Minocycline as a Cause of Drug-Induced Autoimmune Hepatitis Report of Four Cases and Comparison With Autoimmune Hepatitis

Transcription

1 Anatomic Pathology / MINOCYCLINE-INDUCED AUTOIMMUNE HEPATITIS Minocycline as a Cause of Drug-Induced Autoimmune Hepatitis Report of Four Cases and Comparison With Autoimmune Hepatitis Neal S. Goldstein, MD, 1 Nasser Bayati, MD, 2 Ann L. Silverman, MD, 2 and Stuart C. Gordon, MD 2 Key Words: Liver; Biopsy; Minocycline; Autoimmune; Hepatitis; Drug; Drug reaction; Eosinophils Abstract We describe the clinical and liver biopsy morphologic features for 4 patients with minocyclineinduced autoimmune hepatitis (group 1). We compared the serum laboratory values and liver biopsy findings from group 1 with those from 10 patients with sporadic autoimmune hepatitis (group 2). All patients in group 1 had positive serum antinuclear antibody titers, but none had positive serum anti smooth muscle antibody titers. The morphologic findings of group 1 biopsies were those of autoimmune hepatitis in all 4 patients. In addition, 1 of these biopsy specimens also had scattered single eosinophils, unlike autoimmune hepatitis. The mean histologic activity index scores for patients in groups 1 and 2, respectively, were 6.7 and 5.4. No patients in group 1 had marked bridging fibrosis or cirrhosis, compared with 4 of 10 patients in group 2. Minocycline-induced autoimmune hepatitis is usually identical to sporadic autoimmune hepatitis. The absence of eosinophils does not exclude the possibility of a minocycline cause. In the absence of clinical or morphologic differences, a recent ingestion of minocycline should be excluded before the diagnosis of sporadic autoimmune hepatitis is established. Whether the drug is unmasking latent autoimmune hepatitis is unclear. Minocycline is a semisynthetic tetracycline antibiotic that frequently is used to treat acne vulgaris. Syndromic side effects of minocycline include serum sickness with arthritis, 1,2 systemic lupus erythematosus (SLE), 3-10 and autoimmune-like hepatitis. 4,11-14 The liver disease in the latter 2 syndromes seems to be similar. 5,15,16 Although there are numerous case reports of and several review articles about autoimmune hepatitis like or SLE-like minocycline-induced autoimmune hepatitis in the literature, we identified only 10 patients from 5 studies in which the authors reported the histologic findings of this entity, of which 4 have detailed morphologic descriptions. We believe that most pathologists are unaware that minocycline can cause a hepatitis similar to autoimmune hepatitis. In addition, the histologic features of minocyclineinduced autoimmune hepatitis are not completely clear from these previously described patients, including whether there are histologic distinctions between minocycline-induced autoimmune hepatitis and sporadic autoimmune hepatitis. Identification of histologic differences between minocyclineinduced autoimmune hepatitis and the usual type of autoimmune hepatitis would be useful because clinical distinction between the diseases is sometimes difficult, and a history of minocycline use may not be elicited. We describe the clinicopathologic details of 4 patients with minocycline-induced autoimmune hepatitis (group 1), including the morphologic features of the liver biopsy specimens. We also compared the morphologic features of the liver biopsy specimens and laboratory values for group 1 with those for 10 patients with sporadic autoimmune hepatitis (group 2) in a blinded manner to identify histologic features of minocycline-induced autoimmune hepatitis that may distinguish these 2 diseases. American Society of Clinical Pathologists Am J Clin Pathol 2000;114:

2 Goldstein et al / MINOCYCLINE-INDUCED AUTOIMMUNE HEPATITIS Materials and Methods Patients were identified from the clinical hepatology practice of one of us (S.C.G.) for the period January 1990 through December The 4 patients in group 1 were seronegative for the hepatitis A, B, and C viruses. All of the patients underwent serum liver function, erythrocyte sedimentation rate (ESR), antinuclear antibody (ANA), and anti smooth muscle antibody tests. All liver biopsy specimens were paraffin embedded and formalin fixed. They were cut at 0.4-µm thick sections and sequentially stained with H&E, reticulin, periodic acid Schiff, periodic acid Schiff with diastase, Gomori trichrome, Prussian blue, and again with H&E. Appropriate control stains were performed that worked correctly in all cases. All of the liver biopsy specimens were reviewed and scored by one of us (N.S.G.) in a blinded manner. The following features were evaluated in the liver biopsy specimens: 1. Number of portal tracts in each biopsy specimen. 2. Histologic activity index (HAI) using the necroinflammatory grading system of Ishak et al. 17 Periportal or periseptal interface (piecemeal) hepatitis, parenchymal focal (spotty) lytic necrosis, apoptosis, focal inflammation, and portal inflammation were each graded on a scale of 0 to 4. Confluent necrosis was graded on a scale of 0 to Number of acidophilic bodies quantified on a scale of 0 to 3 similar to the criteria of Brunt et al 18 : 0, no acidophilic bodies; 1, occasional acidophilic bodies per 20 field; 2, an average of 2 to 4 acidophilic bodies per 20 field; and 3, an average of more than 4 acidophilic bodies per 20 field. 4. Hepatocyte lobular architectural disarray 19 quantified on a scale of 0 to 3: 0, no hepatocyte architectural disarray; 1, slight disarray with an average of fewer than 3 rosettes, foci of thickened hepatocyte plates, or hepatocyte dropouts per lobule; 2, an average of 3 to 7 of these features per lobule; and 3, more than 7 of these features per lobule. 5. Portal bile ductular proliferation quantified on a scale of 0 to 3: 0, no portal bile ductule proliferation; 1, mild bile ductule proliferation, defined as fewer than 3 peripheral bile ductules in fewer than one third of the portal tracts; 2, moderate bile ductule proliferation, defined as 3 or more bile ductules in one third to two thirds of the portal tracts; and 3, marked bile ductule proliferation, defined as 3 or more bile ductules in more than two thirds of the portal tracts. 6. Number of eosinophils quantified on a scale of 0 to 4: 0, no eosinophils in the biopsy specimen; 1, rare eosinophils, defined as fewer than 3 throughout the entire specimen; 2, mild eosinophils, defined as scattered single eosinophils throughout the lobules; 3, moderate eosinophils, defined as occasional single eosinophils in the lobules and rare 2- to 3-cell collections; and 4, marked eosinophils, defined as numerous single eosinophils with frequent collections and aggregates. 7. Fibrosis stage score quantified using the 0 to 6 scoring system of Ishak et al 17 : 0, no fibrosis; 1, fibrous expansion of some portal areas, with or without short fibrous septa; 2, fibrous expansion of most portal areas, with or without short fibrous septa; 3, fibrous expansion of most portal areas with occasional portal-portal bridges; 4, fibrous expansion of portal areas with marked bridging (portal-portal and portal-central); 5, marked bridging with occasional nodules (incomplete cirrhosis); and 6, cirrhosis (possible or definite). 8. Macrovesicular steatosis quantified on a scale of 0 to 3, using the criteria of Brunt et al, 18 based on the percentage of involved hepatocytes in the biopsy specimen: 0, no steatosis; 1, up to 33% of hepatocytes; 2, 33% to 66%; and 3, more than 66%. 9. Microvesicular steatosis quantified on a scale of 0 to 3 using percentage criteria identical to those for macrovesicular steatosis. 18 Results Group 1 Patient 1 A 16-year-old sought care because of fatigue and nausea. She had been ingesting 100 mg of minocycline daily for the treatment of acne for 1 year before being examined at our clinic. At the time of initial examination, her liver function test results were as follows: aspartate aminotransferase (AST), 1,018 U/L (reference range, U/L); alanine aminotransferase (ALT), 700 U/L (reference range, 8-28 U/L); alkaline phosphatase (ALP), 138 U/L (reference range, U/L); and total bilirubin, 6.5 mg/dl (111 µmol/l; reference range, mg/dl [3-19 µmol/l]). Her erythrocyte sedimentation rate (ESR) was 10 mm/h, and the serum ANA titer was positive (titer, 1:320; homogeneous pattern). A diagnosis of minocycline-induced autoimmune hepatitis was made. The minocycline was discontinued, oral corticosteroid therapy was started, and there was rapid clinical improvement. A follow-up visit, 2 weeks after the corticosteroid therapy was started, revealed the patient was asymptomatic. Her liver function test results normalized during the 4- week period after corticosteroid therapy was started, and a corticosteroid taper schedule was instituted. Her liver function test results were normal at a follow-up visit 3 months after corticosteroid therapy was discontinued. 592 Am J Clin Pathol 2000;114: American Society of Clinical Pathologists

3 Anatomic Pathology / ORIGINAL ARTICLE A liver biopsy was performed 3 days after the initial examination. It showed moderately to markedly active hepatitis, with an HAI score of 10. There was dense portal lymphoplasmacytic inflammation (score, 3) with extensive interface (piecemeal) necrosis (score, 4) and numerous foci of spotty inflammatory parenchymal inflammation associated with hepatocyte dropout and acidophilic bodies (score, 3) Image 1. The overall pattern of injury was that of autoimmune hepatitis. There were no eosinophils in the portal tracts or parenchyma. The parenchymal architecture was distorted and in marked disarray shown by focal collapse and numerous rosettes. The bile ducts and central venules were normal. Trichrome stain showed no abnormal collagen deposition. Patient 2 A 22-year-old woman who had been taking minocycline for 4 months for the treatment of acne sought care because of several weeks of anorexia, fatigue, and new-onset nausea. Her minocycline dosage had been 200 mg/d for the first 2 months and 100 mg/d for the second 2 months. The only abnormality revealed by physical examination was mild epigastric and right upper quadrant abdominal tenderness. Her initial serum liver function test results were as follows: ALT, 1,288 U/L; AST, 433 U/L; ALP, 253 U/L; and total bilirubin, 2.3 mg/dl (39 µmol/l). The serum ANA titer was positive (titer, 1:80; homogeneous pattern). A diagnosis of minocycline-induced autoimmune hepatitis was made. The minocycline was discontinued, oral corticosteroid therapy was started, and she noted a rapid and marked improvement of her symptoms. Her liver function test results resolved to within the reference range 5 weeks after starting corticosteroid therapy. Liver function test results 6 months after corticosteroid therapy was discontinued were also normal. A liver biopsy was performed 3 days after initial examination. It showed moderately to markedly active autoimmune hepatitis, with an HAI score of 10 Image 2. The degree and pattern of activity were identical to those shown in the liver biopsy specimen of Patient 1. In addition, there were scattered single eosinophils admixed with the portal tract and parenchymal lymphoplasmacytic inflammation (eosinophil inflammation score, 2). Trichrome stain showed no abnormal collagen deposition. Patient 3 A 52-year-old woman who had been taking 100 mg/d of minocycline intermittently for 12 years for the treatment of acne sought care because of myalgia and fatigue. She had no abdominal discomfort, and physical examination findings were normal. Her initial serum laboratory results were as follows: ALT, 137 U/L; AST, 78 U/L; ALP, 83 U/L; and total bilirubin, 0.4 mg/dl (7 µmol/l). The ESR was 14 mm/h, and the ANA titer was positive (titer, 1:160; speckled pattern). A diagnosis of minocycline-induced autoimmune hepatitis was made. The minocycline was discontinued, oral corticosteroid therapy was started, and she noted a rapid resolution of her symptoms. Her liver function test results were within the reference range 3 weeks after initiation of corticosteroid therapy. A B Image 1 A, Minocycline-induced autoimmune hepatitis. Architectural disarray and numerous foci of spotty inflammation and piecemeal necrosis are noted along the entire edge of the portal tract (right) (H&E, 26). B, Higher magnification of Image 1A. The inflammation is lymphoplasmacytic without eosinophils. Numerous acidophilic bodies are present. The parenchyma has undergone extensive pseudoacinar transformation secondary to the hepatocyte dropout. Many hepatocytes have rarified cytoplasm, making the sinusoids difficult to identify (H&E, 64). American Society of Clinical Pathologists Am J Clin Pathol 2000;114:

4 Goldstein et al / MINOCYCLINE-INDUCED AUTOIMMUNE HEPATITIS Image 2 Portal tract with moderate lymphoplasmacytic inflammation with admixed eosinophils. Continuous piecemeal necrosis is evident along the length of the portal tract, and numerous inflammatory cells are present between hepatocytes within the parenchyma (H&E, 48). They also were within the reference range at a follow-up visit 8 months after corticosteroid therapy was discontinued. A liver biopsy was performed 1 day after the initial examination. Unlike patients 1 and 2, the liver biopsy from patient 3 revealed only minimal to mild parenchymal, portal, lymphoplasmacytic inflammation and mild interface hepatitis (piecemeal necrosis), producing an HAI score of 4 Image 3. Eosinophils were present in extremely rare numbers; 2 were seen throughout the entire biopsy. There were no acidophilic bodies and extremely rare foci of hepatocyte architectural disarray. There was slight fibrous expansion of the portal tracts by collagen in a stellate pattern but no portal-portal or portal-central venular fibrous bridges. Patient 4 An 18-year-old woman had been taking 100 mg/d of minocycline intermittently for the treatment of acne for 2 years before seeking care because of fatigue and nonspecific arthritis for several months. Her initial physical examination findings were normal. Serum laboratory values were as follows: ALT, 343 U/L; AST, 286 U/L; ALP, 49 U/L; total bilirubin, 1.2 mg/dl (20 µmol/l); and ESR, 33 mm/h. The serum ANA titer was positive (titer, 1:320; speckled pattern). A diagnosis of minocyclineinduced autoimmune hepatitis was made. The minocycline was discontinued, and oral corticosteroid therapy was started. Because the patient did not tolerate oral prednisone, azathioprine was started, and the corticosteroid therapy was discontinued by using a tapering dosage schedule. The patient noted a rapid improvement of her symptoms. At a follow-up visit 10 weeks after the azathioprine was started, serum liver function test results were within the reference range. A liver biopsy was performed 4 days after initial examination. It was remarkable because of the paucity of inflammatory and parenchymal changes, especially in the light of her elevated liver function test results. There were minimal parenchymal, portal lymphoplasmacytic inflammation and minimal interface (piecemeal) hepatitis, producing an HAI score of 3. There were no acidophilic bodies or rosettes. Despite the lack of inflammation and active parenchymal injury, there was fibrous expansion of most of the portal tracts with occasional portal-portal bridges. None of the biopsy specimens from the 4 group 1 patients included microvesicular or macrovesicular steatosis, bile ductule proliferations, intrahepatocyte iron, or periodic acid Schiff diastase resistant globules. Comparison of Groups 1 and 2 The liver biopsy specimens and laboratory test results for group 1 were compared with those for group 2, who were matched with group 1 patients for sex and age at the time of the biopsy. The 10 patients in group 2 had no history of taking a hepatitis-causing medication, and all experienced rapid clinical and biochemical resolution with corticosteroid therapy. All patients in group 1 had positive serum ANA titers (range, 1:80-1:320), but none were anti smooth muscle antibody positive Table 1. Seven of 10 patients in group 2 had positive serum ANA titers (range, 1:40-1:1,280), and 3 had positive anti smooth muscle antibody serum titers. The mean ± SD serum IgG level was 1,585 ± 288 mg/dl (15.8 ± 2.9 g/l) in group 1 and 2,677 ± 1, 096 mg/dl (26.9 ± 11.0) in group 2. There were no appreciable differences in the peak serum AST, ALT, or ALP values between the groups. The median baseline ESR was 14 mm/h in group 1 and 37 mm/h in group 2. The mean liver biopsy HAI scores for groups 1 and 2 were 6.7 and 5.4, respectively. No patients in group 1 had marked bridging fibrosis or cirrhosis, compared with 4 of 10 patients in group 2. One patient in group 2 had rare scattered eosinophils in the portal tracts. With the scoring system for the diagnosis of autoimmune hepatitis, all 10 patients in group 2 had definite or probable autoimmune hepatitis, and all 4 patients in group 1 had probable autoimmune hepatitis (Table 1). 20 Discussion Minocycline-induced autoimmune hepatitis is a chronic hepatitis that clinically can mimic autoimmune hepatitis, 594 Am J Clin Pathol 2000;114: American Society of Clinical Pathologists

5 Anatomic Pathology / ORIGINAL ARTICLE A B C Image 3 A, Mild portal lymphoplasmacytic inflammation and minimal parenchymal architectural distortion are evident (H&E, 26). B, Higher magnification of the portal tract in Image 3A. Mild piecemeal necrosis is present. Eosinophils, which are not seen in this Image, were extremely rare (H&E, 64). C, There is slight fibrous expansion of the portal tracts by collagen in a stellate pattern (H&E, 48). SLE hepatitis, or an overlap syndrome. 1,15,21 Fulminant minocycline hepatitis also has been reported, but this seems to be extremely rare. 1,15,21 Three recent, thorough review articles identified 24 minocycline-induced autoimmune hepatitis single case reports or small series in the literature. 1,21,22 We reviewed all of the articles referenced in the bibliographies of these 3 articles and performed a National Library of Medicine MEDLINE literature search for minocycline and liver. We identified 10 patients from 5 studies in which the authors reported the histologic features of a liver biopsy specimen Table 2. Two of the studies provided photographs of the liver biopsy specimens from patients with minocyclineinduced autoimmune hepatitis. 12,13 Descriptions of the morphologic features of the liver biopsy specimens of 6 of the 10 patients are terse. These descriptions provide little insight into the morphologic features of the liver biopsy specimens and, to some degree, impede rather than forward the characterization of minocycline-induced hepatitis. It is unclear from these publications whether the term chronic active hepatitis refers to a process with piecemeal necrosis or to a specific clinical entity. 23 As pointed out by Ludwig, 23 interpreting the morphologic pattern of piecemeal necrosis (chronic active hepatitis), as used by pathologists as a sign of the clinical disease chronic active hepatitis may have grave consequences for patients. Three studies describing 4 patients provide detailed descriptions of the changes evident in the liver biopsy specimens (Table 2) The descriptions of the changes are similar to the histologic features noted for the 4 patients in group 1, with the exception of 1 patient who had numerous eosinophils. 13 The absence of numerous eosinophils in the liver biopsy specimens of our group 1 patients also contrasts with the statement made by Angulo et al 21 in their review American Society of Clinical Pathologists Am J Clin Pathol 2000;114:

6 Goldstein et al / MINOCYCLINE-INDUCED AUTOIMMUNE HEPATITIS Table 1 Comparison of Clinical Features Between Minocycline-Induced Autoimmune and Usual Type Autoimmune Hepatitis Feature Minocycline Group Autoimmune Hepatitis Group No. of patients 4 10 Mean age, y Median serum alanine aminotransferase, U/L Median erythrocyte sedimentation rate, mm/h Median serum antinuclear antibody titer 1:240 1:180 Mean serum IgG, mg/dl (g/l) 1,585 ± 288 (15.8 ± 2.9) 2,677 ± 1,096 (26.8 ± 11.0) Autoimmune hepatitis diagnosis score Table 2 Extracted Text From Studies of Morphologic Features of Liver Biopsy Specimens in Minocycline-Induced Autoimmune Hepatitis Reference Malcolm et al 13 Gough et al 16 Angulo et al 29 Herzog et al 11 Teitelbaum et al 12 Teitelbaum et al 12 Malcolm et al 13 Description of Morphologic Features of Liver Biopsy Specimens Patient 5 had a biopsy that showed mild lobular hepatitis. Four cases had liver biopsies that showed chronic active hepatitis, acute hepatitis with confluent necrosis, chronic hepatitis with bridging and piecemeal necrosis, and chronic active hepatitis, respectively. A liver biopsy revealed chronic active hepatitis. A liver biopsy showed chronic portal inflammation with lymphocytes, plasma-cells, eosinophils, and a minimal degree of piecemeal necrosis with some intralobular inflammation including councilman bodies were the main histologic features of the biopsy. Patient 2 had a liver biopsy that revealed characteristic features of chronic autoimmune hepatitis, including marked portal lymphoplasmacytic inflammation, diffuse hepatocellular damage, moderate necroinflammatory activity, and marked portal fibrosis. [There was] considerable lobular collapse, nodular transformation, and marked pericellular fibrosis. Patient three s biopsy revealed chronic hepatitis with diffuse hepatocellular damage, marked necroinflammatory activity, and marked portal and pericellular fibrosis with portal-portal bridging was consistent with an autoimmune process. changes of drug-induced chronic hepatitis with a prominent eosinophilic infiltrate. There was marked fibrous expansion of portal tracts with some early bridging. A heavy inflammatory infiltrate was especially prominent at the interface of portal tracts with hepatocytes, with erosion of the limiting plates and piecemeal necrosis. Lymphocytes predominated, but eosinophils and plasma cells were also relatively numerous. Milder lobular inflammation was present in acinar zones 2 and 3. No granulomas were seen, and there was no fatty change. A diagnosis of moderately severe chronic active hepatitis was made. article summarizing the liver biopsy findings for 5 patients in the literature as showing mild to moderate piecemeal necrosis and inflammatory infiltrate on the interface of portal tracts with hepatocytes, consisting mainly of lymphocytes, although plasma cells and eosinophils were also numerous. Combining the previous 4 cases with detailed liver biopsy descriptions and the 4 cases we describe in this study (group 1) suggests that the morphologic features of most minocycline-induced autoimmune hepatitis are histologically identical to those of autoimmune hepatitis. An eosinophil cell component, which may cause a pathologist to suggest a drug cause, seems to be rare and, when present, is usually spare and rarely prominent. The liver biopsies in minocycline-induced autoimmune hepatitis also can show only minimal to mild activity, similar to autoimmune hepatitis. Minimal to moderate autoimmune hepatitis may have no pathognomonic histopathologic features. 24,25 Without the clinical knowledge of minocycline ingestion or an autoimmune hepatitis like or SLE-like clinical syndrome, a liver biopsy in minimally active minocycline-induced autoimmune hepatitis may be interpreted by a pathologist as chronic active hepatitis of unknown cause. The degree of fibrosis in the liver biopsy specimen cannot be used to distinguish between minocycline-induced autoimmune hepatitis and autoimmune hepatitis because rare cases of minocycline-induced autoimmune hepatitis have had marked bridging fibrosis, which is not an unexpected finding in the biopsy specimens in autoimmune hepatitis. 12,19 Fortunately, marked fibrosis and cirrhosis seem to be rare in minocyclineinduced autoimmune hepatitis. Although most studies do not comment on the degree of fibrosis, 1,11,13,21,26 one gets the impression that substantial fibrosis is uncommon in minocycline-induced autoimmune hepatitis because many authors, including those describing patients who did not undergo liver biopsy, state that during follow-up, patients did not develop chronic liver disease, and the liver function test results normalized once the minocycline was discontinued. 4,6,12,16,27-29 The 4 patients we describe (group 1) support this notion. Two patients had no fibrosis, 1 had slight portal fibrous expansion, and 1 had rare portal-portal fibrous bridges. One of the striking features of minocycline-induced autoimmune hepatitis is the temporal relationship between a patient s initial minocycline ingestion and the clinical onset of the autoimmune hepatitis like syndrome. Review articles 596 Am J Clin Pathol 2000;114: American Society of Clinical Pathologists

7 Anatomic Pathology / ORIGINAL ARTICLE have noted that minocycline-related autoimmune disorders develop an average of 2 years (range, 3 days to 6 years) after the initiation of drug therapy. 1,21 In our study, the interval between the start of minocycline ingestion and the patient s seeking care ranged from 4 months to 12 years. A long interval between ingestion and start of autoimmune-type signs and symptoms seems to be a characteristic of drugs that have been reported to cause autoimmune hepatitis, including methyldopa and nitrofurantoin. 30 Both of these drugs also can induce a positive serum ANA titer and produce autoimmune hepatitis like changes in liver biopsy specimens. One major difference between minocycline and these other drugs seems to be the ability to induce significant fibrosis. Minocycline-induced cirrhosis is uncommon, while patients with nitrofurantoin hepatitis not infrequently present in the cirrhotic phase of the disease. 30 It is important that pathologists keep minocycline and these other drugs in mind as a cause of autoimmune hepatitis, especially because the long interval between initiation of drug therapy and disease development may make a drug cause seem unlikely, as may the frequent absence of eosinophils in the biopsy specimen. 1 The mechanism by which minocycline induces hepatitis is unknown. One group of authors showed that the serum from a patient with minocycline-induced autoimmune hepatitis immunoprecipitated a 50- and a 90-kd protein from a human hepatoma cell line. 11 Antibodies in this patient s serum also recognized the antigens CYP3A6 and CYP2C4 and liver microsomes at high concentrations (titer, 1:800). These results led the authors to conclude that the patient s serum contained antibodies that recognized different microsomal cytochromes, supporting the hypothesis that minocycline somehow induced the production of autoantibodies, possibly through its metabolism and generation of the 50- and 90-kd proteins. 11 Possibly, minocycline induces the upregulation or causes the production of specific antigens and acts as an initiating factor for otherwise latent autoimmune hepatitis. If so, one can ask why minocycline-induced autoimmune hepatitis is not more common given the number of patients who are taking minocycline for acne? Possibly this induction and initiation of the autoimmune inflammatory process occurs in patients with unique HLA characteristics. The question of whether minocycline causes or unmasks autoimmune hepatitis in a susceptible population of young women or is unrelated to the development of sporadic autoimmune hepatitis in an otherwise at-risk patient group is unknown. It will be interesting to observe the 4 patients in the present study (group 1) over time to see whether autoimmune hepatitis reactivates, despite not taking minocycline. It is unknown whether the autoimmune hepatitis like process can resolve once the minocycline has been discontinued or whether some of these patients will require long-term immunosuppressive therapy, similar to patients with sporadic autoimmune hepatitis. If corticosteroid therapy can be discontinued without relapse of the hepatitis, the case for minocycline as the cause of hepatitis is strengthened, rather than minocycline simply unmasking otherwise sporadic-type autoimmune hepatitis. We describe 4 patients with minocycline-induced autoimmune hepatitis. The morphologic features of minocycline hepatitis were those of autoimmune hepatitis, except for 1 patient who also had scattered single eosinophils. The serologic features also can be similar except that some patients with autoimmune hepatitis had positive serum anti smooth muscle antibody titers, while none of the patients with minocycline-induced autoimmune hepatitis had a positive titer to this antibody. Pathologists should be aware that the absence of eosinophils in a biopsy specimen that has the morphologic features of autoimmune hepatitis does not exclude the possibility of minocycline-induced autoimmune hepatitis. From the 1 Department of Anatomic Pathology and 2 Division of Gastroenterology-Hepatology, William Beaumont Hospital, Royal Oak, MI. Address reprint requests to Dr Goldstein: Dept of Anatomic Pathology, William Beaumont Hospital, 3601 W Thirteen Mile Rd, Royal Oak, MI References 1. Knowles SR, Shapiro L, Shear NH. Serious adverse reactions induced by minocycline: report of 13 patients and review of the literature. Arch Dermatol. 1996;132: Kaufmann D, Pichler W, Beer JH. Severe episode of high fever with rash, lymphadenopathy, neutropenia, and eosinophilia after minocycline therapy for acne. Arch Intern Med. 1994;154: Lupton JR, Figueroa P, Tamjidi P, et al. An infectious mononucleosis-like syndrome induced by minocycline: a third pattern of adverse drug reaction. Cutis. 1999;64: Golstein PE, Deviere J, Cremer M. Acute hepatitis and drug-related lupus induced by minocycline treatment. Am J Gastroenterol. 1997;92: Crosson J, Stillman MT. Minocycline-related lupus erythematosus with associated liver disease. J Am Acad Dermatol. 1997;36: Pointud P. Minocycline related lupus [letter]. J Rheumatol. 1997;24: Quilty B, McHugh N. Lupus-like syndrome associated with the use of minocycline [letter]. Br J Rheumatol. 1994; 33: Byrne PA, Williams BD, Pritchard MH. Minocycline-related lupus. Br J Rheumatol. 1994;33: Inoue CN, Kondo Y, Suwabe N, et al. Minocycline-related lupus in childhood [letter]. Eur J Pediatr. 1994;153: Gordon PM, White MI, Herriot R, et al. Minocyclineassociated lupus erythematosus. Br J Dermatol. 1995;132: American Society of Clinical Pathologists Am J Clin Pathol 2000;114:

8 Goldstein et al / MINOCYCLINE-INDUCED AUTOIMMUNE HEPATITIS 11. Herzog D, Hajoui O, Russo P, et al. Study of immune reactivity of minocycline-induced chronic active hepatitis. Dig Dis Sci. 1997;42: Teitelbaum JE, Perez-Atayde AR, Cohen M, et al. Minocycline-related autoimmune hepatitis: case series and literature review. Arch Pediatr Adolesc Med. 1998;152: Malcolm A, Heap TR, Eckstein RP, et al. Minocyclineinduced liver injury. Am J Gastroenterol. 1996;91: Matsuura T, Shimizu Y, Fujimoto H, et al. Minocyclinerelated lupus [letter]. Lancet. 1992;340: Farrell GC. Drug-induced hepatic injury. J Gastroenterol Hepatol. 1997;12:S242-S Gough A, Chapman S, Wagstaff K, et al. Minocycline induced autoimmune hepatitis and systemic lupus erythematosus-like syndrome. BMJ. 1996;312: Ishak K, Baptista A, Bianchi L, et al. Histological grading and staging of chronic hepatitis. J Hepatol. 1995;22: Brunt EM, Janney CG, Di Bisceglie AM, et al. Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions. Am J Gastroenterol. 1999;94: Burgart LJ, Batts KP, Ludwig J, et al. Recent-onset autoimmune hepatitis: biopsy findings and clinical correlations. Am J Surg Pathol. 1995;19: Alvarez F, Berg PA, Bianchi FB, et al. International Autoimmune Hepatitis Group Report: review of criteria for diagnosis of autoimmune hepatitis. J Hepatol. 1999;31: Angulo JM, Sigal LH, Espinoza LR. Coexistent minocyclineinduced systemic lupus erythematosus and autoimmune hepatitis. Semin Arthritis Rheum. 1998;28: Boudreaux JP, Hayes DH, Mizrahi S, et al. Fulminant hepatic failure, hepatorenal syndrome, and necrotizing pancreatitis after minocycline hepatotoxicity. Transplant Proc. 1993;25: Ludwig J. Histopathological diagnosis and terminology of chronic hepatitis. J Hepatol. 1995;23(suppl 1): Ishak K. Pathologic features of chronic hepatitis: a review and update. Am J Clin Pathol 2000;113: Czaja AJ, Carpenter HA. Sensitivity, specificity, and predictability of biopsy interpretations in chronic hepatitis. Gastroenterology. 1993;105: Burette A, Finet C, Prigogine T, et al. Acute hepatic injury associated with minocycline. Arch Intern Med. 1984; 144: Bulgen DY. Minocycline-related lupus [letter]. Br J Rheumatol. 1995;34: Davies MG, Kersey PJ. Acute hepatitis and exfoliative dermatitis associated with minocycline. BMJ. 1989;298: Angulo JM, Sigal LH, Espinoza LR. Minocycline induced lupus and autoimmune hepatitis [letter]. J Rheumatol. 1999;26: Farrell GC. Drug-induced chronic hepatitis. In: Drug- Induced Liver Disease. Edinburgh, Scotland: Churchill Livingstone; 1994: Am J Clin Pathol 2000;114: American Society of Clinical Pathologists