A Multicenter Cohort Study in Patients With Primary Empty Sella: Hormonal and Neuroradiological Features Over a Long Follow-Up

Objectiveprimary empty sella (PES) represents a frequent finding, but data on hormonal alterations are heterogeneous, and its natural history is still unclear. Our aim was to evaluate the pituitary function of patients with PES over a long follow-up.Designmulticenter retrospective cohort study enrolling patients referred between 1984-2020 to five Pituitary Units, with neuroradiological confirmed PES and a complete hormonal assessment.Methodswe analyzed hormonal (including basal and dynamic evaluations), clinical and neuroradiological data collected at diagnosis and at the last visit (at least 6 months of follow-up).Resultswe recruited 402 patients (females=63%, mean age=51.5 ± 16 years) with PES (partial, total, undefined in 66%, 13% and 21%, respectively). Hypopituitarism was present in 40.5% (hypogonadism=20.4%, hypoadrenalism=14.7%, growth hormone deficiency=14.7%, hypothyroidism=10.2%, diabetes insipidus=1.5%; multiple deficiencies=11.4%) and hypeprolactinemia in 6.5%. Interestingly, hormonal alterations were diagnosed in 29% of incidental PES. Hypopituitarism was associated with male sex (p=0.02), suspected endocrinopathy (p<0.001), traumatic brain injury (p=0.003) and not with age, BMI, number of pregnancies and neuroradiological grade. A longitudinal assessment was possible in 166/402 (median follow-up=58 months). In 5/166 (3%), new deficiencies occurred, whereas 14/166 (8.4%) showed a hormonal recovery. A progression from partial to total PES, which was found in 6/98 patients assessed with a second imaging, was the only parameter significantly related to the hormonal deterioration (p=0.006).Conclusionsthis is the largest cohort of patients with PES reported. Hypopituitarism is frequent (40%) but hormonal deterioration seems uncommon (3%). Patients need to be carefully evaluated at diagnosis, even if PES is incidentally discovered.

Abstract licence: CC BY

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.