J. Endocrinol. Invest. 2: 379, 1979
Hypothalamic-pituitary function in Cushing's disease D. Le Raith*, M.S. Shapiro**, A. Gutman**, and I.M. Spitz*** *Department of Medicine, Soroka Medical Center, Faculty of Health Science~, Ben-Gurian University of the Negev, Beer-Sheba, **Meir Hospital, Kfar Saba and***Department of Endocrinology and Metabolism, Shaare Zedek Medical Center, Jerusalem, Israel
ABSTRACT. Hypothalamic pituitary function was evaluated in seven patients with Cushing's disease. In all subjects there was an absence of GH elevation following hypoglycemia. Three patients demonstrated basal hyperprolactinemia. Six had an intact PRL rise following TRH. However, four patients failed to show PRL elevation with insulin hypoglycemia and six were unresponsive to chlorpromazine. Five patients showed impaired TSH response to TRH. Many of the subjects had low basal LH and FSH levels. Attenuated or absent gonadotropin responses to LHRH were noted in three females. One male demonstrated an exaggerated FSH response to LHRH. These results indicate that multiple abnormalities of anterior pituitary hormone secretion characterize Cushing's disease. tisol, elevated ACTH levels and the presence of increased urinary 11-hydroxycorticosteroid secretion with lack of suppression following 2 mg dexamethasone for two days, but with good suppression following the administration of 8 mg dexamethasone (Table 1). None of the patients had received specific therapy for their Cushing's disease. Cases 1 and 7 were mildly diabetic and elevated blood pressure levels were recorded in all cases, except in cases 5 and 6. None of the subjects was hypothyroid, and none had galactorrhea. All medication was withheld one week prior to the investigation. Antero-posterior and lateral hypocloidal poly tomography of the pituitary fossa was normal in all subjects.
INTRODUCTION Cushing's disease is primarily a disorder of overproduction of adrenocorticotropin (ACTH) although abnormalities in the secretion of other anterior pituitary hormones have also been noted. Absent growth hormone (GH) and prolactin (PRL) elevation following insulin induced hypoglycemia as well as absent thyrotropin (TSH), luteinizing hormone (LH) and follicle stimulating hormone (FSH) responses to their respective releasing hormones have been described in both Cushing's disease (1-4) and after chronic administration of high dose glucocorticoids in normal subjects (57). Although some of these abnormalities resolve following correction of the hypercortisolism, others do not disappear suggesting an intrinsic hypothalamicpituitary defect (1). This series represents a comprehensive study of hypothalamic-pituitary function in seven patients with Cushing's disease.
Study protocol Informed consent was obtained from each subject. All the tests were performed on separate mornings following an eight hour fast. Blood samples v ere drawn through an indwelling needle, kept patent with a slow infusion of normal saline. After insertion of ,he needle, a 15 minute stabilizing period was allowed juring which two baseline samples were taken. Therbafter the test substance was injected and periodic s,lmpling continued. The blood samples were centrifuged, separated immediately and the serum kept frozen, at -20 C until assayed.
MATERIALS AND METHODS Patients Three male and four female patients, aged between 18 and 55 years were studied. Three of the females (Cases 1,2 and 3), aged 52, 54 and 49 years respectively, were postmenopausal and the fourth, aged 38 (Case 4) was premenopausal. but amenorrheic. The males (Cases 5, 6 and 7) were aged 26, 19 and 46 years respectively. Active Cushing's disease was confirmed by the absence of diurnal rhythm in plasma cor-
Tests i) Insulin hypoglycemia test (IHT). 0.1 - 0.2 units per kg body weight of soluble insulin were administered intravenously to ensure adequate hypoglycemia. Serum glucose, GH and PRL were then measured over a 90 minute period.
Key·words Cush,ng's d'sease. PRL. LH. FSH. TSH. GH
ii) Thyrotropin releasing hormone (TRH). Blood samples for TSH and PRL were taken over a 60 minute period following the iv administration of 200119 TRH.
Correspondence: Dr. I M. SPitZ. Department of Endocrrnology and Metabolism. Shaare Zedek Medical Center. PO. Box 293. Jerusalem 9100. Israel Received December 4, t 978: accepted May 30. t 979
379
0. Le ROith, M.S. Shapiro, A. Gutman, et al. Table 1 - Serum and urinary steroid values. Patient
ACTH pg/ml
1 2 3 4 5 6 7
Plasma cortisol Jlg/l00 ml 08:00
24:00
Baseline
125.4 . 800 700
50 17 21 15 20 26 18
40 18 45 20 26 20 25
930 2083 443 1010 480 1620 921
860 1006 390 902 400 1082 531
230 250 102 200 150 230 ;:>03
20-80
10-25
1-8
55-264
32-128
22-71
840 120.0 2008
Normal range
e
'-
~
c:
:t
... ...
Laboratory
Serum LH, FSH , TSH , PRL and GH were determined by previously described methods (8). Pituitary LH, FSH, TSH, PRL and GH and their respective antisera were kindly supplied by the National Pituitary Agency, National Institute of Arthritis, Metabolism and Digestive Diseases (NIAMDD) The second International Reference preparation for human menopausal gonadotrophins (2nd IRPHMG) provided by Dr. D. R. Bangham was used as ~ reference standard for both LH and FSH. TSH and PRL standards were also supplied by Dr. D. R. Bangham. Nati onal In.stltute lor Bl olo(1 lcal Stcwocll(ls. London, England.
u
, --
~
0 ~
Q:
Dexamethasone 2 mg 8 mg
iii) Luteinizing hormone releasing hormone (LHRH) . 100 pg LHRH was injected as a bolus and LH and FSH measured over a 60 minute period. iv) Chlorpromazine (CPZ). 25 mg was injected im. Halfhourly samples were taken for two hours for PRL. One female (Case 2) had the CPZ test repeated, following the administration of ethinyl estradiol 50 pg daily for 30 days. Seven male and five female controls aged 18 - 26 years were similarly challenged with LHRH, TRH and CPZ. The females were tested during the follicular phase of the menstrual cycle .
MALES
FEMALES
0
Urinary 11 -hydroxycorticosteroid J.I9/24 hours
I HT'
60 '0
Q..20
o
RESULTS
6 !
I M
t
Growth hormone Basal levels were below 1.3 ng / ml in all subjects. During IHT symptomatic hypoglycemia was obtained in all subjects. Despite adequate hypoglycemia (glucose nadir below 40 mg / 100 ml) there was consistent absence of GH response in each of the seven patients.
(~inJ
Fig. 1 - The prolactin response to (i) 200IJgTRH administered iv, ii) 25 mg chlorpromazine (CPl) , and iii) insulin induced hypoglycemia (IHT) in the seven patients with Cushing's disease. The response in each individual is numbered accordingly. Mean ± SE response to TRH and CPZ in controls (broken line) is also shown.
Prolactin The results are shown in Figure 1. Basal PRL levels
380
Hypothalamic-pituitary function in Cushing's disease
MALES
FEMALES 30
~/I
, I
,
o
,,'
I
I
I
I _-1 "" r--'
30
"
"
I
20
I
"0_--_2 __ ---4
10
o
~
___________ 7
Fig. 2 - TSH responses to 200 pg iv TRH in
2b
both male and female patients, numbered accordingly, and compared to controls (broken line, mean ± Sf).
were elevated ( > 20 ng/ ml) on at least one occasion in female subjects 2, 3 and 4. The remaining subjects had normal basal levels. When compared to the controls, six of the subjects demonstrated PRL elevation following TRH. Only female subject 3, with hyperprolactinemia, was unresponsive. Only male patient 6 showed a PRL response to CPZ. The other six subjects failed to
show PRL elevation. Case 2 remained unresponsive to CPZ despite ethinyl estradiol pretreatment. There was an attenuated response to insulin hypoglycemia in female subjects 2, 3 and 4 and male subject 5 (Fig. 1). In the remaining three subjects there was a trebling of their basal levels which is considered a normal response (9).
50
b
TIME (min)
FEMALES
Thyrotropin As shown in Figure 2, basal TSH levels ranged from 1.0 to 6.0 pU/mL When compared to control values there was an impaired TSH response to TRH in all four females and one male (Case 7).
MALES
Gonadotropins The results are shown in Figure 3. When compared to the controls, basal LH levels were low in female cases 1 and 4 and male case 5. Basal FSH levels were low in case 1. Case 6 had elevated basal FSH levels. Following the administration of LHRH female case 1 had LH and FSH uresponsiveness Female cases 2 and 4 had impaired LH responses. LH levels increased in case 3 and FSH in cases 2, 3 and 4. As regards the males, LH and FSH responses were intact with the exception of case 6, who had an exaggerated FSH rise.
/.'
DISCUSSION We have documented multiple abnormalities in pituitary hormone secretion in our seven cases with Cushing's disease. The most consistent defect noted in our study was the absent GH response to insulin induced hypoglycemia. This phenomenon has been described previously both in Cushing's disease (1) as well as in normal subjects given high dose glucocorticoids (5, 10). In the latter condition GH responsiveness returns after removal of the high dose glucocorticoid administration , whereas in Cushing's disease GH remains unresponsive to insulin hypoglycemia despite normalization of the cortisol levels after bilateral adrenalectomy (1). This suggests that in Cushing's disease there may
s , po-• I.. "-"~-A1 ~7
20
'0
60
0
TIM£(min)
20
~
60
Fig. 3 - LH and FSH responses to 100 pg iv LHRH in both male and female patients. Mean ± Sf of controls (broken line) IS also shown.
381
O. Le Roith, M.S. Shapiro, A. Gutman, et al. the glycoprotein hormones (TSH, LH and FSH), with recovery following normalization of cortisol levels after adrenalectomy. Evidence for an additional hypothalamic defect which is an integral part of the disease is seen with respect to GH and possibly PRL. This is shown by the absent GH and PRL response to hypoglycemia which persists despite correction of the hypercortisolism, the elevated basal PRL levels and loss of nocturnal PRL elevation. The lack of PRL response to CPl in our cases and the normalization of GH responsiveness to hypoglycemia following bromoergocryptine administration (3) suggests that these may be related to an alteration in dopamine metabolism, though a direct effect of excess glucocorticoid at the hypothalamic level cannot be excluded.
be a centrally based hypothalamic abnormality in GH secretion. Evidence that dopaminergic depletion may be implicated in this phenomenon was demonstrated by Lamberts et al. (3) who showed that GH responsiveness to insulin hypoglycemia was partially restored in untreated patients following the administration of the dopamine agonist bromoergocryptine. We have assessed PRL dynamics using three stimuli in each patient: i) TRH, which acts directly at the lactotroph; ii) insulin hypoglycemia, whose effect is probably at the hypothalamic level; iii) CPl, which induces PRL elevation by producing dopaminergic blockade at both hypothalamus and pituitary (8). The normal PRL response to TRH attests to the integrity of the lactotroph. However, the high basal levels in three of our patients, together with the absent PRL response to hypoglycemia in three, implies a defect in PRL control, which probably resides at the hypothalamic level or above. Supportive evidence is the loss of nocturnal elevation of PRL which has been described in Cushing's disease (11). The absent PRL response to CPl noted in our case is also in keeping with an abnormality of dopaminergic metabolism which has already been alluded to. It is possible that the defect in GH and PRL secretion noted in Cushing's disease may be intrinsic to the hypothalamic pituitary dysfunction and not a consequence of the hypercortisolism. However, it should be noted that high dose dexamethasone administration suppresses the PRL response to TRH which was intact in our subjects (6). Tha absent TSH response to TRH in our patients has been previously described in Cushing's disease (2, 3) as well as in normal subjects receiving long term high dose glucocorticoids (13), although Hershman and Pittman (14) demonstrated normal responses in their series of Cushing's disease. Since the abnormality is reversible after treatment of the Cushing's disease (2) it would seem that the unresponsiveness of TSH is due to the direct glucocorticoid effect at a pituitary level and not to an abnormality in neurotransmission. The absent gonadotropin response to LHRH has been described in Cushing's disease (4) as well as in normal subjects receiving glucocorticoid therapy (7). Many of our patients showed low basal gonadotropin levels and attenuated or absent responses to LHRH. Although some of our female patients did demonstrate a rise of LH and FSH after LHRH, three of these Cushing's patients were postmenopausal. Postmenopausal females normally have high basal gonadotropin levels and exaggerated responses to LHRH. Since the impaired LH and FSH response to LHRH has been shown to recover after adrenalectomy (15, 1 6), this suggests that the defect is related to the glucocorticoid action at a pituitary level. Thus the basis for the abnormalities in the secretion of anterior pituitary hormones in Cushing's disease is multifactorial. There is evidence for direct interference at pituitary level by the hypercortisolism with regard to
ACKNOWLEDGMENTS Hoffman La Roche kindly supplied the LHRH and TRH which was used This study was supported by grants from the Jewish Agency to D. LeA and the Chief Scientist's Office. Israel Ministry of Health to IMS We thank S Tristian, Z Shemesh and H. Gershman for their techmcal aid and H. Menachemson for typing the manuscript.
REFERENCES 1. Krieger O.T., Glick S.M. Growth hormone and cortisol responsiveness in Cushing's syndrome. Relation to a possible central nervous system etiology. Am. J, Med. 52: 25, 1972. 2. Kuku S.F., Child OF., Nader S, Fraser TR Thyrotropin and prolactin responsiveness to thyrotropin releasing hormone in Cushing's disease, Clin, Endocrino! (Oxt) 4: 437, 1975. 3. Lamberts S.w,J" Timmermans HAT., De Jong F.H .. Birkenhager J.C. The role ot dopaminergic depletion in the pathogeneSiS ot Cushing's disease and the possible consequences for medical therapy. Clin. Endocrino! (Oxf.) 7 185, 1977. 4. Boccuzzi G., Angeli A .. Bisbocci 0 .. Fonzo 0 .. Gaidano GP .. Ceresa F. Effect of synthetic LHRH on the release of gonadotropins in Cushing's disease. J. Clin. Endocrino! Metab. 40: 892. 1975. 5. Hartog M., Gaatar MA. Fraser R. Effect of corticosteroids on serum growth hormone, Lancet 2: 376, 1964. 6. Sowers JR.. Carlson H.E .. Brautbar N .. Hershman J.M. Effect of dexamethasone on prolactin and TSH response to TRH and metoclopramide in man. J. Clin. Endocrino!. Metab. 44: 237. 1977. 7. Sakakura M .. Kazuo T .. Nagasawa S. Inhibition of luteinizing hormone secretion induced by synthetic LHRH by long term treatment with glucocorticoids In human subjects. J. Clin. Endocrino! Metab, 40: 774. 1975. 8 Spitz 1M. Almaliach U .. Rosen E.. Polishuk W. Rabinowitz O.
382
Hypothalamic-pituitary function in Cushing's disease
Dissociation of prolactin responsiveness to TRH and chlorpromazine in women with isolated gonadotropin deficiency. J. Clin. EndocrinoL Metab. 45 1173. 1977.
Effects of glucocorticoids on pituitary hormonal responses to hypoglycemia. Inhibition of prolactin release. J. Clin. EndocrinoL Metab. 40. 442, 1975. 13. Otsukl M, Dakoda M. Baba S Influence of glucocorticoids on TRH-induced TSH response in man. J. Clin. EndocnnoL Metab. 36 95, 1973.
9. WoolH PO, Lee LA, Leebaw W., Thompson D. Lilavivathana U., Brodows R. Intracellular glucopenia causes prolactin release in man. J. Clin. EndocrinoL Metab. 45 377, 1977. 10. Frantz A.G, Rabkin M.T. Human growth hormone. Clinical measurement, response to hypoglycemia and suppression by corticosteroids. N. EngL J. Med. 271. 1375, 1964. 11. Krieger 01., Howanitz P.J., Frantz A.G. Absence of nocturnal elevation of plasma prolactin concentrations in Cushing's disease. J. Clin. EndocrinoL Metab. 42 260. 1976.
14. Hershman JM, Pittman JA Response to synthetic TRH in man. J. Clin. EndocrinoL Metab. 31. 457, 1970 15. Luton J.P. Thieblo P, Valcke JC. Mahoudeau JA. Bricaire H. Reversible gonadotropin deficiency in male Cushing's disease. J. Clin. EndocrinoL Metab. 45 488, 1977. 16. Hackenberg K, Hoff HG. Brouwers HP., Relnwein D. Gonadotropin stimulation test in female Cushing's disease before and after adrenalectomy. Acta. EndocrinoL (SuppL) 208 25. 1977.
12. Copinshi G, L'Hermile M, LeClerq R, Goldstein J, Vanhaelst L., Virasoro E, Robyn C
383
Hypothalamic-pituitary function in Cushing's disease D. Le Raith*, M.S. Shapiro**, A. Gutman**, and I.M. Spitz*** *Department of Medicine, Soroka Medical Center, Faculty of Health Science~, Ben-Gurian University of the Negev, Beer-Sheba, **Meir Hospital, Kfar Saba and***Department of Endocrinology and Metabolism, Shaare Zedek Medical Center, Jerusalem, Israel
ABSTRACT. Hypothalamic pituitary function was evaluated in seven patients with Cushing's disease. In all subjects there was an absence of GH elevation following hypoglycemia. Three patients demonstrated basal hyperprolactinemia. Six had an intact PRL rise following TRH. However, four patients failed to show PRL elevation with insulin hypoglycemia and six were unresponsive to chlorpromazine. Five patients showed impaired TSH response to TRH. Many of the subjects had low basal LH and FSH levels. Attenuated or absent gonadotropin responses to LHRH were noted in three females. One male demonstrated an exaggerated FSH response to LHRH. These results indicate that multiple abnormalities of anterior pituitary hormone secretion characterize Cushing's disease. tisol, elevated ACTH levels and the presence of increased urinary 11-hydroxycorticosteroid secretion with lack of suppression following 2 mg dexamethasone for two days, but with good suppression following the administration of 8 mg dexamethasone (Table 1). None of the patients had received specific therapy for their Cushing's disease. Cases 1 and 7 were mildly diabetic and elevated blood pressure levels were recorded in all cases, except in cases 5 and 6. None of the subjects was hypothyroid, and none had galactorrhea. All medication was withheld one week prior to the investigation. Antero-posterior and lateral hypocloidal poly tomography of the pituitary fossa was normal in all subjects.
INTRODUCTION Cushing's disease is primarily a disorder of overproduction of adrenocorticotropin (ACTH) although abnormalities in the secretion of other anterior pituitary hormones have also been noted. Absent growth hormone (GH) and prolactin (PRL) elevation following insulin induced hypoglycemia as well as absent thyrotropin (TSH), luteinizing hormone (LH) and follicle stimulating hormone (FSH) responses to their respective releasing hormones have been described in both Cushing's disease (1-4) and after chronic administration of high dose glucocorticoids in normal subjects (57). Although some of these abnormalities resolve following correction of the hypercortisolism, others do not disappear suggesting an intrinsic hypothalamicpituitary defect (1). This series represents a comprehensive study of hypothalamic-pituitary function in seven patients with Cushing's disease.
Study protocol Informed consent was obtained from each subject. All the tests were performed on separate mornings following an eight hour fast. Blood samples v ere drawn through an indwelling needle, kept patent with a slow infusion of normal saline. After insertion of ,he needle, a 15 minute stabilizing period was allowed juring which two baseline samples were taken. Therbafter the test substance was injected and periodic s,lmpling continued. The blood samples were centrifuged, separated immediately and the serum kept frozen, at -20 C until assayed.
MATERIALS AND METHODS Patients Three male and four female patients, aged between 18 and 55 years were studied. Three of the females (Cases 1,2 and 3), aged 52, 54 and 49 years respectively, were postmenopausal and the fourth, aged 38 (Case 4) was premenopausal. but amenorrheic. The males (Cases 5, 6 and 7) were aged 26, 19 and 46 years respectively. Active Cushing's disease was confirmed by the absence of diurnal rhythm in plasma cor-
Tests i) Insulin hypoglycemia test (IHT). 0.1 - 0.2 units per kg body weight of soluble insulin were administered intravenously to ensure adequate hypoglycemia. Serum glucose, GH and PRL were then measured over a 90 minute period.
Key·words Cush,ng's d'sease. PRL. LH. FSH. TSH. GH
ii) Thyrotropin releasing hormone (TRH). Blood samples for TSH and PRL were taken over a 60 minute period following the iv administration of 200119 TRH.
Correspondence: Dr. I M. SPitZ. Department of Endocrrnology and Metabolism. Shaare Zedek Medical Center. PO. Box 293. Jerusalem 9100. Israel Received December 4, t 978: accepted May 30. t 979
379
0. Le ROith, M.S. Shapiro, A. Gutman, et al. Table 1 - Serum and urinary steroid values. Patient
ACTH pg/ml
1 2 3 4 5 6 7
Plasma cortisol Jlg/l00 ml 08:00
24:00
Baseline
125.4 . 800 700
50 17 21 15 20 26 18
40 18 45 20 26 20 25
930 2083 443 1010 480 1620 921
860 1006 390 902 400 1082 531
230 250 102 200 150 230 ;:>03
20-80
10-25
1-8
55-264
32-128
22-71
840 120.0 2008
Normal range
e
'-
~
c:
:t
... ...
Laboratory
Serum LH, FSH , TSH , PRL and GH were determined by previously described methods (8). Pituitary LH, FSH, TSH, PRL and GH and their respective antisera were kindly supplied by the National Pituitary Agency, National Institute of Arthritis, Metabolism and Digestive Diseases (NIAMDD) The second International Reference preparation for human menopausal gonadotrophins (2nd IRPHMG) provided by Dr. D. R. Bangham was used as ~ reference standard for both LH and FSH. TSH and PRL standards were also supplied by Dr. D. R. Bangham. Nati onal In.stltute lor Bl olo(1 lcal Stcwocll(ls. London, England.
u
, --
~
0 ~
Q:
Dexamethasone 2 mg 8 mg
iii) Luteinizing hormone releasing hormone (LHRH) . 100 pg LHRH was injected as a bolus and LH and FSH measured over a 60 minute period. iv) Chlorpromazine (CPZ). 25 mg was injected im. Halfhourly samples were taken for two hours for PRL. One female (Case 2) had the CPZ test repeated, following the administration of ethinyl estradiol 50 pg daily for 30 days. Seven male and five female controls aged 18 - 26 years were similarly challenged with LHRH, TRH and CPZ. The females were tested during the follicular phase of the menstrual cycle .
MALES
FEMALES
0
Urinary 11 -hydroxycorticosteroid J.I9/24 hours
I HT'
60 '0
Q..20
o
RESULTS
6 !
I M
t
Growth hormone Basal levels were below 1.3 ng / ml in all subjects. During IHT symptomatic hypoglycemia was obtained in all subjects. Despite adequate hypoglycemia (glucose nadir below 40 mg / 100 ml) there was consistent absence of GH response in each of the seven patients.
(~inJ
Fig. 1 - The prolactin response to (i) 200IJgTRH administered iv, ii) 25 mg chlorpromazine (CPl) , and iii) insulin induced hypoglycemia (IHT) in the seven patients with Cushing's disease. The response in each individual is numbered accordingly. Mean ± SE response to TRH and CPZ in controls (broken line) is also shown.
Prolactin The results are shown in Figure 1. Basal PRL levels
380
Hypothalamic-pituitary function in Cushing's disease
MALES
FEMALES 30
~/I
, I
,
o
,,'
I
I
I
I _-1 "" r--'
30
"
"
I
20
I
"0_--_2 __ ---4
10
o
~
___________ 7
Fig. 2 - TSH responses to 200 pg iv TRH in
2b
both male and female patients, numbered accordingly, and compared to controls (broken line, mean ± Sf).
were elevated ( > 20 ng/ ml) on at least one occasion in female subjects 2, 3 and 4. The remaining subjects had normal basal levels. When compared to the controls, six of the subjects demonstrated PRL elevation following TRH. Only female subject 3, with hyperprolactinemia, was unresponsive. Only male patient 6 showed a PRL response to CPZ. The other six subjects failed to
show PRL elevation. Case 2 remained unresponsive to CPZ despite ethinyl estradiol pretreatment. There was an attenuated response to insulin hypoglycemia in female subjects 2, 3 and 4 and male subject 5 (Fig. 1). In the remaining three subjects there was a trebling of their basal levels which is considered a normal response (9).
50
b
TIME (min)
FEMALES
Thyrotropin As shown in Figure 2, basal TSH levels ranged from 1.0 to 6.0 pU/mL When compared to control values there was an impaired TSH response to TRH in all four females and one male (Case 7).
MALES
Gonadotropins The results are shown in Figure 3. When compared to the controls, basal LH levels were low in female cases 1 and 4 and male case 5. Basal FSH levels were low in case 1. Case 6 had elevated basal FSH levels. Following the administration of LHRH female case 1 had LH and FSH uresponsiveness Female cases 2 and 4 had impaired LH responses. LH levels increased in case 3 and FSH in cases 2, 3 and 4. As regards the males, LH and FSH responses were intact with the exception of case 6, who had an exaggerated FSH rise.
/.'
DISCUSSION We have documented multiple abnormalities in pituitary hormone secretion in our seven cases with Cushing's disease. The most consistent defect noted in our study was the absent GH response to insulin induced hypoglycemia. This phenomenon has been described previously both in Cushing's disease (1) as well as in normal subjects given high dose glucocorticoids (5, 10). In the latter condition GH responsiveness returns after removal of the high dose glucocorticoid administration , whereas in Cushing's disease GH remains unresponsive to insulin hypoglycemia despite normalization of the cortisol levels after bilateral adrenalectomy (1). This suggests that in Cushing's disease there may
s , po-• I.. "-"~-A1 ~7
20
'0
60
0
TIM£(min)
20
~
60
Fig. 3 - LH and FSH responses to 100 pg iv LHRH in both male and female patients. Mean ± Sf of controls (broken line) IS also shown.
381
O. Le Roith, M.S. Shapiro, A. Gutman, et al. the glycoprotein hormones (TSH, LH and FSH), with recovery following normalization of cortisol levels after adrenalectomy. Evidence for an additional hypothalamic defect which is an integral part of the disease is seen with respect to GH and possibly PRL. This is shown by the absent GH and PRL response to hypoglycemia which persists despite correction of the hypercortisolism, the elevated basal PRL levels and loss of nocturnal PRL elevation. The lack of PRL response to CPl in our cases and the normalization of GH responsiveness to hypoglycemia following bromoergocryptine administration (3) suggests that these may be related to an alteration in dopamine metabolism, though a direct effect of excess glucocorticoid at the hypothalamic level cannot be excluded.
be a centrally based hypothalamic abnormality in GH secretion. Evidence that dopaminergic depletion may be implicated in this phenomenon was demonstrated by Lamberts et al. (3) who showed that GH responsiveness to insulin hypoglycemia was partially restored in untreated patients following the administration of the dopamine agonist bromoergocryptine. We have assessed PRL dynamics using three stimuli in each patient: i) TRH, which acts directly at the lactotroph; ii) insulin hypoglycemia, whose effect is probably at the hypothalamic level; iii) CPl, which induces PRL elevation by producing dopaminergic blockade at both hypothalamus and pituitary (8). The normal PRL response to TRH attests to the integrity of the lactotroph. However, the high basal levels in three of our patients, together with the absent PRL response to hypoglycemia in three, implies a defect in PRL control, which probably resides at the hypothalamic level or above. Supportive evidence is the loss of nocturnal elevation of PRL which has been described in Cushing's disease (11). The absent PRL response to CPl noted in our case is also in keeping with an abnormality of dopaminergic metabolism which has already been alluded to. It is possible that the defect in GH and PRL secretion noted in Cushing's disease may be intrinsic to the hypothalamic pituitary dysfunction and not a consequence of the hypercortisolism. However, it should be noted that high dose dexamethasone administration suppresses the PRL response to TRH which was intact in our subjects (6). Tha absent TSH response to TRH in our patients has been previously described in Cushing's disease (2, 3) as well as in normal subjects receiving long term high dose glucocorticoids (13), although Hershman and Pittman (14) demonstrated normal responses in their series of Cushing's disease. Since the abnormality is reversible after treatment of the Cushing's disease (2) it would seem that the unresponsiveness of TSH is due to the direct glucocorticoid effect at a pituitary level and not to an abnormality in neurotransmission. The absent gonadotropin response to LHRH has been described in Cushing's disease (4) as well as in normal subjects receiving glucocorticoid therapy (7). Many of our patients showed low basal gonadotropin levels and attenuated or absent responses to LHRH. Although some of our female patients did demonstrate a rise of LH and FSH after LHRH, three of these Cushing's patients were postmenopausal. Postmenopausal females normally have high basal gonadotropin levels and exaggerated responses to LHRH. Since the impaired LH and FSH response to LHRH has been shown to recover after adrenalectomy (15, 1 6), this suggests that the defect is related to the glucocorticoid action at a pituitary level. Thus the basis for the abnormalities in the secretion of anterior pituitary hormones in Cushing's disease is multifactorial. There is evidence for direct interference at pituitary level by the hypercortisolism with regard to
ACKNOWLEDGMENTS Hoffman La Roche kindly supplied the LHRH and TRH which was used This study was supported by grants from the Jewish Agency to D. LeA and the Chief Scientist's Office. Israel Ministry of Health to IMS We thank S Tristian, Z Shemesh and H. Gershman for their techmcal aid and H. Menachemson for typing the manuscript.
REFERENCES 1. Krieger O.T., Glick S.M. Growth hormone and cortisol responsiveness in Cushing's syndrome. Relation to a possible central nervous system etiology. Am. J, Med. 52: 25, 1972. 2. Kuku S.F., Child OF., Nader S, Fraser TR Thyrotropin and prolactin responsiveness to thyrotropin releasing hormone in Cushing's disease, Clin, Endocrino! (Oxt) 4: 437, 1975. 3. Lamberts S.w,J" Timmermans HAT., De Jong F.H .. Birkenhager J.C. The role ot dopaminergic depletion in the pathogeneSiS ot Cushing's disease and the possible consequences for medical therapy. Clin. Endocrino! (Oxf.) 7 185, 1977. 4. Boccuzzi G., Angeli A .. Bisbocci 0 .. Fonzo 0 .. Gaidano GP .. Ceresa F. Effect of synthetic LHRH on the release of gonadotropins in Cushing's disease. J. Clin. Endocrino! Metab. 40: 892. 1975. 5. Hartog M., Gaatar MA. Fraser R. Effect of corticosteroids on serum growth hormone, Lancet 2: 376, 1964. 6. Sowers JR.. Carlson H.E .. Brautbar N .. Hershman J.M. Effect of dexamethasone on prolactin and TSH response to TRH and metoclopramide in man. J. Clin. Endocrino!. Metab. 44: 237. 1977. 7. Sakakura M .. Kazuo T .. Nagasawa S. Inhibition of luteinizing hormone secretion induced by synthetic LHRH by long term treatment with glucocorticoids In human subjects. J. Clin. Endocrino! Metab, 40: 774. 1975. 8 Spitz 1M. Almaliach U .. Rosen E.. Polishuk W. Rabinowitz O.
382
Hypothalamic-pituitary function in Cushing's disease
Dissociation of prolactin responsiveness to TRH and chlorpromazine in women with isolated gonadotropin deficiency. J. Clin. EndocrinoL Metab. 45 1173. 1977.
Effects of glucocorticoids on pituitary hormonal responses to hypoglycemia. Inhibition of prolactin release. J. Clin. EndocrinoL Metab. 40. 442, 1975. 13. Otsukl M, Dakoda M. Baba S Influence of glucocorticoids on TRH-induced TSH response in man. J. Clin. EndocnnoL Metab. 36 95, 1973.
9. WoolH PO, Lee LA, Leebaw W., Thompson D. Lilavivathana U., Brodows R. Intracellular glucopenia causes prolactin release in man. J. Clin. EndocrinoL Metab. 45 377, 1977. 10. Frantz A.G, Rabkin M.T. Human growth hormone. Clinical measurement, response to hypoglycemia and suppression by corticosteroids. N. EngL J. Med. 271. 1375, 1964. 11. Krieger 01., Howanitz P.J., Frantz A.G. Absence of nocturnal elevation of plasma prolactin concentrations in Cushing's disease. J. Clin. EndocrinoL Metab. 42 260. 1976.
14. Hershman JM, Pittman JA Response to synthetic TRH in man. J. Clin. EndocrinoL Metab. 31. 457, 1970 15. Luton J.P. Thieblo P, Valcke JC. Mahoudeau JA. Bricaire H. Reversible gonadotropin deficiency in male Cushing's disease. J. Clin. EndocrinoL Metab. 45 488, 1977. 16. Hackenberg K, Hoff HG. Brouwers HP., Relnwein D. Gonadotropin stimulation test in female Cushing's disease before and after adrenalectomy. Acta. EndocrinoL (SuppL) 208 25. 1977.
12. Copinshi G, L'Hermile M, LeClerq R, Goldstein J, Vanhaelst L., Virasoro E, Robyn C
383
