Elsevier

The Lancet

Volume 353, Issue 9168, 5 June 1999, Pages 1909-1914
The Lancet

Articles
16-year mortality from breast cancer in the UK Trial of Earaly Detection of Breast Cancer

https://doi.org/10.1016/S0140-6736(98)07412-1Get rights and content

Summary

Background

The UK Trial of Early Detection of Breast Cancer (TEDBC) is a non-randomised study, which was set up in 1979 to investigate the effect of screening and education about breast self-examination on breast-cancer mortality. We report mortality results after 16 years of follow-up, including results by age at trial entry.

Methods

Eight centres (two screening, two breast self-examination, and four comparison) in England and Scotland recruited women aged 45–64 years into the initial cohort, with women reaching age 45 years during the 7-year study period included in later cohorts. The observed number of deaths from breast cancer in each centre was compared with the expected number, which was calculated by Poisson regression model; expected numbers were adjusted for pretrial breast-cancer mortality. Results were analysed by 5-year age-groups, and for women aged 45–46 years and 47–49 years at entry.

Findings

Breast-cancer mortality was 27% lower (rate ratio 0·73 [95% CI 0·63–0·84]), adjusted for pretrial rates, in cohort 1 in the two screening centres combined than in the comparison centres. No reduction in mortality in the two breast self-examination centres combined was seen (rate ratio 0·99 [0·87–1·12]). The mortality reduction in the screening centres did not differ significantly across age-groups; a 35% reduction was seen in women in all cohorts aged 45–46 years at entry (rate ratio 0·65 [0·50–0·86]). Results were similar when deaths were restricted to those in patients diagnosed within 10 years of trial entry.

Interpretation

The results from TEDBC support those from randomised trials in Edinburgh and elsewhere, and show that a reduction in breast-cancer mortality resulting from screening can be achieved in the UK. There was no evidence of less benefit in women aged 45–46 years at the start of screening; the effect of screening in this age-group begins to emerge after 3–4 years.

Introduction

The efficacy of screening for breast cancer by mammography in women aged 50 years and older is now well established from the results of several randomised trials, and several countries have implemented or are planning nationwide screening programmes. However, despite further analyses, the efficacy of screening of women aged younger than 50 years continues to be the subject of considerable debate.

The UK Trial of Early Detection of Breast Cancer (TEDBC) was set up in 1979 to investigate the effect on breast-cancer mortality of both screening (by mammography and clinical examination) and an education programme in breast self-examination.1 TEDBC was non-randomised, and designed to compare eight geographically separate districts in England and Scotland. After a mean 6·6 years of follow-up, breast-cancer mortality was 14% lower in the two populations offered screening than in the comparison populations, the difference increasing to 20% when differences in pretrial mortality were taken into account (p=0·06).2 After 10 years of follow-up, this difference remained similar (p=0·01), with no difference from the comparison centres observed in the two breast self-examination centres combined, despite a significant difference between the two.3

In common with most randomised trials, TEDBC was not designed to study the effect of screening in specific age-groups. However, meta-analyses of randomised trials have been used to argue for and against the existence of an effect of mammography in women younger than 50 years.4, 5 The 10-year analysis of TEDBC showed no significant difference between the effect of screening in women aged 45–49 years at entry and that in older women.3 For a further 4–6 years the intervention centres in TEDBC continued to recruit additional women into the study as they reached 45 years, giving a large population in this age-group. We report the mortality results up to 16 years of follow-up and look in more detail at the effect of screening according to age at entry.

Section snippets

Study design

The details of the trial are described elsewhere.1 The two screening centres, Guildford and Edinburgh, invited women aged 45–64 years to screening sessions every year for 7 years, with screening by mammography and clinical examination every 2 years and by clinical examination only in the intervening years. The two breast self-examination centres, Huddersfield and Nottingham, offered women in the same age-range an invitation to attend an education session on breast self-examination on entry to

Results

Table 2 shows the numbers of women attending their first screening session and the average number of screenings per woman invited by age at entry in cohorts 1 and 2–7. Attendance at first screening was 60% in Edinburgh and 72% in Guildford and fell slightly with increasing age. Not all women invited for screening in each year were invited at subsequent rounds, because some moved away from the district or died. Attendance at subsequent screenings also varied according to the protocol for

Discussion

These results show the long-term effect of screening in TEDBC: in women in the initial cohort breast-cancer mortality was 27% lower in the screening centres than in the comparison districts at 16 years of follow-up. Cumulative mortality in the screening and comparison centres continued to diverge up to 16 years of follow-up.

Follow-up of this study to the end of 1995 meant that several years were included during which the UK NHS breast-screening programme, introduced over a 5-year period from

References (18)

There are more references available in the full text version of this article.

Cited by (140)

  • Evidence for reducing cancer-specific mortality due to screening for breast cancer in Europe: A systematic review

    2020, European Journal of Cancer
    Citation Excerpt :

    Considering the results of all 60 studies included in this review, BC mortality reduction estimates for invited versus non-invited women varied from 4% [89] to 36% [90] in Northern Europe, from 25% [59,82] to 35% [56] in Southern Europe and from 6% [66] to 47% [67] in Western Europe. When comparing BC mortality of screened versus non-screened women, estimates varied from 2% [50] to 89% [38] in Northern Europe, from 43% [86]to 67% [57] in Southern Europe and 12% [64] to 58% [80] in Western Europe. Of the 60 included studies, 40 had statistically significant results.

  • An exploratory analysis of the factors leading to delays in cancer drug reimbursement in the European Union: The trastuzumab case

    2014, European Journal of Cancer
    Citation Excerpt :

    Improvements in breast cancer mortality are related to improvements in several areas of cancer care. Breast cancer screening has been proven to reduce mortality in European populations [41–43]. Breast cancer treatment improvements have been observed in all areas of cancer care, as in the case of surgery and radiotherapy [44], endocrine treatment [37] and chemotherapy [45].

  • An update on breast cancer screening and prevention

    2014, Primary Care - Clinics in Office Practice
    Citation Excerpt :

    A few large trials have analyzed the effect of instructing women in breast self-examination (BSE) on reducing breast cancer mortality. The UK Trialists study, a nonrandomized study with 16 years of follow-up, showed no significant difference in breast cancer mortality between the BSE and control groups.83 A Cochrane review included randomized trials comparing BSE with control groups in both Russia84 and Shanghai,85 with each study showing no significant differences in breast cancer mortality after 13 and 11 years, respectively.86

View all citing articles on Scopus

Members of group listed at end of paper

View full text