Urinary Steroid Estimations in the Prediction of Response to Adrenalectomy or Hypophysectomy.

described the first series of adrenalectoniies amid Luft and Olivecrona (21) reported on hypophysectoniny as a treatment for advanced cancer of the breast. When patients are selected at ran dom for these procedures only 25 % obtain a worthwhile remissioni (12), and attempts to predict response using cli nical features, histologic differentiation, or calcium metab olism (13, 18, 23) have not found gemieral acceptance. The use of steroid measurements mi this connection dates from the introduction of chemical methods for esti mating small amounts of estrogens in the urine (7, 9). i\'Iany attempts were made during the ensuing years to associate preoperative urinary estrogen estimations with response, but none resulted imian accurate method for Se The first suggestion that urinary androgen levels might be of value came from Allen et at. (1), who described a cor relatiomi between prognosis and the ratio of the il-deoxy 17-oxosteroids to the 11-oxy compounds. This work was criticized on methodologic grounds, and reports from other workers (16, 25) did not indicate that the fractionated 17-oxosteroids were of value in prognosis. Subsequently we reexamined the use of various hormone estimations and measured the urinary estrogens, gonadotropin, pregnamie diol, 17-hydroxycorticosteroids (17-OHCS), and the total and fractionated 11-deoxy-17-oxosteroids in a series of patients with advanced breast cancer awaiting hypophy sectomy and adrenalectonny (15). From these estimatiomis it was shown that the levels of etiocholanolone tended to be high in successful cases amid low in failures, whereas the nieasuremenits of the 17-OHCS tended to be low in success ful cases and high mmfailures. Subsequently (12), by means of a discriminant function [80 †" 80 (mgl7-OHCS/ 24 hr) + @g etiocholamiolone/24 hr], we were able to cor relate these measurements with the outcome of the opera tions in 59 patients. Later Atkins (3) reported on a comitinuationi of this work amid described an association be tween the urinary determinations and the survival amid response of 69 patients. The association, however, between the discriminanit function amid the subsequent response of the patient to adrenalectomy or hypophysectorny was found in a retro spective series. If these results were truly representative of a larger population, the discriminant should be of prac tical use not merely for prediction but for selection of pa tients for adrenalectomy amid hypophysectomy. It was, therefore, decided to test the discriminanit in a clinical trial. In this trial patients with a positive dis criminant, who were expected …