Objective Cervical cancer is caused by carcinogenic human papillomavirus (HPV) infections. Prior to the introduction of HPV vaccination in Suriname, we performed a cross-sectional study to estimate the prevalence of and determinants for genital carcinogenic HPV infections.
Methods Women were recruited at a family planning (FP) clinic and a sexually transmitted infections (STI) clinic. Vaginal swabs were used for HPV genotyping by the SPF10 PCR-DEIA-LiPA25 system. Logistic regression was used to identify determinants for carcinogenic HPV infection.
Results The prevalence of any HPV was 54.2% and of carcinogenic HPV was 27.9% among 813 women attending the FP clinic. Among the 188 women attending the STI clinic, the prevalence of any HPV (76.1%) and of carcinogenic HPV (40.4%) was significantly higher. HPV52 was the most prevalent genotype in both clinics. The prevalence of HPV16 and/or 18 was 6.4% in the FP clinic and 12.2% in the STI clinic. The following determinants were independently associated with carcinogenic HPV infection among women visiting the FP clinic: ≥2 recent partners (OR 1.53; 95% CI 1.13 to 2.06), Chlamydia trachomatis co-infection (OR 1.89; 95% CI 1.32 to 2.70), disassortative ethnic sexual mixing (OR 1.50; 95% CI 1.13 to 1.99) and ethnic group (OR 1.90; 95% CI 1.27 to 2.85 for Creole and OR 1.67; 95% CI 1.06 to 2.62 for mixed ethnicity, both compared with Hindustani). No independent determinants were found among women visiting the STI clinic.
Conclusions Carcinogenic HPV is highly prevalent among women in Suriname, and not equally distributed among ethnic groups. These data provide a baseline to assess possible shifts in the prevalence of HPV genotypes following vaccination.
- Cervical Neoplasia
- Chlamydia Infection
- Epidemiology (Clinical)
Statistics from Altmetric.com
A persistent infection with the human papillomavirus (HPV) is considered a necessary cause of cervical cancer.1 The International Agency for Research on Cancer (IARC) has classified mucosal genotypes 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, and 59 as carcinogenic (Class 1), type 68 as probably carcinogenic (Class 2A) and types 26, 30, 34, 53, 66, 67, 69, 70, 73, 82, 85 and 97 as possibly carcinogenic (Class 2B). 2 Other types were not classifiable as to their carcinogenicity (eg, HPV6 and 11) or were considered non-carcinogenic (Class 3). HPV-based strategies for prevention of cervical (pre)cancer are vaccination by a bivalent HPV16/18 vaccine or a quadrivalent HPV6/11/16/18 vaccine3 and HPV testing in screening and triage.4
In Suriname, a Caribbean country on the South American continent, age-standardised incidence and mortality rates for cervical cancer were 27.2 and 11.1 per 100 000 women, respectively, in 2008.5 Suriname has not yet implemented a cervical cancer screening programme. Vaccination of girls aged 9–13 with the quadrivalent vaccine was started in November 2013.
Suriname has a post-colonial population with a variety of ethnicities. The five major ethnic groups are Creole and Maroon (17.7% and 14.7%, respectively; both descendants of African diaspora due to the slave trade, the latter descendants from fugitive slaves settled in the interior forest), Hindustani and Javanese (27.4% and 14.6%, respectively; both descendants of labour immigrants from the former British Indies and Dutch Indies, respectively), and mixed ethnicity (12.5%). Other ethnicities are Chinese, indigenous Amerindians and Caucasians.6
The prevalence of carcinogenic HPV infection among Surinamese women with normal cytological findings was estimated at 80% in one small study (n=100).7 HPV positivity was 35.4% in the Caribbean, 15.3% in South America and 11.7% globally in a recent meta-analysis among 1 million women with normal cytological findings.8 Variations in HPV prevalence are mainly due to differences in study population characteristics (eg, sexual behaviour, age), geographical region and (type-specific) sensitivity of the HPV test used.
In addition, ethno-demographic and sexual behaviour risk factors for carcinogenic HPV infections in Suriname have not yet been defined. A recent study showed that the prevalence of Chlamydia trachomatis differed between ethnic groups in Suriname.9 The same might be true for the prevalence of carcinogenic HPV infection, as ethnic disparities in the prevalence of premalignant cervical lesions have been described.10 In addition, the prevalence of squamous atypia and mild dysplasia is higher among Maroon women than in Javanese, Amerindian and Hindustani women.11
The current study aimed to estimate the pre-vaccination prevalence of carcinogenic HPV infections among women from five major ethnic groups in Suriname and to identify determinants for carcinogenic HPV infections, such as ethno-demographic and sexual behaviour characteristics and concurrent infection with C. trachomatis.
Study design and population
A cross-sectional study was designed to assess the prevalence of and determinants for genital carcinogenic HPV infections among women from five major ethnic groups in Suriname. Women were interviewed about demographic characteristics and sexual behaviour, and vaginal swabs were used for HPV genotyping. Logistic regression analysis was performed to assess the determinants of carcinogenic HPV infection.
Participants had been recruited at two sites in Paramaribo, Suriname, as part of a previous study9:
The Dermatological Service, an integrated outpatient sexually transmitted infections (STI) clinic frequented by men and women, offering free-of-charge examination and treatment of STIs and infectious skin diseases. All women who visited for a STI check-up were invited to participate in the study. Women recruited at this clinic comprise a population that are expected to be at high risk for carcinogenic HPV infection due to sexual behaviour.
The Lobi Foundation, a family planning (FP) clinic frequented by women only. All consecutive women visiting the clinic were invited to participate in the study. Women recruited at this clinic are expected to be more representative for the general sexually active Surinamese population and therefore considered a ‘low-risk’ population.
Recruitment took place between July 2009 and April 2010. Exclusion criteria were age <18 years and previous participation in the study. A nurse interviewed participants about demographic characteristics (including self-reported ethnicity) and sexual behaviour.
Specimen collection and testing procedures
Nurses obtained vaginal swabs using the Aptima Vaginal Swab Specimen Collection Kit (Hologic Gen-Probe, San Diego, USA) prior to routine speculum examination. Details regarding collection, storage, shipment of specimens and testing for C. trachomatis have been described previously.9
A 200 µL aliquot of remaining vaginal swab specimen was tested at the DDL Diagnostic Laboratory, Rijswijk, The Netherlands for the presence of (carcinogenic) HPV. DNA was extracted using a MagNA Pure LC instrument (Roche Diagnostics, Almere, The Netherlands) and a MagNA Pure LC Total Nucleic Acid Isolation Kit (Roche Diagnostics) and collected in 100 µL elution buffer.
The algorithm HPV SPF10 PCR-DEIA (DNA enzyme immunoassay)-LiPA25 (Line probe assay) V.1 (Labo Biomedical Products, Rijswijk, The Netherlands) was used for highly sensitive HPV testing, according to the kit manual. Briefly, 10 μL of extracted DNA was used to amplify a 65 bp fragment of the viral L1 open reading frame by SPF10 primers. A DEIA was used for broad-spectrum detection of HPV using a cocktail of universal probes. The SPF10 DEIA can detect at least HPV2, 3, 4, 5, 6, 7, 8, 11, 13, 14, 16, 18, 20, 26, 27, 28, 30, 31, 32, 33, 34, 35, 37, 39, 40, 42, 43, 44, 45, 51, 52, 53, 54, 55 (re-classified as a subtype of HPV44), 56, 57, 58, 59, 61, 62, 64 (re-classified as a subtype of HPV34), 65, 66, 67, 68, 69, 70, 71, 72, 73, 74, 75, 76, 81, 82, 83, 84, 85, 86, 87, 89, 90, 91, 95, 97, 102, 106, 114 and 115. Detection of the underlined types has been validated by Kleter et al12 while detection of the other types is based on additional sequence analyses13 ,14 and unpublished data.
Subsequently, SPF10 DEIA-positive amplimers were analysed by SPF10 LiPA25 in a ProfiBlot 48 analyser (Tecan Austria GmbH, Salzburg, Austria) to identify 25 anogenital HPV genotypes.15 These genotypes were categorised according to the IARC classification2—that is, HPV16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58 and 59 as carcinogenic (Class 1), HPV34, 53, 66, 68/73 and 70 as probably/possibly carcinogenic (Class 2A/B) and HPV6, 11, 40, 42, 43, 44, 54 and 74 as non-carcinogenic (Class 3).
We assumed that the women recruited at the STI clinic and FP clinic constituted two distinctive populations in terms of sexual behaviour and the prevalence of carcinogenic HPV and we analysed them separately. The χ2 test for independence and the rank sum test were used to examine differences in demographic and behavioural characteristics and HPV prevalence.
To assess determinants of carcinogenic HPV infection, we performed univariable and multivariable logistic regression analysis. Parameters were estimated using generalised estimating equations (GEE) to account for multiple observations within the same individual. An exchangeable correlation structure was assumed. In univariable analysis the effect of the following variables was examined: age, education, ethnicity, difference in ethnicity of sexual partners as perceived by the participant (ie, ethnic sexual mixing), condom use, number of sexual partners in the preceding month, number of partners in the preceding 12 months and having had sex in exchange for money or goods. Age was divided into four categories for women visiting the STI clinic. The level of education of participants was qualified as ‘low’ (primary school and lower vocational training), ‘medium’, ‘high’ (polytechnic and university) or ‘unknown’. Because of the small numbers, we grouped women of Caucasian, Chinese and indigenous Amerindian ethnicity together in univariable and multivariable analyses. Condom use was classified into ‘always condom use’ and ‘never or inconsistent condom use’. Disassortative ethnic sexual mixing was defined as having had sex with at least one partner of another ethnicity (as perceived by the participant) in the preceding 12 months.
Variables that were associated with carcinogenic HPV infection at p≤0.1 in the univariable analysis were entered into a multivariable model. Individuals with missing data were omitted from the models. Young age has been established as a determinant for carcinogenic HPV infection, with maximum positivity rates observed in women less than 25 years old globally.16 Therefore, this determinant was forced into the multivariable model. HPV type was also forced into the multivariable model. To avoid multicollinearity, we only included the variable ‘number of partners in the preceding 12 months’ in the model and not the variable ‘number of partners in the preceding month’. We considered p<0.05 as statistically significant. We checked for interactions between the number of partners in the preceding 12 months and ethnic sexual mixing. SPSS package V.19.0 (SPSS, Chicago, Illinois, USA) and STATA software V.11.2 (Stata Corp, College Station, Texas, USA) were used for the analyses.
In total, 224 women at the STI clinic and 815 at the FP clinic were asked to participate in the study, of whom 36 and 2 women declined to participate, respectively. Reasons for decline were lack of time and lack of interest. Thus, 188 women at the STI clinic and 813 women at the FP clinic were included. Demographics and sexual behaviour of the study participants are shown in table 1 and differed significantly between the two study groups.
Among women attending the STI clinic, the median age was 28 years (IQR 23–34) and the majority had a low (38.3%) or medium (34.6%) level of education. In this group, 58 (30.9%) were of Creole ethnicity, 47 (25.0%) were mixed race, 37 (19.7%) were of Maroon ethnicity, 19 (10.1%) were of Hindustani ethnicity and 13 (6.9%) were of Javanese ethnicity. Women visiting the FP clinic had a median age of 31 (IQR 25–37) and had mostly medium (52.2%) or low (33.7%) education. Women in this group had the following ethnicities: Hindustani (27.6%), Creole (24.5%), Javanese (17.9%), mixed race (16.8%) and Maroon (10.5%). Women visiting the STI clinic reported higher risk behaviour for contracting STI than women visiting the FP clinic, such as >2 partners in the previous year (17.6% vs 6.3%; p<0.001) and sex in exchange for money or goods (9.7% vs 0.8%; p<0.001).
HPV test results are shown in table 2. HPV positivity, assessed by SPF10 PCR-DEIA, was 76.1% (95% CI 69.6% to 81.8%) among women visiting the STI clinic, which was significantly higher than the HPV positivity of 54.2% (95% CI 50.8% to 57.7%) among women visiting the FP clinic (p<0.001). The HPV positivity was also significantly higher among women attending the STI clinic than those frequenting the FP clinic when the comparison was restricted to carcinogenic HPV (40.4% vs 27.9%, p=0.001), probably/possibly carcinogenic HPV (19.1% vs 11.7%, p=0.006), HPV6/11/16/18 together (18.6% vs 9.8%, p=0.001), HPV16/18 together (12.2% vs 6.4%, p=0.006) and multiple HPV infections per sample (23.4% vs 17.0%, p=0.039). No difference was observed for the prevalence of non-carcinogenic types between the STI clinic and the FP clinic (20.2% vs 16.1%, p=0.256).
The HPV genotype was identified by LiPA25 in 118 of 188 (62.8%) women included in the STI clinic. In this group, HPV52 was the most prevalent type (9.0%), followed by HPV16 (7.4%) and HPV35 (7.4%), all characterised as established carcinogenic types (IARC Class 1; figure A in online supplement). The most prevalent types not included in IARC Class 1 were HPV6 (5.9%; Class 3), HPV66 (5.3%; Class 2B) and HPV68/73 (5.3%; Class 2A/2B) (figure B in online supplement). Multiple types were detected among 44 (23.3%) women, of whom 29 (66%) had two types, 9 (20%) had three types, 4 (9%) had four types and 2 (5%) had five types.
Among the 813 women visiting the FP clinic, LiPA25 identified the HPV genotype in 354 (43.5%) of them. In this population, HPV52 was also the most prevalent genotype (7.6%), followed by HPV51 (5.0%; Class 1) and HPV16 (3.9%). The most prevalent types not included in IARC Class 1 were HPV54 (3.9%; Class 3), HPV44 (3.6%; Class 3), HPV66 (3.6%; Class 2B), HPV68/73 (3.6%; Class 2A/2B) and HPV74 (3.6%; Class 3). Multiple types were detected among 138 (17.0%) women, of whom 94 (68%) had two types, 28 (20%) had three types, 11 (8%) had four types and 5 (4%) had five types.
Determinants of HPV infection
In univariable analysis of women visiting the STI clinic, carcinogenic HPV was only significantly associated with C. trachomatis infection. In multivariable analysis, no independent determinants for carcinogenic HPV were found.
Among women visiting the FP clinic, age, ethnic group, ethnic sexual mixing, number of partners in the preceding month, number of partners in the preceding 12 months, sex in exchange for money or goods and C. trachomatis co-infection were all significantly associated with carcinogenic HPV infection in univariable analysis. In multivariable analysis, carcinogenic HPV infection was significantly associated with age (OR 0.78; 95% CI 0.65 to 0.94 per 10 years increase); ethnic group (OR 1.90; 95% CI 1.27 to 2.85 for Creoles and OR 1.67; 95% CI 1.06 to 2.62 for mixed race, both compared with Hindustani); ethnic sexual mixing (OR 1.50; 95% CI 1.13 to 1.99 for disassortative mixing compared with assortative mixing only); number of partners in the preceding 12 months (OR 1.53; 95% CI 1.13 to 2.06 for having ≥2 partners compared with having ≤1 partner); and C. trachomatis co-infection (OR 1.89; 95% CI 1.32 to 2.70) (table 4).
There was no interaction between the number of partners in the preceding 12 months and ethnic sexual mixing in the final model (p=0.721).
The prevalence of genital infection with any HPV (54.2%) and of carcinogenic HPV (27.9%) was high among women attending a FP clinic in Suriname. This population might be considered representative of the general sexually active Surinamese population. The prevalence of any HPV and of carcinogenic HPV infection in women visiting a Surinamese STI clinic was significantly higher than in the FP clinic. This can be explained by the high-risk sexual behaviour among women visiting the STI clinic. The general HPV positivity rate among women recruited at the FP clinic (54.2%) is comparable to the rate among women in Costa Rica (50.7%) and Honduras (51.4%), two other countries in the Latin American and Caribbean region.17 ,18 The same ultrasensitive SPF10 PCR-DEIA-LiPA25 HPV test algorithm was used in both studies.
The HPV positivity rate in the low-risk FP clinic population is high compared with that of a low-risk population of unscreened and unvaccinated young women tested by the same algorithm in the Netherlands (22.5%).19 This is consistent with the observed regional differences in HPV prevalence among women with normal cytology, with the Caribbean region ranking highest worldwide.8
HPV52 was the most prevalent type observed in the Surinamese study population, followed by HPV51 and 16. These were also the three most frequently detected types in other studies in the world that used the same HPV test.17 ,19–21 One exception was a Honduran population, where HPV16, 18 and 51 were the most common.18 Despite the high prevalence of HPV genotypes 52 and 51, the most important types from a clinical point of view are HPV16, 18 and 45. The prevalence of these HPV types is generally increased in women with CIN3 and cervical carcinomas, and they confer an increased risk for cervical cancer.
The prevalence of types targeted by the bivalent HPV16/18 vaccine and the quadrivalent HPV6/11/16/18 vaccine was 6.4% and 9.8%, respectively. However, the potential impact of HPV16/18 vaccination in Suriname for cervical cancer prevention can only be estimated by determining the HPV prevalence in tumour sections of excised or biopsied cervical cancers. The presence of HPV16 or 18 was estimated at 55.4% in 130 Surinamese cervical cancer specimens and at 64.2% in 106 HPV-positive cases.22
Among women attending the FP clinic, carcinogenic HPV infection was significantly more common among women of Creole and mixed ethnicity than among Hindustani women. Ethnic sexual mixing was also an independent determinant for carcinogenic HPV infection. Variation in the prevalence of HPV genotypes and variant lineages among ethnic groups within a population has been observed in previous studies.23 ,24 This could be associated with differences in sexual behaviour, social class (poverty and malnutrition), parity, use of barrier contraceptive protection and use of steroidal contraception.23
Other independent determinants for carcinogenic HPV infection in the FP clinic study population were the number of sexual partners in the preceding 12 months and concurrent infection with C. trachomatis. The association between C. trachomatis and carcinogenic HPV infection is still debated. Although we adjusted for several sexual risk behaviour variables (such as number of partners in the preceding 12 months and sex in exchange for money or goods), residual confounding cannot be excluded as we were unable to adjust for other sexual behaviour factors such as sex techniques and partner characteristics. Some studies have reported that C. trachomatis infection is associated with an increased risk of cervical cancer,25–27 but another study could not confirm this after controlling for carcinogenic HPV-positive status.28 The authors speculated that previous reports of an association might be due to confounding by HPV status or by an increased susceptibility to HPV infection among women with a positive C. trachomatis status.
This study has several limitations. First, vaginal swabs, not suitable for cervical cytology, were used instead of cervical swabs. The absence of cytology results in this study population should be taken into consideration when comparing our data with HPV prevalence studies among women with normal cytology. However, HPV measured in vaginal swabs is generally representative of the HPV status at the cervix.29
Second, only a limited number of known risk factors for carcinogenic HPV infection were investigated. Other (non-sexual) risk factors such as smoking, parity and age of sexual debut30 were not studied.
Third, the number of participants from the STI clinic was lower (n=188) than the number of participants from the FP clinic (n=813). This could explain why no significant associations were found in the multivariable analysis of the STI clinic population.
Finally, women were not recruited from the general population as was done by, for example, Lenselink et al,19 but from outpatient clinics. Clearly, the women visiting the STI clinic were at high risk of having HPV infection. The women attending the FP clinic are likely to be more representative of the general sexually active Surinamese population and the determined HPV prevalence and determinants for HPV infection might be directly extrapolated.
In summary, we present the first data on the prevalence of genital HPV genotypes among women in Suriname in a pre-vaccination era. These results provide a useful baseline to assess possible shifts in HPV genotype prevalence following introduction of vaccination. In addition, independent determinants for carcinogenic HPV infection were the number of sexual partners in the preceding 12 months, concurrent infection with C. trachomatis, ethnic sexual mixing and ethnic group.
Pre-vaccination prevalence of genital carcinogenic human papillomavirus (HPV) genotypes was high among women in Suriname, a Caribbean country with an ethnically diverse population.
These baseline data allow changes in HPV genotype prevalence to be monitored following the introduction of HPV vaccination in 2013.
Independent determinants of HPV infection were the number of recent sexual partners, concurrent infection with Chlamydia trachomatis, ethnic sexual mixing and ethnic group.
The authors are grateful to all the nurses and laboratory technicians at the Dermatological Service and the Lobi Foundation for data collection and to Marga Kamp and Leonie van den Berg at DDL Diagnostic Laboratory for their technical assistance.
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Files in this Data Supplement:
- Data supplement 1 - Online figure
DTG and AWG contributed equally.
Handling editor Jackie A Cassell
Contributors DTG, AWG, JJvdH, MFSvdL, KDQ, LOAS and HJCdV conceived and designed the experiments. AWG and LOAS collected the clinical materials. DTG and JJvdH acquired the experimental data. DTG, AWG, JJvdH, MFSvdL, KQ and HJCdV performed the analysis and interpretation of the data. DTG, AWG, JJvdH, MFSvdL, KDQ, LOAS and HJCdV drafted the manuscript and provided their final approval.
Funding This study was funded by Stichting Pathologie Onderzoek en Ontwikkeling (SPOO), the Research and Development fund of the Public Health Service of Amsterdam (project no 2369 and 2371) and AGIS healthcare insurance (RVVZ no 1417000). The funders had no role in the study design, data collection and analysis, decision to publish or preparation of the manuscript.
Competing interests MSvdL received funding for a study on HPV from Sanofi Pasteur MSD and participated in a Merck-funded investigator-initiated study on Gardasil. The other authors declare that they have no competing interests.
Patient consent Patients participated anonymously and gave written informed consent.
Ethics approval The study was approved by the ethics committee of the Ministry of Health of the Republic of Suriname (VG010–2007, amended VG011–2012) and the ethics committee of the Academic Medical Center, University of Amsterdam, the Netherlands (MEC07/127).
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.