Human papillomavirus cervical infection and associated risk factors in a region of Argentina with a high incidence of cervical carcinoma.

OBJECTIVE: To assess the prevalence and potential risk factors associated with human papillomavirus (HPV) cervical infection among women residing in a region of northeastern Argentina with a high incidence of cervical cancer. METHODS: A case-control study of 330 women participating in a cervical cytological screening program conducted in Posadas city, Misiones, Argentina, from February 1997 to November 1998 was carried out. Standardized questionnaires were administered, and clinical examination including colposcopy was performed. Fresh endocervical specimens for HPV DNA detection by generic polymerase chain reaction were collected and the products typed by dot-blot hybridization. RESULTS: Human papillomavirus DNA was found in 61% of samples analyzed (185/301). Samples with normal cytology had a 43% infection rate (85/199), while those classified as low-grade squamous intraepithelial lesion, high-grade squamous intraepithelial lesion, and invasive cervical carcinoma had an infection rate of 96% (53/55), 100% (29/29), and 100% (18/18), respectively. Human papillomavirus typing showed a 64% (118/185) prevalence of type 16 among all the infected population analyzed; type 16 was detected among 49% (42/85) of infected samples with normal cytology and in an average of 74% (74/100) with abnormal cytology. Sexual behavior, residence in southern Paraguay, and history of a previous sexually transmitted diseases were the main risk factors associated with high-grade cervical lesions. CONCLUSIONS: An elevated prevalence of HPV infection was detected in this population, which also has a high incidence of cervical cancer. The broad distribution of high-risk HPV type 16 in women with normal cytology and colposcopy suggests that viral infection is an important determinant of regional cancer incidence.

Different HPV types have been detected in more than 95% of cases of invasive cervical cancer (ICC) and in varying percentages in cases of lowgrade squamous intraepithelial lesion (LSIL) and high-grade squamous intraepithelial lesion (HSIL). 3-Studies examining risk factors for HPV infection have not been entirely consistent, 9,1 probably due to differences in the studied populations and the detection methods used. It would be of importance to focus on particular characteristics of the population under study, such as demographic growth rate, regional human displacement around the area under analysis, and sexual habits, as well as overall HPV prevalence and presence of specific HPV types.
In Argentina, there are still no complete epidemiologic data to state the actual, global situation regarding HPV infection and its association with cervical neoplasia. Nevertheless, regional studies have shown that cervical cancer incidence varies widely, with values that range from 1-3 cases per 10,000 women residing in metropolitan areas to 4 cases per 1,000 women in peripheral regions. 11 The Province of Misiones is situated at the northeastern tip of Argentina and shares international borders with Paraguay and Brazil. The construction in the last ten years of bridges and roadways and the creation of a common market including these three countries have promoted an intense population migration along the entire territory. This new situation has changed several cultural habits.
Since the implementation of a program of cervical cancer screening by cytology in 1995 in Misiones, an elevated incidence of 3-4 cases of ICC per 1,000 women has been reported consistently, together with high incidences of LSIL (16 per 1,000) and HSIL (11 per 1,000).
These values have defined an area of high cancer incidence where some particular factors probably are involved, lz, 13 The aim of this study was to assess the HPV prevalence and potential risk factors associated with viral cervical infection among women from this region.

Study Population
Three hundred and thirty women were recruited from 5,676 participants attending either the Cen-tral Regional Hospital, two family practice clinics in Posadas, (Misiones, Argentina) or a clinic in Encarnacion (Paraguay) as part of an ongoing cervical cancer screening program from February 1997 to November 1998. A written informed consent form was signed by all participants, and a standardized, self-administered questionnaire concerning sociodemographic factors, sexual and reproductive history, use of birth control, and history of other sexually transmitted diseases was completed. All women underwent a standard gynecologic examination and colposcopy, during which material for Papanicolaou smears was collected with cytobrush. To minimize the bias in sample collection and to ensure a sufficient amount of exfoliated cells, sampling was performed only if the last menstruation had occurred at least one week before the exam. The Pap test diagnosis was used to classify subjects as controls (normal changes) or cases (LSIL, HSIL, or ICC). 4 The case group included all tested women with abnormal cytology and/or abnormal colposcopy ranging from LSIL to ICC. Fifty nine participants were diagnosed with LSIL, 33 with HSIL, and 18 with ICC (all of the squamous cell type).
For each case, two population-based controls were selected and matched for age of subject and time when the sample was collected. This control group included 220 women with normal colposcopy and cytology and no history of sexually transmitted diseases or HPV lesions. For the case-control trios above 65 years old, an age difference of 5 years was tolerated for matching purposes. Biopsies were taken from abnormal areas and under direct colposcopic vision when necessary. Histopathologic examination was carried out, and the final diagnosis was based on the worst morphologic picture.

Human Papillomavirus DNA Analysis
Human papillomavirus detection samples were obtained from the ecto-and endocervix using cytobrush and processed the same day of collection. Cytobrushes were washed with sterile phosphatebuffered saline solution, pH 7.4, and collected cells lysed adding 200 pg/mL of proteinase K in the presence of 0.5% SDS in a total reaction volume of 500 pL of cellular suspension and incubated for hour at 55C. Nucleic acid was extracted as described previously. 5  polymerase chain reaction (PCR) for betaglobin gene to confirm the presence of an adequate and amplifiable template in the samples. 16 The DNA was amplified with the general consensus primers GP5+ and GP6+, targeting a region of approximately 140 base pairs in the highly conserved L10RF of the HPV viral genome. 17 DNA from CaSki and HeLa cells were used as positive controls; negative controls included no template DNA and were included after each tenth specimen in all tests.
The PCR reactions were carried out in 50 pL of solution containing 10 pL of template DNA, 10 mM Tris HC1 (pH 8.3), 50 mM KC1, 3 mM MgC12, 100 pM dATP, dCTP, dTTP and dGTP, 10 pmol of each primer, and 2 U of AmpliTaq TM (5 U/pL; Perkin Elmer, New Jersey). A hot start technique was used; the thermal cycling conditions were 94C for 3 minutes, followed by 40 cycles at 93C for 30 second, 45C for 30 seconds, and 72C for 30 seconds.
The amplification products were analyzed on 1.5% agarose gels, visualized with ethidium bromide staining and ultraviolet light exposure, and photographed. A single band of approximately 140 base pairs indicated a positive sample.

Dot-Blot Hybridization
Thirty-microliter aliquots resulting from positive HPV generic PCR were denatured with 20 pL of IN NaOH and dot-blotted to Zeta TM probe membranes (BioRad, California). Positive and negative controls were included in each assay. Biotinlabeled oligonucleotides type-specific probes for HPV types 6, 11, 16, and 18 were used. 18 Hybridization was performed at 55C for 3 hours. Filters were washed twice for 10 minutes at 56C for HPV types 6 and 11 and at 58C for HPV types 16 and 18. Positive hybridization dots were developed by chemoluminescence (PhotoGene TM nucleic acid detection system, BRL, Grand Island, NY). Specimens that showed a band of approximately 140 base pairs by ethidium bromide visualization but that could not be typed with the type-specific probes were referred to as "other HPV."

Statistical Analysis
The association between risk factors and the presence of HPV or lesion grade was assessed by univariate analysis, 19 yielding estimates of the odds Did not complete high school 36% 32% *Significant difference (P < 0.001). **Significant difference (P < 0.005). ratio (OR); 95% confidence intervals (CI) gauged the precision of the OR estimates. These analyses were carried out using the Epi Info program (Centers for Disease Control, Atlanta, GA).

RESULTS
The general characteristics of the study population at entry are summarized in Table 1  HPV typing results are presented in Table 3.
Infections with two viral types were detected in 17 women (20%) from the cases group; eight with LSIL cytology (15%) and nine with HSIL (31%).
Among the LSIL cases with mixed infections, five had HPV types 11 and 16, and three had HPV types 16 and 18. In those with HSIL cytology, five samples showed HPV 6 and 16, two showed HPV 11 and 16, and two showed HPV 16 and 18. The most prevalent type was HPV 16, which was detected in 49% (42/85) of infected samples from the control group (overall prevalence, 21%), in 77% of those with LSIL, in 72% of those with HSIL, and in 67% of those with ICC.
The association of HPV-16 infection and abnormal cytology is presented in Table 4. The association between selected variables (potential risk factors) of the cases group and abnormal cytology is presented in Table 5. Early age at first sexual intercourse, multiple lifetime number of sex partners, early age of first pregnancy, residence in southern Paraguay, and history of previous sexually transmitted diseases correlated significantly to HSIL and ICC lesions.
Univariate analysis of the same variables in HPV-positive subjects who were in the control group is shown in Table 6. Strong associations are

DISCUSSION
Cervical carcinoma is the leading cause of death by cancer in women in the Province of Misiones, Argentina, with an elevated incidence of 3 cases per 1,000. This value is particularly high compared with other regions of the country (1 case in 10,000 on average) where data are available. This situation prompted us to conduct a case-control study to investigate HPV infection in the female population as a condition necessary but not sufficient for progression to cervical neoplasia 3-8 and to analyze some regional factors probably associated with it. The study group was representative of the regional population and was characterized by married, middle-class, well-educated women.
In the analysis of HPV DNA present in cervical samples from the control group, we found an elevated percentage of viral infection (43%), showing predominantly HPV type 16. As the role of certain  HPV types (16,18,31,33, and 45) in cervical carcinogenesis becomes indisputable, the detection of these HPV types in cervical samples is being considered in conjunction with cytology z to improve diagnostic accuracy and efficiency.
A comparatively high level of HPV generic infection and HPV 16    HSIL (72.4%) is only comparable with that found by Sebbelov et al. in a study of a population at high risk for cancer in Greenland. z6 All ICC samples were HPV positive but 16% of them were infected with viral types other than HPV 6, 11, 16, or 18, suggesting that other carcinogenic types may be involved. Sexual behavior was associated with progression of high-degree lesions in infected subjects; permanent residence in Paraguay was also strongly associated with HSIL and ICC, probably due to the absence of regular Pap test screening programs in that country.
The elevated proportion of women infected with carcinogenic viral types and normal cytology is an important finding that must be considered as a relevant challenge. A follow-up study of this subclinical group may provide important data regarding regional characteristics of HPV genital infection and natural history of cervical cancer precursor lesions. Several particular characteristics of this regional population have been associated with either HPV infection (frequency of subject displacement between Argentina and Paraguay) or lesion severity (place of permanent residence). Implementing coordinated public health policies between Argentina and Paraguay, such as a simultaneous Pap test screening program in borderline cities, may help in lowering cervical cancer cases.