Volume 38, Issue 5 pp. 637-642
Human Cancer
Full Access

Androgen, estrogen, and progesterone receptor levels in malignant and benign breast tumors: A multivariate analysis approach

Maria Mitzi Brentani

Maria Mitzi Brentani

Faculdade de Medicina da Universidade de São Paulo, Laboratório de Oncologia Experimental, Av. Dr. Arnaldo, 455, 01246 São Paulo

Search for more papers by this author
Eduardo Luis Franco

Corresponding Author

Eduardo Luis Franco

Ludwig Institute for Cancer Research—São Paulo Branch, Rua Prof. Antonio Prudente, 109, 01509 São Paulo

Ludwig Institute for Cancer Research—São Paulo Branch, Rua Prof. Antonio Prudente, 109, 01509 São PauloSearch for more papers by this author
Celina Tizuko F. Oshima

Celina Tizuko F. Oshima

Fundação Centro de Pesquisa em Oncologia, Rua Oscar Freire, 2396, 05409 São Paulo, Brazil

Search for more papers by this author
Mercia Medeiros Pacheco

Mercia Medeiros Pacheco

Fundação Centro de Pesquisa em Oncologia, Rua Oscar Freire, 2396, 05409 São Paulo, Brazil

Search for more papers by this author
First published: 15 November 1986
Citations: 19

Abstract

Androgens have been frequently used in the treatment of breast cancer. However, objective responses seem to vary according to the steroid hormone receptor expression of the tumor. We have studied the relationship between concentrations of androgen receptors (AR) and those of estrogen (ER) and progesterone (PR) receptors by multiple regression and stratified analysis techniques in 154 cases of primary breast carcinomas and 39 cases of benign tumors. Both the proportion of AR-positive tumors and the concentration of AR were dependent upon the coexpression of ER and PR by the specimens. This association was evident for malignant tumors with predominance of the positive correlation between AR and ER over that of AR and PR in post-menopausal patients. In premenopausal women, ER and PR concentrations were similarly correlated with AR levels. The PR, but not the ER concentration, was positively correlated to AR levels in benign breast tumors. These findings were confirmed by multiple regression, taking into consideration additional information about the patients to build statistical models allowing prediction of AR. Among the other variables considered in building these models—age, menopausal status, weight, height, and clinical stage—only height (using data from all patients) and age (data from post-menopausal women) emerged in addition to ER and PR as significantly explaining the variability of AR as the dependent variable.

The full text of this article hosted at iucr.org is unavailable due to technical difficulties.