Postoperative jaundice: is anesthesia the cause?

Why examiners ask this

The surgeon calls on postoperative day three: the patient is yellow and wants to know whether the anesthetic did it. Examiners use this question to see whether you can build a differential in an organized way, pre-hepatic, hepatic and post-hepatic, rather than reflexively blaming or defending the volatile agent. It tests organization, application of knowledge, and the communication behind a possible drug reaction.

The 60-second answer

Decision points and the "why"

• Pre-hepatic: bilirubin overload — each unit of stored red cells contains a load of senescent cells that hemolyze, a load classically quoted as about 250 mg of bilirubin per unit; a large hematoma does the same. Hemolytic transfusion reaction and G6PD deficiency or sickle crisis also belong here. I check LDH, haptoglobin, a smear and a direct Coombs test.
• Hepatic: ischemic hepatitis — transaminases in the thousands, often 20–50 times normal, within 24–48 hours of an episode of hypotension or hypoxemia, with only a modest bilirubin rise, falling quickly once perfusion is restored.² I look in the record for the low pressure. Recovery is the rule if perfusion is restored.
• Hepatic: drugs — antibiotics (amoxicillin-clavulanate is the classic), azoles, acetaminophen, TPN and anti-epileptics are far more common culprits than anesthetics.
• Volatile hepatitis — an immune reaction to trifluoroacetylated liver proteins from oxidative metabolism; severe halothane hepatitis follows about 1 in 15,000 first exposures and about 1 in 1,000 repeat exposures, and isoflurane and desflurane cause it far less because they are metabolized much less.¹ Sevoflurane does not form trifluoroacetyl metabolites, though rare cases are reported. Risk factors are prior exposure (especially with fever or jaundice after it), obesity and older age; most reported patients are middle-aged women.¹ It appears 2–14 days after surgery, often with fever and, in up to 30%, eosinophilia.¹
• Post-hepatic: obstruction — a retained stone or a bile duct injury after cholecystectomy; ultrasound shows dilated ducts and the surgeon needs to know today.
• Sepsis and cholestasis — inflammatory cytokines impair bile transport; the bilirubin rises with a modest enzyme change and resolves with source control.
• Gilbert syndrome — fasting and stress unmask a mild unconjugated rise; it is benign and needs no workup beyond a normal liver panel otherwise.

The complication they will inject

The patient had a sevoflurane anesthetic 10 days ago and a desflurane anesthetic for a washout yesterday; today the AST is 1,800, there is fever, eosinophilia and a rash. The surgeon says "anesthesia hepatitis" in front of the family. I do not agree or argue in the hall. I exclude ischemia, viral hepatitis and other drugs; I involve hepatology and check for anti-TFA antibodies if available. I tell the family honestly what we know and what we are testing, document the reaction, report it, and recommend that future anesthetics avoid all halogenated volatile agents, sevoflurane included, and use a total intravenous technique.¹ If the INR rises or encephalopathy appears, this is acute liver failure, and I recommend early transfer to a transplant center.

Pitfalls that cost points

• Blaming or excusing the anesthetic before building a differential.
• Forgetting transfusion and hematoma resorption, the commonest causes.
• Missing ischemic hepatitis by not reviewing the intraoperative record.
• Delaying imaging when obstruction is possible.
• Failing to communicate with the family and to document for future anesthetics.

Rapid-why

Q: Why is halothane different from sevoflurane? A: Roughly 20% or more of halothane is metabolized by CYP2E1 to the trifluoroacetyl hapten;¹ sevoflurane is about 3–5% metabolized to hexafluoroisopropanol, which does not form it.

Q: Why does type 1 halothane injury differ from type 2? A: Type 1 is a mild, common, dose-related transaminase rise; type 2 is the rare immune-mediated fulminant hepatitis.

Q: Why is the pattern so useful? A: An R ratio, ALT over ULN divided by ALP over ULN, of 5 or more says hepatocellular, below 2 says cholestatic, and between is mixed.³

Q: When does ischemic hepatitis peak? A: Transaminases peak in 1–3 days and fall by at least half within 72 hours of restoring perfusion, normalizing in 5–10 days.²

Numbers to know

Links

• Cirrhosis and the patient with liver disease, Massive hemorrhage and transfusion, Transfusion reactions, TRALI and TACO (coming), Volatile agents and nitrous oxide: when and when not, Sepsis and septic shock
• ACG Clinical Guideline on Idiosyncratic Drug-Induced Liver Injury (2021)