Breast cancer and earlier detection efforts. Realized and unrealized impact on stage

M. F. Farwell, R. S. Foster Jr and M. C. Costanza
Department of Surgery, University of Vermont, Burlington.

OBJECTIVE--To test our hypotheses that increased public and professional education in breast cancer detection (screening mammography, clinical breast examination, and breast self-examination) would lead to detection at earlier stages of disease with each of the three methods. DESIGN AND SETTING--A survey study of all pathologically diagnosed breast cancers in a defined geographic area (all nonfederal general hospitals in the state of Vermont) before (1975-1984, n = 1652) and after (1989-1990, n = 683) screening mammography became more commonly used. MAIN OUTCOME MEASURES--Method of detection of breast cancer and stage at detection. RESULTS--The age-adjusted annual incidence rate of breast cancer among adult women was 99 per 100,000 during the years 1975 to 1984 compared with 169 per 100,000 during 1989-1990 (P < .001). Mammography as the method of detection increased in use from 2% to 36% (P < .001). In the later era, invasive breast cancers were detected at more favorable TNM stage (P < .001); mean maximum tumor diameter was smaller (2.2 cm vs 2.7 cm; P < .001); percentage of histologically positive nodes was lower (37% vs 47%; P < .001). When mammographically detected cancers were excluded from the analysis of invasive cancers, the mean maximum tumor diameter and percentage of negative nodes in the two eras were similar even though there was an apparent shift to detection at an earlier clinical stage. CONCLUSION--The earlier stages of detection and the sudden increase in incidence could almost entirely be credited to screening mammography. Mammographic screening had a much greater impact on stage at detection in women aged 50 years and older than in those younger than 50 years.