Editorials

What should be done about interval breast cancers?

The paper by Woodman and colleagues in this week's journal contains the first published data from the NHS breast screening programme on cancers diagnosed in the three years after a negative screen ("interval cancers") (p 224).¹ It reports 15.8 interval cancers per 10000 women screened in the North West region in the 24 months after screening, which is higher than the target that was set for Britain. Preliminary data from 46 screening centres, collected by the National Breast Screening Radiology Quality Assurance Committee, are very similar, suggesting that the North West region is representative of the United Kingdom as a whole.

Comparing rates of interval cancer with those reported in the literature is difficult because different groups have reported the data using different methods and the populations studied have different age compositions. The rates reported today, however, are higher than the rate of 9.4/10000 in the Swedish two county trial² and are very similar to the rates reported from the Nijmegen (15.7/10000)³ and Stockholm (19.2/10000) trials.⁴ In Nijmegen in the eight successive screening rounds since their two yearly screening programme began in 1975, the rates of interval cancer with screening every two years have always exceeded the rates initially observed, and after the seventh round they peaked at 24.9/10000 against a rate of cancer detected by screening of between 2.95/1000 and 3.89/1000.⁵ Understandable concern therefore exists about the sensitivity of the screening test and whether the Health of the Nation's target for reducing deaths from breast cancer in women invited for screening by 25% by 2000 can be achieved.

The protocol for the NHS breast screening programme was heavily influenced by the Swedish two county study⁶; a single mediolateral oblique view of each breast is obtained every three years, with each film being read by one radiologist. At the time this was considered to be the most cost effective method of screening. A comprehensive training and teaching programme for all staff was introduced nationally.

Even as the Department of Health's working party was recommending a national breast screening programme to the government,⁷ evidence was appearing to suggest that the protocol chosen for the programme might not be optimal. Two view mammography may detect 20% more women with breast cancer than one view mammography.^{8 9 10 11} Preliminary comparisons of the results at British centres using one view or two view screening have shown highly significant differences, with increased rates of detection of cancer and decreased recall rates in centres using two views. Preliminary data available from the quality assurance programme indicate that use of two views also reduces the number of interval cancers. The results from the United Kingdom Coordinating Committee on Cancer Research one view versus two view trial are awaited with interest. Screening with mammography films with a mean optical density of more than 1.4 may improve detection--analysis of screening data collected centrally has found that more small cancers are detected in centres using such films.¹² Up to 15% more cancers are detected if the films are read by two radiologists,^{13 14} although this can also reduce specificity.¹⁵ Viewing by two readers was introduced into the two county study at an early stage.²

Reader problems

Preliminary analysis of interval cancers in the United Kingdom suggests that up to 30% of false negative findings on screening result from the film readers' failure to perceive or recognise as suspicious subtle but definite mammographic features, such as asymmetric density or poorly defined masses. Of the remaining interval cancers, about 8.5% are mammographically occult and 62.5% are classified as "true interval cancers." These proportions closely resemble those previously reported from Stockholm⁴ and Nijmegen,⁵ although exact comparisons are difficult because of differences in the methods of classification.

Between 50% and 60% of interval cancers in the United Kingdom occur in the third year after screening. The original organisers of the Swedish two county trial argued that intervals of two years rather than three years should be adopted,⁶ and most other national programmes now repeat screening at two year intervals.

How may the rate of interval cancers be reduced? Two view mammography should be introduced as soon as possible for women having their first screening examination. The recommendation to increase the optical density of mammogram films at no financial cost has already been implemented through the quality assurance network. A working party of the National Breast Screening Radiology Quality Assurance Committee is examining the feasibility of double reading throughout the NHS breast screening programme. A programme to give radiologists additional training in detecting the subtle mammographic signs that account for most "false negative" interval cancers is being established. Decreasing the screening interval to two years clearly has considerable financial and staffing implications; currently Britain has too few suitably trained radiologists. The results of the United Kingdom Coordinating Committee on Cancer Research breast screening frequency trial which is designed to identify the optimum interval for screening should become available next year.

The existence of a properly funded national quality assurance programme with a comprehensive feedback system has ensured that, although the preliminary results from the United Kingdom's national programme seemed satisfactory, the occurrence of a higher than expected rate of interval cancers has been recognised early. Immediate steps are being taken to identify the reasons why and implement corrective action.

Consultant radiologist Kent and Canterbury Hospitals NHS Trust, Canterbury CT1 3NG

Consultant radiologist King's Healthcare NHS Trust, London SE5 9RS

Director Breast Screening Unit, Coventry and Warwickshire Hospital, Coventry CV1 4FH

Director Breast Screening Training Centre, City Hospital NHS Trust, Nottingham NG5 1PB

Dr Field is the radiology representative on the Department of Health's advisory committee for breast cancer screening. Dr Michell is chairman and Drs Wallis and Wilson are regional representatives on the National Breast Screening Radiology Quality Assurance Committee.

S Field, M J Michell, M G W Wallis, A R M Wilson 

1. Woodman CBJ, Threlfall AG, Boggis CRM, Prior P. Is the three year breast screening interval too long? Occurence of interval cancers in NHS breast screening programme's north western region. BMJ1995;310:224-6.
2. Tabar L, Fagerberg G, Day NE, Holmberg L. What is the optimal interval between mammographic screening examinations? An analysis based on the latest results of the Swedish two county breast cancer screening trial. Br J Cancer 1987;55:547-51. [Medline]
3. Peeters PHM, Verbeek ALM, Hendrix JHCL, Holland R, Mravunac M, Vooijs GP. The occurence of interval cancers in the Nijmegen screening programme. Br J Cancer 1989;59:929-32. [Medline]
4. Frisell J, Eklund G, Hellstrom L, Somers A. Analysis of interval breast carcinomas in a randomised screening trial in Stockholm. Breast Cancer Research and Treatment 1987;9:219-25. [Medline]
5. Van Dijck JAAM, Verbeek ALM, Hendrix JHCL, Holland R. The current detectability of breast cancer in a mammographic screening program. Cancer 1993;72:1933-8. [Medline]
6. Tabar L, Fagerberg CJG, Gad A, Baldetorp L, Holmberg LH, Grontoff O, et al. Reduction in mortality from breast cancer after mass screening with mammography. Lancet 1985;i:829-32.
7. Breast cancer screening. Report to the health ministers of England, Scotland and Northern Ireland by a working group chaired by Sir Patrick Forrest. London: HMSO, 1986.
8. Anderson I, Hildell J, Muhlow A, Pettersson H. Number of projections in mammography: influence on detection of breast disease. Am J Roentgenol 1978;130:349-51. [Abstract]
9. Bassett LW, Bunnell DH, Jahanshahi R, Gold R, Arndt RD, Linsman J. Breast cancer detection: one versus two views. Radiology 1987;165:95-7. [Abstract/Free Full Text]
10. Thurfjell E. One versus two view mammographic screening. A prospective trial. Svenska Lakaresaliskapets Handlingar Hygiea 1990;99:220.
11. Van Dijck JAAM, Verbeek ALM, Hendrix JHCL, Holland R. One-view versus two-view mammography in baseline screening for breast cancer: a review. Br J Radiol 1992;65:971-6. [Abstract]
12. Young KC, Wallis MG, Ramsdale ML. Mammographic film density and detection of small breast cancers. Clin Radiol 1994;49:461-5. [Medline]
13. Anderson EDC, Muir BB, Walsh JS, Kirkpatrick AE. The efficacy of double reading mammograms in breast screening. Clin Radiol 1994;49:248-51. [Medline]
14. Thurfjell EL, Lernevall KA, Taube AA. Benefit of independent double reading in a populationbased mammographic screening program. Radiology 1994;191:241-4. [Abstract/Free Full Text]
15. Elmore JG, Wells MPH, Lee CH, Howard DH, Feinstein AR. Variability in radiologists' interpretation of mammograms. N Engl J Med 1994;331:1493-9. [Abstract/Free Full Text]
16. Tabar L, Fagerberg G, Dully SW, Day NE, Gad A, Grontoft O. Update of the Swedish two county program of mammographic screening for breast cancer. Radiol Clin North Am 1992;30:187-210. [Medline]