Endocrinol Japon 1992, 39 (3), 251-257

A Child with Pituitary Gigantism and Precocious Adrenarche: Does GH and/or PRL Advance the Onset of Adrenarche?

NORITAKA IWATANI, MIHOKO KODAMA, AND HIROSHI SETO* Departments of Child Development and *Neurosurgery, Kumamoto University Medical School,Kumamoto 860, Japan

Abstract. We describe a female child with pituitary gigantism and precocious adrenarche. From two years of age she showed unusual overgrowth, and at 5 years old she was 133.5 cm (+ 5.5 SD) tall and weighed 40.5 kg. Her precocious manifestations were pubic hair, acne vulgaris, , and advanced age. Endocrinological examination revealed markedly increased serum (GH) and prolactin (PRL), which responded paradoxically to a TRH test. In addition, the concentrations of serum dehydroepiandrosterone (DHA) and its sulfate (DHAS) were increased to adult levels, moving in accordance with changes in ACTH, which suggested that these androgens were secreted from the adrenal glands functionally. These androgens seemed to be responsible for her partial precocity. Prior reports have suggested that GH and/or PRL overproduction might have played a role in the induction of adrenarche. Also, in previous reports of 9 gigantism patients under 10 years old, the manifestation of precocious adrenarche was suggested in 8. Further investigation of the influence of GH and PRL on adrenal androgen production in children with pituitary gigantism is required. On the other hand, in short children with normal GH secretion, attention should be paid to whether or not the GH therapy in early childhood induces precocious adrenarche.

Key words: Pituitary gigantism, Precocious adrenarche, GH, PRL, Dehydroepiandrosterone (DHA), DHA sulfate (DHAS). (Endocrinol Japon 39: 251-257, 1992)

LITTLE IS known about the mechanism of the hood; however, it also has various other biological activation of adrenal androgen secretion in child- effects in vivo. Again, the physiological significance hood, referred to as adrenarche. Although the of PRL is not fully known. Although gigantism production of these androgens is stimulated by due to pituitary in childhood is a rare ACTH, ACTH may not be the sole factor regulat- disorder, the patient provides much significant ing adrenarche. Several investigators have sug- information for elucidating the biological effects gested that factors other than ACTH are involved of GH and/or PRL in childhood. In this report, we in the control of adrenal androgen production. describe in detail the results of endocrinological However, the nature of these agents remains a studies in this gigantism case. Further, we review matter of controversy. previous reports of pituitary gigantism children The gigantism child presented here, with pituit- [1-8], and discuss the still unknown mechanism of ary adenoma secreting both GH and PRL, was adrenarchal onset. associated with precocious adrenarche. GH is a major hormone promoting linear growth in child-

Received: December 11, 1991 Subject and Methods Accepted: April 17, 1992 Correspondence to: Dr. Noritaka IWATANI, Department of Subject (Case Report) Child Development, Kumamoto University Medical School, 1-1-1 Honjo, Kumamoto 860, Japan. The subject was a 12.6-year-old girl who was 252 IWATANI et al. born after 35 gestational weeks of an uncompli- needed at the ages of 6.2 and 9.2 years to remove cated pregnancy with a birth weight and length of the residual tumor. Bromocriptine administration 2650 g and 46 cm, respectively. Her parents was started after the second operation. In spite of became aware of her accelerated physical growth the administration of large doses of bromocrip- at around 2 years of age. At 3.5 years old her tine, serum GH and PRL continued to be abnor- height was 111.0 cm (+4.0 SD) and weight 28 kg mally high (Table 1). After the third operation (obesity rate according to height, 49%). She was basal GH decreased to a level slightly higher than first admitted to our hospital at 5.4 years old normal, while IGF-I and PRL remained abnormal- because of excessive growth. Her physical man- ly high (Table 1). She continued to grow excessive- ifestations at that time were: height 133.5 cm (+ ly tall and heavy. 5.5 SD); weight 40.5 kg (obesity rate according to At the age of 12.6 years, she reached 176.8 cm in height 36%); a large head (circumference 56.2 height (+ 4.2 SD) and 86.0 kg in weight, and had cm); a prominent forehead and mandible; a wide long extremities (upper to lower segment ratio, nasal bridge; thick lips; disproportionately large 0.78). Her pubertal signs corresponded to Tanner hands and feet (palm-middle finger length 16.0 grade III for breast and IV for pubic hair, but no cm, and foot length 24.5 cm); and acanthosis menstruation or galactorrhea was observed. Her nigricans of the axillae. Her heel pad thickness was bone age was 13.5 years old at the time. 27 mm. Other findings at the time were sparse pubic hair (Tanner grade II) without other exter- Methods nal genital abnormalities, acne on the forehead, hirsutism on the legs, and advanced bone age (7.4 To evaluate the pituitary function, provocative years old); no breast budding or galactorrhea was tests were performed during fasting with in- observed. CT scanning of the head revealed an travenous administration of 0.1 U of insulin per intrasellar tumor extending to the suprasella. kilogram of body weight, 500 ,,cg of TRH and 100 At 5.5 years old, she underwent removal of the μg of LH-RH. Blood samples were drawn before tumor under the diagnosis of a GH- and PRL- and 15, 30, 60, 90 and 120 min after injection. An secreting adenoma. The pathological diagnosis oral glucose tolerance test (OGTT) was performed was an eosinophilic and chromophobic mixed during fasting with 1.75 g of glucose per kilogram adenoma. The pituitary tumor, however, was not of body weight or 75 g as the maximum dose. resected completely; subsequently 30 Gy irradia- Blood samples were drawn before and every 30 tion and two more surgical operations were min after glucose was given, for 180 min. A single

Table 1. Therapy and baseline hormone levels

GH and PRL values indicate means•}1SD (n=number of measured samples). Dose of bromocriptine was increased from 2.5 to 35 mg/day successively until the third surgery. a, range in adulthood; 1), range in childhood. PITUITARY GIGANTISM AND ADRENARCHE 253

dose of 2.5 mg of bromocriptine was administered from 208 to 2513 ng/ml, and PRL increased from orally to assess its influence on serum GH and 1052 to 1640 ng/ml. A single oral dose of 2.5 mg PRL, and 7.5 mg of bromocriptine was adminis- bromocriptine resulted in an apparent reduction tered for 5 days to assess its influence on in the serum levels of both GH and PRL (from 207 androgens. Dexamethasone was administered in 2, and 858 ng/ml to 39 and 182 ng/ml, respectively). 4, and 8 mg doses for 2 days each to achieve Serum GH was not suppressed after oral adminis- sufficient suppression of ACTH. An ACTH test tration of glucose. The ACTH-cortisol axis re- was performed with 250 eug of ACTH-(1-24) sponded normally to insulin-induced hypo- administered intravenously. To assess the changes glycemia (peak cortisol, 15.3 ƒÊg/dl at 15 min). in the serum baselines of GH and PRL, serum Other noticeable findings were increased serum samples were obtained at random. Every blood DHA and DHAS concentrations, already at adult hormone was measured by means of a specific levels (Table 1), and increased urinary 17- RIA: serum GH was measured by a double- ketosteroids (17 KS) and urinary etiocholanolone, antibody technique (HGH-I kit, Eiken), serum one of the metabolites of these androgens. To PRL by an IRMA using monoclonal antibody (Ab determine the origin and nature of these

Bead PRL kit, Eiken), serum LH and FSH by a androgens, an ACTH test, a dexamethasone sup- double-antibody technique (HLH and HFSH kits, pression test, bromocriptine loading, and adrenal Eiken, respectively), serum cortisol by a solid scintigraphy were performed at age 6. In response phase method (Amerlex-Cortisol kit, Amersham), to the ACTH and dexamethasone suppression TSH by an IRMA with monoclonal antibody tests, DHA, DHAS, and etiocholanolone seemed

(Amerwell-TSH kit, Amersham), and serum in- to move in accordance with changes in ACTH sulin by a solid phase method (Ab Bead Insulin kit, (Figs. 1 and 2), and the adrenal scintillation Eiken). Serum IGF-I was measured by the pre- examination showed radioisotope accumulation in viously reported method with an acid-ethanol identical positions in bilateral adrenal glands with extracted sample [9] (SM-C kit, Nichols Institute a normal image and intensity. These findings Diagnostics), and recombinant IGF-I (Fujisawa indicated the origin of the androgens to be the

Pharmaceuticals, Japan) was used as the kit's adrenal glands. standard reference. Estradiol, DHA, DHAS, and In a loading test with 7.5 mg bromocriptine for etiocholanolone were measured at the Special 5 days, the serum GH and PRL concentrations Reference Laboratory (Tokyo, Japan). Serum remained at abnormally high levels even at the 5th samples were stored at -20•Ž until use. Adrenal day (GH: from 82 to 114 ng/ml, PRL: from 519 to scintigraphy was performed with 131I-adosterol. 117 ng/m/), and the serum level of DHAS and the Bone age was estimated by the Tanner- urinary excretion of etiocholanolone were un-

Whitehouse II method. affected (DHAS: from 1072 to 1160 ng/ml, etiocholanolone: from 2.4 to 2.1 mg/day). The LH and FSH responses to an LH-RH test

Results performed at age 5.4 years showed prepubertal increases (Table 2A). The serum estradiol concen-

During her first and second admissions to our hospital at 5.4 and 6.0 years of age, detailed endocrinological evaluations were performed. Complete blood count, routine blood chemistry, and thyroid hormone levels were normal, except for increased alkaline-phosphatase. The baselines of GH and PRL concentrations in her serum were markedly increased and the serum IGF-I level was also very high (Table 1). In pituitary function studies, serum GH and PRL did not respond to either insulin-induced hypo- Fig.1. Adrenal androgen response to ACTH test. Tested glycemia or to an LH-RH test. In response to at age 6. 250 ,ug of ACTH-(1-24) was administered TRH, however, the GH level rose paradoxically intravenously. 254 IWATANI et al.

Discussion

The adrenarche, an increase in adrenal androgens, does not occur in early childhood [10-12]; however, serum DHA and DHAS of this patient had already increased to adult levels at 6 years of age. By means of ACTH and dexametha- sone suppression tests performed at 6 years of age, we confirmed that the origin of the androgens, DHA and DHAS, was the adrenal glands, because the movement of these steroids was dependent on changes in ACTH (Figs. 1 and 2). Furthermore, the results of adrenal scintigraphy also indicated that the origin of these androgens was the Fig. 2. Dexamethasone suppression test. Tested at age 6. adrenals. Dexamethasone was administered orally in doses of Although DHAS is the most abundant steroid 2, 4, and 8 mg for 2 days each. Serum cortisol, DHAS, and urinary etiocholanolone were mea- circulating in the plasma of normal human adults, sured for 7 days. its biological role is not fully known. Some reports have suggested that DHAS may possess weak Table 2. Pituitary function studies androgenic and estrogenic activities [13], and that the adrenarche may be related to the pubic hair growth [14, 15]. Our patient exhibited the follow- ing noticeable manifestations at 5 and 6 years of age: pubic hair, acne vulgaris, hirsutism of the legs, and advanced bone age, which can be considered a reflection of androgenic and/or estrogenic effects. It seems unlikely that these findings resulted from the effects of pituitary- gonadal axis activation, because gonadotropin Examination A was performed at the age of 5.4 years, response to an LH-RH test remained at a pre- and B and C at 12.6 years. Values in parentheses indicate pubertal level and the serum estradiol concentra- peak times in minutes. tion also remained low (Tables 1 and 2A). There- fore, increased adrenal androgens should be re- tration also remained at a prepubertal level (Table sponsible for her partial precocity and bone age 1). acceleration. At 12.6 years of age, we evaluated her pituitary The onset mechanism of adrenarche remains function again. Although basal GH had decreased uncertain. The existence of a pituitary adrenal to a level slightly higher than normal, the serum androgen-stimulating hormone (AASH), which GH still showed a paradoxical response to TRH acts in the presence of ACTH to stimulate and administration (Table 2C). Furthermore, in an control the secretion of adrenal androgens, has OGTT, the GH level was not suppressed signi- been postulated [16-18]. On the other hand, some ficantly, but maintained a slightly higher than researchers do not agree that there is such a factor normal level. LH and FSH responded to an and have suggested that adrenarche results from a LH-RH test better than previously (Table 2B), but shift in the pathway of steroid synthesis in re- the pituitary-gonadal function did not develop to sponse to the ACTH stimulation that follows the the pubertal level because the serum estradiol level development of the zona reticularis [19]. However, remained below 10 pg/ml (Table 1). it is still not known what factor develops the zona reticularis and causes adrenarche. Does pituitary gigantism associated with preco- cious adrenarche, as observed in our patient, suggest a relationship between GH and/or PRL PITUITARY GIGANTISM AND ADRENARCHE 255 and adrenarchal onset? Pituitary gigantism in This possibility is suggested by the following early childhood is exceedingly rare, and most evidence. Carter et al. [20] and Schiebinger et al. patients show overproduction of both GH and [21] suggested that hyperprolactinemia was re- PRL. In our review of the literature, we could find sponsible for the stimulation of adrenal androgen as case reports only nine cases under 10 years old production. They reported that bromocriptine who were not associated with McCune-Albright induced significant decreases in the serum PRL, syndrome, and eight of the nine reports included a DHA, and DHAS concentrations in hyperprolac- description of pubertal signs and evaluations of tinemic patients. The following observations also gonadotropins or steroid hormones [1-8]. Eight suggest the influence of PRL on adrenal function: out of nine cases, including our patient, exhibited adrenal tissue contains PRL receptors [22, 23], and pubic hair or increased urinary 17-KS without cultured human adrenal cortical cells exposed to other pubertal signs or any gonadotropin or sex PRL secrete more DHAS than do control cells [24]. hormone increase (Table 3). Unfortunately, ex- GH may also stimulate the growth of the adrenal cept for our patient, none of their DHA or DHAS glands. Skottner et al. [25] observed augmentation was measured, so there was no direct evidence of of adrenal weight following their treatment of adrenarche. However, the findings of pubic hair experimental rats with GH. Pillion et al. [26] and and urinary 17-KS do not seem to be due to Shigematsu et al. [27] reported that IGF-I receptor gonadal activation, and therefore suggest in- is present in the human adrenal cortex, especially creased adrenal androgens. Hence, these clinical in the zona reticularis [26]. Further, in isolated GH findings associated with pituitary gigantism indi- deficient children, adrenarche does occur; howev- cate a possible relationship between GH and/or er, a lower serum DHAS was observed than in the PRL and the induction of adrenarche. control group [28]. These results indicate that PRL

Table 3. Pituitary gigantism under 10 years of age: review of reported cases*

*One case was omitted since it lacked a pubertal evaluation. Eight out of 9 cases exhibited pubic hair or increased urinary 17-KS without other pubertal signs or any gonadotropin or sex hormone increase. u. 17 KS, urinary 17-ketosteroids; E2, estradiol; ND, not described; Ref. no., reference number. a,b, Normal ranges in childhood before adrenarche are below 1-2 mg/day and 3.5 ,u,mol/day, respectively. 256 IWATANI et al. and also GH, probably mediated by IGF-I, may contribute to the natural onset of adrenarche. contribute to the regulation of adrenal growth and However, the fact that child gigantism with GH- function, and as a consequence play a role in and PRL-producing adenoma was associated with adrenarchal onset. precocious adrenarche suggests that GH and/or In our patient, to determine the influence of GH PRL should be considered a possible causative and PRL on adrenal androgen secretion, bromoc- factor(s) in adrenarchal onset. Our findings also riptine was administered for 5 days. It failed, invite attention to the GH treatment of short however, to suppress the serum DHAS level or children. The recent unlimited supply of GH etiocholanolone excretion in the urine probably preparation derived from recombinant DNA tech- because serum GH and PRL concentrations were nology makes possible studies to assess the efficacy not reduced enough to influence DHAS secretion. of exogenous GH for height gain in non-GH- As shown in Table 1, despite the reduction in the deficient short children [29-31], but such a treat- GH baseline level after the third operation, the ment may induce precocious adrenarche. Further blood IGF-I and also the PRL concentrations were investigation of the influence of GH and PRL on still abnormally high. These findings indicate that adrenal androgen production in children is re- the residual was still active quired. (Table 2C), and this must be the cause of still high DHA and DHAS levels. The mechanism of precocious adrenarche Acknowledgments observed in the gigantism child appears complex. ACTH seems to be necessary to control adrenal We thank Dr. Teruhisa Miike, professor of the androgen production. Further, the possibility of Department of Child Development, Kumamoto the participation of GH or PRL, or the synergic University Medical School, for review of the action of both, should be considered. Again, it is manuscript and helpful discussion. unclear whether or not GH/IGF-I and/or PRL

References 1. Todd RM (1958) in a girl of 8 years. Management of pituitary gigantism; the role of Arch Dis Child 33: 49-54. bromocriptine and radiotherapy. Acta Pediatr Scand 2. Hurxthal LM (1961) Pituitary gigantism in a child 74: 807-814. five years of age: effect of X-radiation, estrogen 9. Suwa S, Katsumata N, Maesaka H, Tokuhiro E, therapy and self-imposed starvation diet during an Yokoya S (1988) Serum insulin-like growth factor I eleven-year period. J Clin Endocrinol Metab 21: (somatomedin-C) level in normal subjects from 343-353. infancy to adulthood, pituitary dwarfs and normal 3. Spence HJ, Trias EP, Raiti S (1972) Acromegaly in variant short children. Endocrinol Japon 35: a 9 1/2-year-old boy. AmerJ Dis Child 123: 504-506. 857-864. 4. Guyda H, Robert F, Colle E, Hardy J (1973) 10. Peretti E, Forest M (1976) Unconjugated plasma Histrogic, ultrastructural, and hormonal charac- dehydroepiandrosterone plasma levels in normal terization of a pituitary tumor secreting both hGH subjects from birth to in human; the and prolactin. J Clin EndocrinolMetab 36: 531-546. use of a sensitive radioimmunoassay.J ClinEndocri- 5. Yamada Y, Okada M, Nohara Y, Fujita J, Endo M, nol Metab 43: 982-991. Matsuura N, Nakayama K, Okuno T (1977) A case 11. Peretti E, Forest M (1978) Pattern of plasma with pituitary gigantism. Horumon To Rinsyo 25: dehydroepiandrosterone sulfate levels in humans 941-948 (in Japanese). from birth to adulthood: evidence for testicular 6. Oikawa H, Yamasaki M, Machida Y, Kusunoki T production. J Clin Endocrinol Metab 47: 572-577. (1977) A case of child gigantism. Horumon To 12. Collu R, Ducharme JR (1978) Role of adrenal Rinsyo 25: 1326-1328. (In Japanese). steroids in the initiation of pubertal mechanisms. 7. Espiner EA, Carter TAH, Abbott GD, Wrightson P In: James VHT, Serio M, Giusti G, Martini L (eds) (1981). Pituitary gigantism in a 31-month-old girl: The Endocrine Function of the Human Adrenal endocrine studies and successful response to Cortex. Academic Press, London, New York, San hypophysectomy. J Endocrinol Invest 4: 445-450. Francisco: 547-559. 8. Ritzen EM, Wettrell G, Davies G, Grant DB (1985) 13. Drucker WD, Blumberg JM, Gandy HM, PITUITARY GIGANTISM AND ADRENARCHE 257

RR, Verde AL (1972) Biologic Activity of dehyd- Specific prolactin binding in the rat adrenal gland: roepiandrosterone sulfate in man. J Clin Endocrinol its characterization and hormonal regulation. J Metab 35: 48-54. Endocrinol 89: 317-325. 14. Rosenfield RL (1971) Plasma 17-ketosteroids and 24. Higuchi K, Nawata H, Maki T, Higashizima M, 17 beta-hydroxysteroids in girls with premature Kato K, Ibayashi H (1984) Prolactin has a direct development of sexual hair. J Pediatr 79: 260-266. effect on adrenal androgen secretion. J Clin Endoc- 15. Korth-Schutz S, Levine LS, New MI (1976) Serum rinol Metab 59: 714-718. androgens in normal prepubertal and pubertal 25. Skottner A, Clark RG, Fryklund L, Robinson ICAF children and in children with precocious adrenar- (1989) Growth responses in a mutant dwarf rat to che. J Clin Endocrinol Metab 42: 117-124. human growth hormone and recombinant human 16. Parker LN, Odell WD (1979) Evidence for exist- insulin-like growth factor I. 124: ence of cortical androgen-stimulating hormone. 2519-2526. Am J Physiol 236: E616-620. 26. Pillion DJ, Arnold P, Yang M, Stockard CR, Grizzle 17. Cutler Jr GB, Davis SE, Johnsonbaugh RE, WE (1989) Receptors for insulin and insulin-like Loriaux DL (1979) Dissociation of cortisol and growth factor-I in the human adrenal gland. adrenal androgen secretion in patients with secon- Biochem Biophys Res Commun 165: 204-211. dary adrenal insufficiency. j Clin Endocrinol Metab 27. Shigematsu K, Niwa M, Kurihara M, Yamasita K, 49: 604-609. Kawai K, Tsuchiyama H (1989) Receptor auto- 18. Sklar CA, Kaplan SL, Grumbach MM (1980) radiographic localization of insulin-like growth Evidence for dissociation between adrenarche and factor-I (IGF-I) binding sites in human fetal and gonadarche: studies in patients with idiopathic adult adrenal glands. Life Sciences 45: 383-389. precocious puberty, gonadal dysgenesis, isolated 28. Tokuhiro E, Suwa S (1981) The study of serum gonadotropin deficiency, and constitutionally de- dehydroepiandrosterone sulfate levels in normal layed growth and adolescence. J Clin Endocrinol children and children with pituitary . Metab 51: 548-556. Folia Endocrinol Jap 57: 1186-1198 (In Japanese). 19. Rich BH, Rosenfield RL, Lucky AW, Helke JC, 29. Rosenfeld RG, Hintz RL, Johanson AJ, Sherman Otto P (1981) Adrenarche: changing adrenal re- B, Brasel JA, Burstein S, Chernausek S, Compton sponse to adrenocorticotropin. J Clin Endocrinol P, Frane J, Gotlin RW, Kuntze J, Lippe BM, Metab 52: 1129-1136. Mahoney PC, Moore WV, New MI, Saenger P, 20. Carter JN, Tyson JE, Warne GL, McNeilly AS, Sybert V (1988) Three-year results of a rando- Faiman C, Friesen HG (1977) Adrenocortical mized prospective trial of methionyl human function in hyperprolactinemic women. J Clin growth hormone and oxandrolone in Turner Endocrinol Metab 45: 973-980. syndrome. J Pediatr 113: 393-400. 21. Schiebinger RJ, Chrousos GP, Cutler GB, Loriaux 30. Takano K, Shizume K, Hibi I, the members of the DL (1986) The effect of serum prolactin on plasma committee for treatment of Turner syndrome adrenal androgens and the production and meta- (1989) Treatment of 94 patients with Turner's bolic clearance rate of dehydroepiandrosterone syndrome with recombinant human growth hor- sulfate in normal and hyperprolactinemic subjects. mone (SM-9500) for two years-the results of a J Clin Endocrinol Metab 62: 202-209. multicentric study in Japan. Endocrinol Japon 36: 22. Posner BI, Kelly PA, Shiu RPC, Friesen HG (1974) 569-578. Studies of insulin, growth hormone and prolactin 31. Genentech Collaborative Study Group (1989) binding: tissue distribution, species variation and Idiopathic : Results of a one-year chracterization. Endocrinology 95 : 521-531. controlled study of human growth hormone treat- 23. Calvo JC, Finocchiaro L, Lathy I, Charreau EH, ment. J Pediatr 115: 713-719. Calandra RS, EngstrOm B, Hansson V (1981)