Mammography moves on

A new analysis of data from four Swedish randomized controlled trials might finally bring an end to the debate surrounding the benefits of mammography screening. In the 16 March issue of The Lancet, Lennarth Nystrom and colleagues report that mammography screening for breast tumours does lead to a statistically significant reduction in cancer mortality in women aged 55 years or over. The study is an updated analysis of the large Swedish clinical trials that have been at the centre of that controversy. In a study published in 2001, Ole Olsen and Peter Gotzsche raised concern that mammography randomized trials — including the Swedish trials — were flawed, and did not provide reliable evidence to support the benefit of this screening procedure. Nystrom et al. therefore extended their analysis, performing a long-term follow-up study of the outcomes of 247,000 women over almost 16 years. They also determined the age-specific and trial-specific effects of mammography on breast cancer mortality, and re-examined their earlier data in light of Olsen and Gotzsche's critiques of their randomization procedures. In comparing the relative risks for breast cancer death (and death from all causes) between women who received mammography screening and controls, the authors reported 584 breast cancer deaths among the 1,688,440 women in control groups, but only 511 breast cancer deaths in 1,864,770 women who were invited for mammography screening. This represents a statistically significant overall reduction in breast cancer mortality of 21%. The reduction was greatest (33%) in the age group 60–69 years at entry to the trials. Nystrom et al. also report statistically significant effects in the age groups 55–59, 60–64 and 65–69 years, but only a small relative risk reduction (5%) in women aged 50–54 years. So, the benefits of mammography increase with age.

In an editorial accompanying the research, Karen Gelmon claims that the study shows “real but modest” benefits for screening, and states that the new analysis “reassures us that the Swedish data are believable”. Hopefully, the research will settle the debate among scientists and statisticians over the value of mammograms, and lend credence to the US government's recent recommendation that women older than 40 have the tests every 1–2 years.

ORIGINAL RESEARCH PAPER Nystrom, L. et al. Long-term effects of mammography screening: updated overview of the Swedish randomised trials. Lancet 359, 909–919 (2002)

FURTHER READING Olson, O. & Gotzsche, P. C. Cochrane review on screening for breast cancer with mammorgraphy. Lancet 358, 1340–1342 (2001) Gelmon, K. A. & Olivotto, I. The mammography screening debate: time to move on. Lancet 359, 904–905 (2002)

WEB SITE NCI statement on mammography screening