Introduction

Cervical cancer incidence and mortality is relatively higher among older women compared with younger women. In addition, women aged ≥65 years are more likely to be diagnosed at an advanced stage, and thus have a poorer prognosis.1 2 This is likely attributed to insufficient screening, diagnostic challenges, inadequate biopsy protocols and insufficient follow-up of older women as well as potential faster disease progression in older women.3–9

Diagnostic work and follow-up include colposcopy-guided biopsies. During colposcopy, it is essential to visualise the transformation zone (TZ), the area between the old and new squamocolumnar junction (SCJ), where precancerous lesions (cervical intraepithelial neoplasia (CIN)) typically arise and biopsies should be sampled. Accurate and timely diagnosis of CIN using colposcopy is, therefore, a crucial part in prevention of cervical cancer in order to detect CIN and prevent the development of invasive disease.

In older women, colposcopy and biopsy are challenging due to postmenopausal-related and age-related epithelial atrophy and retraction of the TZ into the cervical canal, rendering the colposcopic examination inadequate. According to the 2011 International Federation of Cervical Pathology and Colposcopy nomenclature, TZ should be classified in concordance with this retraction as type 1 (fully visible SCJ), type 2 (partially visible SCJ) or type 3 (not visible SCJ).10 TZ type 3 poses a problem during colposcopies: studies revealed inadequate colposcopy in up to 67% of postmenopausal women (mean age: 62 years), mostly due to TZ type 3.8 9 11 This challenge may result in prolonged follow-up involving repeated colposcopies to obtain representative biopsies, causing psychological distress and anxiety among patients,12–14 as well as an increased risk of missing disease in the cervical canal.8 11

This risk is considerable and is supported by a Danish study among screen-positive older women (median age 67.7 years) with TZ type 3 showing that more than half (54.5%) of the CIN grade 2 or higher (CIN2+) (CIN2, CIN3/Adenocarcinoma in Situ (AIS), cancer) cases detected in the cone biopsy were missed by colposcopy-guided biopsies.15 TZ type 3 has also been associated with a fourfold to fivefold increase in risk of cervical cancer.16

Consequently, guidelines in Scandinavia and Australia recommend a cone biopsy to achieve a correct diagnosis in women with TZ type 3 at colposcopy, although this approach entails risk of surgery-related complications, including subsequent cervical stenosis, which may compromise the follow-up.17–19 Moreover, conducting a cone biopsy also entails a recognised risk of overtreatment (61%).15 Therefore, there is an urgent need to overcome this diagnostic challenge and optimise the visualisation of the TZ. This involves converting TZ type 3 into type 1 or 2, thereby improving the diagnostic in older women.

An alternative approach to improve visualisation of the TZ could be medical treatment with oestrogen (oral or vaginal) prior to colposcopy.11 20–26 Unfortunately, the quality of evidence in these studies is low due to the non-comparable study designs, limited sample sizes and missing placebo groups (table 1). Furthermore, comparison of the effect size of the oestrogen treatment across studies is a challenge due to differences in inclusion criteria and age group, but especially as the oestrogen medication (cream, tablets), administration form (oral, vaginal), length of treatment (from 5 to 84 days) and dose vary considerably (table 1). Despite the lack of conclusive evidence, guidelines suggest the use of local oestrogen pretreatment prior to colposcopy in TZ type 3 cases.17 18 These guidelines may be due to encouraging results from studies using vaginal oestrogen treatment where the proportion of women going from TZ type 3 to type 2/1 ranges between 64% and 83%.20 21 25 Nevertheless, evidence derived from randomised controlled trials are needed to gain more knowledge on how to most effectively use vaginal oestrogen as a treatment prior to colposcopy.

Table 1

Overview of studies with use of oral or vaginal oestrogen to enhance visualisation of the TZ (n=6)

Aim

Among screen-positive postmenopausal women aged ≥50 years, this randomised controlled, double-blind, multicentre trial will investigate if treatment with vaginal tablets containing 30 μg oestrogen prior to colposcopy can improve the visualisation of TZ and the diagnosis of CIN2+, as compared with a placebo group that receives placebo tablets with no oestrogen content.

Hypotheses

The pretreatment with vaginal tablets containing 30 μg oestrogen will increase the proportion of women with a visible TZ (type 1 and 2) during colposcopy with minimum 50%. Additionally, this approach will increase detection of CIN2+ cases and increase representation of TZ in at least 25% of the cervical biopsies.

Methods and analysisSetting

According to Danish screening guidelines, women aged 23–64 years are invited for cervical cancer screening, at their general practitioner (GP). Every 5 years, women aged ≥50 years are invited for cytology or high-risk (hr) human papillomavirus (HPV)-based screening depending on even or uneven birthday.27 Women aged 60–64 years will be screened with an hrHPV test, and exit the programme in case of an hrHPV-negative test, without consideration of their screening history.18 In addition, since 2019 women aged 65–69 years in the Central Denmark Region have received an invitation for a catch-up, an extra hrHPV test, as part of a large Danish population-based intervention study.28 29 Depending on the severity of the abnormal screening results, women are referred for either retesting or further examination by colposcopy.18 In Denmark, screening, including eventual treatment and/or follow-up, is free of charge.

When referred for colposcopy, a central visitation centre will randomly mail out timeslots for colposcopies at gynaecology departments based on the women’s residency. The study population will be included from four gynaecology departments in the Central and Southern Regions in Denmark (Randers, Horsens, Gødstrup and Svendborg). They are centralised with three pathology departments.

Study design and eligibility criteria

The Improving Diagnostic in cErvical dysplasia: A randomised study with Local vaginal oestrogen prior to colposcopy is designed as a randomised controlled, double-blind, multicentre, superiority study (figure 1)30 31 and conforms to the Standard Protocol Items: Recommendations for Interventional Trials Statement.30

Women referred for colposcopy with the following criteria will be included: postmenopausal (defined as no menstrual bleeding ≥1 year); aged ≥50 years; a positive HPV test and/or an abnormal cervical cytology; previous abnormal cervical histology with minimum 6 months since the last colposcopy with biopsies. Exclusion criteria will be use of oestrogen within the last 3 months, regardless of administration form; previous cervical radiotherapy, cervical amputation and/or cone biopsy; pregnant women. Eligible women will be booked for colposcopy as usual and will be prospectively enrolled. A total of 150 women will be randomised 1:1 prior to the colposcopic examination to either:

The intervention group, which will receive pretreatment with vaginal application of oestrogen tablets (Vagifem) 30 μg once a day for 14 days.

The control group, which will receive placebo tablets with vaginal application without oestrogen once a day for 14 days.

Figure 1

Study design, eligibility criteria and outcomes. CIN2+, cervical intraepithelial neoplasia grade 2 or higher; hrHPV, high-risk human papillomavirus; SCJ, squamocolumnar junction; TZ, transformation zone.

Eligible women will receive the invitation for colposcopy through the Danish secure digital mail system, together with a study invitation letter that includes information to contact the principal investigator (PI) by phone or email to obtain further information regarding participation in the study. Due to the General Data Protection Regulation, the PI will not be allowed to make the first contact. The information letter will also include a declaration of consent to be signed and returned on participation.

Randomisation

After assignment to interventions, the PI, care providers and pathologists will be blinded. Participating women will be randomised to either the intervention or control group according to a randomisation list. Randomisation will be conducted by a data manager using a locked pseudorandom number function in the Research Electronic Data Capture (REDCap) system.32 33 The randomisation will be stratified by age and divided into two groups in order to ensure a uniform distribution: group 1 (age 50–62 years) and group 2 (age 63 years and above).

The PI will be able to enter patient data in the REDCap system and provide a pseudorandom number for each patient in order to inform the Pharmacy of Central Denmark Region (PCDR). PCDR is responsible for sending the study medication to the women by mail based on a locked randomisation list received from the data manager.

In addition to the randomisation list, the data manager will also create another locked file in the REDCap system which can be used by the PI as an unblinding option for a given patient, without affecting or unblinding the other women included.

Intervention

After informed consent has been obtained, the following information will be collected: age, height, weight, year of latest period, HPV vaccination status, smoking habits, parity, types of delivery, other diseases and use of medication. At the same time, the referral cytology and the dysplasia history will be recorded to ensure correct inclusion. Only one PI will be responsible for inclusion and collection of the relevant patient data. Study enrolment started in August 2023 and will continue until 150 patients have been enrolled, expected around December 2024.

PCDR will manage and mail the study medication to the women by Post Denmark who provides basic postal services to all households and businesses in Denmark every weekday.34 The PI will call the women no more than 5 weekdays after the treatment has been mailed to ensure they have received it. The mail will include a written and picture-based user instruction on how to apply the medication as well as information about the start date and end date for application. The start date will be 14 days prior to the colposcopy at the gynaecological department. The PI’s contact information will appear in the mail as well in case patients need to approach her for any reason.

Compliance will be monitored by asking the patients to return any remaining study medication at the day of colposcopy.

Colposcopy

The routine colposcopy examination will be performed as recommended in the national guideline by trained nurses or physicians who routinely perform colposcopies (Danish Society of Obstetrics and Gynaecology): women referred for colposcopy will have a minimum of four biopsies taken regardless of the referral result and colposcopy assessment,5 for example, whether visible signs of CIN are present or not. Biopsies will be sent for histopathological evaluation at the relevant Department of Pathology in the Central and Southern Region, and will be examined following usual principles and procedures.35 Normally, all biopsies would be collected in one tube. However, in the study, the first biopsy will be collected in one tube (marked 1) and the remaining three biopsies in another tube (marked 2) to evaluate which biopsy (target vs random) represent the most severe histopathological results.

Postcolposcopic evaluations

After the examination, the colposcopist will fill out a questionnaire regarding:

whether the examination was performed by a nurse, a resident doctor or a consultant;

assessing the colposcopic examination by describing the TZ as type 1, 2 or 3;

assessing whether the impression of the examination was ‘normal’, ‘not normal’ or ‘do not know’.

marking whether the patient has handed in any remaining study medication or not.

Patients after the colposcopic examination will receive a link to a validated questionnaire regarding compliance and possible side effects of the pretreatment, including whether they found the approach user-friendly. Optionally, the women can complete the questionnaire by phone with the PI.

The end of trial will be 14 days after the colposcopic examination and within this period it will be possible for the women to contact the PI, for example, to register any side effects/events. There will be no further follow-up of the women after this period. In the event that the women do not show up for their colposcopy examination or no longer wish to participate in the study, they will be excluded from the study.

The histopathological results will be obtained from the patient record and/or the nationwide Danish Pathology Register.36 This will include histological grading using the CIN classification and an evaluation of the cervical biopsies regarding representation of the TZ.37 The most severe histopathological result will be used if more than one result is reported.

Patient and public involvement

Neither patients nor the public were involved in the design of this study. We plan to disseminate the results to the women participating in the study, to the women attending colposcopy examinations and to healthcare professionals working within the field.

OutcomesPrimary outcomes

Between groups, the primary outcomes will be a comparison of the percentage of women with TZ type 1, 2 and 3 scored by the colposcopist and the percentage of women with representation of the TZ in at least 25% of the cervical biopsies described by the pathologist.

Secondary outcomes

Between groups, the secondary outcomes will be the proportion of detected CIN2+ in the cervical biopsies, the proportion of diagnostics cone biopsies, the patients’ report on possible side effects and compliance to the pretreatment. Furthermore, we will explore for each patient which biopsy represents the most severe histopathological result by comparing the first biopsy with the other three.

Data sources

In Denmark all residents are assigned a personal civil registration number, called a CPR number, which gives access to the healthcare system and enables the linkage of all relevant health data at an individual level in a patient record system. Women referred for colposcopy will be identified by using the electronic National Patient Registry,38 which also will provide relevant patient data. This information will be combined with data from the nationwide Danish Pathology Register,36 which will provide the histopathological results of the biopsies.39

Statistical analysis

Descriptive statistics with number and proportions will be used to present baseline characteristics in both groups to determine if the randomisation was successful. Between groups, the outcomes and differences will be estimated both as relative risk and absolute difference, presented with 95% CIs. The questionnaires regarding side effects will be described using descriptive statistics. Statistical analyses will be performed using STATA V.16. All data will be stored in the REDCap database.32 33

Sample size

To detect a clinical relevant difference in the percentage of women with TZ type 1 or type 2 between groups of at least 50%, and assuming that 54%11 21 and 81%20 of the women in the placebo and intervention group, respectively, will have TZ type 1 or 2 at colposcopy, at least 62 women should be randomised to each group, accepting a power of 90% and an alpha value of 5%. To allow for a protocol violation of 20%, 75 women will be included in each group.

Strengths and limitationsStrengths

To our knowledge, this randomised controlled trial will be the first to investigate if visualisation of the TZ and detection of CIN2+ can be enhanced by pretreatment with vaginal oestrogen tablets prior to colposcopy among screen-positive postmenopausal women aged ≥50 years. The randomisation will be stratified for age in order to ensure a uniform distribution. A key strength is also that the trial is blinded for the PI, care providers and pathologists, lowering the risk for information bias in the outcome measures. The trial will include several times more women than previous studies exploring the utility of pretreatment with oestrogen before colposcopy. Selection of the study population is performed randomly by a central visitation centre to minimising risk of selection bias. In addition, the women will be enrolled at four different gynaecological departments representing different regions of Denmark, enhancing the generalisability of the study findings. Another key strength will be the monitoring of the compliance to pretreatment by asking the women to return unused medication the day of colposcopy. The use of data from medical records and the Danish Pathology Registry, which has highly valid records, will ensure completeness of the study outcomes and, thus, reduce the risk of information bias and loss to follow-up. The postcolposcopic validdated questionnaire allows information about the use of the pretreatment from a patients’ perspective, which is relevant for incorporating the pretreatment into clinical practice.

Limitations

A selection problem may occur as patients are required to contact the PI (and not the other way around) in order to be included in the study, which could lower the internal validity of the study. Ideally, inclusion should not be based on the women’s resources and ability to contact the PI. It may also be expected that some women will not receive information about the study from the invitation letter (ie, they never saw the letter or misplaced it). Despite the possible selections bias, we do not expect these limitations to have a large impact on the outcomes, as the patient, PI, care providers and pathologists will not know the patients’ type of TZ at the time of inclusion.

One limitation could be the appearance of the placebo tablets, which are produced without the same logo as on the oestrogen tablets (Vagifem tablets). However, this may only be a limitation if patients recognize the appearance of Vagifem tablets due to previous use. We assume that the proportion of women who have previously been treated with Vagifem should be equal among the two groups and, overall, consider it a minor limitation.

Some women may be misclassified according to their type of TZ. An external reference reviewer could lower the intra-observer and inter-observer variation. Unfortunately, this is not possible in this study. Moreover, this type of study design would not reflect real-life practice. However, this challenge is expected to be equal in the two groups as the colposcopists will be blinded, hence it should have a minor effect on estimated absolute differences in outcomes. Despite the effort to gain a high compliance among participants, compliance can still be a crucial problem which, in this case, will result in an underestimation of an effect.

Ethics and disseminationEthics

The study protocol has been approved by the Central Denmark Region Committee on Biomedical Research Ethics (1-10-72-34-22), the Central Denmark Regions’ Research Unit (1-16-02-72-22) and The Danish Health Authority (Danish Medicine Agency; 2022015030). The study has also been registered with an EudraCT number (1-23-456; 2022-000269-42) and registered as a clinical trial on www.clinicaltrials.gov, NCT05283421 (see table 2 for the WHO trial registration dataset). The local GCP unit will supervise and monitor the study closely before, during and after the study period. Study information will be given to the women verbally and in writing, and they must understand Danish to accept the given information before participating. All information will be anonymised prior to analysis. Participation in the study will not affect the examination and follow-up of the women. The study is investigator-initiated with no conflict of interest and no collaboration with pharmaceutical companies. Novo Nordisk has not been involved in any aspects of the study or the manuscript. The results will be published in international peer-review scientific journals, and presented at national as well as international meetings and conferences. Furthermore, results will be compiled as a thesis for a PhD, which will be submitted for examination at Aarhus University, Denmark.

Table 2

All items from the WHO trial registration dataset

Study medication

The dose of 30 μg local vaginal oestrogen (Vagifem) is within the well-known standard recommendation, with a minimum of side effect and no contraindications.40 It is widely used and recommended by GPs and gynaecologists to women aged ≥50 years. There may be some inconvenience with the application of the study medication but this is considered to be comparable to having to insert a tampon during menstruation. Any side effects and adverse events will be registered and reported according to GCP guidelines.41 The sponsor and PI will assess the causality of all side effects and whether these are related and/or serious. The placebo tablets will be produced by Glostrup Pharmacy,42 who is one of the two Danish pharmacies offering magistral medical production for studies on approval and monitor from the Danish Health Authority (Danish Medicine Agency). The active tablets Vagifem are produced from Novo Nordisk. PCDR will manage both tablets types in order to repack and mail the pretreatment to the patients.

Perspectives

Demographic models predict an increasing cervical cancer incidence within the next 20 years, especially among older women.43 This is influenced by the increasing female life expectancy, the extensions of the upper screening age and by fewer women getting hysterectomies.28 44–47 This may also be a noteworthy public health concern as cervical cancer carries an economic burden. Therefore, this study is important and highly relevant to ensure an accurate and timely diagnosis for those older women.

To our knowledge, the trial will be the first to offer randomisation with vaginal oestrogen pretreatment prior to colposcopy among screen-positive postmenopausal women. The trial will address key issues not previously explored regarding the diagnostic challenges and complexities to a follow-up regimen among this group of patients. It is expected that the results will contribute with important knowledge that can be promptly incorporated into clinical use, minimising unnecessary and repeated colposcopies, achieving early diagnostic and reducing the risk of missing disease. Furthermore, the results are expected to provide data for clinical follow-up screening guidelines. The evidence obtained from this trial is believed, on the long-term, to help reduce the cervical cancer incidence, the number of postmenopausal women aged ≥50 years with late-stage cervical cancer and poor prognosis. The reporting of the trial will conform to the Consolidated Standards of Reporting Trials guidelines.48 The persistence of HPV infections is one of the most significant predictors for the risk of recurrence of cervical diseases.49 50 A follow-up study on these included women is considered in order to reveal HPV persistence among this group.