Endocrine Journal 1996, 43 (6), 665-670

Unexplained Increase in Serum Corticosteroid-B finding Levels in a Patient with Chronic Thyroi ditis, Pituitary Adenoma and Empty Sella

KATSUYOSHITOJO, YUJIOHNO*, MASAHIROKAWAMURA*, ANDOSAMU SAKAI

Second Department of Internal Medicine and * First Department of Pharmacology, The Jikei University School of Medicine, Tokyo 105, Japan

Abstract. Unexplained high serum corticosteroid-binding globulin (CBG) concentrations (mean ± SD, 74.1 ± 12.1,ug/dl; normal women, 32.5 ± 5.6 µg/dl) were found in an unmarried woman who was not pregnant or taking exogenous estrogens. She was also found to suffer from subclinical chronic thyroiditis, pituitary adenoma and empty sella. The increased serum CBG concentrations in this patient were not due to any of the factors known to increase CBG. Consistently high basal serum GH levels and unusual GH responses to GH-releasing factor (GRF) and L -dopa were also noted.

Key words: Hypercortisolemia, Corticosteroid-binding globulin, Pituitary adenoma, Chronic thyroiditis, Abnormal GH secretion (Endocrine Journal 43: 665-670,1996)

CORTICOSTEROID-binding globulin (CBG) is a a variety of diseases [2], and hereditary CBG defi- plasma oc1- of hepatic origin with a ciency has also been described [5-7]. Elevated molecular weight of approximately 52000. CBG is serum CBG concentrations, on the other hand, are the major transport for glucocorticoids in usually caused by estrogen therapy or pregnancy. the blood. Under normal physiological conditions, In addition, increased serum CBG levels have been more than 90% of circulating cortisol is bound to noted in 25-30% of patients with acute lymphocyt- CBG in humans [1]. This CBG-bound cortisol is ic and myelocytic leukemia [8] and 80% of chronic considered to be biologically inactive, whereas the active hepatitis [9L In contrast, there is consider- unbound cortisol constitutes the active form of cor- ably less evidence for high serum CBG levels of tisol. The active fraction of plasma cortisol will unknown origin [10, 11]. The following case re- thus depend on the concentration of CBG. The port details a woman with increased serum CBG serum CBG concentrations vary little diurnally and levels but without the other estrogen-associated are similar in men and women [2]. It is generally changes in blood . Furthermore, chronic agreed that there is little or no variation in serum thyroiditis, pituitary adenoma with empty sella and CBG concentrations during the menstrual cycle [3, consistently high basal serum GH levels with par- 4]. adoxical GH responses to GH-releasing factor Abnormally low serum CBG levels are found in (GRF) and L-dopa were also noted in this patient.

Received: March 28, 1996 Accepted: July 1, 1996 Case Report Correspondence to: Dr. Katsuyoshi TOJO, Second Department of Internal Medicine, The Jikei University A 41-year-old unmarried woman was referred School of Medicine, 3-25-8 Nishi-shinbashi, Minato-ku, to the endocrine clinic for exploration of the pitu- Tokyo 105, Japan itary-adrenal axis after she was found to have a 666 TOJO et a!. high morning serum cortisol concentration (27 j g/ dl; normal, 4.0-18.3 ug/dl). Over the previous 6 months she had been aware of alopecia and con- sulted a dermatologist. She has no particular medical history. She was not taking any medica- tion, had regular menstrual cycles, and was not pregnant. There was no history of diabetes melli- tus or hypertension in her family and no other evidence of endocrine abnormalities. Her height was 160 cm and she weighed 54 kg. Her blood pressure was 134/86 mmHg. On physical exami- nations, she was apparently in good health. The thyroid gland was not palpable. Central obesity, moon face, buffalo hump and cutaneous striae were not observed. Visual acuity was normal. Neuro- logical examination was negative. Routine laboratory examinations showed a count of 4.21 x 106/pJ, and a hemoglo- (Amerlex-M FT4) were within normal limits (3.0 bin count of 13.7 g/dl. The eosinophil count was ,uU/ml; normal range, 0.34-3.5 ,uU/ml, 2.7 pg/ml; 0.22 x 103/pl. The white blood cell count was 4.2 x normal range, 2.47-4.34 pg/ml, and 1.03 ng/dl; 103/,ul. Fasting blood glucose was 105 mg/dl, and normal range, 0.97-1.79 ng/dl, respectively). Se- HbA1c was 4.7%. An oral glucose tolerance test rum thyroxin-binding globulin (TBG) (RIAgnost showed a pattern of impaired glucose tolerance. TBG kit, Hoechst Japan, Tokyo) was 22 µg/ml (nor- Urea nitrogen, creatinine and function tests mal range, 14-28 jug/ml). TSH-binding inhibitory were normal. Serum Na was 142 mmol/l, serum immunoglobulins (TBII) (TRAb kit, Baxter Trave- K, 4.5 mmol/l and serum C1, 106 mmol/l. Serum nol Co., Tokyo) was negative (7.4%; normal range, Ca was 9.6 mg/dl. Urinary sugar and protein were <15%). TgAb (TgAb Cosmic kit, Cosmic Co., To- negative. Baseline endocrine results are as follows. kyo) and TPOAb (TPOAb Cosmic kit, Cosmic Co., Serum cortisol (v-coat cortisol kit, Baxter Travenol Tokyo) were positive (1.9 U/ml; normal range, S Co., Tokyo) was consistently high (29.0 ± 3.9 ug/ 0.3 U/ml and 56.0 U/ml; normal range, 50.3 U/ dl; normal range, 4.0-18.3 ,ug/dl), while serum ml, respectively), suggesting that she suffered from ACTH (Allegro ACTH kit, Nihon Medi-Physics, chronic thyroiditis. Dynamic assessments for the Takarazuka) remained high normal (41 ± 11.8 pg/ pituitary-adrenal axis are shown in Table 1. The ml; normal range, 9-52 pg/ml). Urinary free corti- diurnal rhythm of serum ACTH and cortisol ap- sol was 64.6 ± 9.2 jig/day (normal range, 30-100 peared normal (86 pg/ml and 27.0 µg/dl at 0800 µg/day), urinary 17-hydroxycorticosteroid (17- h, 67 pg/ml and 19.9 ug/dl at 1600 h, respective- OHCS) was 8.4 mg/day (normal range, 2.2-7.3 mg/ ly). Serum cortisol showed normal responses to day), urinary 17-ketosteroid was 6.5 mg/day (nor- 1-25ACTH(0 .25 mg i.v.). Serum ACTH and corti- mal range, 2.4-11.0 mg/day). Serum GH (GH kit, sol showed normal suppression after Daiichi Radioisotope Lab) was also consistently administration of 0.5 and 1 mg dexamethasone at high (21.2 ± 7.6 ng/ml; normal range, 0.66-3.68 2300 h the day before. Pituitary function was test- ng/ml), whereas serum somatomedin C (SmC) (So- ed with TRH (500 jig i.v.), LH-RH (100 ,ug i.v.), matomedin C•II Chiba Coaning kit, Chiba•Coaning corticotropin-releasing factor (CRF; 100 ug i.v.), Diagnostic Co., Tokyo) was within normal limits GH-releasing factor (GRF; 100 µg i.v.), 75 g glu- (140 ng/ml; normal range, 79-383 ng/ml). Uri- cose and L-dopa (500 mg p.o.), and the results are nary GH excretion was very high (41.4 pg/mg-Cr; shown in Table 2. The high GH and normal SmC normal range, 6.2 ± 3.3 pg/mg-Cr). Serum TSH levels in conjunction with paradoxical responses (TSH RIABEAD II kit, Dinabott, Tokyo), free tri- of GH to GRF and L-dopa are isolated since thy- iodothyronine (T3) (Amerlex-M FT3, Amersham roid and adrenal functions were normal, as were International, Tokyo) and free thyroxine (T4) basal and stimulated gonadotropin and PRL lev- HIGH CBG LEVELS OF UNKNOWN ORIGIN 667

Table 2. Provocation tests for pituitary hormone secretion

Table 3. Serum CBG, total cortisol and unbound cortisol levels

Table 4. Serum concentration of the estrogen-responsive centrations were measured by radio-immunoassay proteins (CBG-RIAkit, Medgnix Co., Belgium). Her serum CBG concentration, when first measured, was 94.7 jug/ml (normal women, 32.5 ± 5.6 jug/ml). There- after, her CBG concentration has remained consistently high (74.1 ± 12.1,ug/ml) independent of the menstrual cycle, and serum bioactive un- bound cortisol levels in the present patient were within normal limits (7.36 ± 0.4 ng/ml; normal women, 5.6 ± 3.4 ng/ml) (Table 3). Serum sex els. Pituitary magnetic resonance imaging (MRI) hormone-binding globulin (SHBG),also measured showed pituitary microadenoma and empty sella; by RIA (Delfia SHBG kit, Wallac Oy, Turku, Fin- herniation of suprasellar cistern into anterior sella land), was normal (69 nmol/l; normal women, 65 turcia, whereas abdominal computed tomography ± 39 nmol/l). In addition, serum concentrations of (CT) revealed a normal adrenal gland. Because of other estrogen-responsive proteins such as TBG, the marked discrepancy between serum cortisol , and were and urinary free cortisol levels, serum CBG con- normal (Table 4). 668 TOJO et al.

of these target by using polymerase chain Discussion reaction and subsequent quantification by HPLC. The same mechanism should be also considered in The increased serum CBG concentration in the case of inherited CBG excess. Further studies are present patient was not due to any of the factors necessary to determine the yet unknown causes of known to increase CBG, such as pregnancy or es- high CBG levels. Whether an incidentally discov- trogen administration. The concentration of each ered pituitary microadenoma is functioning or not of the estrogen-responsive proteins including TBG, remains inconclusive at present. Careful follow- ceruloplasmin, haptoglobin and albumin, was with- up study is necessary. in the normal range, supporting the negative The cause of an abnormal GH secretry pattern history of estrogen medication, and the endoge- remains unknown. Basal serum GH levels repeat- nous serum estradiol concentration was normal. edly measured were consistently high, while serum She was apparently in good health. She did not SmC levels were within normal limits. Further- suffer from diabetes mellitus or tuberculosis, which more, a decrease in serum GH in response to GRF have been associated with slightly increased se- and L-dopa administration was observed. The dis- rum CBG concentrations [13, 14]. There was no crepancy between serum GH levels and SmC levels evidence of lymphoprolif erative disease or chron- is usually observed in patients with chronic renal ic active hepatitis, other causes that have been failure, chronic liver disease such as chronic hepa- reported in association with increased serum CBG titis and of the liver, and a hyponutritional [8, 9]. Finally, repeated measurements ruled out state, all of which are unlikely in the present pa- mere coincidence. It would be reasonable to as- tient. Exaggerated urinary GH excretion with cribe the increase in serum CBG in the present normal renal function also reflects overproduction patient to heredity, analogous to the well known of GH. In addition, normal suppression of serum inheritance of high serum TBG [15] and low serum GH levels by glucose tolerance test excludes the CBG concentrations [5, 6], although the serum CBG possibility of acromegaly. One possible reason for concentrations of other family members have not unexplained high GH levels may be a kind of emo- yet been measured. In fact, a similar observation tionally stressful situation caused by suffering from was reported previously [10]. Since no obvious alopecia. GH is responsive to a variety of stressful inherited trait could be detected in any family, the stimuli [19]. Increases in GH occur during sur- inheritance of increased serum CBG levels cannot gery, cardiac catheterization, electroshock therapy, yet be established definitely, although there is good gastroscopy, physical exercise, and stimuli of a evidence for the inheritance of low and high CBG more purely psychological nature [20]. But, even levels in animals [16]. So far, the molecular basis in such cases, the GH response to GRF and L-dopa of high serum CBG levels of unknown origin has may have only a blunted pattern. Furthermore, not yet been elucidated. Several possible mecha- GH is secreted episodically, and the rate of change nisms may be considered. The pathogenesis of in blood levels is rapid. The episodic nature of hemophilia B, one form of the inherited, X-linked, GH secretion complicates interpretation of the re- bleeding disorders, has recently been attributed to sponse of the hormone to various stimuli. To the mutations at - 20 and - 26 in the clotting fac- elucidate a possible involvement of episodic and tor IX promoter, leading to impaired transcription pulsatile manner of GH secretion during provoca- by disrupting the for the liver-enriched tion tests, saline solution was injected as a vehicle transcription factor LF-A1 /HNF4 [17]. Similarly, control and serum GH levels were measured for it may also be possible to suggest that enhanced up to 2 h. As shown in Table 3, serum GH levels promoter activity causes exaggerated transcription, fell spontaneously from 11.7 ng/ml to 2.7 ng/ml, leading to high serum CBG levels. Most recently, suggesting that the episodic and pulsatile GH se- Mori et al. [18] demonstrated that amplifica- cretion may be, at least in part, involved in tion of TBG is the cause of inherited TBG excess in abnormal GH responses to provocation tests. Fi- two Japanese families by evaluating the gene dos- nally, the possibility that biologically inactive GH age of TBG relative to those of fi-globin and is secreted into the general circulation should also Duchenne muscular dystrophy by coamplif ication be considered. The pathophysiological role of emp- HIGH CBG LEVELS OF UNKNOWN ORIGIN 669

ty sella noted in this patient seems unlikely. It is not noted in the present patient. Further studies now generally accepted that the finding of an emp- are necessary to clarify the mechanism of abnor- ty or partially empty sella on routine MRI of the mal GH secretion. brain is of no clinical significance [21]. Dynamic In conclusion, we reported a patient presenting endocrine tests on hypothalamic pituitary function with high serum CBG levels without evidence of in patients with a primary empty sella syndrome excessive endogenous estrogen secretion, exoge- revealed that only 16 of 52 patients had endocrine nous estrogen administration or any other cause. problems [22]. Among them, hyperprolactinemia Careful follow-up study will give new insight into was found in 17 patients and panhypopituitarism the reason for such unexplained high serum CBG in 3 patients, and both of these conditions were concentrations.

References

1. Siiteri PK, Murai JT, Hammond GL, Nisker JA, lin? J Clin Endocrinol Metab 28:1073-1075. Raymoure WJ, Kuhn RW (1982) The Serum trans- 11. Coolens JL, Heyns W (1989) Marked elevation and port of steroid hormones. Recent Prog Horm Res 38: cyclic variation of corticosteroid-binding globulin: 457-510. An inherited abnormalities. J Clin Endocrinol Metab 2. Coolens JL, Van Baelen H, Heyns W (1987)Clinical 68: 492-494. use of unbound plasma cortisol as calculated from 12. Demura R, Terasaki K, Jibiki K, Odagiri E, Demura total cortisol and corticosteroid-binding globulin. J H, Suda T, Shizume K (1987) A case of Cushing's Steroid Biochem26:197-202. syndrome associated with CBG deficiency. J Tokyo 3. Hammond G, Lahteenmaki PL, Lahteenmaki P, Women's Med Coil 57: 377-384. Luukkainen T (1982) Distribution and percentages 13. De Moor P, Heyns W (1967) Cortisol binding ca- of non-contraeptive steroids in human serum. J Ste- pacity of plasma in diabetic subjects. J roid Biochem17: 375-380. Clin Endocrinol Metab 27: 706-708. 4. Moore D, Kawagoe S, Davajan V, Mishell D, 14. De Moor P, Demedts M, Faict D, Gyselen A (1984) Nakamura R (1978) An in vivo system in man for Transient high transcortin levels in patients with quantitation of estrogenicity. I. Physiologic changes sputum positive pulmonary tuberculosis. Acta Clin in binding capacity of serum corticosteroid-bind- Belg 39: 358-362. ing globulin. Am J ObstetGynecol 130: 475-481. 15. Shane S, Seal U, Jones J (1971) X- linked 5. De Moor P, Hendrickx A, Steeno 0 (1965) Idio- inheritance of elevated thyroxine-binding globulin pathic lowering of the cortisol binding capacity of in association with goiter. J Ciin Endocrinol Metab human plasma transcortin. Ann Endocrinol (Paris) 32: 587-594. 26: 488-493. 16. Faict D, De Moor P, Bouillon R (1986) Transcrotin 6. Lohrenz FN, Seal US, Doe RP (1967) Adrenal func- and vitamin D-binding protein levels in mouse se- tion and serum protein concentrations in a kindred rum. J Endocrinol 109:141-147. with decreased corticosteroid binding globulin 17. Crossley M, Ludwig M, Stowell KM, Vos PD, Olek (CBG)concentration. J Clin EndocrinolMetab 27: 966- K, Brownleet GG (1992) Recovery from hemophilia 972. B Leyden: An androgen-responsive element in the 7. Roiman A, Bruchis S, Bauman B, Kaufman H, Laron factor IX promoter. Science 257: 377-379. Z (1984) Total deficiency of corticosteroid-binding 18. Mori Y, Miura Y, Takeuchi H, Igarashi Y, Sugiura J, globulin. Clin Endocrinol(Oxf) 21: 541-548. Saito H, Oiso Y (1995) Gene amplification as a cause 8. De Moor P, Louwagie A, Van Baelen H, Van De of inherited thyroxine-binding globulin excess in Putte I (1980) Unexplained high transcortin levels two Japanese families. J Clin Endocrinol Metab 80: in patients with various hematological disorders 3758-3762. and in their relatives: A connection between these 19. Brown GM, Seggie JA, Chambers JW, Ettigi PG high transcortin levels and HLA antigen B12.J Clin (1978) Psychoendocrinology and growth hormone: EndocrinolMetab 50: 421-426. A review. Psychoneuroendocrinology 3:131-153. 9. Orbach 0, Schussle A (1989) Increased serum corti- 20. Schalch DS (1967) The influence of physical stress sol binding in chronic active hepatitis. Am J Med and exercise on growth hormone and insulin secre- 86: 39-44. tion in man. J Lab Clin Med 69: 256-269. 10. Lohrenz F, Doe RP, Seal US (1968) Idiopathic or 21. Hardjasudarma M, White KE, Nandy I, Burns PL genetic elevation of corticosteroid-binding globu- (1994) Sellar emptiness on routine magnetic reso- 670 TOJO et al.

nance imaging. South Med 187: 340-343. testing of hypothalamic pituitary function in pa- 22. Buchfelder M, Brockmeier S, Pichl J, Schrell U, tients with a primary "Empty" sella syndrome. Fahlbusch R (1989) Results of dynamic endocrine Horm Metab Res 21: 573-576.