Portal de Periódicos da CAPES

Propylene Glycol Toxicity Associated with Lorazepam Infusion in a Patient Receiving Continuous Veno-Venous Hemofiltration with Dialysis

2002; Lippincott Williams & Wilkins; Volume: 94; Issue: 6 Linguagem: Inglês

10.1213/00000539-200206000-00039

1526-7598

Ali Al‐Khafaji, W.E. Dewhirst, Harold L. Manning,

Alcoholism and Thiamine Deficiency

Propylene glycol is a solvent contained in many food and drug formulations. Toxicity, manifested by metabolic acidosis, central nervous system disturbances, and hyperosmolarity, has been reported in patients receiving large amounts of this chemical via IV and topical medications. Similar to ethylene glycol, propylene glycol and its oxidative metabolites undergo renal elimination. Although patients with renal insufficiency are at increased risk for toxicity, hemofiltration and dialysis should be most effective at clearing the chemical and its metabolites. We report a case of a patient who developed the features of propylene glycol toxicity from a continuous infusion of lorazepam while undergoing continuous renal replacement therapy. On cessation of propylene glycol administration, the clinical signs of toxicity abated and the chemical became undetectable in the plasma within 48 h. Case Report A 28-yr old male with a remote history of cadaveric renal transplant for poststreptococcal glomerulonephritis was admitted to the intensive care unit (ICU) suffering from acute pancreatitis. On initial presentation he was febrile to 39°C, with arterial blood pressure 90/60 mm Hg, his heart rate was 128 bpm, and his respiratory rate was 38 breaths/min. His admission laboratory values included blood urea nitrogen of 100 mg/dL, creatinine of 5.1 mg/dL, serum albumin 2.2 g/dL, and triglycerides of 399 mg/dL. The patient developed progressive respiratory failure and required endotracheal intubation and mechanical ventilation shortly after admission. He was sedated with fentanyl and lorazepam administered by continuous infusion, and piperacillin/tazobactam antibiotic therapy was begun. Despite modified antirejection therapy, attempts to optimize systemic perfusion, and avoidance of nephrotoxic drugs, his transplanted kidney failed and he became oliguric, presumably from the systemic inflammatory response to severe pancreatitis. Hemodialysis was begun on the fifth ICU day because of persistent azotemia and oliguria. On the twelfth ICU day he was started on renal replacement therapy by continuous veno-venous hemofiltration with dialysis (CVVHD) because of hypotension occurring during hemodialysis. He required progressively larger dosages of sedatives to control agitation and to achieve adequate synchrony with the ventilator, and by ICU day 30, he had been receiving infusions of fentanyl at 1350 μg/h and lorazepam at 22 mg/h for over 72 h. Despite hemodialysis and later CVVHD, the patient developed a progressive, increased anion gap metabolic acidosis. On ICU day 30, the following laboratory values were obtained: pH 7.22, Pco2 41 mm Hg, HCO3− 16 meq/L, anion gap 20 mmol/L, β-hydroxy-butyrate 0.1 mmol/L, phosphorus 2.7 mg/dL, blood urea nitrogen 67 mg/dL, and creatinine 1.2 mg/dL, glucose 144, serum albumin 2.0 g/dL, total bilirubin 9.9 mg/dL, alkaline phosphatase 330 U/L, aspartate transaminase (AST) 77 U/L, and alanine transaminase (ALT) 160 U/L. The plasma lactate level was 1.2 mmol/L and the serum osmolality was 344 mmol/L with an osmolar gap of 38. Propylene glycol toxicity was suspected because no alternative cause of an increased anion gap metabolic acidosis could be identified. Specifically, ketoacidosis, lactic acidosis, hyperphosphatemia, and other volatile alcohol toxicity were eliminated as potential causes by the laboratory studies presented. The lorazepam was discontinued and the patient was switched to midazolam infusion. The plasma propylene glycol level, drawn at the time the lorazepam was discontinued, was later reported to be 1308 mg/L. Within 48 h after the discontinuation of lorazepam infusion, the metabolic acidosis and anion gap had resolved, the osmolar gap decreased from 38 to 2, and there was no detectable propylene glycol level on repeat plasma assay. Unfortunately, the patient succumbed after a 10-wk course of unremitting multiorgan system failure. Discussion Propylene glycol is a colorless, odorless liquid with a molecular weight of 76.1 Daltons. It was initially thought to be innocuous (1) and is approved by the United States Food and Drug Administration for use as a solvent in food, drugs and cosmetics. Propylene glycol is used as a solvent vehicle for several IV drugs, including lorazepam, diazepam, etomidate, phenytoin, nitroglycerin, hydralazine, esmolol, phenobarbital, trimethoprim-sulfamethoxazole, and chlordiazepoxide. Approximately 50% of a dose of propylene glycol is excreted unchanged by the kidneys. The remainder is metabolized by hepatic alcohol and aldehyde dehydrogenases to glycolic acid and lactate and then subsequently converted to pyruvate or acetone. Thus, patients with hepatic or renal failure are at increased risk of toxicity. The manifestations of propylene glycol toxicity include central nervous system depression, seizures, cardiac arrhythmias, respiratory arrest, hemolysis, and renal failure (2). The presence of an increased anion gap metabolic acidosis (with or without lactic acidosis), hyperosmolarity with an osmolar gap, and intravascular hemolysis may all be clues to the diagnosis of propylene glycol toxicity (2). Treatment of toxicity consists of discontinuing the propylene glycol-containing preparation, administering IV fluids, and dialysis (3). Although fomepizole, an alcohol dehydrogenase inhibitor, has been used in ethylene glycol toxicity, there are no reports of its use in the treatment of propylene glycol intoxication. There have been prior case reports of propylene glycol toxicity associated with the infusion of nitroglycerine (4), lorazepam (2,5,6), diazepam (3), and etomidate (7). This is the first report of a patient who developed propylene glycol toxicity during renal replacement therapy. Although acute renal failure and continuing renal replacement therapy may complicate the interpretation of acid-base status, the appearance and resolution of otherwise unexplained, worsening anion gap acidosis and osmolar gap, coincident with the lorazepam infusion, implicates propylene glycol as the toxic agent in this patient. Renal replacement therapy was continuing and there was no evidence of recovery of allograft renal function during this time period. Dialysis (either intermittent hemodialysis or CVVHD) is effective in clearing small molecular solutes including many organic acids. Although its use is accepted as therapy for severe ethylene glycol intoxication, we are aware of no published reports describing treatment of propylene glycol intoxication by any form of dialysis. Because CVVHD is generally effective at clearing solute molecules of small size, such as propylene glycol, we were surprised that our patient developed propylene glycol toxicity while on CVVHD. One possibility is that CVVHD is less effective than expected in removing the drug. However, clearance of propylene glycol by CVVHD should be efficient owing to its small molecular size, high predicted sieving coefficient, and apparent volume of distribution that approximates body water. Nonetheless, it is evident from our patient that the maximal CVVHD clearance rate of propylene glycol can be exceeded during large-dose lorazepam infusion. It is possible, but not likely, that our patient received a sufficiently rapid rate of lorazepam infusion that propylene glycol toxicity would have occurred even in the presence of normal renal function. Although the maximal clearance rate of propylene glycol by normal humans is unknown, Speth et al. (8) found no evidence of lactic acidosis, hemolysis, or hyperosmolarity in subjects who received IV propylene glycol at 3 to 15 g/m2 over a period of 4 hours. Furthermore, based upon extrapolation from animal data, Morshed et al. (9) predicted a metabolic capacity of more than 1 g/day for a 70-kg human. At the time of the maximal infusion of lorazepam, our patient received propylene glycol at a rate of approximately 9 g/h (830 mg propylene glycol and 2 mg lorazepam per mL, at 11 mL/h), with an estimated cumulative dose of 108 g/m2 per day. Our report demonstrates that CVVHD does not preclude the development of propylene glycol toxicity during continuous infusion of large-dose lorazepam. However, it did appear to treat the toxicity effectively, once the infusion of propylene glycol was stopped. The kinetics of propylene glycol elimination during CVVHD are not published and therefore it is unknown what infusion rate is required to cause toxicity. We suggest that when patients require rapid rates of continuous infusion of drugs containing propylene glycol, the possibility of toxicity from this vehicle must be considered, even in the setting of continuous renal replacement therapy. Alternative drugs that do not contain propylene glycol should be used where possible, especially in the setting of concomitant renal and/or hepatic dysfunction. The authors would like to acknowledge Renee Hebert, RN and Mauri Schwartz for their help in preparing this case report.