Hepatotoxicity after Desflurane Anesthesia

A 65-yr-old woman underwent a left hemithyroidectomy for a thyroid adenoma. The patient had a history of hypertension, which was treated with 50 mg atenolol per day. Her medical history was notable for pneumonia at age 9, pancreatitis at age 52, and recurrent urinary tract infections. Previous surgical and anesthetic history included a tonsillectomy in 1935 for which no records were available; a post partum tubal ligation in 1952, for which she received nitrous oxide, oxygen, cyclopropane, and d-tubocurare; total abdominal hysterectomy and unilateral oophorectomy in 1969, under sodium thiopental, succinylcholine, d-tubocurare, scopolamine, nitrous oxide, oxygen, and cyclopropane; an exploratory laparotomy and ovarian cystectomy in 1976, for which she received sodium thiopental, succinylcholine, pancuronium, nitrous oxide, oxygen, and halothane; and elective cholestectomy in 1982, during which she received sodium thiopental, succinylcholine, pancuronium, nitrous oxide, oxygen, and halothane. Each of these anesthetics were uneventful. The halothane anesthetics in 1976 and 1982 lasted approximately 45 min each. There was no history of jaundice, and the patient had never received a transfusion of blood or blood products. There was no history of intravenous drug abuse, tattoos, acupuncture, or occupational exposure to hepatotoxins. Alcohol consumption was less than 10 g per month. Hives had developed after she took sulfamethoxazole and trimethoprim (Bactrim) 15 yr earlier. She was also allergic to iodine-based dyes. The only other medication was 600 mg Ibuprofen once per day. She was a homemaker who had raised two children, a son who had died in his 40s from what was described as complications of myasthenia gravis, and a daughter who was alive and well. The patient had worked for 5 yr in a manufacturing facility, where she was exposed on occasion to paint products. Her last exposure was 10 yr earlier. There was no family history of liver disease.

The left hemithyroidectomy was performed on September 8, 1994. The patient was 157 cm tall and weighed 71 kg. She was afebrile with normal vital signs. The patient received 2 mg midazolam intravenously for sedation approximately 45 min before transport to the operating room. After placement of a blood pressure cuff, an electrocardiogram, and a pulse oximeter, general anesthesia was induced with 100 mg propofol, 100 micro gram fentanyl citrate, 40 mg atracurium, and 130 mg lidocaine, all given intravenously before tracheal intubation. Anesthesia was maintained with a mixture of 37–39% Oxygen²in nitrous oxide with 3–7% desflurane and intravenous fentanyl. Continuous mechanical ventilation was used without positive end expiratory pressure. Except for a brief decrease in blood pressure (systolic in the mid-80s) just before surgical incision, the patient remained hemodynamically stable. This was treated with 5 mg ephedrine and 0.3 mg glycopyrrolate intravenously. The pulse rate was 55 beats/min at the time of hypotension and increased to 90 beats/min after ephedrine administration. Blood loss was minimal. No blood or blood products were given. The total surgical time was 70 min with an anesthetic time of 90 min. The patient was discharged from the hospital in good condition the following day.

On September 20, 1994, 12 days postoperatively, the patient experienced pruritus, malaise, nausea, and polyarthralgias. There was no fever. At the same time, a macular erythematous rash appeared over the buttocks and thighs. The patient noted that her urine was unusually dark. She received nizatidine for her nausea and 1% hydrocortisone cream for her rash the following day. By the 16th postoperative day, the patient complained of dermal jaundice and epigastric abdominal pain. She was admitted to the hospital.

Physical examination on admission revealed the patient to be afebrile with a blood pressure of 140/80 mmHg and a pulse rate of 90 beats/min. There was an irregular macular erythematous rash over the posterior thorax. Examination results of the head and face were normal with the exception of conjunctival and scleral jaundice. Cardiopulmonary findings were normal. Abdominal examination revealed a nontender liver, which was 8 cm in size by percussion. There was no splenomegaly. No ascites was present. The patient was neurologically intact, without evidence of asterixis or other evidence of encephalopathy. Admission laboratory values are listed in Table 1. Albumin was 3.6 g/dl, hemoglobin was 14.8 g/dl, leukocyte count was 3.9 x 10⁹/l, with platelets of 177 x 10⁹/l and ferritin 554 micro gram/l. Anti-HAV (IgG and IgM), hepatitis B surface antibody, hepatitis B surface antigen, hepatitis B core antibody, and hepatitis C antibody (second generation ELISA) were not detected. Although antinuclear antibody was not detected, antismooth muscle antibodies were detected to a titer of 1:40 dilutions.

On September 30, 1994 (postoperative day 22), the patient was transferred to a hepatology unit for further observation and potential liver transplantation. The physical examination was unchanged except for the presence of ascites. The patient's biochemical and hematologic data are summarized in T1-30. An ultrasound of the upper abdomen revealed diffuse hepatic parenchymal heterogeneity, with normal bile ducts and normal Doppler examination of the hepatic, portal, and splenic vessels. Repeat viral serology for hepatitis A and B (including epsilon antigen and antibody) and hepatitis C were negative. Cultures for cytomegalovirus were negative. Serology for Epstein-Barr and Herpes simplex viruses indicated past infection (anti-VCA/IgG 1:2560, IgM negative, anti-HSV IgG 1:160, IgM negative). By October 5, 27 days after surgery, the patient's liver function and symptoms were improving (T1-30), and she was discharged from the hospital the following day. By December 14 (postoperative day 97), the patient was symptom-free and had a nearly normal liver function. Her medications at this time consisted of spironolactone and furosemide for ascites.

A serum sample collected 48 days after surgery was screened by an enzyme-linked immunosorbent assay for antibodies that reacted to liver microsomes from halothane-treated rats as compared to reactivity to liver microsomes from untreated rats. This assay has been used previously for the detection of serum antibodies in patients with halothane hepatitis. The assay is based on the finding that only patients with halothane hepatitis have serum antibodies that react with one or more liver microsomal proteins that have been trifluoroacetylated (TFA) by the reactive trifluoroacetyl chloride (CF³COCI) metabolite of halothane. As seen in Table 2, compared to a group of 20 normal patients, the patient's antibody reactivity was significantly greater against microsomes from halothane-treated rats than microsomes from control rats. The reactivity with the TFA-proteins was markedly increased when 0.1% sodium deoxycholate (DOC) extracts of liver microsomes were used as test antigen in place of the intact liver microsomes (T2-30). This is due to the preferential extraction by DOC of those proteins from the liver microsomes targeted by the CF³COCI metabolite of halothane. .

Halothane, enflurane, and isoflurane, have been associated with jaundice, hepatitis, and death. Hepatotoxicity associated with halothane is most common and has been studied extensively. Halothane produces two types of hepatotoxicity. The first is a mild form seen in 20% of patients given halothane anesthesia and is characterized by nausea, lethargy, low grade fever, and mild transient elevations in liver amino transferase enzymes (ALT, AST). In contrast, a fulminant hepatic necrosis (halothane hepatitis) occurs in approximately 1/20,000 adult patients exposed to halothane and is characterized by markedly increased serum ALT, AST, and bilirubin concentration, hepatomegaly, hepatic encephalopathy, jaundice, and, often, death. The predominant histologic feature is acute hepatitis with centrilobular necrosis. A variety of risk factors have been identified, which are commonly associated with halothane hepatitis. These include obesity, female sex, middle age, and having multiple anesthetics over a short period. Compared to adults, relatively few cases of halothane hepatitis have been reported in children. .

Current evidence suggests that halothane hepatitis may have an immunologic basis. For example, repeat halothane administrations increase the incidence of hepatitis, with fever, rash, arthralgias, and peripheral eosinophilia often observed. Serum from patients with a clinical diagnosis of halothane hepatitis have been shown to contain circulating IgG antibodies directed against liver microsomal proteins that have been covalently modified by the reactive CF sub 3 COCI metabolite of halothane (Figure 1). In addition, it was found that halothane hepatitis patients' antibodies bind to epitopes consisting of the TFA hapten and structural determinants of the carrier proteins. Several of the TFA-proteins recognized by the serum antibodies of patients with halothane hepatitis have been purified and identified from rat liver microsomes. It also appears that similar TFA-proteins are found in the livers of patients who have been administered halothane. It is thought that these proteins are immunogenic in certain patients and lead to the production of specific antibodies, or possibly specific T cells, which may be the cause of the hepatitis.

Enflurane and isoflurane are oxidatively metabolized by liver cytochrome P-450 to form acylated liver protein adducts by mechanisms similar to that of halothane (F1-30). The adduct formed by isoflurane is expected to be identical to that formed by halothane. Similarly, desflurane could form a TFA-adduct. This raises the possibility that enflurane, isoflurane, and desflurane might cause hepatotoxicity by a mechanism similar to that of halothane, but at a lower incidence, because the degree of anesthetic metabolism appears to be directly related to the potential for hepatic injury. For example, 20% of administered halothane undergoes metabolism, and numerous cases of halothane-associated hepatotoxicity have been reported. Approximately 2.4% of enflurane is metabolized, with 15–24 cases of hepatotoxicity having been well described. Less than 0.2% of isoflurane undergoes metabolism, with five cases of hepatotoxicity having been reported. Desflurane strongly resists biodegradation, with only 0.01% being metabolized. Therefore, from the standpoint of an immune-mediated mechanism of hepatotoxicity, desflurane should represent the safest inhalation anesthetic because it would lead to the lowest level of adduct formation. However, because antibodies reacting with liver microsomal TFA-proteins were detected in the serum of the patient (T2-30), and her clinical symptoms of pruritus, rash, and polyarthralgias suggested that she may have had an immune-mediated reaction, it appears that very small amounts of adduct may be able to precipitate massive hepatotoxicity, particularly if the patient was previously sensitized against TFA-proteins. The desflurane patient had received halothane on two prior occasions. Similarly, halothane may be able to sensitize patients against protein adducts formed by other fluorinated inhalation anesthetics. .

It might be argued that Ibuprofen was a precipitating cause of this patient's hepatotoxicity. However, the evidence argues against this possibility. First, this dose of Ibuprofen (600 mg) is rarely associated with liver injury. Second, the patients' liver function test results were normal in January 1994, while taking Ibuprofen (T1-30). Third, the patient reinstituted her Ibuprofen after surgery, and she continues to take this medication without ill effects. This is analogous to a negative rechallenge test to Ibuprofen. Fourth, her liver function tests and clinical course continue to show improvement, despite her use of Ibuprofen. Finally and most importantly, the patient's antibody response is specific to patients with anesthetic-induced hepatitis and is not seen in patients with other forms of liver disease.

To our knowledge, this represents the first report of desflurane-associated hepatotoxicity. This patient demonstrated several features commonly observed in hepatotoxicity after exposure to fluorinated anesthetics. Liver injury developed 12 days after desflurane anesthesia. This temporal relationship between exposure and injury is consistent with that of patients experiencing liver injury after halothane anesthesia. Also, she was an older female with drug allergies and a history of multiple exposures to halothane. She experienced a rash, jaundice, and marked increases in liver transaminases. The negative serologies, the otherwise negative medical history, and the presence of serum antibodies that reacted with TFA-liver microsomal proteins support the diagnosis of desflurane-induced hepatitis in this patient. It is likely that either or both of the halothane anesthetics received in 1976 and 1982 may have sensitized this patient, such that subsequent reexposure to desflurane or possibly other fluorinated anesthetics may have precipitated hepatotoxicity. One report describes a case of halothane anesthetics given 28 yr apart resulted in hepatitis and death in a patient. The current case serves to remind the clinician that anesthetic-induced hepatotoxicity may occur after anesthesia with any of the fluorinated anesthetic agents.