Mammographic screening from age 40 years

Abstract

Sue Moss and colleagues1Moss SM Cuckle H Evans A Johns L Waller M Bobrow L for the Trial Management GroupEffect of mammographic screening from age 40 years on breast cancer mortality at 10 years' follow-up: a randomised controlled trial.Lancet. 2006; 368: 2053-2060Summary Full Text Full Text PDF PubMed Scopus (401) Google Scholar report a 17% decrease in breast cancer mortality after a mean follow-up of 10·7 years and note that this was consistent with the 3% reduction in all-cause mortality they also found. This result probably reflects biased reporting at a favourable time point. In 1999, Moss and colleagues stated that the first mortality analysis was to be carried out after 7 years,2Moss S A trial to study the effect on breast cancer mortality of annual mammographic screening in women starting at age 40.J Med Screen. 1999; 6: 144-148Crossref PubMed Scopus (55) Google Scholar and in 2005 they reported data after 8·1 years, but curiously, only on all-cause mortality excluding breast cancer.3Moss S Thomas I Evans A Thomas B Johns L Randomised controlled trial of mammographic screening in women from age 40: results of screening in the first 10 years.Br J Cancer. 2005; 92: 949-954Crossref PubMed Scopus (38) Google Scholar This odd outcome did not favour screening, since there was a mortality increase of 4% in the screened group. I estimated the missing number of breast cancer deaths after 8·1 years from a mortality curve,1Moss SM Cuckle H Evans A Johns L Waller M Bobrow L for the Trial Management GroupEffect of mammographic screening from age 40 years on breast cancer mortality at 10 years' follow-up: a randomised controlled trial.Lancet. 2006; 368: 2053-2060Summary Full Text Full Text PDF PubMed Scopus (401) Google Scholar added them to the other deaths, and found a 1% increase in all-cause mortality with screening. I therefore urge Moss and colleagues to report numbers of breast cancer deaths after 8·1 years. Moss and colleagues used a cause of death register which inevitably leads to bias in favour of screening.4Gøtzsche PC Nielsen M Screening for breast cancer with mammography.Cochrane Database Syst Rev. 2006; 4 (CD001877.)Google Scholar It can be particularly difficult to determine the cause of death when the patient has more than one cancer. They should therefore report all-cancer mortality after 8·1 and 10·7 years. Furthermore, 82 women were never invited, although randomised to screening, some because they were dead or already had breast cancer.3Moss S Thomas I Evans A Thomas B Johns L Randomised controlled trial of mammographic screening in women from age 40: results of screening in the first 10 years.Br J Cancer. 2005; 92: 949-954Crossref PubMed Scopus (38) Google Scholar These women were not mentioned in the recent paper.1Moss SM Cuckle H Evans A Johns L Waller M Bobrow L for the Trial Management GroupEffect of mammographic screening from age 40 years on breast cancer mortality at 10 years' follow-up: a randomised controlled trial.Lancet. 2006; 368: 2053-2060Summary Full Text Full Text PDF PubMed Scopus (401) Google Scholar How many were excluded from analysis in the two allocation groups for such reasons? I declare that I have no conflict of interest. Mammographic screening from age 40 years – Authors' replyLuc Bonneux comments on the excess risk of diagnosis in the intervention group in our trial, but it is too soon to estimate the extent of any overdiagnosis, because there will still be an excess in this group owing to lead time. He also compares the absolute level of benefit and the number needed to screen observed in our trial with those associated, for example, with treatment of breast cancer by trastuzumab. However, as we state in our paper, the effect of the later screens in the trial will not have emerged yet. Full-Text PDF