Methimazole-induced cholestatic jaundice.

Abstract: Methimazole is a widely used and generally well-tolerated antithyroid agent. A 43-year-old woman had severe jaundice and itching 1 month after receiving methimazole (10 mg tid) and propranolol (20 mg tid) for treatment of hyperthyroidism. The patient continued treatment for another 4 days after the appearance of jaundice until she finished both medications. When seen at the emergency department 2 weeks later, she still had severe icterus, pruritus, and hyperbilirubinemia, formed mainly of the conjugated fraction. Methimazole-induced cholestasis was diagnosed, and propranolol therapy was resumed. Over the following 9 days, the symptoms improved and plasma bilirubin levels were normal after 12 weeks without methimazole. In rare cases within the first few weeks of therapy, this drug can cause severe and reversible cholestatic jaundice. Physicians and patients should be aware of this adverse effect so that, upon occurrence, they can discontinue methimazole therapy and avoid unnecessary invasive procedures.

**********

In 1949, Stanley and Astwood (1) first introduced methimazole (1-methyl-2-mercaptoimidazole), the most potent antithyroid agent. Since that time, few cases of severe cholestatic jaundice have been described in association with use of the drug. Surprisingly, despite the widespread use of methimazole for treatment of hyperthyroidism, little is known about this rare but serious adverse effect. In even the most detailed reviews (2) and specialized endocrinology textbooks, (3) information regarding various factors related to methimazole-induced cholestasis is lacking. Review of the literature (English, French, and Spanish) yielded 12 cases of cholestatic jaundice induced by methimazole (4-14) and 8 cases by its prodrug carbimazole, (15-22) which is rapidly and completely converted into methimazole in the serum. (2), (3) These cases are summarized in Tables 1 and 2. Patients with established concomitant liver diseases such as viral hepatitis, (23) those with disseminated cryptococcal infection involving the liver, (24) and one patient in whom hyperbilirubinemia was not documented, (25) were excluded. This report presents an additional case and analyzes the features that characterize methimazole-induced cholestatic jaundice.

Discussion

Among the 20 cases of methimazole-induced cholestasis previously reported, 18 occurred in women (Tables 1 and 2). This finding most likely reflects the predominance of hyperthyroidism in the female sex. Most cases of hyperthyroidism were due to Graves disease, but some were due to toxic nodular goiter. (4), (5), (10), (14) The most common symptoms were jaundice (18 cases), followed by itching (10 cases), dark discoloration of urine (7 cases), clay-colored stools (6 cases), fever (3 cases), and generalized weakness (2 cases). Uncommon symptoms were abdominal pain (1 case), diarrhea (4 cases), and bleeding manifestations (8 cases). Abdominal examination, including the liver, was either unremarkable or minimally tender. A latent period between drug exposure and the development of jaundice ranged from 2 days to 3 months (Tables 1 and 2). Abdominal imaging studies, whether invasive (11), (17) or noninvasive (11), (14) as in the present case, did not reveal any evidence of intrahepatic or extrahepatic biliary obstruction. In the case reported by Schwab et al, (14) however, endoscopic retrograde cholangiopancreatography (ERCP) showed rarefaction of the intrahepatic ducts, as seen in primary sclerosing cholangitis. These pathologic changes were reversible 3 months after methimazole therapy was stopped. (14)

Liver biopsy was performed in 17 cases. The most prominent pathologic finding was intrahepatic cholestasis. With one exception of a case characterized by severe inflammation and hepatic necrosis confirmed by liver biopsy, (19) cellular infiltration and signs of inflammation and necrosis were either mild (6-9), (12), (17), (18), (21) or absent. (10), (13), (16), (20)

Agranulocytosis, another rare serious side effect of methimazole, seems to be unrelated to cholestasis. It occurred in only three cases in association with jaundice (4-6) and was the cause of death in one patient. (6) Rechallenging with methimazole was reported in one patient (12) and with carbimazole in three patients. (16-18) In all instances, symptoms of cholestatic jaundice recurred.

In the present case, cholestatic jaundice was most likely caused by methimazole for the following reasons. First, there was a clear chronological relationship between the initiation of methimazole and the development of jaundice 1 month later and the resolution of cholestasis 12 weeks after drug withdrawal. These time periods are consistent with similar cases in the literature (Tables 1 and 2). Second, concomitant liver diseases, such as viral hepatitis, autoimmune hepatitis, and primary biliary cirrhosis, were excluded to a large extent by proper serology. Third, it is unlikely that the severe cholestasis in our case was related to hyperthyroidism. Cholestasis due to hyperthyroidism usually occurs in association with marked elevation of plasma levels of thyroid hormones. (26-29) In our patient, thyroid hormone values had already returned to normal range, while plasma bilirubin levels were extremely elevated. The thyrotropin level was still suppressed, since it lags behind the normalization of thyroid hormones by 4 to 6 weeks. (30) Finally, cholestasis is unlikely to be attributed to propranolol because jaundice resolved despite its reinitiation. Concerning amlodipine, the patient had been taking this antihypertensive agent for years, and use of this drug was not interrupted. Furthermore, cholestatic jaundice attributed to either propranolol or amlodipine has not been reported before.

Liver biopsy and rechallenge with methimazole were not done in the present case because the initial presentation was highly suggestive of methimazole-induced cholestasis and the patient's clinical status and laboratory findings were improving since admission (Table 3). In addition, both procedures are not without risks and significant discomfort to the patient. Therefore, it was believed that liver biopsy and drug rechallenge were neither warranted nor ethical. With respect to the management of the cholestatic jaundice, treatment with corticosteroids (6), (7), (9), (10) or corticotrophin (4) was attempted but did not seem to accelerate recovery. The outcome of methimazole-induced cholestasis was generally benign, with self-resolution of symptoms and normalization of bilirubin levels and other liver function values 5 days to 6 months after stopping methimazole or carbimazole therapy. In one patient (the second case reported by Jansen et al (11), however, methimazole therapy was not stopped, despite the increase in plasma total bilirubin to 7.3 mg/dl 20 days after starting methimazole. Hyperbilirubinemia continued to worsen in the following 4 weeks to reach a peak value of 14.4 mg/dl, then resolved about 2 months later while the patient was still receiving methimazole therapy. (11) The explanation in that case is unclear, but the patient had positive antinuclear antibodies consistent with autoimmune hepatitis. Therefore, the possibility of transient cholestatic jaundice as result of the underlying hepatitis rather than methimazole treatment could not be excluded.

The mechanisms of methimazole-induced cholestasis have not been fully elucidated. However, a number of observations suggest that it is most likely an allergic reaction. Thus, symptoms usually develop within the first few weeks after drug initiation and are reproducible within only hours or a few days upon drug rechallenge. (12), (16-18) In addition, the drug reaction is dose-independent. Furthermore, carbimazole has been shown to induce blast cell transformation of lymphocytes derived from affected patients in vitro (the lymphocyte transformation test). (15), (17) The latter test has been considered a marker of drug allergy or hypersensitivity. Nevertheless, an interaction of the drug with the liver altered by the hyperthyroid state could be a contributing factor.

There are some common features regarding liver toxicity caused by methimazole and the other major, structurally related, antithyroid drug, propylthiouracil. With both agents, the incidence of this serious reaction is rare, dose-independent, and appears to be allergic. (31) Meanwhile, some important differences exist. The latent period preceding the appearance of symptoms after starting the drug is more extended with propylthiouracil, ranging from 1 day (32) to 15 months. (33) Moreover, the liver disease associated with propylthiouracil is generally more severe, showing signs of severe inflammation and necrosis. (31) In fact, death due to fulminant hepatitis has occurred in several patients exposed to propylthiouracil. (31) On the other hand, fulminant hepatitis occurred in only one patient treated with carbimazole. (19) The patient died as a result of bleeding from a duodenal ulcer a few days after hepatic transplantation. (19) Abdominal ultrasonography had shown atrophic liver and ascites, findings that could indicate preexisting liver disease. (19) In three cases of methimazole-induced jaundice, (8), (9), (13) switching to propylthiouracil was successful, suggesting that cross-reaction between the two drugs in terms of liver toxicity is unlikely. However, the number of cases is too limited to draw a definite conclusion. In addition, the fact that propylthiouracil is more hepatotoxic than methimazole makes switching to propylthiouracil a risky approach.

In four hyperthyroid patients, (11), (27-29) methimazole (daily dose of 15-160 mg) was initiated despite obvious icterus and markedly elevated serum total bilirubin levels ranging from 3.4 mg/dl (11) to 26.9 mg/dl. (28) Jaundice resolved and bilirubin levels normalized in parallel to the normalization of thyroid function 6 weeks to 3 months after starting methimazole therapy. These findings lend support to the hypothesis that methimazole-induced cholestasis is most likely the result of allergic reaction that mainly occurs in susceptible individuals and that the presence of pretreatment hyperbilirubinemia probably plays a less important role. However, initiation of methimazole therapy in the setting of clinically evident cholestasis is unsafe and is extremely discouraged. In fact, in two of the previously mentioned four patients, (11), (27) serum bilirubin levels further increased two and a half to fivefold after starting methimazole therapy before serum bilirubin values finally normalized. There are no clearcut guidelines regarding the use of methimazole in case of liver dysfunction. There is indirect evidence that the drug is partly metabolized by the liver, since its half-life was significantly prolonged in cirrhotic patients. (34) In the latest product information, (35) withdrawal of the drug is recommended if liver enzyme values are three times above the upper normal limit. We do not routinely monitor liver function tests in all thyrotoxic patients. The rarity of life-threatening hepatotoxicity due to methimazole or propylthiouracil makes their routine measurement of limited benefit and not cost-effective. Furthermore, the common association of liver function abnormalities with the hyperthyroid state per se and the frequent occurrence of subclinical hepatotoxicity in patients treated with propylthiouracil can create diagnostic confusion. (36) Meanwhile, it is essential that patients be informed about the earliest symptoms of serious adverse effects of antithyroid drugs, such as those of agranulocytosis and hepatic toxicity, and that they be advised to stop taking the drug immediately and contact their physician if such symptoms occur.

Conclusion

Severe cholestatic jaundice is a rare and unpredictable adverse effect of methimazole therapy. The reaction is generally reversible after drug withdrawal. Physicians' awareness of this rare adverse drug reaction and its generally benign course will avoid unnecessary testing and invasive interventions such as ERCP and liver biopsy.

Key Points

* Severe cholestatic jaundice is a rare, and most likely allergic, adverse effect of methimazole that occurs within the first 3 month of use.

* Cholestasis is usually reversible within 3 month after discontinuing drug therapy.

* In case with typical presentation, invasive diagnostic procedures, such as liver biopsy, are not necessary.

* Patients should be instructed to discontinue methimazole if they develop jaundice or other symptoms of liver toxicity.

Wisdom is not wisdom when it is derived from books alone.

--Horace

Table 1. Previously reported cases of jaundice induced by methimazole

                               Latent    Recovery period
Reference  Age/Sex    Dose     Period *  [dagger]

 4          67/F    10 mg tid  29 days   -
 5          62/F    5 mg tid   3 wk      9 days
 6          63/F    15 mg qid  7 wk      10 wk
 7          38/F    20 mg/day  27 days   6 mo
 8          36/F    15 qid     25 days   8 wk
 9          54/F    40 mg/day  25 days   -
10          74/F    10 qid     12 days   2 mo
11          75/F    10 mg qid  20 days   100 days
            62/F    10 mg qid  5 wk      1 mo
12          58/F    20 mg/day  18 days   5 days
13          48/F    10 mg tid  2 wk      3 mo
14          68/M    20 mg tid  9 wk      3 mo

Reference       Comments

 4         Fatal agranulocytosis
 5         Agranulocytosis
 6         Agranulocytosis
 7
 8
 9         Fatal pneumonia
10
11         Methimazole continued

12         Rechallenge
13
14

* Time elapsed from the beginning of treatment with methimazole until
development of icterus or hyperbilirubinemia.
* Time taken for serum bilirubin levels to normalize after discontinuing
methimazole therapy.

Table 2. Previously reported cases of jaundice induced by carbimazole

                               Latent      Recovery
Reference  Age/Sex    Dose     Period (a)   Period (b)

15          64/M    40 mg/day  81 days     5 days

16          24/F    40 mg/day  3 mo        "Rapid"

17          81/F    30 mg/day  6 wk        5 days

18          45/F    20 mg/day  10 days     15 days

19          57/F    40 mg/day  6 wk        -

20          72/F    30 mg/day  5 wk        NR

21          70/F    30 mg/day  2 days      3 mo

22          59/M    60 mg/day  8 days      15 days

Reference          Comments

15         LTT

16         Rechallenge

17         Rechallenge, LTT

18         Rechallenge

19         Fulminant hepatitis

20         Hyperesinophilia

21         Jaundice occurred on
             second  drug exposure

22

(a) Time elapsed from the beginning of treatment with carbimazole until
development of icterus or hyperbilirubinemia.
(b) Time taken for serum bilirubin levels to normalize after
discontinuing carbimazole therapy.
LTT = lymphocyte transformation test.
NR = not reported.

Table 3. Liver function tests (present case)

                       Total      Conjugated
Date                 Bilirubin     Bilirubin    AST    ALT    ALKP

12/23/01 *             16.7          15.4        67    104    289
12/24/01               14.9          13.6        66     91    264
12/25/01               13.3           9.1        72     93    248
12/27/01               12.6          11.5       112    130    233
12/29/01               11.4          10.1       210    269    235
12/31/01 [dagger]       8.8           7.5       179    248    209
1/04/02                 4.7           3.8        67    151    150
1/18/01                 1.6           1.1        42     63    154
3/05/02                 0.6          ND          42     66    111

Total bilirubin normal range, 0.1 to 1.2 mg/dl.
Conjugated bilirubin normal range, 0.0 to 0.4 mg/dl.
AST = aspartate aminotransferase; normal range, 14 to 62 IU/L.
ALT = alanine aminotransferase; normal range, 14 to 75 IU/L.
ALKP = alkaline phosphatase; normal range, 48 to 116 IU/L.
* Date of hospital admission.
[dagger] Date of hospital discharge.
ND = not detectable.

Acknowledgments

I thank Toks Archie, MD, Marylou Dulay, MD, and Stanley Dea, MD, for their excellent care of the patient during her hospitalization.

Accepted June 28, 2002.

Copyright [c] 2004 by The Southern Medical Association

0038-4348/04/9702-0178

References

(1.) Stanley MM, Astwood EB. 1-methyl-2-mercaptoimidazole: An antithyroid compound highly active in man. Endocrinology 1949;44:588-589.

(2.) Kampmann JP, Hansen JM. Clinical pharmacokinetics of antithyroid drugs. Clin Pharmacokinet 1981;6:401-428.

(3.) Cooper DS. Treatment of thyrotoxicosis, in Braveman LE, Utiger RD (eds): Werner and Ingbars The Thyroid. Philadelphia, Lippincott-Raven, 1996, ed 7, pp 713-734.

(4.) Specht NW, Boehme EJ. Death due to agranulocytosis induced by methimazole therapy. JAMA 1952;149:1010-1011.

(5.) Rosenbaum H, Reveno WS. Agranulocytosis and toxic hepatitis from methimazole. JAMA 1953;152:27.

(6.) Shipp J. Jaundice during methimazole (Tapazole) administration. Ann Intern Med 1955;42:701-706.

(7.) Tennenbaum JI, Dreskin H. Toxic hepatitis during treatment with methimazole (Tapazole): Report of a case with apparent recovery. Ohio State Med J 1962;58:306-307.

(8.) Martinez-Lopez JI, Greenberg SE, Kling R. Drug-induced hepatic injury during methimazole therapy. Gastroenterology 1962;43:84-87.

(9.) Becker CF, Gorden P, Robbins J. Hepatitis from methimazole during adrenal steroid therapy for malignant exophthalmos. JAMA 1968;206:1787-1789.

(10.) Fisher MG, Nayer HR, Miller A. Methimazole-induced jaundice. JAMA 1973;223:1028-1029.

(11.) Jansen PL, Froeling PG, Schade RW, et al. Intrahepatic cholestasis in hyperthyroidism and the effect of antithyroid and [beta]-blocking drugs. Neth J Med 1982;25:318-324.

(12.) Schmidt G, Borsch G, Muller KM, et al. Methimazole-associated liver injury: Case report and brief literature review. Hepatogastroenterology 1986;33:244-246.

(13.) Arab D, Malatjalian DA, Rittmaster RS. Severe cholestatic jaundice in uncomplicated hyperthyroidism treated with methimazole. J Clin Endocrinol Metab 1995;80:1083-1085.

(14.) Schwab GP, Wetscher GJ, Vogl W, et al. Methimazole-induced liver injury, mimicking sclerosing cholangitis. Langenbecks Arch Chir 1996; 381:225-227.

(15.) Prost G, Dechavanne M, Levenq P, et al. Cholestatic jaundice caused by carbimazole [in French]. Nouv Presse Med 1973;2:2479 (letter).

(16.) Dinsmore WW, O'Hara MD, Callender ME. Postanesthetic carbimazole jaundice. N Engl J Med 1983;309:438 (letter).

(17.) Blom H, Stolk J, Schreuder HB, et al. A case of carbimazole-induced intrahepatic cholestasis. An immune-mediated reaction? Arch Intern Med 1985;145:1513-1515.

(18.) Ayensa C, Diaz De Otazo R, Cia JM. Carbimazole-induced cholestatic hepatitis. Arch Intern Med 1986;146:1455 (letter).

(19.) Epeirier JM, Pageaux GP, Coste V, et al. Fulminant hepatitis after carbimazole and propranolol administration. Eur J Gastroenterol Hepatol 1996;8:287-288.

(20.) Cales P, Voigt JJ, Suduca JM, et al. Cholestatic jaundice caused by carbimazole [in French]. Presse Med 1987;16:1005.

(21.) Ozenne G, Manchon ND, Doucet J, et al. Carbimazole-induced cholestatic jaundice. J Clin Gastroenterol 1989;11:95-97.

(22.) Sadoul JL, Canivet B, Freychet P. Toxic hepatitis induced by antithyroid drugs: Four cases including one with cross-reactivity between carbimazole and benzylthiouracil. Eur J Med 1993;2:473-477.

(23.) Kang H, Choi JD, Jung IG, et al. A case of methimazole-induced acute hepatic failure in a patient with chronic hepatitis B carrier. Korean J Intern Med 1990;5:69-73.

(24.) Baker B, Shapiro B, Fig LM, et al. Unusual complications of antithyroid drug therapy: Four case reports and review of literature. Thyroidology 1989;1:17-26.

(25.) Lunzer M, Huang S, Ginsburg J, et al. Jaundice due to carbimazole. Gut 1975;16:913-917.

(26.) Yao JD, Gross JB Jr, Ludwig J, et al. Cholestatic jaundice in hyperthyroidism. Am J Med 1989;86:619-620.

(27.) Lersch C, Seige M, Natrath W, et al. Cholestasis induced by hyperthyroidism after liver transplantation. Digestion 1995;56:429-432.

(28.) Babini G. Gurioli L, Rizzi R, et al. Appearance of severe jaundice after radiometabolic treatment of thyrotoxicosis. J Endocrinol Invest 1999;22:209-211.

(29.) Anton Aranda A. Intrahepatic cholestasis in untreated hyperthyroidism [in Spanish]. Rev Esp Enferm Dig 2000;92:49-50.

(30.) Scanlon MF, Toft AD. Regulation of thyrotropin secretion, in Braveman LE, Utiger RD (eds): Werner and Ingbars The Thyroid. Philadelphia, Lippincott-Raven, 1996, ed 7, pp 220-240.

(31.) Ozenirler S, Tuncer C, Boztepe U, et al. Propylthiouracil-induced hepatic damage. Ann Pharmacother 1996;30:960-963.

(32.) Eisen MJ. Fulminant hepatitis during treatment with propylthiouracil. N Engl J Med 1953;249:814-816.

(33.) Maggiore G, Larizza D, Lorini R, et al. Propylthiouracil hepatotoxicity mimicking autoimmune chronic active hepatitis in a girl. J Pediatr Gastroenterol Nutr 1989;8:547-548.

(34.) Cooper DS, Bodes HH, Nath B, et al. Methimazole pharmacology in man: Studies using a newly developed radioimmunoassay for methimazole. J Clin Endocrinol Metab 1984;58:473-479.

(35.) Methimazole, in Murray L (ed): Product Drug Reference. Montvale, NJ, Medical Economics Co., 2002, ed 56, pp 1823-1824.

(36.) Huang M, Li K, Wei JS, et al. Sequential liver and bone biochemical changes in hyperthyroidism: Prospective controlled follow-up study. Am J Gastroenterol 1994;89:1071-1076.

RELATED ARTICLE: Case Report

A 43-year-old Korean woman had hyperthyroidism diagnosed at an outside clinic on November 4, 2001, based on an elevated plasma thyroxine ([T.sub.4]) level of 30.7 [micro]g/dl (normal, 4.5-11.2), triiodothyronine ([T.sub.3]) uptake of 37.7 (normal, 27-37), as well as palpitations and hand tremors. The thyrotropin level was not available at that time. The patient had no significant medical history except for mild hypertension, for which she had taken 5 mg/d amlodipine for several years. She had no history of alcohol intake or liver disease. On November 4, 2001, her baseline liver function values were: total bilirubin 0.8 mg/dl (normal, 0.1-1.5 mg/dl); aspartate aminotransferase (AST) 159 U/L (normal, 1-45 U/L); alanine aminotransferase (ALT) 134 U/L (normal, 1-55 U/L); alkaline phosphatase (ALKP) 201 U/L (normal, 27-142U/L); [gamma]-glutamyl transpeptidase (GGTP) 23 U/L (normal, 5-52 U/L); and lactate dehydrogenase (LDH) 314 U/L (normal, 89-215 U/L). Other laboratory tests were unremarkable. Treatment was started with 10 mg of methimazole and 20 mg of propranolol, each three times daily. One month after starting methimazole (December 4, 2001), the patient began to have jaundice and generalized pruritus. She did not have abdominal pain, nausea, fever, or change in urine or stool color. She continued taking her medication until she finished her supply of methimazole and propranolol 4 days after the appearance of jaundice. Two weeks later, she came to our hospital's emergency department because of persisting jaundice and severe pruritus. At admission, she had severe icterus of the sclerae and skin. She had no signs of heart failure. Abdominal examination was unremarkable. The liver and spleen were not palpable. She had a small diffuse goiter consistent with Graves disease. Results of liver function tests at admission are shown in Table 3. Prothrombin time and partial thromboplastin time values were within normal limits. White blood cell count was slightly elevated at 12.1 X 1[0.sup.9] cells/L (normal, 3.8-1[0.sup.9] X 1[0.sup.9] cells/L). Serum free [T.sub.4] and total [T.sub.3] levels were normal at 1.07 (normal, 0.71-1.85 ng/dl) and 153 ng/ml (normal, 85-185 ng/dl), respectively. Serum thyrotropin level was below 0.02 U/L (normal 0.5-4.5 U/L). Serologic tests for hepatitis A, B, and C were negative. Tests for antimitochondrial antibodies (to rule out primary biliary cirrhosis) and antinuclear and anti-smooth muscle cell antibodies (to rule out autoimmune hepatitis) were negative. Computed tomography (CT) of the abdomen showed unremarkable liver, gallbladder, pancreas, and spleen, without evidence of biliary dilation. To that extent, icterus was attributed to methimazole, and treatment with the drug was not resumed. Propranolol therapy was resumed in the same previous doses, and amlodipine therapy was continued. During hospitalization, the degree of icterus and serum bilirubin levels declined over the next few days. Itching was relieved with cholestyramine and hydroxyzine hydrochloride. Liver enzyme values were closely monitored during the 9-day hospitalization (Table 3). Symptoms had almost completely resolved by the time of the patient's discharge. Serum levels of liver enzymes (ALT and AST) and bilirubin were normal at 6 and 12 weeks, respectively, after the discontinuance of methimazole therapy (Table 3). For treatment of hyperthyroidism, the patient received radioactive iodine ablation of the thyroid gland.

Nasser E. Mikhail, MD, MSC

From the Division of Endocrinology, Department of Medicine, Olive View UCLA Medical Center, Sylmar, CA.

Reprint requests to Nasser E. Mikhail, MD, Division of Endocrinology, Department of Medicine, Olive View UCLA Medical Center, 14445 Olive View Drive, Sylmar, CA 91342-1495. Email: [e-mail]