Dissociation of serum prolactin response to sequential thyrotropin-releasing hormone and chlorpromazine stimulation in patients with primary empty sella syndrome

Vol. 42. No.4. October 1984 Printed in U.SA.

FERTILITY AND STERILITY Copyright © 1984 The American Fertility Society

Dissociation of serum prolactin response to sequential thyrotropin-releasing hormone and chlorpromazine stimulation in patients with primary empty sella syndrome

Walter Futterweit, M.D.* Harry Smith, Jr., Ph.D. t James E. Holt, M.A.t The Mount Sinai School of Medicine of the City University of New York, New York, New York

The presence of galactorrhea and/or hyperprolactinemia in patients with the primary empty sella syndrome (PESS) has been proposed to be of hypothalamic etiology. To further elucidate this possible mechanism, sequential testing of 19 subjects with PESS with 500 M thyrotropin-releasing hormone (TRH), followed by the injection of 0.7 mg/kg chlorpromazine (CPZ) 150 minutes later, was compared with results obtained in 6 patients with idiopathic galactorrhea (lG) and 3 normal adult women in the early follicular phase of the menstrual cycle. The thyroidstimulating hormone and prolactin (PRL) response to TRH was similar in all three groups. The mean maximal increase of serum PRL following CPZ, however, was 16.1 ± 18.5 ng/ml (standard deviation) in the PESS group, whereas the mean maximal PRL response was 68.6 ± 40.9 ng/ml in subjects with IG and 67.7 ± 48.1 ng/ml in the seven normal women. The impaired responsiveness of CPZ in the PESS group was significant (P < 0.05) when compared with the normal CPZ response in the other two groups. The results of this study suggest that patients with PESS may have hypothalamic dysfunction, and that sequential testing of subjects with TRH and CPZ may be of value in differentiating patients with PESS from those with IG. Fertil Steril 42:573, 1984

Hypothalamic dysfunction in the primary empty sella syndrome (PESS) has been discussed as one of several possible mechanisms for the reported occurrence of galactorrhea and/or hyperprolactinemia in this syndrome. 1 , 2 A normal response of serum prolactin (PRL) to thyrotropinreleasing hormone (TRH) and a blunted response of PRL to chlorpromazine (CPZ), a dopamine receptor blocking agent, may suggest a hypothaReceived March 20, 1984; revised and accepted June 14, 1984. *Reprint requests: Walter Futterweit, M.D., Division of Endocrinology, Department of Medicine, Mount Sinai School of Medicine, 1 Gustave Levy Place, New York, New York 10029. tDepartment of Biomathematical Sciences. Vol. 42, No.4, October 1984

lamic defect. 3 It is the purpose of this article to report on 19 patients with PESS, 15 of whom had associated galactorrhea and 4 of whom had hyperprolactinemia, who were tested with sequential TRH-CPZ stimulation tests. The results were compared with those obtained in seven subjects with idiopathic galactorrhea (lG) and seven normal adult women. The data suggest that patients with PESS with galactorrhea and/or hyperprolactinemia may have an associated hypothalamic disturbance. MATERIALS AND METHODS

Nineteen consecutive women, 19 to 62 years of age (mean, 37.9 years) with computerized axial

Futterweit et al. TRH -CPZ testing in primary empty sella

573

r Table 1. Clinical Data of 19 Patients with PESS" Patient

Age

kg

em

1 2 3 4 5

22 33 19 62 46

B B B B B

115 114 92 74 102

157 157 161 157 163

Amen Nl Oligo Panhyst Nl

PI OG PI P3

6 7

49 39

C C

95 54

155 152

Nl Nl

8 9 10

30 44 31

B B H

70 132 56

11 12 13 14 15

34 33 37 41 35

C B B C C

100 58 86 104 55

16 17 18 19

36 53

C B H B

67 100 71 80

Race

G

yr

45

31

Weight Height

Mensesb

Parity"

Galac-

torrhea

Pseudotumor

Chias-

cerebri

CAT

mapexy

+ +

+ +

+

+ + + + +

P2 P2

+

+ +

173 Nl 151 Amen 159 Nl

P4 P3 OG

+ + +

160 162 159 164 163

Oligo Nl Nl Amen

P3 P4 P2 P2 P2

+ + + + +

+ + + + +

160 157 157 169

Nl Menopausal Nl Nl

G3P2 P2 P4 G6P2

+

+ + + +

NI

00

+ +

+

+ +

+ + +

Endocrine data

Thyroiditis, Lthyroxine, 0.1 mg Primary hypothyroidism, L-thyroxine, 0.15 mg Hypothyroidism, jiI-131, L-thyroxine, 0.1 mg

Primary hypothyroidism, L-thyroxine, 0.2 mg

aB, black; C, Caucasian; H, Hispanic. bAmen, amenorrhea; Nl, normal; Oligo, oligomenorrhea; Panhyst, panhysterectomy. cp1, para 1; 00, nulligravida; G3, gravida 3.

tomographic (CAT) evidence of PESS, were studied (Table 1). CAT scans were performed using the General Electric CT 8800 model with sensitivity varying from 0.75 to 3.0 mm. Intracisternal extension of the radiolucent arachnoid space into the sellar fossa in the absence of an enhancing pituitary adenoma was the basic criterion for PESS. The cut which intersected the sella also was studied with an associated high-definition technique. The content of the partially empty sella had a low density equal to that of cerebrospinal fluid. None of the patients studied demonstrated an associated pituitary microadenoma or previous hyperprolactinemia which spontaneously returned to normal. Fifteen of 19 subjects with PESS had galactorrhea, and others were referred for menstrual disturbances, hyperprolactinemia, or an enlarged sella turcica. Three patients with PESS had a history of pseudotumor cerebri, and two were subjected to chiasmapexy for visual impairment. Those subjects with regular menstrual cycles were studied from the early follicular 574

phase of the cycle to minimize the role of endogenous estrogens on PRL response to provocative testing. 4 Other than thyroid therapy in three women with primary hypothyroidism and one patient with subacute thyroiditis (patients 5, 7, 10, and 15) and coexisting PESS, no other medications were taken. Seven normoprolactinemic subjects with IG, 29 to 37 years of age (mean, 33.6 years), were studied, and none had CAT evidence of an enlarged sella turcica/pituitary microadenoma or PESS. Two of these patients had oligomenorrhea (Table 2). Seven normal women volunteers were subjected to PRL stimulation testing during the early follicular phase of the menstrual cycle. There was no evidence of endocrine disturbances in any of these seven subjects. Informed consent was obtained from all patients. An intravenous TRH test was performed using a bolus of 500 J.Lg TRH. The test was performed in the fasting state at 9:00 A.M. After baseline levels for thyroid-stimulating hormone (TSH) and PRL were obtained, serial samples of venous

Futterweit et aI. TRH-CPZ testing in primary empty sella

Fertility and Sterility

Table 2. Clinical Data of Seven Patients with IG and Seven Adult Women Controls Group

Age

Patient

Racea

yr

IG

Normal

20 21 22 23 24 25 26

36 37 36 30 29 32 35

27 28 29 30 31 32 33

26 43 51 23 28 35 50

Weight

Height

Iw

em

C

55 53 72 59 69 51

B

60

169 157 169 170 157 158 163

H

43 75 61 95 56 111 70

157 145 152 173 155 165 170

C B B

C H

C B

C H B B

Menses

Parity

Galactorrhea

+ + + + + + +

Normal Oligomenorrhea Normal Oligomenorrhea Normal Normal Normal

OG OG G5P3 OG P2 OG P1

Oligomenorrhea Normal Normal Normal Normal Normal Normal

OG P1 P2 OG P2 P1 P4

ac, Caucasian; B, black; H, Hispanic. blood were withdrawn at 30, 60, and 90 minutes after TRH administration. An hour later, venous blood was sampled for PRL, and an intramuscular injection of CPZ was given (0.7 mglkg) to all subjects in the deltoid muscle. The maximum CPZ dosage employed was 60 mg intramuscularly. Repeat blood samples for PRL were obtained at 60, 120, 180, and 240 minutes after injection of CPZ. A TRH stimulation test was performed on all 19 patients with PESS, 6 patients with IG, and 3 normal subjects. The CPZ test was also performed on all 19 patients with PESS, 7 patients with IG, and 7 normal control subjects. All of the blood samples were centrifuged, and the serum was stored at -12°C until analyzed. Serum TSH and PRL assays were determined in duplicate by standard radioimmunoassay techniques utilizing Abbott kits (Abbott Laboratories, North Chicago, IL) for PRL and TSH. The normal value for PRL in adult women is 5 to 25 ng/ml, and that for TSH is 0 to 7 /-LIU/ml. The data were analyzed with repeated measures analysis of variance and employed the usual 5% significance tests. Comparisons were made among groups with the least significant difference criterion if the overall F test was significant.

difference in the response of the 3 normal women, 6 patients with IG, or 19 patients with PESS following injection of TRH. The only nonresponders were those with PESS who were treated with thyroid medication at the time ofthe test (patients 5, 7,10, and 15). The maximal response to the TRH stimulus occurred at 30 minutes in all three groups. PROLACTIN RESPONSE TO THYROTROPIN-RELEASING HORMONE

All three groups had an adequate serum PRL response to TRH with a maximum response at 30 minutes following injection of TRH (Fig. 1). Four of the 19 subjects with PESS had basal hyperprolactinemia with a serum PRL range of 27 to 61 ng/ml (patients 1, 8, 12, and 19). None of the latter four patients had baseline elevated levels of serum PRL. An absent serum PRL response to TRH was noted in one subject (patient 8). Although Duncan's multiple range test demonstrated a significant difference between the patients with PESS and patients with IG (P < 0.05), the small number of patients in the IG group makes it impossible to draw firm conclusions about the hyperresponsiveness of serum PRL to TRH in the latter group.

RESULTS PROLACTIN RESPONSE TO CHLORPROMAZINE THYROID-STIMULATING HORMONE RESPONSE TO THYROTROPIN·RELEASING HORMONE

The response of serum TSH to TRH did not differ significantly in any of the three groups (Table 3). Analysis of variance showed no significant Vol. 42, No.4, October 1984

All seven women with IG and the seven normal volunteers had normal serum PRL responsiveness to CPZ, with a mean maximal (~) PRL increase of 68.6 ± 40.9 ng/ml and 67.7 ± 48.1 ng/ ml, respectively. The mean maximal ~ PRL of the

Futterweit et al. TRH-CPZ testing in primary empty sella

575

Table 3. Mean Serum TSH Levels and Response to 500 fLg TRH Intravenously in 19 Patients with PESS, 6 Patients with IG, and 3 Normal Subjects Group

Serum TSH (f1IU/ml) ± standard deviation

No. of patients

PESS IG Normal

19 6 3

Basal

30 minutes

60 minutes

90 minutes

3.6 ± 1.8 4.3 ± 1.6 2.4 ± 0.7

14.7 ± 10.7 22.3 ± 15.7 13.8 ± 7.1

10.6 ± 7.2 11.8 ± 5.8 9.2 ± 3.7

8.6 ± 6.4 12.7 ± 5.5 7.1 ± 2.9

19 patients with PESS was 16.1 ± 18.6 ng/ml, which was significantly different from that of the other two groups (P < 0.05) (Fig. 1). The maximal serum PRL response to CPZ in subjects with IG and normal women usually occurred at 2 to 3 hours following CPZ injection, and the mean maximal Ll PRL of the PESS group varied from 2 to 4 hours following CPZ administration. One of the four hyperprolactinemic patients with PESS (patient 12) responded normally to CPZ with a rise from a basal serum PRL level of 33 to 90 ng/ml 2 hours following administration of CPZ. Only 7 of the 19 patients with PESS demonstrated a twofold increase or more of the basal serum PRL level following CPZ injection. Drowsiness occurred in 8 of the 33 subjects tested with CPZ, and transient hypotension also occurred in 2 of these patients. DISCUSSION

Galactorrhea and/or hyperprolactinemia has been described in association with PESS. 1, 2, 5-16

~~~~6~O~~1~270~~18~O~~2~40~~3~O~O-L~3~6~O~~ 500).l.g TRH iv

0.7mg/kg CPZ im stimulus

Figure 1 The serum PRL response of 7 normal women, 7 patients with IG, and 19 subjects with PESS to 500 fLg TRH intravenously, followed by intramuscular administration of 0.7 mg/kg CPZ 2.5 hours after the injection of TRH. 576

Futterweit et al.

The incidence of galactorrhea in PESS varies to 8.0%.1, 2 Galactorrhea in PESS is frequently associated with regular menses and the absence of hyperprolactinemia, but some patients have oligomenorrhea or amenorrhea associated with PESS. 1, 2, 14, 16 The latter group of patients frequently have associated hyperprolactinemia. If the reports of marked hyperprolactinemia (> 300 ng/mD associated with PESS are excluded, because they probably represent coexisting PRL-secreting pi t)li tary adenomas,8, 13, 14 the usual range of hyperprolactinemia in PESS varies between 30 and 120 ng/ml.1, 2, 7, 12, 14-17 Normal TRH responsiveness in normoprolactinemic subjects with PESS has been well documented in the literature with a twofold or greater increase in basal levels of PRL and TSH. 5, 7, 9, 18-20 Occasional normoprolactinemic patients show a blunting of the serum PRL response to TRH.5, 7, 9 All but 1 of the 19 patients with PESS in this report responded normally to TRH, whereas the 4 subjects on thyroid therapy had an expected blunted response of TSH to the administration of TRH. Although there are no previous reports describing the nature of the TRH response of hyperprolactinemic subjects with PESS, two of four such patients in this series had a normal serum TSH and PRL response to TRH (patients 1 and 19). Six women with IG had a normal TSH response to TRH similar to that of patients with PESS and normal women. The PRL response to TRH, however, appeared to be somewhat exaggerated in three of the six latter subjects tested. Four of five normoprolactinemic subjects with PESS and galactorrhea described in the literature6, 9 failed to double the basal serum PRL level following the administration of 25 mg CPZ intramuscularly, while suppressing normally in response to 500 mg L-dopa. Similarly, a hyperprolactinemic patient with PESS failed to respond to CPZ while suppressing normally to L-dopa 12 while two reported patients also failed to respond to CPZ.12, 16 Thirteen of our 19 patients with

TRH -CPZ testing in primary empty sella

Fertility and Sterility

PESS did not demonstrate an adequate serum PRL response to the administration of CPZ. The importance of using a uniform dosage of 0.7 mgl kg CPZ intramuscularly in assessing the PRL response in our study has also been stressed by Archer et a1. 21 The variable response to CPZ, a nonspecific dopamine receptor antagonist of the hypothalamic as well as pituitary lactotroph receptors, was in contradistinction to the normal PRL response of the seven subjects with IG and seven normal women tested. The absence of a significant serum PRL surge in 62.6% of the patients with PESS following CPZ and normal responsiveness of serum PRL to TRH is highly suggestive of hypothalamic. dysfunction. 3, 22, 23 Similar observations have been noted in a variety of hypothalamic-pituitary disorders. 24 Thus, a possible marker of hypothalamic dysfunction may well be the dissociation of the serum PRL response to sequential TRH and CPZ testing. Occasional instances of an absent serum PRL response to CPZ has been described in patients with IG. 9 Perhaps some of the reported instances of the latter include patients with undiagnosed PESS, or perhaps the dosage of 25 mg CPZ employed was inadequate. Three of the patients with PESS in this series had a history of pseudotumor cerebri, and two of these subjects underwent chiasmapexy. The incidence of increased cerebrospinal fluid pressure in PESS varies from 10.5% to 65%.2,17,20,25-27 Several investigators 1, 2 have suggested that hypothalamic dysfunction in PESS may occur as a result of compression of the hypothalamus or pituitary stalk as a result of elevated cerebrospinal fluid pressure and an associated low anterior end of the third ventricle. Compression or kinking of the pituitary stalk or impairment of its blood supply may result. Distortion of the infundibular stalk with intrasellar cisternal herniation may further compromise the transport or· release of PRL inhibitory factor that reaches the pituitary gland. Evidence of increased intracranial pressure was reported in 24 of 50 patients with PESS described by Brismar and Efendic, 2 with 3 of 6 of their hyperprolactinemic patients demonstrating this sign. Furthermore, the persistence of hyperprolactinemia after chiasmapexy in two of three patients described by Haney et a1.,14 as in one of the patients in this report (patient 1), further suggests a hypothalamic defect in PESS, particularly in that associated with galactorrhea and/or hyperprolactinemia. Vol. 42, No.4, October 1984

It appears that sequential studies of TRH and CPZ may be of diagnostic value in some patients with normoprolactinemic galactorrhea. Although a normal response of serum PRL to TRH and CPZ does not differentiate between those subjects with PESS and IG, an inadequate response to CPZ with a normal response to TRH is suggestive of hypothalamic dysfunction. 3, 22, 23 The high frequency of galactorrhea seen in our series is in agreement with the report of Valenta et a1.,17 who found a significant association of galactorrhea as well as menstrual disturbances in PESS. Acknowledgments. We wish to thank Mrs. Maria T. LaRaque for her expert technical assistance in performing the radioimmunoassay studies. The authors also appreciate the secretarial help of Mrs. Marva D. Williams and Mr. Richard A. Weiss. REFERENCES 1. Bryner JR, EI Gammal T, Acker JD, Asch RH, Greenblatt

2. 3. 4.

5.

6. 7.

8.

9. 10.

11.

12.

13.

RB: Intrasellar subarachnoid herniation or empty sella associated with galactorrhea. Obstet Gynecol 51:198, 1978 Brismar K, Efendic S: Pituitary function in the empty sella syndrome. Neuroendocrinology 32:70, 1981 Friesen H, Guyda H, Hwang P, Tyson JE, Barbeau A: Functional evaluation of prolactin secretion: a guide to therapy. J Clin Invest 51:706, 1972 Buckman MT, Peake GT, Srivastava LS: Endogenous estrogen modulates phenothiazine-stimulated prolactin secretion. J Clin Endocrinol Metab 43:901, 1976 Gates RB, Friesen H, Samaan NA: Inappropriate lactation and amenorrhea: pathological and diagnostic considerations. Acta Endocrinol (Copenh) 72:101, 1973 Bar RS, Mazzaferri EL, Malarkey WB: Primary empty sella, galactorrhea, hyperprolactinemia and renal tubular acidosis. Am J Med 59:863, 1975 Schaison G, Metzger J: The primary empty sella: an endocrine study on 12 cases. Acta Endocrinol (Copenh) 83:483, 1976 Hsu TH, Shapiro JR, Tyson JE, Leddy AL, Paz-Guevara AT: Hyperprolactinemia associated with the empty-sella syndrome. JAMA 235:2002,1976 Kleinberg DL, Noel GL, Frantz AG: Galactorrhea: a study of235 cases, including 48 with pituitary tumors. N Engl J Med 296:589, 1977 Check JH, Rakoff AE, Goldfarb AF, King LC: Amenorrhea-galactorrhea associated with the empty-sella syndrome. Am J Obstet Gynecol 128:688, 1977 Jones JR, DeHempel PAC, Kemmann E, Tenner MS: Galactorrhea and amenorrhea in a patient with an empty sella. Obstet Gynecol 49:S9, 1977 Bryner JR, Greenblatt RB: Primary empty sella syndrome with elevated serum prolactin. Ol:istet Gynecol 50:375, 1977 Archer DF, Maroon JC, Dubois PJ: Galactorrhea, amenorrhea, hyperprolactinemia, and an empty sella. Obstet Gynecol 52:S23, 1978

Futterweit et aI. TRH-CPZ testing in primary empty sella

577

14. Haney AF, Kramer RS, Wiebe RH, Hammond CB: Hypothalamic-pituitary function and radiographic evaluation of women with hyperprolactinemia and an "empty" sella turcica. Am J Obstet Gynecol 134:917, 1979 15. L'Hermite M, Caufriez A, Robyn C: Pathophysiology of human prolactin secretion with special reference to prolactin-secreting pituitary adenomas and isolated galactorrhoea. In Prolactin and Human Reproduction, Edited by PG Crosignani, C Robyn. New York, Academic Press, 1979, p 179 16. Futterweit W: Galactorrhea, amenorrhea, hyperprolactinemia and pseudotumor cerebri in a patient with primary empty sella syndrome: case report with review of the literature. Mt Sinai J Med (NY) 49:514, 1982 17. Valenta LJ, Sostrin RD, Eisenberg H, Tamkin JA, Elias AN: Diagnosis of pituitary tumors by hormone assays and computerized tomography. Am J Med 72:861, 1982 18. Snyder PJ, Jacobs LS, Rabello MM, Sterling FH, Shore RN, Utiger RD, Daughaday WH: Diagnostic values of thyrotropin-releasing hormone in pituitary and hypothalamic disease: assessment of thyrotropin and prolactin secretion in 100 patients. Ann Intern Med 81:751, 1974 19. Ridgway EC, Kourides lA, Kliman B, Bigos T, Maloof F: Thyrotropin and prolactin pituitary reserve in the "empty sella syndrome." J Clin Endocrinol Metab 41:968, 1975

578

20. Jordan RM, Dendall JW, Kerber CW: The primary empty sella syndrome. Am J Med 62:569, 1977 21. .Archer DF, Josimovich JB, Maroon JC: Chlorpromazineinduced changes in serum prolactin in women with oligomenorrhea, amenorrhea, and pituitary adenoma. Fertil Steril 28:1164, 1977 22. Tolis G, Goldstein M, Friesen HG: Functional evaluation ofprolacti" secretion in patients with hypothalamic-pituitary disorders. J Clin Invest 52:783, 1972 23. Langer G, Sachar EJ, Halpern FS, Gruen PH, Solomon M: The ·prolactin response to neuroleptic drugs, a test of dopaminergic blockade: neuroendocrine studies in normal men. J ClinEndocrinol Metab 45:996, 1977 24. Jacobs LS, Daughaday WH: Pathophysiology and control of prolactin secretion in patients with pituitary and hypothalamic disease. In Human Prolactin, Edited by JL Pasteels, C Robyn. Amsterdam, Excerpta Medica, 1973, p 189 25. Foley KM, Posner JB: Does pseudotumor cerebri cause the empty sella syndrome? Neurology 25:565, 1975 26. Jaffer KA, Obbens EAMT, EI Gammal TA: "Empty" sella: review of 76 cases. South Med J 72:294, 1979 27. Neelon FA, Goree JA, Lebovitz HE: The primary empty sella: clinical and radiographic characteristics and endo~ crine function. Medicine (Baltimore) 52:73, 1973

Futterweit et al. TRH-CPZ testing in primary empty sella

Fertility and Sterility