Clinical utility of routine postoperative morning cortisol monitoring in detecting new hypothalamic-pituitary-adrenal axis insufficiency following endoscopic transsphenoidal surgery for sellar lesions

Objective: Hypothalamic-pituitary-adrenal (HPA) axis dysfunction is a well-documented complication of transsphenoidal craniotomy (TSC) for sellar lesions. The authors aimed to assess their multidisciplinary approach to the diagnosis and treatment of postoperative hypocortisolemia utilizing conservative screening methods.

Methods: The authors performed a retrospective review of 257 patients who underwent TSC for pituitary adenoma (PA) or Rathke cleft cyst (RCC) at the University of Southern California between 2012 and 2017. Patients with preoperative adrenal insufficiency, Cushing's disease, or < 3 months of postoperative follow-up were excluded. Patient demographics, pathology, tumor characteristics, and complications were recorded. Postoperative day 1 (POD1) morning serum cortisol was assessed in all patients. Hypocortisolemia on POD1 (serum cortisol < 5 μg/dl) prompted a 7 am cortisol level measurement on POD 2 (POD2). Clinical signs and symptoms of hypocortisolemia were consistently monitored. After two serum cortisol levels < 5 μg/dl, or one serum level < 5 μg/dl plus a high clinical suspicion for HPA dysfunction, high-risk patients received glucocorticoid supplementation.

Results: Data on 165 patients were included in the analysis; there were 101 women (61.2%) and 64 men (38.7%). Preoperative diagnoses included nonfunctional adenoma (n = 97, 58.7%), growth hormone-secreting adenoma (n = 37, 22.4%), RCC (n = 18, 10.9%), prolactinoma (n = 8, 4.8%), and other (n = 5, 3.0%). One hundred thirty-eight patients (63.0%) had either suprasellar extension or cavernous sinus invasion. POD1 hypocortisolemia was diagnosed in 8 patients (4.8%). Of these patients, 2 (1.2%) were clinically asymptomatic and had normalized POD2 cortisol levels. Six patients (3.6%) had clinical symptoms and POD2 cortisol levels confirming HPA axis deficiency. Of these 6 patients treated with early glucocorticoid replacement, 2 patients recovered HPA axis function during follow-up, making the incidence of new, permanent HPA axis deficiency 2.5%.

Conclusions: In the authors' institutional review, all patients warranting postoperative glucocorticoid replacement had both complicated surgical courses and associated clinical symptoms of hypocortisolemia. The authors' algorithm of withholding steroids until patients demonstrate clear evidence of postoperative hypocortisolemia is safe and clinically efficacious. Their data further suggest that routine postoperative cortisol screening may not be necessary following an uncomplicated operative resection, with gland preservation and the absence of clinical symptoms indicative of HPA dysfunction.