It is now well known that an increase or a decrease in the amount of one hormone or another or a change in their balance affects the incidence and rate of growth of certain tumours. Castration, for example, is used therapeutically to delay the progress of breast cancer in the female and prostatic cancer in the male. The interrelationship between the actions of steroid hormones and cancers are of especial interest and the search for connections between the adrenal cortex and cancer led to two dramatic developments: first, the discovery by Huggins (1954) that adrenalectomy will induce a worthwhile remission in approximately 40 per cent of selected female patients with breast cancer; and second, the discovery that the administration of glucocorticoids in high doses will, at least temporarily, check the course of lymphomatous tumours in many patients.

It is thought that some cancers of the human breast are dependent upon a secretory product of the adrenal cortices, possibly an oestrogenic compound. It is known that oophorectomy is followed by temporary regression of some breast cancers and that the administration of oestrogen affects the growth of breast cancers in both laboratory animals and man. There is presumptive evidence to suggest that the adrenal cortices secrete some oestrogens, but it is not yet proved that removal of adrenal oestrogens is the factor which causes tumour remission.

It is possible, but not probable, that the adrenal corticosteroids have a permissive relationship to the growth of breast cancer. Optimal doses of cortisone, or related glucocorticoids, plus 11-deoxycorticosterone do not prevent the remission that may follow adrenalectomy. Perhaps some other steroid that is eliminated by adrenalectomy has a permissive action in tumour growth, or it may be that one or more adrenal steroids-possibly an abnormal secretory product-actively stimulates tumour growth. Sooner or later for unknown reasons the cancer escapes from the favourable effects of adrenalectomy, grows again, and brings the patient to death.

No evidence is known to us that stressful situations operating via the adrenal cortices cause cancer in either laboratory animals or patients. In our own studies of populations of aging rats which develop spontaneous tumours, we have looked in vain for an effect of non-specific stressors upon the incidence of tumours.

General Discussion

A certain mystique seems to have developed around the phrase 'diseases of adaptation' which has led to its wide acceptance but has deflected critical attempts to define its precise meaning. That it has some general meaning we cannot deny, for it would be difficult to define disease without implying some failure of adaptive mechanisms, or to define stress in a way which would exclude any cause of disease as being a stressor. When, however, disease is experimentally reproduced from a complex pattern of causes it is necessary to select those factors which we believe to be of primary importance.

Selye (1956) is the principal advocate of the theory that an alteration in adrenal cortical function as a result of exposure to stress is a primary cause of many human diseases. The hormones of the adrenal cortex are necessary for life itself and indeed for the normal operation of many physiological processes, especially when these are pushed to the limit of their capacity to respond to severe stressors. The presence or absence of the adrenal cortical hormones also affect the processes and signs of disease. The pathological effects of hormonal overdosage are now established, but is should be remembered that most substances are toxic when given in far greater than physiological amounts. In our view too little attention has been directed to the basic question whether exposure to non-specific stressors per se causes disease in laboratory animals or man under natural conditions, since the search for new facts has been more rewarding when done under highly abnormal conditions.

The hypertensive cardiovascular disease which develops in animals stressed by exposure to cold is related, according to our evidence, principally to the effects of overload with the sodium chloride diet rather than to any increase in adrenal cortical function associated with exposure to the stressor as reported by Selye in 1943. Our view is substantiated by the finding that small constant doses of adrenal cortex extract permit the development of hypertension and its pathological lesions in the absence of the adrenal glands.

The relationship between adrenal cortical hormones and the hypertensive disease which develops in salt-loaded animals not exposed to stress and in animals overdosed with sodium-retaining steroids may also depend mainly on the difficulty which the animal has in handling sodium and other electrolytes. The toxicity of sodium is affected by the ratio of sodium intake to that of potassium. An accompanying rise in potassium intake partly ameliorates the pathological effects of high salt loads (Meneely el al., 1956). Adrenal cortical insufficiency also ameliorates the pathological effects of salt loading possibly because the basal secretion of corticosteroids by the adrenal-amounts necessary for life and vigour-has some obligatory sodium-retaining action. The adrenal-insufficient animal, if it is not in shock, may excrete more sodium per unit volume of urine than the animal with intact adrenal glands. Although this idea may be too simple, we believe that it deserves further study.

It seems possible that the toxicity of sodium could be based upon its distribution both inside and outside the cells, together with the accompanying imbalance of other electrolytes. Tobian and Redleaf (1957, 1958) have demonstrated an increased sodium, potassium and water content of the arterial wall in experimental hypertension induced by either renal or endocrine mechanisms. The movement of sodium and water into cells and of potassium into the extracellular phase during acute pressor episodes has been reported (Friedman et al., 1957; Friedman, Nakashima et al., 1958; Friedman, Scherrer et al., 1958) and the close temporal association between these cation shifts and blood pressure changes suggest a possible causal relationship at cellular level in vascular smooth muscle. Some or all of the corticosteroids may affect these peripheral cation changes. Many physiological processes are influenced by their ionic environment and it could be that cation changes affect energy-yielding processes or the actual contractile mechanisms in vascular smooth muscle so leading to an alteration in peripheral vascular resistance and blood pressure level.

Adrenalectomy also ameliorates the experimental hypertension caused by various forms of extensive renal injury which occur in the absence of increased salt-loading. Here the primary cause of hypertensive vascular disease appears to be related to the kidney and yet adrenal hormones must be present to support the overt signs of the disease. The words 'permissive cause' seem preferable to 'active cause' in this situation, although the word 'permissive' does not reveal the exact mechanism by which the corticosteroids support hypertension. We have some insight into the general mechanisms whereby the glucocorticoids support diabetes even when it is due primarily to insulin lack. The genesis of glucose from non-carbohydrate sources is reduced during adrenal cortical insufficiency and adrenalectomy enables the diabetic animal to utilize glucose once more.