Predictors and incidence of central diabetes insipidus after endoscopic pituitary surgery

With the advent of minimally invasive endoscopic pituitary surgery, there has been concern that the technique may be associated with higher rates of complications such as diabetes insipidus (DI) than traditional approaches, particularly early in a center's experience. We report the incidence and predictors of diabetes insipidus in patients after endoscopic transnasal resection (minimally invasive pituitary surgery) of pituitary lesions.

Data were collected from hospital and clinic records on the first 119 consecutive patients undergoing endoscopic pituitary surgery at our center.

The rate of postoperative diabetes insipidus is low in patients undergoing minimally invasive pituitary surgery (permanent, 2.7%; transient, 13.6%). Factors associated with development of DI after minimally invasive pituitary surgery include Rathke's cleft cyst histology, intraoperative cerebrospinal fluid leak, and previous nonendoscopic lesion resection. Elevated serum sodium (>145 mmol/L) within the first 5 days postoperatively has a high sensitivity (87.5%), specificity (83.5%), and negative predictive value (99.5%) for permanent postoperative DI development.

Transitioning from microscopic to endoscopic pituitary surgery can be achieved with a low incidence of DI. An elevated serum sodium level in the first 5 postoperative days using standard monitoring can predict the chance of developing permanent DI. Patients having no elevated serum sodium measurements, defined as >145 mmol/L, in the first 5 days postoperatively will rarely, if ever, develop permanent DI, thereby validating short postoperative inpatient stays with minimal risk of readmission for DI management. Those with a single serum sodium measurement greater than 145 mmol/L have a 15% risk of developing permanent DI.