LIVER PHYSIOLOGY AND DISEASE

Transcription

1 /78/ $02.00/0 GASTROENTEROLOGY 75:66-70, 1978 Copyright 1978 by the American Gastroenterological Association Vol. 75, No.1 Printed in U.s.A. LIVER PHYSIOLOGY AND DISEASE THYROTROPIN-RELEASING HORMONE (TRH)-INDUCED GROWTH HORMONE (hgh) RESPONSES IN CIRRHOTIC MEN DAVID H. VAN THIEL, M.D., JUDITH S. GAVALER, B.A., CARL WIGHT, B.S., WILLIAM 1. SMITH, JR., M.D., AND JUAN ABUID, M.D. Divisions of Gastroenterology and Endocrinology, Department of Medicine, and Department of Patlwlogy, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania The plasma growth hormone (hgh) responses to an intravenous challenge of 400 /-Lg of thyrotropin-releasing hormone (TRH) were evaluated in 14 normal controls and in 29 chronic alcoholic men. The normal controls had either a minimal or no hgh response to TRH, having basal hgh levels of 0.9 ± 0.2 ng per ml and peak hgh levels of 2.0 ± 0.5 ng per ml. In contrast, the chronic alcoholic men had a basal hgh level of 2.8 ± 0.4 ng per ml, 3 times the basal level of the normal controls (P < 0.01). The peak hgh response of the alcoholic men was 7.4 ± 1.5 ng per ml (P < 0.01). The 29 alcoholic men could be divided into two groups based upon the presence or absence of cirrhosis as determined by liver biopsy. The 16 alcoholic men with cirrhosis had greater basal hgh levels (3.5 ± 0.6 ng per ml) and peak hgh levels (9.5 ± 2.3 ng per ml) than did the 13 alcoholic men without cirrhosis (basal hgh 2.1 ± 0.6 ng per ml, peak hgh 4.9 ± 1.5 ng/ ml). Plasma estradiol levels were similar in the normal controls and in the alcoholic men. In contrast, plasma estrone was greater in the alcoholic men (32.2 ± 3.5 pg per ml) than in the normal controls (18.9 ± 1.8 pg per ml) (P < 0.05). However, when the plasma estrone levels of alcoholic men with cirrhosis were compared to those of the alcoholic men without cirrhosis no difference existed. Thus it is difficult to ascribe the increased hgh responses of the cirrhotic alcoholic men when compared to those of the noncirrhotic alcoholic men as being a result of increased basal estrogen levels. Chronic alcoholic men frequently show signs of gonadal failure, 1,2 Considerable evidence has accumulated recently to suggest that, in addition, hypothalamicpituitary dysfunction also exists in these men. &-5 Specifically, reduced gonadotropin responses to clomiphene and luteinizing hormone-releasing factor stimulation and abnormal prolactin (hprl) responses to thyrotropinreleasing hormone (TRH) have been reported. 4, 5 It is of some interest that the hprl responses in some of the alcoholic men studied have been similar to those reported to occur in patients with pituitary adenoma. 5-7 Because disturbances of hprl secretion frequently coexist with abnormalities of growth hormone (hgh) secretion,8, 9 particularly in the case of pituitary adenoma, we have evaluated the hgh responses to provocative TRH stimulation in 29 chronic alcoholic men and have compared them to those of 14 normal volunteers. Received November 21, Accepted February 8, Address requests for reprints to: David H. Van Thiel, M.D., 1000G Scaife Hall, Department of Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania This study was supported in part by Grant AA01450 from the National Institutes of Health. Dr. Van Thiel is the recipient of United States Public Health Service Career Development Award AA Methods Subjects. Twenty-nine chronic alcoholic men ages 42.3 ± 1. 7 years (range 25 to 63) served as experimental subjects for this study. As a group, they admitted to chronic alcohol abuse for 21.0 ± 1.7 years (range 4 to 43). Each had been abstinent for a minimum period of 3 weeks before participation in this study. None were taking or had been taking any medication known to or suspected of affecting hgh secretion for at least 3 months before their participation in this study. A recent liver biopsy was available for each. Mter a layman's description of the study, which included all of the technical details and potential hazards of participation in the study was given, a written consent was obtained from each. Moreover, before the initiation of any studies, approval was obtained from the Committee on Use of Humans for Experimental Purposes of the University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania. The normal controls consisted of 14 adult men ages 28 ± 3 years who agreed to participate in the study as normal volunteers. As in the case of the experimental subjects, each was not taking any medication known to or suspected of affecting hgh secretion, including ethanol, for an interval of at least 3 months before study. Again written informed consent was obtained and approval for use of these volunteers as normal subjects was obtained from the Committee on Use of Humans for Experimental Purposes of the University of Pittsburgh School of Medicine.

2 July 1978 TRH-INDUCED hgh RESPONSES IN CIRRHOSIS 67 Hormone assays. All samples for a given hormone were measured in duplicate in a single assay, thus eliminating interassay variation. Growth hormone was measured by a specific radioimmunoassay supplied by Kallestad Laboratories Inc. (Chaska, Minn.). The detection limit for the assay was 0.4 ng; intraassay variation of known laboratory standards was less than 8%. Estradiol was measured by a specific radioimmunoassay. 10 The detection limit for the assay was 0.2 pg; intraassay variation was less than 6%. Estrone was measured by a specific radioimmunoassay. 1 1 The detection limit for the assay was 1.0 pg; intraassay variation was less than 8%. Experimental design. All subjects and controls were studied after an overnight fast. An indwelling intravenous catheter was inserted in a forearm vein at 0830 hr and a slow infusion of 0.9% saline was initiated. At 0845, 0850, 0855, and 0900 hr blood was obtained for basal hgh, estradiol, and estrone determinations. Four hundred micrograms of TRH (Abbott Laboratories, North Chicago, Ill.) were then rapidly infused as an intravenous bolus injection and additional samples of blood were obtained at 0915, 0930, 0945, 1000, 1030, 1100, and 1200 hr for hgh determinations. Statistical analysis. The basal hgh level was defined as the mean value of the four samples obtained before the TRH infusion. The peak hgh response was defined as the maximum hgh level measured after TRH infusion regardless of the time at which the sample was obtained. Delta hgh was defined as the difference between the peak and basal hgh levels. Analysis of variance was used for statistical analysis, except when X2 analysis was required. 12 The Yates' correction for continuity was used when the X2 analysis was required. 12 AP value of <0.05 was considered probably significant, and a P value of <0.01 was considered signifcant. Results The clinical and biochemical data obtained on the 29 alcoholic men studied are shown in tables 1 and 2. Plasma estradiol levels in the 14 normal controls and 29 alcoholics were similar (17.6 ± 0.3 and 17.8 ± 0.4 pg per ml, respectively). In contrast, plasma estrone levels in the 29 alcoholic men (32.3 ± 3.6 pg per ml) were greater than those of the normal controls (18.9 ± 1.8 pg per ml) (P < 0.05). When the plasma levels of estrone in the alcoholic men with cirrhosis (39.7 ± 4.1 pg per ml) were compared to those in men without cirrhosis (26.6 ± 5.6 pg per ml), the cirrhotics were found to have arithmetically greater estrone levels, although the two groups did not differ significantly. Basal hgh levels in the alcoholic men studied (2.8 ± 0.4 ng per ml) were 3-fold greater than those of the normal controls (0.9 ± 0.2 ng per m}) (P < 0.01) (fig. 1). As was the case with plasma estrone, the basal hgh levels in the alcoholic men with cirrhosis (3.5 ± 0.6 ng per ml) were greater than those of the alcoholic men without cirrhosis (2.1 ± 0.6 ng per mt) (P < 0.05). On the average, in the normal controls, the peak hgh in response to TRH was twice the basal level (fig. 1). In contrast, the 29 alcoholics as a group achieved a peak hgh response of 7.4 ± 1.5 ng per ml, 3 times that of the controls (P < 0.01). Analysis of the individual responses of the 29 alcoholic men studied revealed that 15 men had abnormal peak responses of 5.0 ng per ml or greater and had a mean delta response of 8.2 ± 2.0 ng per ml (table 2), but 14 had flat responses identical to those of the 14 normal controls (delta response 0.9 ± 0.2 pg per m}) (fig. 1, table 1). Interestingly, 11 of the 15 alcoholic men having an abnormal hgh response to TRH had cirrhosis on histological examination of their liver biopsy. In contrast, only 4 of the 14 men who failed to have abnormal hgh responses had cirrhosis (tables 1 and 2) (P < 0.05). No significant difference in biochemical liver function was observed between the two groups as determined by measures of serum bilirubin, alkaline phosphatase, and TABLE 1. Clinical and laboratory data on alcoholic men with normal responses to TRHa Subject no. Age Alcohol abuse Liver histology' Estradiol Estrone BasalhGH PeakhGH yr yr pg/ml pg/ml ng/ml ng/ml Fatty Fatty Alcoholic hepatitis and cirrhosis Cirrhosis Alcoholic hepatitis and cirrhosis Fatty Fatty Fatty Fatty Alcoholic hepatitis and cirrhosis Fatty Fatty Fatty Alcoholic hepatitis Mean SEM Normal range ( a Abbreviations are: TRH, thyrotropin-releasing hormone; hgh, growth hormone. b Histological interpretation by the Surgical Pathology Department of the Department of Pathology, University of Pittsburgh School of Medicine.

3 68 v AN THIEL ET AL. Vol. 75, No.1 TABLE 2. Clinical and laboratory data on alcoholic men with increased hgh responses to TRHa Subject no. Age Alcohol abuse Liver histology" Estradiol Estrone BasalhGH PeakhGH yr yr pg/ml pg/ml ng/ml ng/ml Fatty Alcoholic hepatitis Alcoholic hepatitis and cirrhosis Fatty Cirrhosis Fatty Cirrhosis Cirrhosis Cirrhosis Cirrhosis Cirrhosis Cirrhosis Alcoholic hepatitis and cirrhosis Cirrhosis Cirrhosis Mean SEM Normal range a hgh, growth hormone; TRH, thyrotropin-releasing hormone. b Histological interpretation by the Surgical Pathology Department of the Department of Pathology, University of Pittsburgh School of Medicine ~ II co ~I w 9 ~ ~ 8 a: 0 :x: 7 :x: 6 I- ~ 5 0 a: <!) NORMAL ALL ALCOHOLICS CONTROLS n -29 n:: 14 o BASAL Ill:! PEAK ALCOHOLICS PEAK<5.0 ng/ml n=14 ALCOHOLICS PEAK ~ 5.0 og/ml n=15 FIG. 1. Basal and peak plasma growth hormone (hgh) level in response to thyrotropin-releasing hormone (TRH) in the 14 normal controls and the 29 chronic alcoholic men. The ordinate shows hgh levels in nanograms per milliliter. Open bars, basal hgh levels; cross-hatched bars, peak hgh levels. Brackets represent mean ± SEM for each group. transaminase levels. As might be expected, clinical evidence for endocrine dysfunction as determined by the presence of prostatic atrophy (P < 0.05), spider angiomata (P < 0.01), and gynecomastia (P < 0.01) was more often present in the alcoholic men with abnormal hgh responses than in those with normal responses. Testicular atrophy was present in 6 of 14 men with normal hgh responses and in 10 of 15 with abnormal responses (P > 0.05). Similarly, the frequency of palmar erythema and the presence of a female escutcheon was not different significantly (4 of 14 normal responders and 8 of 15 abnormal responders for both physical signs). Discussion Chronic alcoholic men frequently have gonadal failure characterized by testicular atrophy, reduced or absent spermatogenesis, and reduced plasma levels of testosterone. 1, 2 In addition, they are frequently feminized.1, 13 This is particularly true of those who have established Laennec's cirrhosis. Recent studies have shown that in addition to primary gonadal dysfunction, a second central defect(s) in hypothalamic-pituitary function can be demonstrated in chronic alcoholic men.3-5 This study extends previous studies which have evaluated hypothalamic-pituitary function in chronic alcoholic men by documenting a high incidence of abnormal hgh responses to provocative TRH stimulation in chronic alcoholic men with cirrhosis. Similar abnormal hgh responses to TRH stimulation have been reported to occur in active acromegaly, 14, 15 but are quite unusual in persons without pituitary adenoma. It is of some interest that the alcoholic men having the abnormal hgh response to TRH had elevated basal hgh levels and were more likely to have cirrhosis than those who did not (figs. 1 and 2, tables 1 and 2). Previous studies evaluating hypothalamic-pituitary function in alcoholics have not shown such an association between the degree of histological liver disease and any hormonal abnormality.:t-5 It is well known that women have higher basal levels of hgh than do men. 8, 9 Moreover, estrogen pretreatment is known to sensitize the pituitary to the commonly used hgh stimuli. 8,9 It is therefore noteworthy that the 15 alcoholic men, who had the higher basal and peak hgh responses to TRH, were clinically more feminized, having a statistical greater incidence of spider angiomata (P < 0.01), than were those who had normal hgh responses. Because plasma estradiol levels were similar in the alcoholic and normal control men we studied, it seems

4 July 1978 TRH-INDUCED hgh RESPONSES IN CIRRHOSIS E 12 " ḡo" w 10 z ~ 9 gs 8 I I 7 ~ 6 ~ 5 (,!) <t 4 ~ 3 <t ~ PEAK RESPONSE 2 50ng/ml o PEAK RESPONSE < 5.0ng/ml o MINUTES FIG. 2. Growth hormone (hgh) response to thyrotropin-releasing hormone (TRH) in 29 alcoholic men. The ordinate shows hgh in terms of nanograms per milliliter. The abscissa shows time in minutes. 0-0, response of the 14 alcoholic men who failed to have a peak response >5ng/ml; A-A, response of the 15 alcoholic men who had a peak response >5ng/ml. Brackets represent mean ± SEM for each point. unlikely that this estrogen is responsible for the increased basal hgh and peak hgh responses to TRH we observed in the alcoholics. In contrast, plasma estrone levels were increased in the alcoholic men compared to those of the normal controls (P < 0.05). Moreover, the alcoholic men with cirrhosis had higher levels of estrone than did the alcoholic men without cirrhosis. Whether this relatively small increase in plasma estrone is responsible for the abnormalities of hgh secretion we report in cirrhotic men cannot be assessed fully by the present study. Considering the relatively minor increase in estrone (less than a doubling) between the cirrhotic men and the noncirrhotic alcoholic men, it seems unlikely that the increase in plasma estrone alone is responsible for the abnormal hgh responses observed in the cirrhotics. This is particularly apparent when the difference between estrone levels of the noncirrhotic alcoholic men and normal controls are compared. The noncirrhotic men had a mean plasma estrone level of 26.6 ± 5.6 compared to 18.9 ± 1.8 pg per ml for the 14 normal controls (P < 0.05). Despite this difference in estrone levels, the hgh responses to TRH were similar for the two groups. It is possible that the differences we have observed between the two groups of alcoholic men studied could have been distinguished on the basis of difference in nutritional status. Because of the difficulty in evaluating such factors clinically, we are unable to provide any information relative to such a possible mechanism. Another possible factor which might account for the difference we have observed is the level of circulating thyroid hormones. Although not reported in this manuscript, we have measured both serum T3 and T4 levels in each of these men and are unable to relate the differences in hgh responses to any change in thyroid hormone. Persons with Laennec's cirrhosis, in addition to being feminized, are known to have an increased incidence of diabetes mellitus. 1 &-18 Unlike most cases of maturityonset diabetes mellitus, the glucose intolerant cirrhotic is characterized by hyperinsulinism, peripheral insulin resistance, and hyperglucagonemia,ur-23 Moreover, cirrhotic men are known to have basal hyperprolactinemia Somatostatin is a known suppressor of all four of these hormones (hgh, insulin, glucagon, and hprl). 24 Therefore, it is not unreasonable at least to consider the possibility that a single central isolated defect in somatostatin secretion rather than four separate defects in hgh, glucagon, insulin, and hpri secretion might be responsible for the disturbed regulation of these hormones and the increased incidence of diabetes mellitus in cirrhotics. Such a formulation seems intuitively more likely than the simultaneous development of four separate defects in two anatomically separated organs, the pancreas and the pituitary. Evaluation of both of these hypotheses, however, is beyond the scope of our present studies and awaits further investigation. Regardless of the various mechanisms responsible, based upon our past experience! and our present data, we conclude that: (1) hypothalamic-pituitary dysfunction is common in alcoholic men; (2) in contrast to the abnormalities of prolactin and gonadotropin secretion which are seen in alcoholic men in general, disturbed hgh secretion is seen primarily in alcoholic men with established cirrhosis; and (3) disturbed hgh secretion in alcoholic men with cirrhosis may explain, at least in part, the increased incidence of diabetes mellitus in such men. REFERENCES 1. Van Thiel DH, Lester R: Alcoholism: its effect on hypothalamicpituitary-gonadal function. Gastroenterology 71: , Baker WG, Burger HG, DeKretser DM, et al: A study of the endocrine manifestations of hepatic cirrhosis. Q J Med 45: , Van Thiel DH, Lester R, Sherins RJ: Hypogonadism in alcoholic liver disease: evidence for a double defect. Gastroenterology 67: , Van Thiel DH, Gavaler JS, Vaitukaitis J, et al: Evidence for an isolated defect in pituitary secretion of LH in chronic alcoholic men (abstr). Gastroenterology 71:933, Van Thiel DH, McClain CJ, Elson MK, et al: Evidence for autonomous secretion of prolactin in some alcoholic men with cirrhosis and gynecomastia (abstr). Gastroenterology 73:1251, Kleinberg DL, Noel GL, Frantz AG: Galactorrhea: a study of235 cases including 98 with pituitary tumors. N Engl J Med 296: , MartinJB, Reichlin S, Brown GM: Clinical Neuroendocrinology. First edition. Philadelphia, FA Davis, 1977, p Daugheday WH: The adenohypophysis. In Textbook of Endocrinology. Edited by RH Williams. Philadelphia, WB Saunders Co, 1974, p MartinJB, Reichlin S, Brown GM: Clinical Neuroendocrinology. First edition. Philadelphia, FA Davis, 1977, p Hotchkiss JL, Atkinson E, Knobil E: Time course of serum estrogen and luteinizing hormone (LH) concentrations during

5 70 V AN THIEL ET AL. Vol. 75, No.1 the menstrual cycle of the rhesus monkey. Endocrinology 89: , Loriaux DL, Ruder HJ, Lipsitt MB: The measurement of estrone sulfate in plasma. Steroids 18: , Freeman LC: Elementary Applied Statistics. First edition. New York, Wiley & Sons Inc, Van Thiel DH, Gavaler JS, Lester R, et al: Plasma estrone, prolactin, neurophysin and sex steroid binding globulin in chronic alcoholic men. Metabolism 24: , Irie M, Tsushima T: Increase of serum growth hormone concentration following thyrotropin releasing-hormone injection in patients with acromegaly or gigantism. J Clin Endocrinol Metab 35:97-100, Faglia G, Beck-Peccoz P, Ferrari C, et al: Plasma growth hormone response to thyrotropin-releasing hormone in patients with active acromegaly. J Clin Endocrinol Metab 36: , Liuzzi A, Chiodini PG, Botalla L, et al: Growth hormone (GH) releasing activity of TRH and GH-Iowering effect of dopaminergic drugs in acromegaly: homogeneity in the two responses. J Clin Endocrinol Metab 39: , Muting D, Lackas N, Reikowski H: Cirrhosis of the liver and diabetes mellitus: a study of 140 combined cases. Dtsch Med Wochenschr 91: , Conn HO, Schreiber W, Elkington SG, et al: Cirrhosis and diabetes. I. Increased incidence of diabetes in patients with Laennec's cirrhosis. Am J Dig Dis 14: , Conn HO, Schreiber W, Elkington SG: Cirrhosis and diabetes. II. Association of impaired glucose tolerance with portal-systemic shunting in Laennec's cirrhosis. Am J Dig Dis 16: , Conn HO: Cirrhosis. In Diseases of the Liver. Edited by L Schiff. Philadelphia, JB Lippincott Co, 1976, p Sherwin R, Joshi P, Hendler R, et al: Hyperglucagonemia in Laennec's cirrhosis. N Engl J Med 290: , Megyesi C, Samolos E, Marks V: Glucose tolerance and diabetes in chronic liver disease. Lancet 2: , Collins, JR, Crofford OB: Glucose intolerance and insulin resistance in patients with liver disease. Arch Intern Med 124: , Reichlin S, Sapestein R, Jackson IMD: Hypothalamic hormones. In Annual Review of Physiology, vol 38. Edited by E Knobil, RR Sonnenschein, IS Edelman. Palo Alto, Annual Reviews Inc, 1976, p