How To Learn from Patients: Fuller Albright's Exploration of Adrenal Function

The Cushing Syndrome

Albright's strongly held concept of parathyroid action was based on his investigation of a single patient with idiopathic hypoparathyroidism [3]. In contrast, Albright cast a wide net in initiating his studies of adrenocortical abnormalities. He and his associates “measured something” in a spectrum of patients with endocrine disorders and then focused on those in whom there was an unexpected aberrant result. Their findings were reported in three papers, all published in 1941. In the first of these [4], a survey of changes in carbohydrate metabolism in several endocrine diseases, Albright and his colleagues showed that the Cushing syndrome was characterized by impaired glucose tolerance and decreased insulin sensitivity, the latter measured by the glucose-insulin tolerance test.

In the second paper [5], the Boston group surveyed the changes in the 24-hour urinary excretion of neutral 17-ketosteroids, which was thought to be a rough measure of androgen secretion, in a similarly wide variety of endocrine abnormalities. One hundred thirty-five patients were tested and more that 1500 assays were done. By simply examining mean 17-ketosteroid values in pertinent disease states, Albright adduced convincing evidence that urinary 17-ketosteroids were the end-products of secretions from not only the testes but also the adrenal glands. He reasoned that because normal 17-ketosteroid excretion averaged 14 mg in men and only 9 mg in women, it was likely that 9 mg was secreted from the adrenal glands and 5 mg from the testes. He predicted correctly that among patients with Addison disease, women would excrete essentially no 17-ketosteroids and men would excrete only about 5 mg per 24 hours. Similarly, he found that patients with panhypopituitarism and loss of both adrenal and gonadal function showed negligible urinary 17-ketosteroid excretion. He concluded, again correctly, that the contribution of the ovaries to 17-ketosteroid excretion must be small because he found no change from normal urinary excretion values in patients who were surgically castrated or had undergone natural menopause.

Albright also clearly asserted that all forms of the Cushing syndrome were caused by hyperfunction of the adrenal cortex. Although this assertion is for us an endocrinologic truism, in 1941 it was a bold effort to dispel a decade of confusion. Cushing, in his ground-breaking report of 1932 [6], tentatively concluded that the syndrome was due to “pituitary basophilism” because 2 of his 12 patients had shown only pituitary basophilic adenomas at autopsy. Whether the adrenal cortex was also involved was uncertain, although 1 patient had an adrenal adenoma and another had adrenal hyperplasia. Cushing's view was challenged 3 years later by Oppenheimer and colleagues [7], who concluded that “adrenal changes are practically essential for the development of the clinical features of basophilism.” Others did not clarify matters when they suggested that the adrenal overactivity might stimulate the growth of pituitary basophils, and the waters were muddied further by the conclusion of Heinbecker [8] that “either an adrenal tumor or hypofunction of paraventricular hypothalamic nuclei may be primary causes of the basophil degeneration which in turn is the immediate cause … of Cushing's syndrome.”

Still another confounding factor was the variation in histologic findings in the adrenal glands. These findings included tumors, both adenomas and carcinomas, bilateral hyperplasia, and, most puzzling, instances in which the adrenal glands were found to be histologically normal. Albright maintained that “normal” or not, such adrenal glands were hyperfunctioning. (At the time of these discussions it was not known that the Cushing syndrome could also result from ectopic adrenocorticotropic hormone [ACTH] production, although Albright must be credited with proposing the concept of tumoral ectopic hormone production in his description of a patient with hypercalcemia accompanying a hypernephroma [9].) Although contemporary investigators [10-12] agreed that some cases of the Cushing syndrome were attributable to adrenal hyperfunction it was Albright who cut the Gordian knot by insisting that all cases were caused by adrenal overactivity regardless of whether histologic abnormalities of the adrenal glands were shown or whether a pituitary tumor was present.

Several years later [13], Albright happily accepted the suggestion of Kessel that “the term ‘Cushing's disease’ is reserved for those cases in which a basophil tumor of the pituitary is present, whereas the term ‘Cushing's syndrome’ is used for the syndrome as a whole.”

Having cleared the etiologic air, at least to his own satisfaction, Albright turned his attention to his favorite preoccupation: pathophysiology. The third paper, published in 1941 [14], was devoted to the treatment of the Cushing syndrome with testosterone. Albright pointed out that although the 24-hour urinary excretion of 17-ketosteroids was greatly elevated in patients with adrenocortical carcinoma, it was only slightly elevated in patients with adrenal hyperplasia and, finally, that the measurement of urinary “cortin” in J.S.L. Browne's laboratory showed definite elevations. (It is unusual that Albright, such a sharp-eyed clinician, never recorded the differentiating observation that patients with pituitary adenoma showed typical Addisonian pigmentation of the skin but that patients with adrenocortical tumor did not.)

When considering the pathogenesis of the Cushing syndrome, Albright drew heavily on the seminal studies of Long [15], who showed with irrefutable clarity that adrenocortical hormone enhanced protein catabolism through a process of gluconeogenesis from protein, with a resulting impairment of carbohydrate metabolism. Albright asked the rhetorical question, “Are not many features of the Cushing syndrome the result of protein shortage?” He responded with a resounding “Yes,” singling out muscular weakness, thinning of the skin, easy bruisability and osteoporosis.

If the Cushing syndrome were indeed a state of “hypergluconeogenesis,” Albright reasoned that patients would improve when treated with an anabolic hormone. Again, his conjecture was found to be correct; long-term testosterone therapy was clearly efficacious: “All the patients gained in weight and strength, their abdomens became less protuberant, the skin noticeably thicker and, with the disappearance of the reddish hue, they bruise less easily.” The salutary effect of testosterone in the treatment of the Cushing syndrome was confirmed in the laboratory of J.S.L. Browne [13], but this form of therapy never gained much acceptance; surgical approaches remain predominant.

Albright reviewed his findings on the Cushing syndrome in a 1943 Harvey lecture [13]. He credited Long with “the importance of adreno-cortical hormone. in converting proteins and hence tissues into sugar (gluconeogenesis).” Albright suggested the term “S” hormone (today's cortisol) to designate this sugar hormone. He pointed out that osteoporosis was the result of protein depletion rather than a disease of calcium metabolism and that an excess of “S” hormone would therefore lead to a decrease in bone density by depleting the proteinaceous bone matrix. He emphasized the fact that testosterone was more than just a stimulator of sexual organ growth; it promoted a generalized somatic anabolic effect. He was struck by Browne's observations of a patient with an interstitial cell testicular cancer. This individual had “a 17 ketosteroid excretion of over 1000 mgs. per 24 hours and … (failed) … right up to his death, to lose weight or to become debilitated in spite of extensive metastases.” Finally, Albright showed, using the elegant but arduous metabolic balance studies for which he was famous, that treatment with testosterone resulted in both the production of a strikingly positive nitrogen balance in these patients with the Cushing syndrome and the previously mentioned considerable improvement in their clinical state.

He went on to speculate about the mode of action of the “S” hormone. He rejected the obvious possibility that the “S” hormone provoked a direct protein catabolic effect because baseline balance studies on patients with the Cushing syndrome showed negligible protein depletion. Instead, he offered the hypothesis that the action was anti-anabolic. Albright recognized how tenuous his new idea might be. “The author is aware of no data which do not harmonize with this new theory. However, he feels that such data will be forthcoming and that a new hypothesis or a further modification of the present one will be necessary.” Today, more than 50 years later, this question is still being scrutinized; Louard and others [16] have reported recently that “glucocorticoids antagonize insulin's antiproteolytic action on skeleton muscle in humans.”

Congenital Adrenal Hyperplasia

Called the adrenogenital syndrome by Albright, congenital adrenal hyperplasia presents often as pseudohermaphroditism in the female and as macrosomia praecox in the male as detailed in the classic descriptions by Young [17]. These striking disorders became of particular interest to Albright when he found that, in afflicted patients, urinary excretion of 17-ketosteroids was unequivocally elevated. Because the syndrome was associated with either adrenal tumor or adrenocortical hyperplasia, it seemed clear to Albright that an adrenocortical hormone, quite distinct from the catabolic “S” hormone, must be secreted by the adrenal cortex as well. He concluded then that “it would appear, therefore, that the adrenal cortex is capable of producing at least, under certain conditions, 2 different hormones: one which inhibits growth we have already termed the ‘sugar hormone’ or ‘S’ hormone; the other which stimulates growth we might call the ‘nitrogen hormone’ or the ‘N’ hormone.” This was a notion that initially received little credence by his peers, including the respected Dr. Gregory Pincus [18].

To bolster this assertion, it was necessary to argue that 17-ketosteroids, previously shown to be in part excretion products of the adrenal cortex, were more specifically end-products of adrenal androgen secretion. Set forth in a later review [18], these arguments are worth summarizing here as examples of elegant deductive reasoning by an astute clinician: 1) the adrenogenital syndrome itself with “virilism, somatic precosity, and a high excretion of 17-ketosteroids. suggests that the adrenal cortex produces a testosterone-like steroid.” 2) Castration in either sex has only a small effect on urinary 17-ketosteroid excretion, indicating that the gonads are not a major source of these steroid metabolites. 3) After puberty, axillary and pubic hair grows in either sex despite gonadectomy, suggesting that the adrenals, rather than the gonads, are responsible for this pubescent change. 4) Addison disease in females is accompanied by very low 17-ketosteroid excretion. 5) Addison disease in males results in only a modest decrease in urinary 17-ketosteroid excretion. Thus, because both axillary and pubic hair are sparse or absent in both male and female adrenal-deficient patients, their presence would reflect the action of adrenal, rather than gonadal, hormones. 6) Urinary 17-ketosteroid excretion in both sexes does not appear until puberty. “In the female this observation is consistent with the thesis that the adrenal cortex starts producing a second hormone at puberty.” [This observation led Albright to coin the term “adrenarche.”] 7) Because pubic and axillary hair appear at puberty along with urinary 17-ketosteroids, it is reasonable to suppose that this association is causative. 8) Other conditions in the female that lead to the absence of axillary and pubic hair [panhypopituitarism, myxedema and old age] are characterized by low or absent 17-ketosteroid excretion. 9) In a patient with panhypopituitarism, the rubbing of an ointment containing testosterone, “a steroid closely akin … to the ‘N’ hormone,” into the skin of one axilla caused axillary hair to grow in that axilla and not in the other. This maneuver is a fine example of Albright's ingenuity in constructing experiments at the bedside.

Having assembled these closely reasoned arguments that an adrenocortical androgen is a normal constituent of the human economy, Albright could lay to rest the widely held view that the adrenogenital syndrome occurred only because of the persistence of the fetal androgenic zone of the adrenal cortex. His stance was strongly supported by the definitive histologic studies of Blackman [19].

Next, Albright and his associates [20] examined the effect of methyltestosterone administration on the urinary excretion of 17-ketosteroids in various hormonal disorders. In selecting methyltestosterone administration for study, Albright had in mind something more than the mere evaluation of the effect of yet another androgen. When Albright learned that methyltestosterone is not excreted as a 17-ketosteroid, it put him in a position to “study the effect of a testosterone compound on the endogenous production of urinary 17 KS.” He simplified his experiment further by confining his studies to women, and to one man who had a congenital absence of testicular tissue—subjects in whom 17-ketosteroids could originate only in the adrenal cortex. He found that two normal women, two women with the Cushing syndrome and the agonadal man responded to methyltestosterone administration with a reduction in 17-ketosteroid excretion. He also found a “suggestive but not conclusive” decrease in urinary 17-ketosteroid excretion in two male patients with Addison disease.

Albright assembled these facts to construct a hypothesis that led to an astonishingly prophetic prediction. He reasoned that because the production of both adrenal and gonadal androgens were inhibited by a synthetic androgen, “it is suggested that the mechanism of inhibition is the same for both glands.” Although he did not rule out the possibility of a direct inhibition by methyltestosterone, he much preferred the idea that the inhibition be “attributed to a decreased production of some pituitary tropic hormone.” Adrenocorticotropic hormone (ACTH) seemed to him an unlikely candidate, because in congenital adrenal hyperplasia with elevated 17-ketosteroid excretion there is no excess production of “S” hormone [13]. He suggested pituitary luteinizing hormone as the most likely candidate, a quite reasonable guess in view of the limited number of tropic hormones known to exist in 1945. We have learned in the fullness of time that the control of the secretion of adrenal androgens is inordinately complex [21]. Apparently, several tropic hormones, including gonadotropins, play a role. Recently, it has been suggested that human joining peptide, a component of pro-opiomelanocorticotropin, may also stimulate androgen secretion [22].

Turning his attention to congenital adrenal hyperplasia, Albright seized on the fact that the syndrome was, indeed, the result of adrenal hyperplasia and, therefore, probably the result of stimulation by that same tropic hormone. Although methyltestosterone administration reduced adrenal androgen secretion, as reflected in the decrease in 17-ketosteroid excretion, it was obvious that methyltestosterone would not serve as a form of treatment because it is itself a potent androgen. However, the implications of the conclusion that adrenal androgen secretion could be inhibited were clear to Albright. He stated [20], “If a steroid could be found which inhibits the endogenous source without itself being an androgen, an important advance in the therapy of these unfortunate patients would probably be at hand.” Thus, through clear deductive reasoning, Albright predicted, 5 years before the event, the successful therapy for this remarkable disease.

Fuller Albright was not alone in his search for a specific form of therapy to reverse the profound changes provoked by congenital adrenal hyperplasia. He was competing with a friendly rival, Dr. Lawson Wilkins, the pioneer pediatric endocrinologist and revered intellectual father of a whole generation of accomplished academicians who in 1971 memorialized their mentor by establishing the Lawson Wilkins Pediatric Endocrine Society.

Wilkins's entry into this area of research began in 1940 [23] when he described in detail a boy with macrogenitosomia praecox who died from adrenal insufficiency. For the first time, it was documented that this disorder was associated with diffuse bilateral adrenal hyperplasia, “analogous to pseudohermaphroditism in the female.” The pathophysiology of this illness remained a puzzle. When increased basophilic pituitary cells were found at autopsy, Wilkins speculated “that pituitary basophilism might be the result rather than the cause of the adrenal hyperplasia.” Wilkins was also the first to report in detail a case of a feminizing adrenocortical tumor in a boy [24]. The boy recovered uneventfully after extirpation of the tumor.

Wilkins meticulously charted the clinical course of many patients with pseudohermaphroditism who, “in an earlier era, would have gone to Hugh Young's clinic” [25]. Although he was caught up in problems of pathophysiology, his approach to patients was directed primarily toward therapeutic goals; he was “never willing to gamble for the sake of new knowledge” [26]. A possible departure from this dictum was his experimental treatment of a pseudohermaphroditic infant with associated adrenal insufficiency [27]. After completing studies that suggested that the adrenal glands were secreting a salt-losing hormone, and with the knowledge that in such infants “prospects … for continued life (were) almost hopeless,” bilateral adrenalectomy was done. Despite concerted hormonal replacement efforts, the child died of pneumonia. It is a testament to Dr. Wilkins's integrity that he proceeded with the publication of this case study after he became aware of the beneficial effects of cortisone in congenital adrenal hyperplasia.

Before his successful trial with cortisone therapy he believed that the best hope for treatment of congenital adrenal hyperplasia was to find a drug that would have an “antagonistic effect upon adrenal androgen” [28] and, indeed, he attempted unsuccessfully to suppress 17-ketosteroid excretion by administering several testosterone and androstenediol derivatives [29].

Inevitably, the paths of Wilkins and Albright converged so that what was “at first a friendly rivalry … between the Johns Hopkins and the Massachusetts General Hospital … became a neck and neck race. It was in the same month, December, 1949, that the competitors tested their first hyperadrenocortical hermaphrodite with the newly available, newly synthesized adrenocortical steroid, cortisone” [26].

Unlike Lawson Wilkins, Albright, not surprisingly, followed a more pathophysiologic route to the selection of cortisone as a form of therapy for congenital adrenal hyperplasia [30, 31]. He administered ACTH to three such patients and found that there was a relatively deficient response in the “S” hormone with an obvious increase in the already high 17-ketosteroid excretion. Albright postulated that there was a limitation in response in glucocorticoid production, along with “an increased production of ACTH resulting in an abnormally high production of adrenal androgen.” It then followed logically “that a carbohydrate active adrenocortical hormone such as cortisone should depress the production of ACTH and thus diminish adrenal androgen production.” The administration of cortisone to these patients revealed a dramatic reduction in 17-ketosteroid excretion with subsequent clinical improvement. Albright concluded that “the primary defect in this syndrome is the decreased production of ‘sugar’ hormone by the adrenal cortex in response to ACTH,” leading to “a lack of sugar hormone, the failure of normal inhibition of ACTH, an increased output of ACTH and a resulting increase in adrenal androgen.” This hypothesis accounted for the adrenal insufficiency encountered in some patients with this syndrome and also for adequate adrenal function found in others because it “does not require that patients show an absolute deficit of ‘sugar’ hormone since the increased production of ACTH may completely compensate for the inherent insensitivity to ACTH” [31].

This formulation, as true today as it was in 1950, stands as a stellar example of the power of clinical investigation when applied with rigorous logic and knowledge-based intuition. Current therapy for congenital adrenal hyperplasia remains essentially unchanged—although treatment regimens are now individualized—and, remarkably, even prenatal diagnosis and treatment have become available [32].

In pursuing these studies, Fuller Albright simply determined urinary 17-ketosteroid excretion and administered methyltestosterone in an array of patients with endocrine abnormalities. These unsophisticated measures, coupled with his clinical acumen, sufficed to ensure his success in elucidating the pathophysiology of both the Cushing syndrome and congenital adrenal hyperplasia. Furthermore, in the case of congenital adrenal hyperplasia, he used his new-found knowledge to devise a rational and effective treatment.

Admittedly, it was easier in those days to learn from patients, and Fuller Albright functioned in a setting that optimized his opportunities [2]; nonetheless, he had the acumen to take full advantage of his good fortune and thus has provided for us a standard that few could equal and none could surpass.