[Biomorphological features of "minimal" carcinomas selected from a series of 2077 breast cancer cases]. / Caratteri biomorfologici dei carcinomi "minimi" selezionati da una serie di 2077 casi di carcinoma mammario.

Two hundred eighty nine pT1a/pT1b (less than < or = 1 cm in diameter) mammary carcinomas were selected from a series of 2077 consecutive cases of breast carcinoma. When compared with carcinomas of a larger size, they were significantly associated with a lower histological grade (SBR), a lower growth fraction (Ki-67 antigen in less than 20% of neoplastic cells) and a lower number of positive cases (more than 10% of neoplastic cells) as far as p53, c-erbB-2 oncoproteins and EGF-R, as detected by immunohistochemical methods, are concerned. Moreover, a significantly higher number of estrogen and progesterone receptors positive cases (more than 10% of positive cells), and a higher frequency of "other" histotypes with a favourable prognosis, was detected. No difference between pT1a and pT1b cases was noted so that "1 cm size" only stratifies a class with a better prognosis. It is possible that small cases, although invasive, did not undergone sufficient mytotic cycles to produce the sequence of genetic changes that characterize the "no return" phase of breast cancer. A 17% of pT1a and pT1b cases displayed lymph node metastases (mean 26 lymph node/case examined): the number is so high that cautions about simple lumpectomies and about sentinel lymphadenectomy, even in cases of small cancers, are necessary, until a larger number of studies will become available.

Quantitative growth fraction evaluation with MIB1 and Ki67 antibodies in breast carcinomas.

Ki67 and MIB1 monoclonal antibodies are directed against different epitopes of the same proliferation-related antigen. Whereas Ki67 works only on frozen sections, MIB1 may be used also on fixed sections. The authors immunostained a series of 40 breast carcinomas with MIB1 and Ki67 antibodies on serial frozen sections and on fixed material. The Ki67 labeling index (LI) was 12.9 +/- 8.9 and 12 (mean +/- SD and median, respectively). MIB1 LI was 21.2 +/- 11.9 and 19.5 on frozen sections and 24 +/- 15.2 and 21.5 on fixed sections (mean +/- SD and median, respectively). Ki67 LI and MIB1 LI on frozen and fixed sections were strictly correlated (P < .001). The results are in keeping with the reported coincidental nuclear staining pattern of Ki67 and MIB1, but the mean and median values of MIB1 LI are almost twice the values of Ki67 LI. The cut-off values to define high and low proliferative activity with the two antibodies are therefore different. The differences in immunolabeling may be due to better survival of the MIB1 epitope in freezing and acetone fixation or to differing accessibilities of the MIB1 and Ki67 epitopes during the cell cycle due to molecular conformational modifications. The MIB1 monoclonal antibody is a reasonable substitute for the Ki67 monoclonal antibody. The advantages of MIB1 immunostaining on paraffin sections include the feasibility of retrospective studies and of obtaining clear morphologic specimens that are optimal for use with computer-assisted image analysis systems. Our image-processing system allows automatic nuclear counting, detects positive nuclei and measures their staining intensity.