For Thai women, infection with human papillomavirus (HPV) type 16 or 18 is the most important factor in the progression of precancerous cervical lesions to invasive cancer; the risk of cancers associated with these viral types rises with decreasing age at first intercourse and with increasing numbers of pregnancies.1 The great majority of women with cervical cancer have had intercourse only with their husbands, and their risk of cancer is strongly related to the frequency of their husbands' visits to prostitutes as young men; according to the investigators, most are therefore likely to have acquired carcinogenic types of HPV from their husbands.2These are the major findings of three interrelated studies--two involving women with cervical cancer and one involving commercial sex workers in Bangkok and husbands of a subgroup of women with cervical cancer and controls from the first two studies.
All participants in the first two studies were recruited from Siriraj Hospital in Bangkok. Women with newly diagnosed precancerous lesions or invasive cervical cancer were eligible if they had been admitted to the hospital between September 1991 and September 1993, had been born after 1929 and had lived in Thailand for at least one year before their interview. These women were matched to other hospitalized women without either of these cervical abnormalities, on the basis of date of admittance to the hospital, five-year age-group and area of residence. Investigators interviewed all participating women to obtain information regarding their use of contraceptives, sexual behavior, medical history, tobacco and alcohol use, and socioeconomic status. Blood samples and cervical scrapings were analyzed to detect markers for sexually transmitted diseases and to identify specific types of HPV. For the third study, the investigators interviewed and obtained biologic samples from a subset of the husbands of women with invasive cancer and of control women, as well as commercial sex workers in one brothel and one massage parlor in Bangkok.
Cofactors in HPV-Related Cervical Cancer
To investigate risk factors for specific types of invasive cervical cancer, the investigators analyzed data from 232 women with invasive cervical cancer and 291 controls.3 They also examined tissue samples from cases to classify tumors as either squamous or adenomatous. They then compared women with each type of tumor to controls, and compared cases with each of two strains of HPV (HPV-16 and HPV-18) to controls with no evidence of HPV. All odds ratios were adjusted for age using logistic regression analysis.
Women with squamous and adenomatous cancers were much more likely than controls to be infected with a carcinogenic type of HPV (79% and 76% vs. 3%). HPV-16, the most common carcinogenic type in this sample, occurred nearly twice as often in women with squamous tumors as in those with adenomatous tumors (60% vs. 33%), and was rare among controls (2%). HPV-18 was much more strongly associated with adenomatous cancers (43%) than with squamous ones (14%), and was not detected in any of the controls. Women with squamous and adenomatous cancers did not differ significantly with regard to any other risk factors.
The risk of both tumor categories increased as age at first intercourse declined: For example, compared with women who had first had sex at age 24 or older, those who had done so at ages 17-18 had three times the odds of developing an HPV-16- related tumor (odds ratio of 3.1) and more than four times the odds of having an HPV-18-related tumor (4.6). Women's pregnancy history also influenced their risk: Those who had experienced three or four pregnancies had a higher risk of developing both types of tumors than did women who had been pregnant only once or twice (odds ratios of 2.1 for HPV-16-related tumors and 4.1 for HPV-18-associated tumors).
Women with a Pap smear in the previous 12 months were at much lower risk for both HPV-16-related and HPV-18-related tumors (odds ratios of 0.2 and 0.1, respectively) than were women who had never had a Pap smear, while women who had ever had a chest X ray were at lower risk (0.6 and 0.4) than those who had not. The odds of HPV-16-related cancer were significantly lower for women who had ever attended school (0.5) or used an IUD (0.4) than for women without those characteristics. Among women who had been pregnant, those with a history of spontaneous abortion had an elevated risk of HPV-18-related cancer (2.2).
Each of the associations with HPV-16-related cancer remained after the effects of other factors had been accounted for; the number of women with HPV-18-related tumors was too small to permit a similar analysis. No significant association was observed between either tumor category and number of sexual partners; use of oral contraceptives; smoking; or prior infection with herpes simplex virus types 1 or 2, syphilis or hepatitis B.
The investigators note that the association between early age at first intercourse and HPV-16-related and HPV-18-related tumors may be confounded by the sexual habits of the women's husbands and that the reduced risk of cancer related to HPV-16 observed among IUD users is probably the result of cervical cancer screening at the time of IUD insertion or removal. They also point out that school attendance and chest X rays are most likely markers for important protective socioeconomic factors that have yet to be identified. The investigators conclude that "the similarity in risk factors for cervical carcinomas with HPV-16 and HPV-18 DNA provides strong evidence that the same cofactors operate to enhance the carcinogenicity of these two viral types."
Progression to Invasive Cancer
To investigate risk factors for progression from precancerous lesions to invasive cervical cancer, the investigators analyzed data from 190 women with invasive squamous-cell cancer and 291 matched controls, and from 75 women with precancerous lesions and 124 matched controls.4 They compared cases to controls to determine the factors associated with each type of cervical abnormality. They then examined the risk of invasive disease associated with each factor in relation to the risk of precancerous lesions to identify factors associated with progression to invasive cancer. All logistic regression results were adjusted for age.
The women with invasive cancer and their controls were somewhat older (median age, 43-45 years) than the women with precancerous disease and their controls (37-38 years). The proportions of women with invasive and precancerous disease reporting vaginal bleeding were similar (65% and 63%), as were the proportions reporting no specific symptoms (33% and 31%). Women with precancerous lesions were much more likely than controls to be infected with any type of HPV (77% vs. 10%), any carcinogenic type of HPV (57% vs. 6%), HPV-16 (38% vs. 5%) or HPV-18 (8% vs. 1%).
Women with invasive disease were significantly more likely than those with precancerous lesions to test positive for any HPV type (odds ratio of 2.7), any cancer-causing type of HPV (3.5), HPV-16 (4.2) or HPV-18 (4.0), indicating that HPV infection is a significant factor in the progression to invasive disease. No factor other than HPV infection had a significantly stronger association with invasive cancer than with precancerous lesions.
The Role of Husbands and Sex Workers
More than 90% of the women with cervical cancer and the controls participating in these studies reported only one lifetime sexual partner. To investigate the possible role of husbands in women's risk for HPV and cervical cancer, investigators interviewed and obtained serum samples from the husbands of 50 of the 175 women with squamous-cell cervical cancer who reported one lifetime sexual partner, and the husbands of 98 of the 272 monogamous controls.5 Penile scrapings were obtained from 57 husbands of women with precancerous lesions or invasive disease and 68 husbands of controls. The association found in previous research between Thai men's sexual contact with prostitutes and their wives' risk of cervical cancer was confirmed in this study and suggests that the men acquire carcinogenic HPV from prostitutes and transmit the infection to their wives.
To determine the prevalence of carginogenic HPV and high-grade precancerous lesions among commercial sex workers, the researchers interviewed and obtained blood samples and cervical smears and scrapings from 170 female massage parlor workers and 84 female brothel workers in Bangkok.
Women whose husbands reported more than 280 lifetime visits to prostitutes had more than three times the risk of invasive disease among women whose husbands reported no visits (odds ratio of 3.2). The husband's annual number of visits before the age of 30 and his use of condoms less than 10% of the time during these visits were also associated with an elevated risk of invasive cancer.
On average, the brothel workers interviewed were much younger than massage parlor workers (median age of 19 vs. 30 years) and were more likely to be infected with each of nine carcinogenic types of HPV tested. For example, among 20-24-year-old women, the prevalence of HPV-16 was 18% for brothel workers and 10% for massage parlor workers.
High-grade precancerous lesions were found in 8% of brothel workers and 4% of massage parlor workers. Women who had high-grade lesions were more likely than those who did not to test positive for HPV-16 (odds ratio of 19.2).
While acknowledging that the response rate among husbands was low, the investigators note that their findings are consistent with those from studies with higher response rates and provide "direct evidence for the role of husbands as vectors of HPV transmission from prostitutes."--A. Hirozawa
1. Thomas DB et al., Human papillomaviruses and cervical cancer in Bangkok. I. Risk factors for invasive cervical carcinomas with human papillomavirus types 16 and 18 DNA, American Journal of Epidemiology, 2001, 153(8):723-731; and Thomas DB et al., Human papillomaviruses and cervical cancer in Bangkok. II. Risk factors for in situ and invasive squamous cell cervical carcinomas, American Journal of Epidemiology, 2001, 153(8): 732-739.
2. Thomas DB et al., Human papillomaviruses and cervical cancer in Bangkok. III. The role of husbands and commercial sex workers, American Journal of Epidemiology, 2001, 153(8):740-748.
3. Thomas DB et al., Human papillomaviruses and cervical cancer in Bangkok. I. Risk factors for invasive cervical carcinomas..., 2001, op. cit. (see reference 1).
4. Thomas DB et al., Human papillomaviruses and cervical cancer in Bangkok. II. Risk factors for in situ..., 2001, op. cit. (see reference 1).
5. Thomas DB et al., 2001, op. cit. (see reference 2).