A clinical index for selection of breast cancer patients for adrenalectomy.

Selection of patients with disseminated breast cancer for endocrine surgery has not been facilitated by sophisticated biochemical techniques. The present study was designed to produce a method of prognosis based exclusively on clinical data. Eighteen features were studied in 95 patients who underwent adrenalectomy and oophorectomy, and correlated with their degree of response to produce predictive variables. In a further group of 90 patients a prognostic score (the sum of the predictive variables) was calculated. There were 17 patients who scored less than 0 and in no case was a good response to endocrine surgery achieved. There were 40 who scored over 10, 29 of whom had a good response. The remaining 33 scored between 0 and 10 and there was no correlation between score and response. This predictive system offers significant advantages over accepted methods of clinical assessment (chi2 = 6-61, P less than 0-02). When patients with a score between 0 and 10 are excluded, its advantages are even more obvious (chi2 = 11-26, P less than 0-001). Patients who are likely to respond to endocrine surgery can be selected by clinical parameters alone, and total failure to respond can be forecast with complete accuracy. Between these two there is a third group whose response is unpredictable.