Content uploaded by Diane M Harper
Author content
All content in this area was uploaded by Diane M Harper on Jul 14, 2018
Content may be subject to copyright.
100
www.thelancet.com/oncology Vol 13 January 2012
Articles
Lancet Oncol 2012; 13: 100–10
Published Online
November 9, 2011
DOI:10.1016/S1470-
2045(11)70287-X
This publication
has been corrected.
The corrected version fi rst
appeared at thelancet.com/
oncology on January 3, 2011
See Comment page 10
See Articles page 89
*For the HPV PATRICIA Study
Group see webappendix p 9
Departments of Pathology and
Obstetrics and Gynecology,
University of New Mexico
Health Sciences Center,
Albuquerque, NM, USA
(Prof C M Wheeler PhD);
Biomedical Research Centre
Network for Epidemiology and
Public Health (CIBER-ESP;
X Castellsagué PhD), Network
on Cooperative Cancer
Research (RTICC; F X Bosch PhD),
and Unit of Infections and
Cancer, Cancer Epidemiology
Research Program, Institut
Català d’Oncologia,
L’Hospitalet de Llobregat,
IDIBELL, Catalonia, Spain
(X Castellsagué, F X Bosch);
Department of Microbiology
and Infectious Diseases, Royal
Women’s Hospital,
Department of Microbiology,
Royal Children’s Hospital,
Murdoch Children’s Research
Institute, and Department of
Obstetrics and Gynaecology,
University of Melbourne,
Parkville, VIC, Australia
(Prof S M Garland FRCPA);
Centre for Cancer Prevention,
Wolfson Institute of Preventive
Medicine, Queen Mary
University of London, London,
UK (A Szarewski PhD);
Department of Obstetrics and
Gynaecology, University of
Helsinki, Helsinki, Finland
Cross-protective effi cacy of HPV-16/18 AS04-adjuvanted
vaccine against cervical infection and precancer caused by
non-vaccine oncogenic HPV types: 4-year end-of-study
analysis of the randomised, double-blind PATRICIA trial
Cosette M Wheeler, Xavier Castellsagué, Suzanne M Garland, Anne Szarewski, Jorma Paavonen, Paulo Naud, Jorge Salmerón, Song-Nan Chow,
Dan Apter, Henry Kitchener, Júlio C Teixeira, S Rachel Skinner, Unnop Jaisamrarn, Genara Limson, Barbara Romanowski, Fred Y Aoki,
Tino F Schwarz, Willy A J Poppe, F Xavier Bosch, Diane M Harper, Warner Huh, Karin Hardt, Toufi k Zahaf, Dominique Descamps, Frank Struyf,
Gary Dubin, Matti Lehtinen, for the HPV PATRICIA Study Group*
Summary
Background We evaluated the effi cacy of the human papillomavirus HPV-16/18 AS04-adjuvanted vaccine against non-
vaccine oncogenic HPV types in the end-of-study analysis after 4 years of follow-up in PATRICIA (PApilloma TRIal
against Cancer In young Adults).
Methods Healthy women aged 15–25 years with no more than six lifetime sexual partners were included in PATRICIA
irrespective of their baseline HPV DNA status, HPV-16 or HPV-18 serostatus, or cytology. Women were randomly
assigned (1:1) to HPV-16/18 vaccine or a control hepatitis A vaccine, via an internet-based central randomisation
system using a minimisation algorithm to account for age ranges and study sites. The study was double-blind. The
primary endpoint of PATRICIA has been reported previously; the present analysis evaluates cross-protective vaccine
effi cacy against non-vaccine oncogenic HPV types in the end-of-study analysis. Analyses were done for three cohorts:
the according-to-protocol cohort for effi cacy (ATP-E; vaccine n=8067, control n=8047), total vaccinated HPV-naive
cohort (TVC-naive; no evidence of infection with 14 oncogenic HPV types at baseline, approximating young
adolescents before sexual debut; vaccine n=5824, control n=5820), and the total vaccinated cohort (TVC; all women
who received at least one vaccine dose, approximating catch-up populations that include sexually active women;
vaccine n=9319, control=9325). Vaccine effi cacy was evaluated against 6-month persistent infection, cervical
intraepithelial neoplasia grade 2 or greater (CIN2+) associated with 12 non-vaccine HPV types (individually or as
composite endpoints), and CIN3+ associated with the composite of 12 non-vaccine HPV types. This study is registered
with ClinicalTrials.gov, number NCT00122681.
Findings Consistent vaccine effi cacy against persistent infection and CIN2+ (with or without HPV-16/18 co-infection)
was seen across cohorts for HPV-33, HPV-31, HPV-45, and HPV-51. In the most conservative analysis of vaccine effi cacy
against CIN2+, where all cases co-infected with HPV-16/18 were removed, vaccine effi cacy was noted for HPV-33 in all
cohorts, and for HPV-31 in the ATP-E and TVC-naive. Vaccine effi cacy against CIN2+ associated with the composite of
12 non-vaccine HPV types (31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68), with or without HPV-16/18 co-infection, was
46·8% (95% CI 30·7–59·4) in the ATP-E, 56·2% (37·2–69·9) in the TVC-naive, and 34·2% (20·4–45·8) in the TVC.
Corresponding values for CIN3+ were 73·8% (48·3–87·9), 91·4% (65·0–99·0), and 47·5% (22·8–64·8).
Interpretation Data from the end-of-study analysis of PATRICIA show cross-protective effi cacy of the HPV-16/18
vaccine against four oncogenic non-vaccine HPV types—HPV-33, HPV-31, HPV-45, and HPV-51—in diff erent trial
cohorts representing diverse groups of women.
Funding GlaxoSmithKline Biologicals.
Introduction
Infection with oncogenic human papillomavirus (HPV)
types is a necessary cause of invasive cervical cancer
(ICC).1,2 Roughly 15 HPV types have been classifi ed as
oncogenic. Among these, HPV-16 and HPV-18 are the
most prevalent, and cause around 70% of ICC
worldwide.3 The next most prevalent oncogenic type is
HPV-45.3 HPV-16 (A9 species) together with HPV-18
and HPV-45 (A7 species) cause 75% of squamous-cell
carcinoma (SCC) and 94% of adenocarcinoma.3 The
next fi ve most common oncogenic HPV types are all
from the A9 species (HPV-31, HPV-33, HPV-35,
HPV-52, and HPV-58) and together cause another 15%
of ICC.3 The remaining oncogenic HPV types
individually cause a very small proportion of ICC
worldwide (<2%) and include HPV-51 (A5 species),
HPV-56 (A6 species), and HPV-39 and HPV-59
(A7 species).3–5 The possible carcinogenicity of HPV-66
(A6 species) is uncertain, whereas HPV-68 (A7 species)
is probably oncogenic.4
Articles
www.thelancet.com/oncology Vol 13 January 2012
101
(Prof J Paavonen MD);
Department of Gynecology and
Obstetrics, Federal University
of Rio Grande do Sul, UFRGS/
HCPA, Hospital de Clínicas de
Porto Alegre, Brazil
(Prof P Naud PhD); Unidad de
Investigación Epidemiológica y
en Servicios de Salud, Instituto
Mexicano del Seguro Social,
Morelos, Mexico
(Prof J Salmerón PhD);
Department of Obstetrics and
Gynecology, College of
Medicine and the Hospital,
National Taiwan University,
Taipei, Taiwan
(Prof S-N Chow MD); Family
Federation of Finland, Sexual
Health Clinic, Helsinki, Finland
(D Apter MD); Manchester
Academic Health Science
Centre, Central Manchester
University Hospitals NHS
Foundation Trust, St Mary’s
Hospital, Manchester, UK
(Prof H Kitchener MD);
Departamento de
Tocoginecologia da Unicamp,
University of Campinas,
Campinas, Brazil
(J C Teixeira MD); Vaccines Trials
Group, Telethon Institute for
Child Health Research, Perth,
WA, and Sydney University
Discipline of Paediatrics and
Child Health, Children’s
Hospital Westmead, Sydney,
NSW, Australia
(S R Skinner PhD); Department
of Obstetrics and Gynaecology,
Faculty of Medicine,
Chulalongkorn University,
Bangkok, Thailand
(Prof U Jaisamrarn MD); College
of Medicine, University of the
Philippines, Philippine General
Hospital, Makati Medical
Centre, Makati City, Philippines
(Prof G Limson MD); Division of
Infectious Diseases,
Department of Medicine,
Faculty of Medicine and
Dentistry, University of Alberta,
Edmonton, AB, Canada
(Prof B Romanowski MD);
Department of Medical
Microbiology, University of
Manitoba, Winnipeg, MB,
Canada (Prof F Y Aoki MD);
Central Laboratory and
Vaccination Centre, Stiftung
Juliusspital, Academic Teaching
Hospital of the University of
Wuerzburg, Wuerzburg,
Germany (Prof T F Schwarz MD);
Department of Gynaecology,
University Hospital KU Leuven
Gasthuisberg, Leuven, Belgium
(Prof W A J Poppe PhD);
Dartmouth Medical School,
Prophylactic HPV vaccines are administered in vaccin-
ation programmes targeted at young adolescent girls
before sexual exposure, and in catch-up programmes for
young women in some countries. Since non-vaccine
HPV types account for around 30% of cervical cancers,
cross-protection against these types would potentially
enhance primary cervical cancer prevention eff orts.
The HPV-16/18 AS04-adjuvanted vaccine (Cervarix,
GlaxoSmithKline Biologicals) and HPV-6/11/16/18 vaccine
(Gardasil, Merck) consist of virus-like particles (VLPs)
composed of relatively well conserved L1 capsid proteins.
The neutralising antigenic sites (epitopes) defi ned so far
are mainly situated on one of fi ve variable loops of the
L1 capsomer. These are exposed on virion surfaces and
should be readily accessible to neutralising antibodies.6–8 In
theory, aminoacid sequence or con formational diff erences
determine the type-specifi city of any HPV-neutralising
epitope. Some oncogenic HPV types that are phylogenetically
related to vaccine types presumably share epitopes that can
elicit cross-reactive immune responses, although cross-
neutralising anti bodies might be induced by HPV
vaccination at much lower levels than type-specifi c
antibodies.9
This report summarises cross-protection data with the
HPV-16/18 vaccine in the end-of-study analysis of the
PApilloma TRIal against Cancer In young Adults
(PATRICIA). In general, cervical intraepithelial neoplasia
grade 2 or greater (CIN2+) is the accepted clinical endpoint
to evaluate HPV vaccine effi cacy. However, analyses can
be biased if a lesion is co-infected with both a vaccine and
a non-vaccine oncogenic HPV type, since defi nitive
causality to a single HPV type cannot be readily
assigned.10–12 This confounding bias particularly applies to
analyses of cross-protection, because HPV-16 and HPV-18
are common co-infections and are more prevalent than
other HPV types in cervical lesions. Additionally, as a
Vaccine Control Effi cacy (95% CI)
N Cases Rate N Cases Rate
6-month persistent infection
Non-vaccine A9 species (composite HPV-31/33/35/52/58) 7671 608 2·49 7656 770 3·19 22·0% (13·2 to 30·0)
HPV-31 7400 58 0·24 7414 247 1·01 76·8% (69·0 to 82·9)
HPV-33 7534 65 0·26 7513 117 0·47 44·8% (24·6 to 59·9)
HPV-35 7579 67 0·27 7569 56 0·22 –19·8% (–74·1 to 17·2)
HPV-52 7289 346 1·46 7237 374 1·59 8·3% (–6·5 to 21·0)
HPV-58 7518 144 0·58 7511 122 0·49 –18·3% (–51·8 to 7·7)
Non-vaccine A7 species (composite HPV-39/45/59/68) 7672 419 1·69 7656 472 1·91 11·6% (–1·0 to 22·7)
HPV-39 7429 175 0·71 7428 184 0·75 4·8% (–17·7 to 23·1)
HPV-45 7594 24 0·09 7556 90 0·36 73·6% (58·1 to 83·9)
HPV-59 7536 73 0·29 7530 68 0·27 –7·5% (–51·8 to 23·8)
HPV-68 7450 165 0·67 7424 169 0·69 2·6% (–21·5 to 21·9)
Other
HPV-51 7190 349 1·50 7165 416 1·79 16·6% (3·6 to 27·9)
HPV-56 7467 226 0·92 7451 215 0·88 –5·3% (–27·5 to 13·1)
HPV-66 7412 211 0·87 7375 215 0·89 2·3% (–18·7 to 19·6)
CIN2+ with or without co-infection with HPV-16/18
Non-vaccine A9 species (composite HPV-31/33/35/52/58) 7854 55 0·21 7846 108 0·42 49·1% (29·0 to 63·9)
HPV-31 7575 5 0·02 7592 40 0·16 87·5% (68·3 to 96·1)
HPV-33 7712 13 0·05 7700 41 0·16 68·3% (39·7 to 84·4)
HPV-35 7760 3 0·01 7757 8 0·03 62·5% (–56·5 to 93·6)
HPV-52 7455 24 0·10 7409 33 0·14 27·6% (–26·3 to 59·1)
HPV-58 7701 15 0·06 7696 21 0·08 28·5% (–45·5 to 65·7)
Non-vaccine A7 species (composite HPV-39/45/59/68) 7855 18 0·07 7846 43 0·17 58·2% (25·9 to 77·3)
HPV-39 7602 4 0·02 7608 16 0·06 74·9% (22·3 to 93·9)
HPV-45 7774 2 0·01 7738 11 0·04 81·9% (17·0 to 98·1)
HPV-59 7713 1 0·00 7716 5 0·02 80·0% (–79·1 to 99·6)
HPV-68 7626 11 0·04 7606 15 0·06 26·8% (–70·7 to 69·6)
Other
HPV-51 7356 21 0·09 7341 46 0·19 54·4% (22·0 to 74·2)
HPV-56 7638 7 0·03 7631 13 0·05 46·1% (–45·2 to 81·8)
HPV-66 7583 7 0·03 7559 16 0·06 56·4% (–12·1 to 84·8)
(Continues on next page)
Articles
102
www.thelancet.com/oncology Vol 13 January 2012
Hanover, NH, USA
(D M Harper MD); Division of
Gynecologic Oncology,
University of Alabama at
Birmingham, Birmingham, AL,
USA (W Huh MD);
GlaxoSmithKline Biologicals,
Wavre, Belgium (K Hardt PhD,
T Zahaf PhD, D Descamps MD,
F Struyf MD); GlaxoSmithKline
Biologicals, King of Prussia, PA,
USA (G Dubin MD); and
University of Tampere, School
of Public Health, Tampere,
Finland (Prof M Lehtinen PhD)
Correspondence to:
Dr Cosette M Wheeler,
Departments of Pathology and
Obstetrics and Gynecology,
University of New Mexico Health
Sciences Center, Albuquerque,
NM, USA
cwheeler@salud.unm.edu
See Online for webappendix
result of vaccination, HPV-16 and HPV-18 infections are
diff erentially removed from the vaccine and control
groups. To overcome these biases, we did analyses of
CIN2+ and the more stringent endpoint, CIN3+, which
either include or exclude cases co-infected with a vaccine
type. We also did complementary analyses using virological
endpoints. Persistent HPV infection usually precedes
cervical cancer and its precursor lesions (CIN grade 2
and 3),13–16 and therefore provides a relevant marker for the
risk of developing these precancerous lesions.
To estimate the extent of cross-protection, we did the
analyses in various cohorts: the according-to-protocol
cohort for effi cacy (ATP-E), the total vaccinated HPV-naive
cohort (TVC-naive), and the total vaccinated cohort (TVC).
The ATP-E cohort represents a population of women who
at baseline had no evidence of infection with the HPV type
under analysis and who received all three vaccine doses.
In terms of exposure to and acquisition of HPV types, the
TVC-naive approximates the current primary target of
HPV vaccination programmes. The TVC includes women
with evidence of current or previous infection with
oncogenic HPV types, and approximates a population
currently targeted by catch-up HPV vaccination
programmes. Data regarding other measures of effi cacy
in the TVC-naive and TVC are reported in an accompanying
article by Lehtinen and colleagues.17
Methods
The trial methods have been previously described
in detail, and the results of event-driven analyses
Vaccine Control Effi cacy (95% CI)
N Cases Rate N Cases Rate
(Continued from previous page)
CIN2+ excluding co-infection with HPV-16/18
Non-vaccine A9 species (composite HPV-31/33/35/52/58) 7854 55 0·21 7846 78 0·30 29·5% (–0·9 to 51·0)
HPV-31 7575 5 0·02 7592 32 0·13 84·3% (59·5 to 95·2)
HPV-33 7712 13 0·05 7700 32 0·13 59·4% (20·5 to 80·4)
HPV-35 7760 3 0·01 7757 5 0·02 39·9% (–208·9 to 90·7)
HPV-52 7455 24 0·10 7409 19 0·08 –25·8% (–142·9 to 34·0)
HPV-58 7701 15 0·06 7696 14 0·06 –7·3% (–139·9 to 51·7)
Non-vaccine A7 species (composite HPV-39/45/59/68) 7855 16 0·06 7846 23 0·09 30·4% (–37·5 to 65·7)
HPV-39 7602 4 0·02 7608 7 0·03 42·7% (–125·4 to 87·7)
HPV-45 7774 2 0·01 7738 4 0·02 50·1% (–247·9 to 95·5)
HPV-59 7713 1 0·00 7716 3 0·01 66·6% (–316·1 to 99·4)
HPV-68 7626 9 0·04 7606 10 0·04 10·1% (–146·2 to 67·7)
Other
HPV-51 7356 19 0·08 7341 22 0·09 13·7% (–67·1 to 55·8)
HPV-56 7638 6 0·02 7631 10 0·04 40·0% (–82·3 to 82·1)
HPV-66 7583 7 0·03 7559 11 0·04 36·5% (–79·3 to 79·1)
Women could be infected with multiple HPV types (therefore the number of cases for the composite endpoints might not equal the sum of the cases for each individual
type included in the composite). Types tested for were HPV-16, HPV-18, HPV-31, HPV-33, HPV-35, HPV-39, HPV-45, HPV-51, HPV-52, HPV-56, HPV-58, HPV-59, HPV-66,
and HPV-68. Women included in the analysis were HPV DNA negative for the HPV type under consideration at month 0 and month 6, irrespective of initial serostatus,
and had negative or low-grade cytology at month 0. For the composite endpoints, women had to be infected with, or have a lesion associated with, at least one of the
HPV types in the composite, and had to be HPV DNA negative for the corresponding HPV type at month 0 and month 6. CIN2+ was defi ned histologically as CIN2, CIN3,
adenocarcinoma in situ, or invasive carcinoma. CIN=cervical intraepithelial neoplasia. HPV=human papillomavirus. ATP-E=according-to-protocol for effi cacy. N=number
of evaluable women in each group. Cases=number of evaluable women reporting at least one event. Rate=number of cases divided by sum of follow-up period
(per 100 woman years); follow-up period began the day after the third vaccine dose.
Table 1: Cross-protective effi cacy against 6-month persistent infection and CIN2+ associated with non-vaccine HPV types, in women who were DNA
negative for the corresponding HPV type at baseline (ATP-E cohort)
–10
0
10
20
30
40
50
60
70
80
90
100
With or without HPV-16/18 co-infection Excluding HPV-16/18 co-infection
46·8%
62·1%
24·1%
73·8%
Vaccine efficacy (%)
Number of cases vaccine:
Number of cases control:
88
165 42
11 85 11
29
112
CIN2+
CIN3+
Figure 1: Vaccine effi cacy against CIN2+ and CIN3+ associated with a composite of 12 non-vaccine HPV types,
with or without HPV-16/18 co-infection and excluding HPV-16/18 co-infection, in the ATP-E cohort
Women had to have a lesion associated with at least one of the HPV types included in the composite. In the
analysis of the ATP-E, women were HPV DNA negative for the corresponding HPV type at month 0 and month 6,
irrespective of initial serostatus, and had negative or low-grade cytology at baseline. Types tested for were HPV-16,
HPV-18, HPV-31, HPV-33, HPV-35, HPV-39, HPV-45, HPV-51, HPV-52, HPV-56, HPV-58, HPV-59, HPV-66, and
HPV-68. Follow-up period started the day after the third vaccine dose. CIN2+ was defi ned histologically as CIN2,
CIN3, adenocarcinoma in situ, or invasive carcinoma; CIN3+ did not include CIN2. Vaccine effi cacy point estimates
are shown above each bar, and error bars represent 95% CI. CIN=cervical intraepithelial neoplasia. HPV=human
papillomavirus. ATP-E=according to protocol for effi cacy.
Articles
www.thelancet.com/oncology Vol 13 January 2012
103
presented.10,11 In this end-of-study analysis after 48 months
of follow-up, we report vaccine effi cacy against types
other than HPV-16 and HPV-18, using persistent
infection, CIN2+, and CIN3+ as endpoints.
Participants
Healthy women aged 15–25 years with no more than six
lifetime sexual partners (this exclusion criterion was not
applied in Finland, in accordance with local regulatory
and ethical requirements18) were included in the trial; full
inclusion and exclusion criteria, trial locations, and dates
have been described previously.10,11 Women were included
regardless of their baseline HPV DNA status, HPV-16 or
HPV-18 serostatus, or cytology. Written informed consent
was obtained from all adult participants. For minors,
written informed assent was obtained from the participant
and written informed consent from their parents. The
protocol and other materials were approved by
independent ethics committees or institutional review
boards at each location.
Procedures
Women were randomised in a 1:1 ratio to receive
either the HPV-16/18 AS04-adjuvanted vaccine (Cervarix,
GlaxoSmithKline Biologicals) or control hepatitis A vaccine
(GlaxoSmithKline Biologicals) at 0, 1, and 6 months in a
double-blind manner. The study protocol prescribed that
both groups were to be unblinded after the month 48 visit
and off ered the cross-over vaccine. Cervical samples were
obtained every 6 months for HPV DNA detection and
typing. Gynaecological examinations were performed and
cy tology samples were obtained every 12 months. Collection
of cytology and histopathology specimens and the clinical
management algorithm for abnormal cytology results and
colposcopy referral have been described elsewhere.10,11,17
A broad spectrum PCR SPF10-LiPA25 (version 1 based on
licensed Innogenetics SPF10 technology; Labo Biomedical
Products, Rijswijk, Netherlands) and type-specifi c PCR
for HPV-16 and HPV-18 DNA were used to test cervical
samples and biopsy material for HPV DNA from
14 oncogenic HPV types (16, 18, 31, 33, 35, 39, 45, 51, 52,
56, 58, 59, 66, and 68).10,19 These types were selected based
on the International Agency for Research on Cancer
(IARC) list of oncogenic HPV types that was current at the
time of writing the study protocol, although the evidence
in humans for the carcinogenicity of HPV-66 is now
considered to be limited.20 All CIN cases were reviewed by
an independent endpoint committee.
In the present analysis, we describe secondary and
exploratory endpoints of vaccine effi cacy against per-
sistent infection, CIN2+, and CIN3+ associated with
non-vaccine HPV types (31, 33, 35, 39, 45, 51, 52, 56,
58, 59, 66, 68) tested by PCR. For persistent infection
and CIN2+, we evaluated vaccine effi cacy against
individual non-vaccine types, and against composite
endpoints of four HPV types from the A7 species
(HPV-39/45/59/68), fi ve HPV types from the A9 species
(HPV-31/33/35/52/58), and 12 non-vaccine HPV types
(all HPV types tested by PCR except HPV-16/18—ie,
HPV-31/33/35/39/45/51/52/56/58/59/66/68). For CIN3+,
we evaluated vaccine effi cacy against the 12-type
composite. For the composite endpoints, women had
to be infected with, or have a lesion associated with, at
least one of the HPV types included in the composite.
6-month persistent infection was defi ned as detection
of the same HPV type in consecutive samples over a
minimum of 5 months. 12-month persistent infection
was defi ned as detection of the same HPV type in
consecutive samples over a minimum of 10 months.
CIN2+ was defi ned as CIN2, CIN3, adenocarcinoma in
situ (AIS), or invasive carcinoma; CIN3+ excluded CIN2.
Since multiple HPV types are often found in cervical
Vaccine Control Effi cacy (95% CI)
Cases Rate Cases Rate
6-month persistent infection (N=5427 vaccine vs 5399 control)
Non-vaccine A9 species
(composite HPV-31/33/35/52/58)
407 2·05 550 2·84 27·6% (17·6 to 36·5)
HPV-31 38 0·18 163 0·81 77·1% (67·2 to 84·4)
HPV-33 53 0·26 92 0·45 43·1% (19·3 to 60·2)
HPV-35 38 0·18 31 0·15 –21·8% (–102·5 to 26·2)
HPV-52 231 1·14 281 1·41 18·9% (3·2 to 32·2)
HPV-58 93 0·45 87 0·43 –6·2% (–44·0 to 21·6)
Non-vaccine A7 species
(composite HPV-39/45/59/68)
263 1·31 334 1·68 22·3% (8·4 to 34·2)
HPV-39 111 0·54 139 0·69 20·9% (–2·3 to 38·9)
HPV-45 13 0·06 61 0·30 79·0% (61·3 to 89·4)
HPV-59 45 0·22 43 0·21 –3·9% (–61·7 to 33·1)
HPV-68 112 0·55 122 0·60 8·9% (–18·8 to 30·1)
Other
HPV-51 253 1·26 334 1·68 25·5% (12·0 to 37·0)
HPV-56 147 0·72 148 0·73 1·4% (–24·8 to 22·0)
HPV-66 141 0·69 138 0·68 –1·5% (–29·3 to 20·3)
CIN2+ with or without co-infection with HPV-16/18 (N=5466 vaccine vs 5452 control)
Non-vaccine A9 species
(composite HPV-31/33/35/52/58)
31 0·15 71 0·35 56·6% (33·0 to 72·5)
HPV-31 3 0·01 28 0·14 89·4% (65·5 to 97·9)
HPV-33 5 0·02 28 0·14 82·3% (53·4 to 94·7)
HPV-35 1 0·00 6 0·03 83·4% (–36·6 to 99·6)
HPV-52 14 0·07 20 0·10 30·4% (–45·0 to 67·5)
HPV-58 9 0·04 14 0·07 36·1% (–58·6 to 75·6)
Non-vaccine A7 species
(composite HPV-39/45/59/68)
8 0·04 28 0·14 71·6% (36·0 to 88·8)
HPV-39 3 0·01 11 0·05 72·9% (–2·7 to 95·1)
HPV-45 0 0·00 8 0·04 100% (41·7 to 100)
HPV-59 0 0·00 2 0·01 100% (–429·6 to 100)
HPV-68 5 0·02 11 0·05 54·8% (–41·2 to 87·7)
Other
HPV-51 9 0·04 30 0·15 70·2% (35·6 to 87·6)
HPV-56 0 0·00 7 0·03 100% (31·0 to 100)
HPV-66 3 0·01 11 0·05 72·9% (–2·7 to 95·1)
(Continues on next page)
Articles
104
www.thelancet.com/oncology Vol 13 January 2012
lesion samples and it is not always possible to assign
causality to a particular type, we considered two analyses:
a prespecifi ed analysis in which all cases were included,
irrespective of whether they were co-infected with
HPV-16/18 (referred to hereafter as CIN2+ or CIN3+);
and an additional post-hoc analysis in which all CIN2+ or
CIN3+ cases that were co-infected with HPV-16/18 were
excluded (referred to hereafter as CIN2+ or CIN3+
excluding co-infection with HPV-16/18).
Statistical analysis
Endpoints were evaluated in three cohorts: the ATP-E, the
TVC-naive, and the TVC cohorts.17 The ATP-E cohort
included women who were evaluable for effi cacy (ie, had a
baseline PCR or cytology sample and one further sample
available), met all eligibility criteria, complied with the
protocol, received all three vaccine doses, and had negative
or low-grade cytology at baseline. In the ATP-E cohort,
endpoints were assessed in women who were HPV DNA
negative at baseline and at month 6 for the HPV type
analysed. The TVC-naive included women who had
received at least one vaccine dose, were evaluable for
effi cacy, were HPV DNA negative at baseline for all 14 HPV
types tested for, seronegative for HPV-16 and HPV-18, and
had negative cytology. Excluding women who were positive
for any of the 14 HPV types at baseline is the main reason
for the substantially lower number of women included in
the TVC-naive than in the ATP-E cohort. The TVC-naive
represents the least HPV-exposed analytical group. The
TVC included all women who received at least one vaccine
dose and were evaluable for effi cacy. Endpoints were
assessed in the TVC irrespective of women’s baseline
HPV DNA, cytological status, and serostatus. Licensure of
the vaccine was based on analysis of the ATP-E cohort to
fully describe the vaccine’s profi le. However, the TVC and
TVC-naive are more relevant from a public health
perspective, and we have therefore included all three
cohorts in the end-of-study analysis of cross-protection.
The end-of-study analysis was intended to support the
effi cacy results of the fi nal event-driven analysis.11 Vaccine
effi cacy and 95% CIs were calculated using a conditional
exact method (webappendix p 8). 95% CIs were calculated
for the end-of-study analysis, whereas 97·9% and
96·1% CIs were used for the interim and fi nal event-
driven analyses, respectively.10,11 Results were considered
to support statistically signifi cant vaccine effi cacy
observed in the fi nal event-driven analysis if end-of-study
estimates and their 95% CIs were above zero.
Event rates were calculated as the number of cases divided
by the total follow-up in years and were expressed per
100 woman years. In the TVC and the TVC-naive, follow-up
started the day after the fi rst vaccine dose. In the ATP-E,
follow-up started the day after the third vaccine dose.
Follow-up for each outcome ended at the time the outcome
occurred or at the last available sample (up to month 48).
Statistical analyses were done with SAS version 9.1 and
Proc StatXact-7 on Windows XP.
Vaccine Control Effi cacy (95% CI)
Cases Rate Cases Rate
(Continued from previous page)
CIN2+ excluding co-infection with HPV-16/18 (N=5466 vaccine vs 5452 control)
Non-vaccine A9 species
(composite HPV-31/33/35/52/58)
31 0·15 43 0·21 28·3% (–16·5 to 56·3)
HPV-31 3 0·01 18 0·09 83·4% (43·3 to 96·9)
HPV-33 5 0·02 21 0·10 76·3% (35·5 to 93·0)
HPV-35 1 0·00 3 0·01 66·8% (–313·0 to 99·4)
HPV-52 14 0·07 6 0·03 –132·3% (–637·5 to 16·2)
HPV-58 9 0·04 8 0·04 –11·9% (–233·4 to 61·7)
Non-vaccine A7 species
(composite HPV-39/45/59/68)
8 0·04 10 0·05 20·4% (–124·0 to 72·7)
HPV-39 3 0·01 5 0·02 40·3% (–206·8 to 90·7)
HPV-45 0 0·00 2 0·01 100% (–429·7 to 100)
HPV-59 0 0·00 1 0·00 100% (–3779·6 to 100)
HPV-68 5 0·02 3 0·01 –65·8% (–967·9 to 67·7)
Other
HPV-51 9 0·04 9 0·04 0·5% (–182·9 to 65·0)
HPV-56 0 0·00 2 0·01 100% (–429·7 to 100)
HPV-66 3 0·01 7 0·03 57·4% (–86·8 to 92·9)
Women could be infected with multiple HPV types (therefore the number of cases for the composite endpoints might
not equal the sum of the cases for each individual type included in the composite). Types tested for were HPV-16,
HPV-18, HPV-31, HPV-33, HPV-35, HPV-39, HPV-45, HPV-51, HPV-52, HPV-56, HPV-58, HPV-59, HPV-66, and
HPV-68. Women included in the analysis were DNA negative for all 14 HPV types tested for, seronegative for HPV-16
and HPV-18, and had negative cytology at month 0. CIN2+ was defi ned histologically as CIN2, CIN3, adenocarcinoma
in situ, or invasive carcinoma. CIN=cervical intraepithelial neoplasia. HPV=human papillomavirus. TVC-naive=total
vaccinated HPV-naive cohort. N=number of evaluable women in each group (vaccine vs control). Cases=number of
evaluable women reporting at least one event. Rate=number of cases divided by sum of follow-up period
(per 100 woman years); follow-up period began the day after the fi rst vaccine dose.
Table 2: Cross-protective effi cacy against 6-month persistent infection and CIN2+ associated with
non-vaccine HPV types, in women who were HPV-naive at baseline (TVC-naive)
–30
–20
–10
0
10
20
30
40
50
60
70
80
90
100
56·2%
81·9%
17·1%
91·4%
Number of cases vaccine:
Number of cases control:
45
102 23
245 2
11
54
Vaccine efficacy (%)
With or without HPV-16/18 co-infection Excluding HPV-16/18 co-infection
CIN2+
CIN3+
Figure 2: Vaccine effi cacy against CIN2+ and CIN3+ associated with a composite of 12 non-vaccine
HPV types, with or without HPV-16/18 co-infection and excluding HPV-16/18 co-infection, in the
TVC-naive
Women had to have a lesion associated with at least one of the HPV types included in the composite. Women
included in the analysis of the TVC-naive were HPV DNA negative for the 14 HPV types tested for, seronegative
for HPV-16/18, and had negative cytology at month 0. Types tested for were HPV-16, HPV-18, HPV-31,
HPV-33, HPV-35, HPV-39, HPV-45, HPV-51, HPV-52, HPV-56, HPV-58, HPV-59, HPV-66, and HPV-68.
Follow-up period started the day after the fi rst vaccine dose. CIN2+ was defi ned histologically as CIN2, CIN3,
adenocarcinoma in situ, or invasive carcinoma; CIN3+ did not include CIN2. Vaccine effi cacy point estimates are
shown above each bar, and error bars represent 95% CI. CIN=cervical intraepithelial neoplasia. HPV=human
papillomavirus. TVC-naive=total vaccinated HPV-naive cohort.
Articles
www.thelancet.com/oncology Vol 13 January 2012
105
The trial is registered with ClinicalTrials.gov, number
NCT00122681.
Role of the funding source
The trial was funded by GlaxoSmithKline Biologicals, who
designed the study in collaboration with investi gators, and
coordinated collection, analysis, and inter pretation of data.
Investigators from the HPV PATRICIA Study Group
collected data for the trial and cared for the participants.
The authors had full access to all the trial data and had fi nal
responsibility for the decision to submit for publication.
Results
A total of 16 114, 11 644, and 18 644 women were included in
the ATP-E (vaccine n=8067, control n=8047), TVC-naive
(vaccine n=5824, control n=5820), and TVC cohorts
(vaccine n=9319, control n=9325), respectively. Compliance
was high. Roughly 16% of participants (3034 of 18 644)
were lost to follow-up by the end of the study; the number
of participants who did not complete the study was
balanced between the vaccine and control groups.17 In the
ATP-E cohort, mean and median follow-up times after
dose 3 were 39·8 months (SD 8·0) and 41·6 months
(range 0–55·5), respectively (3·3 and 3·5 years).
In the ATP-E cohort, vaccine effi cacy was consistently
high against all endpoints associated with HPV-33: 44·8%
(95% CI 24·6 to 59·9) for 6-month persistent infection,
68·3% (39·7 to 84·4) for CIN2+, and 59·4% (20·5 to 80·4)
for CIN2+ excluding HPV-16/18 co-infection (table 1).
Cross-protective effi cacy against all endpoints associated
with HPV-31 was also observed: 76·8% (69·0 to 82·9)
against 6-month persistent infection, 87·5% (68·3 to 96·1)
against CIN2+, and 84·3% (59·5 to 95·2) against CIN2+
excluding HPV-16/18 co-infection (table 1). Few events
associated with HPV-45 were observed; however, vaccine
effi cacy was 73·6% (58·1 to 83·9) for 6-month persistent
infection and 81·9% (17·0 to 98·1) for CIN2+.
Corresponding values against HPV-51 were 16·6%
(3·6 to 27·9) and 54·4% (22·0 to 74·2; table 1). In the
ATP-E cohort, vaccine effi cacy against the composite of
12 non-vaccine HPV types was 46·8% (30·7 to 59·4) for
CIN2+, 24·1% (–1·5 to 43·5) for CIN2+ excluding co-
infection with HPV-16/18, 73·8% (48·3 to 87·9) for
CIN3+, and 62·1% (21·8 to 82·9) for CIN3+ excluding
co-infection with HPV-16/18 (fi gure 1).
In the TVC-naive, vaccine effi cacy estimates with
95% CIs above zero were consistently noted for all
endpoints associated with HPV-33 and HPV-31 (table 2).
For HPV-45, vaccine effi cacy was 79·0% (95% CI
61·3 to 89·4) for 6-month persistent infection, 100%
(41·7 to 100) for CIN2+, and 100% (–429·7 to 100) for
CIN2+ excluding HPV-16/18 co-infection, although the
number of events was again small (table 2). Vaccine
effi cacy with 95% CIs above zero was also found against
HPV-51 for 6-month persistent infection (25·5%
[12·0 to 37·0]) and CIN2+ (70·2% [35·6 to 87·6]), but not
for CIN2+ excluding HPV-16/18 co-infection (0·5%
[–182·9 to 65·0]; table 2). For the composite endpoint of
12 non-vaccine HPV types, vaccine effi cacy was 56·2%
(37·2 to 69·9) against CIN2+, and 17·1% (–25·5 to 45·4)
when CIN2+ cases co-infected with HPV-16/18 were
excluded. Corresponding values for CIN3+ were 91·4%
(65·0 to 99·0) and 81·9% (17·1 to 98·1; fi gure 2).
The pattern of vaccine effi cacy across endpoints in
the TVC was similar to the ATP-E and TVC-naive, with
lower point estimates as expected in this broader cohort,
which included sexually active women with previous or
current HPV type-specifi c infections or lesions under
consideration at study entry. Compared with the ATP-E
and TVC-naive, more moderate vaccine effi cacy, albeit
with 95% CIs above zero, was seen in the TVC for 6-month
persistent infection with HPV-33, HPV-31, HPV-45, and
Vaccine Control Effi cacy (95% CI)
Cases Rate Cases Rate
6-month persistent infection (N=8863 vaccine vs 8870 control)
Non-vaccine A9 species
(composite HPV-31/33/35/52/58)
1179 4·00 1364 4·67 14·5% (7·5 to 20·9)
HPV-31 235 0·73 433 1·37 46·3% (36·9 to 54·3)
HPV-33 156 0·48 211 0·66 26·3% (8·9 to 40·4)
HPV-35 128 0·40 106 0·33 –21·1% (–58·2 to 7·1)
HPV-52 643 2·07 698 2·25 8·1% (–2·4 to 17·6)
HPV-58 245 0·76 216 0·67 –13·7% (–37·2 to 5·7)
Non-vaccine A7 species
(composite HPV-39/45/59/68)
769 2·50 838 2·73 8·5% (–1·1 to 17·1)
HPV-39 340 1·07 347 1·09 2·0% (–14·2 to 15·8)
HPV-45 70 0·22 153 0·47 54·5% (39·2 to 66·2)
HPV-59 130 0·40 117 0·36 –11·2% (–44·0 to 14·1)
HPV-68 284 0·89 296 0·92 4·0% (–13·3 to 18·7)
Other
HPV-51 636 2·05 732 2·37 13·7% (3·8 to 22·5)
HPV-56 351 1·10 357 1·12 1·6% (–14·4 to 15·3)
HPV-66 345 1·08 358 1·12 3·7% (–11·9 to 17·2)
CIN2+ with or without co-infection with HPV-16/18 (N=8694 vaccine vs 8708 control)
Non-vaccine A9 species
(composite HPV-31/33/35/52/58)
133 0·41 191 0·59 30·4% (12·7 to 44·6)
HPV-31 36 0·11 68 0·21 47·0% (19·5 to 65·7)
HPV-33 31 0·10 64 0·20 51·5% (24·5 to 69·5)
HPV-35 9 0·03 16 0·05 43·7% (–35·3 to 78·1)
HPV-52 55 0·17 61 0·19 9·8% (–32·1 to 38·5)
HPV-58 33 0·10 37 0·11 10·7% (–46·7 to 45·9)
Non-vaccine A7 species
(composite HPV-39/45/59/68)
34 0·10 71 0·22 52·1% (27·0 to 69·2)
HPV-39 13 0·04 24 0·07 45·8% (–10·8 to 74·7)
HPV-45 2 0·01 21 0·06 90·5% (61·0 to 98·9)
HPV-59 5 0·02 8 0·02 37·4% (–116·9 to 83·9)
HPV-68 15 0·05 23 0·07 34·7% (–30·6 to 68·3)
Other
HPV-51 38 0·12 76 0·23 50·0% (25·3 to 67·1)
HPV-56 11 0·03 23 0·07 52·2% (–2·1 to 79·0)
HPV-66 13 0·04 25 0·08 48·0% (–5·6 to 75·6)
(Continues on next page)
Articles
106
www.thelancet.com/oncology Vol 13 January 2012
HPV-51 (26·3% [8·9 to 40·4], 46·3% [36·9 to 54·3], 54·5%
[39·2 to 66·2], and 13·7% [3·8 to 22·5], respectively;
table 3). Vaccine effi cacy against CIN2+ associated with
HPV-33 was 51·5% (24·5 to 69·5), and 50·0% (14·7 to 71·4)
against CIN2+ excluding co-infection with HPV-16/18.
Vaccine effi cacy against CIN2+ was 47·0% (19·5 to 65·7)
for HPV-31, 90·5% (61·0 to 98·9) for HPV-45, and 50·0%
(25·3 to 67·1) for HPV-51; in the analysis of CIN2+
excluding co-infection with HPV-16/18, the lower limits of
the 95% CIs were below zero for these HPV types (table 3).
Once again, few events associated with HPV-45 were
observed. Vaccine effi cacy against the composite of 12 non-
vaccine HPV types was 34·2% (20·4 to 45·8) for CIN2+
and 6·2% (–18·1 to 25·6) for CIN2+ excluding co-infection
with HPV-16/18 (fi gure 3). Vaccine effi cacy estimates with
95% CIs above zero were consistently seen for CIN3+
associated with the 12-type composite: 47·5% (22·8 to 64·8)
and 40·0% (1·1 to 64·2) including and excluding
HPV-16/18 co-infection, respectively (fi gure 3).
Negative vaccine effi cacy with both 95% CIs below zero
was seen for CIN2+ associated with HPV-52 excluding
HPV-16/18 co-infection in the TVC (table 3), and for
12-month persistent infection with HPV-58 in the ATP-E
and TVC (webappendix p 3). For other endpoints and
cohorts, results for these two HPV types were in-
consistent, with both positive and negative vaccine effi cacy
point estimates.
Table 4 summarises vaccine effi cacy estimates across
the three diff erent cohorts (ATP-E, TVC-naive, and TVC).
Vaccine effi cacy estimates for 12-month persistant
infection and CIN3+ associated with individual HPV types
are shown in the webappendix p 3 and p 5, respectively.
Discussion
Data from the end-of-study analysis of PATRICIA show
that the HPV-16/18 vaccine provides cross-protective
effi cacy against 6-month persistent infection and CIN2+
associated with HPV-33, HPV-31, HPV-45, and HPV-51.
Consistent vaccine effi cacy for all endpoints across all
cohorts was seen only for HPV-33. As expected, estimates
of vaccine effi cacy were generally higher in the TVC-naive
and ATP-E cohorts than in the TVC. The TVC-naive and
ATP-E cohorts represent the primary target population
for the vaccine, in terms of little to no genital HPV
exposure, and show the potential eff ect of the vaccine
against new infections and lesions. By contrast, the TVC
includes women with pre-existing infections or lesions
associated with the HPV types considered in the analyses,
which are not expected to be aff ected by the prophylactic
vaccine. Factors that might have limited the general-
isability of the study results were enrolment of 80% of
the 15–17 year old stratum in a single country (Finland),
enrolment of 47% of the 18–25 year old stratum from
Asia-Pacifi c, and exclusion of women with more than six
lifetime sexual partners (this criterion did not apply to
15–17 year olds in Finland). Additionally, cross-protective
vaccine effi cacy could be diff erent in diff erent populations
as a result of host (eg, race or ethnicity) and viral factors
(eg, variant aminoacids potentially relevant to HPV cross-
protective epitopes).
Vaccine Control Effi cacy (95% CI)
Cases Rate Cases Rate
(Continued from previous page)
CIN2+ excluding co-infection with HPV-16/18 (N=8694 vaccine vs 8708 control)
Non-vaccine A9 species
(composite HPV-31/33/35/52/58)
112 0·35 115 0·35 2·4% (–27·7 to 25·5)
HPV-31 30 0·09 46 0·14 34·7% (–5·7 to 60·2)
HPV-33 22 0·07 44 0·14 50·0% (14·7 to 71·4)
HPV-35 6 0·02 7 0·02 14·2% (–198·2 to 76·2)
HPV-52 45 0·14 24 0·07 –87·9% (–222·4 to –12·1)
HPV-58 30 0·09 22 0·07 –36·6% (–148·5 to 23·8)
Non-vaccine A7 species
(composite HPV-39/45/59/68)
28 0·09 34 0·10 17·6% (–40·1 to 51·8)
HPV-39 10 0·03 13 0·04 23·0% (–90·1 to 69·8)
HPV-45 2 0·01 9 0·03 77·8% (–7·4 to 97·7)
HPV-59 5 0·02 5 0·02 –0·1% (–335·1 to 77·0)
HPV-68 12 0·04 8 0·02 –50·2% (–323·6 to 43·5)
Other
HPV-51 30 0·09 34 0·10 11·7% (–48·7 to 47·8)
HPV-56 7 0·02 7 0·02 –0·1% (–234·5 to 70·0)
HPV-66 8 0·02 16 0·05 –50·0% (–23·9 to 81·5)
Women could be infected with multiple HPV types (therefore the number of cases for the composite endpoints might
not equal the sum of the cases for each individual type included in the composite). Types tested for were HPV-16,
HPV-18, HPV-31, HPV-33, HPV-35, HPV-39, HPV-45, HPV-51, HPV-52, HPV-56, HPV-58, HPV-59, HPV-66, and
HPV-68. Women were included in the analysis irrespective of their HPV DNA or serostatus at month 0. CIN2+ was
defi ned histologically as CIN2, CIN3, adenocarcinoma in situ, or invasive carcinoma. CIN=cervical intraepithelial
neoplasia. HPV=human papillomavirus. TVC=total vaccinated cohort. N=number of evaluable women in each group
(vaccine vs control). Cases=number of evaluable women reporting at least one event. Rate=number of cases divided by
sum of follow-up period (per 100 woman years); follow-up period began the day after the fi rst vaccine dose.
Table 3: Cross-protective effi cacy against 6-month persistent infection and CIN2+ associated with
non-vaccine HPV types in women irrespective of their baseline HPV DNA and serostatus (TVC)
–20
–10
0
10
20
30
40
50
60
70
80
90
100
34·2%
40·0%
6·2%
47·5%
Number of cases vaccine:
Number of cases control:
182
276 80
42 148 27
45
158
Vaccine efficacy (%)
With or without HPV-16/18 co-infection Excluding HPV-16/18 co-infection
CIN2+
CIN3+
Figure 3: Vaccine effi cacy against CIN2+ and CIN3+ associated with a composite of 12 non-vaccine HPV types,
with or without HPV-16/18 co-infection and excluding HPV-16/18 co-infection, in the TVC
Women had to have a lesion associated with at least one of the HPV types included in the composite. Types tested
for were HPV-16, HPV-18, HPV-31, HPV-33, HPV-35, HPV-39, HPV-45, HPV-51, HPV-52, HPV-56, HPV-58, HPV-59,
HPV-66, and HPV-68. Women were included in the analysis of the TVC regardless of their HPV DNA or serostatus
at month 0. Follow-up period started the day after the fi rst vaccine dose. CIN2+ was defi ned histologically as CIN2,
CIN3, adenocarcinoma in situ, or invasive carcinoma; CIN3+ did not include CIN2. Vaccine effi cacy point estimates
are shown above each bar, and error bars represent 95% CI. CIN=cervical intraepithelial neoplasia. HPV=human
papillomavirus. TVC=total vaccinated cohort.
Articles
www.thelancet.com/oncology Vol 13 January 2012
107
The assessment of protection against non-vaccine
oncogenic HPV types poses a considerable challenge in
clinical trials, as shown by our analyses. Although CIN2+
is generally the preferred endpoint to evaluate vaccine
effi cacy against vaccine types, it has several limitations
for the evaluation of cross-protection; these limitations
also apply to CIN3+, which is less common. First, very
large sample sizes and extensive follow-up periods are
needed, because lesions are less often associated with
non-vaccine types than with HPV-16 or HPV-18. Second,
HPV DNA of non-vaccine and vaccine types might be
detected in biopsies taken from lesions, and causality of
the lesion cannot be defi nitively assigned to one HPV
type. Microdissection of lesions has been proposed as
a method for attributing causality, but it is impractical,
and it is unclear whether it would be useful.21 Third,
HPV-16 infections are more likely than infections with
any other HPV type to progress to a detectable lesion.22
The carcinogenicity of HPV-16 is unique, according to
recognised risk criteria, including attrib utable fraction
of prevalent CIN3 or cancer, probability of persistence,
and detection of incident CIN3 or cancer.23 CIN3+
associated with HPV-16 is diagnosed much earlier than
other HPV types,24 presumably as a result of greater
chromosomal instability and corresponding cellular
transformation induced by deregulated HPV E6 and
E7 oncogene expression.25 Thus, the increased likelihood
of progression of HPV-16 infection to detectable lesions
leads to increased referral and therapy, which could
diff erentially aff ect the development and detection of
lesions caused by other HPV types among vaccinated
versus non-vaccinated women.
We attempted to address the confounding eff ect of
multiple infections in lesions by calculating two separate
estimates of vaccine effi cacy against CIN2+ and CIN3+.
Our analyses of CIN2+ and CIN3+ with or without co-
infection with HPV-16/18 are likely to overestimate cross-
protective effi cacy, since some cases counted in the
analyses were probably caused by HPV-16 or HPV-18.
However, our analyses excluding cases co-infected with
HPV-16 or HPV-18 are very conservative. For example,
cases were excluded if two independent lesions were
found in the same woman, one infected with HPV-16/18
and one infected with a non-vaccine type, even if the
lesions occurred at diff erent sampling timepoints.
Additionally, because of the effi cacy of the vaccine against
HPV-16/18 infections, fewer cases were co-infected with
HPV-16/18 in the vaccine group than in the control group.
As a result, most cases removed were from the control
group, including some that might have been caused by a
non-vaccine type. Thus, true vaccine effi cacy against
CIN2+ or CIN3+ associated with non-vaccine HPV types
possibly lies somewhere between the two estimates.
Reports of cross-protection conferred by HPV vaccines
against lesions associated with oncogenic HPV types have
not always taken this conservative approach.10,26
ATP-E TVC-naive TVC
6-month
persistent
infection
CIN2+ with
or without
HPV-16/18
co-infection
CIN2+
excluding
HPV-16/18
co-infection
6-month
persistent
infection
CIN2+ with
or without
HPV-16/18
co-infection
CIN2+
excluding
HPV-16/18
co-infection
6-month
persistent
infection
CIN2+ with
or without
HPV-16/18
co-infection
CIN2+
excluding
HPV-16/18
co-infection
Non-vaccine
A9 species (composite
HPV-31/33/35/52/58)
22·0%* 49·1%* 29·5% 27·6%* 56·6%* 28·3% 14·5%* 30·4%* 2·4%
HPV-31 76·8%* 87·5%* 84·3%* 77·1%* 89·4%* 83·4%* 46·3%* 47·0%* 34·7%
HPV-33 44·8%* 68·3%* 59·4%* 43·1%* 82·3%* 76·3%* 26·3%* 51·5%* 50·0%*
HPV-35 –19·8% 62·5% 39·9% –21·8% 83·4% 66·8% –21·1% 43·7% 14·2%
HPV-52 8·3% 27·6% –25·8% 18·9%* 30·4% –132·3% 8·1% 9·8% –87·9%†
HPV-58 –18·3% 28·5% –7·3% –6·2% 36·1% –11·9% –13·7% 10·7% –36·6%
Non-vaccine
A7 species (composite
HPV-39/45/59/66)
11·6% 58·2%* 30·4% 22·3%* 71·6% 20·4% 8·5% 52·1%* 17·6%
HPV-39 4·8% 74·9%* 42·7% 20·9% 72·9% 40·3% 2·0% 45·8% 23·0%
HPV-45 73·6%* 81·9%* 50·1% 79·0%* 100%* 100% 54·5%* 90·5%* 77·8%
HPV-59 –7·5% 80·0% 66·6% –3·9% 100% 100% –11·2% 37·4% –0·1%
HPV-68 2·6% 26·8% 10·1% 8·9% 54·8% –65·8% 4·0% 34·7% –50·2%
Other
HPV-51 16·6%* 54·4%* 13·7% 25·5%* 70·2%* 0·5% 13·7%* 50·0%* 11·7%
HPV-56 –5·3% 46·1% 40·0% 1·4% 100%* 100% 1·6% 52·2% –0·1%
HPV-66 2·3% 56·4% 36·5% –1·5% 72·9% 57·4% 3·7% 48·0% –50·0%
ATP-E=according-to-protocol for effi cacy. TVC-naive=total vaccinated HPV-naive cohort. TVC=total vaccinated cohort. CIN=cervical intraepithelial neoplasia. HPV=human
papillomavirus. *Vaccine effi cacy values with lower limit of the 95% CI above 0. †Negative vaccine efficacy values with entire 95% CI below 0.
Table 4: Summary of cross-protective effi cacy across endpoints and cohorts
Articles
108
www.thelancet.com/oncology Vol 13 January 2012
Virological endpoints are valuable in evaluation of
cross-protective vaccine effi cacy. Persistent infection with
oncogenic HPV types usually precedes the detection of
CIN2+13–16 and occurs substantially more often, allowing
suffi cient statistical power using a realistic sample size in
vaccine trials. We chose to use persistent infection of at
least 6 months since this was comparable to persistent
infection of at least 12 months as a pre dictor of developing
CIN2+ (webappendix p 7),27 although occurring at higher
rates and providing additional statistical power. Unlike
CIN2+ and CIN3+, persistent infection directly refl ects
the eff ect of vaccin ation on an individual HPV type, is
presumably not biased by multiple infections, and does
not pose the problem of assessing causality. Therefore,
using viro logical endpoints as indicative of cross-
protection, and clinical endpoints as confi rmatory
evidence, is a scien tifi cally sound and robust approach.
In our study, cross-protection against endpoints
associated with HPV-33—6-month persistent infection,
CIN2+, and CIN2+ excluding HPV-16/18 co-infection—
was consistently observed across cohorts. HPV-33 is the
fourth most prevalent HPV type after HPV-16, HPV-18,
and HPV-45 in ICC worldwide (4% prevalence),3 and
growing evidence supports the high-risk nature of HPV-33.
For example, HPV-33 incident infections are at a very high
risk of progression to CIN2 or CIN3,28,29 and among
women positive for HPV-33, the percent risk of carcinoma
in situ or adenocarcinoma in situ, relative to HPV-16, has
been reported to be 101% (95% CI 62 to 163)—ie, essentially
equal to that for women positive for HPV-16.24
Vaccine effi cacy against HPV-31 was consistently shown
across most cohorts and endpoints. High effi cacy estimates
for 6-month persistent infection and CIN2+ associated
with HPV-31 were obtained in the ATP-E and TVC-naive
cohorts. As expected, estimates of vaccine effi cacy were
lower in the TVC, which was the only cohort where
analysis of CIN2+ associated with HPV-31 excluding co-
infection with HPV-16/18 resulted in an effi cacy estimate
with the lower limit of the 95% CI below zero. For HPV-45,
the type most closely related to HPV-18, cross-protective
vaccine effi cacy was seen against 6-month persistent
infection and against CIN2+ with and without HPV-16/18
co-infection, across all cohorts. Vaccine effi cacy against
the composite of tested non-vaccine A7 species, including
HPV-45, was seen for some endpoints, probably driven by
effi cacy against HPV-45. HPV-45 is the third most prevalent
type after HPV-16 and HPV-18, and causes roughly 6%
of all ICC.3 HPV-45 is more frequent in adenocarcinoma
(12% of cases) than in SCC (5% of cases).3 Although rarer
than SCC, adenocarcinoma represents up to 25% of
cervical cancers.3,30
We also noted consistent cross-protection against
6-month persistent infection and CIN2+ associated with
HPV-51, although at lower levels than for HPV-33,
HPV-31, and HPV-45. HPV-51 is ranked as the tenth or
eleventh most common HPV type associated with ICC
worldwide (about 1% prevalence),3,5 but ranks higher in
precursor cervical lesions such as high-grade squamous
intra epithelial lesions (3·6% worldwide).31 Although there
was suggestion of cross-protection against other HPV
types, such as HPV-52, HPV-56, and HPV-39, the results
for virological and clinical endpoints were not consistent,
and might therefore represent chance observations.
Negative vaccine effi cacy was noted for some endpoints
and cohorts, for HPV-52 and HPV-58. In theory, negative
vaccine effi cacy could represent a chance fi nding, a reduced
sensitivity of the PCR for some non-vaccine HPV types in
case of multiple infections that are more common in the
control arm because of effi cacy of the vaccine, or the
occurrence of HPV type replacement; however, the latter is
unlikely in the context of a clinical trial. Moreover, vaccine
effi cacy results for HPV-52 and HPV-58 were generally
inconsistent across endpoints and cohorts. In the long
term, postmarketing surveillance programmes will be
important to properly characterise any changes in type-
specifi c HPV prevalence and disease incidence, particularly
among cervical precancers (ie, CIN2 and CIN3).
In addition to effi cacy against the individual HPV types
described above, vaccine effi cacy was consistently shown
across cohorts for the composite of 12 non-vaccine HPV
Panel: Research in context
Systematic review
The present article reports part of a prophylactic HPV vaccine development programme.
Studies in the programme were done to achieve licensure of the vaccine and to examine
how the vaccine might be best used in real-world settings, and were developed in
conjunction with leading experts in HPV vaccine research and with regulatory bodies.
Literature related to HPV vaccination studies was systematically followed before the start
of the study, during the trial, and during development of the publication (1997–2011).
The volume of literature has now increased, and we used our knowledge and expertise to
select the trials we thought were most relevant for the present report.
Interpretation
About 30% of invasive cervical cancer is caused by HPV types not included in current
prophylactic HPV vaccines. The level of cross-protection against these non-vaccine HPV
types is therefore an important component of the overall level of protection against
cervical cancer off ered by an HPV vaccine. This end-of-study analysis of PATRICIA reports a
comprehensive evaluation of cross-protection conferred by the HPV-16/18 vaccine in
diverse populations of women, including against the most stringent endpoint, CIN3+.
Selection of appropriate endpoints to evaluate cross-protection is a challenge; we included
analyses of both virological and clinical endpoints.
Our analysis showed that the HPV-16/18 vaccine off ers cross-protective effi cacy against
HPV-33, HPV-31, HPV-45, and HPV-51. HPV-16/18 and these four types cause about
85% of cervical cancer; moreover, there is a particularly high risk of HPV-33 infections
progressing to cervical lesions, and HPV-45 is over-represented in adenocarcinoma.
Vaccine effi cacy against the most stringent endpoint, CIN3+ associated with
12 non-vaccine HPV types excluding co-infection with HPV-16/18, was around 80% in
HPV-naive women. Our results show that cross-protective effi cacy might provide
substantial additional protection against cervical cancer beyond protection conferred
against HPV-16/18. These are important data for doctors and public health bodies when
estimating the overall reduction in cervical lesions and invasive cancer likely to result from
immunisation programmes using the HPV-16/18 vaccine.
Articles
www.thelancet.com/oncology Vol 13 January 2012
109
types. Notably, removing HPV-16/18 co-infected lesions
from the analysis had little eff ect on observed cross-
protective vaccine effi cacy. In the TVC-naive, the point
estimates of vaccine effi cacy were 91·4% for CIN3+, and
81·9% when CIN3+ cases co-infected with HPV-16/18
were excluded. In the TVC, vaccine effi cacy was 47·5%
for CIN3+ and 40·0% for CIN3+ excluding HPV-16/18
co-infected cases. Because non-vaccine HPV types are
responsible for about 30% of cervical cancers, these data
suggest that cross-protection could provide substantial
additional protection against cervical cancers beyond
protection conferred against HPV-16/18.
Vaccine effi cacy was highest for CIN3+ and lower for
CIN2+ and persistent infection. Additionally, the diff erence
in vaccine effi cacy with exclusion versus non-exclusion of
HPV-16/18 co-infections was greater for CIN2+ than for
CIN3+. These results are expected because the attributable
proportion of A9 and A7 species (which include HPV
types for which cross-protection was observed) rises from
infection to increasingly severe lesions,23 and detection of
co-infections with any oncogenic HPV type progressively
decreases as lesion severity increases.32,33 In analyses
excluding HPV-16/18 co-infection, cases are removed
more often from the control group than from the vaccine
group, because of the effi cacy of the vaccine against
HPV-16/18. This introduces a bias against the vaccine (ie,
lowers the vaccine effi cacy point estimate). Because fewer
CIN3+ cases are co-infected than CIN2+ cases, fewer
CIN3+ cases are removed from the analysis, so there is
less eff ect on the vaccine effi cacy point estimate for CIN3+
than for CIN2+. It is important to note that, although extra
benefi t off ered by cross-protection has important public
health value, it is impossible to predict whether individual
women will be protected against vaccine or non-vaccine
HPV types. Additionally, the duration of cross-protective
effi cacy is unknown; this should be assessed in long-term
follow-up, including population surveillance and
eff ectiveness studies in real-world settings.18,34
A possible explanation for the high cross-protection
seen with the HPV-16/18 vaccine is the presence of the
AS04 Adjuvant System in the vaccine formulation, which
enhances the overall immune response. The HPV-16/18
vaccine induces cross-neutralising antibodies for HPV-31
and HPV-45, raising the possibility that such antibodies
might eff ect cross-protection.35 However, it is unknown
whether the levels of cross-reactive antibodies will help
sustain vaccine-induced cross-protection against non-
vaccine HPV types over the long term. The immune
mechanisms of vaccine-induced cross-protection are not
fully understood, but are most likely linked to conserved
aminoacid sequences or structural similarities within
shared neutralising epitopes among HPV types.6–8,36–39 Only
a few HPV types that belong to the same species as HPV-16
(A9: HPV-31 and HPV-33) or HPV-18 (A7: HPV-45) were
associated with cross-protection. This suggests that minor
diff erences in aminoacid sequences or structure could be
important in the recognition of neutralising epitopes by
vaccine-induced antibodies. However, the results for
HPV-51 (A5 species) show that the cross-protection
induced by the vaccine extends outside the A9 and
A7 species, possibly due to sequence-based or functional
similarities at critical aminoacids in shared (most likely
conformational) epitopes, although effi cacy for HPV-51
was lower than for HPV-31, HPV-33, and HPV-45.
In conclusion, our analyses show some of the challenges
in evaluating cross-protective effi cacy of HPV vaccines.
They highlight the importance of using both virological
and clinical endpoints, and of observing consistency
between these endpoints before concluding on cross-
protection. Our analyses also provide additional evidence
for cross-protective effi cacy of the HPV-16/18 vaccine
against HPV-33, HPV-31, HPV-45, and HPV-51 in
diff erent cohorts representing diverse groups of women.
Overall, we anticipate that the cross-protective effi cacy of
the HPV-16/18 vaccine when administered to HPV-naive
women might provide substantial additional protection
against cervical cancer over and above that achieved by
effi cacy against HPV-16/18 (panel), but long-term follow-
up is needed to confi rm this.
Contributors
CMW, ML, SMG, AS, XC, DD, FS, and GD formed the core writing team
for the manuscript. All authors reviewed and commented on a draft of
the manuscript and gave fi nal approval to submit for publication. All
authors contributed to study design, acquisition of data or statistical
analyses, and interpretation of data. For the HPV PATRICIA Study
Group, see webappendix p 9.
Confl icts of interest
DD, GD, FS, KH, and TZ are employees of GlaxoSmithKline Biologicals.
DD, GD, FS, and KH own stock in GlaxoSmithKline Biologicals, and GD
holds a relevant patent. All investigators at study clinical sites were funded
through their institutions to do the study protocol. CMW, DMH, DA, JP,
PN, HK, FYA, FXB, SRS, SMG, ML, TFS, AS, XC, JCT, WH, and BR have
received funding through their institutions to do HPV vaccine studies for
GlaxoSmithKline Biologicals or Merck Sharp & Dohme (Sanofi Pasteur
MSD). JP received a research grant through the Helsinki University
Hospital Research Institute to conduct clinical trials on HPV vaccination.
SRS has also received funding through her institution from CSL to do
research on school-based adolescent HPV vaccination. Through the
University of New Mexico, CMW has received equipment and reagents for
HPV genotyping from Roche Molecular Systems and funding for HPV
vaccine studies from GlaxoSmithKline (in addition to the present study)
and Merck & Co. FXB is an editor of the international newsletter (HPV
TODAY) and guest editor of the journal Vaccine to prepare international
reviews on topics related to HPV.WAJP, FXB, WH, XC, SMG, PN, JCT,
BR, TFS, and AS have received consulting fees. DMH, SMG, SRS, FYA,
PN, and TFS have received honoraria; TFS, BR, and FXB have been paid
for expert testimony; BR, FYA, SRS, DMH, JCT, and WAJP have received
payment for board membership; JCT, FYA, XC, PN, FXB, BR, and TFS
have received payment for lectures, including service on speakers bureau;
AS, FYA, FXB, and BR have received payment for development of
educational presentations; and JS, WAJP, JCT, SRS, PN, XC, FXB, WH,
UJ, FYA, JH, SMG, AS, and CMW have received travel reimbursements
from GlaxoSmithKline Biologicals or Merck Sharp & Dohme (Sanofi
Pasteur MSD), or both. DA has received support for travel from
Väestöliitto. S-NC and GL declare that they have no confl icts of interest.
Acknowledgments
This study (NCT00122681/580299/008) was funded and coordinated
by GlaxoSmithKline Biologicals. We thank study participants and
their families. We thank Mary Greenacre for writing and editorial assistance,
and Jenny Andersson (Cromsource) for editorial assistance and manuscript
coordination on behalf of GlaxoSmithKline Biologicals, Wavre, Belgium.
Articles
110
www.thelancet.com/oncology Vol 13 January 2012
References
1 Walboomers JM, Jacobs MV, Manos MM, et al. Human
papillomavirus is a necessary cause of invasive cervical
cancer worldwide. J Pathol 1999; 189: 12–19.
2 Muñoz N, Bosch FX, de Sanjosé S, et al. Epidemiologic
classifi cation of human papillomavirus types associated
with cervical cancer. N Engl J Med 2003; 348: 518–27.
3 de Sanjosé S, Quint WGV, Alemany L, et al, on behalf of the
Retrospective International Survery and HPV Time Trends Study
Group. Human papillomavirus genotype attribution in invasive
cervical cancer: a retrospective cross-sectional worldwide study.
Lancet Oncol 2010; 11: 1048–56.
4 Schiff man M, Cliff ord G, Buonaguro FM. Classifi cation of
weakly carcinogenic human papillomavirus types: addressing the
limits of epidemiology at the borderline. Infect Agent Cancer 2009;
4: 8.
5 Li N, Franceschi S, Howell-Jones R, Snijders PJ, Cliff ord GM.
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: 927–35.
6 Bishop B, Dasgupta J, Klein M, et al. Crystal structures of four types
of human papillomavirus L1 capsid proteins: understanding the
specifi city of neutralizing monoclonal antibodies. J Biol Chem 2007;
282: 31803–11.
7 Ryding J, Dahlberg L, Wallen-Ohman M, Dillner J. Deletion of a
major neutralizing epitope of human papillomavirus type 16
virus-like particles. J Gen Virol 2007; 88: 792–802.
8 Chen Z, DeSalle R, Schiff man M, Herrero R, Burk RD.
Evolutionary dynamics of variant genomes of human
papillomavirus types 18, 45, and 97. J Virol 2009; 83: 1443–55.
9 Smith JF, Brownlow M, Brown M, et al. Antibodies from women
immunized with Gardasil cross-neutralize HPV-45 pseudovirions.
Hum Vaccin 2007; 3: 109–16.
10 Paavonen J, Jenkins D, Bosch FX, et al, for the HPV PATRICIA
Study Group. Effi cacy 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: 2161–70.
11 Paavonen J, Naud P, Salmerón J, et al, for the HPV PATRICIA
Study Group. Effi cacy of human papillomavirus (HPV)-16/18
AS04-adjuvanted vaccine against cervical infection and precancer
caused by oncogenic HPV types (PATRICIA): fi nal analysis of a
double-blind, randomised study in young women. Lancet 2009;
374: 301–14.
12 Muñoz N, Kjaer SK, Sigurdsson K, et al. Impact of human
papillomavirus (HPV)-6/11/16/18 vaccine on all HPV-associated
genital diseases in young women. J Natl Cancer Inst 2010;
102: 325–39.
13 Hildesheim A, Schiff man MH, Gravitt ME, et al. Persistence of
type-specifi c human papillomavirus infection among cytologically
normal women. J Infect Dis 1994; 169: 235–40.
14 Ho GY, Burk RD, Klein S, et al. Persistent genital human
papillomavirus infection as a risk factor for persistent cervical
dysplasia. J Natl Cancer Inst 1995; 87: 1365–71.
15 Koshiol J, Lindsey L, Pimenta JM, Poole C, Jenkins D, Smith JS.
Persistent human papillomavirus infection and cervical neoplasia:
a systematic review and meta-analysis. Am J Epidemiol 2008;
168: 123–37.
16 Kjær SK, Frederiksen K, Munk C, Iftner T. Long-term absolute
risk of cervical intraepithelial neoplasia grade 3 or worse
following human papillomavirus infection: role of persistence.
J Natl Cancer Inst 2010; 102: 1478–88.
17 Lehtinen M, Paavonen J, Wheeler CM, et al, for the HPV
PATRICIA Study Group. Overall effi cacy of HPV-16/18
AS04-adjuvanted vaccine against grade 3 or greater cervical
intraepithelial neoplasia: 4-year end-of-study analysis of the
randomised, double-blind PATRICIA trial. Lancet Oncol 2011;
published online Nov 9. DOI:10.1016/S1470-2045(11)70286-8.
18 Lehtinen M, Apter D, Dubin G, et al. Enrolment of 22,000
adolescent women to cancer registry follow-up for long-term human
papillomavirus vaccine effi cacy: guarding against guessing.
Int J STD AIDS 2006; 17: 517–21.
19 van Doorn L-J, Molijn A, Kleter B, Quint W, Colau B. Highly
eff ective detection of human papillomavirus 16 and 18 DNA by
a testing algorithm combining broad-spectrum and type-specifi c
PCR. J Clin Microbiol 2006; 44: 3292–98.
20 Bouvard V, Baan R, Straif K, et al. A review of human
carcinogens—Part B: biological agents. Lancet Oncol 2009; 10: 321–22.
21 Quint W, Molijn A, Guimera N, et al. Which HPV type caused
that lesion? The value of the laser capture microscopy technology.
HPV Today 2011; 23: 8–9.
22 Cliff ord GM, Smith JS, Aguado T, Franceschi S. Comparison of
HPV type distribution in high-grade cervical lesions and cervical
cancer: a meta-analysis. Br J Cancer 2003; 89: 101–05.
23 Schiff man M, Herrero R, DeSalle R, et al. The carcinogenicity of
human papillomavirus types refl ects viral evolution. Virology 2005;
337: 76–84.
24 Wheeler CM, Hunt WC, Joste NE, Key CR, Quint WGV,
Castle PE. Human papillomavirus genotype distributions:
implications for vaccination and cancer screening in the
United States. J Natl Cancer Inst 2009; 101: 475–87.
25 Vinokurova S, Wentzensen N, Kraus I, et al. Type-dependent
integration frequency of human papillomavirus genomes in cervical
lesions. Cancer Res 2008; 68: 307–13.
26 Wheeler CM, Kjaer SK, Sigurdsson K, 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 sexually active
women aged 16–26 years. J Infect Dis 2009; 199: 936–44.
27 Brown DR, Kjaer SK, Sigurdsson K, 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: 926–35.
28 Bulkmans NW, Bleeker MCG, Berkhof J, Voorhorst FJ, Snijders PJF,
Meijer CJLM. Prevalence of types 16 and 33 is increased in high-risk
human papillomavirus positive women with cervical intraepithelial
neoplasia grade 2 or worse. Int J Cancer 2005; 117: 177–91.
29 Insinga RP, Perez G, Wheeler CM, et al, FUTURE I Investigators.
Incident cervical HPV infections in young women:
transition probabilities for CIN and infection clearance.
Cancer Epidemiol Biomarkers Prev 2011; 20: 287–96.
30 Smith HO, Tiff any MF, Qualls CR, Key CR. The rising incidence
of adenocarcinoma relative to squamous cell carcinoma of the
uterine cervix in the United States—a 24-year population-based
study. Gynecol Oncol 2000; 78: 97–105.
31 Bosch FX, Burchell AN, Schiff man M, et al. Epidemiology and
natural history of human papillomavirus infections and type-specifi c
implications in cervical neoplasia. Vaccine 2008; 26: 1–16.
32 Wentzensen N, Schiff man M, Dunn ST, et al. Multiple HPV
genotype infections in cervical cancer progression in the Study
to Understand Cervical Cancer Early Endpoints and Determinants
(SUCCEED). Int J Cancer 2009; 125: 2151–58.
33 Zuna RE, Allen RA, Moore WE, Lu Y, Mattu R, Dunn ST.
Distribution of HPV genotypes in 282 women with cervical
lesions: evidence for three categories of intraepithelial lesions based
on morphology and HPV type. Mod Pathol 2007; 20: 167–74.
34 Herrero R, Hildesheim A, Rodriguez AC, et al. Rationale and design
of a community-based double-blind clinical trial of an HPV-16 and
18 vaccine in Guanacaste, Costa Rica. Vaccine 2008; 26: 4795–808.
35 Kemp T, Hildesheim A, Safaeian M, et al. HPV16/18 L1 VLP
vaccine induces cross-neutralizing antibodies that may mediate
cross-protection. Vaccine 2011; 29: 2011–14.
36 Giroglou T, Sapp M, Lane C, et al. Immunological analyses
of human papillomavirus capsids. Vaccine 2001; 19: 1783–93.
37 Combita AL, Touzé A, Bousarghin L, Christensen ND, Coursaget P.
Identifi cation of two cross-neutralizing linear epitopes within the
L1 major capsid protein of human papillomaviruses. J Virol 2002;
76: 6480–86.
38 Roth SD, Sapp M, Streeck RE, Selinka HC. Characterization of
neutralizing epitopes within the major capsid protein of human
papillomavirus type 33. Virology J 2006; 3: 83.
39 Fleury MJ, Touzé A, Maurel MC, Moreau T, Coursaget P.
Identifi cation of neutralizing conformational epitopes on
the human papillomavirus type 31 major capsid protein
and functional implications. Protein Sci 2009; 18: 1425–38.
