Endocrine Abstracts

Introduction: Endonasal endoscopic transsphenoidal surgery (EETS) for a sellar or suprasellar mass poses potential complications, including transient or permanent hypopituitarism. Adrenocortical insufficiency is especially worrisome given its potentially life-threatening course, if untreated. Usual clinical practice includes administration of perioperative “stress doses” of steroids followed by long term steroid replacement until the hypothalamic-pituitary-adrenal axis (HPA) is reevaluated. High-dose steroid treatment carries potential side effects.

Objective: To determine if post-surgical morning serum cortisol levels can reliably predict development of long-term hypocortisolism (LT-hC), and the risk factors associated with post-surgery LT-hC.

Materials and methods: Retrospective review of patients who underwent EETS from January 2016- December 2017 in our hospital. Data on tumor size, histology, presurgical hormone deficits and baseline hormonal levels were recollected, and logistic regression analysis was performed to calculate the odds ratio (OR) for development of LT-hC. Patients with Cushing’s disease were excluded. Morning serum cortisol level was measured on postoperative day-3 (POD3), and replacement steroid therapy was initiated if deemed necessary. Diagnosis of LT-hC was established based on HPA-axis tests at follow-up. An area under the ROC (AUROC) curve was calculated to determine the cortisol level that best predicts the development of LT-hC.

Results: Forty patients underwent EETS for a sellar or suprasellar mass: 67.5% non-functioning adenomas, 20% functioning adenomas, 5% cysts, 5% craneopharyngiomas, 2.5% meningiomas,. A diagnosis of permanent post-surgical hypocortisolism was made in 11 patients (27.5%). Patients who develop LT-hC have masses >30 mm (mean 35±10 mm). The presence of a presurgical hormone deficit was associated with the development of LT-hC (P<0.05]; OR 31.99, 18.33 and 18.00 for central hypothyroidism, central hypogonadism and GH deficit, respectively). Conversely, patients who developed LT-hC were more likely to develop a new-onset post-surgical hormone deficit (OR 57.49 and 14.67% for central hypothyroidism and central hypogonadism, respectively). There were no patients with LT-hC who developed permanent diabetes insipidus. The AUROC curve analysis found that a POD3 morning serum cortisol level >11.2 ug/dl yields a 95.65% sensibility, a 58.82 specificity and a 90.91% negative predictive value for ruling out long-term hypocortisolism.

Conclusions: POD3 morning serum cortisol level can predict development of LT-hC. Levels >11.2 ug/dl reliably rule out LT-hC and may help to avoid over-treatment with steroid replacement in most patients who conserve HPA-axis integrity. Patients with large tumor size and other presurgical hormonal deficits have a higher risk of developing LT-hC after EETS.