Prostatic cancer is a common disease of the elderly human male; its natural course has been well established. In the study of Bumpus (3) there were 485 patients who received no form of treatment. The average duration of the disease from first symptoms to death was about 81 months; four patients lived more than 3 years, and two of them more than 10 years. When metastases had occurred at the time of examination, two-thirds of the patients were dead within 9 months. In the study of Nesbit and Plumb (40) there were 260 patients with metastasis at the time of diagnosis. The average survival following diagnosis in this untreated group was 16.9 months, while the median survival time was 9.6 months; the extremes were 1 month and 176 months.

The endocrine treatment of prostatic carcinoma consists of anti-androgenic measures. This idea arose from physiologic studies of the prostate.

The function of the prostatic glands is external secretion, and a method (23) was devised to obtain and to measure the output of the prostatic glands in dogs over a period of years. It was found that the secretion of the prostate of healthy dogs often remains relatively constant, varying only slightly in its amount over many months. This constancy indicated that during these periods the pituitary-testis-target complex remains in a steady chemical state. Orchiectomy abolishes the prostatic secretion; the prostate undergoes atrophy.

The prostatic secretion of dogs is eliminated also by the administration of phenolic estrogens (18); here, too, the prostate shrinks in size. The suppression of prostatic function is due to manifold actions of phenolic estrogens. Among these, estrogens cause metaplasia in the prostatic epithelial cells. Furthermore, the production of go-nadotrophin is depressed. The simultaneous administration of gonadotrophin (13) and phenolic estrogens preserves the prostatic secretion, although the output is reduced in amount. In dogs in which the prostatic output has been abolished by orchiectomy or by the injection of estrogenic substances, the administration of testosterone or related compounds restores the secretion to normal, both in quantity and in quality.

The next observation in the development of therapy for human cancer was the effect of anti-androgenic measures on cystic hyperplasia of the canine prostate. In the dog, as in man, neoplasms of the prostate develop in senility with great frequency. These tumors were found (18) to shrink profoundly after orchiectomy or the administration of estrogenic substances, whereas testosterone restored their size and accelerated their growth.

These findings in the dog proved to be pertinent to the human. Translated to man with prostatic carcinoma, they turned out to be directly applicable en bloc and permitted the introduction of anti-androgenic therapy in clinical practice. The treatment was not empirical; it had an experimental basis. Both orchiectomy and phenolic estrogens (21, 27) were effective in controlling cancer of the prostate in certain cases, while, conversely, the administration of testosterone intensified the growth of the neoplasm. Historically, the synthetic compound diethylstilbestrol was the first agent of known chemical constitution, aside from radium, to ameliorate carcinomatosis. These observations happened to be the beginning of chemotherapy of cancer.

The method of proof of a proposition can be of greater interest than that which is proved. The effects of the administration or withdrawal of steroids on human prostatic cancer were first proved by a systematic study of enzymes in the serum of patients with advanced cancer of the prostate. The proof concerned acid and alkaline phosphatases. Cancer of the prostate metastasizes to bone marrow and lymph nodes inter alia; in the bone it commonly causes osteoblastic metastases. Kay (31) discovered that the alkaline phosphatase content of serum was increased in the presence of increased osteoblastic activity accompanying some metastatic neoplasms; this explains the increase of alkaline phosphatase in prostatic cancer which has invaded the skeleton. Kutscher and Wolbergs (82) discovered that acid phosphatase was present in human prostatic epithelium in large amounts; that finding was confirmed and extended to include prostatic cancer by Gutman, Sproul, and Gutman (8). Moreover, in the serum of patients with cancer of the prostate Gutman and Gutman (7) found that acid phosphatase was increased, but only when the neoplasm had metastasized. It was evident to Huggins and Hodges (21) that the simultaneous estimation of acid and alkaline phosphatases of the serum over many days could provide quantitative information concerning the metabolic activity of the prostatic neoplasm and also, when metastases were present in bone, the reactivity of the host to the presence of the neoplasm in the skeleton. It was convenient and informative to study in a single sample of serum both the activity of cancer and the corresponding response of normal cells to the presence of that cancer. The administration of testosterone or, conversely, the institution of anti-androgenic measures caused characteristic reflections of hormonal activity in the serum enzymes, and these provided simple mathematical proof of the stimulatory or the inhibitory action, respectively, of these modifications of the hormonal milieu interieur.

These findings were confirmed through the study of another enzyme in the serum of patients with disseminated prostatic cancer; this enzyme was fibrinolysin. It had been discovered (24) earlier that human prostatic epithelium secretes large quantities of a proteolytic enzyme which is specially active against fibrin as a substrate. Human prostatic secretion contains several proteolytic enzymes, of which fibrinolysin is the most active; fibrin is digested far more rapidly by prostatic fluid than is fibrinogen or denatured hemoglobin.

Tagnon et al. (49) observed that the blood of some patients with metastasized cancer of the prostate becomes incoagulable because of its concentration of prostatic fibrinolysin; fibrinolysis, as defined by these workers, is the complete dissolution of the blood clot within 24 hours at 87° C. The content of fibrinolysin in serum is eliminated (48) by the administration of estrogenic substances or by orchiectomy; testosterone causes fibrinolysin to reappear in such patients. The entry of prostatic fibrinolysin into the blood is similar to that of acid phosphatase. The anti-androgenic measures restore the coagulability of the blood.

The first series of patients with prostatic cancer treated by orchiectomy (27) comprised twenty patients with wide-spread metastases; only four of them survived more than 12 years. Despite regressions of great magnitude, it is obvious that there were many failures of endocrine therapy to control the disease.

Nesbit and Baum (39) reviewed the effects of orchiectomy and estrogenic substances on 1,818 cases of prostatic cancer. They state: "The present survey has demonstrated conclusively that patients with prostatic cancer who respond favorably to castration and/or estrogen therapy live more comfortably and longer than patients not treated by these methods."

It was postulated by Huggins and Scott (26) that, in some of the "failure" cases, the adrenal glands were the source of growth-promoting steroids in sufficient amounts to maintain activity of prostatic cancer in the absence of the testes. Bilateral adrenalectomy was carried out, and one patient survived for 4 months. With the advent of cortisone in large amounts the problem was reinvestigated (16, 17). There was evidence of regression of prostatic cancer in four patients after adrenalectomy, and a complete remission has persisted for 5 years in one patient so treated. Remissions following adrenalectomy have been reported by others (52); it is now known that adrenalectomy benefits only a small percentage of patients with prostatic cancer in relapse after orchiectomy and the administration of estrogens.