BULLETIN OF THE NEW YORK ACADEMY OF MEDICINE

APRIL 1977

VOL. 53, No. 3

THYROID-HORMONE EFFECTS ON STEROID-HORMONE METABOLISM* GARY G. GORDON, M. D., F. A. C. P. Professor of Medicine

A. Louis SOUTHREN, M.D., F.A.C.P. Professor of Medicine Chief of Endocrinology New York Medical College New York, N.Y.

THE level of thyroid function significantly alters the metabolism of steroid and other hormones in man, including their rates of secretion or production, plasma concentrations and plasma protein binding, flux through various compartments or "pools," and rate and pathway of catabolism in the liver or other metabolizing tissues. These changes in steroid-hormone metabolism are attributed to the action of thyroid hormones on controlling various rate-limiting enzymes and to the concentration of critical cofactors in the metabolizing tissues. However, direct *Presented as part of a Symposium on the Thyroid held by the Section on Medicine of the New York Academy of Medicine November 20, 1975. This research was supported in part by Public Health Service Research Grant No. AM12845 from the National Institute of Arthritis, Metabolism, and Digestive Diseases, Bethesda, Md. Address for reprint requests: New York Medical College, 1249 Fifth Avenue, New York, N.Y. 10029.

242

G.G. GORDON AND A.L. SOUTHREN

effects of thyroid hormones on the hypothalmic-pituitary-target gland axes also must be considered in evaluating changes in steroid metabolism. The biological significance of these alterations in steroid dynamics (i.e., the flux of hormone through the blood) depends on the ultimate concentration of active steroid hormone delivered to the receptor site of target tissues. Clinical evidence suggests that in disorders of thyroid function under certain circumstances the concentration of hormones in the target tissues may be altered. The application of isotopic tracer studies and radioimmunoassay methods in conjunction with classic measurements of urinary hormonal metabolites and certain critical animal experiments has permitted at least a partial understanding of the multiplicity of effects of altered thyroid function on the metabolism of such groups of hormones as the glucocorticoids, mineralocorticoids, and sex steroid hormones. This information, of course, is of theoretical interest, not only for an understanding of hormonal homeostasis and in ultimately relating hormone secretion to the control of metabolic processes and perhaps hormonal action, but also for its clinical significance in the management of steroid-hormone therapy and in the diagnosis of diseases of steroid-hormone secretion in the presence of altered thyroid-hormone function. The study of steroid-hormone metabolism in states of thyroid dysfunction also has been a useful tool in furthering our understanding of the mechanisms of control of hormone metabolism and has provided insights into some aspects of the pathophysiology of thyroidal disease.

C21 STEROIDS Cortisol. An increase in thyroid function to hyperthyroid levels results in a marked increase in the rate of secretion and catabolism of cortisol. The latter is evidenced by the shortened half-life and increased pool "turnover" rate of the steroid hormone' (Table I). The metabolic clearance rate (MCR) of cortisol from plasma, using Peterson's data,1 can be calculated and is elevated significantly (Table I). The increased rate of secretion of cortisol is a result of the greatly increased plasma clearance of the hormone (i.e., MCR) rather than the converse. This is a specific effect of the increased thyroid function since in instances where cortisol secretion is increased, such as during "stress," ACTH stimulation, or euthyroidal hypermetabolism, a decrease rather than an increase in clearance rate parameters is noted.2 Further, changes in hepatic blood flow cannot exBull. N.Y. Acad. Med.

THYROID EFFECTS ON STEROID METABOLISM

243

TABLE I. THE EFFECT OF HYPERTHYROIDISM AND HYPOTHYROIDISM ON VARIOUS PARAMETERS OF CORTISOL METABOLISM. THE VALUES GIVEN REPRESENT THE MEAN OF PUBLISHED DATA. Cortisol Hyperthyroidism Hypothyroidism

Secretion rate (mg./day) Miscible pool (mg.) Turnover rate (k) (pools/hr.) Half-life (T½/2) (min.) Metabolic clearance rate (liters/day) Plasma level (Ag./dl.) Protein binding Diurnal variation N 17-OHCS (mg./day) 17-KS (mg./day) Fractionated urinary metabolites THE (%) THF (%)

T(70)

J(8)

N (1.3)

N (1.6)

T(2.6)

J(0.2)

4(25) T(700*)

T(172)

=

N (17) N (?) Altered

1 (2.9)

20 1.7 0.5 84 140* 14

k(5.6) T(35)

1 (3.6)

4-14 6-25

J(17)

25

0(l0)

T(33)

16

T =increase, N=normal, l =decrease, * = calculated, 17-OHCS

17-ketosteroids, THE

J(47*)

N (10) N (?) Altered or T (16.6)

Normal

17-hydroxycorticoids, 17-KS

tetrahydrocortisone, THF = tetrahydrocortisol

plain the increased MCR, since the hepatic blood flow does not increase significantly in thyrotoxicosis.3 The increased MCR mostly is offset by an increased secretion rate of the hormone, resulting in normal plasma levels and miscible pool size1 (Table I). The converse is seen in hypothyroidism, where the metabolic clearance (calculated) and secretion rate of cortisol decreases and its half-life is prolonged1 (Table I). Again the miscible pool and plasma levels of the hormone are maintained within the normal range, since the decreased plasma clearance of the steroid is offset by an almost proportional fall in its rate of synthesis.1 It must be appreciated, however, that only single samples of cortisol have been obtained in most studies of cortisol metabolism in states of thyroid dysfunction. In light of recent findings that the secretion of cortisol and other hormones is episodic, further delineation of plasma cortisol levels in these disorders by means of frequent sampling is important. Some information in this area has been obtained by Martin et al.4 and Gallagher et al.5 Martin and his associates4 studied patients at four-hour intervals and showed that alterations in thyroid function affect the diurnal variation of plasma free and conjugated 17-hydroxycorticoids (17-OHCS) and urinary 17-OHCS. Hyperthyroidism was associated with an accelerated disVol. 53, No. 3, April 1977

244

G.G. GORDON AND A.L. SOUTHREN

appearance and lowered mean level of free 17-OHCS from the plasma and resulted in an exaggerated circadian variation and elevated mean plasma levels of the conjugated 17-OHCS. The converse was seen in myxedema. The urinary 17-OHCS showed a markedly exaggerated diurnal pattern in hyperthyroidism and a flattened curve in hypothyroidism. Gallagher et al.,5 using a 20-minute sampling frequency for a 24-hour period, reported on cortisol levels in four hyperthyroid females. The following was found: there was an increase in the number of cortisol secretory episodes per day, the duration of adrenal secretory episodes was prolonged, an increased amount of cortisol was produced, the half-life of the hormone was shortened markedly, the adrenocortical functional capacity was nearly doubled (i.e., mg./unit time increased almost twofold), and there was retention of the normal postmidnight concentration of plasma cortisol at or near zero. Gallagher et al.5 estimated the MCR of cortisol and found it to be increased; the magnitude of the increase is in close agreement with the clearance calculated from Peterson's data.' In the two studies reported to date there was no evidence of an alteration in the plasma protein binding of cortisol in either hyperthyroidism or hypothyroidism.6'7 However, in the study by Biesel et al.6 a low specific activity C14 cortisol tracer was used; increases in inert plasma cortisol concentration were found to approximate 200 ,-g.%, far beyond the physiologic range. The high concentration of inert cortisol tends to saturate its specific plasma-binding sites and obscure changes in binding parameters. The study by Doe et al.7 measured the amount of plasma transcortin in a few patients; again no change from normal was found in either hyperthyroidism or hypothyroidism. However, in view of the observations by Gala and Westphal,8"9 in the rat, of a controlling influence of thyroid activity on the cortisol-binding globulin in plasma, further studies by modern methods and in adequate numbers of patients with thyroid dysfunction are necessary to clarify this point. The urinary metabolites of cortisol, the 17-OHCS, generally reflect the increased rate of cortisol synthesis and are normal to elevated in hyperthyroidism; occasionally they reach the ranges seen in Cushing's syndrome.1 The urinary neutral 17-ketosteroids (17-KS) usually are decreased significantly in both hyperthyroidism and hypothyroidism.1 This probably reflects a nonspecific response seen in many disease states.10 Hellman et al. , using isotopic tracer methods, studied the major urinary metabolites of cortisol in states of altered thyroid function. A marked increase in the

Bull. N.Y. Acad. Md.

THYROID EFFECTS ON STEROID METABOLISM

245

relative and absolute excretions of the 1 -ketonic A-ring reduced metabolite, tetrahydrocortisone (THE), was noted in thyrotoxicosis. The percentage excretion of the 1 1-hydroxy A-ring reduced metabolite, tetrahydrocortisol (THF), decreased, although its absolute amount tended to increase somewhat. Similarly, the 5ca metabolite, allo-THF, increased in absolute amounts, although its percentage excretion did not change. Further, there was an increase in the percentage of tracer excreted in the 1 1-keto-20reduced metabolite, cortolone, with a decrease in the excretion of the 1 1-hydroxy 20-reduced metabolite, cortol. The increased excretion of THE in the urine is unique for hyperthyroidism, since in other circumstances associated with an increased secretory rate of cortisol, such as in Cushing's syndrome or after ACTH stimulation, the predominant urinary metabolite is THF rather than THE.11 These data suggest that thyroid hormone controls the rate of conversion of the active hormone, cortisol (an 1 1-hydroxy compound) to its inactive metabolite, cortisone (an 1 1-ketonic steroid). Both cortisol and cortisone then are metabolized further in the liver by A-ring reduction to their respective tetrahydro-metabolites; these are conjugated and excreted in the urine. The increase in urinary cortolone excretion reflects the increased conversion of cortisol to cortisone. While it has been suggested that thyroid hormone increases the 5a reduction of cortisol,12'13 review of the reported data does not appear to support this conclusion. Gold and Crigler12 showed an increase in the 5a/5,( ratio (allo-THE/THF) in one male patient with Cushing' s syndrome treated with triiodothyronine. Beale13 recently studied the urinary metabolites of cortisol by a gas-liquid chromatographic method and showed an increase in the 5a metabolites of cortisol, as did Hellman et al.,11 but without an increase in the relative percentage of these compounds formed. In view of the definitive studies of Hellman et al., it seems unlikely that the thyroid hormone actively regulates 5cr, 5/3 reduction of cortisol; the changes in the metabolite pattern merely reflect the increased amount of cortisol available for metabolism through these pathways. Thus, it seems that the role of thyroid hormones in controlling cortisol metabolism centers on the conversion of the active hormone, cortisol, to its inactive metabolite, cortisone, which occurs predominantly in the liver. This step is under the control of 1 1-hydroxysteroid dehydrogenase (11oxidoreductase) requiring nicotinamide adenine dinucletide phosphate (NADP) as cofactor. Although the pattern of urinary metabolites is Vol. 53, No. 3, April 1977

246

G.G. GORDON AND A.L. SOUTHREN

strongly suggestive, there is no direct evidence that thyroid function alters the level of this enzyme. Indeed, Koerner and Hellman14 demonstrated in the rat that thyroxine decreases the level of activity of this enzyme system. Further studies are needed to clarify these inconsistencies. Whatever the mechanism, increased thyroid activity leads to markedly increased plasma clearance of cortisol, which in turn decreases the negative feedback effect of the steroid on the central nervous system, at least transiently, until a new higher set point for the secretion of adrenocorticotropin (ACTH) is established. The increased secretion of ACTH leads to enhanced secretion of cortisol to compensate for the increased rate of removal of the steroid. Consistent with this hypothesis are the findings of Hilton et al.,15 who used a bioassay technique in hypophysectomized dogs and demonstrated increased levels of ACTH-like activity in the blood of thyrotoxic patients. Additional evidence in support of increased blood ACTH activity in thyrotoxicosis is the increased size of the adrenal gland in humans16 and the increased cutaneous pigmentation in some patients with this disorder. 15,17 It is apparent that the increased plasma clearance of cortisol prevents the hormone from exerting an effect proportional to its secretory rate, since it is likely that the active hormone does not reach the target tissue in increased amounts. Thus, there are no overt signs of Cushing's syndrome in this disorder despite the markedly increased secretory rate of cortisol. The reverse occurs in hypothyroidism. In this disorder there is a decreased MCR of cortisol with a presumedly suppressed release of ACTH, leading to a decrease in the cortisol secretory rate until a new dynamic steady state develops with a lower set-point for synthesis of cortisol. In hypothyroidism as in hyperthyroidism there is no evidence of adrenal insufficiency, since the amount of hormone secreted usually is adequate for the patient's needs. The alterations in the metabolism of cortisol associated with changes in thyroid function have clinical significance. It should be appreciated from the prior discussion that the administration of thyroid. hormone to man will lead to increased inactivation of cortisol and in situations in which compensatory increase in cortisol secretory rate cannot occur adrenal insufficiency may supervene. This is the pathophysiologic mechanism for the well-known hazard in using thyroid hormone in patients with untreated Addison's disease. A similar phenomenon can occur in hypopituitary patients with secondary adrenocorticol and thyroid insufficiency who undergo thyroid therapy without adequate replacement of cortisol. In these

Bull. N.Y. Acad. Med.

247

THYROID EFFECTS ON STEROID METABOLISM

TABLE II. THE EFFECT OF HYPERTHYROIDISM AND HYPOTHYROIDISM ON VARIOUS PARAMETERS OF ALDOSTERONE METABOLISM. THE VALUES GIVEN REPRESENT THE MEAN OF PUBLISHED DATA. Aldosterone Hyperthyroidism HFvpoNa ad lib 10 mEq. Na + thyroidism N (108) J(87)

Secretion rate (qg./day)

1(645)

Plasma renin activity (ng./ml./3 hr.) Miscible pool (gg.) Turnover rate (k) (pools/day) Half-life (71/2) (min.) Metabolic clearance rate (liters/day)

Plasma concentration (ngd/dl.) Urinary metabolites

T= increase,

N= normal, 4=

T(33)

T(11.8)

i(2.1) 1(18)

1(2.8)

T(5.8)

T(43)

1( 20')

T(54')

T(1919)

N* Normal gross pattern

Normal 100 936 18 4.5 4.8 24 39

1,336

T(2,230*) j(47)

1(1,000)

N* Normal gross

1,625 96

pattern

decrease, *= calculated

circumstances it is advisable to institute thyroid replacement cautiously with low initial doses in patients pretreated with cortisol. In severely ill or debilitated elderly patients with myxedema it is our practice to give cortisol in replacement doses prior to the administration of thyroid hormone. In addition, it has been suggested that some features of thyroid storm may be the result of increased cortisol metabolism with consequent adrenocortical insufficiency. This has led to the use of glucocorticoids in this disorder.'8 The success of /8-adrenergic blockade in the management of thyroid storm18 suggests that adrenocortical insufficiency may not be of prime importance in the pathogenesis of this disorder. However, the use of cortisol is still advocated in the treatment of thyroid storm. In the management of myxedema coma, relatively large doses of parenteral thyroxine are suggested.'9 However, this would lead to increased metabolic removal of cortisol, although the time course of this effect is unclear, and would necessitate the administration of cortisol to prevent the possible occurrence of adrenocortical insufficiency due to compromised hypothalamic-pituitary-adrenocortical function in these severely ill patients. Aldosterone. The effects of thyroid dysfunction on the metabolism of aldosterone mimic those found with cortisol, although they are of lesser magnitude (Table II). It can be seen in Table II that the half-life of aldosterone is shortened and the pool turnover and MCR are increased in Vol. 53, No. 3, April 1977

248

G.G. GORDON AND A.L. SOUTHREN

thyrotoxicosis.20 The mean secretion rate of aldosterone in hyperthyroid patients given an ad lib sodium diet is usually in the upper normal range but may be elevated, and the calculated plasma level of aldosterone is usually normal. The miscible pool of the hormone is decreased slightly with inconstant changes in the volume of distribution of the steroid.20 These data suggest a modest increase in the metabolism of aldosterone in hyperthyroidism. Hypothyroid subjects showed opposite effects. Luetscher et al.20 also studied the urinary metabolites of aldosterone after the injection of a radioactive tracer dose and demonstrated a normal pattern of urinary radioactivity in the 3-oxo conjugate and tetrahydro metabolite. Milech et al.2" and Hauger-Klevene22 showed elevated plasma renin activity in thyrotoxicosis with a decreased activity in hypothyroidism. These changes are in a sense analogous to those described for ACTH (vide supra). When the renin-angiotension-aldosterone system in thyrotoxic men was studied during a low sodium diet (10 mEq. Na/day) the normal increase in aldosterone secretion did not occur despite a markedly increased plasma renin level23 (Table II). These patients have lower than expected plasma levels of aldosterone and increased estimated MCR of the hormone. Renin and aldosterone responses during ACTH and saline loading were normal; oral potassium loading (200 mEq.) normalized aldosterone secretion and lowered plasma renin activity. Propranolol significantly lowered renin activity but did not reduce aldosterone secretion. The elevated plasma renin activity in this group of patients probably is secondary to increased sympathetic stimulation and potassium depletion. It is likely that potassium depletion contributed to the blunted aldosterone secretion during deprivation of sodium.23 The mechanism in hyperthyroidism which is responsible for the changes in renin activity and in the aldosterone secretory response to sodium deprivation is unknown, but it appears to be independent of the alterations in the peripheral metabolism of the steroid. The possibility that potassium depletion or increased ,3-adrenergic stimulation may play a role in these changes is suggested by the response to administration of potassium and ,B-adrenergic blocking agents. The clinical significance of these alterations is not apparent, however, since despite the subnormal increase in the rate of aldosterone secretion during restriction of sodium, the sodium balance remains normal.

Bull. N.Y. Acad. Med.

THYROID EFFECTS ON STEROID METABOLISM

249

TABLE II1. THE EFFECT OF HYPERTHYROIDISM AND HYPOTHYROIDISM ON VARIOUS PARAMETERS OF TESTOSTERONE METABOLISM. THE VALUES GIVEN REPRESENT THE MEAN OF PUBLISHED DATA. Testosterone Hypothyroidism

Hyperth'roidism Male

Production rate (mg./day) CRTA (%)

CRTE1 (%) CRTE2 (%)

[p]TA (%)

Normal

Female Male Female Male Female

N (5. 10) N (0.34) ( 1.0) W(1.0)

T(0.72) T(0.70) N (0. 16) N (6.9) 1(325)

N (0.30) N (5. 1) T(530) T(0.98) T(95.8)

N (0.33) 6.8 T(4.2) 3.6 0.12 0.27 N (10.6)

Metabolic clearance rate (liters/day) T(883) (1,059) Plasma concentration (,ig.Idl.) T(0. 1 1) N (0.03) 0.66 Protein binding (%) T(97.2) 1(86.6) 1(89.8) 89.7 Urinary 17-KS fractionation (67)* (68) (14) (14) 43 (A/A + EX 1 00) (%) T=increase, N =normal, I=decrease, *= calculated, 17-KS = 17-ketosteroids

0.44 2.1 0.07 0.13

525 0.04 93.2 44

C19 STEROIDS In constrast to the C2, steroids, the effect of thyroid dysfunction on the metabolism of the C,9 steroids (testosterone, androstenedione, and dihydrotestosterone) is to decrease their over-all rate of metabolism. Testosterone. The most striking effect of hyperthyroidism on the metabolism of testosterone (T), in both men and women, is a decline of approximately 50% in the clearance of the steroid (MCRT)24 25 (Table III). Conversely, in hypothyroidism plasma clearance of the hormone is increased.25 These findings are in sharp contrast to those found in cortisol and aldosterone metabolism (see above). The mechanism for the decreased MCRT has been related to the significant increase in the specific highaffinity plasma binding protein of the sex steroid that occurs in hyperthyroidism (Table III).26,29 The increased plasma binding is associated with an increased total plasma concentration of testosterone.252628 However, Chopra and Tulchinsky29 estimated the concentration of unbound or "free" testosterone by a dialysis technique and found that it was not significantly different from that seen in normal men (21.9 compared with a normal of 17.3 ng./100 ml.). These data are somewhat discordant with the widely confirmed effects of hyperthyroidism on the MCRT. The nearly normal concentration of unbound hormone found by Chopra et al.29 ap-

Vol. 53, No. 3, April 1977

G.G. GORDON AND A.L. SOUTHREN

250

o DERIVED FROM ANDROSTENEDIONE DERIVED FROM TESTOSTERONE

MALES 4

20

2.30

or

NORMAL

~ 2~ ~

*4I

E0.24~~~LG

FEMALES

1.0oo o°0.02

03

'O.053

25 13-54

HYPERTHYROI DISM

20 is

I0

~~~~~~~~~~~7.51

8.57

ESTRONE (n9%)

3.23

Fl

3.52

I

ESTRADIOL (ng %)

ESTRONE (ng %)

1.50

ESTRADIOL (ng %)

Fig. 1. The contribution (CRPre-Pro.PCPre) of plasma testosterone and androstenedione to estrone and estradiol in normal individuals and in those with hyperthyroidism. Reproduced by permission from Southren, A. L., Olivo, J., Gordon, G. G., Vittek, J., Brener, J., and Rafii. F.: The conversion of androgens to estrogens in hyperthyroidism. J. Clin. Endocrinol. Metab. 38:211, 1974.

pears to be related to the marked increase in plasma levels of testosterone which they observed in hyperthyroid men. If their observation is further substantiated, it raises the possibility that the decrease in plasma clearance of testosterone in hyperthyroidism is not only a function of an alteration in protein binding of the hormone but may be a more direct effect of the thyroid hormone on the peripheral metabolism of the steroid. We have previously described other circumstances where there is a dissociation between the plasma level and, presumably, the "free" hormone concentration and the MCRT.3l031 The production rate of the hormone is normal in both sexes (Table III).24`25 Despite the maintenance of a "normal" unbound concentration and production rate of testosterone, several investigators have reported moderate elevations in plasma levels of luteinizing hormone (LH)32in induced and spontaneous29'33'34 hyperthyroidism, although one report has failed to

Bull. N.Y. Acad. Med.

THYROID EFFECTS ON STEROID METABOLISM

251

confirm this observation.35 Since these studies were carried out with only single samples and plasma LH is known to undergo episodic fluctuation, more detailed studies are indicated to settle this question. If, indeed, plasma LH is elevated despite the "normal" unbound fraction and production rate of the hormone it would suggest that thyroxine has direct effects on the central nervous system or testes. In hypothyroidism the findings in testosterone metabolism are essentially the reverse of those in hyperthyroidism (Table III). In women there is a significant decrease in the plasma binding protein for testosterone with an increase in the MCR of the steroid.25'27 There are no data for hypothyroid men. The levels tend to return to normal after treatment.25'27 In vivo hormone conversion of testosterone to androstenedione (CR11) is decreased in hyperthyroidism and increased in hypothyroidism (Table III).25,36 The conversion ratio of testosterone to estrone (CR11 ) is significantly increased in both men and women (Table III).36 The product of the conversion ratio (CR) and plasma concentration (PC) of the hormone permits calculation of the instantaneous contribution of a precursor to a product steroid (CRBB-PrO . PC"~r). There is a significant increase in the instantaneous contribution of plasma testosterone to estrone and estradiol (Figure, I see above).36 In contrast to cortisol, there is evidence of a marked increase in the 5a reductase pathway for testosterone in hyperthyroidism.37 This was demonstrated by the increase in the ratio of urinary androsterone (5a) to etiocholanolone (I,8) in hyperthyroid patients given labelled testosterone (Table III). The converse was noted in hypothyroidism.37 McGuire and Tompkins38 showed that administration of thyroxine in the rat is associated initially with increased NADPH activity and that more prolonged treatment leads to significant increases in 5a reductase activity. While it is attractive to postulate that a similar event occurs in human hyperthyroidism, since the 5ax/5,f ratio changes in a consistent fashion, mechanisms other than increased activity may produce changes in the ratio of metabolites formed without a specific change in enzymic protein.1' Thus, caution should be exercised in attributing changes in urinary metabolite patterns to alterations in levels of enzyme activity. Direct measurement of hepatic reductase activity in hyperthyroid patients would be necessary to verify one or the other of these hypotheses. Androstenedione. Androstenedione (A) is an important precursor of the sex steroids. It is only weakly bound to plasma proteins39 and alterations in Vol. 53, No. 3, April 1977

252

G.G. GORDON AND A.L. SOUTHREN

TABLE IV. THE EFFECT OF HYPERTHYROIDISM ON VARIOUS PARAMETERS OF ANDROSTENEDIONE METABOLISM. THE VALUES GIVEN REPRESENT THE MEAN OF PUBLISHED DATA.

Androstenedione Hyperthyroidism Normal Male Female Male Female

Production rate (mg./day) CRAT (%)

CRAEI (%) CRAE () Metabolic clearance rate (liters/day) Plasma level

(,ig.Idl. plasma)

T(8.59) 4(30.5) T(4.0) T(0.9) N (2481)

T(6.72)

2.50 13.2

T(2.6) N (0.5) N (2329)

2.1

T(0.34)

T(0.29)*

T(32.9)

0.2 2487 0.11

3.52 13.6 1.0 0.2 2001 0.19

I-increase, N=normal, 4=decrease, *=calculated

thyroid function do not seem to alter its MCR. However, the plasma level and production rate of this hormone are markedly elevated in hyperthyroidism (Table IV).25'36 The origin of the increased production of androstenedione in this disorder has not been defined as yet. The markedly increased production rate of androstenedione in males with hyperthyroidism cannot be accounted for by increased conversion from testosterone. In the normal male 21% of the production rate of androstenedione derives from testosterone while in hyperthyroid men only 4% originates from this steroid.36 Thus, the increased production of androstenedione is most probably the result of direct glandular secretion, although increased conversion from another precursor cannot be excluded. In both normal and hyperthyroid females approximately 70% of the production rate of testosterone is derived from androstenedione ([A]TA). In hyperthyroid males this factor is increased from the normal of 2.4% to 10.2%. However, it should be appreciated that 90% of the plasma-production rate of testosterone is of testicular origin and is not derived from peripheral conversion. The conversion ratio of androstenedione to testosterone (CRyA) is increased twofold in both men and women with hyperthyroidism (Table IV).25,36 The conversion of androstenedione to the estrogens (CRA1 and CRA2) is increased significantly in both men and women with hyperthyroidism36 (Table IV). Thus, the instantaneous contribution of androstenedione to estrone and estradiol is markedly increased (Figure 1). The fraction of the plasma production of testosterone derived from the plasma androstenedione pool ([A]TA) Bull. N.Y. Acad. Med.

THYROID EFFECTS ON STEROID METABOLISM

253

Am ANDROSTENEDIONE T= TESTOSTERONE

MALES NORMALS

10.0 cd

HYPERTHYROIDISM

9

co

8

Cj s

I-

z CP

7 6

0 uj

5

ti a-

4

C C.> im

dC

3

2

a-

0

FEMALES NORMALS

10.0

HYPERTHYROIDISM

9.7

8I* 7 CY w

I 6 A: a

C,

0

ca-

S

4

3

CS co -i

2

C-.

0. A

af

T

AwA

T

Fig. 2. The plasma production rate (PR) of androstenedione (A) and testosterone (T) and the percentage of the respective production rates /[9]TA an d PRA { []BB*.P R'T * 100and LVJBB PR derived by interconversion. Reproduced by raAT

PRA

PRT

/

permission from Southren, A. L., Olivo, J. Gordon, G. G., Vittek, J., Brener, J., and Rafii, F.: The conversion of androgens to estrogens in hyperthyroidism. J. Clin Endocrinol. Metab. 38:212, 1974. Vol. 53, No. 3, April 1977

254

G.G. GORDON AND A.L. SOUTHREN

per day in hyperthyroid males is increased more than threefold (Figure 2). Conversely, the fraction of the daily plasma androstenedione production derived from plasma testosterone ([A]AT) in these patients decreased significantly (Figure 2). There were no significant changes in these parameters in female patients with hyperthyroidism.36 Dihydrotestosterone. Dihydrotestosterone (DHT) is an important androgen since it mediates the intracellular action of testosterone in certain target tissues. This hormone is derived predominantly by peripheral conversion from testosterone by the action of steroid 5a-reductase. Since dihydrotestosterone is bound more highly to the specific high-affinity binding protein of the sex steroid in plasma than testosterone,40 we might expect alterations in its metabolism in disorders of thyroid function to be quantitatively similar to those of testosterone. Indeed, the plasma level of dihydrotestosterone in hyperthyroid men is elevated more than sixfold (279 ng.% compared with a normal of 44 ng.%).29 The MCRDHT has been reported only in hyperthyroid females and is decreased significantly (95 liters/ day/mi.2 versus a normal of 209 liters/day/mi.2).40 The latter study also reported a significant increase in the peripheral conversion of testosterone to dihydrotestosterone as manifested by an increase in the CRT 1HT and the [p] TBBDHT. Since data on the plasma level and MCRDHT in patients of the same sex are not available, the production rate of the hormone remains undefined. The effects of thyroid dysfunction on the plasma binding of dihydrotestosterone (presumably increased in hyperthyroidism and decreased in hypothyroidism) also has not been reported as yet.

C18 STEROIDS The effects of thyroid dysfunction on the metabolism of C18 steroids tend to resemble those of the C19 rather than the C21 steroids. 17/3-Estradiol (E2). Data on E2 metabolism are only available in hyperthyroidism. Chopra et al.,2933 Bercovici and Mauvais-Jarvis,4' and our group31 (Table I) have reported significantly elevated plasma E2 levels in hyperthyroid men. The values varied between 2½/2 and 15 times normal (Chopra et al.:33 6 ng.% compared with a normal of 2.4 ng.% [N=10]; Chopra and Tulchinsky: 2 9 10.8 ng.% versus a normal of 2.7 ng.% [N=15]; Bercovici and Mauvais-Jarvis:41 1,450 ng.% versus a normal of 30 ng.% [N=I] and Olivo et al.,31 20 ng.% versus a normal of 3.4 ng.% [N=4]). Ridgway, Longcope, and Maloof35 found only slight increases in plasma E2 levels in hyperthyroidism. The increases were not statistically Bull. N.Y. Acad. Med.

255

THYROID EFFECTS ON STEROID METABOLISM

TABLE V. THE EFFECT OF HYPERTHYROIDISM ON VARIOUS PARAMETERS OF ESTRADIOL METABOLISM. THE VALUES GIVEN REPRESENT THE MEAN OF PUBLISHED DATA.

Estradiol Normal Hyperthyroidism Male Female Male Female

Production rate (pug./day) CRE2E1 (%) Metabolic clearance rate (liters/day) Plasma concentration (ug./liter) Protein binding (SS BG) (jkg./dl.)

T(156)

N (203)

N (1 1.9)

T(18.9) 1(804) N (0.260) T(8.6)

1(733) 1(0 200) 1(3.9)

57 11.0

1,664

0.034 1.4

150 10.5 1,200 0.117 2.3

t=increase, N=normal, 4 =decrease, *= calculated

significant (4.6 ng.% compared with a normal of 2.9 ng.%). However, in three of their patients who were studied after they returned to the euthyroid state significant decreases in plasma E2 levels were noted. The increased plasma E2 in hyperthyroidism results in large part from increased peripheral conversion from the androgens (testosterone and androstenedione) (Figure 1)36 rather than from direct glandular secretion. The plasma levels of E2 in female hyperthyroidism tend to be within-the normal range31'35 (Table V). The MCRE2 in our studies of hyperthyroidism was decreased significantly in both sexes31'32 (Table V). This was confirmed recently by Ridgway, Longcope, and Maloof.35 We found the plasma levels of E2 to be elevated significantly, resulting in an increased plasma production rate of the hormone31 (Table V). The presence of an increased production rate tends to be supported by the significant increases in plasma E2 found by other investigators.293342 The magnitude of these increases (21/2- to 15fold) is greater than can be accounted for by a decrease in the MCRE2 of 30%35 to 50% .31 The increased plasma production rate of E2 in male hyperthyroidism may be of some significance in explaining the gynecomastia seen in this disorder. This relation is further strengthened by the observation of Chopra and Tulchinsky29 and Chopra et al.33 that hyperthyroid men with gynecomastia had a higher mean unbound or "free" plasma concentration of E2. The high-affinity sex steroid binding globulin in plasma which, as previously stated, is increased in hyperthyroidism, binds not only testosterone but also E2. The decreased MCRE2 and the increased plasma level Vol. 53, No. 3, April 1977

256

G.G. GORDON AND A.L. SOUTHREN

of E2 may be a consequence of this increased binding. Somewhat discordant with this hypothesis, however, is the observation of Loriaux, Kono, and Lipsett,43 which showed that the sex-steroid binding protein does not bind E2 at body temperature in vitro and, therefore, may not effect its metabolism in vivo. This observation casts doubt as to the role of the sex-steroid binding globulin in regulating E2 metabolism. Another inconsistency in the relation between plasma binding and E2 metabolism stems from the observation of Chopra et al.3 on the unbound or "free" concentration of E2. These investigators showed a fall in the dialyzable fraction of plasma E2, from 2.15% in normal individuals to 1.32% in those with hyperthyroidism. Since the total plasma E2 concentration was elevated fourfold, the unbound concentration of E2 (percent dialyzable times total E2 in hyperthyroid men was elevated almost threefold (135 pg./ml. compared with a normal of 55.6 pg./ml.). Thus, the MCRE2, which should reflect the unbound or "free" mass of steroid available for removal, may be decreased in hyperthyroidism by a mechanism other than changes in plasma binding. It has been speculated that hyperthyroidism may alter E.B metabolism by decreasing the amount of subcutaneous fat and muscle mass available for metabolizing the steroid.35 Estrone (El). Little information is available on EB metabolism in disorders of the thyroid. Bercovici and Mauvais-Jarvis41 found elevated plasma E1 levels in one patient with hyperthyroidism and gynecomastia (8,400 pg./ml. versus a normal of 870 pg./ml.). We demonstrated elevated plasma E1 levels in hyperthyroid men (252 pg./ml. compared with a normal of 63 pg./ml.).3I The MCREI was decreased significantly in both men (1,427 + 163 [SD] liters/day [N = 4]) and women (1,526 and 1,677 liters/day [N = 2]) with hyperthyroidism.44 The normal mean MCRE1 is approximately 2,500 liters/day. The production rate of E1 was increased in hyperthyroid men and women (358 + 68 [SD] ng./day and 293 [average of 2 values] ng./day respectively). The increased production of E1 also reflects the increased conversion from the androgens.3f The increased circulating E1 also may contribute to the gynecomastia seen in hyperthyroid men. Urinary estrogen metabolites. Hyperthyroidism markedly alters the pattern of the urinary metabolites of E2. Fishman et al.4546 showed that there is a significant decrease in the conversion of 16-C14-estradiol to estriol and a significant increase in the conversion to 2-methoxyestrone and 2-hydroxyestrone. In myxedema the amount of 2-hydroxyestrone is diminBull. N.Y. Acad. Med.

THYROID EFFECTS ON STEROID METABOLSIM

257

ished. The authors suggest that the hydroxylations at C-2 and C- 16 of estrone are competitive reactions and are regulated reciprocally by the level of thyroid function. SUMMARY

The effects of altered thyroid function on the metabolism of the steroid hormones in man and in experimental animals have been reviewed. The metabolic clearance and secretory rates of the C-21 steroids (cortisol and aldosterone) are increased in hyperthyroidism and decreased in myxedema. In contrast, the metabolism of the C-19 and C-18 steroids (sex hormones) is decreased in hyperthyroidism and increased in hypothyroidism. The effect of thyroid dysfunction on the sex hormones appears to be mediated, at least in part, by alteration in the plasma protein binding of the steroids. The clinical implications of these findings have been discussed and areas for further investigation have been suggested. REFERENCES 1. Peterson, R. E.: The influence of the thyroid on adrenal cortical function. J. Clin. Invest. 37:736, 1958. 2. Garren, L. D. and Lipsett, M. B.: The effect of euthyroidal hypermetabolism on cortisol removal rates. J. Clin. Endocrinol. Metab. 21:1248, 1961. 3. Myers, J. D., Brannon, E. S., and Holland, B. C.: A correlative study of the cardiac output and the hepatic circulation in hyperthyroidism. J. Clin. Invest. 29: 1069, 1950. 4. Martin, M. M., Mintz, D. H., and Tamagaki, H.: Effect of altered thyroid function upon steroid circadian rhythms in man. J. Clin. Endocrinol. Metab. 23:242, 1963. 5. Gallagher, T. F., Hellman, L., Finkelstein, J., Yoshida, K., Weitzman, E. D., Roffwarg, H. D., and Fukushima, D. K.: Hyperthyroidism and cortisol secretion in man. J. Clin. Endocrinol. Metab. 34:919, 1972. 6. Beisel, W. R., DiRaimondo, V. C., Chu, P. Y., Rosner, J. M., and Forsham, P. H.: The influence of plasma protein binding on the extra-adrenal metabolism of cortisol in normal,

Vol. 53, No. 3, April 1977

7.

8.

9.

10.

11.

hyperthyroid and hypothyroid subjects. Metabolism 13:942, 1961. Doe, R. P., Fernandez, R., and Seal, U. S.: Measurement of corticosteroidbinding globulin in man. J. Clin. Endocrinol. Metab. 24:1029, 1964. Gala, R. R. and Westphal, U.: Influence of anterior pituitary hormones on the corticosteroid-binding globulin in the rat. Endocrinology 79:55, 1966. Gala, R. R. and Westphal, U.: Further studies on the corticosteroidbinding globulin in the rat. Proposed endocrine control. Endocrinology 79:67, 1966. Zumoff, B., Bradlow, H. L., Gallagher, T. F., and Hellman, L.: Decreased conversion of androgens to normal 17-ketosteroids metabolites: A nonspecific consequence of illness. J. Clin. Endocrinol. Metab. 32:824, 1971. Hellman, L., Bradlow, H. L., Zumoff, B., and Gallagher, T.F.: The influence of thyroid hormone on hydrocortisone production and metabolism. J. Clin. Endocrinol. Metab. 21:1231, 1961.

258

G.G. GORDON AND A.L. SOUTHREN

12. Gold, N. I. and Crigler, J. F., Jr.: Influence of L-triiodothyronine on steroid hormone metabolism: Studies in a patient with adrenal hyperplasia (Cushing's Syndrome). J. Clin. Endocrinol. Metab. 23:156, 1963. 13. Beale, R. N., Croft, D., and Powell, D.: Some effects of thyroid disease on neutral steroid metabolism. J. Endocrinol. 57:3 17, 1973. 14. Koerner, D. R. and Hellman, L.: Effect of thyroxine administration on the 1 -beta-hydroxysteroid dehydrogenases in rat liver and kidney. Endocrinology 75:592, 1964. 15. Hilton, J. G., Black, W. C., Athos, W., McHugh, B., and Westermann, C. D.: Increased ACTH-like activity in plasma in patients with thyrotoxicosis. J. Clin. Endocrinol. Metab. 22:900, 1962. 16. Holst, J.: Pathologische anatomie der organe ausser der schildruse bei der Basedowschen Krankheit. Int. Kropfkonf. Zweite, Bern, 1933, Verhandlungsbericht. Bern, H. Huber, 1935, p. 62. 17. Kirkeby, K., Hangaard, G., and Lingjaerde, P.: The pigmentation of thyrotoxic patients. Acta Med. Scand. 174:257, 1963. 18. Ingbar, S. H.: Thyrotoxic storm. In: The Thyroid, 3d ed., Werner, S. C. and Ingbar, S.H., editors. New York, Harper and Row, 1971, p. 659. 19. Werner, S. C.: Myxedema Coma. In: The Thyroid, 3d ed., Werner, S.C. and Ingbar, S. H. editors. New York, Harper and Row, 1971, p. 838. 20. Luetscher, J. A., Cohn, A. P., Camargo, C. A., Dowdy, A. J., and Callaghan, A. M.: Aldosterone secretion and metabolism in hyperthyroidism and myxedema. J. Clin. Endocrinol. Metab. 23:873, 1963. 21. Milech, A., Robin, N. I., Dluhy, R. G., and Williams, G. H.: Altered responsiveness of the renin aldosterone system in hyperthyroidism. Clin. Res. 18: 367, 1970. 22. Hauger-Klevene, J. H., Brown, H., and Zavaleta, J.: Plasma renin activity in hyper- and hypothyroidism: Effect

23.

24.

25.

26.

27.

28.

29.

30.

31.

of adrenergic blocking agents. J. Clin. Endocrinol. Metab. 34:625, 1972. Cain, J. P., Dluhy, R. G., Williams, G. H., Selenkow, H. A., Milech, A., and Richmond, S.: Control of aldosterone secretion in hyperthyroidism. J. Clin. Endocrinol. Metab. 36:365, 1973. Dray, F., Sebaoun, J., Mowszowicz, I., Delzant, G., Desgrez, P., and Dreyfus, G.: Facteurs influencant le taux de la testosterone plasmatique chez l'homme: Role des hormones thyroidiennes. C. R. Acad. Sci. (Paris) 264:2578, 1967. Gordon, G. G., Southren, A. L., Tochimoto, S., Rand, J. J., and Olivo, J.: Effect of hyperthyroidism and hypothyroidism on the metabolism of testosterone and androstenedione in man. J. Clin. Endocrinol. Metab. 29:164, 1969. Crepy, O., Dray, F., and Sebaoun, J.: Role des hormones thyroidiennes dans les interactions entre la testosterone et les proteins seriques. C. R. Acad. Sci. (Paris) 264:2651, 1967. Olivo, J., Southren, A. L., Gordon, G. G., and Tochimoto, S.: Studies of the protein binding of testosterone in plasma in disorders of thyroid function: Effect of therapy. J. Clin. Endocrinol. Metab. 31:539, 1970. Tulchinsky, D. and Chopra, I. J.: Competitive ligand-binding assay for measurement of sex hormone-binding globulin (SHBG). J. Clin. Endocrinol. Metab. 37:873, 1973. Chopra, I. J. and Tulchinsky, D.: Status of estrogen-androgen balance in hyperthyroid men with Graves disease. J. Clin. Endocrinol. Metab. 38:269, 1974. Southren, A. L., Gordon, G. G., and Tochimoto, S.: Further study of factors affecting the metabolic clearance rate of testosterone in man. J. Clin. Endocrinol. Metab. 28:1105, 1968. Olivo, J., Gordon, G. G., Rafli, F. and Southren. A. L.: Estrogen metabolism in hyperthyroidism and in cirrhosis of the liver. Steroids 26:47, 1975.

Bull. N.Y. Acad. Med.

THYROID EFFECTS ON STEROID METABOLSIM

32. Ruder, H., Corval, P., Mahoudeau, J. A., Ross, G. T., and Lipsett, M. B.: Effects of induced hyperthyroidism on steroid metabolism in man. J. Clin. Endocrinol. Metab. 33:382, 1971. 33. Chopra, I. J., Abraham, G. E., Chopra, U., Solomon, D. H., and Odell, W. D.: Alterations in circulating estradiol-17,3 in male patients with Graves disease. N. Engl. J. Med. 286:124, 1972. 34. Akande, E. 0. and Hockaday, T. D. R.: Luteinizing hormone, oestrogen and progesterone levels in thyrotoxic menstrual disturbances. IV Int. Congr. Endocrinol. Int. Congress Series, No. 256, 1972. Abstract 68. 35. Ridgway, E. C., Longcope, C., and Maloof, F.: Metabolic clearance and blood production rates of estradiol in hyperthyroidism. J. Clin. Endocrinol. Metab. 41:491, 1975. 36. Southren. A. L., Olivo, J., Gordon, G. G., Vittek, J., Brener, J., and Rafii, F.: The conversion of androgens to estrogens in hyperthyroidism. J. Clin. Endocrinol. Metab. 38:207, 1974. 37. Hellman, L., Bradlow, H. L., Zumoff, B., Fukishima, D., and Gallagher, T. F.: Thyroid-androgen interrelations and the hypocholesterolemic effect of androsterone. J. Clin. Endocrinol. Metab. 19:936, 1959. 38. McGuire, J. S. and Tompkins, G. M.: The effects of thyroxin administration on the enzymic reduction of A4-3ketosteroids. J. Biol. Chem. 234:791, 1959. 39. Rivarola, M. A., Forest, M. G., and

Vol. 53, No. 3, April 1977

40.

41.

42.

43.

44. 45.

46.

259

Migeon, C. J.: Testosterone, androstenedione, and dehydroepiandrosterone in plasma during pregnancy and delivery: Concentration and protein binding. J. Clin. Endocrinol. Metab. 28:34, 1968. Saez, J. M., Forest, M. G., Morera, A. M., and Bertand, J.: Metabolic clearance rate and blood production rate of testosterone and dihydrotestosterone in normal subjects, during pregnancy and in hyperthyroidism. J. Clin. Invest. 51:1226, 1972. Bercovici, J. P. and Mauvais-Jarvis, P.: Hyperthyroidism and gynecomastia: Metabolic studies. J. Clin. Endocrinol. Metab. 35:671, 1972. Olivo, J., Gordon, G. G., Rafii, F., and Southren, A. L.: Metabolic clearance rates (MCR) of estradiol (E2) in cirrhosis and hyperthyroidism. Endocrine Soc. Program 56th Meeting, Atlanta, 1974. Abstract 454. Loriaux, D. L., Kono, S., and Lipsett, M. B.: Plasma estradiol (E2) binding and renal clearance: The effect of testosterone-estradiol binding globulin (TeBG). Endocrine Soc. Program 56th Meeting, Atlanta, 1974. Abstract 92. Olivo, J., Gordon, G. G., and Southren, A. L.: Unpublished observations. Fishman, J., Hellman, L., Zumoff, B., and Gallagher, T. F.: Influence of thyroid hormone in estrogen metabolism in man. J. Clin, Endocrinol. Metab. 22:389, 1962. Fishman, J., Hellman, L., Zumoff, B., and Gallagher, T. F.: Effect of thyroid on hydroxylation of estrogen in man. J. Clin. Endocrinol. Metab. 25:365, 1965.