Human papillomavirus (HPV) infection is common, with an estimated worldwide prevalence of cervical HPV infection of 10%.1 HPV has been established to be a necessary, but not sufficient cause of cervical cancer, as only a fraction of infections progress to cervical cancer.2 The majority of infections are asymptomatic and transient, with most new infections no longer detectable within 1 to 2 years.3 Viral persistence of high-risk (HR) or carcinogenic HPV (particularly HPV-16 and HPV-18) longer than 1 year strongly predicts cervical precancer.4
Understanding the natural history of HPV is critical for the evaluation of public health initiatives such as the integration of HPV vaccination into cervical cancer prevention programs. Despite this, there have been a limited number of studies focusing on incidence and persistence in Canada.5,6 Little research has been conducted on Aboriginal populations, which represent an important target for public health initiatives because of their high risk of cervical cancer.7 Across in the United States, Australia, and Canada, cervical cancer incidence and mortality have been shown to be higher among Aboriginal females compared with non-Aboriginal females.8–11
Nunavik is the arctic and subarctic region of Quebec, where almost 90% of the approximately 11,000 residents who populate its 14 villages self-identify as Inuit.12 Among the Canadian Inuit, the age-standardized incidence rate of cervical cancer between 1989 and 2003 was 14.7 per 100,000, approximately 3 times higher than the Canadian average.7 A school-based HPV vaccination program was implemented in the province of Quebec, including Nunavik, in the fall of 2008. The quadrivalent vaccine Gardasil was offered to girls aged 10 to 18 years in Nunavik in the first year and girls aged 10 years (fourth grade) in subsequent years.13 Before the introduction of the HPV vaccine, a cohort of Inuit women from Nunavik was formed to study the natural history of HPV in this population. Among this population, the baseline prevalence of any HPV and HR-HPV was 29% and 20%, respectively, suggesting that the risk of HPV is elevated compared with other non-Aboriginal Canadian populations.14 It was estimated that 30% of this population had not been screened for cervical cancer in the previous 3 years and 6.5% had abnormal cytology at enrollment. To further understand the natural history of HPV among this high-risk population, this study aims to determine the incidence and persistence of HPV infection, and the predictors of HPV infection acquisition and clearance among this cohort of Inuit women in Nunavik, Quebec.
A prospective cohort of Inuit women living in communities on Ungava Bay and Hudson Bay in Nunavik, Québec, was recruited between January 2002 and December 2007 and followed up until April 2009. Details on recruitment, eligibility, and data collection have been described previously.14,15 Briefly, 621 Inuit women aged 15 to 69 years presenting for a Papanicolaou (Pap) test at community health clinics with an intact uterus and no diagnostic suspicion of cervical cancer were recruited by nurse practitioners. In addition to fulfilling cohort eligibility, women had to have completed a baseline questionnaire, have at least 2 HPV-DNA test results of acceptable quality, and not have withdrawn their consent at any time during follow-up to be eligible for this analysis of HPV infection incidence and persistence. All eligible women provided written informed consent via a standardized consent form; any woman wishing to withdraw from the study was able to do so at any time. Ethics approval was obtained from both the McGill University Institutional Review Board and the Tulattavik Health Centre.
Follow-up of the cohort in this study was passive, based on indications for cervical cancer screening in the local communities; no preset testing interval was used. Specimens for HPV-DNA typing were collected at any subsequent, regularly scheduled clinic visit for which a gynecologic examination was warranted.
At the time of enrollment, a nurse practitioner administered a baseline questionnaire and a medical chart review was conducted by the research team. Cervical specimens were collected at each visit requiring a Pap test, including at entry into the study. Ectocervical and endocervical cells were collected with a Dacron swab and used to prepare a Pap smear and for HPV-DNA typing.
The protocol for cell preparation and HPV-DNA testing by polymerase chain reaction amplification using PGMY09/11 consensus primers and quality-controlled reverse line blot assay (Roche Diagnostics) has been described elsewhere.14 Twenty-seven genital HPV genotypes (6, 11, 16, 18, 26, 31, 33, 35, 39, 40, 42, 44, 45, 51, 52, 53, 54, 56, 57, 58, 59, 66, 68, 73, 82, 83, and 84) were identified by oligonucleotide hybridization. In April 2004, an extended line blot was introduced into the study to probe for 10 additional HPV genotypes (34, 61, 62, 67, 69, 70, 71, 72, 81, and 89). Standard precautions were taken to avoid contamination and yielded a high adequacy of samples for analysis.14
Human papillomavirus types were classified as either HR or low-risk (LR) based on their oncogenic potential. Probable and possible HR genotypes were classified with established HR genotypes (16, 18, 26, 31, 33, 35, 39, 45, 51, 52, 53, 56, 58, 59, 66, 67, 68, 69, 70, 73, and 82), and genotypes with unknown risk were classified with established LR genotypes (6, 11, 34, 40, 42, 44, 54, 61, 62, 64, 71, 72, 81, 83, 84, and 89).16 Human papillomavirus types were also classified according to species of the Alphapapillomavirus (α) genus. Of particular interest were species α7 (HPV-18 and related types), α9 (HPV-16 and related types), and α3/15 (LR types).
The incidence rates and their corresponding Fisher exact 95% CIs were calculated for overall HPV infection and HPV infection by oncogenic risk grouping, and Alphapapillomavirus species as the number of new cases per 1000 women-months (WM). Person-time was calculated for each participant based on the number of months from enrollment to either a first positive test result for the infection of interest or the end of follow-up, defined as the date of the most recent HPV-DNA test, if the woman remained negative for the infection of interest. Participants did not contribute person-time to the incidence rate calculation for an infection for which they tested positive at enrollment, but contributed to the overall incidence rate if they tested positive for a different HPV type during follow-up.
The time of acquisition was assumed to occur halfway between the date of the first positive HPV-DNA test result and the date of the preceding (negative) HPV test result. Time of clearance was also estimated in a conservative fashion, assumed to occur at halfway point between the date of the last positive HPV-DNA test result and the date of the subsequent negative test result. Participants who failed to clear an incident infection before the last available result were censored at the date of the last study visit.
The mean and median duration (in months) of HPV infection was calculated for overall HPV infection and by HR, LR, α3/15, α7, and α9 groupings. The HPV infection persistence by oncogenic grouping and species was calculated 2 ways. First, the overall persistence was calculated by using the number of months during which a woman tested positive, regardless of the HPV type. The second method took into account the type of HPV for which a woman tested HPV-DNA positive. In this case, persistence was calculated based on the longest enduring type-specific HR or LR infection present in the initial incident positive result (termed type-specific HR/LR infections). These 2 calculations were also used for species-specific persistence. Time-to-clearance curves and median and mean persistence times (with corresponding 95% CIs) were generated for overall HPV infection and HR-HPV infection. Incidence and persistence curves were created in SAS 9.3 using the CIF macro and PROC LIFETEST, respectively.
Relationships between all covariates were explored for collinearity through cross-tabulations of categorical variables and correlation matrices of continuous variables. Covariates included age, marital status, employment status, education, smoking status, use of birth control, type of birth control, age at first sexual intercourse, number of lifetime sexual partners, number of live births, and self-reported history of sexually transmitted infections (STIs). All covariates were measured at baseline. The potential associations between all independent variables and each outcome (any incident HPV infection, any incident HR-HPV infection, and clearance of any incident HPV infection regardless of HPV type) were explored using univariate and age-adjusted (data not shown) unconditional logistic regression. Odds ratios (ORs) and their associated 95% CIs were calculated for each outcome. Model selection for the fully adjusted model was then performed using the Bayesian Information Criterion (BIC), where the model with the smallest BIC was chosen as the best model. The BMA package for R version 2.9.1 was used for the multivariate analyses. Multiple imputations for missing covariates were carried out using the command “mi impute” in STATA version 11. The results of the complete case analysis and the imputed analysis did not differ greatly, and therefore, only the results of the imputed analysis were reported.
A total of 416 women recruited into the original cohort (n = 621) had completed the baseline questionnaire and had at least 2 adequate HPV-DNA test results and were therefore eligible to be included in this analysis of acquisition and persistence of incident HPV infections. Women returned an average of 3 times (range, 2–12), resulting in a total of 1457 adequate HPV-DNA test results and 15612.1 months of follow-up. The median total follow-up per woman was 36.3 (range, 1.95–79.5) months. The interval between most visits was 1 year (median, 12.5 months); however, the interval between visits ranged from less than 30 days (n = 10) to more than 5.5 years. The sociodemographic characteristics of the study population are shown in Table 1. The mean (SD) age of the population at baseline was 32 (11.15) years.
HPV Infection Incidence
Overall, 39.9% (166/416) of the participants acquired an incident infection at some point during follow-up. The incidence rate of any new infection was 14.44 per 1000 WM. The cumulative incidence of any HPV type at 12 months was 10.3% (95% CI, 7.9–13.1). The HR-HPV incidence rate was 8.63 per 1000 WM (95% CI, 6.89–10.67), and the LR-HPV incidence rate was 6.13 per 1000 WM (95% CI, 4.81–7.65; Table 2). Species-specific incidence rates and incident rates of the 4 HPV types in the quadrivalent vaccine are shown in Table 2.
A U-shaped age-specific curve was observed for incidence of both any HPV infection and HR-HPV infection (Fig. 1). The highest incidence was found among women 15 to 19 years of age at baseline, followed by a decreasing incidence with increasing age. It seems that a second peak of incidence occurs among women 60 years and older at baseline, although it was not statistically significant. Figure 2 displays the cumulative incidence curves shown for any-HPV and HR-HPV.
Overall, 36.1% of participants cleared their newly acquired infections. The mean and median duration of overall incident HPV infection were 34.79 months (95% CI, 29.50–40.08 months) and 25.78 months (95% CI, 20.01–33.48 months), respectively. The mean persistence time (i.e., time to infection clearance) for an incident type-specific HR-HPV infection was longer than that for an incident type-specific LR-HPV infection (Table 3). The infection survival curves displaying the persistence of any-HPV infection and HR-HPV infection are shown in Figure 3. Persistence of overall infection was greater than type-specific infection for α3/15 species and similar across infections of α7 and α9 species (Table 3). However, there was no significant difference between the duration of infections across the species.
Determinants of HPV Infection Acquisition
In the univariate analysis, single marital status at baseline and higher educational attainment were significantly associated with HPV infection acquisition irrespective of genotype, whereas older age, older age at first sexual intercourse, a higher number of lifetime deliveries, and being employed were significant protective factors (Table 4). Similarly higher educational achievement and single marital status were also significantly associated with incident HR-HPV infections, and older age, being employed, and more than 3 lifetime deliveries were significantly protective in the univariate analysis (Table 4). A higher number of lifetime sexual partners were also found to be protective against HR-HPV infection; however, it did not maintain significance when adjusted for age (data not shown).
Based on the lowest BIC, only age and marital status were included as independent variables in the final model for any incident HPV infection. In this model, older age maintained its protective effect (OR, 0.92; 95% CI, 0.9–0.95) and single marital status maintained its association with HPV infection (OR, 2.04; 95% CI, 1.30–3.19). Age (OR, 0.93; 95% CI, 0.91–0.96) was the only covariate selected for the final model incident HR-HPV infection.
Determinants of HPV Infection Clearance
None of the sociodemographic, lifestyle, sexual, or medical factors were found to be significantly associated with clearance of any incident HPV infections in the univariate or age-adjusted (not shown) analysis (Table 4). There was a nonsignificant trend toward increased clearance with increasing age and having a self-reported history of STI.
The overall incidence proportion of HPV infection in this population was 39.9% across all ages. This estimate of overall incidence proportion is higher than what was reported for a representative screening population from various Ontario health regions,6 but similar to an HR population of Canadian university students.5 Although the HR-HPV and LR-HPV incidence rates among our study population were lower when compared with rates found among the student population, the age-specific incidence among Inuit women in the younger age group was higher. The species-specific incidence rates found in this study are almost equal to those found in a study of cytologically normal Hawaiian women,17 although in that population, α3 infections were acquired at the highest rate.
Studies have reported a decline in HPV incidence with age, followed by an increase in older women. Studies of Colombian women18 and women from the former Soviet Union19 both reported this age pattern of infection acquisition. In our study, a second peak in women aged 60 years due to HR-HPV infection rate was observed. Given the small number of older women in this sample, statistical differences were not identified among these age-specific incidence rates. It has been hypothesized that this can be explained by a cohort effect,20 reactivation of latent infections,21,22 or acquisition of new infections from sexual contact with new partners later in life.3 The age-specific incidence rates for our study population suggest that the peak in prevalence previously reported in older groups14 is at least partially the result of newly detected infections. This is consistent with the association found between prevalence and sexual activity markers in this older group.14
Of the women in the study population who acquired a new infection during follow-up, 36.1% cleared the infection by the last study visit. This is lower than what has been found in other Canadian populations, where approximately 50% of infections were cleared within 1 year.5,6 The present study's estimates of median duration are higher than those reported previously for any-type HPV infections5,17,23,24 and HR- and LR-HPV.5,17,18,25 Our sample included women with abnormal cytology at baseline (6.5%), which may partly explain the longer infection duration we found compared with studies that included only women with normal cytology.17,18 However, compared with a population with a high proportion (64%) of women with abnormal cytology, the median type-specific duration of HR infection was 7 months longer in our sample.24 For type-specific infections in our population, the median duration of type-specific HR-HPV infections was found to be longer than that of type-specific LR-HPV infections, which is consistent with previous research.3,5,17,18,25 Only one other study17 reported species-specific persistence analyses. Infection with HPV species α3, α7, and α9 were found to clear much more quickly in Hawaiian women with normal cytology than in our study.
It is surprising that this study found substantially greater infection persistence compared with other population given (1) that only (presumed) incident infections were included in our analyses, whereas many other studies have used mixed incident and prevalent infections, and (2) that follow-up was incomplete for many women, suggesting that our estimates are likely to be underestimations because some infections are presumed to continue to persist beyond the last available study visit. The median follow-up time in our study was shorter than one other study18 that found a shorter average duration of infection. Two other studies6,19 failed to report median follow-up time. Inuit women in Québec seem, therefore, to be at much greater risk for persistent HPV infections across all infection classifications discussed here, which may help explain the higher rate of cervical cancer found among this population.
In multivariate models, variables that showed a significant association with incident any-type HPV infections were age and single marital status and only age was associated with incident HR-HPV in the final model. These results are consistent with the literature, as many studies have shown age and markers of sexual activity to be consistent risk factors for HPV.6,18,19,26,27 In this population, there was no evidence of an association between incident HPV infection and the use of oral contraceptives or smoking. Although these have been identified as risk factors previously in other populations,6 our null result may relate to the measurement of covariates at baseline, the high proportion of ever smoking in this population, and insufficient power. Interestingly, more than 10 lifetime sexual partners at baseline were found to be protective of HPV infection in the univariate analysis, but did not maintain a significant association when adjusted for age. This finding should be interpreted with caution because number of sexual partners was crudely dichotomized (≥10 or < 10) and sexual partners was only measured at baseline, so it did not capture women's sexual activity before their incident infection.
No significant associations were found between sociodemographic and lifestyle factors and clearance of any-type HPV infections in either the univariate or age-adjusted analyses, which likely reflects the low number of cleared infections in this population during follow-up. Factors associated with HPV persistence have been less frequently studied, but there is evidence that viral load,3,28 infection with multiple types,29 molecular variant type,3 and factors such as older age,30 lifetime pregnancies,17 and certain dietary factors31,32 may be associated with persistence.
Limitations of this study included a relatively small study population, nonrandom recruitment, and missing data. Despite the nonrandom recruitment of the study population, the study population has been shown to be representative of the general population in terms of age distribution,12 education,12 and marital status33 and captured more than 40% of the target population. Furthermore, only 313 (75%) of the 416 study participants had complete information on all variables used in the analysis. However, missing data were not regarded as a major limitation, given that those with and without missing data had similar demographic characteristics and the complete-case analysis was similar to the imputed analysis. Only a baseline questionnaire was administered, as it was not feasible to readminister the questionnaire at each subsequent visit. The associations between these baseline characteristics and subsequent acquisition or persistence of HPV infections must, therefore, be interpreted with caution and most likely serve as hypothesis-generating information for future studies.
This study is the first known analysis of HPV incidence, persistence, and determinants of HPV acquisition and clearance among Inuit women. These data, when taken with previously published prevalence data, provide further insight into the burden and patterns of HPV infection in this population at high risk for cervical cancer, thus filling an existing gap in knowledge regarding HPV infection in important population subgroups in North America. The results of this study may also be used to help evaluate vaccination strategies currently used for this HR population.
1. de Sanjose S, Diaz M, Castellsague X, et al. Worldwide prevalence and genotype distribution of cervical human papillomavirus DNA in women with normal cytology: A meta-analysis. Lancet Infect Dis
2007; 7: 453–459.
2. Walboomers JM, Jacobs MV, Manos MM, et al. Human papillomavirus is a necessary cause of invasive cervical cancer worldwide. J Pathol
1999; 189: 12–19.
3. Trottier H, Franco EL. The epidemiology of genital human papillomavirus infection. Vaccine
2006; 24(suppl 1): S1–S15.
4. Castle PE, Rodríguez AC, Burk RD, et al. Short term persistence of human papillomavirus and risk of cervical precancer and cancer: Population based cohort study. BMJ
2009; 339: b2569.
5. Richardson H, Kelsall G, Tellier P, et al. The natural history of type-specific human papillomavirus infections in female university students. Cancer Epidemiol Biomarkers Prev
2003; 12: 485–490.
6. Sellors JW, Karwalajtys TL, Kaczorowski J, et al. Incidence, clearance, and predictors of human papillomavirus infection in women. CMAJ
2003; 168: 421–425.
7. Circumpolar Health Inuit Cancer Review Working Group, Kelly J, Lanier A, et al. Cancer among the circumpolar Inuit, 1989–2003. II. Patterns and trends. Int J Circumpolar Health
2008; 67: 408–420.
8. Luke C, Nguyen AM, Heard A, et al. Benchmarking epidemiological characteristics of cervical cancer in advance of change in screening practice and commencement of vaccination. Aust N Z J Public Health
2007; 31: 149–154.
9. Marrett LD, Chaudhry M. Cancer incidence and mortality in Ontario First Nations, 1968–1991 (Canada). Cancer Causes Control
2003; 14: 259–268.
10. Healey SM, Aronson KJ, Mao Y, et al. Oncogenic human papillomavirus infection and cervical lesions in aboriginal women of Nunavut, Canada. Sex Transm Dis
2001; 28: 694–700.
11. Espey DK, Wu XC, Swan J, et al. Annual report to the nation on the status of cancer, 1975–2004, featuring cancer in American Indians and Alaska Natives. Cancer
2007; 110: 2119–2152.
14. Hamlin-Douglas LK, Coutlee F, Roger M, et al. Prevalence and age distribution of human papillomavirus infection in a population of Inuit women in Nunavik, Quebec. Cancer Epidemiol Biomarkers Prev
2008; 17: 3141–3149.
15. Hamlin-Douglas LK, Coutlee F, Roger M, et al. Determinants of human papillomavirus infection among Inuit women of northern Quebec, Canada. Sex Transm Dis
2010; 37: 377–381.
16. Bouvard V, Baan R, Straif K, et al. A review of human carcinogens—Part B: Biological agents. Lancet Oncol
2009; 10: 321–322.
17. Goodman MT, Shvetsov YB, McDuffie K, et al. Prevalence, acquisition, and clearance of cervical human papillomavirus infection among women with normal cytology: Hawaii Human Papillomavirus Cohort Study. Cancer Res
2008; 68: 8813–8824.
18. Munoz N, Méndez F, Posso H, et al. Incidence, duration, and determinants of cervical human papillomavirus infection in a cohort of Colombian women with normal cytological results. J Infect Dis
2004; 190: 2077–2087.
19. Syrjanen S, Shabalova I, Petrovichev N, et al. Age-specific incidence and clearance of high-risk human papillomavirus infections in women in the former Soviet Union. Int J STD AIDS
2005; 16: 217–223.
20. 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: 464–474.
21. Lazcano-Ponce E, Herrero R, Munoz N, et al. Epidemiology of HPV infection among Mexican women with normal cervical cytology. Int J Cancer
2001; 91 (3): 412–420.
22. Castle P, Jeronimo J, Schiffman M, et al. Age-related changes of the cervix influence human papillomavirus type distribution. Cancer Res
2006; 66: 1218–1224.
23. Woodman CB, Collins S, Winter H, et al. Natural history of cervical human papillomavirus infection in young women: A longitudinal cohort study. Lancet
2001; 357: 1831–1836.
24. Kulmala SM, Shabalova IP, Petrovitchev N, et al. Type-specific persistence of high-risk human papillomavirus infections in the New Independent States of the former Soviet Union Cohort Study. Cancer Epidemiol Biomarkers Prev
2007; 16: 17–22.
25. 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.
26. Giuliano AR, Harris R, Sedjo RL, et al. Incidence, prevalence, and clearance of type-specific human papillomavirus infections: The Young Women's Health Study. J Infect Dis
2002; 186: 462–469.
27. Velicer C, Zhu X, Vuocolo S, et al. Prevalence and incidence of HPV genital infection in women. Sex Transm Dis
2009; 36: 696–703.
28. Munoz N, Hernandez-Suarez G, Méndez F, et al. Persistence of HPV infection and risk of high-grade cervical intraepithelial neoplasia in a cohort of Colombian women. Br J Cancer
2009; 100: 1184–1190.
29. Ho GY, Bierman R, Beardsley L, et al. Natural history of cervicovaginal papillomavirus infection in young women. N Engl J Med
1998; 338: 423–428.
30. Castle PE, Schiffman M, Herrero R, et al. A prospective study of age trends in cervical human papillomavirus acquisition and persistence in Guanacaste, Costa Rica. J Infect Dis
2005; 191: 1808–1816.
31. Giuliano AR, Siegel EM, Roe DJ, et al. Dietary intake and risk of persistent human papillomavirus (HPV) infection: The Ludwig-McGill HPV Natural History Study. J Infect Dis
2003; 188: 1508–1516.
32. Richardson H, Abrahamowicz M, Tellier PP, et al. Modifiable risk factors associated with clearance of type-specific cervical human papillomavirus infections in a cohort of university students. Cancer Epidemiol Biomarkers Prev
2005; 14: 1149–1156.
© Copyright 2015 American Sexually Transmitted Diseases Association
33. Anctil M. Survey Highlights. Nunavik Inuit Health Survey 2004, Qanuippitaa? How Are We?
Quebec: Institut national de sant‚ publique du Québec & Nunavik Regional Board of Health and Social Services; 2008.