Worldwide, invasive cervical cancer (ICC) is the second most common cancer in women, with approximately 470,000 new cases and 230,000 deaths annually.1 Approximately 80% of ICC-related deaths occur in developing countries. The incidence of ICC is highest in Latin America, the Caribbean, Sub-Saharan Africa and South Asia and considerably lower in North America and Western Europe. In Vietnam, the ICC incidence rate is intermediate (17.3 per 100,000 women) but, on account of the relatively low incidence for all cancer (101.6/100,000), it is the most common cancer in women with approximately 5,600 new cases and 2,500 deaths per year.2 The incidence of ICC varies from 6.8/100,000 women in Hanoi, in the North, to 26/100,000 women in Ho Chi Minh City, in the South.2, 3
Infection with high-risk types of human papilloma virus (HPV) is established to be the necessary cause of ICC.4, 5 Recent efforts have focused on the development of both prophylactic and therapeutic HPV vaccines, chiefly targeting the most common oncogenic types HPV 16 and 18.6 An effective HPV vaccine could benefit populations worldwide, but data on the age- and type-specific distribution of HPV infection in women from different areas are needed for the planning of future vaccination strategies. Population-based information on the distribution of HPV types among women is still limited,7, 8, 9 particularly in Asia.10
Hungvuong Obstetric and Gynaecology Hospital in Ho Chi Minh City and the National Cancer Institute in Hanoi conducted 2 prevalence surveys of HPV infection in the urban district of Ho Chi Minh City and in a peri-urban district in Hanoi, Vietnam, in the framework of a multicentre study coordinated by the International Agency for Research on Cancer (IARC).10 Our study aims were to contribute to the knowledge on the prevalence of, and risk factors for, HPV cervical infection.
MATERIAL AND METHODS
Study population and enrollment
The 2 present HPV prevalence surveys were located in one area in South and one in North Vietnam. The list of all married women aged 15 years or older was obtained from family planning clinic records. A random sample of 120 married women was selected from each of 11 5-year age groups from 15–19 to 65 years or more. Women were eligible if they: (i) were not pregnant; (ii) had not undergone a hysterectomy or conization; (iii) did not suffer from mental impairment. In the South, 3 communes were chosen out of the 15 that constitute District 10, in the central part of Ho Chi Minh City. A sample of 1,279 eligible women was thus obtained. In the North, Socson, one of the 12 districts of the capital, Hanoi, was chosen. Socson is a peri-urban district located 40 km northeast of Hanoi. Four of 26 communes (Hongky, Kimlu, Phulo and Minhphu) were randomly selected. A sample of 1,320 eligible women was obtained.
In both study sites, health workers visited each selected woman at her home to (i) check exclusion criteria; (ii) explain study aims; and (iii) obtain a written consent form from all women willing to participate. The study protocol was reviewed and cleared by the Ethical Committee of the IARC, Lyon, and the Ministry of Health Board of Vietnam.
Overall, 1,246 women in Hanoi (94.4% of those contacted) and 1,122 from Ho Chi Minh City (87.7% of those contacted) came to the commune health centre for an interview and a gynaecological examination, which took place from April–May 1997 in Hanoi and from April–October 1997 in Ho Chi Minh City. Personal interviews were obtained using a structured questionnaire, which included information on sociodemographic variables and lifestyle factors, including smoking habits and reproductive and menstrual characteristics. Information on sexual habits, history of sexually transmitted diseases (STDs) and contraceptive use among participating women were collected.
Clinical examination and specimen collection
A pelvic examination was carried out by a gynaecologist. Samples of exfoliated cells from the ectocervix were collected with 2 wooden Ayre spatulas and from the endocervix with a cytobrush (Cervibrush, CellPath, Herte, United Kingdom). After the preparation of a Papanicolau (Pap) smear, the remaining exfoliated cervical cells were placed in tubes with phosphate-buffered saline (PBS). A sample of 10 ml of blood was collected from each consenting participant in a heparinized tube. On the day of collection, the specimens were sent to the laboratory of Hung Vuong Hospital, Ho Chi Minh City, and to the National Cancer Institute, Hanoi, for processing and storing.
The samples of exfoliated cervical cells were centrifuged at 3,000g and the resulting pellet diluted in PBS and poured into labeled nunc tubes. Blood samples were centrifuged at 1,500g for 10 min and separated into plasma, buffy coats and red blood cells that were distributed in labeled nunc tubes. All samples were stored at −30°C on dry ice until the shipping to IARC.
Pap smears were stained and read by locally trained cytopathologists and classified according to the Bethesda system. Women with abnormal cytologic findings underwent confirmatory biopsies and treatment, if necessary. One woman in Ho Chi Minh City and 72 in Hanoi (mostly elderly women) did not wish to have a Pap smear and were excluded from the following analyses.
HPV and HSV-2 detection techniques
HPV testing was performed on exfoliated cervical cells from 1,120 women in Ho Chi Minh City and 1,139 in Hanoi, after excluding those (1 and 35, respectively) for whom a specimen was not available. The cell pellets were resuspended in 1 ml of 10 PBS and 10 μl were used for PCR purposes as described previously.11 HPV DNA detection was performed using the HPV GP5+/6+ primer-mediated PCR enzyme immunoassay (EIA) targeting 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 using β-globin gene-specific primers.11 Women whose samples were negative for the β-globin probe were excluded from the HPV analysis (198 from Ho Chi Minh City and 145 from Hanoi).
Cervical specimens were subject to HPV DNA genotyping, which was carried out as follows: A first screening was performed to determine the overall presence of HPV DNA using 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.12 HPV positivity was assessed by enzyme immunoassay (EIA) using 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/IS39, 83 [equivalent to MM7], 84 [equivalent to MM8] and CP 6108).13 In addition, HPV positivity was assessed by low-stringency Southern blot analysis of PCR products with a cocktail probe of HPV-specific DNA fragments.14 Subsequently, GP5+/6+ PCR was repeated on positive samples in triplicate to generate sufficient products for further typing. After pooling these PCR products, typing was performed using EIA and HPV type-specific oligoprobes for the HR and LR types described above.12 Three samples from Ho Chi Minh City and 3 from Hanoi 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 (Table I). Special precautions were taken to minimize false-positive results in the PCR, as has been described in detail elsewhere.14
Table I. Prevalence of Type-Specific HPV Infection among Women in Ho Chi Minh City and Hanoi, Vietnam
The presence of type-specific plasma IgG antibodies against HSV-2 were tested in a central laboratory in Seattle, WA. All serologic testing on stored plasma was conducted blindly using an HSV-2 ELISA assay developed by Focus Technology/formerly MRL (Cypress, CA).15 All HSV-2-positive sera were retested to confirm results.
Age-standardized prevalence of HPV DNA detection were computed using, as standard, the female population of Ho Chi Minh City3 and Hanoi2 or the world standard population. Odds ratios (OR) for HPV DNA detection and corresponding 95% confidence intervals (CIs) were computed by means of unconditional multiple logistic regression equations separately for the 2 study areas. The variables that showed the strongest association with HPV DNA detection in the age-adjusted analysis in either Ho Chi Minh City or Hanoi were eventually fitted into the same model. This was done for each area and, on all women, separately for women <45 and ≥45 years, single and multiple HPV infections and high- and low-risk ones. Tests for linear trend of the ORs were performed, giving an increasing score for each level of the categorized variable and fitting them in the model as continuous variables.
HPV prevalence and type-specific distribution
A total of 922 women from Ho Chi Minh City (median age, 41 years) and 994 from Hanoi (median age, 45 years), for whom a cytologic smear and HPV testing were available, were included in the analyses that follow.
Table I presents the prevalence of HPV DNA detection overall and by HPV type by area. In Ho Chi Minh City, HPV DNA was detected in 101 cases (10.9%). Fifty-nine women (58.4% of HPV DNA-positive ones) were infected with a single HPV type and 42 (41.6%) with multiple types. Twenty-eight different HPV types were detected, the most common being HPV 16 (alone in 12 women and in multiple infections in 18 women; overall prevalence 3.3%; 29.7% of HPV infections). The next most common were HPV 58 (2 single and 12 multiple infections; overall prevalence 1.5%) and HPV type 18 (1 single and 10 multiple infections; overall prevalence 1.2%). The majority of single (75%) and multiple infections (88%) involved 1 or more high-risk HPV types.
In Hanoi, HPV DNA was detected in 20 cases (2.0%); 16 women (80.0% of HPV DNA-positive ones) were infected with a single HPV type and 4 (20.0%) with multiple types. Sixteen different HPV DNA types were detected. Each of HPV 16, 18 and 58 were found in 2 single infections. In 10 of 16 single and 3 of 4 multiple infections, HPV high-risk types were involved (Table I).
The majority of study women (90.1% in Ho Chi Minh City and 99.5% in Hanoi) had never had a Pap smear before participating in our surveys. The prevalence of abnormal cytologic findings was 15 of 922 (1.6%) in Ho Chi Minh City (including 2 women with cervical intraepithelial neoplasia (CIN)-3 and 5 with carcinoma in situ) and 7 of 994 (0.7%) in Hanoi (including one CIN-2). Ten of 15 (67%) and 1 of 7 (14%) women, respectively, with abnormal cytologic findings were positive for HPV DNA. If women with cytologic abnormalities had been excluded, overall HPV prevalence would have been 10.0% in Ho Chi Minh City and 1.9% in Hanoi.
Determinants of HPV infection
In Ho Chi Minh City, the prevalence of HPV DNA declined significantly with age from 22.3% <25 years to 10.9% at 25–34 and 8.4% at 55–64 years of age (Table II, Fig. 1). It was 9.4% at ≥65 years. In Hanoi, no clear age pattern emerged, but the highest HPV DNA prevalence (3.3%) was found in the age group 35–44 years (Table II). In Ho Chi Minh City, widows showed a lower HPV prevalence (OR = 0.5) than currently married women. Religion was not related to HPV DNA detection in either area. The prevalence of women with no education was lower in Ho Chi Minh City (8.6%) than in Hanoi (22.7%). Education tended to be directly associated with HPV DNA detection in both areas (ORs for secondary education vs. none: 1.7 in Ho Chi Minh City and 1.8 in Hanoi), although not significantly so. In Ho Chi Minh City, an OR of 2.8 among ever-smoking women (2.3%) was found. In Hanoi, smoking was reported by only 1 woman (Table II).
Table II. OR of HPV Infection and Corresponding 95% CI by Age and Selected Characteristics in Ho Chi Minh City and Hanoi, Vietnam
Age at menarche ≥17 vs. ≤16 years was associated with an OR of 0.6 in Ho Chi Minh City and 0.4 in Hanoi (Table III). Age at first pregnancy was not significantly associated with HPV DNA detection. Nulliparous women were significantly likelier to be HPV DNA-positive than women with 1 or 2 children (OR = 3.0 in Ho Chi Minh City and 6.0 in Hanoi), even after allowance for age in a single year (not shown). Conversely, the OR of HPV DNA detection did not change substantially according to the number of children among parous women (92.4% of women in Ho Chi Minh City and 98.0% in Hanoi). A history of spontaneous abortions was unrelated to HPV DNA, whereas a trend of increase in HPV DNA detection with the increase in the number of induced abortions was seen (OR for ≥2 vs. 0 = 1.8 in Ho Chi Minh City and 2.2 in Hanoi). Menopausal status was unrelated to HPV DNA detection in the age-adjusted model (Table III).
Table III. OR of HPV Infection and Corresponding 95% CI According to Menstrual and Reproductive Factors in Ho Chi Minh City and Hanoi, Vietnam
In Ho Chi Minh City, women who reported 2 or ≥3 sexual partners compared to 1 had ORs for HPV DNA detection of 2.0 (95% CI 0.9–4.5) and 4.0 (1.4–11.6), respectively (Table IV). A similar, although not significant, association was found in Hanoi (OR for ≥2 vs. 1 = 2.0). In Ho Chi Minh City, the inverse association with age at first intercourse was of borderline significance (OR for <19 compared to ≥21 years = 1.7). Among contraceptive methods, only current use of oral contraceptives (OCs) seemed to affect significantly the probability of HPV DNA detection in Ho Chi Minh City (OR = 3.2, 95% CI 1.4–7.4). OC use was very rarely reported in Hanoi. An inverse association with condom use was of borderline significance in Ho Chi Minh City (OR = 0.6). Use of intrauterine device (IUD) was reported by 42.0% of women in Ho Chi Minh City and 55.4% in Hanoi. It was unrelated to HPV DNA detection. HSV-2 antibodies were found among 34.5% of women in Ho Chi Minh City and they were significantly associated with HPV DNA detection (OR = 2.4; 95% CI 1.5–3.8). In Hanoi, only 8.9% of women were HSV-2 seropositive and no relationship between HSV-2 and HPV DNA emerged (Table IV).
Table IV. OR of HPV Infection and Corresponding 95% CI According to Sexual Indicators, Contraceptive Methods and HSV2 Antibodies in Ho Chi Minh City and Hanoi, Vietnam
When the variables most strongly related to HPV DNA detection were fitted into the same model (Table V), associations were slightly attenuated, most notably those aged <25 years in Ho Chi Minh City and number of lifetime sexual partners. HSV-2 antibodies (OR = 2.2) and current OC use (OR = 2.9) seemed to be the strongest independent risk factors for HPV DNA detection in Ho Chi Minh City. In Hanoi, nulliparity was the only significant risk factor.
Table V. Major Risk Factors for HPV DNA Detection in the Multivariate Model in Ho Chi Minh and Hanoi, Vietnam
ORs were similar in women <45 and ≥45 years of age (Table VI), although the associations with the number of lifetime sexual partners and current OC use were somewhat stronger among older women. Major determinants of HPV DNA detection were also similar for single and multiple infections (Table V) and for low- and high-risk HPV infections (Table VI). The peak in HPV infections seen in women <25 was, however, chiefly due to multiple rather than single infections.
Table VI. Major Risk Factors for HPV DNA Detection in the Multivariate Model by Age Group, HPV Type and Multiplicity of HPV Infections, Vietnam
Our study has shown that populations exist, as in Hanoi, where HPV DNA detection in women is very rare. In Ho Chi Minh City, HPV DNA detection was 5-fold higher than in Hanoi. Infection with multiple HPV types was 9-fold more common in Ho Chi Minh City than in Hanoi. A peak in the prevalence of HPV DNA detection among women <25 was found only in Ho Chi Minh City. Applying age-specific prevalence from our 2 surveys to the female population of Ho Chi Minh City and Hanoi, respectively, gives, again, an overall HPV prevalence in women aged ≥20 of 11.1% and 2.4% (10.6% and 2.3% if the world standard population is used). The difference in HPV prevalence between Ho Chi Minh City and Hanoi cannot be due to technical reasons because women were randomly chosen in the same way in the 2 areas and all samples were examined in parallel in the same laboratory. Thus, our HPV-related findings show a good correlation with the 4-fold difference in cervical cancer incidence rates between the 2 study areas.2, 3
Major characteristics associated with HPV DNA detection in multivariate analysis included nulliparity, indicators of sexual habits (most notably the presence of HSV-2 antibodies) and current use of OCs. Risk factors for HPV DNA detection in Hanoi were consistent with the ones from Ho Chi Minh City, despite the lower prevalence. The distribution of HPV DNA detection across age groups in Hanoi suggests that, in low-risk populations, the early peak of HPV infection may be missing or delayed to 35–44 years. The prevalence of HSV-2 antibodies was also higher in Ho Chi Minh City (34%) than in Hanoi (9%), despite similarly low proportions of women reporting ≥2 lifetime sexual partners (7% and 6%, respectively) and a greater percentage of women who had sexual intercourse before age 19 years in Hanoi. Unless underreporting of high-risk sexual partners is more severe in Ho Chi Minh City than in Hanoi, the difference in prevalence of HPV and HSV-2 between the 2 areas cannot be explained by our questionnaire-based findings. It may rather be due to the greater isolation North Vietnam had endured compared to South Vietnam, during many decades of war and socialist economy.
Few population-based studies of HPV DNA prevalence have been published and, notably, none have been from Asia. Four large surveys similar to our present one were carried out in Costa Rica,8 Mexico,9 Argentina16 and South Korea.17 Overall prevalences of HPV DNA detection were 16%, 15% (women with normal cytology only), 18% and 10%, respectively. As in Ho Chi Minh City, prevalences of ≥20% were found in young women. Among 1,000 cytologically normal women aged 20–29 years, randomly chosen from the population of Copenhagen, Denmark, overall HPV DNA detection was 15%.7 Among 878 cytologically normal women aged 20–49 years, randomly selected from general practice lists in Ontario, Canada, HPV prevalence was 13% overall but 24% in the 20–24 age group.18 Among 3,305 cytologically normal women aged 15–69 years who attended cervical cancer screening programs in the Netherlands, HPV prevalence was 4.6%, with a peak of 20% at age 25–29.13
Previous studies are also consistent with our findings in respect to the direct association between HPV DNA detection and number of sexual partners.7, 9, 16, 17, 18, 19, 20 In some,7, 18, 21, 22 but not all,9, 16, 20, 23, 24, 25 previous studies an association between HPV DNA and OC use (especially current use) was found, after allowance for sexual habits. Previous results are also inconsistent for an association with condom use and smoking habits,7, 23 which showed, respectively, an inverse association of borderline significance and a direct one in our Vietnam surveys. In respect to the possible role of reproductive characteristics, nulliparous women have often been found to be at a higher risk of HPV infection, also after carefully allowing for age.7, 23, 26 It is possible that some residual confounding effect of sexual habits accounts for the association of HPV infection with OC use, smoking and nulliparity. Oral contraceptives,27 cigarette smoking23 and a high number of births26 seem, however, more consistently associated with the risk of cervical carcinoma than with HPV DNA detection.
Major determinants of HPV DNA detection were similar in Vietnam for single and multiple infections for high-risk and low-risk HVP types and for women younger than and older than age 45. In some,7, 16 but not all,9 previous studies, the direct association of HPV DNA detection with the number of lifetime sexual partners seemed especially strong among young women.
As in most population-based16 and clinic-based13, 28, 29 samples of the general female population, a broad spectrum of HPV types was found, with HPV 16 being marginally, if at all, the most frequent type. In Vietnam, HPV 58 was the second most common type after HPV 16, thus confirming the relative frequency of this type in the general female population.8 Previous reports have also shown that HPV 58 was among the most common types found in cervical cancer specimens in China, Thailand and the Philippines.30 In agreement with previous reports, the high-risk types were substantially more frequent than low-risk ones and multiple infections were common (36% of all HPV infections).8 This suggests that multiple HPV infections, which do not seem associated with an increased risk of cervical neoplasias compared to single infections,8 are frequent, especially in samples of exfoliated cells, which come from a broader area of the cervix than biopsies. Multiple HPV infections were also found substantially more often in women age <25 years than among older women, which is in agreement with previous studies.17
Our present findings from an urban district of Ho Chi Minh City and a peri-urban district of Hanoi cannot be considered representative of the whole Vietnamese population. They should, however, provide an accurate indication of the broad variation in HPV prevalence among sexually active (married) women in Vietnam. Overall, >90% of contacted women agreed to participate and were interviewed. Eventually, 19% of participating women could not be included in our present report on account of a lack of an adequate exfoliated cell sample for HPV testing and/or of a Pap smear. The analysis of the women who were interviewed, but for whom HPV findings or a Pap smear were not available, showed that they were similar to the women included in the present analysis in respect to age and major risk factors for HPV infection (e.g., education, parity, number of induced abortions, and number of lifetime sexual partners. As we had access to a population-based, rather than clinic-based, sample, we chose not to exclude the relatively few women with cytologic abnormalities. The exclusion of these women would have slightly decreased the prevalence of HPV detection (i.e., 10.0% in Ho Chi Minh City and 1.9% in Hanoi).
In conclusion, our present findings, the first ever reported from Vietnam, add to the accumulating evidence that: (i) the prevalence of HPV infection varies greatly between populations worldwide; (ii) HPV prevalence and incidence rates of cervical cancer in a population are closely correlated; and (iii) indicators of sexual activity, especially those that reflect more accurately high-risk sexual behaviour of a woman and her partner (e.g., HSV-2 antibodies) and perhaps OC use are moderately associated with HPV prevalence.
The 2 surveys were conducted by P.T.H.A. (in Hanoi) and N.T.H. (in Ho Chi Minh) with the assistance of N.H.N. and N.B.D., and N.T.T., respectively. R.H. and N.M. designed the protocol and initiated the study, assisted by J.S.S. and D.M.P. S.V. checked data accuracy and did statistical analyses. R.A. carried out testing for anti-HSV-2 antibodies and P.J.F.S. and C.J.L.M.M. testing for HPV-DNA. P.T.H.A. and S.F. wrote the first draft of the manuscript. All investigators contributed to the interpretation of the data and to the writing of the paper. We acknowledge the collaboration of all women and staff of the Gynecology Outpatient Clinic and the Cytology Unit of Hung Vuong Hospital. We also thank Dr. D.L.D. Hanh, Dr. N.V.M. Linh and the medical staff of Communes 2, 21 and 14, District 10, Ho Chi Minh City, the Hanoi Cancer Institute and Cancer Registry, the Center of Population and Family Planning of the Socson District. The People's Committee and the staff of the Health Center of the Communes: Hongky, Kimlu, Minh phu, Phulo supported the study. Mrs. A. Arslan, Ms. S. Hussain and Mr. C. Hsu helped with the preparation of the data.