Article Text
PDF
Abstract
Background Human papillomavirus (HPV) is recommended as a primary screening tool for cervical screening. Assessment of age-specific performance of newer HPV careHPV DNA testing is important as risk of cervical intraepithelial neoplasia (CIN) varies at different ages. We aim to evaluate careHPV in comparison to Papanicolaou (Pap) test and visual inspection of the cervix with acetic acid (VIA) cervical screening tests for the detection of high-grade CIN.
Methods The cross sectional study was conducted in a rural population of North India. Ever-married women 30–59 years of age were invited for screening by careHPV (self-collected vaginal and physician-collected cervical samples), Pap test and VIA. Associations for trend in age for detecting histological-confirmed CINII+ and CINIII+ for each screening test were evaluated. Age-specific association with each screening test was evaluated.
Results Of a total of 7761 women invited, 5032 were screened and analysis was performed on 4658 with all screen test results. No significant (p>0.05) association of age for any screening test in the detection of CINII+ or CINIII+ was observed. For the older age group, cervical HPV (CHPV) showed high sensitivity and specificity for CINII+ detection. Specificity of CHPV or vaginal HPV (VHPV) was equal or higher than Pap in all age groups.
Conclusions Cervical screening options of CHPV or VHPV, or Pap, performed equally in the younger age group while CHPV might be an option for all ages in the detection of high-grade CIN.
- CANCER: CERVIX
- Cancer epidemiology
- SCREENING
Statistics from Altmetric.com
Introduction
Human papillomavirus (HPV) is an established causal agent in the process of cervical carcinogenesis.1 As the interpretation of Papanicolaou (Pap) smear is subjective and less sensitive, there is an increasing trend towards more objective HPV DNA methods of cervical screening.2 Cervical HPV (CHPV) DNA testing by Hybrid Capture (HCII) method along with the conventional methods of screening, such as Pap and visual inspection of cervix using acetic acid (VIA) for the detection of high-grade cervical intraepithelial neoplasia (CIN), is studied in different clinical settings.3–7 HPV testing is shown to be more sensitive and less specific than Pap testing for the detection of high-grade CIN.8 HPV testing on self-collected vaginal samples by women and cervical samples collected by healthcare provider found similar performances.9 ,10 In a systematic review and meta-analysis conducted to compare the detection rates of genital HPV infection in self-collected and provider-collected samples, it was concluded that self-sampling is as sensitive as provider-collected sampling to detect high-risk HPV.9 ,11 Liquid-based cytology (LBC) in comparison to self-sampling HPV testing as a primary cervical cancer screening in low-resource settings also resulted in comparable sensitivity between the methods in a pooled analysis.12 Recent studies on accuracy of HPV testing on self-collected versus clinician-collected samples13 ,14 recommended the sampling by clinicians for screening programmes, as the self-sampling approach was found to be less sensitive and specific.
As the risk of cervical cancer often differs substantially among generations, difference in detection of CIN among age groups in a cross-sectional study can be due to different risks among different generations. Thus it is important to study age-specific differences of CIN detection. Various studies reported age-specific performance of HPV testing by HCII method in primary cervical cancer screening.15–18 Sensitivity of HPV testing is uniformly high in all ages while that of Pap is substantially better in older women.16 It is observed that in HPV as well as Pap testing, specificities of screening increased with increasing age.12 ,16
A new variant of traditional HCII, called careHPV, developed to detect 14 high-risk types of cervical carcinogenic HPV, demonstrated careHPV as a promising primary screening method for cervical cancer prevention.19 Studies19–26 on a new careHPV testing on self-collected vaginal (vaginal HPV, VHPV) and physician-collected cervical samples (CHPV) also demonstrated better performance as compared to conventional screening methods. CareHPV demonstration is carried out for all ages combined in low-resources settings.20 ,21 Age-specific performance of careHPV test on self-collected and physician-collected samples in comparison with VIA and Pap has not been reported earlier. This study reports age-specific performances of careHPV tests in self-collected as well as health provider-collected samples in comparison to other conventional approaches of screening.
Methods
The present study from a North Indian rural community, performed between September 2010 and April 2012, was part of a multicentre cross-sectional study involving four centres in three countries, namely, India, Uganda and Nicaragua. A baseline survey and motivation of all women were conducted in the selected target population aged between 30 and 59 years. Health education material such as flip charts, information brochures and pamphlets were developed and used to motivate the women for screening. Auxiliary Nurse Midwife (ANM) and grass-root level workers of a community health centre (CHC) were trained on motivation, screening and counselling. The duration of training given was for 1 month on motivational aspects, consent taking, and screening aspects of theoretical and hands on training, and quality control on screening methods was carried out by assessing the agreement between the ANMs and medical doctor. Women who reported at screening centres were informed about the screening procedures. Baseline-demographic details were noted and written informed consent obtained prior to screening. More details of methodology have been reported elsewhere.20 ,21 Collection of vaginal careHPV samples by the women themselves was first in sequence at the screening clinic. An ANM assisted each woman on the procedure of collecting a self-vaginal sample. An ANM collected a cervical sample for careHPV testing and a Pap smear sample by Ayres spatula, and performed VIA after per speculum examination was completed. Both vaginal and cervical samples were collected with careHPV cervical samplers into DCM collection medium by Qiagen company-provided brush. For careHPV testing, a ratio of viral upload expressed in relative light units (RLU) and positive control set at 1 pg/mL cut-off (CO) was considered positive if the RLU/CO value was ≥1.0. Bethesda system was followed for cytology in reading the Pap smear.27 A Pap smear report with atypical squamous cells of undetermined significance (ASC-US) or more was considered a positive test. VIA was performed by application of 5% acetic acid with a cotton swab on cervix and by allowing sufficient time (1 min) to observe a colour change on the transformation zone. VIA was considered positive if the colour turned white against the pinkish background of normal epithelium, or negative as per standard criteria.28 A positive screen by any of the tests was referred to colposcopy and directed biopsy.
Colposcopic diagnosis was made as per IARC guidelines.27 Biopsy/ECC (endocervical curettage) was performed wherever necessary. Treatment of precancerous lesions was carried out as per IARC guidelines.27 CINII+ (CINII, CINIII and cancer) cases were treated with cryotherapy, where eligible, by a doctor at a CHC, or referred to a tertiary care hospital for surgical and radiotherapy procedures. Quality control of screening test, colposcopy and histology evaluations was ensured by retraining of the ANMs, doctors and external histopathology personnel, respectively.
Sensitivity and specificity and positive predictive value (PPV) of the various screening tests, VHPV, CHPV, Pap and VIA, were computed for detecting histological-confirmed CINII+ or CINIII+ lesions in specified age groups of 30–39, 40–49 and ≥50 years. Along with histological and screening test positives, it was assumed that negative results on all four screening tests were truly negative for precancer or cancer.20 Sensitivity of a clinical test refers to the ability of the test to correctly identify those patients with the disease, and the specificity of a screening test refers to the ability of the test to correctly identify those patients without the disease. PPV is the percentage of patients with a positive test who actually have the disease. A sample size of 5000 women for the present study screening site was determined by assuming 2% prevalence of CINII+ in the general population, with 0.4% allowable error, 6% attrition and 95% CI. The χ2 tests were performed to assess trend in age groups in relation to the detection of CINII+ or CINIII+ for various screening tests. The study was scientifically and ethically approved by collaborating institutions. Analysis of data was performed by using SPSS V.21.0.
Results
A total of 5032 (65%) women aged 30–59 years responded for cervical screening out of 7761 women invited from the community. Women who did not complete all the four screening tests or those missing histology reports were excluded. Data on 4658 women, with results of all screening tests available after exclusion of missing tests, were subjected to further analysis. The baseline characteristics collected were summarised as follows. Of a total of 4658 women screened, 4537 (97%) women were married. Menstrual history was regular in 3852 (83%) women and irregular in 805 (17%) women. None of the women had any health risks that could influence the results. Age-specific distribution of various parameters of women screened is shown in table 1.
Age group-wise distribution of various parameters of women screened
Of the total 4658 women, 61.7% were between 30 and 39 years of age, followed by 27.1% in the 40–49 year category and 11.2% ≥50 years age. Any screen test positivity (18.9%) was highest in the 30–39 years age group. Significant (p<0.05) declining trends in screen-positive rates with increasing age were observed. Table 2 depicts the diagnostic performance of various screening tests, namely, careHPV testing, on self-collected and provider-collected samples, and Pap and VIA tests, in different age categories, for detecting CINII+ and CINIII+. For all screening tests a non-significant (p>0.05) association of age with CINII+ or CINIII+ was observed. In the age group of 30–39 years, the performance of VHPV was comparable to that of CHPV and Pap in the detection of CINII+ and CINIII+. For the age group 40–49 years, performance of VHPV was inferior to that of CHPV and Pap in the detection of both CINII+ and CINIII+. In this age group, both CHPV and Pap testing showed 100% sensitivity. On the contrary, VIA failed to pick up any CINII+ lesion. For the ≥50 years age group, performance of detecting of CINII+ showed by CHPV was remarkable, with a sensitivity of 80% and specificity of 97.5% as compared to VHPV, with 40% sensitivity and 97.7% specificity. For detection of CINIII+, CHPV had high sensitivity (100%) as compared to VHPV (50%). Pap was not able to detect any CINII+ cases in ≥50 years age group. VIA failed to detect any case of CINIII+ in the 40–49 and 50 years and above age groups. Specificity of CHPV was better (97.9%) as compared to Pap (89.5%). Ninety-five per cent CIs for various estimates were extremely wide in older age groups (table 2).
Age-specific performance of various screening tests
The total screen-positive tests in all ages for which colposcopy referral was required varied from 111 (2.4%) to 257 (5.5%) for different modes of screening. Biopsy was performed on 62–83% of cases who screened positive by various modes. Positivity indicated on colposcopy for 47.6%, 83.3% and 56.3% of histological CINII+ cases in the age groups 30–39, 40–49 and ≥50 years, respectively. In general, those screen-positive women recalled for further management were seen more in the younger age group of 30–39 years, for all the tests. Maximum referral was for VIA 6.8% followed by CHPV (3.4%), Pap (3.0%) and VHPV (2.8%). For the 40–49 years age group, VIA referrals were also maximum (4.3%) followed by 3.4% for Pap, 2.1% for CHPV and 1.5% for VHPV. In the age group ≥50 years, referrals by Pap (0.8%) were minimum, followed by VIA (1.3%), CHPV and VHPV (2.3%).
Discussion
The study showed no significant association between age and high-grade CIN for careHPV testing on provider-collected samples or self-collected samples, or for Pap or VIA. In the younger age group, CHPV, VHPV and Pap showed almost equal performance for detection of CINII+. In the 40–49 years age group, performance of VHPV was lower than Pap in the detection of CINII+ and CINIII+. Both CHPV and VHPV performed equally or higher than Pap, with respect to specificity in all age groups. In the older ≥50 years age group, CHPV showed highest sensitivity. Performance of VHPV was half of CHPV's performance. Pap showed 20% or zero sensitivity, respectively, in detection of CINII+ and CINIII+, for the older age group. The careHPV test recommended for low resource countries (LRCs)19 is an affordable test, and can run with minimum lab requirements. The present study showed better performance of CHPV as compared to other screening tests in all the age groups. The overall sensitivity for all ages combined was higher for CHPV followed by Pap, VHPV and VIA: 53.1%, 43.8%, 40.6% and 21.9%, respectively.
In the present study, poor overall sensitivity of the different tests, including HPV, was observed, as compared to other studies.16 ,20 ,29 ,30 The combined data of all centres in the multicentre careHPV study20 reported higher sensitivities for CHPV (82%) and VHPV (70%). A variation of 46–81% across study sites on HPV sensitivities for the detection of CINII+ were reported in another multicentre study from India.31 The present study is in agreement with the observations of a meta-analysis13 that showed lower sensitivity for VHPV as compared to CHPV.
Our study considered only three broad categories of age groups for studying association with the performance of screening tests, instead of narrower age ranges, due to the limitation of fewer high-grade lesions detected for screening from an adequately large population size. The main limitation of the study was low screen positivity and fewer high-grade CIN lesions. The other limitation was fewer samples in the older age groups for performance evaluation and the fact that the estimated rates had less precision. The major strength of the study was its conduct in a truly representative rural community, rather than a hospital-based or camp-based setting. In a rural-based setting such as ours, women of all age groups were personally invited for screening after providing them with health education. The target population responding for screening in the present study was sufficiently high, 65%.
A split sample analysis of various North American and European countries found consistently higher HPV sensitivity and was unaffected by the ages of the women.16 A pooled study from China18 reported that sensitivity did not vary by age or study site. On the contrary, an HPV study from women in a high-risk province of Costa Rica15 reported higher sensitivity in the ≥41 years age group (93.2%) as compared to the 31–40 years age group (80.8%). The present study demonstrated that careHPV had performance identical to CHPV and to VHPV in women of younger ages (below 40 years). It was also found that the performance was equal. This might be due to higher proportion of cervicovaginal infections in sexually active women. In the present study, no high-grade CIN could be detected by Pap in the age group of ≥50 years. An overview of various studies on HPV reported1 higher PPV for detecting CINII+ (17.8%) in the age group >35 years as against 12.8% in younger women. In the present study, the PPVs in the detection of CINII+ for CHPV, VHPV and Pap for the age group of 30–39 years were 12.5%, 10.3% and 11.8%, respectively.
The poor performance of the Pap test in the age group ≥50 years might have been due to collection of HPV samples prior to the Pap smear, which could have resulted in further depletion of exfoliated cells in postmenopausal women with already diminished cellularity. The other possibility might be the low response rate of older women in the screening programme in a previously unscreened population. A study from China12 on pooled analysis of various studies showed significantly lower sensitivity of cytology by LBC for CINII+ and CINIII+ lesions as compared to HPV samples (both for self-collected and physician collected). In contrast, Pap demonstrated16 substantially better sensitivity in the ≥50 years age group for CINII+ detection. For the younger age groups of 30–39 and 40–49 years, our study showed that Pap sensitivity was comparable to CHPV and superior to VHPV. VIA was adequately specific but demonstrated poor sensitivity in all ages in the detection of both CINII+ and CINIII+. The reason for low VIA sensitivity could be due to its high false negativity in a parallel HPV setting.
The study also observed a maximum number of women responding for screening and a higher per cent of women detected as screen positive in the younger age group compared to the other age groups. This higher per cent of younger reproductive aged and sexually active women responding for cervical cancer screening could be due to harbouring of associated low-grade genital infections. In this context, an earlier survey from the same community32 stressed the need for special and focused efforts to direct health education towards illiterate and older women to improve their response to screening.
In conclusion, the performance of CHPV was found to be better in the detection of CINII+ and CINIII+, irrespective of the age of women. For those women below 40 years of age, CHPV, VHPV and Pap tests performed almost identically in the detection of high-grade CIN. The performance of Pap was comparable with CHPV among women aged 30–39 years and those aged 40–49 years. For postmenopausal women ≥50 years, CHPV showed better performance. As VHPV avoids the cost of setting up a primary screening clinic, it may be a feasible option for LRCs, if targeted at those up to the age of 40 years. Pap testing is still an option for the target group up to 50 years of age if LRCs can gear up and make it possible to implement. However, for the control of cancer of the cervix, the better screening option for all age groups is undoubtedly CHPV.
What is already known on this subject
The performance of careHPV, a newer primary cervical screening tool, is known to be superior to Papanicolaou (Pap) test and visual inspection of the cervix with acetic acid (VIA). However, not much is known about the age-specific performance of careHPV in the community setting.
What this study adds
The study shows age-specific performance of careHPV in self-collected vaginal and physician-collected cervical samples and its comparison with Papanicolaou (Pap) test and visual inspection of the cervix with acetic acid (VIA). Both careHPV (cervical and vaginal) and Pap test are suitable for women up to 50 years of age, but for women >50 years of age, physician-collected cervical careHPV is the only option.
Acknowledgments
The authors thank Dr J Jernimo, PATH coordinator for multicountry monitoring in data collection and Dr S Bhambhani, Dr S Sodhani and Dr S Gupta, Department of Cytopathology, ICPO, for histological diagnoses in the study.
References
Footnotes
Contributors SA and SL contributed to study design, planning, execution and interpretation. SA carried out clinical management and field monitoring. SL contributed to data collection, monitoring and analysis of the data.
Funding This work was supported by PATH, Seattle, USA.
Competing interests None declared.
Patient consent Obtained.
Ethics approval The study was approved by the Health Ministry Steering Committee (HMSC), Ministry of Health, Government of India, as well as by the Institutional Scientific Advisory Committee (ISAC) and Institutional Ethical Committee (IEC) of ICPO and PATH and other collaborating hospitals.
Provenance and peer review Not commissioned; externally peer reviewed.