The diagnosis of ductal carcinoma in situ of the breast (DCIS) has become common with the advent of breast screening programs.
Methods: Proliferation indices (S-phase fraction) were studied in 76 cases of pure DCIS. Tumors were classified according to conventional criteria and also according to a novel simplified classification based on cellular necrosis and morphology. This new classification defines three distinct tumor groups: pure comedo in 19 (25.0%) cases, DCIS with necrosis (non-pure comedo) in 21 (27.6%) patients, and DCIS without necrosis in 36 (47.4%) of cases, the latter group comprising largely classical cribriform or micropapillary architectural subtypes.
Results: Flow cytometric DNA analysis showed a significantly higher S-phase fraction in comedo DCIS than in the subgroup of DCIS tumors without necrosis (P < 0.01 [anova]). A preliminary analysis of disease recurrence and disease-free survival in a large series of 391 cases of pure DCIS showed that of 181 cases of pure comedo DCIS there were 19 local recurrences at the 7-year stage (82% 7-year disease-free survival), with 5 local recurrences in 51 cases of DCIS with necrosis (non-pure comedo) (85% 7-year disease-free survival) and only 6 local recurrences in the 159 cases of the DCIS-without-necrosis subgroup (94% 7-year disease-free survival). The chi 2 value for the frequency of disease recurrence of all cases of DCIS with necrosis (i.e., combining the groups of comedo DCIS and DCIS with necrosis (non-pure comedo)) as compared to DCIS without histological evidence of necrosis was 5705 (df = 2; P = 0.0001), and the chi 2 for disease-free survival of types of DCIS with necrosis as compared to cases without necrosis was 178 (df = 2; P = 0.0001). This analysis indicates that the histological presence of necrosis appears to be a relatively powerful predictor of increased disease recurrence and poorer disease-free survival after treatment for DCIS.
Conclusions: Necrosis in DCIS in the absence of pure classical comedo morphology is a feature of more biologically aggressive in situ breast cancer with an intermediate proliferative fraction as compared with the high proliferative fraction of pure comedo DCIS and the low proliferative fraction of DCIS without necrosis. There was no significant difference in DNA ploidy (diploid or aneuploid) between the subgroups as assessed by chi 2 analysis. Further larger studies are required to establish if DCIS with necrosis (non-pure comedo) also shows a greater tendency to local recurrence after breast conservation treatment than do subtypes of DCIS without necrosis. DCIS with necrosis (non-pure comedo) should be adopted as a distinct histological subgroup of DCIS in future clinical studies of in situ mammary carcinoma.