ORIGINAL ARTICLES LIVER, PANCREAS, AND BILIARY TRACT

CLINICAL GASTROENTEROLOGY AND HEPATOLOGY 2009;7:98 103 ORIGINAL ARTICLES LIVER, PANCREAS, AND BILIARY TRACT Serologic Markers Do Not Predict Histologic Severity or Response to Treatment in Patients With Autoimmune Hepatitis VAIBHAV MEHENDIRATTA,* PRADNYA MITROO, ALESSANDRO BOMBONATI, VICTOR J. NAVARRO, SIMONA ROSSI, RAPHAEL RUBIN, and STEVEN K. HERRINE *Department of Medicine, Division of Gastroenterology and Hepatology, and Department of Pathology, Anatomy and Cell Biology, Thomas Jefferson University Hospital, Philadelphia, Pennsylvania See Lok AS et al on page 138 for companion article in the January 2009 issue of Gastroenterology. Background & Aims: Autoimmune hepatitis (AIH) is characterized by the presence of circulating autoantibodies, hypergammaglobulinemia, necroinflammatory histology, and a response to immunosuppressive drugs. The goal of this retrospective study was to determine whether the presence of antinuclear antibodies (ANAs) or anti smooth muscle antibodies (ASMAs) in patients with AIH correlated with clinical presentation, histologic findings, or response to immunosuppressive therapy. Methods: Fifty-two patients diagnosed with AIH, on the basis of the revised scoring system of International Autoimmune Hepatitis group, were reviewed. Data on age, gender, aminotransferase levels, autoantibody titers, treatment regimens, and response to treatment were recorded. Seropositivity was defined as ANA >1:40 or ASMA >1:40. Percutaneous liver biopsies obtained at the initial presentation were reviewed. Results: Forty-two patients with AIH (81%) were seropositive, and 10 (19%) were seronegative. Both groups were similar with respect to demographics, treatment regimens, and response to therapy. Histologic parameters were similar among the 2 groups, including portal and lobular inflammation, piecemeal necrosis, and centrilobular necrosis. There were no significant differences in aminotransferase levels at diagnosis or after treatment. Conclusions: The prevalence of ANAs or ASMAs did not correlate with the clinical or histologic severity of AIH at diagnosis. Furthermore, there was no correlation between antibody status and response to immunosuppressive therapy. Therefore, patients who meet the diagnosis of AIH on the basis of the revised scoring system of International Autoimmune Hepatitis Group should be given immunosuppressive therapy, regardless of antibody status. Autoimmune hepatitis (AIH) is a chronic hepatitis of unknown cause that occurs in children and adults of all ages. The disease is characterized by the presence of circulating autoantibodies, hypergammaglobulinemia, necroinflammatory histology, and response to immunosuppressive drugs (eg, prednisone and azathioprine). The diagnosis of AIH, which requires the presence of characteristic clinical features and the exclusion of other conditions that resemble AIH, has been codified and updated by the International Autoimmune Hepatitis Group (IAIHG). 1,2 Differences between a definite and probable diagnosis of AIH relate mainly to the degree of serum -globulin or immunoglobulin G elevation, levels of autoantibodies, and exposures to alcohol, medications, or infections that could cause liver injury. Histologic features of AIH variably include portal and lobular lymphoplasmacytic necroinflammatory infiltrates, piecemeal necrosis, and bridging necrosis. Severe cases are associated with the formation of liver cell rosettes and nodular regeneration. 2 These individual features (interface hepatitis, predominantly lymphoplasmacytic necroinflammatory infiltrate, and rosetting of liver cells) have been incorporated in the diagnostic scoring system of IAIHG and constitute 19% of the maximum score allowable for a definite diagnosis of AIH. 1 4 Absence of the above mentioned features and presence of biliary and other incompatible changes raise doubt about the diagnosis and reduce the score by 42%. Although it is reaffirmed that there are no morphologic findings characteristic of AIH, absence of compatible histologic features or presence of incompatible histologic features preclude a definite diagnosis of AIH, regardless of other supportive clinical findings. Antinuclear antibodies (ANAs) and anti smooth muscle antibodies (ASMAs) are the most common markers for type I AIH, but about 20% 30% of AIH patients present without significant titers ( 1:40) of these autoantibodies, whereas up to 20% have no detectable autoantibodies. 2,5 7 Although antibodies play a pivotal role in the initial diagnosis of AIH, they are neither pathogenic nor disease-specific. 8 Autoantibody titers at presentation correlate poorly with disease severity, degree of liver injury, and response to treatment. 8 Whether the absence of Abbreviations used in this paper: AIH, autoimmune hepatitis; ANA, antinuclear antibody; ASMA, anti smooth muscle antibody; ELISA, enzyme-linked immunosorbent assay; HEp-2, human epithelioid 2; IAIHG, International Autoimmune Hepatitis Group; IIF, indirect immunofluorescence; LKM, anti liver-kidney microsomes. 2009 by the AGA Institute 1542-3565/09/$36.00 doi:10.1016/j.cgh.2008.08.043

January 2009 SEROLOGY IN AUTOIMMUNE HEPATITIS 99 autoantibodies has a favorable prognosis on the degree of liver injury or response to immunosuppression is unclear. The goals of the present study were to calculate the prevalence of autoantibodies in patients diagnosed with AIH and to determine whether their presence correlates with clinical presentation, histologic findings, or response to immunosuppressive therapy. Materials and Methods Patient Selection Electronic billing records at our university hepatology practice were used to identify patients diagnosed with International Classification of Diseases, ninth revision code 573.3 (Hepatitis, Autoimmune/Not otherwise specified) between 1999 and 2004. The revised scoring system of the IAIHG was used to identify patients who met the probable diagnosis of AIH by having a pretreatment score of more than 10. Patients were included in the study if they were treated with immunosuppressive therapy, were followed for 6 months after treatment initiation, and had liver biopsy slides available for review. Exclusion criteria included pretreatment aggregate score 10, pretreatment score 10 but insufficient available follow-up data, features of other liver diseases such as hemochromatosis, Wilson s disease, primary biliary cirrhosis, alcoholic hepatitis, or viral hepatitis, or absence of liver tissue specimens for review. Patients who were found to have antibodies against hepatotropic viruses were included if they were nonviremic and the clinical features of their liver disease were consistent with AIH. Data Collection Charts of patients who met inclusion criteria were retrospectively reviewed to record age, gender, serum aminotransferase levels, serum autoantibody levels, and treatment regimens. ALT and AST levels at diagnosis and then at 1 month, 6 months, and 1 year after treatment were collected. Treatment failure was defined as worsening clinical (development of jaundice, ascites, or hepatic encephalopathy) or laboratory (increase in serum aminotransferase levels by 67%) features despite compliance with therapy. Data collection methods were approved by the Institutional Review Board of Thomas Jefferson University. Autoantibody Testing The technique used to test for routine antibodies associated with AIH was indirect immunofluorescence (IIF). Slides prepared from human epithelioid 2 (HEp-2) cell lines as substrate were incubated with diluted serum, and the presence of ANAs was detected by fluorescent anti-ig antibody. Similarly, a mouse stomach-kidney substrate was used to test for ASMAs, and specific staining pattern was recognized with the addition of diluted sera. If staining pattern was suspicious for presence of anti liver-kidney microsome type-1 (anti-lkm-1) antibody, further testing with a liver-kidney-stomach substrate was used to confirm presence of the autoantibody. Seropositivity was defined as ANA 1:40 and/or ASMA 1:40. Histologic Evaluation Liver biopsy slides obtained before treatment initiation were reviewed. Sections were stained with hematoxylin-eosin and Masson trichrome. Two pathologists, who were blinded to clinical and serologic data, reviewed all slides. The degree of liver injury on histology was defined by 8 individual categories including plasma cells, portal inflammation, lobular inflammatory activity, piecemeal necrosis, centrilobular necrosis, bile duct proliferation, cholestasis, and pericholangitis. Each category was graded on a scale of 0 4. In addition, the stage of fibrosis was assessed on a scale of 0 6 (Figure 1). Statistical Analysis Comparison between means was performed by Student t test. The Mann-Whitney U test was used to evaluate the difference between means for non-normally distributed data. All statistical analyses were conducted with SAS statistical software (version 9.1; SAS Institute, Cary, NC), with P.05 considered to be statistically significant. Results Patient Population Two hundred patients were identified by using International Classification of Diseases, ninth revision code 573.3, of whom 87 met the diagnosis of AIH according to revised scoring system of the IAIHG. Twelve patients were excluded as a result of insufficient clinical information, 17 were excluded as a result of absence of liver biopsy specimens for review, and 6 were excluded as a result of concomitant infection with hepatotropic viruses and presence of viremia. Two patients (1 seropositive and 1 seronegative) with positive anti-hcv antibody and negative HCV polymerase chain reaction RNA were included in the study. Results from 52 patients who met all inclusion criteria are discussed below. Demographics The mean age of patients was 53 15 years, with 41 (79%) being female. The patients with autoantibodies were similar to the patients without autoantibodies with respect to age and gender (Table 1). Serology Positive ANA titers ( 1:40) were found in 35 patients (67%), 33 patients (63%) were ASMA-positive ( 1:40), 26 (50%) were both ANA/ASMA-positive, and 10 (19%) were both ANA/ ASMA-negative. None of the patients in seropositive or seronegative groups had staining patterns suggestive for presence of anti-lkm-1 antibody by using IIF testing on a mouse stomachkidney substrate. Treatment Regimens At the time of diagnosis, patients were treated with prednisone either alone (9, 17%) or in combination with azathioprine (42, 79%). One patient was treated with hydroxychloroquine for underlying lupus, which led to improvement of liver function. The patients with autoantibodies were similar to the patients without autoantibodies with respect to initial treatment regimen used (Table 1). Because no preset protocol is followed at our institution, drug doses used at initiation of treatment varied widely among patients and ranged from 10 to 60 mg for prednisone and 50 to 100 mg for azathioprine. Treatment regimen and dosage used depended on treating physician s preference. The hepatology practice at our institution is run by 3 board-certified hepatologists. The number of

100 MEHENDIRATTA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 1 Figure 1. Degree of liver injury on histology was defined by 8 individual categories including plasma cells, portal inflammation, and piecemeal necrosis (A), lobular inflammatory activity (B), centrilobular necrosis (C), and fibrosis (D). patients treated by each individual hepatologist was similar between the seronegative and seropositive groups. Response to Therapy Four (10%) patients in the seropositive group had suboptimal responses to initial treatment, requiring addition of immunosuppressants to treatment regimen (2 treated with mycophenolate mofetil and 2 treated with tacrolimus). One patient (10%) in the seronegative group had worsening clinical and laboratory features requiring orthotopic liver transplantation. There was no statistically significant difference between the mean serum ALT and AST levels at diagnosis and at 1, 6, and 12 months after treatment between the seropositive and the seronegative groups (Table 1). Histologic Data The degree of lobular inflammatory activity was slightly higher in the seronegative group versus the seropositive group (mean score, 1.3 vs 0.95), but the results were not statistically significant (Table 2). Both groups had a similar degree of portal inflammation (1.3 vs 1.5), piecemeal necrosis (0.9 vs 1.2), and plasma cells (1.21 vs 0.9). Centrilobular necrosis was found in liver biopsy specimens of 2 (4%) seropositive patients and in no seronegative patients. Overall, 6 patients had cirrhosis at the time of diagnosis (5 patients with autoantibodies vs 1 patient without autoantibodies). However, the fibrosis stage was similar in patients with or without autoantibodies (1.6 vs 1.4). The extent of biliary changes such as bile duct proliferation (mean score, 0.5 vs 0.6), pericholangitis (0.36 vs 0.6), and cholestasis (0.31 vs 0.7) was comparable between seronegative and seropositive groups. However, changes in biliary architecture were not sufficiently prominent to raise doubt about the diagnosis of AIH. None of the biopsy specimens examined showed atypical features (such as steatosis, iron overload, viral features, or inclusions) to suggest an alternative diagnosis. Discussion The diagnosis of AIH incorporates a set of clinical, serologic, and histologic features that are not specific to the

January 2009 SEROLOGY IN AUTOIMMUNE HEPATITIS 101 Table 1. Patient Characteristics in Seropositive and Seronegative AIH Seropositive (n 42) Seronegative (n 10) P value Mean age 52 y 58 y.212 Female sex 35 (83%) 6 (60%).101 Treatment Prednisone alone 7 (16%) 2 (20%).322 Prednisone azathioprine 34 (79%) 8 (80%).367 Other 1 (2%) 0.500 Mean ALT At diagnosis 623 806.259 1 mo 122 68.227 6 mo 68 59.812 12 mo 70 43.675 Mean AST At diagnosis 567 725.412 1 mo 71 56.488 6 mo 48 58.630 12 mo 40 45.758 Alkaline phosphatase at diagnosis 143 163.454 disease. The diagnosis of AIH in patients who present without serum autoantibodies can be difficult, but it might be made on the basis of a combination of hepatic pattern of serum biochemical abnormalities, marked hypergammaglobulinemia with selective elevation of serum IgG, typical histologic findings, immunogenetic background, and a careful exclusion of other possible causes of liver diseases. 2 The proposed scoring system does not preclude the diagnosis of AIH in these patients, but there is limited literature to date explaining the natural progression of disease and response to conventional treatment in this cohort. Thus, our retrospective study aimed to determine whether the presence or absence of autoantibodies had any correlation or prognostic significance with regard to presentation, disease severity, or treatment response. We found that the presence or absence of autoantibodies did not correlate with the clinical or histologic severity of AIH at diagnosis. Furthermore, there was no correlation between antibody status and response to immunosuppressive therapy. Nineteen percent of patients diagnosed with AIH at our university practice were seronegative for conventional markers of AIH. There was no statistically significant difference between the degree of laboratory and histologic manifestations of hepatocellular inflammation between the seronegative cohort and their seropositive counterparts. ANAs and ASMAs are the most common markers for AIH, but about 20% 30% of AIH patients present without significant titers ( 1:40) of these autoantibodies. There is little evidence that autoantibodies play a part in the pathogenesis of AIH, 9 and titers correlate poorly with disease severity, degree of liver injury, and response to treatment. 8 Titers of ANA and ASMA contribute in calculating the IAIHG diagnostic score for patients with a probable or definite diagnosis of AIH. 2 A score of 1, 2, and 3 is assigned to autoantibody titers of 1:40, 1:80, and 1:80, respectively. A cut-off titer of 1:40 was used to define seropositivity in our study. Similar cutoff has been used by Feld et al 10 while studying the effects of symptoms on the natural history of AIH. They concluded that asymptomatic patients with AIH have a favorable prognosis and might not require immunosuppression. Although both groups had similar proportion of seropositive patients, no correlation was made between autoantibody status or titers and treatment response. In our study 6 patients in the seropositive group had antibody titer of 1:40. These patients did not differ from patients with higher titers ( 1:80) in terms of clinical presentation or response to immunosuppression. IIF was used to test for routine antibodies associated with AIH. A consensus statement in 2004 from the Committee for Autoimmune Serology of the IAIHG provided guidelines on how to test for autoantibodies relevant to AIH and concluded that IIF on fresh sections of multi-organ (liver, kidney, stomach) from rodents (usually rat) should be the first-line screening. 11 Problems associated with the IIF technique include the limited availability and variable quality of the tissue substrate, time-consuming technique, and intraoperator variability requiring highly trained and experienced personnel. This has given way to newer, less time-consuming, and easy to perform automated enzyme-linked immunosorbent assays (ELISAs). Either technique has its pros and cons and gives answers to different questions, such that results are not directly comparable. 12 14 Most liver-related autoantibodies can be detected by IIF when using a triple rodent tissue. HEp-2 cells can help to differentiate ANA patterns, and ethanol-fixed neutrophils can be used for the detection of ANCAs. In contrast, ELISAs give answers for (usually) preselected individual autoantibody specificities. Although the analytical sensitivity of ELISAs is satisfactory, their specificity varies according to the manufacturer, 14 whereas such problems are rather infrequent by IIF testing based on a triple rodent tissue substrate. 15 At the present time, ELISAs should be used to complement but not replace IIF. 16 Histologic features of AIH variably include portal and lobular lymphoplasmacytic necroinflammatory infiltrates, piecemeal necrosis, and bridging necrosis. Complex numeric systems have been proposed for grading and staging of chronic hepatitis, with the intent that these could be used in the statistical analysis in clinical trials. 17 Although there is no gold standard grading and staging system for AIH, the most widely accepted scoring system is the Histological Activity Index, either in its original form, Knodell score, 18 or the modified form, Ishak score. 19 The histologic scoring system used in the current study has been prepared by the authors along the lines of the modified Histological Activity Index and maintains a range of numbers without any missing numbers. Few additional criteria (such as bile duct proliferation, pericholangitis) have been in- Table 2. Histologic Features at Diagnosis Seropositive Seronegative P value Portal inflammation 1.3 1.5.617 Lobular inflammatory activity 0.95 1.3.416 Plasma cells 1.21 0.9.334 Piecemeal necrosis 0.9 1.2.617 Centrilobular necrosis 4/43 (10%) 0/10.100 Bile duct proliferation 0.5 0.6.700 Pericholangitis 0.36 0.6.303 Cholestasis 0.31 0.7.165 Fibrosis 1.6 1.4.753 Cirrhosis 5/42 (12%) 1/10 (10%).076

102 MEHENDIRATTA ET AL CLINICAL GASTROENTEROLOGY AND HEPATOLOGY Vol. 7, No. 1 corporated to generate biologically plausible and clinically useful information. Since the early 1990s, various authors have proposed an underlying autoimmune mechanism on patients with chronic hepatitis without significant titers of conventional serum autoantibodies. Johnson et al 20 reported a case series of 12 patients who presented with histologically florid chronic active hepatitis but without significant titers of conventional serum autoantibodies. Clinical and histologic remission was achieved in all cases within 8 months of immunosuppressive therapy. Six patients subsequently relapsed during or after withdrawal of immunosuppression. Czaja et al 21,22 and Kaymakoglu et al 23 studied the clinical features, genetic phenotype, and corticosteroid responsiveness in patients with severe cryptogenic chronic hepatitis and found no difference when compared with AIH patients. Thiele et al 24 studied the effect of corticosteroid-containing therapy in patients with non-b chronic hepatitis and varying anti hepatitis C virus status. Neither the presence nor the relative titers of conventional autoantibodies were significantly associated with steroid responsiveness in patients with chronic hepatitis in this study. More recently, Gassert et al 25 described the criteria for diagnosis of seronegative AIH in 6 patients earlier classified as having cryptogenic chronic hepatitis. The clinical and histologic features and response to corticosteroids in these 6 patients were similar to those of 22 patients with conventional type I AIH. Czaja 8 studied 107 patients with seropositive AIH and followed them for 128 9 months for response to immunosuppression. Eighty-one (76%) patients lost either one or both autoantibodies (ANA, ASMA) after initial treatment. Disappearance of autoantibodies was associated with lower serum aminotransferase, bilirubin, -globulin, and IgG levels. Interface hepatitis, bridging or lobular necrosis, and active cirrhosis were more commonly associated with ANA and/or ASMA seropositivity. However, in our study there was a slightly higher degree of lobular inflammation, portal inflammation, and piecemeal necrosis, which was not statistically significant. Furthermore, the aminotransferase levels at diagnosis were slightly higher in the seronegative group than in the seropositive group. It is possible that a delay in the initial diagnosis associated with seronegativity accounted for a slightly greater degree of disease severity at presentation. In the study by Czaja, 8 patients were followed from seropositivity to seronegativity after treatment, and changes in clinical and histologic characteristics were studied. Our study assessed the clinicopathologic features of patients with seronegative AIH when compared with patients with seropositive AIH before initiation of immunosuppression. Hence, our findings add a new dimension to the understanding of AIH, specifically as regards to the clinical and histologic characteristics of seronegative AIH. Ten percent of patients within both seropositive and seronegative groups showed suboptimal response to initial immunosuppressive therapy. This finding is similar to a recent study by Montano-Loza et al, 26 who reported a failure rate of 7% (14 of 214) in patients with type 1 AIH treated with conventional corticosteroid treatment. 27 Patients who failed therapy in this study 27 were younger, had higher serum bilirubin levels, and presented more acutely. In our study, we did not find any significant difference between responders and nonresponders. When patients were followed for 1 year after the institution of immunosuppression, there was no significant difference between the aminotransferase levels at 1, 6, and 12 months after treatment between seronegative and seropositive groups. Limitations of our study include smaller sample size when compared with earlier studies. 8,26 Because of the retrospective nature of our study, certain information like IgG level was not available on all patients. We have not done subgroup analysis on the basis of autoantibody titers. However, this study compares the clinical and histologic characteristics of seronegative and seropositive AIH before initiation of therapy with AIH, adding a new dimension to our understanding of AIH. In conclusion, our data suggest that antibody status should not be used to predict AIH activity or outcome. Patients who meet the diagnosis of AIH on the basis of the revised scoring system of IAIHG should be treated with immunosuppression, irrespective of antibody status. References 1. Johnson PJ, McFarlane IG. Meeting report: International Autoimmune Hepatitis Group. Hepatology 1993;18:998 1005. 2. Alvarez F, Berg PA, Bianchi FB, et al. International Autoimmune Hepatitis Group report: review of criteria for diagnosis of autoimmune hepatitis. J Hepatol 1999;31:929 938. 3. Czaja AJ, Carpenter HA. Validation of scoring system for the diagnosis of autoimmune hepatitis. Dig Dis Sci 1996;41:305 314. 4. Carpenter HA, Czaja AJ. The role of histologic evaluation in the diagnosis and management of autoimmune hepatitis and its variants. Clin Liver Dis 2002;6:685 705. 5. Czaja AJ, Norman GL. Autoantibodies in the diagnosis and management of liver disease. J Clin Gastroenterol 2003;37:315 329. 6. Czaja AJ, Carpenter HA, Santrach PJ, et al. The nature and prognosis of severe cryptogenic chronic active hepatitis. Gastroenterology 1993;104:1755 1761. 7. Czaja AJ. The variant forms of autoimmune hepatitis. Ann Intern Med 1996;125:588 598. 8. Czaja AJ. Behavior and significance of autoantibodies in type 1 autoimmune hepatitis. J Hepatol 1999;30:394 401. 9. Krawitt EL. Autoimmune hepatitis. N Engl J Med 2006;354:54 66. 10. Feld JJ, Dinh H, Arenovich T, et al. Autoimmune hepatitis: effect of symptoms and cirrhosis on natural history and outcome. Hepatology 2005;42:53 62. 11. Vergani D, Alvarez F, Bianchi FB, et al. Liver autoimmune serology: a consensus statement from the committee for autoimmune serology of the International Autoimmune Hepatitis Group. J Hepatol 2004;41:677 683. 12. Tan EM, Smolen JS, McDougal JS, et al. A critical evaluation of enzyme immunoassay kits for detection of antinuclear autoantibodies of defined specificities: II potential for quantitation of antibody content. J Rheumatol 2002;29:68 74. 13. Fritzler MJ, Wiik A, Tan EM, et al. A critical evaluation of enzyme immunoassay kits for detection of antinuclear autoantibodies of defined specificities: III comparative performance characteristics of academic and manufacturers laboratories. J Rheumatol 2003;30:2374 2381. 14. Fritzler MJ, Wiik A, Fritzler ML, et al. The use and abuse of commercial kits used to detect autoantibodies. Arthritis Res Ther 2003;5:192 201. 15. Rondeel JM, van Gelder W, van der Leeden H, et al. Different strategies in the laboratory diagnosis of autoimmune disease: immunofluorescence, enzyme-linked immunosorbent assay or both? Ann Clin Biochem 1999;36:189 195. 16. Bogdanos DP, Invernizzi P, Mackay R, et al. Autoimmune liver serology: current diagnostic and clinical challenges. World J Gastroenterol 2008;14:3374 3387. 17. Goodman ZD. Grading and staging for inflammation and fibrosis in chronic liver diseases. J Hepatol 2007;47:598 607.

January 2009 SEROLOGY IN AUTOIMMUNE HEPATITIS 103 18. Knodell RG, Ishak KG, Black WC, et al. Formulation and application of a numerical scoring system for assessing histological activity in asymptomatic chronic active hepatitis. Hepatology 1981;1:431 435. 19. Ishak K, Baptista A, Bianchi L, et al. Histological grading and staging of chronic hepatitis. J Hepatol 1995;22:696 699. 20. Johnson PJ, McFarlene IG, McFarlene BM, et al Auto-immune features in patients with idiopathic chronic active hepatitis who are seronegative for convenional auto-antibodies. J Gastroenterol Hepatol 1990:5:244 251. 21. Czaja AJ, Hay JE, Rakela J. Clinical features and prognostic implications of severe corticosteroid-treated cryptogenic chronic active hepatitis. Mayo Clin Proc 1990;65:23 30. 22. Czaja AJ, Carpenter HA, Santrach PJ, et al. The nature and prognosis of severe cryptogenic chronic active hepatitis. Gastroenterology 1993;104:1755 1761. 23. Kaymakoglu S, Cakaloglu Y, Demir K, et al. Is severe cryptogenic chronic hepatitis similar to autoimmune hepatitis. J Hepatol 1998;28:78 83. 24. Thiele D, DuCharme L, Cunningham MR, et al. Steroid therapy of chronic hepatitis: characteristics associated with response in anti-hepatitis C virus-positive and negative patients. Am J Gastroenterol 1996;91:300 308. 25. Gassert DJ, Garcia H, Tanaka K, et al. Corticosteroid-responsive cryptogenic chronic hepatitis: evidence of seronegative autoimmune hepatitis. Dig Dis Sci 2007;52:2433 2437. 26. Montano-Loza AJ, Carpenter HA, Czaja AJ. Features associated with treatment failure in type 1 autoimmune hepatitis and predictive value of the model of end-stage live disease. Hepatology 2007;46:1138 1145. 27. Czaja AJ, Freese DK. Diagnosis and treatment of autoimmune hepatitis. Hepatology 2002;36:479 497. Address requests for reprints to: Steven K. Herrine, MD, Division of Gastroenterology and Hepatology, Thomas Jefferson University, 132 S 10th Street, #450, Philadelphia, Pennsylvania 19107. e-mail: steven. herrine@jefferson.edu; fax: (215)503-2146. The authors disclose no conflicts.