An Unusual Case of Transient Diabetes Insipidus After Prolonged Vasopressin Infusion

Abstract

Background: Vasopressin is a hormone produced in the hypothalamus and secreted by the posterior pituitary. Underproduction or resistance to vasopressin can lead to diabetes insipidus (DI) resulting in the imbalance of fluid status and serum sodium levels. Synthetic vasopressin is a common second line agent used in the critical care setting for its vasopressor effects with an added clinical benefit by increasing cerebral perfusion pressure. There have been reported cases of transient DI after discontinuation of vasopressin in neurosurgical patients, however only few cases have been reported in septic shock patients without neurological involvement. We present a case of transient DI during treatment for septic shock after attempts to wean off vasopressin infusion. Presentation: A 41-year-old female with history of type 2 diabetes mellitus, quadriplegia following a motor vehicle accident at age 11 was admitted to the surgical intensive care unit for urosepsis after placement of right ureteral stent. The patient was started on broad spectrum intravenous (IV) antibiotics and vasopressors norepinephrine and vasopressin. After 8 days of vasopressin treatment, with each effort to wean IV vasopressin, the patient was noted to have significant polyuria with urine output reaching as much as 800 mL/hr and increase in her serum sodium to a maximum of 148 mmol/L (n 145 mmol/L). During episodes of excessive diuresis, urine sodium reached 87 mmol/L, urine osmolality was 72 mOsm/kg (50-1400 mOsm/kg), and serum osmolality was 288 mOsm/kg (n 300 mOsm/kg). The patient received three doses of 1 microgram of desmopressin 24 hours apart while weaning off vasopressin. This helped slow excessive diuresis while also maintaining normal serum sodium levels. Urine output decreased to approximately 150 mL/hr after each administration of desmopressin. The serum sodium levels eventually stabilized between 135-140 mmol/L and urine output held steady around 200 cc/hr. The patient did not require additional desmopressin and was hemodynamically stable off all vasopressors. Discussion: The pathophysiology behind transient diabetes insipidus following vasopressin infusion is still unclear. It is known that in septic shock, there is depletion of vasopressin stores which suggests central DI. Yet, pharmacologic doses of vasopressin are speculated to downregulate V2 receptors in the renal distal convoluted tubules and collecting ducts which suggests nephrogenic DI. Our patient responded to desmopressin which indicates that she at least had a central component of DI. There is no consensus on the duration of vasopressin required to precipitate transient DI, but vasopressin infusion was administered for at least 24 hours in other cases prior to onset. We present a rare case of transient diabetes insipidus after prolonged vasopressin infusion that clinicians should be aware of in the critical care setting.