Ventricular Tachycardia Caused by Hyperkalemia after Administration of Hypertonic Mannitol

A 52-yr-old, 174-cm, 62-kg man with a subarachnoid hemorrhage was admitted to our hospital for clipping of an anterior communicating artery aneurysm. The patient had been in good health and had been administered no medication until the morning of admission, when he presented with a severe headache. He had a Glasgow Coma Scale score of 15, without neurologic abnormalities. Preoperative 12-lead electrocardiography, chest radiography, and routine laboratory evaluations showed no abnormalities.

The next day, the patient was premedicated with triazolam and pirenzepine. At the time of arrival to the operating room, he was fully alert. Electrocardiography showed a normal sinus rhythm with a rate of 55 beats/min, a normal QRS duration, and a normal QT interval. Normal saline was infused through a venous catheter. A catheter then was inserted into the radial artery, and arterial blood analysis while the patient was breathing room air showed a pH of 7.39, a partial pressure of oxygen (Po²) of 88 mmHg, a partial pressure of carbon dioxide (Pco²) of 41 mmHg, a bicarbonate concentration of 26 mm, a sodium concentration of 140 mm, and a potassium concentration of 4.8 mm. General anesthesia was induced with use of propofol and fentanyl, and muscle relaxation was produced with use of vecuronium. After tracheal intubation, the patient underwent ventilation to achieve an end-tidal carbon dioxide value of 22–25 mmHg. Anesthesia was maintained with use of 1–1.5% sevoflurane and 60% nitrous oxide in oxygen. Fentanyl and vecuronium were added as needed.

The patient was hemodynamically stable throughout induction and at the start of surgery. Mannitol, 60 g, was administered intravenously during 20 min after surgical incision. Approximately 1 h after the beginning of mannitol infusion, when the electrocardiograph showed peaked T-waves and a wide QRS complex, an arterial blood sample was obtained, which showed a pH of 7.49, an arterial oxygen tension (Pao²) of 174 mmHg, an arterial carbon dioxide tension (Paco²) of 27 mmHg, a bicarbonate concentration of 21 mm, a sodium concentration of 124 mm, and a potassium concentration of 6.8 mm. Infusion of 500 ml glucose, 10%, in water containing 20 U insulin was started. Five min later, VT at a rate of 150 beats/min occurred, and systolic blood pressure decreased to less than 50 mmHg. Lidocaine, 90 mg, was administered, followed by continuous infusion at 2 mg/min, and 10 ml calcium gluconate, 8.4%, and 5 U insulin with 20 ml glucose, 50%, were administered, which restored VT to sinus rhythm at a rate of 60 beats/min. Systolic blood pressure increased to 115 mmHg. At the time, electrocardiography and echocardiography showed no evidence of an acute myocardial ischemic event. Twenty min later, plasma potassium concentration was 6.3 mm. The width of the QRS complex gradually normalized. Surgery was completed without further incident. At the end of surgery, sodium and potassium concentrations were 136 mm and 5.0 mm, respectively. During the 230-min procedure, blood loss was estimated to be 100 ml, and 2,000 ml normal saline was infused. No potassium-containing solutions were administered. Urine output was 1,200 ml, and the color of the urine was clear and amber. Urinary potassium excretion was 60 mmol during surgery. The patient had an uneventful postoperative course. The next day, plasma potassium concentration was 4.9 mm and 12-lead electrocardiography showed no changes compared with preoperative electrocardiography. The patient was discharged on the eighteenth postoperative day, and there were no sequelae.

There are several contributing factors to the development of VT, such as ischemic heart disease, valvular heart disease, cardiomyopathy, drugs, and electrolyte imbalance. 5We do not think that cardiac disease or drugs caused VT because the patient had no diseases and had been administered no drugs. There was no evidence of an acute myocardial ischemic event when VT occurred. We believe that hyperkalemia was associated casually to VT because the characteristic electrocardiographic changes of hyperkalemia preceded VT, and plasma potassium concentration was high when VT occurred. In addition, VT did not recur when plasma concentration of potassium decreased. In this patient, VT occurred even though the degree of hyperkalemia was not marked. Hyponatremia is reported to increase the cardiac toxicity of hyperkalemia; however, this mechanism is incompletely identified. 6–8Hyponatremia after administration of mannitol may participate in the occurrence of VT caused by hyperkalemia. Attention should be given also to plasma sodium concentration during hyperkalemia.

The causes of hyperkalemia are divided into the following three categories: increased intake of potassium, decreased urinary excretion of potassium, and transcellular movement of potassium from cells into the extracellular fluid. 8In this patient, the first two were ruled out because of the intravenous administration of potassium-free solutions and urinary potassium excretion during surgery. There are some factors that contribute to the transcellular movement of potassium, which include an acute increase in plasma osmolality, acidosis, malignant hyperthermia, rhabdomyolysis, hemolysis, insulin deficiency, and drugs. 8Several investigators have shown that hyperosmolality induced by mannitol increases plasma potassium concentration. 1–4Plasma potassium concentration has been reported to increase until 2 h after hypertonic mannitol infusion, which is consistent with the time course of the occurrence of VT in this patient. 3Acidosis, malignant hyperthermia, rhabdomyolysis, and hemolysis were ruled out because of the lack of suitable clinical and laboratory findings. The patient did not have diabetes mellitus, and no drugs other than mannitol were administered before hyperkalemia occurred. Therefore, we believe that hyperosmolality after administering hypertonic mannitol is causative of hyperkalemia.

Dilutional acidosis has been proposed as a possible mechanism of hyperkalemia because of the transcellular movement of potassium after administration of hypertonic mannitol. 1Maintenance of steady acid–base status, however, does not prevent the increase in plasma potassium concentration. 2–4,8Two alternative mechanisms that shift potassium out of the cells after mannitol administration have been suggested. First, a solvent drag phenomenon that moves potassium-rich intracellular water into the hypertonic extracellular compartment through the water pores is involved. 2–4,8Second, the loss of water in the cells, caused by hyperosmolality in the extracellular fluids, increases the intracellular potassium concentration, which creates a favorable gradient for passive potassium exit through potassium channels. 8These mechanisms may have caused hyperkalemia after administration of hypertonic mannitol in this patient.

Administering mannitol in a dose of 1 g/kg caused hyperkalemia in this patient. A dose range of 0.25–1 g/kg mannitol is considered standard for reduction of intracranial pressure. 9A previous study demonstrated that an increase in plasma osmolality of 10 mOsm necessitates reduction of intracranial pressure, which can be generated by 0.25 g/kg mannitol. 10Because the extreme osmotic gradient produced by mannitol may cause electrolyte imbalance, we recommend administration of a smaller dose (0.25–0.5 g/kg) and measuring of plasma osmolality. When large doses of mannitol are given, careful monitoring of the electrolyte status is essential.

This case shows a potential risk of life-threatening ventricular tachyarrhythmia caused by hyperkalemia after administration of hypertonic mannitol. It is essential to observe changes in the electrocardiogram and in plasma electrolyte concentration after administering hypertonic mannitol.