Both desflurane and sevoflurane have individually been reported to induce hepatic dysfunction; however hepatic dysfunction after administration of both of them separately in a single patient has not previously been reported. As their metabolites differ in nature, we considered that it would be unlikely that their combined use would cause sensitization and induce hepatic dysfunction. We report on the first patient with reproducible liver dysfunction after sevoflurane and desflurane. This 54-year-old man sequentially received 3 anesthetics over a 1-year period. The first anesthetic was isoflurane, and the course was uneventful. The second anesthetic was sevoflurane, and this resulted in fever with chills and elevated aspartate aminotransferase (543 U/l) 17 days later. The third anesthetic was desflurane which resulted in a similar clinical picture after 17 days. The symptoms improved, and the serum transaminase level returned to normal after conservative therapy. The similar time interval between the operation date and the onset of hepatic dysfunction, after excluding other possibilities, made us highly suspicious that the hepatic dysfunction was induced by sevoflurane on 1 occasion and desflurane on the other. We suggest that inhaled anesthetics should be totally replaced by intravenous anesthetics for future operations in patients with such a diagnosis.
(Chang Gung Med J 2003;26:357-62)

Key words:
isoflurane, sevoflurane, desflurane, liver function, hypersensitivity.

A 54-year-old male, 157 cm tall, weighing 71 kg, and with an ASA physical status of II, was admitted to our surgical ward for excision of a soft tissue tumor in the posterior area of the right mandibular alveolar ridge. The operation was performed under general anesthesia, which was maintained with nitrous oxide and isoflurane and which lasted for 2 hours 15 min on August 3, 1999 (Table 1). He denied having any major systemic disease including hypertension or liver disease, had no history of drug allergies, blood transfusions, intravenous drug abuse, or exposure to hepatotoxin, and alcohol consumption was only social. Preoperative laboratory data showed mild anemia and normal liver function. The patient was discharged uneventfully 2 days after the surgery. He was readmitted about 3 weeks later because the pathologist reported a verrucose carcinoma.
Radical neck dissection and radical flap reconstruction were performed on August 30, 1999. The blood pressure on arrival in the operating theatre was 229/112 mmHg, and 35 mg labetalol was administered intravenously. Anesthesia was induced with fentanyl, 2.5% sodium thiopental, and sevoflurane and was maintained with sevoflurane in oxygen (0.5 l/min) for a duration of 14 hours without immediate morbidity. Vecuronium was used to facilitate tracheal intubation and for muscle relaxation. The heart rate and O2 saturation remained normal, and the electrocardiogram (ECG) revealed no abnormality throughout the operation. He was discharged uneventfully on September 15. Two days after discharge, he experienced fever with chills, and visited our emergency room on September 19, where impaired liver function with elevated aspartate aminotransferase (AST) (543 U/l) was noted (Table 1). An abdominal sonogram was performed on September 23, which revealed mild hepatomegaly and mild fatty liver. No chemotherapy was administered in this period. The symptoms spontaneously subsided in 5 days.
One year later, on August 22, 2000, this patient underwent a radical thyroidectomy and left neck dissection under general anesthesia for a left thyroid papillary carcinoma with left neck lymph node metastasis. Anesthesia was induced with fentanyl, 2.5% sodium thiopental, and atracurium, and was maintained with desflurane in oxygen (1 l/min). Hypertension was also noted upon arrival in the operating theatre (238/117 mmHg), during induction of anesthesia (190/105 mmHg), and at the end of the operation (230/105 mmHg). Nicardipine at 1 mg and 10 mg labetalol were administered intravenously, and the ECG throughout the entire period showed no abnormality. The operation lasted for approximately 4 hours. However, on September 8, 2000, 18 days after the operation, he again experienced chills and fever. He returned to the hospital 8 days later, and impaired liver function was revealed by the elevated AST, alanine aminotransferase (ALT), alkaline phosphatase (ALK-P), and g-glutamyltransferase (g-GT) (Table 1). Mild jaundice developed in the following days, and an abdominal sonogram performed on September 19, 2000 revealed a moderate fatty liver of normal size; hepatitis B surface-antigen (HBsAg) and hepatitis C virus-antibody (HCV-Ab) were negative. Toxic hepatitis secondary to inhalation anesthetics was highly suspected by the GI specialists, but the patient refused a liver biopsy. The symptoms improved with conservative therapy. The concomitant medications used during and after anesthesia are presented in Table 2, and the changes in AST and ALT levels during desflurane and sevoflurane anesthesia are shown in Fig. 1.
DISCUSSION

In the first episode of hepatitis, the causative factors were difficult to identify as the course was short, and the quick spontaneous recovery did not alert anyone. However, after the second episode of hepatitis, the similar time interval of 18 days between the operation and the onset of signs of hepatitis made us curious about whether the hepatitis was related to the anesthetics.
Common causes of hepatic dysfunction include hepatic trauma, shock-induced hepatic tissue hypoxia, viral hepatitis, or drug-induced toxic hepatitis. After excluding the possibilities of hepatic trauma or intraoperative shock-induced tissue hypoxia, we considered a virus, preoperative medications, or intraoperative anesthetics to be the possible causative agents.
Acute viral hepatitis B and C can be excluded as a cause of hepatic dysfunction in this patient due to the negative findings of HBsAg and HCV-Ab. As hepatitis can cause liver damage of varying severity, we could not completely exclude viral hepatitis solely from the pattern of serum enzymes in our patient, since hepatitis A, hepatitis D, or another rarer hepatitis may still have been possible. However, the possibility of viral hepatitis infections after both operations was very low.
In searching for the possible toxic agents, we focused on the common medications administered in the second and third operations. Acetaminophen was administered preoperatively in both operations. Although it is reported to induce hepatic dysfunction,(6) it was safely administered after the last attack of hepatic dysfunction with no reported side effects. No antibiotic (second operation: metronidazole and cephalothin; third operation: penicillin G) was administered during these 2 periods, and there is no literature reporting a relationship between these antibiotics and hepatitis.
Our patient had undergone 3 operations to the head and neck region within 1 year. Low-flow anesthesia with sevoflurane in 0.5 l/min of oxygen flow was administered for 14 hours in the second operation. Sevoflurane anesthesia with low oxygen flow at 0.5-0.8 l/min had been safely used for years in our hospital, and no liver-related side effects of low-flow sevoflurane anesthesia had been reported.(7) Therefore it seemed unlikely that the impaired liver function was related to the low-flow anesthesia per se.
Although this patient denied a history of hypertension, his blood pressure on arrival in the operating theatre for the second operation was high. During the third operation when anesthesia was maintained with desflurane, hypertension was also noted upon arrival in the operating theatre (238/117 mmHg), during induction of anesthesia (190/105 mmHg), and at the end of the operation (230/105 mmHg). Anyway, his heart rate and O2 saturation remained normal, and the ECG revealed no abnormalities throughout the operation. Furthermore, hepatic dysfunction induced by fluctuations in intraoperative blood pressure without cerebral or cardiac injury has not been reported.
Desflurane and sevoflurane are both fluorinated anesthetics, and they have been reported to induce hepatitis.(1-3) Desflurane is oxidatively metabolized by liver cytochrome P-450 to form TFA proteins, which may produce immuno-responses; thus there is the possibility of inducing acute hepatitis by a mechanism similar to that of halothane and isoflurane.(4) Desflurane-induced hepatic dysfunction has been reported with pre-exposure to halothane but not to isoflurane.(3) Although desflurane and isoflurane can only produce very low levels of TFA formation, this small amount of TFA is enough to induce massive hepatotoxicity, particularly if a patient has previously been sensitized against TFA proteins.(3) In the present case, it is believed that the TFA antibody which presented after exposure to isoflurane in the first anesthesia sensitized the liver and resulted in hepatic function impairment after exposure to desflurane at a later time. Although the presence of the anti-trifluoroacetyl antibody (anti-TFA) may provide strong support for our diagnosis,(8) we were unable to detect this antibody, since the specific laboratory kit was unavailable at that time in Taiwan.
The mechanism of sevoflurane-induced hepatitis is still not completely understood. The primary organic metabolite of sevoflurane is HFIP, which is readily and rapidly conjugated with glucuronic acid.(5) As HFIP seldom forms liver protein adducts, TFA-protein sensitization was not likely the a cause of hepatic dysfunction in this patient.
The similar time interval between the operation date and the onset of hepatic dysfunction with both the sevoflurane and desflurane anesthesia, after excluding other possibilities, made us highly suspicious that the hepatic dysfunction was induced by the fluorinated anesthetics, sevoflurane and desflurane.
The unique aspect of this report is the nature of reproducible hepatic dysfunction following prolonged sevoflurane or desflurane anesthesia in a patient with no prior history of liver disease. While the diagnosis following both anesthetics was made by exclusion, inhalation anesthetics should be totally replaced by intravenous anesthetics for future operations.

REFERENCES

1. Shichinohe Y, Masuda Y, Takahashi H, Kotaki M, Omote T, Shichinohe M, Namiki A. A case of postoperative hepatic injury after sevoflurane anesthesia. Masui 1992;41:1802-5. (Japanese)
2. Watanabe K, Hatakenaka S, Ikemune K, Chigyo Y, Kubozono T, Arai T. A case of suspected liver dysfunction induced by sevoflurane anesthesia. Masui 1993; 42:902-5. (Japanese)
3. Martin JL, Plevak DJ, Flannery KD, Charlton M, Poterucha JJ, Humphreys CE, Derfur G, Pohl LR. Hepatotoxicity after desflurane anesthesia. Anesthesiology 1995;83:1125-9
4. Sinha A, Clatch RJ, Stuck G, Blumenthal SA, Patel SA. Isoflurane hepatotoxicity: a case report and review of the literature. Am J Gastroenterol 1996;91:2406-8
5. Kharasch ED. Biotransformation of sevoflurane. Anesth Analg 1995;81:S27-38
6. Monto GL, Scheuer PJ, Hansing RL, Burroughs AK. Attenuation of acetaminophen hepatitis by prostaglandin E2. A histopathological study. Digest Dis Sci 1994;39: 957-60.
7. Goeters C, Reinhardt C, Gronau E, Wusten R, Prien T, Baum J, Vrana S, Van Aken H. Minimal flow sevoflurane and isoflurane anaesthesia and impact on renal function. Eur J Anaesthesiol 2001;18:43-50.
8. Njoku D, Laster MJ, Gong DH, Eger EI 2nd, Reed GF, Martin JL. Biotransformation of halothane, enflurane, isoflurane, and desflurane to trifluoroacetylated liver proteins: association between protein acylation and hepatic injury. Anesth Analg 1997;84:173-8.

From the Department of Anesthesiology, 1Department of Metabolism, 2Department of Gastroenterology, Chang Gung Memorial Hospital, Keelung.
Received: Aug. 22, 2002;
Accepted: Oct. 21, 2002
Address for reprints: Dr. Chung-Hang Wong, Department of Anesthesiology, Chang Gung Memorial Hospital. 222, Maijin Rd., Anle Chiu, Keelung, Taiwan 204, R.O.C.
Tel.: 886-2-24313131 ext. 2777;
Fax: 886-2-24313161;
E-mail: wong@cgmh.org.tw