Is carcinoma in situ a precursor lesion of invasive breast cancer?

This study measures the probability of development of invasive breast cancer ( BC) following the diagnosis of carcinomas in situ (CIS). A 25-year prospective follow-up was conducted by linking the Canadian National Breast Screening Study (CNBSS) to cancer registries and a national vital statistics database. Subsequent BC incidence was identified in CNBSS women who were diagnosed with CIS. CIS was classified into ductal (DCIS) and lobular carcinoma in situ (LCIS). Cumulative cancer incidence probabilities were calculated and a 1: 5 matched nested case control study was conducted to estimate the odds of BC development. Of the 146 women diagnosed with CIS, 26 developed invasive BC (17.8%) and 12 died of BC (8.2%). The average time from the diagnosis of CIS to invasive BC was 6.3 years (+/-5.6). The 20-year cumulative incidence probabilities for DCIS and LCIS were 19.0% (95% CI: 11.2, 26.8) and 21.3% (95% CI: 7.1, 35.4) respectively. The odds of development of BC in CIS women was significantly elevated compared with controls (OR=2.6, 95% CI: 1.5, 4.5). While women with CIS had a higher odds of development of BC compared to those without CIS, at 20-year post CIS diagnosis, more than 80% of them remained free of invasive BC. This low probability of developing invasive BC post CIS diagnosis does not support the notion that CIS of the breast is an obligate precursor lesion of invasive BC.