Diabetes insipidus (DI) is an uncommon perioperative complication that can occur secondary to medications or surgical manipulation and can cause significant hypovolemia and electrolyte abnormalities. We reviewed and evaluated the current literature and identified 29 cases of DI related to medications commonly used in anesthesia such as propofol, dexmedetomidine, sevoflurane, ketamine, and opioids. This review summarizes the case reports and frequency of DI with each medication and presents possible pathophysiology. Medication-induced DI should be included in the differential diagnosis when intraoperative polyuria is identified. Early identification, removal of the agent, and treatment of intraoperative DI are critical to minimize complications.

Diabetes Insipidus (DI) is characterized by polydipsia and polyuria due to an inability to autoregulate water balance. There are two general classes of DI; central and nephrogenic. In nephrogenic DI, there is a decreased responsiveness of the kidneys to arginine vasopressin (AVP). Central DI results from a deficiency of AVP secretion by the posterior pituitary. This can result in hypovolemia secondary to polyuria with hypernatremia, high plasma osmolality and low urine osmolality. Anesthetic-associated DI in the perioperative period is rare but is an emerging topic in the adult literature. 1 We report a case of DI in a pediatric patient associated with intraoperative dexmedetomidine administration. | C A S E REP ORTA 16-year-old, 88 kg male, with recurrent juvenile nasopharyngeal angiofibroma with extradural skull base involvement presented for embolization followed by endoscopic nasal resection. His surgical history included prior embolization and resection of the angiofibroma the previous year without any complications. His medical history included attention deficit hyperactivity disorder and obstructive sleep apnea. The patient underwent embolization for the recurrent angiofibroma under general anesthesia, where he was induced with propofol and maintained on sevoflurane with intermittent dexmedetomidine boluses, totaling 32 mcg, without polyuria.During the endoscopic nasal resection, which occurred the following day, the angiofibroma was successfully removed without disruption of the blood-brain barrier despite extensive involvement of the basisphenoid, left sphenoid floor, infraorbital nerve, greater palatine nerves and pterygopalatine ganglion. The patient received a propofol induction followed by maintenance of anesthesia with sevoflurane, propofol and remifentanil infusion (Table 1). During the first 2 h postinduction, urine output (UOP) was 7.9 ml/kg/h (700 ml) and 5.7 ml/kg/h (500 ml), respectively. During the second hour of the case, a dexmedetomidine infusion was started at 0.3 mcg/kg/h to provide additional analgesia. Following initiation of the dexmedetomidine, UOP ranged from 9.1 ml/kg/h (800 ml) to 14.8 ml/kg/h