccine group; 21 were recorded in the control group. Vaccine efficacy against CIN2+ containing HPV-16/18 DNA was therefore 90.4%.
However, of the 23 cases, 14 (two in the HPV-16/18 vaccine group,
12 in the control group) contained DNA of several oncogenic HPV
types. Additional analyses were done to attribute a likely causal
association to an HPV type. The attribution of causality was based
Table 2
Population impact of human papillomavirus (HPV) types 6, 11, 16, and 18 (HPV-6/11/16/18) vaccination in young women. Median duration of follow-up: three years after
vaccination dose one. Adapted from Barr and Tamms [18].
End point, analysis
HPV vaccine
No. of subjects
Placebo
No. of subjects
No. of cases
3
139
142
9400
–
9897
121
134
255
97.5 (92.6–99.5)
–
44.3 (31.4–55.0)
8641
–
8954
1
8
9
8668
–
8964
29
2
31
96.5 (79.1–99.9)
HPV-6-, HPV-11-, HPV-16-, or HPV-18-related CIN or AIS
Prophylactic efficacy (unrestricted susceptible population)
HPV-6, HPV-11, HPV-16, and/or HPV-18 positive at day 1
General population impact (all randomized subjects with follow-up data)
8628
–
8817
12
180
192
8673
–
8847
219
195
414
94.5 (90.3–97.2)
–
53.9 (45.1–61.3)
HPV-6-, HPV-11-, HPV-16-, or HPV-18-related genital warts
Prophylactic efficacy (unrestricted susceptible population)
HPV-6, HPV-11, HPV-16, and/or HPV-18 positive at day 1
General population impact (all randomized subjects with follow-up data)
8760
–
8954
10
51
61
8787
–
8964
215
51
266
95.4 (91.3–97.8)
–
77.2 (69.8–83.0)
HPV-16- or HPV-18–related CIN2 or 3 or AIS
Prophylactic efficacy (unrestricted susceptible population)
HPV-16 and/or HPV-18 positive at day one
General population impact (all randomized subjects with follow-up data)
9344
–
9834
HPV-16- or HPV-18-related VIN2 or 3 and VaIN 2 or 3
Prophylactic efficacy (unrestricted susceptible population)
HPV-16 and/or HPV-18 positive at day 1
General population impact (all randomized subjects with follow-up data)
No. of cases
Reduction, % (95% CI)
70.9 (37.4–87.8)
P. Bonanni et al. / Vaccine 27 (2009) A46–A53
A49
Table 3
Detection of CIN2+ lesions with HPV-16 or HPV-18 DNA in the vaccine group and in the control group after administration of the bivalent (16/18) HPV vaccine (assessment
of efficacy over a mean period of 15 months from start of vaccination). Adapted from Paavonen et al. [21].
CIN2+ with HPV-16 or HPV-18 DNA in lesion
CIN2 with HPV-16 or HPV-18 DNA in lesion
CIN3 with HPV-16 or HPV-18 DNA in lesion
CIN2+ with HPV-16 or HPV-18 plus other oncogenic types of HPV
CIN2+ with several oncogenic types of HPV in which HPV-16 or HPV-18 detected for the first time
CIN2+ with HPV-16 or HPV-18 DNA in lesion and in preceding cytology sample
on the presence of an oncogenic HPV infection preceding the development of CIN. If more than one HPV DNA type was detected in a
lesion, the presence of HPV types in one of two immediately preceding cytology sample(s) was considered; where the HPV type
present in both the lesion and in one of two immediately preceding cytology sample(s) was the same, this type was considered to
be causally associated with that lesion. According to this approach,
the two lesions assigned to vaccine types in vaccinees were in reality attributable to other co-infecting HPV types, thus giving 100%
efficacy values (Tables 3–4) [21].
Vaccine group
Control group
Total
2
1
1
2
2
0
21
16
5
12
1
20
23
17
6
14
3
20
ration, Austin, TX, USA). Briefly, antibody titres were determined
in a competitive format–i.e., known, type-specific phyco-erythrinlabelled, neutralising antibodies compete with serum antibodies
from the vaccinee for binding to conformationally-sensitive, neutralising epitopes on VLPs [22].
On the other hand, serological testing for antibodies to HPV16 and HPV-18 virus-like particles for the bivalent vaccine was by
ELISA. Recombinant HPV-16 or HPV-18 virus-like particles were
used as coating antigens for antibody detection [20]. All this means
that, for published data, direct comparison of titres achieved after
immunisation with the two vaccines is not possible. Efforts are
currently underway to reach standardisation of methods for the
measurement of anti-L1 antibodies. In this respect, the availability
of VLPs from the two vaccine producers to be used as coating antigen for the solid phase of assays would allow to obtain comparable
data. Furthermore, in the last months, the technique of pseudovirions neutralisation has gained much popularity among experts in
the field.
Secondly, no minimum protective antibody level was determined up to now, due to the excellent efficacy and immunogenicity of the vaccine and to the absence of breakthrough
lesions also in immunised subjects who lost anti-HPV over
time.
4. Immunogenicity of HPV vaccines, antibody
measurement and persistence of protection
There are several aspects of the immune response to HPV vaccines that must be taken into account in order to understand the
available information and the problems to face in the collection of
further data on duration of protection.
As a first point, methods used in the assessment of immunogenicity for the two vaccines are different, so that comparability of
results is not warranted.
As a matter of fact, the response to the quadrivalent vaccine
was measured using a competitive immunoassay (Luminex Corpo-
Table 4
Efficacy against CIN2+ and CIN1+ associated with HPV-16/18 in the total vaccinated cohort for efficacy after administration of the bivalent (16/18) HPV vaccine (assessment
of efficacy over a mean period of 15 months from start of vaccination). Adapted from Paavonen et al. [21].
Endpoint
Group
N
CIN2+ pre-specified case definition based on PCR detection in lesion only
HPV
7788
CIN2+ HPV-16/18
Control
7838
HPV
6701
CIN2+ HPV-16
Control
6717
HPV
7221
CIN2+ HPV-18
Control
7258
n
Vaccine efficacy (97.9% CI)
2
21
1
15
1
6
%
LL
90.4
53.4
99.3
<0.0001
93.3
47.0
99.9
0.0005
83.3
−78.8
99.9
0.1249
CIN2+ additional-post hoc-analysis considering patterns of HPV types in preceding cytological samples
HPV
7788
0
CIN2+ HPV-16/18
100
Control
7838
20
HPV
6701
0
CIN2+ HPV-16
100
Control
6717
15
CIN2+ HPV-18
HPV
Control
7221
7258
CIN1+ pre-specified case definition based on PCR detection in lesion only
HPV
7788
CIN1+ HPV-16/18
Control
7838
HPV
6701
CIN1+ HPV-16
Control
6717
HPV
7221
CIN1+ HPV-18
Control
7258
0
5
3
28
2
18
1
11
UL
P-value
74.2
100
<0.0001
64.5
100
<0.0001
100
−49.5
100
0.0625
89.2
59.4
98.5
<0.0001
88.9
44.6
99.2
0.0004
90.9
22.1
99.9
0.0063
CIN1+ additional-post hoc-analysis considering patterns of HPV types in preceding cytological samples
HPV
7788
1
CIN1+ HPV-16/18
96.1
Control
7838
26
HPV
6701
1
94.1
CIN1+ HPV-16
Control
6717
17
HPV
7221
0
CIN1+ HPV-18
100
Control
7258
9
71.6
54.3
33.8
100
99.9
100
<0.0001
<0.0001
0.0039
A50
P. Bonanni et al. / Vaccine 27 (2009) A46–A53
In an extended follow-up (60 months after the first dose) of 241
recipients of three doses of quadrivalent vaccine, immunoglobulin
levels were measured using a competitive Luminex immunoassay
(cLIA) and were reported in arbitrary units (milli-Merck units per
milliliter or mMU/mL) relative to the standard curves generated for
each individual HPV type. The results showed a decline of serum
anti-HPV levels that was maximum for the first 24 months, while
a trend to plateau (much slower decline) was registered thereafter.
Antibody levels remained highly above the titre reached after natural infection for HPV type 16, and, at a lesser extent, for type 6,
while for types 11 and 18, the final titres were comparable to those
registered in naturally infected subjects [23].
The study of the antibody kinetic after administration of the
bivalent vaccine has been recently reported for a follow-up of
6.4 years in 776 vaccinees [24]. Participants to the initial efficacy
study from North America and Brazil who were aged 15–25 years,
seronegative to HPV-16 and -18, DNA negative for 14 oncogenic
types and received three vaccine doses were enrolled in the followup study. Antibody levels were assessed by ELISA (EU/ml) and
neutralisation assay in the per-protocol population. More than 98%
of vaccinees remained HPV-16 and -18 seropositive, and antibody
levels were several fold higher than natural infection levels for both
total IgG and neutralising antibodies. During this follow-up time,
efficacy against HPV-16/18 related CIN2+ lesions was 100%.
The long-term duration of the antibody response induced by a
monovalent HPV-16 vaccine was estimated by mathematical modelling of the antibody levels measured during a 48 month period
following vaccination, using a power-law model of antibody decline
based upon the biological dynamics of B cell turnover, and a modification of this model, which additionally allows for the long-term
persistence of a memory B cell subpopulation. Although both models acceptably fitted the data and provided a range of long-term
predictions, a better fit was provided by the modified model, which
predicted a near life-long persistence of detectable antibodies following HPV-16 vaccination in a majority of women. As a matter
of fact, using the antibody decay model, it was estimated that following administration of a three-dose regimen of HPV-16 vaccine
in women aged 16–23 years, anti-HPV-16 levels will remain above
those induced naturally by HPV-16 infection for 12 years, and above
detectable levels for 32 years in 50% of vaccinees. With the modified model, which fitted the data better (p < 0.001), it was estimated
that near life-long persistence of anti-HPV-16 following vaccination
is expected at titre levels above those associated with reduction
of natural HPV-16 infection in 76% of these subjects, and above
detectable levels in 99% of these subjects [25]. However, also the
kinetics of the other three HPV types included in vaccines need to
be studied and modeled when sufficiently long-term follow-up will
be available.
A crucial aspect for the evaluation of long-term protection is the
creation, the persistence and the way the immune memory to HPV
is stimulated.
Although hepatitis B is a systemic pathogen, while HPV remains
confined to the genital mucosa, the 25 year experience with the
hepatitis B vaccine suggests that specific lymphoproliferation, the
in vivo humoral response, and the generation of memory B and T
cells were the reasons of the long-term persistence of protection
against HBV infection in spite of the fall of anti-HBs to undetectable
levels in some immunised subjects [26–28].
According to the guidelines of WHO, induction of immune memory should be assessed by means of evaluating immune responses
to additional doses of vaccine administered at planned intervals
following completion of the primary series [29].
The boosting of quadrivalent vaccine recipients with a new dose
five years after initiation of the immunisation schedule, and the
demonstration of a clear anamnestic effect, is a proof that immune
memory had been established and persisted during the follow-up
period, also taking into account that a proportion of subjects in
the group that participated in the extension of the phase 2 trial
were found to be seronegative to one or more vaccine HPV types at
month 60 [23]. It is noteworthy that no breakthrough cases of confirmed infection with HPV vaccine types occurred during the five
year period in the group of vaccinees, while continuing occurrence
of infections in subjects belonging to the placebo group indicates a
sustained protection for the duration of the follow-up period.
An important issue concerns the fact the route by which the
immune memory is stimulated through a challenge dose of vaccine (parenteral) is different form that of a possible viral challenge
occurring through a sexual intercourse (mucosal). Whether natural exposure to HPV induces an anamnestic response to L1 protein
in vaccinees remains to be demonstrated. At the same time, the
mechanism of protection conferred by vaccination has not been
completely elucidated. The most likely hypothesis is that antibodies
pass through from blood to the cervical mucus, where they would
be able to neutralise vaccine types HPV virions, as show for the
bivalent vaccine [30].
How seronegative vaccinees continued to be protected during
the 60 month follow-up is the subject of a lively debate. It could be
hypothesised that even at very low, undetectable titre, neutralising
antibodies are still effective, or that other neutralising unknown
antibodies induced by vaccination exist [23]. On the other side, the
recall effect of viral challenge at the mucosal level might play an
important role. The implications of the response to this question
are crucial to understand whether an elevated titre of anti-L1 is
an obliged pre-requisite for long-term protection, or if immunity
can rely on the anamnestic response of memory B cells to nascent
infection.
The prevention of pre-cancerous lesions (CIN2 and CIN3) is considered as the standard surrogate of protection of HPV vaccines
against cancer. However, we need to demonstrate also the impact
on cervical cancer, which is obviously only measurable as a longterm outcome.
Since about 25 million people living in Nordic European countries (Denmark, Finland, Iceland, Norway and Sweden) are subject
to countrywide registration of health events, they represent an ideal
population for the evaluation of HPV vaccine long-term effectiveness in cancer prevention. For this reason, two population based
phase 3–4 trials are currently underway in those countries in order
to provide evidence of long term protection against invasive cervical cancer and CIN3 using the cancer registry follow-up. Results are
awaited between 2015 and 2020 [19,31].
5. Cross protection and cross neutralisation
Immunity to HPV is type-specific. However, if we look at the phylogenetic tree including the different HPV types, we realise that a
certain degree of cross protection is possible, given the high homology of some viral types with vaccine types (Figure 1)[32]. This is the
case, for instance, for HPV-31 and -35 (strictly related to HPV-16),
and for HPV-45 (strictly related to HPV-18).
In order to understand this issue, and the approaches to study
the problem, we need to distinguish between cross-neutralisation
and cross protection. Cross neutralisation means that antibodies
elicited by vaccination with a HPV type neutralise virions of another
HPV type at a variable degree in vitro. Cross protection means that
immunisation with a certain vaccine type provides clinically significant protection against infection or disease (or both) due to another
HPV type.
A study conducted on ten subjects seronegative and HPV–DNA
negative at baseline for HPV-6, -11, -16, -18, -31 and -45, who were
P. Bonanni et al. / Vaccine 27 (2009) A46–A53
A51
Fig. 1. Type specificity of HPV: phylogenetic tree of Human Papilloma Virus. Adapted from Wieland et al. [32].
immunised with three doses of the quadrivalent vaccine, showed
that serum antibodies from 10/10 women neutralized HPV-18 pseudovirions, six out of 10 neutralized HPV type 45 pseudovirions,
and eight out of 10 neutralized HPV type 31 pseudovirions [33].
Therefore, cross neutralisation of vaccine-induced antibodies versus related HPV types was demonstrated.
Data were also reported in the literature on cross protection
against incident infection, persistent infection and pre-cancerous
lesions (CIN2+) in an analysis conducted as single type.
The study on the bivalent vaccine (phase 2 study) at 4.5
years of follow-up showed a significant protection against incident infection with HPV-45 ((one case/528 vaccinated women
and 17 cases/518 controls; vaccine efficacy: 94.2%; 95% c.i.:
63.3–99.9%) and HPV-31 (14 versus 30 cases, in the above groups,
respectively; vaccine efficacy: 54.5%; 95% c.i.: 11.5–77.7%) [34].
Similar results were also reported at 5.5 years of follow-up
[35].
More recently, data were reported for the bivalent vaccine on
efficacy against persistent infection at month 6 and 12 of follow-up
after vaccination [21]. The point efficacy value at month six was
59.9%, for HPV-45, and 46.1% for HPV-31. The overall point efficacy
against five oncogenic types other than 16 and 18 (namely, 45, 31,
33, 52 and 58) at month 12 of follow-up was 27.1%. However, 95%
confidence intervals for these estimates were very wide (Table 5).
For the quadrivalent vaccine, data were measured as efficacy
against CIN2/3 or Adenocarcinoma In Situ (AIS) during a three year
follow-up. As shown in Table 6, data were reported for combined
HPV-31/45 types (point efficacy: 62%), five combined HPV types,
Table 5
Vaccine efficacy against persistent infections with oncogenic HPV types in the total vaccinated cohort for efficacy. Adapted from Paavonen et al. [21].
Endpoint
6-month persistent infection with HPV-16/18
DNA negative and
seronegative at study entry
Group
N
n
Vaccine efficacy
(97.9% CI)
P-value
N
n
Vaccine Efficacy
(97.9% CI)
P-value
Type 16/18
HPV
Control
HPV
Control
HPV
Controls
6344
6402
5493
5520
5896
5939
38
193
23
144
15
58
80.4%
(70.4 to 87.4)
84.1%
(73.5 to 91.1)
74.0%
(49.1 to 8.8)
<0.0001
3386
3437
2945
2972
3143
3190
11
46
7
35
4
12
75.9%
(47.7 to 90.2)
79.9%
(48.3 to 93.8)
66.2%
(−32.6 to 94.0)
<0.0001
Type 16
Type 18
12 month persistent infection with HPV-16/18
<0.0001
<0.0001
Endpoint
6-month persistent infection with oncogenic HPV types
Type specific DNA negative
at study entry
Group
N
n
Vaccine efficacy
(97.9% CI)
P-value
Type 45
HPV
Control
HPV
Control
HPV
Control
HPV
Control
HPV
Control
HPV
Control
HPV
Control
6724
6747
6615
6667
6702
6736
6532
6573
6688
6734
6773
6804
6773
6804
10
25
47
74
31
49
79
116
43
33
505
554
545
691
59.9%
(2.6 to 85.2)
36.1%
(0.5 to 59.5)
36.5%
(−9.9 to 64.0)
31.6%
(3.5 to 51.9)
−31.4%
(−132.1 to 24.7)
9.0%
(−5.1 to 21.2)
21.9%
(10.7 to 31.7)
0.0165
Type 31
Type 33
Type 52
Type 58
Oncogenic HPV other than
vaccine types
Oncogenic HPV
<0.0001
0.0766
12-month persistent infection with oncogenic HPV types
0.0173
0.0560
0.0093
0.2515
0.1410
<0.0001
N
n
Vaccine Efficacy
(97.9% CI)
3584
3601
3527
3568
3574
3603
3489
3508
3563
3601
3611
3632
3611
3632
3
8
15
17
6
11
16
30
6
6
100
137
112
180
62.3%
(−93.2 to 95.4)
10.8%
(−115.2 to 63.6)
45.1%
(−91.8 to 86.5)
46.5%
(−12.3 to 75.8)
−1.1%
(−372.0 to 78.4)
27.1%
(0.5 to 46.8)
38.2%
(18.0 to 53.7)
P-value
0.2262
0.8598
0.3318
0.0533
1.000
0.0174
<0.0001
A52
P. Bonanni et al. / Vaccine 27 (2009) A46–A53
Table 6
Disease cross-protection analysis: efficacy against high grade cervical dysplasia (CIN2/3 or AIS) – three years analysis.
CIN 2/3 or AIS due to
Quadrivalent vaccine N = 4616
Placebo N = 4675
Efficacy
95% CI
HPV-31/45
HPV-31/33/45/52/58
10 oncogenic HPV types (non-vaccine types) 31, 33, 35, 39, 45, 51, 52, 56, 58, 59
8
27
38
21
48
62
62%
43%
38%
(10, 85)
(7, 66)
(6, 60)
31/33/45/52/58 (point efficacy: 43%) and 10 combined HPV types,
31/33/35/39/45/51/52/56/58/59 (point efficacy: 38%) [36].
Those 10 additional serotypes are responsible for about 16% of
cervical cancer cases in Europe and about 22% worldwide [34].
Should HPV vaccines demonstrate the ability to prevent at least a
fraction of pre-cancerous lesions due to additional HPV types, they
might add further value to their efficacy profile. However, although
data presented above are strongly suggestive of a cross-protection
effect, its definitive demonstration is very hard to reach, given the
low numbers of CIN2/3 due to HPV types other than 16 and 18,
and for the time being it has been accepted by EMEA regarding
the quadrivalent vaccine for some HPV types. As a matter of fact,
the clinical trials performed to date were not designed to evaluate
vaccine impact on rare HPV type-related lesions, and this explains
the very wide confidence intervals registered for efficacy against
additional single oncogenic types.
6. The Italian policy of HPV vaccination
Italy was the first European country to give a positive opinion
to the introduction of HPV vaccination in 2006, taking the decision
to recommend and to offer the vaccine free of charge to all 12year old girls (between the 11th and the 12th birthday). However,
since implementation of immunisation policies is a responsibility of
regional health authorities, the start of the vaccination programme
was completed in all 21 regions/autonomous provinces only in
2008.
Some regions have already decided to offer also a free of charge
catch up for one or more other age cohorts of their population, and
some others are expected to expand the cohorts involved in the
immunisation offer in the near future.
Both vaccines (quadrivalent and bivalent) are available in Italy
and are acquired through regional tenders. It is remarkable that HPV
vaccination is performed by public health services in all Regions,
which makes Italy an ideal country for the long-term follow-up of
HPV immunisation programmes.
7. Summary and conclusions
The two HPV vaccines showed excellent immunogenicity and
efficacy for the prevention of lesions related to vaccine types in
phase 2 and 3 clinical trials. Virtually all naïve participating women
responded to three doses of vaccine. A decline in antibody titres was
observed in follow-up studies of immunised subjects, especially
in the first months after completion of the vaccination schedule,
reaching a plateau in the longer time.
For the quadrivalent vaccine, titres of anti-HPV remained substantially higher than those obtained after natural infection for
types 16 and 6, while they were similar to those observed in naturally infected subjects at month 60 of follow-up for types 11 and 18.
Only few vaccinees lost their antibodies during the five years after
start of vaccination. However, no breakthrough disease occurred
even in those subjects. The administration of a challenge dose at
the end of follow-up resulted in a strong anamnestic response for
all vaccine types, demonstrating the induction of immunological
memory following the primary series.
For the bivalent vaccine, data up to 6.4 years show persistence of
antibodies to both vaccine types in >98% subjects. Antibody titres
(total and neutralising) remained at levels higher than those seen
after natural infection, with no lesion occurring in vaccinees.
The mechanism by which vaccination induces protection and
the reason for continuing vaccine efficacy to date also in subjects
who lost anti-HPV over time remains to be elucidated. The same is
true for the possibility to induce an anamnestic response following
a viral challenge occurring through a sexual intercourse.
Data strongly suggest that both vaccines can have a variable level
of cross protection against HPV types genetically and antigenicallyclosely related to vaccine types. Demonstration of cross protection
against combined endpoints (CIN2/3 and AIS) for combined HPV
types, and, as a single type, for HPV-31, has been reached so far for
the quadrivalent vaccine, and there is evidence of cross protection
against HPV 31 and 45 persistent infections (as single types) for the
bivalent vaccine.
Assays used for antibody detection were different for the two
vaccines, and standardisation of methods for anti-HPV L1 protein
detection is presently the object of substantial effort from international agencies and specialised laboratories. The possibility to use
universally accepted tests for antibody measurement would make
comparison between vaccines and different studies much easier.
In conclusion, an impressive amount of data on HPV vaccines
has accumulated in the last few years, and results of all researches
performed so far demonstrate that immunisation against HPV has
the potential to substantially impact on HPV-related disease and to
become (together with continuing screening programmes) the key
to success in the battle against one of the most important public
health priorities at the global level. Future research will allow to
highlight the still open questions, thus contributing to design the
most appropriate strategies of intervention.
References
[1] Muñoz N, Bosch FX, de Sanjosé S, Herrero R, Castellsagué X, Shah KV, et al.
International Agency for Research on Cancer Multicenter Cervical Cancer Study
Group. Epidemiologic classification of human papillomavirus types associated
with cervical cancer. N Engl J Med 2003;348(6):518–27.
[2] Chiaradia G, La Torre G, Ricciardi W. Epidemiologia dell’infezione da HPV e delle
malattie correlate. IJPH 2007;4(2):1–13.
[3] Smith JS, Lindsay L, Hoots B, Keys J, Franceschi S, Winer R, et al. Human papillomavirus type distribution in invasive cervical cancer and highgrade cervical
lesions: a meta-analysis update. Int J Cancer 2007;121(3):621–32.
[4] Koutsky L. Epidemiology of genital human papillomavirus infection. Am J Med
1997;102(5A):3–8.
[5] Schiffman M, Kjaer SK. Chapter 2: natural history of anogenital human papillomavirus infection and neoplasia. J Natl Cancer Inst Monogr 2003;31:14–9.
[6] Garnett GP, Kim JJ, French K, Goldie SJ. Chapter 21: modelling the impact of HPV
vaccines on cervical cancer and screening programmes. Vaccine 2006;24(Suppl
3):S178–86.
[7] Frazer I. Correlating immunity with protection for HPV infection. Int J Infect Dis
2007;11(2):S10–6.
[8] Carter JJ, Koutsky LA, Hughes JP, Lee SK, Kuypers J, Kiviat N, et al. Comparison of
human papillomavirus types 16, 18, and 6 capsid antibody responses following
incident infection. J Infect Dis 2000;181(6):1911–9.
[9] Kreider JW, Bartlett GL. The Shope papilloma-carcinoma complex of rabbits:
a model system of neoplastic progression and spontaneous regression. Adv
Cancer Res 1981;35:81–110.
[10] Suzich JA, Ghim SJ, Palmer-Hill FJ, White WI, Tamura JK, Bell JA, et al. Systemic
immunization with papillomavirus L1 protein completely prevents the development of viral mucosal papillomas. Proc Natl Acad Sci USA 1995;92:11553–7.
[11] Coleman N, Birley HDL, Renton AM, Hanna NF, Ryait BK, Byrne M, et al. Immunological events in regressing genital warts. Am J Clin Pathol 1994;102:768–74.
P. Bonanni et al. / Vaccine 27 (2009) A46–A53
[12] Stanley M. Immune responses to human papillomavirus. Vaccine
2006;24(Suppl 1):S16–22.
[13] Leigh IM, Glover MT. Skin cancer and warts in immunosuppressed renal transplant recipients. Recent Results Cancer Res 1995;139:69–86.
[14] Ozsaran AA, Ateş T, Dikmen Y, Zeytinoglu A, Terek C, Erhan Y, et al. Evaluation of
the risk of cervical intraepithelial neoplasia and human papilloma virus infection in renal transplant patients receiving immunosuppressive therapy. Eur J
Gynaecol Oncol 1999;20(2):127–30.
[15] Koshiol JE, Schroeder JC, Jamieson DJ, Marshall SW, Duerr A, Heilig CM, et
al. Time to clearance of human papillomavirus infection by type and human
immunodeficiency virus serostatus. Int J Cancer 2006;119(7):1623–9.
[16] Scott M, Nakagawa M, Moscicki AB. Cell-mediated immune response to human
papillomavirus infection. Clin Diagn Lab Immunol 2001;8:209–20.
[17] Garland SM, Hernandez-Avila M, Wheeler CM, Perez G, Harper DM, Leodolter
S, et al. Females United to Unilaterally Reduce Endo/Ectocervical Disease
(FUTURE) I Investigators. Quadrivalent vaccine against human papillomavirus
to prevent anogenital diseases. N Engl J Med 2007;356(19):1928–43.
[18] Barr E, Tamms G. Quadrivalent human papillomavirus vaccine. Clin Infect Dis
2007;45(5):609–17.
[19] FUTURE II Study Group. Quadrivalent vaccine against human papillomavirus to
prevent high-grade cervical lesions. N Engl J Med 2007;356(19):1915–27.
[20] Harper DM, Franco EL, Wheeler C, Ferris DG, Jenkins D, Schuind A, et al. Efficacy
of a bivalent L1 virus-like particle vaccine in prevention of infection with human
papillomavirus types 16 and 18 in young women: a randomised controlled trial.
Lancet 2004;364(9447):1757–65.
[21] Paavonen J, Jenkins D, Bosch FX, Naud P, Salmerón J, Wheeler CM, et al. Efficacy of a prophylactic adjuvanted bivalent L1 virus-like-particle vaccine against
infection with human papillomavirus types 16 and 18 in young women: an
interim analysis of a phase III double-blind, randomised controlled trial. Lancet
2007;369(9580):2161–70 [Erratum in: Lancet 2007;370(9596):1414]s.
[22] Villa LL, Costa RL, Petta CA, Andrade RP, Ault KA, Giuliano AR, et al. Prophylactic
quadrivalent human papillomavirus (types 6, 11, 16, and 18) L1 virus-like particle vaccine in young women: a randomised double-blind placebo-controlled
multicentre phase II efficacy trial. Lancet Oncol 2005;6(5):271–8.
[23] Olsson SE, Villa LL, Costa RL, Petta CA, Andrade RP, Malm C, et al. Induction
of immune memory following administration of a prophylactic quadrivalent
human papillomavirus (HPV) types 6/11/16/18 L1 virus-like particle (VLP) vaccine. Vaccine 2007;25(26):4931–9.
[24] Wheeler CM, Teixeira J, Romanowski B, De Carvalho N, Dubin G, Schuind A.
High and sustained HPV-16 and 18 antibody levels through 6.4 years in women
vaccinated with CervarixTM (GSK HPV-16/18 AS04 vaccine). Poster presentation
698, 26th Annual Meeting of the European Society for Paediatric Infectious
Diseases (ESPID), 13–17 May, 2008, Graz, Austria.
A53
[25] Fraser C, Tomassini JE, Xi L, Golm G, Watson M, Giuliano AR, et al. Modeling the long-term antibody response of a human papillomavirus (HPV)
virus-like particle (VLP) type 16 prophylactic vaccine. Vaccine 2007;25:4324–
33.
[26] Leclerc C, Sedlik C, Lo-Man R, Charlot B, Rojas M, Dériaud E. Stimulation of a
memory B cell response does not require primed helper T cells. Eur J Immunol
1995;25(9):2533–8.
[27] Zanetti AR, Mariano A, Romanò L, D’Amelio R, Chironna M, Coppola RC, et al.
Long-term immunogenicity of hepatitis B vaccination and policy for booster:
an Italian multicentre study. Lancet 2005;366(9494):1379–84.
[28] Ault KA. Long-term efficacy of human papillomavirus vaccination. Gynecol
Oncol 2007;107(2, Suppl 1):S27–30.
[29] WHO. Expert Committee On Biological Standardization. Guidelines to assure
the quality, safety and efficacy of recombinant human papillomavirus virus-like
particle vaccines. Geneva, 23–27 October 2006. WHO/BS/06.2050.
[30] Poncelet S, Cambron P, Giannini SL, Colau B, Dessy F, Zahaf T, et al. Induction
of cervical mucosal HPV IgG in women 15–55 years old following systemic
vaccination with GSK cervical cancer candidate vaccine. Abstracts of the 25th
Annual Meeting of the European Society for Paediatric Infectious Diseases.
Porto, Portugal, May 2–4, 2007.
[31] Lehtinen M, Apter D, Dubin G, Kosunen E, Isaksson R, Korpivaara EL, et al. Enrolment of 22,000 adolescent women to cancer registry follow-up for long-term
human papillomavirus vaccine efficacy: guarding against guessing. Int J STD
AIDS 2006;17(8):517–21.
[32] Wieland U, Pfister H. Papillomaviruses in human pathology: epidemiology,
pathogenesis and oncogenic role. In: Gross, Barrasso, editors. Human papilloma
virus infection: a clinical atlas. Ullstein: Mosby; 1997. p. 1–18.
[33] Smith JF, Brownlow M, Brown M, Kowalski R, Esser MT, Ruiz W, et al. Antibodies
from women immunized with Gardasil cross-neutralize HPV 45 pseudovirions.
Hum Vac 2007;3(4):109–15.
[34] Harper DM, Franco EL, Wheeler CM, Moscicki AB, Romanowski B, Roteli-Martins
CM, et al. Sustained efficacy up to 4.5 years of a bivalent L1 virus-like particle vaccine against human papillomavirus types 16 and 18: follow-up from a
randomised control trial. Lancet 2006;367(9518):1247–55.
[35] Jenkins D. A review of cross-protection against oncogenic HPV by an HPV-16/18
AS04-adjuvanted cervical cancer vaccine: Importance of virological and clinical
endpoints and implications for mass vaccination in cervical cancer prevention.
Ginecology Oncology 2008;110:S18–25.
[36] Brown D. HPV type 6/11/16/18 vaccine: first analysis of cross-protection against
persistent infection, cervical intraepithelial neoplasia (CIN), and adenocarcinoma in situ (AIS) caused by oncogenic HPV types in addition to 16/18. Poster
presentation G-1720b. In: Interscience Conference on Antimicrobial Agents and
Chemotherapy (ICAAC). 2007.