Hepatotoxicity after sevoflurane anaesthesia: a new twist to an old story

Editor—Anaesthetic-induced hepatic dysfunction is a complex diagnostic problem. Because of the use of multiple drugs that can cause liver dysfunction, it is often difficult to pinpoint the culprit. Diagnosis of inhalation anaesthetic-related hepatic injury is usually made after other causes have been excluded. Unlike other drugs of this class, sevoflurane has the least hepatotoxic potential as it is not metabolised to reactive intermediates that lead to the formation of hepatotoxic proteins.¹ Using a case of acute hepatitis after general anaesthesia as an example, we review the mechanisms of liver injury and suggest measures to avoid recurrence. We obtained written consent from the patient for publication and have reported this adverse event to our hospital pharmacy and the US Food and Drug Administration.

A 58-yr-old male with a BMI of 23.5 kg m⁻² was undergoing L5-S1 laminectomy under general anaesthesia. His anaesthetic history was significant for acute hepatitis after two previous spine surgeries. In both cases, his postoperative course was marked by elevated liver enzymes accompanied by nausea and anorexia. His medical history was significant for hypertension, hyperlipidaemia, exercise-induced asthma, and Factor V Leiden. His medications included enalapril, atorvastatin, enoxaparin, warfarin, pregabalin, ibuprofen, tramadol, and paracetamol. He had no history of jaundice, blood transfusions, tattoo, high risk sexual activity, injected drug use, or alcohol use.

He had previously received propofol, fentanyl, rocuronium, and sevoflurane for anaesthesia. He had been diagnosed as being allergic to propofol which had been incriminated as the offending agent during his previous admissions. Hence, we planned to use alternate anaesthetics such as etomidate for induction along with opioids and a neuromuscular blocking agent, with sevoflurane for maintenance.

His preoperative evaluation was unremarkable with all investigations, including liver enzymes, within normal limits. Preoperative vital signs were normal. On the day of surgery, anaesthesia was induced with midazolam, etomidate, and fentanyl, and tracheal intubation was facilitated with cisatracurium. Anaesthesia was maintained with sevoflurane, and ketamine infusion to decrease the requirement of an inhalation agent. Monitoring included an ECG, pulse oximetry, radial arterial line, and bispectral index to titrate the anaesthetic. The surgery lasted 2.5 h and intraoperative course was uneventful with no episodes of hypotension, hypoxaemia, or excessive blood loss. His trachea was extubated in the operating room.

Postoperative Day 1 was uneventful. On postoperative Day 2, the patient's serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), and alkaline phosphatase increased to 506, 724, and 153 U L⁻¹, respectively. His coagulation profile and bilirubin were normal. His ALT and AST values peaked on postoperative Day 3 at 1373 and 1309 U L⁻¹, respectively, and total bilirubin peaked at 2.3 mg dl⁻¹. Hepatitis and human immunodeficiency virus serologies were negative for acute infection. Alpha-1 antitrypsin phenotype and iron panel findings were in the normal range. An ultrasound of the liver and gallbladder was unremarkable. With conservative management, there was gradual improvement and his liver enzymes normalised by postoperative Day 9, after which he was discharged home.

Anaesthesia and surgery can be followed by liver function abnormalities. Anaesthesia-induced hepatotoxicity is a diagnosis of exclusion, so ruling out other causes is essential. Several factors can contribute to postoperative hepatic dysfunction, including reduced oxygen delivery to the liver as a result of hypoperfusion or hypoxaemia, transfusion, direct liver compression during surgery, viral hepatitis, use of hepatotoxic drugs, and pre-existing hepatic dysfunction. In this case, these factors were ruled out except for the use of sevoflurane.

The hepatotoxic potential of different volatile anaesthetics is halothane>enflurane>isoflurane>desflurane, and sevoflurane is considered the least hepatotoxic.¹ Except for sevoflurane, all other halogenated anaesthetics are metabolised by cytochrome P450 2E1 to produce trifluoroacetyl chloride which binds to hepatic proteins and elicits antibody and T-cell mediated responses.² Hexafluoroisopropanol, which comprises 85% of the organic metabolites of sevoflurane, has a low binding affinity for liver macromolecules and is therefore rapidly converted to glucuronidates that are excreted in the urine.³ Proposed mechanisms of hepatic injury induced by sevoflurane include production of compound A, however, with use of Amsorb^® Plus (Armstrong Medical, Coleraine, Northern Ireland), as in this case, its production is minimal. Compound A causes cellular damage by activating free radical metabolising enzymes, leading to severe injury on re-exposure,⁴ and increases in cytosolic free Ca²⁺ possibly leading to hepatocyte necrosis.⁵

An important factor to consider is the effect of repeat exposure of halogenated anaesthetics. The acyl metabolites of the volatile anaesthetics covalently bind to hepatic proteins and can serve as neoantigens, which may lead to sensitisation and idiosyncratic hepatotoxicity upon re-exposure to volatile anaesthetics.⁶ Our patient had two recent surgeries and was exposed to sevoflurane anaesthesia with both. About 95% of patients with anaesthetic-induced hepatitis have a history of fluorinated anaesthetic exposure, which is considered as a major risk factor.⁶

Our patient had been taking paracetamol 2.6 g day⁻¹ for 5-6 months. While the dose was <4 g day⁻¹, considered the safe upper limit, there have been case reports where hepatic dysfunction is reported at much lower doses.⁷ The mechanism for this dysfunction is speculative and could be because of depleted stores of glutathione with susceptibility to injury from paracetamol at usual therapeutic doses.

Various risk factors have been described for halogenated volatile anaesthetic-induced hepatotoxicity. These include re-exposure, middle age, obesity, female sex, renal failure, history of atopy or multiple drug allergies, induction of the cytochrome P450 system, and chronic ingestion of ethanol/paracetamol.⁸ Our patient was middle-aged with a history of past exposure to sevoflurane and chronic ingestion of paracetamol.

The unique aspect of this report is the nature of reproducible hepatic dysfunction after sevoflurane anaesthesia in a patient with no prior history of liver disease. While the diagnosis was made by exclusion, volatile anaesthetics should be replaced by other modes of anaesthesia for future surgeries. Our recommendations for this patient are use of regional anaesthesia when feasible, total i.v. anaesthesia with processed EEG monitoring, a medical alert bracelet in case of emergent surgery, and consideration of CYP 2E1 gene polymorphism testing.

Declarations of interest

The authors declare that they have no conflicts of interest.