Serial Changes in Liver Function Tests in Patients with Thyrotoxicosis Induced by Graves’ Disease and Painless Thyroiditis

Home , Liver function tests

THYROID ORIGINAL STUDIES, REVIEWS AND Volume 18, Number 3, 2008 ª Mary Ann Liebert, Inc. DOI: 10.1089=thy.2007.0189 SCHOLARLY DIALOG

Sumihisa Kubota, Nobuyuki Amino, Yuka Matsumoto, Naoko Ikeda, Shinji Morita, Takumi Kudo, Hidemi Ohye, Eijun Nishihara, Mitsuru Ito, Shuji Fukata, and Akira Miyauchi

Context: When the liver function tests are aggravated after starting antithyroid drugs (ATDs) in Graves’ hyperthyroidism, discontinuation of ATDs is generally considered. However, a question arises whether such aggra- vation constitutes an adverse effect of the drugs or not. Objective: The aim of this study was to clarify the inﬂuence of thyrotoxicosis on liver function tests, comparing the results with those in thyrotoxicosis induced by painless thyroiditis. Design: We prospectively studied liver biochemical tests in 30 patients with Graves’ disease and in 27 patients with painless thyroiditis. Main outcomes: Twenty-three (76.7%) untreated Graves’ disease patients and 14 (51.9%) untreated painless thyroiditis patients were found to have at least one liver function test abnormality. One month after starting ATD therapy in patients with Graves’ disease, aspartate aminotransferase (AST) and alanine aminotransferase (ALT) elevations from initial values were observed in 16 (53.3%). Similar elevations of AST and ALT from initial values at 1 month were observed in 10 (37.0%) and 7 (25.9%) patients with painless thyroiditis, respectively. Alkaline phosphatase (ALP) increased gradually after starting ATD therapy and maintained an elevated value for 3–5 months in Graves’ disease. In painless thyroiditis, ALP also increased gradually, similarly to that in Graves’ disease, but changes were mild. Elevation of ALT after 1 month of ATD therapy in Graves’ disease was signiﬁcantly higher in patients whose estimated disease duration was 6 months or more compared to those with duration of less than 6 months. Elevated AST and ALT at 1 month after ATD therapy decreased to normal ranges, even though patients were receiving the same ATDs in Graves’ disease. Conclusion: Similar serial changes in liver function tests in both Graves’ disease and painless thyroiditis strongly suggest that increases of AST and ALT after starting ATD therapy may not be due to ATD side effects but may be induced by changes in thyroid function.

Introduction PTU-induced liver injury, and previous reports suggest that treatment should be stopped in these patients with ‘‘PTU t is well known that liver biochemical abnormalities hepatitis’’ (5–10). On the other hand, several studies reported Ihave been shown in untreated patients with thyrotoxicosis that the discontinuation of PTU is not necessarily required (1–3). These abnormalities are thought to be induced by the unless overt hepatitis develops, because transient mild asymp- metabolic effects of excess thyroid hormone and hepatic tis- tomatic ‘‘PTU hepatitis’’ frequently occurs in patients with sue hypoxia, which occur as a result of enhanced splanchnic PTU-treated Graves’ disease (11–13). oxygen consumption and an increase in the hepatic require- To clarify the inﬂuence of thyrotoxicosis it may be helpful ment for oxygen (1,4). In Graves’ disease, abnormalities of to compare the liver function tests in thyrotoxicosis induced liver function tests are not only observed in the untreated by painless thyroiditis, since thyrotoxicosis spontaneously condition but also frequently found after starting antithyroid resolves without any medication (14). However, there have drugs (ATDs). Liver biochemical abnormalities after pro- been no reports on the serial changes in liver function tests pylthiouracil (PTU) administration are thought to represent in patients with painless thyroiditis. In the present work, we

Kuma Hospital, 8–2–35, Shimoyamate-dori, Chuo-ku, Kobe 650–0011, Japan.

283 284 KUBOTA ET AL. prospectively studied liver biochemical tests before and after antigen and antihepatitis C virus antibodies were assayed starting ATD therapy in patients with Graves’ disease and in using immunoassay kits (ARCHITECT HBsAg QT [Abbott patients with painless thyroiditis to compare the parameter Japan, Tokyo, Japan] and Ortho HCVAb IRMA test 3 changes in both diseases. [Ortho-Clinical Diagnostics, Tokyo, Japan], respectively). TSH, FT4, and FT3 concentrations were measured by chemi- Materials and Methods luminescent immunoassays (ARCHITECT TSH, ARCHI- TECT FREE T4, and ARCHITECT FREE T3 [Abbott Japan], Subjects respectively). Anti-TSH receptor antibodies were measured We prospectively recruited 30 patients with newly diag- as TSH-binding inhibitor immunoglobulin (TBII) by a TSH nosed Graves’ disease and 27 patients with painless thyroid- receptor antibody ELISA assay kit (Cosmic, Tokyo, Japan). itis in 2006. Graves’ hyperthyroidism was diagnosed on the basis of diffuse goiter, thyrotoxicosis, elevated 3- or 24-hour Statistics radioiodine uptake (RAIU) of the thyroid, and=or positive anti–thyroid-stimulating hormone (anti-TSH) receptor anti- The baseline characteristics and results (serum levels of bodies. Painless thyroiditis is deﬁned by thyrotoxicosis, no- AST, ALT, and ALP) between each patient and control group tender goiter, and markedly decreased RAIU of the thyroid. were compared using the Mann–Whitney U-test. Correlation Patients who had evidence of cardiovascular complications analyses were performed using the Spearman’s rank correla- or past histories of hepatobiliary disorder, and excess alcohol tion test. Differences were considered to be signiﬁcant at consumption were excluded from this study. Patients who p < 0.05. had changed from methimazole (MMI) to PTU or vice versa during treatment were excluded, and all patients received the Results same drug during the study period. Patients who were posi- The initial laboratory results of all cases are summarized in tive for hepatitis B surface antigen and antihepatitis C virus Table 1. Serial changes in values of serum AST, ALT, ALP, and antibodies were also excluded. We selected 35 healthy age- FT4 in the patients with Graves’ disease and painless thy- and sex-matched persons as normal controls. The durations of roiditis are shown in Table 2. Individual changes in values of Graves’ disease before starting ATD therapy were estimated serum AST, ALT, and ALP in the patients with Graves’ dis- from the clinical histories of patients. Informed consent was ease and painless thyroiditis are also shown in Figure 1. obtained from every subject. Protocols were approved by the ethics committee of our hospital. Graves’ disease

Treatment The initial mean values of ALT, ALP, and g-GTP of patients were signiﬁcantly higher than those of normal controls, but All patients with Graves’ disease were treated with ATDs. AST was not. On the other hand, the initial mean values of = = Initially, 15–30 mg day of MMI or 100–300 mg day of PTU ALB, BIL, LDH, and CK of patients were signiﬁcantly lower was administered, and tapered gradually to maintain a euthy- than those of normal controls. Serum AST, ALT, g-GTP, and roid state. MMI was used for 25 patients, and PTU was used ALP elevations above the upper limit of the normal range for 5 patients. Patients with painless thyroiditis were observed were observed in 2 (6.6%), 8 (26.7%), 8 (26.7%), and 18 (60.0%) without medication. patients, respectively. Twenty-three (76.7%) untreated pa-

Follow-up Table 1. Liver Function Tests in Untreated Patients We measured serum levels of albumin (ALB), total bilirubin with Graves’ Disease and Painless Thyroiditis (BIL), aspartate aminotransferase (AST), alanine aminotrans- and Healthy Controls ferase (ALT), alkaline phosphatase (ALP), gammaglutamyl transpeptidase (g-GTP), lactic dehydrogenase (LDH), creati- Graves’ Painless Healthy nine phosphokinase (CK), TSH, free thyroxine (FT4), and free disease thyroiditis controls triiodothyronine (FT3) before and 1, 2, and 3–5 months after starting ATD therapy in patients with Graves’ disease. Iden- No. of patients 30 27 35 tical parameters were measured at the time of diagnosis and Female=male 25=524=331=4 1, 2, and 3–5 months after the initial visit in untreated thyro- Age (years) 40.3 Æ 13.6 38.8 Æ 17.3 43.7 Æ 12.2 ALB (g=dL) 4.1 Æ 0.2a 4.1 Æ 0.3a 4.7 Æ 0.3 toxic patients with painless thyroiditis. CK levels were mea- BIL (mg=dL) 0.5 Æ 0.3b 0.6 Æ 0.4 0.7 Æ 0.3 sured to clarify the effect of the hypermetabolic state on serum AST (IU=L) 23.7 Æ 9.5 30.6 Æ 16.5b 20.3 Æ 5.4 enzymes and to compare changes in hepatic enzymes. ALT (IU=L) 30.8 Æ 11.3a 47.2 Æ 42.1a 17.4 Æ 7.3 ALP (IU=L) 368.7 Æ 140.6a 206.1 Æ 78.4 187.7 Æ 58.7 Laboratory methods g-GTP (IU=L) 38.4 Æ 36.4b 28.5 Æ 30.0 26.3 Æ 27.1 LDH (IU=L) 147.8 Æ 25.0b 154.9 Æ 35.3 161.8 Æ 28.7 Serum levels of ALB (normal range: 3.8–5.2 g=dL), BIL (nor- CK (IU=L) 51.4 Æ 21.3a 70.5 Æ 82.2a 98.6 Æ 32.7 mal range: 0.2–1.0 mg=dL), AST (normal range: 5–40 IU=L), ALT (normal range: 5–35 IU=L), ALP (normal range: 105– Signiﬁcantly different from healthy controls at ap < 0.001 and b 340 IU=L), g-GTP (normal range: 0–40 IU=L), LDH (normal p < 0.05. = = Data indicate mean Æ SD. ALB: albumin; BIL: bilirubin; AST: as- range: 105–215 IU L), and CK (normal range: 50–180 IU L) partate aminotransferase; ALT: alanine aminotransferase; ALP: alka- were measured using a routine automated analyzer (Hitachi line phosphatase; g-GTP: gammaglutamyl transpeptidase; LDH: lactic 7170S Clinical Analyzer, Tokyo, Japan). Hepatitis B surface dehydrogenase; CK: creatinine phosphokinase. Table 2. Serial Changes in Values of Serum AST, ALT, and ALP in Patients with Graves’ Disease and Painless Thyroiditis

Months after ﬁrst visit Healthy 0 1 2 3–5 controls

Graves’ disease AST (IU=L) 23.7 Æ 9.5 27.8 Æ 15.4a 21.2 Æ 6.5 18.4 Æ 6.7 20.3 Æ 5.4 ALT (IU=L) 30.8 Æ 11.3b 40.7 Æ 29.6b 25.9 Æ 14.3a 19.2 Æ 13.0 17.4 Æ 7.3 ALP (IU=L) 368.7 Æ 140.6b 438.7 Æ 171.9b 471.0 Æ 200.1b 447.2 Æ 204.1b 187.7 Æ 58.7 b b b FT4 (ng=dL) 3.23 Æ 0.86 1.56 Æ 0.53 1.25 Æ 0.52 1.02 Æ 0.52 0.89 Æ 0.11 Painless thyroiditis AST (IU=L) 30.6 Æ 16.5a 32.7 Æ 17.7b 24.6 Æ 11.3 20.3 Æ 6.0 20.3 Æ 5.4 ALT (IU=L) 47.2 Æ 42.1b 48.4 Æ 35.2b 28.2 Æ 18.9a 20.1 Æ 9.5 17.4 Æ 7.3 ALP (IU=L) 206.1 Æ 78.4 283.4 Æ 127.2a 310.2 Æ 123.4b 296.0 Æ 112.9b 187.7 Æ 58.7 b a b FT4 (ng=dL) 2.82 Æ 0.73 1.56 Æ 1.12 0.73 Æ 0.35 0.83 Æ 0.20 0.89 Æ 0.11

Signiﬁcantly different from healthy controls at ap < 0.05 and bp < 0.001. Data indicate mean Æ SD. Normal range of FT4: 0.7–1.6 ng=dL. AST: aspartate aminotransferase; ALT: alanine aminotransferase; ALP: alkaline phosphatase; FT4: free thyroxine.

Graves’ disease A B Painless thyroiditis 100 100

80 80

60 60

AST (IU/l) 40 AST (IU/l) 20 20

0 0 180 220 150 150 120 120 90 90

ALT (IU/l) 60 ALT (IU/l) ALT 60 30 30 0 0

1000 1000

800 800

600 600

400

ALP (IU/l) ALP 400 ALP (IU/l) ALP 200 200

0 0123-5 0 0123-5 Months after first visit Months after first visit

FIG. 1. Individual serial changes in values of serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), and alkaline phosphatase (ALP) in patients with Graves’ disease and painless thyroiditis. 286 KUBOTA ET AL. tients with Graves’ disease were found to have at least one from the initial value were observed in 10 (37.0%). Similarly, liver function test abnormality. Serum CK reductions below serum ALT elevations from the initial value were observed in 7 the lower limit of the normal range were observed in 17 (25.9%) at 1 month after the initial evaluation. The mean value (56.7%) patients. of serum ALP increased gradually, and the peak of serum ALP The mean values of AST and ALT increased 1 month after occurred within 2 months of the initial evaluation. These ALP starting ATD therapy and then decreased to the normal range changes were similar but mild compared to those of Graves’ in 3–5 months. One month after starting ATD therapy, serum disease. We performed correlation analyses between initial AST and ALT elevations from initial values were observed in values of ALT and FT4; however, there were no significant 16 (53.3%) and 16 (53.3%) patients, respectively. Serum AST correlations. and ALT were normalized in all and 27 (90.0%) patients, re- There was no significant difference between Graves’ dis- spectively, at 3–5 months after starting ATD therapy. On the ease and painless thyroiditis in mean values of AST and ALT contrary, ALP increased gradually after commencing ATD at all points. The mean values of ALP in Graves’ disease were therapy, and maintained an elevated value for 3–5 months. significantly higher ( p < 0.001) than those in painless thy- We performed correlation analyses between initial values roiditis at all points. of ALT or ALP and FT4,FT3, or TBII; however, there were no significant correlations. Serum ALT elevations shown 1 month Discussion after starting ATD therapy were significantly higher in the patients whose estimated duration of disease was 6 months or Liver biochemical abnormalities were marked in both more compared to those with duration of less than 6 months untreated patients with Graves’ disease and painless thy- (Fig. 2). The changes in liver function tests were similar in both roiditis in this study. In general, the duration of thyro- PTU-treated patients and MMI-treated patients, and showed toxicosis induced by Graves’ disease is longer than that no significant difference. induced by painless thyroiditis, because thyrotoxicosis is transient in painless thyroiditis. A comparatively short du- Painless thyroiditis ration of thyrotoxicosis seems to be able to increase serum AST and ALT levels. A major difference between these two The initial mean values of AST and ALT of patients were diseases is the initial mean value of serum ALP, which is significantly higher than those of normal controls; however, mainly of bone origin and sometimes helpful in the differ- ALB and CK were significantly lower than those of normal ential diagnosis of thyrotoxicosis (11,12,15,16). The rise in controls. Serum AST, ALT, g-GTP, and LDH elevations above serum ALP observed after the treatment of Graves’ hyper- the upper limits of the normal range were observed in 5 thyroidism related to heightened osteoblastic activity is well (18.5%), 12 (44.4%), 5 (18.5%), and 2 (7.4%) patients, respec-

known (16). The rise of serum ALP was observed not only in tively. Fourteen patients (51.9%) were found to have at least the treated patients with Grave’s disease but also in those one liver function test abnormality. Serum ALB and CK with painless thyroiditis in this study. These data indicate reductions below the lower limits of the normal range were that even the short duration of thyrotoxicosis inﬂuences bone observed in 3 (11.1%) and 16 (59.3%) patients, respectively. metabolism. One month after the initial evaluation, serum AST elevations Transient liver biochemical abnormalities after starting PTU therapy have been reported. In these reports, transient elevations of serum ALT were observed in 16–30% of PTU- P 0.01 160 treated patients 6–8 weeks after starting therapy, and the abnormalities were considered to be induced by PTU (13,16,17). 140 Interestingly, similar transient liver biochemical abnormalities were observed in patients with painless thyroiditis who 120 did not receive any medication in this study. These results suggest that liver biochemical abnormalities observed after 100 starting ATD therapy in this study may not be due to ATDs. Kim et al. reported that the incidence of symptomatic 80 clinically overt hepatic injury due to PTU was 1.2% (18). This supports our notion. We may say that we do not have to ALT (IU/l) 60 discontinue ATDs unless serum AST or ALT is elevated to more than 150 IU=L. Nonetheless, close observation of hepatic 40 enzymes is necessary after starting ATD therapy in Graves’ disease. 20 One month after the initial evaluation, serum changes in AST and ALT varied in painless thyroiditis. Serum AST and 0 ALT decreased after 1 month in the patients whose initial Less than 6 months 6 months or more values were high. It is assumed that the variation was related FIG. 2. Elevation of serum alanine aminotransferase (ALT) to the phase of painless thyroiditis. The patients whose serum at 1 month after starting antithyroid drug (ATD) therapy in AST and ALT decreased after 1 month were possibly evalu- patients with Graves’ disease. Comparison between patients ated in the late phase of the disease; on the contrary, the pa- whose estimated duration of disease was less than 6 months tients whose serum AST and ALT increased were evaluated in and 6 months or more. the early phase. CHANGES IN LIVER FUNCTION TESTS IN THYROTOXICOSIS PATIENTS 287

We studied various factors that may inﬂuence transient 9. Mihas AA, Holley P, Koff RS, Hirschowitz BI 1976 Fulminant serum AST and ALT elevations shown after ATD therapy in hepatitis and lymphocyte sensitization due to propylthio- Graves’ disease to elucidate its mechanism. Only the esti- uracil. Gastroenterology 70:770–774. mated duration of Graves’ disease before starting therapy was 10. Parker WA 1982 Propylthiouracil-induced hepatotoxicity. related to the transient serum AST and ALT elevations. It is Clin Pharm 1:471–474. possible that a prolonged hypermetabolic state regarding liver 11. Liaw YF, Huang MJ, Fan KD, Li KL, Wu SS, Chen TJ 1993 function is responsible. Possibly, both the production and Hepatic injury during propylthiouracil therapy in patients turnover rate of hepatic enzymes may be increased in the with hyperthyroidism. A cohort study. Ann Intern Med 15: condition of hyperthyroidism, and if the turnover rate nor- 424–428. malizes faster than production, serum hepatic enzymes may 12. Huang MJ, Li KL, Wei JS, Wu SS, Fan KD, Liaw YF 1994 Sequential liver and bone biochemical changes in hyperthy- increase transiently. roidism: prospective controlled follow-up study. Am J Gas- In summary, similar serial changes in liver function tests troenterol 89:1071–1076. were shown in both untreated patients with Graves’ disease 13. Aydemir S, Bayraktaroglu T, Demircan N, Sert M, Acikgoz S, and painless thyroiditis. Transient liver biochemical elevations Tekin IO, Ustundag Y 2005 Effect of hyperthyroidism and were observed not only in ATD-treated patients with Graves’ propylthiouracil treatment on liver biochemical tests. Int J disease but also in patients with painless thyroiditis who did Clin Pract 59:1304–1308. not receive any medication. These results strongly suggest 14. Papapetrou PD, Jackson IM 1975 Thyrotoxicosis due to that liver biochemical abnormalities observed after starting ‘‘silent’’ thyroiditis. Lancet 1:361–363. ATD therapy may not be induced by ATDs but by the dy- 15. Cooper DS, Kaplan MM, Ridgway EC, Maloof F, Daniels GH namic changes in thyroid function. 1979 Alkaline phosphatase isoenzyme patterns in hyperthyroidism. Ann Intern Med 90:164–168. References 16. Pantazi H, Papapetrou PD 2000 Changes in parameters of bone and mineral metabolism during therapy for hyperthy- 1. Bayraktar M, Van Thiel DH 1997 Abnormalities in measures roidism. J Clin Endocrinol Metab 85:1099–1106. of liver function and injury in thyroid disorders. Hepatogas- 17. Gurlek A, Cobankara V, Bayraktar M 1997 Liver tests in hy- troenterology 44:1614–1618. perthyroidism: effect of antithyroid therapy. J Clin Gastro- 2. Biscoveanu M, Hasinski S 2000 Abnormal results of liver enterol 24:180–183. function tests in patients with Graves’ disease. Endocr Pract 18. Kim HJ, Kim BH, Han YS, Yang I, Kim KJ, Dong SH, Kim HJ, 6:367–369. Chang YW, Lee JI, Chang R 2001 The incidence and clinical 3. Huang MJ, Liaw YF 1995 Clinical associations between thy- characteristics of symptomatic propylthiouracil-induced he- roid and liver diseases. J Gastroenterol Hepatol 10:344–350. patic injury in patients with hyperthyroidism: a single-center

4. Sellin JH, Vassilopoulou-Sellin R 2005 The gastrointestinal retrospective study. Am J Gastroenterol 96:165–169. tract and liver in thyrotoxicosis. In: Braverman LE, Utiger RD (eds) The Thyroid, 9th edition. Lippincott Williams & Wilkins, New York, pp 622–626. 5. Hanson JS 1984 Propylthiouracil and hepatitis. Two cases and Address reprint requests to: a review of the literature. Arch Intern Med 144:994–996. Sumihisa Kubota 6. Limaye A, Ruffolo PR 1987 Propylthiouracil-induced fatal Kuma Hospital hepatic necrosis. Am J Gastroenterol 82:152–154. 8-2-35 Shimoyamate-Dori 7. Jonas MM, Eidson MS 1988 Propylthiouracil hepatotoxicity: Chuo-Ku two pediatric cases and review of the literature. J Pediatr Kobe 650-0011 Gastroenterol Nutr 7:776–779. Japan 8. Fedotin MS, Lefer LG 1975 Liver disease caused by propylthiouracil. Arch Intern Med 135:319–321. E-mail: [email protected]