Mammographic screening: significance of minimal breast cancers

Analysis of cancers occurring during an aggressively performed screening program indicates: (1) In this study, during 3 years of active incidence screening of women under age 50, 24 cancers occurred, 14 of which were minimal. In the 3 years of very limited screening, 23 cancers occurred, ony five of which were minimal (P = 0.02393). (2) It was projected that 110 breast cancers should have occurred in the incidence years of observation of this self-selected population of 10,531 women. In the screened population, 124 cancers occurred. This is not significantly different from the expected number. (3) In a similar period of observation of a similar sized screened population in Louisville, Ky., there were 112 cancers, and during a similar period in Seattle, Wash., 113. There is no significant difference from the 124 reported in Cincinnati, Ohio, or the 110 expected. However of all the cancers (prevalent, incident, interval) in Cincinnati, 67 were minimal; in Louisville, 35; and in Seattle, 23 (P less tha 0.0001). All in all, these data support the concept that screening for breast cancer does not in any significant way increase the number of cancers detected; it only lowers the stage of detection. The data also suggest that, for an aggressive screen, length-biased sampling does not seem to be insurmountable obstacle.