Search

The distribution of PECAM-1/CD31 molecule was investigated in 133 breast carcinomas using monoclonal antibody and frozen sections. Anti-CD31 labels endothelial cells and reflects stromal angiogenesis. The CD31 immunoreactivity was evaluated by computer-assisted analysis of digitized microscopic images. The automatic screening of the whole preparation and the measurements of the mean CD31 immunostained surface was performed in each case. A similar procedure was achieved for p53, cathepsin D, P-gp, pHER-2/neu, Ki67, pS2 estrogen and progesterone antigenic sites immunodetection. The image analysis of positive CD31 surface was variable, ranging from 4% to 33% (mean 14.7%, SD = 5.43). The CD31 positive surface correlated (P < .01) with the Nottingham prognostic index, but not with the tumor size, the node status, the tumor grade, nor with the patient age. Also the CD31 immunoreactivity was independent of the pHER-2/neu, Ki67 antigen, p53, ER, PR and pS2 immunodetectable expression in tumors, but correlates with that of cathepsin D (P = .024) and P-gp (P = .028), which reflects the multi-drug resistance capacity of tumor cells. In conclusion, CD31 positive vessels assessed on frozen sections by image analysis constitute an excellent method of evaluating tumor stromal angiogenesis, and can be further used for clinical purposes. The results also suggest that the CD31/PECAM molecule may be involved in the spread of tumor by interacting with extracellular matrix lysis that results from the tumor cell proteasic activity and with multidrug resistance. (Key words : PECAM/CD31 ; Immunohistochemical assays ; Image analysis) Am J Clin Pathol 1995 ;103 :443-448.

A series of 222 tumor samples stored at –80 °C in the authors’ tumor library were investigated with anti- p53 (PA 1801) and streptavidin-biotin- peroxidase complex. The p53 immunoprecipitates were quantified by densitometry assessed by image analysis of digitized microscopic images. Two parameters, percentage of positive surface and mean optical densities, were compared with the patient’s outcome (follow-up = 96.8 months) (life table method, Mantel Cox test, BMDP statistical software). The p53 expression significantly correlated with a poor overall survival ( P = .0063), metastasis-free survival ( P = .024), and recurrence-free survival ( P = .022) at a 20% cutoff point of positive immunoreactive tumor surface. A strong prognostic significance was observed in the node-positive subset of patients but not in the nodenegative subset, except for recurrence-free survival ( P = .047). The results indicate the clinical relevance of p53 evaluated by quantitative immunocytochemistry.