Description. Drug induced liver injury with autoimmune features (also called drug induced autoimmune hepatitis) is marked by acute or chronic liver injury accompanied by development of autoantibodies, a hepatocellular pattern of serum enzyme elevations, and liver biopsy features suggestive of idiopathic autoimmune hepatitis. Drug induced autoimmune hepatitis resembles idiopathic autoimmune hepatitis, but typically resolves completely once the medication is withdrawn, although recovery may be slow and lead to a limited course of corticosteroid therapy.

Latency to Onset. The time to onset of drug induced autoimmune hepatitis is usually more than 6 months and can be up to several years after initiation of therapy. In some situations, onset is more rapid and the clinical phenotype resembles acute hepatitis, but onset with appearance of autoantibodies is rare before 2 months.

Symptoms. The onset of drug induced autoimmune hepatitis is typically insidious, sometimes with extrahepatic manifestations such as rash or joint pains that precede symptoms of hepatic injury (fatigue, nausea, poor appetite). Jaundice arises late and may follow several months of nonspecific symptoms.

Serum Enzyme Elevations. Autoimmune hepatitis due to medications is typically manifested by moderate serum aminotransferase elevations (200 to 800 IU/L: 5 to 20 times ULN) with minimal increase in alkaline phosphatase levels (<230 U/L: <2 times ULN), suggestive of chronic rather than acute hepatitis. Serum autoantibodies are usually (but not invariably) present, including antinuclear (ANA), smooth muscle antibody (SMA) or antibody to liver-kidney microsomes (anti-LKM). Immunoglobulin levels and total globulins are often but not always raised. Medications that cause autoimmune hepatitis associated with chronic hepatitis can often also cause an acute viral hepatitis-like syndrome (minocycline, nitrofurantoin, methyldopa, hydralazine) and these two phenotypes may overlap. Recovery after stopping the medication in autoimmune hepatitis may be delayed and lead to use of corticosteroids, which typically results in prompt improvement. Autoantibody titers and immunoglobulin levels also decrease with recovery, but low levels of autoantibodies may persist long term. In most instances, drug induced autoimmune hepatitis ultimately resolves once the medication is stopped, although this may take 6 months or more; in some cases, however, particularly with biologic response modifiers (interferons, tumor necrosis factor antagonists), the autoimmune hepatitis appears to have been triggered rather than caused by the drug and can be self sustaining and require long term immunosuppressive therapy.

Drugs. Medications that typically cause autoimmune hepatitis include minocycline, nitrofurantoin, hydralazine, methyldopa, statins, fenofibrate, alpha and beta interferon, infliximab and etanercept. Individual case reports of an autoimmune hepatitis-like syndrome attributed to medications have implicated many other agents, including nonsteroidal antiinflammatory agents, azathioprine, and herbals. In addition, some medications induce autoantibodies even in the absence of hepatic injury (isoniazid, procainamide) and their presence during liver injury may be coincidental.

Criteria for Definition. There are no established diagnostic criteria for drug induced autoimmune hepatitis; however, based upon similarity to idiopathic autoimmune hepatitis the following are important elements in the diagnosis:

1.

Time to onset of 2 months or more

2.

Rash, arthralgias or extrahepatic manifestations

3.

Hepatocellular pattern of serum enzyme elevations (R >5 at onset)

4.

Presence of an autoantibody in titers of 1:80 or greater (ANA, SMA or anti-LKM)

5.

Raised immunoglobulin (IgG >1800 mg/dL) or total globulin levels (>3.0 grams/dL)

6.

Exposure to an agent that is typically associated with drug induced autoimmune hepatitis: nitrofurantoin, methyldopa, minocycline, hydralazine, alpha interferon, beta interferon, cholesterol lowering agents

7.

Liver biopsy showing features of chronic hepatitis with interface hepatitis and prominence of plasma cells

8.

Prompt response to corticosteroid therapy (decrease in ALT levels by half within 2 weeks)

9.

Ultimate resolution upon stopping the medication (and corticosteroid therapy), although resolution may be slow and may call for corticosteroid therapy for several months.

Proof that a drug is the cause of an autoimmune hepatitis-like syndrome is sometimes difficult, particularly if the medication is not a common cause. Most diagnostic is a resolution of the syndrome with withdrawal of the agent. Improvements, however, often take several weeks to begin and the presence of features of autoimmunity often prompt treatment with corticosteroids, the usual therapy of idiopathic autoimmune hepatitis. While controlled trials of corticosteroid therapy have not been conducted in drug induced autoimmune hepatitis, case reports and small case series suggest that corticosteroids induce a prompt improvement in serum bilirubin and aminotransferase elevations. In this regard as well, drug induced resembles idiopathic autoimmune hepatitis.

Distinction between drug induced and idiopathic autoimmune hepatitis can also be made if corticosteroids can be withdrawn without a relapse in disease. Relapse upon withdrawal of immunosuppression is common with idiopathic but atypical of drug induced autoimmune hepatitis (at least if the course of corticosteroids was adequate). Also, typical of drug induced injury is disappearance of the autoantibodies with resolution. Spontaneous improvements can occur in idiopathic autoimmune hepatitis, but relapse is ultimately common, although sometimes months or years later. Once resolved, relapse of an autoimmune hepatitis initially triggered by a drug is unusual unless the drug is restarted.

Clinical Features and Course. Drug induced autoimmune hepatitis typically presents with the insidious onset of fatigue, weakness, nausea, poor appetite, right upper quadrant discomfort and jaundice or pruritus. The pattern of enzyme elevations is predominantly hepatocellular with prominent elevations in serum aminotransferase and modest increases in alkaline phosphatase levels. The severity of drug induced autoimmune hepatitis ranges widely from mild serum enzyme elevations with few if any symptoms, to symptomatic but anicteric hepatitis, to an acute or chronic hepatitis, to acute liver failure. If severe and prolonged injury occurs, evidence of hepatocellular dysfunction can arise including low serum albumin and prolongation of prothrombin time, and the injury can result in cirrhosis and end stage liver disease. In rare instances, the injury is acute and severe, leading rapidly to acute liver failure and death or need for liver transplantation.

Most characteristic of drug induced autoimmune hepatitis are the autoimmune features, including increases in serum immunoglobulins (particularly IgG) and autoantibodies. The most typical autoantibodies are directed at nuclear (ANA), smooth muscle (SMA) and liver-kidney microsomal antigens (anti-LKM). The pattern and specificity of these autoantibodies may provide a clue to the pathogenesis of the apparent autoimmune process induced by the medication, occasionally having specificity for a metabolic intermediate or drug-protein adduct such as antibody to altered P450 enzymes or anti-CYP 2D6. However, by far the most common autoantibodies found are nonspecific ANA reactivity or SMA with the typical anti-actin reactivity.

Drug induced autoimmune hepatitis typically arises 2 months to several years after starting the implicated medication and the clinical pattern of disease is more like chronic than acute liver injury. Patients may have extrahepatic manifestations of autoimmunity as well, such as arthralgias and skin rash or even other major organ involvement such as pneumonitis, nephritis or cerebritis. A history of other autoimmune diseases may be present but is not typical. Liver biopsy, if performed, usually shows elements that are typical of autoimmune hepatitis such as interface hepatitis with a lymphocytic or lymphoplasmocytic infiltrate and hepatic rosette formation. There may be variable degrees of fibrosis. In some instances, advanced fibrosis or even cirrhosis is present at the time of diagnosis, particularly if diagnosis and discontinuation of the medication is delayed.

Many of the agents that can induce autoimmune hepatitis can also induce other autoimmune conditions or symptoms such as a lupus-like syndrome (minocycline), pneumonitis (nitrofurantoin), hemolytic anemia (methyldopa) and acute thyroiditis (interferon alfa). Indeed, in these conditions there may be minor serum aminotransferase elevations indicative of a mild component of hepatic injury.

The important differential diagnosis for drug induced autoimmune hepatitis is, of course, spontaneous or idiopathic autoimmune hepatitis, a disease that can have spontaneous remissions and relapses, but that is typically severe and progressive without adequate immunosuppressive therapy. Some cases of drug induced autoimmune hepatitis do not resolve completely upon withdrawal of the medication, and some require long term immunosuppressive therapy and have had relapses (some severe) when corticosteroids were stopped. In these instances, it is not clear whether the autoimmune hepatitis-like syndrome was actually caused by the medication or was a co-incidence, arising spontaneously in a patient being treated long term with a common medication. An intermediate hypothesis is also possible, that the medication triggered the onset of autoimmune hepatitis in a susceptible patient that may have ultimately developed the disease, even without the medication. Often, this last possibility cannot be excluded. Spontaneous, idiopathic autoimmune hepatitis obviously has triggering factors for its onset and a medication or liver injury may be such a trigger. These issues call for careful clinical investigation of cases of autoimmune hepatitis associated with medications for genetic and environmental factors that might reveal the cause of this important syndrome.

Mechanism of Injury. The pathogenesis of autoimmune hepatitis due to medications is not clear. However, it is likely that hepatocytes, owing to their ability to metabolize drugs, can form drug-protein adducts which may be immunogenic. These newly formed antigens may, in turn, incite CD4+ and CD8+ cytotoxic lymphocyte and natural killer cell proliferative responses. The drug metabolizing enzymes (P450; CYP) are likely targets for neoantigen formation. Additionally, drugs may interact with T cell receptors to induce, and thus induce an immune response without forming cellular protein adducts.

There appears to be a genetic predisposition to autoimmune hepatitis from medications, and this may be the same genetic predisposition as for idiopathic autoimmune hepatitis. Indeed, in some instances, the hepatic injury from a medication appears to trigger autoimmune hepatitis and medications may be one of the environmental features that underlie the onset of this syndrome in susceptible patients. Genetic testing has shown that autoimmune hepatitis is closely linked to HLA types, in particular the extended haplotype of A3, B8, Dr3 and DrB03*0301 in Caucasians and northern Europeans with autoimmune hepatitis. This haplotype may also predispose to drug induced autoimmune hepatitis, although this has not been well defined.

Management. The most important element in management of drug induced autoimmune hepatitis is to recognize the possible role of the medication and discontinue it promptly. An adequate clinical and laboratory evaluation to assess the severity and characteristics of the injury is also important, as is testing for other possible causes. Liver biopsy is not necessary for diagnosis, but can be helpful and will provide information on the severity of the injury and the presence of any fibrosis.

Active therapy for autoimmune hepatitis using corticosteroids is often used in cases where a medication is thought to be the cause, although without controlled medical evidence for efficacy. Nevertheless, corticosteroid therapy is appropriate if recovery does not start within 1 to 2 weeks of stopping the implicated agent or if there is any evidence of hepatic failure, such as prolongation of the prothrombin time, deepening jaundice, hepatic encephalopathy, serum hypoalbuminemia with edema or ascites. An appropriate regimen would be prednisone (or its equivalent) in doses of 20 to 60 mg daily, decreasing the dose rapidly upon evidence of improvement with the goal to discontinue therapy completely within 3 to 6 months. High dose corticosteroid therapy can have significant and even permanent side effects, so that use of lower initial doses (20 to 30 mg daily) is appropriate for cases without signs of hepatic failure. An important element in using corticosteroids is careful follow up for at least six months after they have been discontinued to document full recovery and the absence of relapse, which can be severe and even fatal in patients with spontaneous autoimmune hepatitis in whom corticosteroids are stopped early.