Differences in Ki67 and c-erbB2 Expression between Screen-detected and True Interval Breast Cancers

• Published in: Clinical cancer research 1999-10, Vol.5 (10), p.2682-2688
• Main Authors: CROSIER, M, SCOTT, D, WILSON, R. G, GRIFFITHS, C. D. M, MAY, F. E. B, WESTLEY, B. R
• Format: Article
• Language: English
• Subjects: Aged; Biological and medical sciences; Biomarkers, Tumor - analysis; Breast Neoplasms - chemistry; Female; Gynecology. Andrology. Obstetrics; Humans; Immunohistochemistry; Ki-67 Antigen - analysis; Logistic Models; Mammary gland diseases; Medical sciences; Middle Aged; Receptor, ErbB-2 - analysis; Retrospective Studies; Tumor Suppressor Protein p53 - analysis; Tumors
• ISSN: 1078-0432; 1557-3265

Breast cancer screening facilitates the early detection of breast cancer, although a significant number of tumors still arise in the interval between screening. The objective of this study was to measure the expression of five markers of proven prognostic significance in symptomatic breast cancer (estrogen receptor, progesterone receptor, p53, Ki67, and c-erbB2) in screen-detected and interval breast cancers to identify biological markers that may be associated with the emergence of symptomatic breast cancer in the screening interval. The expression of estrogen receptor, progesterone receptor, p53, Ki67, and c-erbB2 was assessed in a series of 51 true interval and 84 screened-detected invasive tumors by immunohistochemistry. Interval cancers tended to be of higher histological grade and were of larger pathological size than screen-detected cancers. Expression of estrogen receptor was 1.7-fold lower ( P < 0.001), whereas expression of p53 was 2.5-fold ( P < 0.01), Ki67 2.4-fold ( P < 0.001), and c-erbB2 3.6-fold higher ( P < 0.01) in true interval cancers compared with screen-detected invasive cancers. There was no significant difference in progesterone receptor expression. The most important differences identified by multiple logistic regression analysis were in the expression of Ki67 and c-erbB2. The differences in the expression of these markers were more important than clinical features such as pathological grade and size. Using the logistic regression model, 83% of the tumors analyzed in this study could be correctly assigned as interval or screen-detected tumors on the basis of Ki67 and c-erbB2 expression. The importance of high expression of Ki67 in interval cancers compared with screen-detected cancers suggests that tumors may become symptomatic in the screening interval as a result of increased levels of cell proliferation. The inclusion of c-erbB2 in the regression equation suggests that this growth factor receptor may play a significant role in stimulating the rapid growth of interval cancers.