Matos, Elena PhD*; Loria, Dora PhD*; Amestoy, Gustavo M. MD, PhD† a; Herrera, Lily MD‡; Prince, Miguel A. MD‡; Moreno, Juan MD‡; Krunfly, Cristina MD‡; van den Brule, A. J. C. PhD§; Meijer, Chris J. L. M. MD§; Muñoz, Nubia MD, MPH∥; Herrero, Rolando MD, PhD∥; Proyecto Concordia Collaborative Group
CERVICAL CANCER is one of the leading cancers among women worldwide, constituting an important public health problem. 1 Developed countries have experienced important reductions in the incidence of and mortality associated with cervical cancer over the past several decades, but it remains the most common cancer among women in many developing countries. 2
In Argentina, where breast cancer and cancer of the uterus are the first and second most frequent causes of cancer death among women, 3 age-adjusted mortality rates associated with cervical cancer are higher in provinces with a lower socioeconomic level. 4
It is now clear that certain types of human papillomavirus (HPV) are the central cause of cervical cancer and its precursor lesions. 5,6 HPV DNA is detectable in the majority of cervical intraepithelial neoplasia and cancer, with attributable fractions >80%, 7 and in some studies of invasive cervical cancer HPV DNA has been detected in virtually all cases. 8 It has been proposed that infections with certain HPV types, called oncogenic or high-risk (mainly types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66 and 68), are more likely to progress to cancer than infections with other HPV types.
Current research is oriented toward finding determinants of persistent infection with HPV and progression from infection to CIN, 9,10 as well as developing appropriate vaccines against HPV, 11 which appear to be the most promising strategies for cervical cancer prevention in the future.
Due to the limited population-based information on the age-specific prevalence of HPV infection among women, 7,12 the International Agency for Research on Cancer (IARC) is coordinating a series of population-based prevalence surveys of type-specific HPV infections in preparation for vaccine trials and programs. The studies are being carried out in areas of the world with high and low incidences of cervical cancer.
We report the results of a population-based study in the city of Concordia in the Province of Entre Rios, Argentina. The age-specific prevalence and determinants of HPV detection were evaluated in a random population sample of 1028 women. A PCR-based assay capable of detecting over 35 HPV types was used. The population-based nature of this study provided previously unavailable estimates of the prevalence of the full spectrum of HPV infections in this area of the world.
Materials and Methods
This study was conducted in Concordia, a town located in the province of Entre Ríos with a population of about 140,000 inhabitants and a high incidence of cervical cancer. The annual age-standardized incidence rate (ASIR) of invasive cervical cancer, reported by the Tumor Registry of Concordia, was 32.0 per 100,000 women for the period 1990 to 1994. 13
The National Institute of Statistics and Census (INDEC) selected for this study a random subsample of 1786 households from a larger random sample of 7189 households used in different surveys performed over the country. The sample was stratified by socioeconomic level into six categories, characterized by the occupation of the head of the household and the type of house.
A mass media campaign was performed through local newspapers, journals, and television and radio programs before and during the survey. Four female interviewers visited the selected households to invite all resident women aged ≥15 years to participate in the study.
Ten percent of the 1786 households selected could not be contacted, and 30% refused participation, leaving 1118 participating households. In 1052 of them there were 1473 women fulfilling the inclusion criteria, and 1028 of them (70%) agreed to participate, for an overall participation rate of about 50%. Exclusion criteria for participation were current pregnancy, previous hysterectomy or conization, and mental incompetence.
The number of women included in the current study, by age category, is shown in Table 1. The study was carried out from August 24 to December 7, 1998, and women aged ≥15 years who agreed to participate were seen at the hospital or at a clinic facility of the Red Cross for an interview and a gynecologic examination, including a Papanicolaou smear and colposcopy. No clinical examinations were done at home.
Participants were asked to sign an informed consent including details about the objectives of the study, benefits and risks associated with participation, and assurance of the confidentiality of the information provided. Data collected in the interview by trained female interviewers were on demographic characteristics, sexual and reproductive behavior, contraceptive practices, and medical and smoking histories. The questionnaires were checked and coded by the investigators in Buenos Aires (EM, DL) and returned to the interviewers for data retrieval when inconsistencies were detected.
Clinical Examination and Specimen Collection
Twenty gynecologists were involved in pelvic and colposcopic examinations. All the gynecologists were trained in special sessions by the gynecologic coordinator (G. A.) and examined patients with a fixed weekly schedule.
After visual inspection of the vulva, a nonlubricated sterile speculum was inserted, and exfoliated cells were collected by firmly rotating a wooden Ayre spatula on the cervical external os and a cytobrush in the endocervix. Immediately, a conventional Papanicolaou smear specimen was obtained and fixed in ethanol. An additional specimen of exfoliated cells was collected with a second Ayre spatula by the same technique. Both spatulas and the cytobrush were eluted in 20 ml of PBS solution. The examination of participants who reported not having initiated sexual activity included only an external pelvic examination and a vaginal swab sampling with a sterile cotton swab for a Papanicolaou smear and cells for HPV testing. In addition, 10 ml of blood was collected from all participants for immunologic studies.
The cervical exfoliated cell pellets were frozen at −20°C after centrifugation at 3000 g. Aliquots of plasma, buffy coat, and red blood cells were prepared and stored at −20°C after blood centrifugation at 1500 g. All the samples were stored in prelabeled vials. All the biologic specimens were sent to IARC, in Lyon, France, on dry ice for storage, where immunologic, nutritional, and genetic studies were performed.
The Papanicolaou smears were stained and interpreted locally by a team of cytopathologists, coordinated by L. H., and the results were classified according to the Bethesda System.
HPV testing was performed on exfoliated cervical cells from 1022 women, after exclusion of those with inadequate specimens or without specimens at all. The cell pellets were resuspended in 1 ml of 10 mmol/l Tris buffer, and 10 μl was used for PCR as described previously. 14 HPV DNA detection was performed with the HPV GP5+/6+ primer-mediated PCR enzyme immunoassay (EIA), on a small fragment of the L1 gene, in combination with HPV type-specific oligoprobes. To analyze the quality of target DNA for PCR purposes, cervical specimens were tested by PCR with β-globin gene-specific primers. 14 Cervical specimens were subject to HPV DNA genotyping, which was carried out as follows.
Briefly, a first screening was performed to determine the overall presence of HPV with a general primer GP5+/6+-mediated PCR, which permits the detection of a broad spectrum of sequenced and still-unsequenced genital HPV types at the subpicogram level. 15 HPV positivity was assessed by EIA with a high-risk (HR) HPV oligoprobe cocktail (14 HR types: HPV 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68) and a low-risk (LR) HPV oligoprobe cocktail (23 LR types: HPV 6, 11, 26, 34, 40, 42, 43, 44, 53, 54, 55, 57, 61, 70, 71 [equivalent to CP8061], 72, 73, 81 [equivalent to CP8304], 82/MM4, 82/IS39, 83 [equivalent to MM7], 84 [equivalent to MM8] and CP 6,108). 16 In addition, HPV positivity was assessed by low-stringency Southern blot analysis of the PCR product with a cocktail probe of HPV-specific DNA fragments. 17
Subsequently, GP5+/6+ PCR was repeated on positive samples in triplicate to generate sufficient products for further typing. After pooling of these PCR products, typing was performed with EIA and HPV type-specific oligoprobes for the HR and LR types described above. 16 Six samples were GP5/6+-positive by low-stringent Southern Blot analyses but could not be identified by the above-mentioned HR/LR EIAs and were considered as HPV X, uncharacterized HPV types. We took special precautions to minimize false-positive results in the PCR, as has been described in detail elsewhere. 17 Subjects whose samples were negative for the β-globin probe were excluded from the HPV analysis (n = 35).
In this analysis, the group of a priori high-risk HPV types includes 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, 66, and 68. 7,16 The group of “non-cancer-associated” HPV types includes all other HPV types tested. Some of them are recognized non-cancer-associated types (e.g., HPV types 6, 11, 40, 42, 43, and 44), and others are HPV types with undetermined oncogenic potential (e.g., HPV types 26, 34, 53, 54, 70, 72, and 73).
Most of the analyses were carried out by age group. Binomial 95% confidence intervals for HPV prevalence by age group were calculated.
Continuous variables were categorized, and the risk associated with HPV positivity was calculated for each category of a variable, adjusted by age. Socioeconomic level was analyzed according to the INDEC strata and the level of education achieved. Multivariate logistic regression analysis, in order to identify the independent variables that influence the relative risk of HPV positivity, was performed with stratification by age group.
Odds ratios with their 95% confidence intervals were calculated. When appropriate, a linear trend test was used. STATA software for Windows 18 (STATA, Chicago, IL) was used for statistical analysis.
Characteristics of the Population
The distribution of households by socioeconomic level in the final sample was similar to the distribution in the original sample (data not shown).
A total of 1028 women were included. Table 1 shows some of the characteristics of the population studied. Almost 50% of the women were ranked in the two lowest categories of socioeconomic level. Most women were currently married, and 6% reported being divorced. More than 50% reached high school (studied for more than 7 years). Sixty women (8%) had never had sexual intercourse, and gynecological examination confirmed a conserved hymen. The age at first sexual intercourse was <17 years for 30% of the women, and 60% of the sexually active women reported only one lifetime sexual partner. The median number of pregnancies was three. Forty percent of women reported four or more pregnancies.
Oral contraceptives were ever used by 46% of the women and condoms by 42% of them. About 36% of the women reported ever having smoked, and 23% were current smokers. Histories of sexually transmitted diseases were rare; 2.5% reported genital herpes, 1.8% condyloma, 0.9% syphilis, and 0.6% gonorrhea. No women reported having AIDS or having had a partner with AIDS.
Young women tended to be more educated than older women (Table 2) and to report much earlier initiation of sexual activity. Women in the age group of 25 to 34 years had more partners than their older counterparts, and there was a clear trend of decreasing age at first pregnancy with decreasing age; more than 90% of women younger than age 25 years reported a first pregnancy before age 21 years, versus only 25% of women aged 55 years or older.
PCR results were available for 987 women: HPV was detected in 166 (16.6%;Table 3). Of them, 90 had an infection with a single HPV type (9.0%, or 54.2% of the positives), while 76 had multiple HPV types detected (7.6%, or 45.8% of the positives). The prevalence of cancer-associated HPV types among sexually active women was 12.1% (68.4% of all infections). The most common type was HPV 16 (4.0%), followed by HPV 35 (2.2%), HPV18 (1.9%), and other high-risk types (including those detected in single or multiple infections).
Prevalence of HPV DNA among the 60 virgins tested was 3% (n = 2). One girl aged 16 years was positive for HPV 16, and a second female, 26 years old, had a double infection with HPV 31 and 58; for both, cytologic findings were normal. For sexually active women, positivity for HPV DNA of any type peaked among those <25 years of age, with a prevalence of 25.2%, and decreased consistently with age, to a minimum of 9.1% in the age group of ≥65 years (Figure 1). The age-related decrease was driven mainly by the cancer-associated HPV types, because non-cancer-associated types had a constant prevalence by age group of around 5% (Figure 2).
The main difference regarding distribution of types between women <45 and ≥45 years of age was the prevalence of HPV 16, which was clearly predominant among the former, while among the latter, the prevalence of HPV 16 was very similar to the prevalence of all the other types (Figure 2). A similar but less marked pattern was observed for HPV 35, while HPV 18 was equally common among women of all ages and HPV 31 predominated in young women but was less frequent than most types among older women.
Risk Factors for HPV Detection
Table 4 shows the results of the multivariate logistic regression models of risk of prevalent HPV infection for all ages combined and for women <45 and ≥45 years of age. Education was not significantly associated with detection of HPV in the multivariate model when all ages combined were considered, but higher education was significantly associated with lower risk among women <45 years of age. For women aged ≥45 years, a higher but not significant risk was observed among those who reached the university level of education.
The lifetime number of sex partners was significantly associated with detection of HPV: women reporting three or more partners were three times more likely to carry HPV than were monogamous women. This association was even stronger among women <45 years of age, with risk increasing to 4.4 for women reporting three or more partners. However, no increase in detection of HPV was observed for number of sex partners among women ≥45 years of age.
Use of oral contraceptives in this dataset was associated with a significant two-fold reduction in HPV detection among current users. This association was present only among younger women, but as expected, there were very few current users among older women. Among women <45 years of age, the reduction in risk of HPV detection was maximal for users of oral contraceptives for 3 to 5 years, but it was of a lesser magnitude for longer-term users.
Severe vaginal discharge was associated with a twofold to threefold increase in detection of HPV in both age groups, although the association was significant only in the combined group and among women <45 years of age.
As expected, cytologic diagnoses of HSIL or invasive cancer were associated with increased HPV detection (data not shown). Information about the number of partners in the last year was not available from 66 women from the pilot study, and relatively few women (n = 14) reported more than two partners in the last year. No significant association with HPV detection was detected for this variable.
In the studied women, HPV DNA detection was not associated with marital status, smoking, age at first intercourse, condom use, visits to prostitutes by the partner, age at first pregnancy, parity, or inflammation as reported for the Papanicolaou smear.
To our knowledge, this is the first study to investigate the prevalence of type-specific HPV infection in a population-based sample of Argentine women. The area investigated, Concordia, is an urban community in the province of Entre Ríos, with a predominantly low- or middle-income population. Most women report monogamy (60%) and high parity (40%, more than four pregnancies), and contrary to other Latin American populations, a relatively high proportion of women report smoking (36% ever smokers and 23% current smokers).
The age distribution of HPV infection needs to be studied in the context of changing cultural patterns, and this is a population with a clearly changing sexual and reproductive behavior. Younger women are now more educated and tend to start sexual activity and have children much earlier than their older counterparts.
Overall HPV prevalence was around 16%, a figure that is similar to the prevalence observed with use of comparable detection methods in other areas with similar incidences of cervical cancer, such as Guanacaste, in Costa Rica, 7 or Morelos, in Mexico. 12
In this study, we had the opportunity to test specimens from women reporting no previous sexual intercourse, as well as from sexually active women of all ages. In concordance with the notion of HPV being a sexually transmitted disease, 19 detection of the virus was very uncommon among virgins (3%). If their classification as not sexually active is correct, they could have acquired these infections through noncoital sexual activity. Vertical transmission from mother to child has been reported but very rarely accounts for any significant proportion of HPV infections. 20,21 There is also a theoretical possibility of infection via fomites. 22
As in most populations investigated, 7,12,16,23,24 we observed that the prevalence of HPV infection was highest among women <25 years of age. This can be considered further indication of the sexual transmission of HPV, because women in this group are closer to initiation of sexual activity.
In Concordia, prevalence of HPV decreases consistently with age to its lowest level among women ≥65 years of age, supporting the common notion that initial infections tend to be cured and women possibly develop immunity to individual HPV types over time. However, this is in contrast with the findings of some recent studies, 7,12 in which a second peak of positivity was observed among women after menopause. We hypothesize that the low prevalence of HPV detection among older women in Concordia can be related to the recent spread of HPV infection in this population, as evidenced by the marked changes in sexual behavior reported by the women. In other words, older women in Concordia were less exposed to HPV when they were young, and therefore we would not expect, in a cross-sectional study, a second peak of infection, even if there is reactivation of past infections.
The predominant HPV types at all ages are cancer-associated ones, and the decline observed with age is mainly driven by declines in cancer-associated types. Non-cancer-associated types tend to have a constant prevalence in all age groups. Thus, the relative proportion of infections with non-cancer-associated HPV infections increases with age, as has been described in other studies.
Prevalence of multiple infections was 7.7%, corresponding to 46% of infections—higher than the 2.3% reported in Mexico 12 but similar to the prevalence in Guanacaste, where 39% of infections involved multiple types when women with normal diagnoses were analyzed. Most studies to date have revealed a high frequency of multiple infections, particularly when the specimens analyzed were exfoliated cervical cells.
Among the HPV types detected there was a predominance of HPV 16, with a prevalence of 2.5% in single infections and 1.5% in multiple infections, for a total of 4.0%, corresponding to 22% of all infections. The second most common type was HPV 35, which was detectable alone or in combination in 2.2%, corresponding to 13.3% of all infections. In general, the most common types in the population are also the most common types detected in cervical lesions in most areas of the world, and this is confirmed in the few high-grade lesions and cancers detected in this population (data not shown). Notably, HPV 35 has recently been reported as the most common HPV type in the population and in cervical neoplasia in Mozambique. 25
We analyzed the prevalence of type-specific HPV infection by age group, and a pattern emerged in women younger than 45 years of age in comparison with older women: there was a relatively higher prevalence of HPV types 16, 31, and 35 but not other types, providing indirect evidence that infections with those HPV types are less likely to be persistent at older ages in the absence of disease.
Risk factors for HPV detection in a multivariate model were clearly delineated among women <45 years of age, while none of the factors investigated were associated with HPV detection among older women, suggesting that detection of HPV may be related in the older age groups to other, yet-unidentified aspects of the immune or endocrine system.
Among the risk factors in young women was education, with women who had a university education having a twofold reduction in HPV detection. This may be an indicator of characteristics of the partners or other aspects of sexual hygiene of the more-educated women.
A clear trend of increasing detection of HPV was associated with increasing number of sex partners among women <45 years of age, in concordance with the sexual transmission of the virus. Similarly, risk was increased among women reporting vaginal discharge, with an increase in HPV detection associated with the severity of the discharge. Vaginal discharge may be an additional marker of sexual activity or may indicate the possibility of an interaction between the virus and other infections in the cervical milieu, which may favor viral persistence.
The reduced detection of HPV among users of oral contraceptives is intriguing and may be related to chance or confounding, but this deserves further investigation.
Although this study was population-based and women were not selected from among those attending specific gynecologic services, limited participation rates hampered the study. However, the distribution by socioeconomic level of the households included in the final sample was similar to the distribution in the initial list of households, indicating a low potential for selection bias.
HPV infection is relatively common in this population, and there are indications that changing sexual behavior patterns are producing a tendency for HPV prevalence to increase. This, in addition to relatively high smoking rates, a known cofactor of HPV infection for progression to high-grade cervical neoplasia and cancer, 26,27 could harbinger future increases in cervical cancer incidence, unless preventive efforts are implemented.
1. Parkin DM, Whelan SL, Ferlay J, Raymond L, Young J. Cancer Incidence in five Continents. Vol 2. Lyon, France: International Agency for Research on Cancer, 1997.
2. Beral V, Hermon C, Muñoz N, Devesa S. Cervical cancer. Cancer Surveys 1994; 19/20: 265–285.
3. Matos E, Loria D, Vilensky M, García C. Cancer Mortality Atlas: Argentina 1989-1992. Buenos Aires: Comité Argentino de Coordinación, Programa Latinoamérica contra el Cáncer, 1997.
4. Matos E, Loria D, Vilensky M. Cancer mortality and poverty in Argentina: a geographic correlation study. Cancer Epidemiology Biomarkers and Prevention 1994; 3: 213–218.
5. Human Papillomaviruses (IARC Monogr Eval Carcinog Risks Hum 64). Lyon, France: International Agency for Research on Cancer, 1995.
6. Muñoz N, Bosch FX, Shah KV, Meheus A, eds. The Epidemiology of Cervical Cancer and Human Papillomavirus. Lyon, France: International Agency for Research on Cancer, 1995.
7. Herrero R, Hildesheim A, Bratti C, Sherman ME, Hutchinson M, Morales J, et al. Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa Rica. J Natl Cancer Inst 2000; 92: 464–474.
8. Walboomers JMM, Jacobs MV, Manos MM, Bosch FX, Kummer JA, Shah KV. J Pathol 1999; 189: 12–19.
9. Nobbenhuis MA, Walboomers JM, Helmerhorst TJ, et al. Relation of human papillomavirus status to cervical lesions and consequences for cervical cancer screening: a prospective study. Lancet 1999; 354: 20–25.
10. Franco EL, Villa LL, Sobrinho JP, et al. Epidemiology of acquisition and clearance of cervical human papillomavirus infection in women from a high-risk area for cervical cancer. J Infect Dis 1999; 180: 1415–1423.
11. Lowy DR, Schiller JT. Papillomaviruses and cervical cancer: pathogenesis and vaccine development. J Natl Cancer Inst Monographs 1998; 23: 27–30.
12. Lazcano-Ponce EC, Miquel JF, Muñoz N, et al. Epidemiology and molecular pathology of gallbladder cancer. Cancer 2001; 51: 349–364.
13. Prince MA, Loria D, Herrera L, Matos E. Argentina: Concordia. In: Parkin DM, Whelan SL, Ferlay J, Raymond L, Young J, eds. Cancer Incidence in Five Continents. Vol. 2. Lyon, France: International Agency for Research on Cancer, 1997; 1997: 94–97.
14. Jacobs MV, Snijders PJ, van den Brule AJ, Helmerhorst TJ, Meijer CJ, Walboomers JM. A general primer GP5+/GP6(+)-mediated PCR-enzyme immunoassay method for rapid detection of 14 high-risk and 6 low-risk human papillomavirus genotypes in cervical scrapings. J Clin Microbiol 1997; 35: 791–795.
15. Jacobs MV, de Roda Husman AM, van den Brule AJC, Snijders PJF, Meijer CJLM, Walboomers JWW. Group-specific differentiation between high- and low-risk human papillomavirus genotypes by general primer-mediated PC and two cocktails of oligonucleotide probes. J Clin Microbiol 1995; 33: 901–905.
16. Jacobs MV, Walboomers JM, Snijder PJ, et al. Distribution of 37 mucosotropic HPV types in women with cytologically normal cervical smears: the age-related patterns for high-risk and low-risk types. Int J Cancer 2000; 87: 221–227.
17. Walboomers JWW, Melkert P, van den Brule AJ Snijders PJF, Meijer CJLM. The polymerase chain reaction for human papillomavirus screening in diagnostic cytopathology of the cervix. In: Herrington CS, McGee O, eds. Diagnostic molecular pathology: a practical approach. Oxford: University Press Oxford, 1992: 152–172.
18. Stata Corporation. Stata Statistical Software: Release 5.0. College Station, Texas: Stata Corporation, 1997.
19. Kjaer SK, Chackerian B, van den Brule AJ, et al. High-risk human papillomavirus is sexually transmitted: evidence from a follow-up study of virgins starting sexual activity (intercourse). Cancer Epidemiol Biomarkers Prev 2001; 10: 101–106.
20. Watts DH, Koutsky LA, Holmes KK, et al. Low risk of perinatal transmission of human papillomavirus: results from a prospective cohort study. Am J Epidemiol 1998; 178: 365–373.
21. Dillner J, Andersson-Ellstrom A, Hagmar B, Schiller J. High risk genital papillomavirus infections are not spread vertically. Rev Med Virol 1999; 9: 23–29.
22. Roden RB, Lowy DR, Schiller JT. Papillomavirus is resistant to dessication. J Infect Dis 1997; 176: 1076–1079.
23. Melkert PW, Hopman E, van den Brule AJ, et al. Prevalence of HPV in cytomorphologically normal cervical smears, as determined by the polymerase chain reaction, is age-dependent. Int J Cancer 1993; 53: 919–923.
24. Burk RD, Kelly P, Feldman J, et al. Declining prevalence of cervicovaginal human papillomavirus infection with age is independent of other risk factors. Sex Transm Dis 1996; 23: 333–341.
25. Castellsagué X, Menéndez C, Loscertales MP, et al. Human papillomavirus genotypes in rural Mozambique. Lancet 2001; 358: 1429–1430.
26. Hildesheim A, Herrero R, Castle PE, et al. HPV co-factors related to the development of cervical cancer: results from a population-based study in Costa Rica. Br J Cancer 2001: 1–8.
27. Deacon JM, Evans CD, Yule R, et al. Sexual bahaviour and smoking as determinants of cervical HPV infection and of CIN3 among those infected: a case-control study nested within the Manchester cohort. Br J Cancer 2000; 83: 1565–1572.