Detection of cervical precancer and cancer in a hospital population: benefits of testing for human papillomavirus

doi:10.1016/j.ejca.2004.01.021

European Journal of Cancer

Volume 40, Issue 8, May 2004, Pages 1225-1232

https://doi.org/10.1016/j.ejca.2004.01.021 Get rights and content

Abstract

The aim was to determine the relevance of human papillomavirus (HPV) testing in identifying high-grade cervical intraepithelial neoplasia or worse (CIN2/3+) in a hospital population (n=3574) characterised by a high rate of cytological abnormalities and high-risk HPV infections. According to the results of the initial Papanicolaou and HPV test, women were directly referred for colposcopy/biopsy or recalled for a control visit. Sensitivity and specificity were corrected for verification bias. HPV-testing sensitivity was 94.3%, higher than that of cytological testing at any cut-off point (65.1%–86.8%), while specificity was greater for cytology than for HPV testing (99.3% or 91.8% versus 83.4%). The combination of both tests allowed 100% sensitivity and negative predictive value. We conclude that HPV testing is a relevant tool for the detection of cervical disease. The best way of combining cytology and HPV detection in screening programmes should be evaluated in large-scale studies.

Introduction

To date, the detection of precancerous cervical lesions by Papanicolaou (Pap) smear is widely recognised as the most effective method for preventing carcinoma of the cervix. The implementation of cytological screening has led to a major reduction in deaths from cervical cancer in certain countries. In France, the incidence of invasive carcinoma decreased by 3.5% per year during the period 1982–1992, while the numbers of carcinoma in situ rose sharply [1].

However, cervical screening based on conventional cytology is far from perfect. In addition to its low sensitivity, conventional cytology has poor reproducibility, even with the introduction of the Bethesda system 2, 3. In the early 1990s, liquid-based cytology was introduced, and large-scale studies demonstrated a slight improvement in the detection of high-grade lesions and carcinomas 4, 5. Moreover, it has advantage of preserving cells, which can be used later for human papillomavirus (HPV) DNA testing.

Indeed, it is now well established that high-risk (HR-) HPV are the main causal agent of intraepithelial and invasive cancer of the cervix 6, 7. However, only persistent infections may be of predictive value for present as well as future cervical intraepithelial neoplasia (CIN) 8, 9. The commercially available Hybrid Capture II (HCII) method for HPV detection, reproducible [10] and sensitive [11], provides objective results and could complement cytology in improving cervical screening. Numerous studies have evaluated HPV testing in primary screening 12, 13, 14, 15, 16, 17, 18, for the triage of atypical squamous cells of undetermined significance (ASCUS) and low-grade squamous intraepithelial lesions (LGSIL) 19, 20, and as a predictive marker of residual or recurrent disease following surgical treatment [21]. Although many researchers and clinicians are deeply convinced of the utility of HPV testing 22, 23, others are not 24, 25. These matters are also confusing because they involve important financial interests and deeply fixed habits. Currently, the point would be to find the best cost-efficient and clinically effective procedure for improving the prevention of cervical cancer.

In this study, we report the 5-year use of HPV testing in routine clinical practice in a French hospital, located in a county that is also a pilot area for organised cervical screening. Indeed, since 1993, Doubs County has been one of the pilot counties in France for such screening, targeting women between 20 and 65 years of age living in this administrative area 26, 27. Here we compare the results of different combinations of screening tests (standard cytology and HPV testing) in identifying precancerous cervical lesions.

Section snippets

Study participants

Between August 1997 and July 2002, 9436 cervical specimens from 6691 women were tested for HPV DNA by the HCII system in the Cellular and Molecular Biology Laboratory of Besançon University Hospital. All the women who participated were carefully informed about HPV testing, the benefits of participation and the possibility of being recalled for additional visits. The present analysis included only women attending the Gynaecology Department of the University Hospital of Besançon, because of the

Results

The mean age of the 3574 women eligible at entry into the study was 37.2±11.6 years (range 15.8–87.5 years).

Pap smears were as follows: 3019 normal (84.47%), 62 ASCUS (1.73%), 249 LGSIL (6.97%), 209 HGSIL (5.85%), 9 suggestive of invasive carcinoma (0.25%) and 26 unsatisfactory (0.73%). LGSIL smears were mostly observed at a young age (15.8% in the 20- to 24-year group), while HGSIL+ peaked between 35 and 39 years (8.8%).

3165 specimens out of 3574 (88.6%) were tested for both HR- and LR-HPV.

Discussion

Here we report the 5-year routine use of HPV testing in addition to conventional cytology in women attending the University Hospital of the Doubs County, and evaluate the performance of both tests in detecting cervical precancer or cancer. Obstetrics/gynaecology specialists of this hospital are referent practitioners in the area for screening, management and treatment of cervical pathology. They are thus in charge of a large majority of the women participating in the Doubs’ pilot screening

Acknowledgments

We are indebted to Sylviane Coumes-Marquet for excellent technical assistance with HPV testing. We thank all the women who participated in the study. The grant sponsors were Ligue contre le Cancer (comité du Doubs). Mrs V. Dalstein is the recipient of a predoctoral scholarship from the Fondation de France.

References (31)

S.J. Bernstein et al.
Liquid-based cervical cytologic smear study and conventional Papanicolaou smearsa metaanalysis of prospective studies comparing cytologic diagnosis and sample adequacy
Am. J. Obstet. Gynecol.
(2001)
K.U. Petry et al.
Factors associated with an increased risk of prevalent and incident grade III cervical intraepithelial neoplasia and invasive cervical cancer among women with Papanicolaou tests classified as grades I or II cervical intraepithelial neoplasia
Am. J. Obstet. Gynecol.
(2002)
J. Cuzick
Time to consider HPV testing in cervical screening
Ann. Oncol.
(2001)
A.L. Herbst et al.
The management of ASCUS cervical cytologic abnormalities and HPV testinga cautionary note
Obstet. Gynecol.
(2001)
P. Schaffer et al.
Cervical cancer screening in France
Eur. J. Cancer
(2000)
M.A. Nobbenhuis et al.
Cytological regression and clearance of high-risk human papillomavirus in women with an abnormal cervical smear
Lancet
(2001)
J. Cuzick et al.
Management of women who test positive for high-risk types of human papillomavirusthe HART study
Lancet
(2003)
C. Weidmann et al.
L'incidence du cancer du col de l'utérus régresse régulièrement en France
Bull. Epidémiol. Hebdomadaire
(1998)
P. Sasieni
Cervical sampling devices
Br. Med. J.
(1996)
M.H. Stoler et al.
Interobserver reproducibility of cervical cytologic and histologic interpretationsrealistic estimates from the ASCUS-LSIL Triage Study
JAMA
(2001)

J.L. Papillo et al.

Evaluation of the ThinPrep Pap test in clinical practice. A seven-month, 16,314-case experience in northern Vermont

Acta Cytol.

(1998)

F.X. Bosch et al.

Prevalence of human papillomavirus in cervical cancera worldwide perspective. International biological study on cervical cancer (IBSCC) Study Group

J. Natl. Cancer Inst.

(1995)

J.M. Walboomers et al.

Human papillomavirus is a necessary cause of invasive cervical cancer worldwide

J. Pathol.

(1999)

K.L. Wallin et al.

Type-specific persistence of human papillomavirus DNA before the development of invasive cervical cancer

N. Engl. J. Med.

(1999)

V. Dalstein et al.

Persistence and load of high-risk HPV are predictors for development of high-grade cervical lesionsA longitudinal French cohort study

Int. J. Cancer

(2003)

Cited by (35)

High proportion of abnormal pap smear tests and cervical dysplasia in women with cystic fibrosis
2018, European Journal of Obstetrics and Gynecology and Reproductive Biology
Citation Excerpt :
Although no epidemiological data neither specific natural history of HPV infection in case of inadequate cervical mucus have been described, an unusual HPV interaction and the role of a local inflammation might be suspected. Another finding of this study was the difficulty of cervical screening in women with CF. Non-interpretable PS were not rare (4.3%, compared to a 0.5% to 0.75% proportion reported in the general population) [9,10,13]. Local inflammation and easily induced cervical bleeding might explain these high figures.
Insufficient gynecological follow-up and cervical screening has been reported in women with cystic fibrosis (CF). Some of these patients will require a pulmonary transplantation, known to be associated with a higher risk of cervical dysplasia. The aim of this study was to explore the results of cervical screening in adult women with CF, and to report the prevalence of abnormal pap smear tests in this population.
We retrospectively analyzed medical records of sexually active women with CF who attended a gynecological consultation in Lyon University CF referral center between June 2014 and December 2015. The primary outcome was the result of the pap smear test.
Forty-seven women (32 non-transplanted and 15 transplanted) were included in the study. The median age of the patients was 28 (range 18–53). The clinical examination revealed that 20 (42.5%) women presented an abnormality (inflammatory cervix, cervical or vulvovaginal condyloma). An abnormal pap smear was found in 8/32 (25%) non transplanted women and in 5/15 (33.3%) transplanted women, with no significant difference between the two groups (p = 0.75): seven atypical squamous cells of undetermined significance (ASC-US), five low grade squamous intraepithelial lesion (LSIL), one atypical glandular cells (AGC). Six (12.8%) (four non transplanted, and two transplanted) women had an histologically proven dysplasia (four Cervical Intraepithelial Neoplasia (CIN)1, one CIN2, and one endocervical adenocarcinoma in situ). Overall, ten (21.3%) women had a Human Papilloma Virus (HPV) related disease (cervical and/or vulvovaginal).
A high proportion of transplanted and non-transplanted women with CF had abnormal pap smear tests and cervical dysplasia. A regular gynaecological follow-up, periodic cervical screening, and routine HPV vaccination are strongly recommended in this population.
Molecular diagnosis of human papillomaviruses (HPV): What test(s) in clinical practice?
2016, Journal de Gynecologie Obstetrique et Biologie de la Reproduction
La prescription d’un test HPV en pratique doit permettre au clinicien d’optimiser le suivi et la prise en charge de ses patientes, en particulier dans le cadre du dépistage du cancer du col de l’utérus. Un nombre conséquent de tests sont disponibles qui présentent des sensibilité et spécificité analytiques et cliniques propres. Des recommandations internationales concernant les performances cliniques des tests HPV utilisables pour le dépistage ont été publiées par un groupe d’experts et il convient d’utiliser les tests répondant à ces critères de performance. Au-delà de la trousse de détection des HPV, c’est l’ensemble du circuit, du prélèvement au rendu de résultat, qui doit être maîtrisé. Cela implique que les étapes pré-analytique (prise d’échantillon, qualité du milieu de recueil de l’échantillon, condition de stockage et d’acheminement des prélèvements…) et post-analytique (qualité du compte-rendu, prestation de conseil…) soient elles aussi normalisées. À cette fin, les laboratoires sont soumis à une accréditation par le COFRAC selon la norme internationale ISO 15189 qui impose un cadre, opposable, définissant les critères de qualité requis pour la réalisation de tout test de biologie médicale, et du test HPV en particulier.
Prescription of an HPV test in practice will enable the clinician to optimize the monitoring and the management of patients, especially in the context of cervical cancer screening. Numerous HPV tests are available that present different analytical and clinical sensitivity and specificity. International recommendations on clinical performance of HPV tests used for cervical cancer screening have been published by a group of experts, and tests that meet these performance criteria should be used. Apart from the HPV detection kit, the whole circuit from sampling to report of the results must be considered. This implies that the pre-analytical (sampling, quality of sample collection medium, storage condition and sample transportation…) and post-analytical steps (quality of result reporting, providing expert advices…) are also standardized. For this purpose, medical-biology laboratories are subjected to a COFRAC certification, as defined by the international standard ISO 15189 providing quality criteria for any clinical laboratory test and HPV test in particular.
High prevalence of abnormal cervical smears in a hospital cohort of French women beyond the upper age limit screening program
2015, Preventive Medicine
Citation Excerpt :
As a whole we considered that women were not adequately screened. Women underwent a pelvic examination and two cervical samples were obtained and used for conventional cytology and HPV testing as we previously described (Dalstein et al., 2004). When necessary, a short course (3–4 weeks) of intravaginal estrogen cream was given.
To determine the prevalence of cytological abnormalities and high risk Human PapillomaVirus (hrHPV) in cervical smears from French women aged over 65 years who attended the referent Gynecology Clinic of the Besançon University Hospital.
Between 2002 and 2012, 796 French women aged 66–99 years were cotested for cytology and hrHPV by Hybrid Capture 2 (hc2). hc2-positive cases were subjected to real time PCR for specific HPV 16/18/45 genotyping. Women with normal Pap smears and positive for hrHPV were followed-up every 12 months.
Cytological abnormalities were detected in more than 30% of women and cervical cancers (CC) in 2.9% of women. Benign lesions were more frequent in women aged 66–75 years while (pre)-malignant lesions were preferentially found in women over 76. The prevalence of hrHPV was 22.7%. HPV 16 was the most frequent (23.8%), followed by HPV 45 (7.7%) and HPV 18 (3.9%). The rate of hrHPV increased with the lesion severity and HPV 16 was identified in 50% of CC. Among the followed-up women, those who developed CIN3 were HPV16 positive at study entry.
The study provides important estimates of the prevalence of cervical abnormalities and hrHPV positivity in a French hospital based-population over 65. Findings suggest to consider this high risk population in regards to cervical cancer.
A new method to address verification bias in studies of clinical screening tests: Cervical cancer screening assays as an example
2014, Journal of Clinical Epidemiology
Citation Excerpt :
The limited use of these methods is reflected in recent studies of cervical cancer screening assays. In our review of 15 recent comparative cervical screening trials which performed colposcopy (gold standard) in a subset of screen negatives [11–25], only one study used the likelihood method by Zhou [7] to compare the sensitivity and specificity of the assays (Table 1). A simple and easily implemented statistical method is therefore needed to estimate and compare the performance of screening assays while accounting for verification bias.
Studies to evaluate clinical screening tests often face the problem that the “gold standard” diagnostic approach is costly and/or invasive. It is therefore common to verify only a subset of negative screening tests using the gold standard method. However, undersampling the screen negatives can lead to substantial overestimation of the sensitivity and underestimation of the specificity of the diagnostic test. Our objective was to develop a simple and accurate statistical method to address this “verification bias.”
We developed a weighted generalized estimating equation approach to estimate, in a single model, the accuracy (eg, sensitivity/specificity) of multiple assays and simultaneously compare results between assays while addressing verification bias. This approach can be implemented using standard statistical software. Simulations were conducted to assess the proposed method. An example is provided using a cervical cancer screening trial that compared the accuracy of human papillomavirus and Pap tests, with histologic data as the gold standard.
The proposed approach performed well in estimating and comparing the accuracy of multiple assays in the presence of verification bias.
The proposed approach is an easy to apply and accurate method for addressing verification bias in studies of multiple screening methods.
Prevalence and risk factors of human papillomavirus infection types 16/18/45 in a cohort of French females aged 15-23years
2014, Journal of Epidemiology and Global Health
Investigation of the prevalence and risk factors of human papillomavirus (HPV) infection is the basis for developing prophylactic strategies against cervical cancer, especially for young women. This study aimed to assess the prevalence and risk factors of HPV infection among a cohort of sexually active young French women eligible for catch-up vaccination. Between 1997 and 2007, 2163 women aged 15–23 years attending consultations at the department of gynecology in the Hospital of Besançon (France) were screened for high risk HPV (HR HPV) infection. Risk factors were investigated through a questionnaire sent to all participants in 2010. HPV DNA was detected by HC2 and Probe Set assays. The overall prevalence for HR HPV and HPV16, 18 and/or HPV45 was 44.6% (95% CI, 42.5–46.7%) and 19% (95% CI, 17.3–20.7%), respectively. The response rate to the questionnaire was 22.6%. The prevalence of independent risk factors (age older than 19, smoking, and oral contraception) for HPV 16/18/45 infection in this population was less than 20%. Based on this study, HPV vaccination should be offered not only to teenage girls, but also to young women, regardless of their sexual activity.
HrHPV-testing in a university hospital gynecology outpatient clinic: Recommendations for clinical practice
2012, Gynecologic Oncology
To study the effect of hrHPV-testing on the detection of CIN2/3+ in women referred to a gynecology outpatient clinic, and to assess a useful risk profile in relation to the referral reason to identify who should be tested for cervical pathology.
This study was designed as an observational cohort study. In the first six months of 2007, we categorized the referral reason of 1149 consecutive women who visited our gynecology outpatient clinic and assessed the risk for CIN2/3+ as found by cytology or co-testing with a hrHPV-test and cytology.
Three different categories of referral reasons were identified; women with presumed cervix pathology, women with presumed endometrial pathology and women with other referral indications. The cumulative 18-month CIN2+ and CIN3+ risks were highest in the group with presumed cervical disease (adjusted risks 11.1% and 5.4% respectively) and lowest in the miscellaneous group with no suspicion of cervical and/or endometrial pathology (adjusted risks 4.1% and 1.8% respectively). HrHPV-testing detected significantly more CIN2/3+ lesions than cytology (relative detection rate: 1.42 (95%CI 1.05–1.92) and 1.38 (95%CI 0.95–2.05) respectively).
The high (> 2%) cumulative 18-month CIN2/3+ risk in patients with presumed cervical and/or endometrial pathology warrants routine cervical testing. In these women a hrHPV-test should be added to cytology because this identifies a significant number of additional women with a substantial risk of CIN2/3+ lesions who would not be identified with cytology alone. Women referred for other reasons should not have cervical testing beforehand, because of their low risk of CIN2/3+.

View all citing articles on Scopus

View full text

Detection of cervical precancer and cancer in a hospital population: benefits of testing for human papillomavirus

Abstract

Introduction

Section snippets

Study participants

Results

Discussion

Acknowledgments

Am. J. Obstet. Gynecol.

Am. J. Obstet. Gynecol.

Ann. Oncol.

Obstet. Gynecol.

Eur. J. Cancer

Lancet

Lancet

L'incidence du cancer du col de l'utérus régresse régulièrement en France

Bull. Epidémiol. Hebdomadaire

Cervical sampling devices

Br. Med. J.

Interobserver reproducibility of cervical cytologic and histologic interpretationsrealistic estimates from the ASCUS-LSIL Triage Study

JAMA