Liver fibrosis attributed to lipid lowering medications: two cases

Authors


Dr Ooi Division of Endocrinology and Metabolism, The Ottawa Hospital – Civic Campus, 1053 Carling Avenue, Ottawa, Ontario, Canada, K1Y 4E9 (Fax: 613 761-5316; e-mail: tcooi@ottawahospital.on.ca).

Abstract

Abstract. Punthakee Z, Scully LJ, Guindi MM, Ooi TC (Department of Medicine, Division of Gastroenterology, Department of Pathology and the Laboratory of Medicine and the Division of Endocrinology and Metabolism, The Ottawa Hospital – Civic Campus, University of Ottawa, Ottawa, Canada). Liver fibrosis attributed to lipid lowering medications: two cases (Case Report). J Intern Med 2001; 250: 249–254.

We identified two cases of chronic active hepatitis with liver fibrosis induced by lipid lowering drugs of the statin and fibrate classes despite regular monitoring of transaminases. There are few reports of clinically significant hepatitis induced by these drugs and even fewer cases of fibrosis. Given the growing use of these drugs, there are implications for monitoring patients on long-term therapy for liver damage.

Introduction

Two classes of lipid lowering drugs, statins and fibrates, have now been proven effective in secondary prevention of cardiovascular outcomes [1, 2]. They are considered relatively safe, but they are known to cause reversible, asymptomatic elevation of serum transaminases [1, 3]. Of patients on statins, high transaminases have been found in 2–5% with 1% requiring cessation of the drug [4]. Symptomatic liver disease with these agents is very uncommon and documented liver fibrosis is a rarity. We present two cases of symptomatic liver injury secondary to hypolipidemic agents after prolonged exposure. One patient on a statin alone had chronic active hepatitis and mild hepatic fibrosis on histopathology whilst our second patient developed severe fibrosis on a combination of statin and fibrate which has never been described.

Case 1

A 39-year-old man with a strong family history of heart disease was seen in lipid clinic for a lipid profile as follows: total cholesterol 6.43 mmol L–1, triglycerides 1.08 mmol L–1, HDL-Cholesterol 0.97 mmol L–1, and LDL-Cholesterol 5.04 mmol L–1. After suboptimal control with gemfibrozil and cholestyramine, his medication was changed to lovastatin which brought his total cholesterol to 4.9 mmol L–1. After 9 months, lovastatin was stopped because of myalgias despite a lack of concomitant rise in creatine kinase.

He was off all medication for 9 months and was then started on pravastatin which brought his cholesterol to 4.4 mmol L–1. After 9 months on pravastatin with no adverse effects, he presented to the emergency department complaining of 3 weeks of fever and asthenia. He stopped taking pravastatin after 1 week of being ill. In the months preceding the illness, he had several protected homosexual contacts but denied excessive alcohol use. He had an episode of jaundice in childhood, the details of which he was unaware. There was no family history of liver disease. He was not jaundiced and had no stigmata of chronic liver disease. His liver enzymes were consistent with acute liver injury and are presented in Fig. 1(a).

Figure 1.

 (a) Time course, in case 1, of medications and liver tests: ALT in U/L (◆), AST in U/L (▮), ALP in U/L (▴), total bilirubin in μmol L–1 (×). Symptoms were present for a defined period. (b) Time course, in case 2, of medications and liver tests: ALT in U/L (◆), AST in U/L (▮), ALP in U/L (▴), total bilirubin in μmol L–1 (×). Symptoms were ill-defined and prolonged. Inserts highlight the episodes of acute symptomatic hepatitis in both cases.

Tests for viral hepatitis were negative (anti-HAV IgM, anti-HBc and anti-HBc IgM, HBsAg, anti-HCV). Eosinophil count, ANA, IgG and IgA were normal. IgM was high at 6.15 g L–1 (reference interval 0.44–3.40 g L–1) and antismooth muscle antibody titre was 1 : 40. Ultrasound showed a normal liver, gallbladder and biliary tree. Liver enzymes returned to normal 10 weeks after discontinuing pravastatin (Fig. 1a).

One week after his liver enzymes normalized, simvastatin was started. He remained asymptomatic on simvastatin with normal transaminases for 22 months, at which time, he developed the same symptoms he had 2 years previously. This time, he had even greater elevations of liver enzymes (Fig. 1a) with negative HBsAg, anti-HCV, anti-EBV IgM, and normal ceruloplasmin, C3, C4, IgA and IgG but an elevated IgM at 12.10 g L–1. He stopped taking simvastatin and treated his symptoms with acetaminophen for 1 week. Liver enzymes dropped within a week.

Fourteen days after stopping simvastatin, a needle biopsy of the liver (Fig. 2) showed preserved architecture, chronic inflammation of the portal tracts with mild piecemeal necrosis and mild, focal periportal fibrosis. Lobules contained mild necroinflammatory activity. A few incidental lipogranulomata were noted in the portal tracts and centrilobular areas. Mild parenchymal and Kuppfer cell siderosis was noted. This was consistent with but not diagnostic of drug-induced chronic active hepatitis with fibrosis. No other aetiologies were suggested histologically. His only significant long-term drug exposure was simvastatin.

Figure 2.

 Liver biopsy of the patient in case 1. (a) Edge of a portal area showing mild piecemeal necrosis – lymphocytes destroying individual periportal hepatocytes (haematoxylin, phloxin and saffron stain – 400× magnification). (b) Mild periportal fibrosis (reticulin stain – 250× magnification).

Statins were never re-introduced. Six and a half years after stopping simvastatin, ANA, antismooth muscle antibody and antimitochondrial antibody were negative and his quantitative IgM was in the normal range. He has had no recurrence of hepatitis.

Case 2

A 63-year-old obese lady with a family history of heart disease had a myocardial infarction in 1986 and was found to have a lipid profile as follows: total cholesterol 6.2 mmol L–1, triglycerides 6.39 mmol L–1 and HDL-Cholesterol 0.80 mmol L–1, so she was started on cholestyramine and gemfibrozil. Nine years later, when her LDL-Cholesterol was 3.4 mmol L–1, she was changed to simvastatin and gemfibrozil. Two years later, she was switched from gemfibrozil to fenofibrate and was continued on simvastatin. Her other medications were oestrogen, progesterone and vitamins.

Liver enzymes were normal when monitored every 6 to 10 months up to and including 1 year after starting fenofibrate and 3 years after starting simvastatin (Fig. 1b). Two years after starting fenofibrate and 4 years after starting simvastatin, on routine bloodwork, her ALT and AST were 443 U L–1 and 560 U L–1, respectively (reference intervals ALT 5–40 U L–1, AST 10–30 U L–1). At that time, she described minimal fatigue and some weight loss. Simvastatin and fenofibrate were discontinued. She denied alcohol excess. Past history included a cholecystectomy, a blood transfusion years previously during surgery for a cerebral aneurysm and transient jaundice during her teens. She exhibited mild palmar erythema and a large firm liver.

Anti-HAV IgM, HBsAg, anti-HBsAb and anti-HCVAb were negative, and eosinophils, ferritin, IgA and IgM were normal. IgG was high at 22.1 g L–1 (reference interval 6.9–16.2 g L–1) and ANA was positive with a titre of 1 : 40 and a speckled nucleolar pattern, but antismooth muscle and antimitochondrial antibodies were negative. Abdominal ultrasound revealed a cholecystectomy, an elongated right liver lobe (Reidel’s lobe, a normal variant), an anatomical variant of the intrahepatic portal vein, but a normal biliary tree and liver parenchyma. Liver biopsy (Fig. 3) carried out 7 weeks after simvastatin and fenofibrate cessation showed portal and intralobular inflammation with periportal piecemeal necrosis and severe bridging fibrosis. Eosinophils were rare but present and the bile ducts were normal. Iron stains were normal and there were no α-1 antitrypsin globules. These findings were those of chronic active hepatitis with severe fibrosis, and with no other obvious aetiology besides simvastatin and fenofibrate exposure.

Figure 3.

 Liver biopsy of the patient in case 2. (a) Portal area showing lymphocytic inflammation and piecemeal necrosis (haematoxylin, phloxin and saffron stain – 250× magnification). (b) Two portal areas connected by bridging fibrosis (severe fibrosis) (Masson trichrome stain – 100× magnification).

One month after discontinuing simvastatin and fenofibrate, her liver enzymes returned to near normal (Fig. 1b). Two months after stopping the pills, transaminases were completely normal without any other intervention or treatment. The ANA became negative, the IgG level was normal at 13.3 g L–1 and she was asymptomatic.

Discussion

In both cases, the pattern of liver injury was that of chronic active hepatitis with fibrosis. Apart from drug intake, there was no other apparent cause. Although chronic active hepatitis is not specific for drug-induced liver disease, it is one of the patterns of drug injury to the liver.

Our patients fulfil the International Consensus Criteria for Drug-Induced Liver Disorders [5], with the only significant drugs being pravastatin and simvastatin in case 1, and simvastatin and fenofibrate in case 2. The time to onset of the reactions is compatible, being more that 90 days. Other reports of statin- or fibrate-induced liver injury occurred after a treatment period anywhere from 2 months to 3 years [6–18]. In both our patients, the 50% decrease in ALT within 30 days of cessation, along with the absence of other hepatotoxic drugs or alcohol, and absence of serological or ultrasonographic evidence to suggest other causes of liver damage allow attribution of causation to pravastatin and simvastatin in case 1 and simvastatin and/or fenofibrate in case 2. In case 1, although the same drug was not re-administered, one might consider the second event as a positive rechallenge with a statin.

The most striking feature of our cases is the fibrosis. In the statin literature, Medline lists about 22 cases of symptomatic liver injury induced by statins. Ten of these include histopathology [6–14] and only two identified mild liver fibrosis [13, 14]. Only eight cases linking fibrates to fibrosis have been published previously [15–18]. Fibrosis with elevated IgM, as in case 1, is also seen in autoimmune diseases of the liver. In this case, an autoimmune injury induced by the drug or its metabolites is most probable, as his symptoms, enzymes, IgM and antismooth muscle antibody titre normalized after withdrawing the statin, and without any other treatment he has not relapsed in 6.5 years. In case 2, symptoms, liver enzymes, IgG and ANA normalized after drug withdrawal, suggesting a similar mechanism. Fibrosis was nonetheless present in both cases.

There are important implications here for monitoring for adverse drug reactions to lipid lowering agents. Despite diligent monitoring of liver enzymes, in both cases, every 6 to 12 months, enzyme elevations were unpredictable and brief but the degree of fibrosis seen would normally take months to years to develop. Perhaps these drugs can cause ongoing inflammation and fibrosis without much enzyme elevation, as occurs with methotrexate [19]. If this is true, liver failure after years of lipid lowering therapy may occur without warning unless other methods of monitoring for such injury are undertaken. Biopsy or postmortem studies of patients on long-term hypolipidemic therapy may be necessary to further delineate the prevalence of asymptomatic statin- or fibrate-induced liver fibrosis, the concern being that as more patients receive these drugs to stave off cardiovascular mortality, patients who develop liver fibrosis may survive long enough to suffer the effects of liver damage.

In summary, we identified two patients with fibrosis of the liver with histological chronic active hepatitis as a consequence of statins and/or fibrates. Previously, fibrosis induced by statins has been reported in only two patients and by fibrates in only eight patients. There have been no reports in patients on a statin and fibrate combination. Our cases developed hepatitis and fibrosis despite regular surveillance for transaminase elevation. This has implications for monitoring such patients for liver damage and may suggest that more invasive investigations are necessary for patients on long-term lipid lowering therapy.

Ancillary