Primary empty sella a retrospective study

Abstract

Introduction

Primary empty sella (PES) consists of herniation of the suprasellar subarachnoid space within the sella turcica in patients with no history of sellar pathology. Diagnosis is mostly made incidentally.

Objective

Describe the characteristics of patients with PES.

Methods

Patients with PES from 6 Argentine centers specialized in Neuroendocrinology diagnosed between 1997-2021 were included. Diagnosis was based in sellar magnetic resonance, pituitary functional evaluation by baseline hormonal measurement and visual function assessed by neuro-ophthalmological evaluation and campimetry.

Results

Sixty-three patients with a mean age of 52.3 ± 13.2 years and a predominance of female gender (79%, 4:1 ratio) were included. The average BMI was 32.7 ± 8.13 kg/m2 and 68% were overweight or obese. HTA was also found in 59%. Among women, the parity frequency was 78%: 92% had multiple pregnancies. The diagnosis was incidental in 22%. In symptomatic patients (n= 49), the reasons for consultation were headache in 61%, symptoms of endocrine dysfunction in 51% and visual disturbances in 33%: 43% of the patients presented multiple symptoms. MRI showed partial PES in 63% and total PES in 37% of the patients. In the biochemical evaluation, hypopituitarism was observed in 46%.

Conclusions

Although the diagnosis of PES can be made incidentally, in our series most of the patients presented symptoms. Most of them were multiparous women, obese and hypertensive patients. Biochemical evaluation should be carried out to investigate isolated or multiple pituitary deficits in order to indicate the corresponding replacement treatment.

CONCEPTOS CLAVE

Qué se sabe sobre el tema.

La silla turca vacía primaria (STVP) es una patología poco frecuente cuyo diagnóstico se realiza principalmente de forma incidental, pudiendo manifestarse con una amplia variedad de síntomas, como déficits pituitarios o alteraciones visuales.

Qué aporta este trabajo.

En nuestro trabajo describimos las características de los pacientes tratados en centros argentinos especializados en patología neuroendocrina, así como las manifestaciones más frecuentes.

Divulgación

Se analizaron 63 pacientes tratados en centros argentinos especializados en Neuroendocrinología con STVP entre 1997-2021. El diagnóstico fue realizado mediante resonancia magnética selar (RM), la evaluación funcional hipofisaria por medición hormonal basal y la función visual mediante evaluación neuro-oftalmológica y campimetría. Se analizaron las características de la población.

Introduction

Empty sella is the herniation of the suprasellar subarachnoid space into the sella turcica ⁽¹⁾ . The term was first used by Bush in 1951 ⁽²⁾ . It can be total, when more than 50% of the sellar volume is occupied by cerebrospinal fluid and the gland measures less than 2 mm, or partial if the occupation is less than 50% ⁽³⁾ . Likewise, it can be classified as primary empty sella, in patients without a history of previous sellar pathology and secondary, in patients with arachnoidocele as a consequence of hypothalamic-pituitary disease and/or its surgical, pharmacological and radiotherapy treatment; consequence of bleeding or spontaneous necrosis, as well as due to traumatic brain injury, infectious processes or autoimmune pituitary disease ⁽⁴⁾ .

The physiopathogenesis is not fully elucidated and the hypotheses include incomplete formation or incompetence of the sellar diaphragm, associated or not with intracranial hypertension and pituitary volumetric changes ^(1,5) . Conditions such as obesity, sleep apnea, arterial hypertension, pregnancy, and childbirth can intermittently increase intracranial pressure, but it can also present idiopathically ⁽⁶⁾ . It presents a frequency of 2-20% in autopsies and up to 35% in clinical series ⁽⁷⁾ , with a female predominance (4:1) ⁽³⁾ . The diagnosis is mostly incidental, although endocrine and/or neuro-ophthalmological involvement can be found with variable frequency, as well as some non-specific symptoms ⁽⁸⁾ .

The aim of this study was to describe the clinical, radiological, and hormonal characteristics of patients with diagnosis of primary empty sella (PES).

Materials and methods

We retrospectively analyzed data of 63 patients with PES from 6 Argentine centers specialized in Neuroendocrinology in Buenos Aires, Mar del Plata, San Juan and Catamarca diagnosed between 1997-2021. The principles of the Helsinki declaration were respected. Diagnosis was made by sellar magnetic resonance (MRI), pituitary functional evaluation by baseline hormonal measurement, and visual dysfunction by neuro-ophthalmological evaluation and campimetry. Patients with a previous history of hypothalamic-pituitary disease were excluded. We evaluate the reasons for consultation and potential risk factors (parity, obesity, blood pressure), the radiological pattern (total PES: occupation of >50% of the sellar volume by cerebrospinal fluid and partial PES: ≤50% sella occupation) and the frequency of endocrine dysfunction using thyrotropin -TSH-, free thyroxine -fT4-, basal cortisol, follicle stimulating hormone -FSH-, luteinizing hormone -LH-, testosterone/estradiol, prolactin -PRL-, insulin-like growth factor 1 - IGF-1. No dynamic tests were performed. We defined overweight as body mass index (BMI) greater than 25, and obesity as BMI greater than 30. Multiple pregnancies were defined as two or more pregnancies.

Results

We included 63 patients with an average age at diagnosis of 52.3 ± 13.2 years. Among them, 50 were women, with 79% prevalence of female gender (4:1). The average age of female patients was 51.2±13.2 and men 56±12.7 years. The patients had an average BMI of 32.7 ± 8.13 kg/m2 and 68% were overweight or obese (43 patients). Of those, 17 (40%) were overweight and 26 (60%) had obesity. Hypertension was found in 59% (37 patients). Among women, parity frequency was 78 % (39/50): 92% (36/39) reported multiple pregnancies (range 2-14).

The diagnosis was made incidentally in 14 (22%) patients. The reasons for consultation among the 49 symptomatic patients (78%) were: headache 61% (30/49), symptoms of endocrine dysfunction 51% (25/49) including amenorrhea and decreased libido, and visual disturbances 33% (16/49); 43% of patients (21/49) had more than one symptom. The ophthalmology evaluation did not evidence visual field defects in none of the patients. If headaches were excluded, incidental diagnosis would have reached 60% of patients.

The MRI showed partial PES in 63% (40 patients) of the patients and total PES in 37% (23 patients).

Biochemical evaluation revealed hypopituitarism in 46% (29/63), with female prevalence of 72% (21/29). Sixty nine percent (20/29) presented isolated hormonal deficiencies and 31% (9/29) combined hormonal deficiencies (6 with panhypopituitarism). Overall, central hypogonadism was reported in 76% (22/29), secondary adrenal insufficiency in 41% (12/29), and central hypothyroidism in 28% (8/29). One patient presented with ADH deficiency. IGF-1 decreased for age and sex was found in 34% (10/29). No dynamic tests were performed to assess isolated decreased IGF-1. Of the 10 patients who presented decreased IGF-1, only 4 were associated with panhypopituitarism confirming the diagnosis of GH deficiency.

Additionally, hyperprolactinemia was detected in 21%, 8 women and 5 men, with an average PRL level of 60.2 ng/ml (Range 20-116 ng/ml). Of these 8/13, 61% (4 men and 4 women) concomitantly presented with central hypogonadism.

In patients with at least one axis deficit (n = 29), only 9 (31%) had evidence of total PES, the remaining had partial PES. Of the 6 patients with panhypopituitarism, 4 presented total PES.

Hormone replacement treatment was required in 72% (21/29). Only 3/13 patients with hyperprolactinemia received treatment with cabergoline.

Discussion

The PES epidemiology is variable according to the bibliography analyzed. Its diagnosis has been rising given the appearance of MRI ⁽⁵⁾ . In autopsies, it is described as an incidental finding in 2-20% of cases ^(2,8,9) . However, in imaging analysis, the prevalence can be up to 35% ^(10,11) . In our population, we observed a diminished number of incidentally diagnosed cases, probably attributed to referral bias, considering the specialized nature of the centers involved. A higher prevalence was observed in multiparous women, obese and hypertensive patients, known risk factors. The pituitary size can increase during pregnancy, mainly in multiple pregnancies, contributing to the development of PES afterwards, which would explain the higher prevalence in this population ⁽¹¹⁾ . We observed a predominance of multiparous females, similar to that described in the literature ^(5,11) .Regarding obesity, it is believed that the hypercapnia induced by morbid obesity can lead to increased intracranial pressure or the development of pseudotumor cerebri ^{( 12} ) . The relationship between PES, obesity and arterial hypertension is not fully understood. It is believed that it is associated with increased intracranial pressure and hypercapnia; increased intra-abdominal pressure participates, producing an increase in cardiac filling, obstructing venous return from the brain, and increasing blood pressure ^(13,14) as well as the proinflammatory state generated by cytokines that activate the 11β-HSD1 enzyme, increasing local cortisol production with the consequent increase in cerebrospinal fluid production ⁽¹⁵⁾ . De Marinis et al. showed 73% of patients were overweight, and most of the patients with BMI over 30 kg/m2 also had respiratory abnormalities such as sleep apnea syndrome ⁽¹¹⁾ .

Analyzing other series, we noticed that the diagnosis was mainly incidental, however, in our population most patients presented symptoms ⁽⁵⁾ . This may be due to a referral bias, since the patients were all evaluated in neuroendocrine centers, and probably the coexistence of empty sella and symptoms was a reason for consultation. The most frequent symptom was headache, followed by some degree of endocrine dysfunction and visual disturbances. If headaches were excluded, due to non-specific nature, asymptomatic diagnosis predominated. The literature reported, PES has been associated with visual disturbances such as oculomotor nerve defects, optic neuritis, and papilledema associated with compression of the optic nerve and/or the optic chiasm ⁽¹⁶⁾ . None of these could be evidenced in our population, as long as ophthalmologic evaluation was normal among our patients.

Regarding endocrine involvement, hypogonadism was the most frequent affection, presenting as oligomenorrhea in women and symptoms of sexual dysfunction and decreased libido in men.

Some studies describe the somatotropic axis as one of the most affected ⁽¹⁷⁾ . In our case, given that only 4 of 10 patients had GH deficiency confirmed, it is possible that this number would have been higher if dynamic tests were done.

Table 1 summarized our data and data extracted from another recent series of PES published by Ekhzaimy and col ⁽⁵⁾ .

Tabla 1. Main characteristics of comparative studies.

	Ekhzaimy 2020	Leon 2022
Patients (n)	765	63
Women (%)	79.4	79.3
W:M ratio	04:01	04:01
Age (years, average ± DE)	48.4 ± 14.8	52.3± 13.2
Incidental diagnosis (%)	79	28.5
ADH deficiency (%)	6.2	1.6
GH deficiency (%)	0.5	16
TSH deficiency (%)	15.5	12.6
ACTH deficiency (%)	5	19
FSH/LH deficiency (%)	5	35
Hyperprolactinemia (%)	6.2	20.6
Panhypopituitarism (%)	0.8	8

Comparing our series with another study carried out in Argentina ⁽¹⁸⁾ , we observed similar prevalence in terms of age and sex, highlighting as differences an increase in the prevalence of partial PES (63 vs 31%) and a greater diagnosis of arterial hypertension in our population.

Conclusions

The PES has a wide variety of clinical presentations. Although the diagnosis of PES can be made incidentally, in our series most of the patients presented symptoms, regardless of the magnitude of radiological alteration (total or partial PES). A clear predominance was observed in multiparous women and in obese and hypertensive patients as described in the literature. We highlight the importance of biochemical evaluation to investigate isolated or multiple pituitary deficits and indicate the corresponding replacement treatment.