Study demographics

A total of 178 studies were eligible for inclusion, comprising 427, 659 patients (Fig. 1). There were 52 retrospective studies (29%), 118 prospective studies (66%) and 9 randomised controlled trials (5%). One study included both retrospective and prospective patients. The number of studies reporting on transsphenoidal surgery over time has increased, from 14 studies in the years 1990–1999, to 36 studies in 2000–2009, and 129 studies in 2010–2021 (Table 1). North America was the continent with the most studies (67, 38%), whilst Europe and Asia had 50 and 48 studies respectively.

Fig. 1

PRISMA Flow Diagram demonstrating inclusion of studies

Table 1 Breakdown of studies which met the inclusion criteria by decade, number of patients and study typePatient demographics

The majority of studies reported descriptive characteristics (age: 165 studies, 93%; sex: 164 studies, 92%) (Table 2). Forty-five studies (25%) reported patient co-morbidities and forty-two studies (24%) reported if the patient had undergone previous pituitary adenoma surgery, respectively. Only 21 studies (12%) reported BMI and 16 studies (9%) reported patient ethnicity (Table 3).

Table 2 Summary of data elements reported related to patient demographics, tumour histopathology and number of histopathological subtypes reported in the studiesTable 3 Baseline variable reporting (absolute number and % of total studies) across two different eras (1993–1998 & 2016–2021)Tumour histopathology

The most commonly reported tumour histopathology was non-functioning (101 studies, 57%) and GH-secreting (106 studies, 60%) (Table 1). Thirty-six studies (20%) did not specify the tumour pathology. The studies that did not report a specific pathology often wrote ‘pituitary adenoma’, ‘functioning’ or ‘hormone secreting tumour’. Ninety-one studies (51%) reported 2 or more tumour histopathological subtypes in the series.

Clinical baseline variables

Pre-operative clinical presentation data included endocrine, ophthalmic, nasal, cognitive, or headaches (Fig. 2). Endocrine baseline variables were reported in 100 studies (56%; Fig. 2a). Evidence of hormonal hypersecretion was reported in 78 studies (44%) and hyposecretion in 31 studies (17%). Of the 101 papers reporting functioning adenomas, 37 studies (37%) reported hypersecretion and 29 studies (29%) reported hyposecretion. Of the 132 papers reporting non-functioning adenomas, 68 studies (52%) reported hypersecretion, while 22 studies (17%) reported hyposecretion. 88 studies included both functioning and non-functioning adenomas. 81 studies (46%) mentioned endocrine testing with 70 (39%) of them reporting or describing the specific test performed. 48 studies (27%) explicitly defined the hormone concentration. 13 studies (7.3%) reported patients who required pre-operative hormonal replacement due to one or more hormonal deficits. 23 (13%) studies reported patients who required medical suppression therapy pre-operatively.

Fig. 2

Summary of endocrine, ophthalmic, and radiological baseline variables. a Summary of endocrine baseline variables. b Summary of ophthalmic baseline variables. c Summary of radiological baseline variables

For ophthalmic baseline variables, visual fields were reported in 61 studies (34%) and visual acuity in 32 studies (18%; Fig. 2b). Visual field was reported slightly more frequently in studies reporting functioning adenomas (42 studies, 42%), compared to studies reporting non-functioning adenomas (46 studies, 35%). Visual acuity was reported in 22 studies (22%) reporting functioning tumours, and in 24 studies (18%) reporting non-functioning tumours. For visual fields, the specific modality used was only reported in 16 of the 61 studies (26%): Humphrey (13 studies), Goldman (2 studies), both (1 study). Use of the Snellen Chart to assess acuity was reported in 9 of the 32 studies. Diplopia was reported in 17 studies (10%).

Nasal baseline variables were reported in 13 studies (7%) and a specific measure was reported in 10 studies: SNOT-22 (4 studies), Sniffin’ Sticks (3 studies), BAST-24 (1 study), (UPSIT, 1 study), NSS (1 study), and SIT40 (1 study). Cognitive baseline variables were reported in 9 studies (5%) and included mental disorders, altered mental status, depression, or mood change. One study that reported cognitive presentation used a validated measurement tool – the Mini Mental State Exam. Headaches were reported in 45 studies (25%).

Imaging baseline variables

Preoperative imaging was reported in 132 studies (74%; Fig. 2c). MRI alone was the most reported imaging modality (89 studies, 67%), with 33 studies (25%) reporting both CT and MR imaging. CT alone was reported in 3 studies (2%). Only 12 studies (10%) detailed a specific pituitary imaging protocol. Designation of tumours as micro-/macro-adenomas was reported in 93 studies (70%). Of the studies that reported macroadenomas, 44 studies defined macroadenoma explicitly as > 10mm. Further specific radiological baseline variables reported included: tumour diameter (52 studies, 39%); tumour volume (28 studies, 21%); cavernous sinus invasion (53 studies, 40%); Wilson Hardy grade (25 studies, 19%); Knosp grade (36 studies, 27%; Fig. 2c). Eight studies (4%) reported the surgeon’s preoperative resection intention.

Other baseline variables reported

The transsphenoidal approach was used in all studies. The endoscopic transsphenoidal technique was used in 78 studies, microscopic technique in 58 studies but 58 studies did not report the specific technique used to resect the pituitary adenoma. 30 studies reported collaboration with Ear, Nose and Throat surgeons. 37 studies (21%) reported the use of pre/intra- operative adjuncts. Of these, 13 studies reported insertion of a pre-operative lumbar drain. Other intra-operative adjuncts utilised were intraoperative CT (3 studies); intraoperative MRI (14 studies); intraoperative Doppler (2 studies); and neuronavigation (13 studies).

Heterogeneity in baseline variables across two eras (1993–1998 & 2016–2021)

Comparing 1993–1998 with 2016–2021, study design was reported in 100% vs 86% respectively, patient demographics 92% vs 88% respectively, tumour pathology 85% vs 88% respectively, clinical presentation 69% vs 51% respectively, endocrine baseline variables 31% vs 54% respectively, pre-operative imaging 77% vs 62% respectively, and peri-operative variables 77% vs 87% respectively. These first and last five-year epochs were chosen to assess if there had been a change in reporting over time.