Metabolic Acidosis Associated with a New Formulation of Propofol

A 21-yr-old, 70-kg woman with a left hemispheric arteriovenous malformation presented to the emergency room intubated and unresponsive after reporting headaches and increased right-sided weakness. She was admitted to the neurosurgery service and treated using phenytoin, mannitol, and dexamethasone, and ventriculostomy was performed. Cerebral angiography showed a diffuse, left parietal arteriovenous malformation with hemorrhage from the anterior communicating and middle cerebral arteries. The arteriovenous malformation was embolized, and the patient was treated during deep propofol sedation and hypothermia to 34°C, with reevaluation in 3 or 4 days. An infusion of the Baxter brand propofol was started at 75 μg · kg⁻¹· min⁻¹and increased to 150 μg · kg⁻¹· min⁻¹. Initial measurement of arterial blood gas showed a pH of 7.32, with a partial pressure of carbon dioxide (Pco²) of 42.7 mmHg (all values corrected to 37°C) and bicarbonate (HCO³) concentration of 22.3 mm (calculated). After 6 h, pH was 7.44, with a Pco²of 29, but the HCO³had decreased to 19.7. By 12 h, pH was 7.35, HCO³was 17.6, and a base excess of −6.1 mEq was measured. The patient continued to become increasingly acidotic and was treated with intravenous sodium bicarbonate. Additionally, she deteriorated hemodynamically, requiring 10 μg · kg⁻¹· min⁻¹dopamine for blood pressure support. The source of acidosis was unclear and thought to be related to hypothermia, hypovolemia, or sepsis. Renal function was normal, with a urine output less than 1 ml · kg⁻¹· h⁻¹and serial creatinine measurements within normal limits. Central venous pressure was maintained in the range of 8–10 mmHg. Arterial blood gas measurement was done approximately 48 h after initiation of the propofol infusion showed a pH of 7.12, a Pco²of 24, an HCO³of 7.9, and a base excess of −19.1. These abnormalities persisted despite aggressive use of sodium bicarbonate. Serum lactate measurement was not performed until 50 h after admission. This showed an increased level of 7.3 mm, which continued to increase to a level of 14 to 15 mm during the propofol infusion, decreased to 8.9 mm after discontinuation of the infusion, and again increased to 14.2 mm just before the patient died. Serum chloride concentration was consistently in the range of 107–115 mm. It was speculated that perhaps the new formulation of propofol could be contributing to the problem and the infusion was discontinued. Within 4 h, arterial blood gas improved and in 12 h pH was 7.37, PCO²was 24, HCO³was 17, and base excess was −4.5 (fig. 1). The patient also improved hemodynamically, and the administration of dopamine was discontinued. Unfortunately, the patient’s neurologic condition continued to deteriorate, followed by a recrudescence of hemodynamic instability (fig. 2). A pulmonary artery catheter was inserted, and pulmonary artery pressures were seen to rapidly increase and cardiac output to decrease, consistent with the patient’s overall decline. The patient was pronounced brain dead the next day.

Propofol has been used in the intensive care population as a short-acting sedative hypnotic agent administered easily by continuous intravenous infusion. At sedative dose rates (10–50 μg · kg⁻¹· min⁻¹), it has the advantage of rapid elimination despite prolonged administration, thereby allowing neurologic examinations at frequent intervals and facilitating the weaning of the patient from ventilatory support. Doses necessary for electroencephalographic burst suppression (100– 200 μg · kg⁻¹· min⁻¹) use substantial volumes of propofol, but still offer rapid neurologic examination after discontinuation. At the University of Mississippi School of Medicine, we have not experienced hemodynamic or metabolic derangement associated with use of the Diprivan brand of propofol.

Several difficulties have occurred with the use of propofol, both in the United States and abroad. Initially, there was significant concern regarding its lipid emulsion preparation and ability to support bacterial growth. 1After extensive education of healthcare professionals about the safe handling of propofol and performing the aseptic technique, this is less of an issue. The nature of the emulsion also has produced numerous reports of pain during injection and patient discomfort, 2which again has largely been resolved by addition of lidocaine and other local anesthetics by practitioners to the solution before administration. Most recently, the new preparation of propofol by Baxter has been criticized because of the use of sulfite preservative, 3which may pose a risk of allergic reactions in susceptible patients. This preparation has the advantage, however, of lower cost than the Diprivan brand in many markets and has been chosen by some hospital pharmacies on this basis. Propofol was criticized widely in the United Kingdom and Canada in the late 1990s as a sedative for children in the intensive care unit. 4–7Despite some studies that showed the safety of propofol, 8,9numerous reports of unexplained metabolic and lactic acidosis and cardiovascular collapse were published As a result, propofol sedation of children in the intensive care unit was associated with these complications. The children described were aged 3–12 yr. The use of propofol infusion for the sedation of children in the intensive care unit is controversial.

Diprivan brand 10of propofol emulsion has a pH of 7.5–8.0, whereas the Baxter 11formulation has a pH of 4.5–6.4. In the current case, the patient’s clinical change after discontinuation of the Baxter solution suggests that this formulation caused her metabolic acidosis. Contributing factors may have been hypothermia, mild hypovolemia, and administration of the greater, neuroprotective dose of propofol rather than the lesser dose used for routine sedation. We estimate that, at a electroencephalographic burst suppression dose of 150 μg · kg⁻¹· min⁻¹, the patient received approximately 3,000 ml propofol over 48 h. In the current case, lactate concentration measurements were not obtained until relatively late in the patient’s course, decreasing their diagnostic value. Similarly, pulmonary artery catheterization was not performed until the patient was moribund, but central venous pressure values obtained earlier suggested adequate vascular volume and cardiac performance. It is interesting to speculate about the formulation of propofol used in the children described in the UK case reports. It is reasonable to assume that an exogenous acid load would be tolerated more poorly in a pediatric population and would present the clinical picture described, but the actual formulation used in the UK cases is unclear. The acid load received from an unbuffered solution is usually insignificant, but may contribute to metabolic acidosis and multiple other factors, such as hypothermia and hypovolemia, in a critically ill patient. This is the first case report of metabolic acidosis associated with propofol infusion in an adult. In summary, we present a case in which severe metabolic acidosis developed in an adult patient after the initiation of an infusion of a new brand of propofol. The acidosis resolved after discontinuation of the infusion.