The effect of changes in tumor size on breast carcinoma survival in the U.S.

The article by Elkin et al.1 provides interesting insights into the decrease in breast cancer deaths that has occurred in the United States, for the first time in 50 years, beginning in 1990. It would be interesting to learn whether they could refine their analysis to take into account the onset of mammography screening in the United States that occurred during 1983 to 1984 in sufficient numbers to affect National Statistics (Surveillance, Epidemiology, and End Results).2 At that time, not only was there a sudden increase in the detection of invasive cancers, but there was also an abrupt increase in the detection of ductal carcinoma in situ, which is detected virtually only by mammography, indicating that 1983 to 1984 was the beginning of widespread mammography screening. In fact, their graphs suggest that there was little size shifting within stages until 1985, when the sizes began to shift consistent with the onset of mammography screening. Unfortunately, the authors have made the mistake that has been made repeatedly by others3 by combining the data for women ages 49 and younger. I presume they did this thinking that these were the “premenopausal” women and that they may be grouped legitimately. This is not the case with regard to screening,4 and it causes the results to be biased5 by the very young women. I would encourage the authors, at the least, to evaluate women ages 40 to 49 years separately from the younger women, because these are the women who have been encouraged to be screened in the United States. Although breaking the data into even smaller age groups increases statistical fluctuation, it may provide greater insight into the benefits of screening that do not change suddenly at the age of 50 years, or ay any other age. Daniel B. Kopans MD*, * Avon Foundation Comprehensive Breast Evaluation Center, Wang Ambulatory Care Center, Massachusetts General Hospital, Boston, Massachusetts.