Determinants of Cervical Human Papillomavirus Infection: Differences ...

2 downloads 0 Views 93KB Size Report
Dec 5, 2001 - infections among 2080 women who participated in cervical cancer ... low- and unknown-risk HPVs had a second peak in older women.
28

Determinants of Cervical Human Papillomavirus Infection: Differences between High- and Low-Oncogenic Risk Types Paul K. S. Chan,1 Alexander R. Chang,2 Jo L. K. Cheung,1 Denise P. C. Chan,1 L. Y. Xu,4 Nelson L. S. Tang,3 and Augustine F. Cheng1

Departments of 1Microbiology, 2Anatomical and Cellular Pathology, and 3Chemical Pathology and 4Centre for Clinical Trials and Epidemiological Research, Chinese University of Hong Kong, Prince of Wales Hospital, Shatin, New Territories, Hong Kong SAR, China

This cross-sectional survey assessed the determinants of human papillomavirus (HPV) infections among 2080 women who participated in cervical cancer screening. HPVs were typed by restriction and sequencing analyses. The prevalence of HPV was 7.3% (4.2% for high-risk, 1.9% for low-risk, and 2.1% for unknown-risk types). High-risk HPV prevalence decreased with age, whereas low- and unknown-risk HPVs had a second peak in older women. Young age was the only common variable associated with the 3 groups of HPV infections. Lifetime number of sex partners was associated with high- and low-risk types but not with unknown-risk HPVs. Previous Pap smear, treatment for cervical lesions, induced abortion, smoking and having smoker(s) in the family were risk factors for high-risk HPVs. Barrier contraception was protective for low-risk HPVs; current vaginal discharge had a negative association with unknown-risk HPVs. The results indicate that different risk profiles exist for infections with different HPV groups.

Human papillomaviruses (HPVs) are a heterogeneous group of DNA viruses encompassing 100 genotypes, of which 40 are found in the genital tract [1]. Epidemiologic and experimental data indicate that genital HPV is the main etiologic factor for cervical cancer [2, 3]. Genital HPVs have been classified further according to their malignant potential. High-risk (oncogenic) types (e.g., HPV-16 and -18) have been linked to high-grade cervical lesions and invasive carcinoma, whereas low-risk (nononcogenic) types (e.g., HPV-6 and -11) are associated with benign or low-grade intraepithelial lesions [4]. More than 30 years of epidemiologic research have established a strong association between sexual activity and the development of cervical cancer [5, 6]. Although genital HPV infection, the biologic intermediate cause of cervical neoplasia, is considered to be sexually transmitted, not all studies have reproduced this result uniformly. The association between

sexual activity and overall HPV prevalence varies from strong [7–9], to moderate [10, 11], to no link [12]. In addition, several studies have suggested that different HPV types may have different degrees of sexual transmissibility [13, 14]. Thus, studies based on different populations may produce dissimilar results because of variations in the relative prevalence of HPV types. An understanding of the epidemiologic risk factors for genital HPV infection and, in particular, of the differences between high-risk and low-risk types is crucial for controlling HPV-related diseases. In Hong Kong women, cervical cancer ranks fourth in cancer incidence. Annually, there are 500 new cases and 150 deaths from the malignancy. The age-standardized incidence rate of cervical cancer is 11.6 cases per 100,000 persons, and the median age at diagnosis is 56 years. The cumulative lifetime incidence up to age 75 years is 1 in 75 [15]. In this study, we assessed the epidemiologic risk profiles of genital HPV infections among Hong Kong Chinese women.

Received 21 May 2001; revised 17 September 2001; electronically published 5 December 2001.

Presented in part: 19th International Papillomavirus Conference, Florianopolis, Brazil, September 2001 (abstract P-232). Informed consent was obtained from all study participants, and human experimentation guidelines of the Chinese University of Hong Kong were followed in the conduct of the clinical research. Financial support: Lee Hysan Foundation (grant); United College, Chinese University of Hong Kong endowment fund (research grant 1998/99); Hong Kong Cancer fund (to Shatin Community Clinic for the Prevention of Cervical Cancer). Reprints or correspondence: Dr. Paul K. S. Chan, Chinese University of Hong Kong, Prince of Wales Hospital, Dept. of Microbiology, Shatin, New Territories, Hong Kong SAR, China ([email protected]). The Journal of Infectious Diseases 2002;185:28–35 q 2002 by the Infectious Diseases Society of America. All rights reserved. 0022-1899/2002/18501-0004$02.00

Materials and Methods Study population and epidemiologic data collection. A crosssectional study was conducted of 2080 consecutive Chinese women who enrolled for a Pap smear at the Shatin Community Clinic for the Prevention of Cervical Cancer (Hong Kong SAR). All subjects were interviewed face-to-face so that their epidemiologic characteristics could be determined. Detection and identification of HPVs. A cervical sample was collected with a Cervex brush (Rovers Medical Devices) from each woman for routine cytologic examination. The brush then was immersed in 10 mL of PBS and was agitated vigorously so that residual cells were dislodged. Cell pellets were collected and kept

JID 2002;185 (1 January)

Determinants of HPV infection

frozen at 270 C until analyzed. Total DNA was extracted as described elsewhere [16]. To determine the quality and adequacy of extracted DNA and to confirm the absence of nonspecific inhibitors, polymerase chain reaction (PCR) that used b-globin primers PC03 and PC04 [17] was done. HPV DNA was detected by “hotstart” PCR with degenerate HPV consensus primers MY09 (50 -CGTCCMARRGGAWACTGATC-30 ) and MY11 (50 -GCMCAGGGWCATAAYAATGG-30 ) (M ¼ A or C; R ¼ A or G; W ¼ A or T; and Y ¼ C or T) [18]. These primers are capable of amplifying a 450-bp DNA fragment from the L1 open-reading frame of > 40 genital HPV types [19, 20]. The PCR was performed in a 50-mL reaction mix containing 50 mM KCl, 1.5 mM MgCl2, 200 mM each dNTP, 0.25 mM each primer, 1.25 U of DNA polymerase (HotStarTaq; Qiagen), and 5 mL of extracted DNA. The thermal cycling conditions were denaturation of DNA template and activation of DNA polymerase at 95 C for 15 min; 40 cycles at 94 C for 1 min, 55 C for 1 min, and 72 C for 1 min; and a final extension at 72 C for 8 min. Amplicons were detected by agarose gel electrophoresis. All PCRs were done under conditions to minimize contamination [21]. A negative control containing all PCR reagents, except DNA template, was added after every fifth specimen, to monitor for contamination. A positive control consisting of 50 copies of fulllength HPV-16 genome was included in each PCR run, to monitor for sensitivity. Serial dilutions of recombinant plasmids containing the whole genome of HPV-6, -11, -16, and -18 were used to estimate the analytic sensitivity of the PCR. The amplification reactions allowed for consistent and reproducible detection of >50 copies of these HPV genomes. To further reassure the absence of crosscontamination, 50 positive samples were selected randomly for repeat testing and all showed consistent results. HPV types were identified by restriction fragment– length polymorphism (RFLP) that used restriction enzymes Rsa I and Dde I, as described elsewhere [20, 22]. This approach is capable of identifying . 40 types of genital HPVs. Samples showing ambiguous RFLP patterns were subjected to direct sequencing. The MY09/MY11 PCR products were purified by using MicroSpin columns S-400 (Pharmacia Biotech) and were subjected to cycle sequencing with the ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction Kit (Perkin-Elmer). Sequencing reactions were run on the ABI Prism 310 Genetic Analyzer (Perkin-Elmer). A BLAST server was used to match all sequences available in GenBank. HPV type was identified on the basis of >98% sequence homology in 400– 450 nt. Statistical analysis. The association between epidemiologic variables and the presence of HPV was examined initially by age-adjusted analyses with odds ratios, 95% confidence intervals, and P values calculated by logistic regression. To identify independent associations, variables with P < :08 by age-adjusted analyses were included in the multivariate analyses by multiple logistic regressions. Statistical analysis was done with SPSS software (version 10.1.0; SPSS). In addition to analyzing associations for all HPV types combined, the same analyses were performed with HPVs categorized by oncogenicity [2]. HPV types 16, 18, 31, 33, 39, 52, 56, 58, and 68 were categorized as high-risk types; HPV types 6, 11, 53, 57, and 66 were categorized as low-risk types. We categorized HPV-61, CP6180, CP8061, CP8304, LVX160, and MM8 as unknown-risk

29

types. Women with coinfection with multiple types were considered to be in each risk category, on the basis of detection of any of the high-, low-, or unknown-risk HPV types.

Results HPV type-specific prevalence. In total, 2080 Chinese women were recruited. All had cervical samples with DNA of adequate quality for PCR. HPV DNA was detected in 152 samples (7.3%), of which 48 (31.6%) showed ambiguous RFLP patterns and required sequencing for genotype identification. Multiple infections with 2 HPV types were found in 22 women: 4 were coinfected with different high-risk types, 5 with a highrisk and a low-risk type, 6 with a high-risk and an unknown-risk type, and 7 with a low-risk and an unknown-risk HPV. The overall prevalences of high-, low-, and unknown-risk HPV infections were 4.2%, 1.9%, and 2.1%, respectively. Table 1 shows the prevalence of HPV types. HPV age-specific prevalence. The mean age of the 2080 women was 43.8 years (range, 19–83 years; SD, 9.56 years): 337 (16.2%) were < 35 years old, 912 (43.8%) were 36–45 years old, 596 (28.7%) were 46–55 years old, and 235 (11.3%) Table 1. Prevalence of human papillomavirus (HPV) types in 2080 Hong Kong Chinese women. HPV type

No. (%) of women

a

152 (7.3) 87 (4.2) 39 (1.9) 43 (2.1) 33 (1.6) 24 (1.2) 14 (0.7) 14 (0.7) 12 (0.6) 11 (0.5) 10 (0.5) 10 (0.5) 9 (0.4) 9 (0.4) 8 (0.4) 4 (0.2) 4 (0.2) 2 (0.1) 2 (0.1) 2 (0.1) 2 (0.1) 1 (, 0.1) 1 (, 0.1) 1 (, 0.1)

Any High-risk HPVb Low-risk HPVc Unknown-risk HPVd HPV-16 HPV-58 HPV-11 CP8304 HPV-33 HPV-18 HPV-6 HPV-53 HPV61 LVX160 MM8 HPV-31 HPV-66 HPV-39 HPV-56 HPV-68 CP8061 HPV-52 HPV-57 CP6108

NOTE. The 22 women coinfected with 2 HPV types are counted more than once. a Regardless of single or multiple HPV infection. b High-risk types include 16, 18, 31, 33, 39, 52, 56, 58, and 68. c Low-risk types include 6, 11, 53, 57, and 66. d Unknown risk types include 61, CP6180, CP8061, CP8304, LVX160, and MM8.

30

Chan et al.

Figure 1.

JID 2002;185 (1 January)

Age-specific prevalence of human papillomavirus (HPV) infection in 2080 Hong Kong Chinese women

were > 56 years old. The prevalence of high-risk HPV infection peaked (11.9%) in the 26–30-year-old age group and then gradually declined, reaching a level of 1.7%–3.1% for women .41 years old. A small increase in prevalence in older women (46–55 years old) was observed for low-risk HPV infection. A more obvious second peak for unknown-risk HPV infection was observed in women 51–55 years old (figure 1). Cervical cytology. Of the 2080 women, 1802 (86.6%) had normal cervical cytology, 160 (7.7%) showed reactive changes, 77 (3.7%) had atypical squamous cells of undetermined significance, 31 (1.5%) had low-grade squamous intraepithelial lesions, and 10 (0.5%) had high-grade squamous intraepithelial lesions. Table 2 shows the association between cytologic abnormality and HPV infection. Age-adjusted analyses. After adjustment for age, HPV (any type) infection was associated with years of residence in Hong Kong, previous treatment for cervical lesions, marital status, number of induced abortion, menopause, age at first sexual intercourse, number of lifetime sex partners, barrier contraception, postcoital bleeding, and smoking status of the subject or of a family member. Different risk profiles emerged when the analyses focused separately on the oncogenic HPV infection risk group (table 3). Multivariate analyses. To identify independent associations, all variables with an age-adjusted P < :08 for any HPV oncogenic risk group were subjected to multivariate analyses.

Multiple logistic regressions were done by using the enter method for infections with any HPV type in combined, high-risk, lowrisk, and unknown-risk groups, respectively (table 4). The only common variable with an independent risk association with all 3 groups of HPVs was young age. Number of lifetime sex partners was associated with both high- and low-risk HPVs, but not with unknown-risk HPVs. Previous Pap smear, history of treatment for cervical lesion, induced abortion, cigarette smoking, and having a family member who smoked (passive smoking) were independent risk factors for high-risk HPV infection but not for low- or unknown-risk groups. Regular use of barrier contraception had a significant protective effect for low-risk HPV infection, but the effect was only near statistical significance for high-risk types. Having current vaginal discharge had a negative association with unknown-risk HPV infection but not with other risk groups. When multivariate analyses were repeated with the backward step-wise method, consistent results were obtained (data not shown).

Discussion An understanding of the epidemiology of HPV infection is an important first step toward the development of strategies for preventing cervical cancer. Although HPVs can be classified as high- and low-risk types according to their oncogenic potential, most epidemiologic studies have analyzed HPV infection data

JID 2002;185 (1 January)

Determinants of HPV infection

Table 2. Age-adjusted prevalence odds ratios (ORs) for any, high-risk, low-risk, and unknown-risk human papillomavirus (HPV) infections, according to cervical cytologic abnormalities. HPV group, cytologic outcome (n) Any Normal (1802) Reactive (160) ASCUS (77) SIL (41) High risk Normal (1802) Reactive (160) ASCUS (77) SIL (41) Low risk Normal (1802) Reactive (160) ASCUS (77) SIL (41) Unknown risk Normal (1802) Reactive (160) ASCUS (77) SIL (41)

Prevalence, %

Age-adjusted OR (95% CI)

5.0 11.9 27.3 51.2

1.0 2.52 (1.49–4.28)a 6.54 (3.76–11.37)a 19.51 (10.08–37.76)a

2.6 6.3 15.6 43.9

1.0 2.49 (1.23–5.06)b 6.15 (3.08–12.29)a 28.91 (14.37–58.14)a

1.4 3.1 9.1 2.4

1.0 2.19 (0.82–5.82) 6.11 (2.51–14.83)a 1.52 (0.20–11.67)

1.6 4.4 6.5 7.3

1.0 2.84 (1.22–6.64)b 4.16 (1.55–11.19)a 4.78 (1.38–16.55)b

NOTE. Women with normal cytology serve as the reference for ageadjusted analysis by logistic regression. ASCUS, atypical squamous cells of undetermined significance; CI, confidence interval; SIL, squamous intraepithelial lesions including high and low grade. a P < :01. b P < :05.

in an aggregated form. In this study, we tested the hypothesis that the 2 risk groups of HPV infections might have different epidemiologic risk profiles. We used a consensus PCR to detect HPV DNA that was followed by RFLP and sequencing analyses for genotype identification. This approach allowed a “catch-all” detection and identification of a broad spectrum of genital HPV types. However, consensus PCR is less efficient than typespecific PCR in detecting multiple infections, especially when a coinfecting HPV type exists in a low concentration. In the present study, a significant proportion of the HPV positive women harbored genotypes that have not been fully characterized. This finding is in accordance with studies of populations with a low incidence of abnormal cervical cytology, where the majority of HPVs detected were of uncharacterized genotypes [14, 23]. The association of these uncharacterized genotypes with cervical cancer is uncertain. For the purpose of analysis, these uncharacterized genotypes were regarded as having “unknown risk” for oncogenesis. Analyses of age-specific prevalence revealed different patterns for high-, low-, and unknown-risk HPV infections. The prevalence of high-risk HPV infection was highest among young women and then declined with increasing age, whereas the prevalence of unknown-risk HPV infection showed a bimodal distribution, with peaks occurring at ages 26–30 and 51–55 years.

31

A slight increase in prevalence in older women also was observed for low-risk HPV infection. A difference in age distribution for different groups of HPVs also has been reported in recent studies. Herrero et al. [24] in Costa Rica showed that HPV infections had 2 peaks: the second was in women > 55 years old and was predominately due to infection with lowrisk or uncharacterized types, although a slight increase in high-risk types also was observed. Jacobs et al. [25] in Amsterdam found that the prevalence of high-risk HPV declined with increasing age, whereas the low-risk types remained constant. A common observation in these reports and in the current study is that the prevalence of low-risk and uncharacterized HPV infection in older women is higher than expected, but the reason is not clear. Most HPV infections are transient: 70% are no longer evident within a year, and up to 91% are cleared within 2 years [26, 27]. Low-risk HPV infections are cleared equally well and probably faster than high-risk types [26, 28]. However, the correlation of HPV clearance or persistence with age is not fully understood. Our observation could be related to changes in immune and hormonal status of older women that impair HPV clearance or favor HPV reactivation. Another possible explanation could be differences in susceptibility to different HPV groups among older women as a result of previous exposure. In our study cohort, younger subjects had 2 times more infections due to high-risk HPVs, compared with other types. It is possible that older women are more likely to be immunologically naive to low-risk or unknownrisk HPVs than to high-risk HPVs. It is generally thought that most HPV infections detected in older women represent persistent infections with high-risk viruses. In contrast, we found that the majority (75%) of HPVs detected in Hong Kong women 51–55 years old were non– high-risk types. This emphasizes the importance of using HPV detection methods that are type specific when applied to cervical cancer screening, regardless of the age of the target population. Since cervical lesions are a consequence rather than a risk factor for HPV infection, cytology results were not included in the subsequent multivariate analyses. The overall results of multivariate analyses revealed different risk profiles for the 3 groups of HPV infections. The only common risk factor found to have independent association for all 3 groups of HPVs was young age. Lifetime number of sex partners was an independent risk for both high- and low-risk HPVs, indicating the importance of sexual transmission. Reports linking age at first sexual intercourse and HPV infection have been inconclusive [13, 23, 29]. In the present study, women who had their first sexual intercourse at or before age 17 years had a higher detection rate for HPV (any type and unknown risk types) by age-adjusted analysis. However, the association was not significant when it was subjected to multivariate analysis, which suggests that the age at first sexual intercourse may not be a good predictor for subsequent sexual behavior among Hong Kong women.

Table 3. Age-adjusted prevalence odds ratios (ORs) for any, high-risk, low-risk, and unknown-risk human papillomavirus (HPV) infections, according to epidemiologic variables. Any HPV Variable (n) Years in Hong Kong 11 (1941) Education Secondary or below (1954) Tertiary or above (126) Previous Pap smear Never (593) Ever (1487) Treatment for cervical lesionb Never (1894) Ever (186) Marital status Married (1902) Single (178) No. of full-term pregnancies 0 (211) > 1 (1869) No. of induced abortions 0 (1117) > 1 (963) No. of spontaneous abortions 0 (1836) >1 (244) Menopause No (1647) Yes (433) Age at first sexual intercourse, years >18 (1952) 56 Years in Hong Kong < 10 > 11 Previous Pap smear Never Ever Treatment for cervical lesiona Never Ever Marital status Married Single No. of full-term pregnancies 0 >1 No. of induced abortions 0 >1 Menopause No Yes Age at first sexual intercourse, years > 18 < 17 Lifetime no. of sex partners 4 Regular barrier contraception Yes No Postcoital bleeding No Yes Current vaginal discharge No Yes Cigarette smoking No Yes Smoker in family No Yesc

OR (95% CI)

High-risk HPV P

OR (95% CI)

.001 1.0 0.47 (0.29–0.78) 0.45 (0.25–0.84) 0.13 (0.05–0.40)

P

.103

.083

.117

.088

.649

.022

1.0 2.34 (1.45–3.77)

.824

b

33

.532 1.0 1.50 (0.42–5.40)

.226 1.0 1.53 (0.77–3.07)

NOTE. Variables showing age-adjusted P < :08 were included in multiple logistic regression analyses. CI, confidence interval. Local surgery, including cauterization, cryotherapy, laser ablation, or loop electrosurgical excision procedure. 95% CI not applicable. c 90% were husbands. a

.024 1.0 0.26 (0.08-0.83)

1.0 0.87 (0.25–3.06) , .001

.012

.203 1.0 1.89 (0.71–5.06)

1.0 145.24 b

1.0 2.40 (1.13–5.09)

.225 1.0 1.72 (0.72–4.11)

1.0 2.56 (0.88–7.41)

1.0 0.40 (0.12–1.26)

1.0 1.77 (0.92–3.42)

.003

.747

.376

.981 1.0 1.02 (0.20–5.23)

1.0 9.60 (2.16–42.76)

1.0 1.14 (0.51–2.53)

1.0 0.66 (0.26–1.65)

.002

.054

.032

.076 1.0 2.56 (0.91–7.22)

1.0 6.10 (1.94–19.15)

1.0 1.74 (0.99-3.05)

1.0 1.85 (1.06–3.26)

.746

, .001

.001

.270 1.0 1.72 (0.66–4.51)

1.0 0.82 (0.25–2.67)

1.0 5.07 (2.06-12.44)

1.0 2.17 (1.35–3.48)

.106

.066

, .001

.105 1.0 1.71 (0.90–3.25)

1.0 2.38 (0.83–6.80)

1.0 0.41 (0.16-1.06)

1.0 4.85 (2.30–10.23)

.815

.594

.931

.857 1.0 2.26 (0.55–9.24)

1.0 1.09 (0.55–2.16)

1.0 0.75 (0.27–2.14)

1.0 0.97 (0.50–1.89)

.285

.032

.078

.362 1.0 1.67 (0.55–5.04)

1.0 0.55 (0.19–1.64)

1.0 1.67 (1.05–2.67)

1.0 1.73 (0.94–3.17)

.436

.559

.125

.966 1.0 1.00 (0.34–2.92)

1.0 1.50 (0.54–4.14)

1.0 0.79 (0.35–1.76)

1.0 1.32 (0.93–1.88)

.743

.081

.539

.621 1.0 1.21 (0.57–2.54)

1.0 1.22 (0.38–3.93)

1.0 2.00 (0.92–4.32)

1.0 0.82 (0.44–1.54)

.445

.005

.226 1.0 1.46 (0.79–2.67)

.180 1.0 0.52 (0.20–1.35)

1.0 0.75 (0.35–1.58)

1.0 2.42 (1.31–4.49)

P .045

.193

.020

.012

OR (95% CI) 1.0 0.41 (0.16–1.04) 0.63 (0.22–1.82) 0.11 (0.02–0.76)

1.0 0.51 (0.19–1.40)

1.0 2.06 (1.12–3.80)

1.0 1.93 (1.15–3.22)

P .022

.574

.117

Unknown-risk HPV

1.0 0.27 (0.10–0.76) 0.48 (0.15–1.55) 0.06 (0.01–0.68)

1.0 0.81 (0.38–1.72)

1.0 1.41 (0.92–2.17)

OR (95% CI)

.038 1.0 0.49 (0.26–0.91) 0.32 (0.14–0.72) 0.24 (0.05–1.16)

1.0 0.63 (0.36–1.10)

1.0 1.57 (1.10–2.25)

Low-risk HPV

.885 1.0 1.05 (0.55–1.99)

34

Chan et al.

Another finding was a protective effect of barrier contraception for low-risk HPVs, whereas the effect for high-risk HPVs was only close to statistical significance, and no effect was observed for unknown-risk HPVs. Published data on the effectiveness of barrier contraception in preventing HPV infection are inconsistent. Richardson et al. [14] reported that regular lifetime condom use had a protective effect for high-risk HPVs but not for low-risk HPVs. Kjaer et al. [30] found that condom use within the past year was an independent risk factor for low-risk HPV infection. These conflicting observations could be due to the fact that, although condoms confer some degree of protection, their use may also be associated with more frequent sexual activity. Nevertheless, further studies are required to clarify whether these observations indicate different transmission mechanisms between high-risk and low-risk HPVs. We also evaluated the effect of oral contraceptive use on HPV infection but found no significant association. In our study, cigarette smoking was an independent risk factor for high-risk HPV infection but not for low-risk or unknown-risk HPVs. The association with smoking was further confirmed by the observation that being a passive smoker (having a family member who smoked) was also an independent risk factor for high-risk HPV infection. This supports the hypothesis that smoking promotes persistent infection, which occurs more often, if not exclusively, in high-risk HPVs. Although most studies have shown that smoking confers > 2fold increase in risk for cervical neoplasia even after adjustment for sexual variables [31], its association with HPV infection is less consistent [29]. The latter could be due, in part, to the fact that many of these analyses were based on infection with all HPV types combined, and yet the association with smoking varies with the type of HPV. We found a negative association between current vaginal discharge and infection with unknown-risk HPVs. This observation has not been reported by other studies. Further investigations are required to clarify whether this is spurious or has a biologic or epidemiologic implication. Women who have received treatment for cervical lesions are more likely to have been infected with high-risk HPVs. In line with this finding, our data confirmed a positive association between a history of cervical lesion treatment and high-risk HPV infection. However, no similar association was observed for low-risk or unknown-risk HPV infection. If this observation is reproducible, it provides evidence that infections with different risk groups of HPVs are linked to different events. In conclusion, this study shows that both high-risk and lowrisk HPV infections are primarily transmitted sexually, as indicated by their strong associations with lifetime number of sex partners. However, the overall epidemiologic risk profiles for high- and low-risk and uncharacterized HPV infections differ. This observation needs to be considered when interpreting study results in which all HPV infections are analyzed in an aggregated form. Finally, a thorough understand-

JID 2002;185 (1 January)

ing of the factors that may facilitate infection with clinical relevant HPV types will be important in the implementation of effective public health programs aimed at preventing cervical cancer. Acknowledgments

We thank Chau Chau Wong, Lai Wai Lie, and Ying Ping Lam (Shatin Community Clinic for the Prevention of Cervical Cancer), for assistance in the collection of cervical samples and for epidemiologic data. We greatly appreciate the support of Sally Lo. References 1. Franco E. Epidemiology of anogenital warts and cancers. In: Reid R, Lorincz AT, eds. Human papillomaviruses. I. Obstet Gynecol Clin North Am 1996; 23:597–623. 2. Bosch FX, Manos MM, Munoz N, et al. Prevalence of human papillomavirus in cervical cancer: a worldwide perspective. International biological study on cervical cancer (IBSCC) Study Group. J Natl Cancer Inst 1995; 87:796–802. 3. Walboomers JM, Jacobs MV, Manos MM, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol 1999; 189:12–9. 4. Lorincz AT, Reid R, Jenson AB, Greenberg MD, Lancaster W, Kurman RJ. Human papillomavirus infection of the cervix: relative risk associations of 15 common anogenital types. Obstet Gynecol 1992; 79:328–37. 5. Brinton LA. Epidemiology of cervical cancer—overview. In: Munoz N, Bosch FX, Shah KV, Meheus A, eds. The epidemiology of cervical cancer and human papillomavirus. Oxford, UK: Oxford University Press, 1992:3–23. 6. Munoz N, Bosch FX. HPV and cervical cancer: review of case-control and cohort studies. In: Munoz N, Bosch FX, Meheus A, eds. The epidemiology of cervical cancer and human papillomavirus. Oxford, UK: Oxford University Press, 1992:251–61. 7. Ley C, Bauer HM, Reingold A, et al. Determinants of genital human papillomavirus infection in young women. J Natl Cancer Inst 1991; 83:997–1003. 8. Bauer HM, Hildesheim A, Schiffman MH, et al. Determinants of genital human papillomavirus infection in low-risk women in Portland, Oregon. Sex Transm Dis 1993; 20:274– 8. 9. Wheeler CM, Parmenter CA, Hunt WC, et al. Determinants of genital human papillomavirus infection among cytologically normal women attending the University of New Mexico student health center. Sex Transm Dis 1993; 20:286–9. 10. Rohan T, Mann V, McLaughlin J, et al. PCR-detected genital papillomavirus infection: prevalence and association with risk factors for cervical cancer. Int J Cancer 1991; 49:856–60. 11. Hildesheim A, Gravitt P, Schiffman MH, et al. Determinants of genital human papillomavirus infection in low-income women in Washington, DC. Sex Transm Dis 1993; 20:279–85. 12. Kjaer SK, de Villiers EM, Caglayan H, et al. Human papillomavirus, herpes simplex virus and other potential risk factors for cervical cancer in a highrisk area (Greenland) and a low-risk area (Denmark)—a second look. Br J Cancer 1993; 67:830–7. 13. Franco EL, Villa LL, Ruiz A, Costa MC. Transmission of cervical human papillomavirus infection by sexual activity: differences between low and high oncogenic risk types. J Infect Dis 1995; 172:756–63. 14. Richardson H, Franco E, Pintos J, Bergeron J, Arella M, Tellier P. Determinants of low-risk and high-risk cervical human papillomavirus infections in Montreal university students. Sex Transm Dis 2000; 27: 79–86.

JID 2002;185 (1 January)

Determinants of HPV infection

15. Hong Kong Cancer Registry. Cancer incidence and mortality in Hong Kong 1995– 1996. Hong Kong: Hospital Authority, 1996:28. 16. Shibata D. PCR-based detection of the herpesviruses and the human papillomaviruses. In: Ehrlich GD, Greenberg SJ, eds. PCR-based diagnostics in infectious disease. Boston: Blackwell Scientific Publications, 1994: 475–6. 17. Saiki RA, Bugawan TL, Horn GT, Mullis KB, Erlich HA. Analysis of enzymatically amplified b-globin and HLA DQ alpha DNA with allele specific oligonucleotide probes. Nature 1986; 324:163– 6. 18. Ting Y, Manos MM. Detection and typing of genital human papillomavirus. In: Innis MA, Gelfond DH, Sninsky JJ, White TJ, eds. PCR protocol: a guide to methods and applications. San Diego: Academic Press, 1990:356–7. 19. Bauer HM, Ting Y, Greer CE, et al. Genital human papillomavirus infection in female university students as determined by a PCR-based method. JAMA 1991; 265:472–7. 20. Bernard HU, Chan SY, Manos MM, et al. Identification and assessment of known and novel human papillomaviruses by polymerase chain reaction amplification, restriction fragment length polymorphisms, nucleotide sequence, and phylogenetic algorithms. J Infect Dis 1994; 170:1077– 85. 21. Kwok S, Higuchi R. Avoiding false positives with PCR. Nature 1989; 339:237–8. 22. Chan PK, Li WH, Chan MY, Ma WL, Cheung JL, Cheng AF. High prevalence of human papillomavirus type 58 in Chinese women with cervical cancer and precancerous lesions. J Med Virol 1999; 59:232– 8. 23. Svare EI, Kjaer SK, Worm AM, et al. Risk factors for HPV infection in women from sexually transmitted disease clinics: comparison between two areas with different cervical cancer incidence. Int J Cancer 1998; 75:1–8.

35

24. Herrero R, Hildesheim A, Bratti C, et al. Population-based study of human papillomavirus infection and cervical neoplasia in rural Costa Rica. J Natl Cancer Inst 2000; 92:464–74. 25. Jacobs MV, Walboomers JM, Snijders 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; 87:221–7. 26. Hildesheim A, Schiffman MH, Gravitt PE, et al. Persistence of type-specific human papillomavirus infection among cytologically normal women. J Infect Dis 1994; 169:235– 40. 27. Ho GY, Bierman R, Beardsley L, Chang CJ, Burk RD. Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med 1998; 338:423– 8. 28. Elfgren K, Kalantari M, Moberger B, Hagmar B, Dillner J. A populationbased five-year follow-up study of cervical human papillomavirus infection. Am J Obstet Gynecol 2000; 183:561–7. 29. Deacon JM, Evans CD, Yule R, et al. Sexual behaviour 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; 83:1565–72. 30. Kjaer SK, van den Brule AJ, Bock JE, et al. Determinants for genital human papillomavirus (HPV) infection in 1000 randomly chosen young Danish women with normal Pap smear: are there different risk profiles for oncogenic and nononcogenic HPV types? Cancer Epidemiol Biomarkers Prev 1997; 6:799–805. 31. Szarewski A, Cuzick J. Smoking and cervical neoplasia: a review of the evidence. J Epidemiol Biostat 1998; 3:229–56.