Primary Empty Sella Syndrome with Visual Field Defects
T. BLJCKMAN, M.D.
MUHAMMED HUSAIN, M.D.’ THOMAS J. CARLOW,
GLENN T. PEAKE, M.D. Albuquerque, New Mexico
From the Veterans Administration Hospital and University of New Mexico School of Medicine, Albuquerque, New Mexico 87131. This research was supported by the Veterans Administration (Research and Education Associate Award to Dr. M. Buckman) and by NIH Grant HD05794. Requests for reprints should be addressed to Dr. f&ire T. Buckman, Research Veterans Administration Hospital, Service, 2100 Ridgecrest Drive, S.E., Albuquerque. New Mexico 87108. Manuscript accepted November 18. 1975. * Present address: Veterans Administration Hospital, 1055 Clermont Street, Denver, Colorado 80220.
A patient with an enlarged, asymmetrlc sella turclca and vlsual field defects suggestive of a pituitary or parasellar tumor underwent extenslve roentgenographlc and pltultary function studles. No abnormalltles In pltultary lutelnlrlng hormone, folllcle-stlmulatlng hormone, thyroid-stlmulatlng hormone, ACTH, prolactln or vasopressln secretion were detected. Growth hormone secretion was provoked by arglnlne Infusion but not by hypoglycemia. Pneumoencephalography revealed air In the sella turclca, and no evidence of tumor. Thus, an enlarged sella turclca In a patient with visual field defects but normal pltultary function may suggest the presence of an “empty sella syndrome.” A large asymmetric sella turcica in association with a visual field defect suggests the presence of a pituitary or parasellar tumor. Some of these enlarged sellas have been found to be “empty” on pneumoencephalography or during transfrontal or transsphenoidal pituitary surgery. Weiss and Raskind [l] emphasized the need to distinguish between “primary” and “secondary” empty sella syndrome. The secondary syndrome is associated with previous surgical or radiation treatment for pituitary or parapituitary tumor. The primary syndrome, on the other hand, is idiopathic and not associated with such therapeutic maneuvers. Although visual field defects have regularly been described in the secondary empty sella syndrome, only a few cases of chiasmal defects have been reported in the primary syndrome [2-61. Pituitary function in the primary empty sella syndrome with visual field impairment has been reported to be (1) compromised by clinical evaluation [2,3], (2) compromised by unspecified pituitary function studies , or (3) not evaluated [4,5]. More detailed endocrine studies were carried out in one patient who had diabetes insipidus . In addition to vasopressin deficiency, this patient had modestly impaired growth hormone and cortisol responses to insulin hypoglycemia. The pituitary-thyroidal axis was intact. We describe here an additional patient with the empty sella syndrome who presented with a large asymmetric sella and bitemporal superior quadrantanopsia. Unlike the previously reported cases of primary empty sella syndrome with visual field impairment in which endocrine status was incompletely evaluated, extensive endocrine studies in our patient revealed normal pituitary function. Since neither irradiation nor surgery, except in the rare instance of progressive visual field loss, is indicated in the empty sella syn-
The Amerlcen Journal of Mediche
PRIMARY EMPTY SELLA SYNDROME-BUCKMAN
air in the sella.
drome, pneumoencephalography is an essential diagnostic procedure in any patient presenting with a visual field defect and an enlarged sella turcica. CASE REPORT A 47 year old Caucasian man was admitted to the Albuquerque VA Hospital in September 1974 because of impending delirium tremens and blackout spells. The patient admitted to being a heavy alcohol consumer for many years, although the few days prior to admission he had only consumed a few bottles of beer and essentially no other calories. On the morning of admission, the patient stated that he felt dizzy and he lost consciousness for approximately 30 minutes. He described two such prior blackouts, both temporally related to heavy drinking. There was no history suggestive of seizure activity. The patient related a long history of poorly defined headaches in the frontal region and on top of the head, sometimes lasting one to two days. He denied any recent change in the character, location or duration of the headaches. Diminution of visual acuity and occasional diplopia, progressive over the past two years, had forced him to quit his job as a diamond cutter. Three months prior to admission, he had a skull roentgenogram at a local hospital and was told that he had a “brain tumor.” Physical examination revealed a disheveled, tremulous, middle-aged man with a strong odor of alcohol on his breath. Vital signs showed a blood pressure of 160190 mm Hg, supine; a pulse rate of 104/min; a respiratory rate of 16/min: and an oral temperature of 96.6’F. Habitus was that of a normal man. He had abundant body and facial hair which he had not shaved for several days. Eye examination is subsequently described in detail. His thyroid was normal to palpation. There was no gynecomastia. Genital examination revealed a normal penis and each testicle measured 2 by 4 cm. Neurologic examination, in-
cluding cranial nerves, motor strength, sensory examination, deep tendon reflexes and cerebellar function, was intact except for a drift of his outstretched right hand and a wide-based gait. Both of the latter abnormalities resolved during the course of his hospitalization. The remainder of the physical examination was unremarkable. Initial laboratory data revealed normal urinalysis, electrolytes and blood urea nitrogen. Liver function studies showed modestly increased lactic dehydrogenase (LDH), serum glutamic oxaloacetic transaminase (SGOT), alkaline phosphatase and total bilirubin levels consistent with alcoholic liver disease. A random blood sugar level was 126 mg/ 100 ml. Complete blood count was normal except for a hematocrit value of 55 per cent which reflected a mild state of dehydration. A chest roentgenogram revealed old granulomatous calcifications and old pleural scars, but no acute changes. An electrocardiogram was within normal limits. A lateral skull roentgenogram showed an enlarged sella turcica with a double floor (Figure 1A). Tomograms of the sella turcica confirmed the presence of a double floor with the anterior portion of the left side depressed compared to the right. Measurements of sellar volume revealed it to be markedly enlarged at 2,040 mm3 (normal range 240 to 1,092 mm3) . Carotid artery siphon calcification was seen in the region of the sella. A pneumoencephalogram with polytomography showed extension of subarachnoid air into the posterior portion of the sella turcica, a finding consistent with the empty sella syndrome (Figure IB). There was no evidence of tumor in the sellar or parasellar region. In addition, there was minimal enlargement of the third and lateral ventricles, and minimal widening of the cerebral sulci. considered consistent with his long history of alcoholic abuse. Eye Examination. Goldman quantitative perimetry documented a bitemporal quadrantanopsia on three different occasions within an eight month period (Figure 2). There
The American Journal of Medicine
PRIMARY EMPTY SELLA SYNDROME-BUCKMAN
Goldman quantitative perimetry illustrating
was no progression of the visual field defect during this time span. The visual field defect was more pronounced in the left eye than in the rigfit, being demonstrable with a 1 mm* object with unit intensity in the left eye. To demonstrate a temporal defect in the right eye a smaller object of 0.25 mm* unit intensity was necessary. Testing of visual acuity revealed a refractile error which could easily be corrected with corrective lenses. Pupillary response to light, ocular motility and ophthalmoscopy were normal. Endocrine Evaluation. In order to determine the functional reserve of the compressed pituitary gland, an investigation of pituitary function was undertaken. Normal serum thyrosine (T4), Ts resin uptake and thyroid stimulating hormone (TSH) (Table I) suggested an intact hypothalamic-pituitary-thyroidal axis. To determine the TSH reserve in this patient more precisely, 500 pg thyrotropin-releasing hormone (TRH) was administered intravenously and TSH levels were determined at timed intervals. Serum TSH rose from a basal level of 1.4 pU/ml to a maximum of 13 pU/ml 45 minutes after the administration of TRH (Figure 3C), indicating adequate pituitary TSH reserve. Hypothalamic-pituitary-testicular function was assessed by serum luteinizing hormone (LH), follicle-stimulating hormone (FSH) and testosterone assay. These were found to be normal (Table I). To assess pituitary-adrenal function, a standard insulin tolerance test was performed; 0.1 unit of regular crystalline insulin per kg body weight was administered intravenously. Blood sugar fell from a basal level of TABLE I
Endocrine Studies Test
by R IA (clg/lOO ml) T, resin uptake (%I TSH (flu/ml) l_H (mlU/ml) FSH (mlU/ml) Testosterone (ng/lOO ml) Urine osmolality following overnight water deprivation (mOsm/kg)
3.8 6.0 8.0 701.0 1,063 .O
The American Journal of Medicine
100 mg/lOO ml to a nadir of 45 mg/iOO ml 40 minutes after the administration of insulin, resulting in an adequate hypoglycemic stimulus for ACTH secretion. Serum cortisol rose from a basal level of 9 cLg/iOO ml to a peak level of 24 pg/lOO ml 90 minutes after the administration of insulin (Figure 3A). This is considered an adequate cortisol response and suggests an intact pituitary-adrenal axis. Human growth hormone (HGH) was also measured during insulin-induced hypoglycemia, and the response was blunted (Figure 3A). Because hypoglycemia-induced growth hormone response is blunted in some normal subjects , a second growth hormone provocative study was performed. Thirty gramsof arginine hydrochloride was infused intravendusly over a 30 minute period. After this stimulus, HGH rose from a basal level of 8 ng/ml to a peak of 30 ng/ml (Figure 3B), indicating normal pituitary growth hormone reserve to this amino acid stimulus. Pituitary prolactin reserve was studied after the oral administration of a phenothiazine, perphenazine, thought to act primarily via hypothalamic suppression of prolactin inhibitory factor. Radioimmunoassayable serum prolactin increased from 4 ng/ml to a peak of 17 ng/ml (Figure 3D). The normal male response is an increase greater than 8 ng/ml to a value above 15 ng/ml (unpublished data). Prolactin was also measured after direct pituitary stimulation with TRH, and the response to this stimulus was also indicative of normal pituitary prolactin reserve (Figure 3C). Finally, the urine osmolality of 1,083 mOsm/kg following overnight water deprivation was sufficiently high to rule out vasopressin deficiency (Table I). Thus, seven of the nine pituitary hormones were evaluated and no trophic hormone deficiencies or excesses were found to suggest the presence of a functioning pituitary adenoma or significant destruction of pituitary tissue in this patient. COMMENT Busch  introduced the term “empty sella” in an investigation of 788 consecutive postmortem specimens to describe the empty appearance of 40 sellas, an incidence of 5.1 per cent. Subsequently,
PRIMARY EMPTY SELLA SYNDROME-EUCKMAN
other inuestigators have confirmed the common occurrence of an empty appearing sella in postmortem studies of subjects without known endocrine disease [ 10.111. In addition to the autopsy findings, the term has been used to describe empty appearing sellas on surgical transfrontal or transsphenoidal exposure, and the appearance of air or contrast material in the sella at pneumoencephalography. An incomplete diaphragma sellae appears to be a precondition to the development of this syndrome, permitting extension of the subarachnoid space below the diaphragm [ 121. In the primary empty sella syndrome it is postulated that a congenital defect in the diaphragm, coupled with normal and constant pulsations of cerebrospinal fluid, pushes the subarachnoid space farther and farther into the sella, resulting, in time, in compression of the pituitary gland and enlargement of the sella [ 121. Alternatively, Sutton and Vezina [ 131 suggested, in their patients with empty sellas in association with a pituitary tumor, that pituitary tumor apoplexy may result in herniation of the subarachnoid space into the sella through an incompetent diaphragm. Whether primary intrapituitary processes such as cysts can secondarily establish communications with the subarachnoid space in the presence of a competent diaphragm is not known. Although pituitary function is stated to be normal in most patients with the primary empty sella syndrome [ 141, thorough and systematic studies of pituitary function have not been reported in these patients. Few studies exist utilizing sensitive and specific radioimmunoassay technics to evaluate not only basal hormone secretion but also pituitary hormone reserve to provocative stimulation. No studies exist in which a complete pituitary evaluation, including TSH, ACTH, FSH, LH, HGH, PRL (prolactin) and AVP (arginine vasopressin) function has been made in the same patient. The status of pituitary function in the primary empty sella syndrome is, therefore, largely unknown. Despite the conspicuous absence of reports of systematic pituitary evaluation, abnormalities in pituitary function have occasionally been described. The empty sella is known to occur in association with pituitary tumors presenting with acromegaly or the galactorrhea-amenorrhea syndrome [ 131. Deficiencies of pituitary LH [ 151, TSH [ 161, ACTH [ 161 and vasopressin  have been reported. Abnormalities in growth hormone dynamics have been described by several investigators. Caplan and Dobben [ 171 reported a blunted growth hormone response to argtnine infusion in one patient. Brisman et al.  observed a diminished growth hormone response to hypoglycemia in eight of 13 patients. Since the degree of hypoglycemia was not documented in this study,
PerpAenozine8 mg p. O. 20
3 4 HOURS
Dynamic studies of pituitary function. The horizontal axes for panels A, C3and C are depicted in minutes and for panel 0 in hours. A, cortisol and growth hormone response ta insulin hypoglycemia. 6, growth hormone reFponse to arginine infusion. C, TSH and prolactin response to TRlj administration. D, prolactin response to perphenazine ingestion.
American Journal of Medicine Volume61
these data cannot be interpreted as indicating growth hormone unresponsiveness. Faglia et al. [ 181 and Neelon et al.  described an additional five patients with the primary empty sella syndrome who failed to respond to insulin hypoglycemia. Our patient also failed to show an increase in HGH following insulin hypoglycemia although he had normal HGH dynamics following arginine infusion. Since some normal subjects fail to respond to insulin or to arginine but have a normal response to another stimulus , the diagnosis of growth hormone deficiency cannot be established on the basis of only one provocative study. The most striking abnormality in our patient was the visual field defect manifested by a mild bitemporal superior quadrantanopsia. Although visual field defects are common in secondary empty sella syndrome (secondary to radiotherapy or surgery of the sellar region), they have only rarely been reported in the primary empty sella syndrome. Thorough endocrine evaluation has not been accomplished in the reported cases associated with visual field defects. Robertson  described a patient with the primary empty sella syndrome who had clinical evidence of panhypopituitarism and loss of temporal vision. Walsh and Hoyt  described a 35 year old man with bitemporal field defects and optic atrophy whose endocrine status was not evaluated although he was stated to be impotent. An additional two patients were described by Mortara et al. : one had an incomplete bitemporal hemianopsia and was stated to be panhypopituitary (no data given); the other patient had a complete bitemporal hemianopsia with
incomplete binasal involvement (endocrine function not studied). A fifth patient was described by Metzger and associates  with partial binasal and bitemporal field defects whose endocrine status was not described. Matisonn and Pimstone  described a sixth patient with diabetes insipidus who had gross visual impairment of the left eye and a temporal field defect in the right. Surgical exploration of the suprasellar region has been described in three patients with the primary empty sella syndrome. The optic nerves and chiasm dipped into the sella  or were in their normal position but thinned [2,4]. These anatomic variations were probably responsible for the visual field defects described. In summary, we present a patient with a chiasmal visual field defect and an empty sella who was subjected to extensive endocrine evaluation. In contrast to all previously published cases of empty sella syndrome associated with visual field defects, in which there was either evidence of pituitary deficiency or in which the endocrine status was not evaluated, we documented normal pituitary function in our patient. The diagnosis of primary empty sella syndrome should be considered in any patient presenting with an enlarged sella turcica, visual field defects and normal pituitary function. ACKNOWLEDGMENT We thank Abbott Laboratories for supplying the TRH and Mr. Vernon Heade, BCMC Clinical Laboratory, for expert technical assistance.
a. 9. 10.
Weiss SR, Raskind R: Non-neoplastic intrasellar cysts. Int Surg 51: 282, 1969. Robertson EG: Pneumoencephalography, 2nd ed. Springfield, Charles C Thomas, 1967, p 314. Walsh FB, Hoyt WF: Clinical Neuro-Ophthalmology, 3rd ed. Baltimore, Williams & Wilkins Co., 1969, p 2098. Mortara R, Norell H: Consequences of a deficient sellar diaphragm. J Neurosurg 32: 565, 1970. Metzger J, Helias A, Messimy R, et al.: Nouveaux cas de selles turciques dites vides avec signes neuro-opthalmologiques ou rhinorrhee. Rev Neurol 123: 435, 1970. Matisonn R, Pimstone B: Diabetes insipidus associated with an empty sella turcica. Postgrad Med J (South Africa) 49: 274, 1973. DiChiro G, Nelson KB: The volume of the sella turcica. Am J Roentgen01 87: 989, 1962. Parker ML, Hammond JM, Daughaday WH: The arginine provocative test: an aid in the diagnosis of hyposomatotropism. J Clin Endocrinol Metab 27: 1129. 1967. Busch W: Die morphologie der sella turcica und ihre beziehungen zur hypophyse. Virchows Arch 320: 437. 1951. Bergland RM, Ray BS, Torack RM: Anatomical variations in the pituitary gland and adjacent structures in 225 human autopsy cases. J Neurosurg 28: 93, 1968.
The American Journal of Medlclne
Kaufman B, Chamberlin WB Jr: The ubiquitous “empty” sella turcica. Acta Radio1 (Diagn) (Stockh) 13: 413, 1972. Kaufman B: The “empty” sella turcica-a manifestation of the intrasellar subarachnoid space. Radiology 90: 931, 1968. Sutton TG, Vezina JL: Co-existing pituitary adenomas and intrasellar arachnoid invagination. Am J Roentgen01 Radium Ther Nucl Med 122: 508, 1974. Hodgson SF, Randall RV, Holman CB, MacCarty CS: Empty sella syndrome: report of 10 cases. Med Clin North Am 56: 897, 1972. Brisman R, Hughes JEO, Holub DA: Endocrine function in nineteen patients with empty sella syndrome. J Clin Endocrinol 34: 570, 1972. Neelon FA, Goree JA, Lebovitz HE: The primary empty sella: clinical and radiographic characteristics and endocrine function. Medicine 52: 73, 1973. Caplan RH, Dobben GD: Endocrine studies in patients with the empty sella syndrome. Arch Intern Med 123: 611, 1969. Faglia G. Ambrosi B, Beck-Peccoz P, Giovanelli M: Disorders of growth hormone and corticotropin regulation in patients.with empty sella. J Neurosurg 38: 59, 1973.