Through the efforts of many workers, it has been established that steroids and related polycyclic hydrocarbons have a pertinent relationship to cancer. Under special conditions aromatic hydrocarbons can incite cancer, can prevent it, and can cause a profound regression of, or actually kill, cancer cells-in the animals and in man. Clearly, hormones occupy a position near the center of the cancer problem.

In certain patients it is possible to induce a profound regression (Figs. 1-4) of widely disseminated cancer even in the latest stages of the disease by modification (33)1 of the hormonal status of the patient. These wither-ings of extensive cancer in multiple loci are accompanied pari passu by clinical improvement which has lasted for more than a decade in favorable cases, the neoplasm having been destroyed-in whole or in part. Worthwhile relief has been rendered thereby to many people. It is now certain that hormones are of high significance in the maintenance of four types of neoplasms. The cancer problem is far from solved; many patients in these select categories do not respond to attempts at hormonal control, and in others the relief is of short duration. It would appear that the control of cancer by polynuclear hydrocarbons is in its incipience, but, significantly, a crack has developed in the facade of cancer. From a priori reasoning it is not surprising that, limited as its usefulness often is in the control of the great growth process called cancer, physiology has been more successful than pharmacology in its control.

The control of cancers by hormonal means rests on two principles. (1) Cancer is not necessarily autonomous and intrinsically self-perpetuating. Some neoplasms, perhaps all, retain biochemical characteristics of the normal cells from which they arose. When the original cells are dependent upon hormonal subvention for metabolic activity at a high rate, its cancers can be similarly dependent, and these cancer cells atrophy when hormonal support is withdrawn. Be the causes of the cancers what they may, removal of supporting hormones causes the death or atrophy of hormone-dependent cancer cells. Therefore in cancers of this class, the hormones are a component of critical importance. (2) Cancer can be sustained and propagated by hormonal function that is not abnormal in kind or exaggerated in rate.

1 The numbered references relate to publications from the Ben May Laboratory, 1951-61 (pp. 131-44).

These concepts arose during a study of prostatic tumors in living creatures-'first in the dog, subsequendy in man. We found inter alia that the administration of diethylstilbestrol, a synthetic substance which does not occur in nature, caused regression of widely disseminated prostatic cancer in man (Fig. 5). This observation happened to be the beginning of chemotherapy of cancer.