We appreciate Dr. Kopans's comments regarding our analysis of shifts over time in the size distribution of newly diagnosed breast tumors and the impact of these shifts on within-stage breast cancer survival.1 To address Dr. Kopans's concern that we inappropriately grouped women ages 25 to 39 years with those ages 40 to 49 years in our susbet analysis by age at diagnosis, we repeated the analysis excluding patients under age 40 years. In women ages 40 to 49 years with localized breast cancer, standardization of relative survival on the basis of tumor size explained 43% of the increase in relative survival between the cohort of women who were diagnosed from 1975 to 1979 and the cohort of women who were diagnosed from 1995 to 1999; this is slightly greater than the 38% of survival improvement explained by tumor size standardization in the entire group of women ages 25 to 49 years. In those with regional breast cancer, tumor size standardization explained 21% of the survival improvement over time, slightly less than the 23% of survival improvement explained by tumor size standardization in all women ages 25 to 49 years. In both localized and regional breast cancer, in all years studied, women ages 40 to 49 years accounted for at least two-thirds of all patients ages 25 to 49 years.

We do not suggest that the benefits of screening or of any other breast cancer intervention change abruptly at age 50 years. Rather, we chose our age categories (25–49 years, 50–64 years, and 65 years and older) so that our findings could be interpreted in the context of the many screening, treatment, epidemiologic, health services, and other studies that use similar or identical age categories. In addition, as we stated in our article, within-stage shift in the tumor size distribution likely reflects the effects of increased use of screening mammography in the United States, but its association with improvement in breast cancer survival is not definitive evidence of the effectiveness of screening. This applies to Dr. Kopans's observation that most of the change in crude relative survival for localized disease appears to occur in the mid-1980s. We believe that tumor size standardization is a methodologic refinement that improves our ability to interpret secular trends in breast cancer survival.


  1. Top of page

Elena B. Elkin PhD*, Clifford Hudis MD*, Colin B. Begg PhD*, Deborah Schrag MD, MPH*, * Department of Epidemiology and Biostatistics, Memorial Sloan-Kettering Cancer Center, New York, New York.