Trial of management of borderline and other low-grade abnormal smears (TOMBOLA): Trial design

https://doi.org/10.1016/j.cct.2006.04.001Get rights and content

Abstract

Cervical screening reduces the risk of cervical cancer by detecting and treating cervical intraepithelial neoplasia (CIN). The management of women with low-grade cervical abnormalities is controversial. Two management policies exist: repeat smears in primary care and colposcopy examination. It is not clear which of these is the more effective and efficient. There is also uncertainty as to the most effective and efficient management of women at colposcopy when an area of abnormality is seen on the cervix — immediate treatment or biopsy and selective recall for treatment if the biopsy result suggests this is necessary. The result of a human papillomavirus (HPV) test might assist in deciding the appropriate management of women with low-grade abnormalities. TOMBOLA, a pragmatic randomised-controlled trial set within the cervical screening programmes in Scotland and England, addresses these three areas of uncertainty. Almost four and a half thousand women aged 20–59 with a low-grade cervical abnormality have been recruited and randomised to either repeat smears or colposcopy examination. Women in the colposcopy arm of the trial are further randomised to a policy of either immediate treatment or biopsy and selective recall for treatment if they have an abnormal transformation zone. Women are followed up to an exit examination at 3 years. HPV testing is undertaken at recruitment and at the exit examination. The primary endpoint is cumulative incidence of CIN2/3. A range of other clinical, psychosocial and economic outcomes is being considered. This paper describes the design of the trial, and discusses the rationale underlying aspects of the design and the challenges faced in designing and implementing the trial.

Introduction

By detecting and treating cervical intraepithelial neoplasia (CIN), cervical screening has reduced the incidence of, and mortality due to, cervical cancer. In the UK, screening is based on 3–5 yearly cytological smear tests for women aged 20 or 25 to 60 or 65. Smears are classified as “unsatisfactory”, “normal”, “borderline nuclear abnormalities” (BNA), “mild”, “moderate” or “severe dyskaryosis” or “indicating invasive cancer”. Most clinicians consider smears showing BNA or mild dyskaryosis as “low-grade” and those showing moderate or severe dyskaryosis as “high-grade”. There are well-established guidelines for management of women with high-grade abnormalities but the management of women with low-grade abnormalities is controversial. Each year in the UK, more than 250,000 cytological smears show low-grade abnormalities [1], [2], so this uncertainty regarding management has important implications for both women and the National Health Service.

Two methods of management for women with low-grade abnormalities are routinely employed — repeat smears in primary care at regular intervals (cytological surveillance) or colposcopic examination in a hospital outpatient clinic, with biopsy if an abnormal area is seen on the cervix. It is not clear which of these is the more effective or efficient management policy. Revised guidelines on colposcopy and programme management for the NHS Cervical Screening Programme in England reflect this uncertainty, suggesting that “women should be referred for colposcopy after one test reported as mild dyskaryosis, but it is acceptable to recommend a repeat test” [3]. Referral to colposcopy after a single mild smear will increase the number of referrals to colposcopy, which may increase waiting times. In Scotland, there has been reluctance to deviate from the previous guidelines which applied to the whole of the UK, deferring colposcopy until two smears showing mild dyskaryosis or three smears showing BNA had been reported [4]. In addition, there is considerable heterogeneity in management across the UK. Furthermore, over time and without treatment, some low-grade lesions will progress to higher-grade lesions, others will remain stable, and others will regress spontaneously; follow-up and treatment may therefore be of no benefit to the group of women who have a low-grade lesion that is going to regress. At the moment, however, it is not possible to identify which abnormalities will regress, and which will progress.

Human papillomavirus (HPV) has been shown to have a role in the aetiology of cervical cancer and pre-cancer, with certain types leading to a particularly high risk [5], [6]. This has led to suggestions that HPV testing would be of potential value in the triage of women with low-grade abnormalities [7], [8]. If HPV testing could be used to discriminate between women with a low-grade smear who are likely to have underlying high-grade CIN and those who are unlikely to have high-grade disease, then it might be possible to manage the two groups of women in different ways. For example, those likely to have underlying disease might be referred straight to colposcopy, while those unlikely to have underlying disease could be kept under cytological surveillance, thus reducing the burden on colposcopy clinics in terms of new referrals.

There is also uncertainty as to the most effective and efficient method of managing women with a low-grade smear when an area of abnormality is seen on the cervix during colposcopy. There are two main methods — biopsy followed by recall for treatment if the histological result confirms CIN or immediate treatment [9]. The recent English guidelines discourage immediate treatment for women with low-grade smears, because of the potential for over-treatment [3]. However, trial evidence is lacking on whether immediate treatment or biopsy and recall is the most effective and efficient policy.

In 1995 the Medical Research Council and Department of Health issued a joint commissioning brief for a randomised-controlled trial of alternative referral and management policies for women with low-grade smears, to include a HPV component (BMJ, 10 June 1995, vol 310). TOMBOLA (Trial Of Management of Borderline and Other Low-grade Abnormal smears) is the result of that call.

This paper describes the design of TOMBOLA, the outcomes being considered, and discusses some of the issues arising in the design and implementation of this pragmatic trial nested within the NHS Cervical Screening Programmes in England and Scotland.

Section snippets

Aims

The aims of TOMBOLA are, in women with a low-grade smear:

  • To determine whether cytological surveillance or a colposcopic examination is the more effective and efficient management policy;

  • In those in whom an abnormal transformation zone is noted during colposcopy, to determine whether immediate treatment (using large loop excision of the transformation zone; LLETZ) or biopsies and recall if necessary for LLETZ is the more effective and efficient treatment policy;

  • To evaluate the contribution of

Trial design

TOMBOLA is a pragmatic randomised-controlled trial nested within the Cervical Screening Programmes in two areas of Scotland (Grampian and Tayside) and one area of England (Nottingham). Fig. 1 shows the trial design. Eligible women were recruited through hospital-based clinics. A swab sample from the cervix was taken at recruitment for HPV testing. Women were classified as high-risk HPV positive if they tested positive for any of 14 high-risk HPV types, or high-risk HPV negative if they did not

Outcomes

A range of clinical, psychosocial and health economic outcomes are being considered within TOMBOLA. These are summarised in Table 4. Table 5(a) and (b) describes the timing of data collection in relation to (a) clinical outcomes and (b) other outcomes ascertained by questionnaire.

Quality assurance

The cytopathology and histopathology laboratories in the three trial centres participate in national quality assurance schemes. In addition, quality assurance exercises are being undertaken to quantify levels of consistency in histological and cytological diagnosis within and between the trial centres.

A slide circulation has been undertaken to assess consistency in cytological grading. Slides from different smear gradings were circulated between cytopathologists in each of the trial centres.

Statistical power

Power calculations were undertaken when the trial was being planned. These were revised in light of the changes to eligibility criteria. In the comparison of cytological surveillance versus initial colposcopy, TOMBOLA has:

  • 80% power to detect a relative risk of 1.15 in cumulative incidence of CIN2/3, and 95% power to detect a relative risk of 1.2 (assuming an overall cumulative incidence of CIN2/3 of 15% and α = 0.05 for a two-sided test).

  • 80% power to detect a difference of 6% in the cumulative

Analysis

Analysis will be conducted on the basis of intention-to-treat. Comparison will be made between (i) the cytological surveillance and initial colposcopy arms and (ii) the immediate LLETZ and biopsy and selective recall arms. In the evaluation of cytological surveillance versus initial colposcopy, the following outcomes will be compared: cumulative incidence of CIN2/3 (ie CIN2/3 or more severe disease), cumulative proportion of clinically significant anxiety and/or depression, and default from

Trial organisation

The trial is managed according to MRC guidelines [26]. A Trial Steering Committee provides overall supervision for the trial. The Committee includes independent members and a chairperson who are not otherwise involved in the trial. A Data Monitoring and Ethics Committee reviews unblinded data from the trial, and has opportunity to make recommendations to the Steering Committee on any ethical or safety issues relating to the trial. Members of this committee are independent both from the trial

Current status

Recruitment to the trial closed on 31 January 2003. Almost four and a half thousand women were randomised at R1 (ie to either colposcopy or cytological surveillance). Thirty percent of these were recruited during the first recruitment phase and the remainder during the second recruitment phase. Around one-third of women randomised had a mild index smear. The recruitment rate was 52%. Of those randomised to colposcopy, just over 2000 women were randomised at R2 (ie to either immediate treatment

Discussion

When the joint commissioning brief for a randomised-controlled trial of alternative referral and management policies for women with low-grade smears was issued, the original design proposed by the MRC included an arm where HPV status would be disclosed and women who were HPV positive would be referred for immediate treatment. In designing TOMBOLA, the investigators felt that this would be scientifically sub-optimal for a number of reasons. First, such a design would result in confounding

References (65)

  • N. Munoz et al.

    Epidemiologic classification of human papillomavirus types associated with cervical cancer

    N Engl J Med

    (2003)
  • J. Cuzick et al.

    A systematic review of the role of human papillomavirus testing within a cervical screening programme

    Health Technol Assess

    (1999)
  • Refining the management of low-grade cervical abnormalities in the UK National Health Service, and defining the potential for HPV testing: a commentary on emerging evidence

    J Low Genit Tract Dis

    (2006)
  • P.L. Martin-Hirsch et al.

    Surgery for cervical intraepithelial neoplasia (Cochrane Review). Issue 2

    (2004)
  • J. Wise

    UK pilot scheme for HPV testing announced

    Br Med J

    (2000)
  • Current Controlled Trials. A randomised trial of human papillomavirus (HPV) testing in primary cervical screening...
  • S. Cotton

    Smear history in women with low-grade abnormal smears

  • M.V. Jacobs et al.

    A general primer GP5+/GP6(+)-mediated PCR-enzyme immunoasasy method for rapid detection of 14 high-risk and 6 low-risk human papillomavirus genotypes in cervical scrapings

    J Clin Microbiol

    (1997)
  • M.J. Garton et al.

    Estimating the point accuracy of population registers using capture–recapture methods in Scotland

    J Epidemiol Community Health

    (1996)
  • A.S. Zigmond et al.

    The hospital anxiety and depression scale

    Acta Psychiatr Scand

    (1983)
  • N.M. Gray et al.

    Developing a questionnaire to measure the psychosocial impact of an abnormal cervical smear result and its subsequent management: the TOMBOLA (Trial Of Management of Borderline and Other Low-grade Abnormal smears) trial

    Qual Life Res

    (2005)
  • EuroQol — a new facility for the measurement of health-related quality of life

    Health Policy

    (1990)
  • K.A. Wallston et al.

    Development of the multidimensional health locus of control (MHLC) scales

    Health Educ Monogr

    (1978)
  • T.M. Marteau et al.

    The development of a 6-item short-form of the state scale of the Spielberger State Trait Anxiety Inventory (STAI)

    Br J Clin Psychol

    (1992)
  • H.J. Eysenck

    A short questionnaire for the measurement of two dimensions of personality

    J Appl Psychol

    (1958)
  • M. Horowitz et al.

    Impact of event scale: a measure of subjective stress

    Psychosom Med

    (1979)
  • E. Fenwick et al.

    Representing uncertainty: the role of cost-effectiveness acceptability curves

    Health Econ

    (2001)
  • Medical Research Council

    MRC guidelines for good clinical practice in clinical trials

    (1998)
  • M. Schiffman et al.

    ASCUS-LSIL triage study. Design, methods and characteristics of trial participants

    Acta Cytol

    (2000)
  • M. Evander et al.

    Human papillomavirus infection is transient in young women: a population-based cohort study

    J Infect Dis

    (1995)
  • T.M. Becker et al.

    Sexually transmitted diseases and other risk factors for cervical dysplasia among southwestern Hispanic and non-Hispanic white women

    JAMA

    (1994)
  • Cited by (48)

    • Long-Term Worries after Colposcopy: Which Women Are at Increased Risk?

      2015, Women's Health Issues
      Citation Excerpt :

      The analysis was nested within the colposcopy arm of the multicenter Trial of Management of Borderline and Other Low-grade Abnormal smears (TOMBOLA). Full details are provided elsewhere (Cotton et al., 2006; TOMBOLA Group, 2009a, 2009b). Briefly, eligible women were aged 20 to 59 years and had a recent low-grade cytology test (mild dyskaryosis or borderline nuclear abnormalities [BNA]) taken as part of routine screening from October 1999 to October 2002.

    • Who defaults from colposcopy? A multi-centre, population-based, prospective cohort study of predictors of non-attendance for follow-up among women with low-grade abnormal cervical cytology

      2012, European Journal of Obstetrics and Gynecology and Reproductive Biology
      Citation Excerpt :

      Table 1 summarises other characteristics of the full study population and the subsample invited to complete the psychosocial questionnaire. As a result of changes to the eligibility criteria part-way through recruitment, to include women with a single BNA cytology result [22], there were some modest differences between the full study population and the subsample. Most notably, a higher proportion of the latter group had BNA cytology at recruitment.

    • Human papillomavirus deoxyribonucleic acid testing in developed countries

      2012, Best Practice and Research: Clinical Obstetrics and Gynaecology
      Citation Excerpt :

      Large-scale randomised-controlled trials with long-term follow up are needed for this. The Trial of Management of Borderline and other Low grade Abnormal (TOMBOLA) smears trial46 started recruiting women in December 1999 and was a 7-year, multi-centre trial run from Dundee, Aberdeen and Nottingham. It evaluated the contribution of HPV testing to the effectiveness and efficiency of the existing procedures for management of women with mild or borderline dyskaryosis.

    View all citing articles on Scopus

    TOMBOLA is funded by the Medical Research Council, the NHS in Scotland and the NHS in England.

    View full text