Hôpital
de la Pitié, 75013
Paris -
THE PRIMARY EMPTY SELLA
AN ENDOCRINE STUDY ON 12 CASES
By G. Schaison and
J. Metzger
ABSTRACT Twelve patients (10 women and 2 men) with a primary empty sella turcica were studied. Endocrine function tests were performed as follows: growth hormone (GH) was measured after insulin-induced-hypoglycaemia, luteinizing hormone (LH) and follicle-stimulating hormone (FSH) after LH-releasing hormone, thyrotrophin (TSH) and prolactin after thyrotrophin-releasing hormone; pituitary reserve of adrenocorticotrophin (ACTH) was determined by measurement of plasma cortisol after lysinevasopressin and 11 deoxycortisol after metyrapone. Five of the patients (group A) had no endocrine disturbance. Seven patients (group B) had a hypothalamo-pituitary disorder. Two of them had panhypopituitarism which appeared in one case after meningoencephalitis and in the other after a severe cranial trauma. In two cases an amenorrhoea-galactorrhoea syndrome with increased prolactin level (68 and 230 ng/ml) led to a diagnosis of a prolactin producing adenoma, which was confirmed by surgery. Finally three cases of amenorrhoeagalactorrhoea, with normal prolactin level, and/or diabetes insipidus remained unexplained. However, no causal relationship could be demonstrated between the pituitary disturbance and the "empty sella". Primary empty sella turcica is therefore a neuroanatomical and neuroradiological entity with no endocrine implication. A pituitary disorder might suggest a microadenoma or an incidentally associated disease.
The term primary empty sella is used to denote intrasellar arachnoid invagina¬ tion of the chiasmatic cistern probably due to incomplete congenital development of the sellar diaphragm (Busch 1951). It also includes intrasellar arachnoidal cyst and even spontaneous necrosis of pituitary adenomas which are often
(Caplan et al. 1969). This definition however exludes sella empty secondary to radiotheraphy or surgical removal of the hypophysis Ras (Weiss 8c kind 1969). Expansion of the subarachnoidal space leads to a reshaping of sellar anatomy, the hypophysis being pushed back into the posterior-inferior part of the sella. The question arises however as to whether pituitary insult is sufficient to induce endocrine disease. In spite of many publications however, this problem remains unresolved. Apparently everything has already been described: no concurrent endocrine disorder at all, panhypopituitarism (Caplan et al. 1969; Neelon et al. 1973), GH, LH or ACTH deficiency (Neelon et al. 1973; Brisman et al. 1972; Faglia et al. 1973), amenorrhoea associated with galactorrhoea (Gates et al. 1973; Jacobs 8c Daughaday 1969), acromegaly (Hardy 1969; Andrioli et al. 1970; Zatz et al. 1969), diabetes insipidus (Matissonn 8c Pimstone 1973), primary hypothyroidism (Neelon et al. 1973), Turner's syndrome (Odell 1973) and virilization (Shore et al. 1974). This study reports the investigation of 12 cases of primary empty sella. The conclusions are the following: 1) intrasellar arachnoidal diverticula have no notable consequence on pituitary function, 2) when endocrine symptoms do exist there is either an associated micro-adenoma or some other pathological disorder. difficult to demonstrate
MATERIALS AND METHODS
1) Neurological exploration Although the definition does not necessarily imply an increased sellar size, the sella turcica in the 12 cases considered had a volume of more than 1092 mm3 as assessed by the procedure of Di Chiro Se Nelson (1962). In pneumoencephalography the arachnoidal diverticulum filled the sella almost com¬ pletely or extended far below the limits of the sellar diaphragm as described by the land marks of McLachlan et al. (1970). Thus the simple intrasellar cisterns, minimal subarachnoidal expansions which develop in the anterior and superior part of the sella but which are not likely to push back or crush the hypophysis were eliminated.
2)
Endocrine
exploration
The study of posterior pituitary function was carried out only in cases with polyuria and polydipsia. The deficiency in anti-diuretic hormone was checked indirectly by measurement of urinary and plasma osmolarities before and after water restriction and intramuscular injection of 5 U of lysine-vasopressin (LVP) according to the procedure of Miller et al. (1970). Growth hormone (GH) level was determined by double antibody radioimmuno¬ assay (Click et al. 1963) at times 0, 30, 60 and 90 min of an insulin-induced hypoglycaemia (0.1 U/kg intravenously). Basal normal values are: 2.1 + 0.9 ng/ml (mean + se). A GH increase of more than 5 ng/ml was considered as a normal posi¬ tive response. The thyroid function was assessed by determining the free thyroxine index (FTI),
the normal range being between 3.5 and 5.5 (Savoie & Massin 1970). TSH was measured by radioimmunoassay under basal conditions (mean + se 1.17 + 0.29 ng/ml) and after iv injection of 200 /eg of synthetic thyrotrophin releasing hormone (TRH) (Odell et al. 1967). The mean peak value in normal subjects is 3.27 + 1.38 ng/ml; the conversion factor from ng/ml to pU/ml being five. The adrenal function was studied by measuring plasma 11-deoxycortisol (S) and cortisol (F) by competitive protein binding assay (Guillemant et al. 1969). A stimula¬ tion by metyrapone (4.5 g per os/24 h) and lysine-vasopressin (LVP Sandoz 10 U im) was performed in each case. A cortisol increase after LVP of more than 16 /eg/100 ml and S after metyrapone of more than 8 /eg/100 ml was considered as a normal re¬ =
sponse. was investigated according to sex by radioimmunoassay of (Oe,>) or plasma testosterone (T) (Castanier Se Scholler 1970; Mowszowicz Se Dray 1967); normal values being respectively 58-228 pg/ml in follicular phase, and 2.8-9.9 ng/ml. Pituitary reserve in gonadotrophins was determined by synthetic LH-releasing hormone (LH-RH) given intravenously at a dose of 100 /eg (in the follicular phase of pre-menopausal patients who were not amenorrhoeic). Blood samples were taken at -15, 0, 15, 30, 60 and 120 min. LH and FSH were measured by radioimmunoassay expressed as mIU/ml 2nd IRP HMG (Franchimont 1969). Normal basal values are the following: 1.92 ± 0.16 mIU/ml for LH, and 1.77 ± 1.6 mIU/ml (mean ± se) for FSH. After LH-RH injection the maximal peak is respectively 18.4 ± 1.6 mIU/ml
Gonadal function
oestradiol
and 3.98 ± 1.26 mIU/ml. Prolactin was measured by double antibody radioimmunoassay by Dr. L'Hermite 0, 15, 30, 60 and 120 min after injection of TRH (200 /eg iv) (L'Hermite Se Midgley 1971): results were expressed in ng of human prolactin H. Pr. V.-L.-S. No. 1 per ml of serum. The basal rate varies between 2 and 25 ng/ml. A twofold increase in serum prolactin after TRH stimulation was considered to be a normal response (Jacobs Se
Daughaday 1969). RESULTS
Clinical study (Table 1) Studies were carried out on 2 men and 10 women whose ages ranged from 18 to 60 years. Four cases were overweight. Thus the definite female pre¬ dominance and the frequency of obesity were confirmed. But none of our female patients had had more than three children. Headache, although often the presenting sign was noted in only 2 cases. In 3 cases the first clinical manifestation was some visual disorders such as bitemporal hemianopsia or optic atrophy on one side, with a temporal scotoma on the other. No rhinnorrhoea or hypertension was observed in our patients. Five patients had no evidence of pituitary disorder. (Group A, case 1 to 5), but this was not the case in the other patients (group B, case 6 to 12). Four patients had an amenorrhoea-galactorrhoea syndrome and in cases 9 and 10 the incompletely "empty sella" and biological explorations suggested an asso¬ ciated prolactin producing microadenoma. Transsphenoidal adenomectomy was
1)
Table 1.
Clinical characteristics of Sex
Age
M M F F F
Case 6
patients
with
primary empty
sella.
Pregnancies
Pituitary dysfunction
Associated disorders
60 26 51 45 35
2
0 0 0 0 0
headache visual disturbances visual disturbances
F
35
3
7 8 9
F F F
23 29 29
0 3 0
amenor.-galactorrhoea obesity diabetes insipidus 0 amenor.-galactorrhoea diabetes insipidus obesity amenor.-galactorrhoea prolactin producing
10
F
29
0
amenor.-galactorrhoea prolactin producing
11
F
18
0
hypopituitarism
microadenoma cranial trauma visual disturbances
12
F
23
0
hypopituitarism
meningo-encephalitis
Group
A Case 1 2 3 4 5
Group
1 0
obesity obesity, headache
B
microadenoma
carried out and pathologic finding confirmed the diagnosis. In cases 6 and 8 there was diabetes insipidus due to deficiency of antidiuretic hormone. No histiocytosis or sarcoidosis could be detected. Cases 11 and 12 showed a picture of panhypopituitarism. But this picture intervened a few months after a meningoencephalitis in case 12 and immediately following a severe cranial trauma in case 11. In this latter case there was possibly a necrosis, secondary to a pré¬ existent pituitary adenoma.
2) Biological study (Tables 2 and 3) a) Group A, patients with no evidence of pituitary disease. No abnormality of pituitary function was detectable in five patients (cases 1-5). Mean (± se) basal GH levels were 2.1 ± 0.9 ng/ml and increased to 8.9 ± 4.1 ng/ml in insulin induced hypoglycaemia. Case 5 had impaired GH release (5.2 ng/ml) but was 60 kg overweight. Free thyroxine index was normal (mean 4.6 ± 0.6). Basal serum TSH was 1.0 ± 0.1 ng/ml and reached a peak level of 3.56 + 1.15 ng/ml after TRH. In case 3 without clinical symptoms of hyperthyroidism the TSH peak was im¬ paired (1.65 ng/ml) but FTI was 6.8. -
pa Oh 00
"
rt
rt" -^ GO CO •-* Vf) OS CO CM 00 CM CM CO
©
S
£03
>• U Dh
H
S 'S
X
CM
m
Ö 1-Î
I"*
«-
de la Pitié, 75013
Paris -
THE PRIMARY EMPTY SELLA
AN ENDOCRINE STUDY ON 12 CASES
By G. Schaison and
J. Metzger
ABSTRACT Twelve patients (10 women and 2 men) with a primary empty sella turcica were studied. Endocrine function tests were performed as follows: growth hormone (GH) was measured after insulin-induced-hypoglycaemia, luteinizing hormone (LH) and follicle-stimulating hormone (FSH) after LH-releasing hormone, thyrotrophin (TSH) and prolactin after thyrotrophin-releasing hormone; pituitary reserve of adrenocorticotrophin (ACTH) was determined by measurement of plasma cortisol after lysinevasopressin and 11 deoxycortisol after metyrapone. Five of the patients (group A) had no endocrine disturbance. Seven patients (group B) had a hypothalamo-pituitary disorder. Two of them had panhypopituitarism which appeared in one case after meningoencephalitis and in the other after a severe cranial trauma. In two cases an amenorrhoea-galactorrhoea syndrome with increased prolactin level (68 and 230 ng/ml) led to a diagnosis of a prolactin producing adenoma, which was confirmed by surgery. Finally three cases of amenorrhoeagalactorrhoea, with normal prolactin level, and/or diabetes insipidus remained unexplained. However, no causal relationship could be demonstrated between the pituitary disturbance and the "empty sella". Primary empty sella turcica is therefore a neuroanatomical and neuroradiological entity with no endocrine implication. A pituitary disorder might suggest a microadenoma or an incidentally associated disease.
The term primary empty sella is used to denote intrasellar arachnoid invagina¬ tion of the chiasmatic cistern probably due to incomplete congenital development of the sellar diaphragm (Busch 1951). It also includes intrasellar arachnoidal cyst and even spontaneous necrosis of pituitary adenomas which are often
(Caplan et al. 1969). This definition however exludes sella empty secondary to radiotheraphy or surgical removal of the hypophysis Ras (Weiss 8c kind 1969). Expansion of the subarachnoidal space leads to a reshaping of sellar anatomy, the hypophysis being pushed back into the posterior-inferior part of the sella. The question arises however as to whether pituitary insult is sufficient to induce endocrine disease. In spite of many publications however, this problem remains unresolved. Apparently everything has already been described: no concurrent endocrine disorder at all, panhypopituitarism (Caplan et al. 1969; Neelon et al. 1973), GH, LH or ACTH deficiency (Neelon et al. 1973; Brisman et al. 1972; Faglia et al. 1973), amenorrhoea associated with galactorrhoea (Gates et al. 1973; Jacobs 8c Daughaday 1969), acromegaly (Hardy 1969; Andrioli et al. 1970; Zatz et al. 1969), diabetes insipidus (Matissonn 8c Pimstone 1973), primary hypothyroidism (Neelon et al. 1973), Turner's syndrome (Odell 1973) and virilization (Shore et al. 1974). This study reports the investigation of 12 cases of primary empty sella. The conclusions are the following: 1) intrasellar arachnoidal diverticula have no notable consequence on pituitary function, 2) when endocrine symptoms do exist there is either an associated micro-adenoma or some other pathological disorder. difficult to demonstrate
MATERIALS AND METHODS
1) Neurological exploration Although the definition does not necessarily imply an increased sellar size, the sella turcica in the 12 cases considered had a volume of more than 1092 mm3 as assessed by the procedure of Di Chiro Se Nelson (1962). In pneumoencephalography the arachnoidal diverticulum filled the sella almost com¬ pletely or extended far below the limits of the sellar diaphragm as described by the land marks of McLachlan et al. (1970). Thus the simple intrasellar cisterns, minimal subarachnoidal expansions which develop in the anterior and superior part of the sella but which are not likely to push back or crush the hypophysis were eliminated.
2)
Endocrine
exploration
The study of posterior pituitary function was carried out only in cases with polyuria and polydipsia. The deficiency in anti-diuretic hormone was checked indirectly by measurement of urinary and plasma osmolarities before and after water restriction and intramuscular injection of 5 U of lysine-vasopressin (LVP) according to the procedure of Miller et al. (1970). Growth hormone (GH) level was determined by double antibody radioimmuno¬ assay (Click et al. 1963) at times 0, 30, 60 and 90 min of an insulin-induced hypoglycaemia (0.1 U/kg intravenously). Basal normal values are: 2.1 + 0.9 ng/ml (mean + se). A GH increase of more than 5 ng/ml was considered as a normal posi¬ tive response. The thyroid function was assessed by determining the free thyroxine index (FTI),
the normal range being between 3.5 and 5.5 (Savoie & Massin 1970). TSH was measured by radioimmunoassay under basal conditions (mean + se 1.17 + 0.29 ng/ml) and after iv injection of 200 /eg of synthetic thyrotrophin releasing hormone (TRH) (Odell et al. 1967). The mean peak value in normal subjects is 3.27 + 1.38 ng/ml; the conversion factor from ng/ml to pU/ml being five. The adrenal function was studied by measuring plasma 11-deoxycortisol (S) and cortisol (F) by competitive protein binding assay (Guillemant et al. 1969). A stimula¬ tion by metyrapone (4.5 g per os/24 h) and lysine-vasopressin (LVP Sandoz 10 U im) was performed in each case. A cortisol increase after LVP of more than 16 /eg/100 ml and S after metyrapone of more than 8 /eg/100 ml was considered as a normal re¬ =
sponse. was investigated according to sex by radioimmunoassay of (Oe,>) or plasma testosterone (T) (Castanier Se Scholler 1970; Mowszowicz Se Dray 1967); normal values being respectively 58-228 pg/ml in follicular phase, and 2.8-9.9 ng/ml. Pituitary reserve in gonadotrophins was determined by synthetic LH-releasing hormone (LH-RH) given intravenously at a dose of 100 /eg (in the follicular phase of pre-menopausal patients who were not amenorrhoeic). Blood samples were taken at -15, 0, 15, 30, 60 and 120 min. LH and FSH were measured by radioimmunoassay expressed as mIU/ml 2nd IRP HMG (Franchimont 1969). Normal basal values are the following: 1.92 ± 0.16 mIU/ml for LH, and 1.77 ± 1.6 mIU/ml (mean ± se) for FSH. After LH-RH injection the maximal peak is respectively 18.4 ± 1.6 mIU/ml
Gonadal function
oestradiol
and 3.98 ± 1.26 mIU/ml. Prolactin was measured by double antibody radioimmunoassay by Dr. L'Hermite 0, 15, 30, 60 and 120 min after injection of TRH (200 /eg iv) (L'Hermite Se Midgley 1971): results were expressed in ng of human prolactin H. Pr. V.-L.-S. No. 1 per ml of serum. The basal rate varies between 2 and 25 ng/ml. A twofold increase in serum prolactin after TRH stimulation was considered to be a normal response (Jacobs Se
Daughaday 1969). RESULTS
Clinical study (Table 1) Studies were carried out on 2 men and 10 women whose ages ranged from 18 to 60 years. Four cases were overweight. Thus the definite female pre¬ dominance and the frequency of obesity were confirmed. But none of our female patients had had more than three children. Headache, although often the presenting sign was noted in only 2 cases. In 3 cases the first clinical manifestation was some visual disorders such as bitemporal hemianopsia or optic atrophy on one side, with a temporal scotoma on the other. No rhinnorrhoea or hypertension was observed in our patients. Five patients had no evidence of pituitary disorder. (Group A, case 1 to 5), but this was not the case in the other patients (group B, case 6 to 12). Four patients had an amenorrhoea-galactorrhoea syndrome and in cases 9 and 10 the incompletely "empty sella" and biological explorations suggested an asso¬ ciated prolactin producing microadenoma. Transsphenoidal adenomectomy was
1)
Table 1.
Clinical characteristics of Sex
Age
M M F F F
Case 6
patients
with
primary empty
sella.
Pregnancies
Pituitary dysfunction
Associated disorders
60 26 51 45 35
2
0 0 0 0 0
headache visual disturbances visual disturbances
F
35
3
7 8 9
F F F
23 29 29
0 3 0
amenor.-galactorrhoea obesity diabetes insipidus 0 amenor.-galactorrhoea diabetes insipidus obesity amenor.-galactorrhoea prolactin producing
10
F
29
0
amenor.-galactorrhoea prolactin producing
11
F
18
0
hypopituitarism
microadenoma cranial trauma visual disturbances
12
F
23
0
hypopituitarism
meningo-encephalitis
Group
A Case 1 2 3 4 5
Group
1 0
obesity obesity, headache
B
microadenoma
carried out and pathologic finding confirmed the diagnosis. In cases 6 and 8 there was diabetes insipidus due to deficiency of antidiuretic hormone. No histiocytosis or sarcoidosis could be detected. Cases 11 and 12 showed a picture of panhypopituitarism. But this picture intervened a few months after a meningoencephalitis in case 12 and immediately following a severe cranial trauma in case 11. In this latter case there was possibly a necrosis, secondary to a pré¬ existent pituitary adenoma.
2) Biological study (Tables 2 and 3) a) Group A, patients with no evidence of pituitary disease. No abnormality of pituitary function was detectable in five patients (cases 1-5). Mean (± se) basal GH levels were 2.1 ± 0.9 ng/ml and increased to 8.9 ± 4.1 ng/ml in insulin induced hypoglycaemia. Case 5 had impaired GH release (5.2 ng/ml) but was 60 kg overweight. Free thyroxine index was normal (mean 4.6 ± 0.6). Basal serum TSH was 1.0 ± 0.1 ng/ml and reached a peak level of 3.56 + 1.15 ng/ml after TRH. In case 3 without clinical symptoms of hyperthyroidism the TSH peak was im¬ paired (1.65 ng/ml) but FTI was 6.8. -
pa Oh 00
"
rt
rt" -^ GO CO •-* Vf) OS CO CM 00 CM CM CO
©
S
£03
>• U Dh
H
S 'S
X
CM
m
Ö 1-Î
I"*
«-
