To the Editor: Using the Surveillance, Epidemiology, and End Results (SEER) database from 1976 to 2009, Dr Johnson and colleagues1 reported a statistically significant increase in the incidence rate of breast cancer with distant involvement in young women, but no significant increase in the incidence of localized or regional disease. The authors also stratified the analysis by estrogen receptor (ER) and progesterone receptor (PR) status. They reported an annual percent change (APC) of 8.15 for ER positive (ER+) and PR+ disease (P<.001), 8.89 for ER+ and PR negative (PR-) (P<.001), -0.51 for ER-PR+ (P=.79), and 5.44 for ER-PR- (P< .001). We question whether these observations are accurate because there was a considerable proportion of missing values on ER and PR status, especially during the earlier years, in the SEER database.
展开▼