Vous êtes sur la page 1sur 4

Available online at www.sciencedirect.


Gynecologic Oncology 110 (2008) S4 – S7


Natural history and epidemiology of HPV infection and cervical cancer

Xavier Castellsagué
Unit of Infections and Cancer, Institut Català d'Oncologia (ICO), IDIBELL, CIBER-ESP
Received 23 July 2008


Cervical cancer is the most common cancer affecting women in developing countries. It has been estimated to have been responsible for almost
260 000 deaths annually, of which about 80% occurred in developing countries.
Persistent infection by certain oncogenic HPV types is firmly established as the necessary cause of most premalignant and malignant epithelial
lesions of the cervix and of a variable fraction of neoplastic lesions of the vulva, vagina, anus, penis, and oropharynx.
There are more than 100 known HPV genotypes, at least 15 of which can cause cancer of the cervix and other sites. HPV 16 and 18, the two
most common oncogenic types, cause approximately 70% of all cervical cancers worldwide. HPV, especially genotypes 6 and 11, can also cause
genital warts. HPV is highly transmissible and it is now considered the most common sexually transmitted infection in most populations. Although
most women infected with the virus become negative within 2 years, women with persistent high-risk HPV infections are at greatest risk for
developing cervical cancer.
Since the identification of HPV as the necessary cause of cervical cancer, HPV-based technology has become the centre of novel primary and
secondary cervical cancer prevention strategies by the introduction of HPV testing in screening and of HPV vaccines in preadolescent girls and
young women. If implemented widely and wisely the deployment of these protocols has the potential to complete Papanicolaou's goal of cervical
cancer eradication by extending the benefits of prevention to the developing populations of the world.
© 2008 Elsevier Inc. All rights reserved.

Introduction have cohabited with the human specie over dozens of millennia
suffering relatively few changes in their genetic composition.
One of the most important discoveries in the etiologic These nearly 100 different types of papillomavirus identified
investigation of cancer over these last 25 years has been the express a characteristic tropism. Some types are cutaneotropic
demonstration that cervical cancer is caused by the persistent (HPVs 1, 4, 5, 8, 41, 48, 60, 63 and 65) and are isolated frequently
infection by certain genotypes of the Human Papilloma Virus in cutaneous and plantar warts, in cutaneous lesions in the patients
(HPV). The scientific evidence accumulated from virological, with verruciform epidermodisplasia, in cutaneous lesions in
molecular, clinical and epidemiological studies has demon- immuno-depressed patients after a transplant and in some
strated unequivocally that cervical cancer is in fact a sequel to a epithelial tumours. Another group of HPVs are mucosotropic
long term unresolved infection by certain genotypes of the HPV (HPVs 6, 11, 13, 44, 55, 16, 31, 33, 35, 52, 58, 67, 18, 39, 45, 59,
[1]. Thus, we can now affirm that cervical cancer is the final 68, 70, 26, 51, 69, 30, 53, 56, 66, 32, 42, 34, 64, 73, 54) and they
result of a viral infection and, as such, vaccination is a strategy are identified in benign and malignant lesions of the anogenital
to consider in the primary prevention of cancers and other tract in both sexes. Occasionally, these viral types are isolated in
diseases caused by HPVs. tissues and lesions of the oral cavity, oropharynx, larynx and
oesophagus. Finally, another group of HPVs is isolated
The Human Papilloma Virus (HPV) indifferently in cutaneous or mucous tissues and lesions (HPVs
2, 3, 7, 10, 27, 28, 29, 40, 43, 57, 61, 62 and 72) and its association
The human papilloma viruses (HPVs) are DNA double strand with malignant lesions is less established.
viruses and of small size (approximately 8000 pairs of bases) that The best known clinical expression of the viral infection is
condylomas or genital warts, associated in approximately 90%
E-mail address: xcastellsague@iconcologia.net. of the cases by infections by HPVs 6 and 11 and, more rarely,
0090-8258/$ - see front matter © 2008 Elsevier Inc. All rights reserved.
X. Castellsagué / Gynecologic Oncology 110 (2008) S4–S7 S5

with HPVs 42 and 16. These viral types can be transmitted from High prevalence groups can be identified socially in the
the mother to the newborn at the time of childbirth and cause population of women who practice prostitution and in persons
laryngeal papillomatosis, a strange complication but of infected by the Human Immunodeficiency Virus (HIV). HPV
recidivist prognosis and difficult to treat. transmission takes place, mainly, by sexual contact and the
The neoplastic cervical (CIN), vulvar (VIN), vaginal (VaIN), organs most susceptible to infection with potential of starting a
penile (PIN) and anal (AIN) lesions are associated occasionally neoplastic transformation are the cervix (transformation zone)
with “benign” or “low risk” HPVs such as HPV 6 and HPV 11, and the pectineal line of the anal canal. HPV infections are
but more frequently with the typically “high risk” carcinogenic frequently in sheet and HPV DNA can be detected in the cervix,
or oncogenic HPVs like HPV 16, 18, 45 and 31. More than 35 vagina, and vulva in the woman, the glans, prepuce and skin of
types of HPV have been isolated in neoplastic lesions of the the penis and scrotum in the man, and in the anal canal and
anogenital tract. perianal area in women and men.
At the ages of greatest sexual activity, the prevalence of
Natural History of Infections by HPV subclinical HPV infections (presence of viral DNA with normal
morphology or minimal changes) can be up to 40% in the
Infection by HPV is basically a sexually transmitted disease. female population, with an annual infection rate of 10–15%. In
As such, both men and women are involved in the the age groups beyond 30 years, the prevalence decreases to 5–
epidemiological chain of infection and are able at the same 10%. The half life of infections by HPV has been estimated at
time to be asymptomatic carriers, transmitters and also victims 8–10 months for the high risk types and of approximately half
of the infection by HPV [2]. In this sense, the risk factors that for low risk viral types. The infections by HPV 16 are those
associated with the infection by HPV are clearly related with the that present the most prolonged longevities with average
individual's sexual behaviour. The most important are: early age persistence values of 16 months in some studies [6]. The
at the start of the first sexual relationships, high number of resolution of the infection seems to offer a certain degree of
sexual partners throughout life, sexual contacts with high risk protection with respect to re-infections for the same type of
individuals (in men, frequent contact with women that practice HPV, with a certain degree of cross immunity between viral
prostitution and in women, frequent contacts with men with types being described (in few studies).
multiple sexual partners). Male circumcision and the strict and It is important to emphasize that at young ages and at the
systematic use of condoms are factors that can reduce, although most sexually active ages in spite of very frequent infection by
without totally preventing, the risk of transmission of HPV HPV, the great majority of infected women (more than 90%)
between sexual partners [3–5]. resolve the infection spontaneously and the infection persists in

Fig. 1. Estimated percentage of cases of cervical cancer attributed to the most frequent HPV types in all the regions of the world combined. Estimated from: (A) the
combined analysis of the IARC with 3,085 cases [(adapted from [17]], and (B) a meta-analysis including more than 14,500 cases [(adapted from [18]].
S6 X. Castellsagué / Gynecologic Oncology 110 (2008) S4–S7

only a small fraction of women [7]. It is this small group of

women, chronic carriers of high risk HPVs, who have a high
risk of progression and development of neoplastic lesions of the
anogenital tract [8].
The well-known determinants of progression are the viral
type, the persistence of the infection upon repeated examina-
tions and, probably, the viral load per cellular unit, as well as
the integration of the viral DNA into the cellular DNA.
Infection by HIV constitutes a risk factor for infection as well
as for neoplastic progression, in particular during the immuno-
suppresion periods. Additional established progression factors
include prolonged use of oral contraceptives, high parity and
tobacco smoking [9,10]. Possible factors include diet, in
particular diets poor in fruits and vegetables, and co-infection
by other sexually transmitted agents such as Chlamydia
trachomatis and Herpes Simplex Virus type 2 [11–13].

HPV and Cervical Cancer

The epidemiological and clinical studies that have incorpo-

rated molecular biology techniques of high sensitivity in
appropriate biological specimens, detect oncogenic or high Fig. 2. Estimation of cervical cancer risk by specific HPV types.
risk HPVs in practically 100% of the cervical cancers. Formally
it has resulted in questioning the existence of cervical cancers
unassociated with HPV. Equally, viral DNA is detected in the
majority (70 – 90%) of the precursor lesions or intraepithelial high, between 95 and 99%. The associations observed between
lesions of high degree (CIN II-III) and, in a smaller fraction the infection by HPV and cervical cancer are among the highest
(20 – 50%), in the low grade lesions (CIN I). Finally, in the ever identified in human cancerology, with a growing consensus
category of cytological lesions of uncertain nature (ASCUS and existing in qualifying HPV as the “necessary cause” (absence
AGUS) the detection of HPV is close to 50%. of disease in absence of infection) although insufficient
The most frequent viral types in cases of invasive carcinoma (presence of infection without presence of disease), due to the
are reflected in Fig. 1. Among the cases, the cumulative great number of infections of cervical cancer that are resolved
prevalence of 4 types (HPVs 16, 18, 45 and 31) would explain spontaneously [1,15–17].
80% of the cases approximately. This information is of interest Prospective studies demonstrate that persistent cervical
for defining the type-specific antigen composition of HPV infection by high-risk HPVs precedes the appearance of the
vaccines for the prevention of cervical cancer. CIN and is required for the development, maintenance and
Case-control studies of invasive cervical carcinoma yield progression of these lesions. Epidemiological studies in multi-
relative risks (multiplying factor of the probability of developing ple populations show consistently that the bulk of the infections
disease on being exposed over the probability of developing by HPV always precedes the bulk of the neoplasias by one or
disease on not being exposed) of between 50 and 100 for the two decades.
DNA detection of HPV in general and risks (“odds ratios”) of
between 100 and 500 for the HPVs 16 and 18. In some studies New Options in the Prevention and Early Detection of
these figures reach values of between 500 and 1000. In Fig. 2 the Cervical Cancer
estimations of the odds ratios are presented for the 15 most
frequent viral types in invasive cervical carcinoma. The description of the viral origin of cervical cancer and
The estimation of the risk for the 15 most frequent types the refinement of techniques of clinical diagnosis has opened
reinforces the concept that in screening programs, the assess- new and interesting options to improve screening programs.
ment of type-specific positivity for a cocktail of all of them is One of the first evaluated proposals has been the use of
sufficiently discriminative of the risk (high/low) and potential HPV DNA detection assays in the triage of women with
for progression to lesions, being therefore useful for the cervical abnormalities. In this approach, HPV DNA detection
appropriate population control. These results do not suggest is used as a prognostic marker in cases of ambiguous
the requirement of differentiated specific clinical protocols for cytology results (ASCUS, CIN1, mild discariosis…) The
women with infections of the HPV 16 type or any other of the conclusions of these studies that include among other a large
high risk viral types that are rather infrequent [14]. controlled randomized trial, indicate that the viral detection
The fractions of cervical cancer attributable to HPV in cases of ASCUS predicts the co-existence of high-grade
(proportion of cases in a population in which HPV is considered lesions with greater sensitivity and better cost-benefit ratio
as a causal agent) calculated starting from these studies, are very than the repetition of the cytology or even than the immediate
X. Castellsagué / Gynecologic Oncology 110 (2008) S4–S7 S7

colposcopy with or without directed biopsy. These conclusions [3] Castellsague X, Bosch FX, Munoz N, Meijer CJ, Shah KV, de Sanjose S, et
are being evaluated by the medical societies for adaptation and al. Male circumcision, penile human papillomavirus infection, and cervical
cancer in female partners. N Engl J Med 2002;346:1105–12.
adoption into routine clinical protocols. [4] Bleeker MC, Hogewoning CJ, Voorhorst FJ, van Den Brule AJ, Snijders
In women older than 30–35 years, viral detection is being PJ, Starink TM, et al. Condom use promotes regression of human
evaluated as the primary screening test associated with cytology papillomavirus-associated penile lesions in male sexual partners of women
in countries with established screening programs, or as primary with cervical intraepithelial neoplasia. Int J Cancer 2003;107:804–10.
[5] Hogewoning CJ, Bleeker MC, van Den Brule AJ, Voorhorst FJ, Snijders
test in populations where the cytological screening programs are
PJ, Berkhof J, et al. Condom use promotes regression of cervical
lacking, in which case the cytology or the biopsy are considered intraepithelial neoplasia and clearance of human papillomavirus: a
as secondary screening tests and confirmation of the lesion. In randomized clinical trial. Int J Cancer 2003;107:811–6.
all the cases, it has been demonstrated that the sensitivity of the [6] Woodman CB, Collins S, Winter H, Bailey A, Ellis J, Prior P, et al. Natural
viral detection is greater than that of the specialized cytology to history of cervical human papillomavirus infection in young women: a
detect prevalent lesions. The Food and Drug Administration, longitudinal cohort study. Lancet 2001;357(9271):1831–6.
[7] Elfgren K, Kalantari M, Moberger B, Hagmar B, Dillner J. A population-
FDA recognized in 2003 the value of the cytology associated based five-year follow-up study of cervical human papillomavirus
with the HPV test in women older than 30 years for the infection. Am J Obstet Gynecol 2000;183:561–7.
population from the United States. [8] Schlecht NF, Kulaga S, Robitaille J, Ferreira S, Santos M, Miyamura RA,
The development of prophylactic vaccines, either therapeu- et al. Persistent human papillomavirus infection as a predictor of cervical
tical or combined is a new option for the prevention of intraepithelial neoplasia. J Am Med Assoc 2001;286:3106–14.
[9] Castellsague X, Munoz N. Chapter 3: Cofactors in human papillomavirus
infections by HPV and maybe for the treatment of established carcinogenesis-role of parity, oral contraceptives, and tobacco smoking. J
infections. Some research lines exists that are evaluating new Natl Cancer Inst Monogr 2003:20–8.
molecules for the treatment of the infections by HPV and [10] Castellsague X, Bosch FX, Munoz N. Environmental co-factors in HPV
associated lesions, but the evidence is still limited. Some new carcinogenesis. Virus Res 2002;89:191–9.
inmunomodulators have shown effectiveness in the treatment of [11] Garcia-Closas R, Castellsague X, Bosch X, Gonzalez CA. The role of diet
and nutrition in cervical carcinogenesis: A review of recent evidence. Int J
condylomas and preparations are in the development phase Cancer 2005.
adapted to the treatment of infections on mucous surfaces. [12] Smith JS, Bosetti C, Munoz N, Herrero R, Bosch FX, Eluf-Neto J, et
On the other hand, two prophylactic vaccines, Cervarix and al. Chlamydia trachomatis and invasive cervical cancer: a pooled
Gardasil, are already in a very advanced phase of implementa- analysis of the IARC multicentric case-control study. Int J Cancer
tion at the world level, and both vaccines have demonstrated not 2004;111:431–9.
[13] Smith JS, Herrero R, Bosetti C, Munoz N, Bosch FX, Eluf-Neto J, et al.
only safety and immunogenicity but also efficacy for the Herpes simplex virus-2 as a human papillomavirus cofactor in the etiology
prevention of cervical neoplastic lesions as well as vaginal, of invasive cervical cancer. J Natl Cancer Inst 2002;94:1604–13.
vulval and genital warts for Gardasil. [14] Munoz N, Bosch FX, de Sanjose S, Herrero R, Castellsague X, Shah KV, et
al. Epidemiologic classification of human papillomavirus types associated
with cervical cancer. N Engl J Med 2003;348:518–27.
Conflict of interest statement
XC has received research grants from GlaxoSmithKline, Merck Sharp & [15] Bosch FX, Manos MM, Muñoz N, Sherman M, Jansen A, Peto J, et al.
Dohme and Sanofi Pasteur MSD, served on Speakers' Bureaus for Prevalence of human papillomavirus in cervical cancer: a worldwide
GlaxoSmithKline and Sanofi Pasteur MSD, and Steering Committees for perspective. J Natl Cancer Inst 1995;87:796–802.
[16] Walboomers JM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah
GlaxoSmithKline and Sanofi Pasteur MSD.
KV, et al. Human papillomavirus is a necessary cause of invasive cervical
cancer worldwide. J Pathol 1999;189:12–9.
[17] Muñoz N, Bosch FX, Castellsague X, Diaz M, de Sanjose S,
References Hammouda D, et al. Against which human papillomavirus types shall
we vaccinate and screen? The international perspective. Int J Cancer
[1] Bosch FX, Lorincz A, Muñoz N, Meijer CJLM, Shah KV. The causal 2004;111:278–85.
relation between human papillomavirus and cervical cancer. J Clin Pathol [18] Smith JS, Lindsay L, Hoots B, Keys J, Franceschi S, Winer R, et al.
2002;55:244–65. Human papillomavirus type distribution in invasive cervical cancer
[2] Castellsague X, Bosch FX, Munoz N. The male role in cervical cancer. and high-grade cervical lesions: a meta-analysis update. Int JCancer
Salud Publica Mex 2003;45(Suppl 3):S345–53. 2007;121:621–32.