A 55-year-old Caucasian male developed a well characterized autoimmune hepatitis after completing treatment with pegylated interferon and ribavirin for recurrent hepatitis C. We hypothesize that pegylated interferon triggered a severe form of immune-mediated hepatitis.
Recurrent hepatitis C viral infection (HCV) is universal post–liver transplantation (LT) and may result in considerable graft dysfunction and mortality. The most effective current therapy for recurrent HCV infection post-LT is pegylated interferon (peg-IFN) in combination with ribavirin (RBV).
Interferon (IFN) and RBV exert a variety of immunomodulatory effects. Antiviral therapy may confer a potential increased risk of inducing or unmasking underlying autoimmune diseases. Prolonged exposure to exogenous IFN has been implicated in the induction of organ-specific autoimmune diseases such as autoimmune thyroiditis and rheumatoid arthritis.1, 2 Furthermore, IFN has been demonstrated to trigger and exacerbate autoimmune hepatitis when used to treat chronic HCV infection in the nontransplant setting.3–6 Lastly, HCV itself is independently associated with a high incidence of autoimmune diseases, such as thyroiditis.7
We describe a patient who developed an aggressive plasma-cell-predominant hepatitis post-LT for HCV-related cirrhosis and hepatocellular carcinoma. This condition occurred in the absence of a prior history of autoimmune hepatitis, while the patient was being treated with peg-IFN and had undetectable HCV RNA. We hypothesize that peg-IFN was the likely trigger for a severe form of immune mediated hepatitis.
A 55-year-old Caucasian male underwent LT for HCV-related cirrhosis and early-stage hepatocellular carcinoma in September 2002. He had not received IFN antiviral therapy for HCV prior to LT and had no manifestations of autoimmunity. His immediate postoperative course was unremarkable. Induction immunosuppression consisted of daclizumab and prednisolone with tacrolimus titrated to maintain trough levels of 8–12 μg/L during the first 2 months, then 6–10 μg/L thereafter. Corticosteroids were rapidly tapered and discontinued by 1 month post-LT with the patient maintained on tacrolimus monotherapy.
Five months later, a rise in liver enzymes was noted: alanine aminotransferase (ALT), 931 IU/mL (normal, <40 IU/mL); aspartate aminotransferase, 495 IU/mL (normal, <35 IU/mL); alkaline phosphatase, 212 IU/mL (normal, <120 IU/mL); bilirubin, 0.7 mg/dL (normal, <1.0 mg/dL). In addition, the tacrolimus level was 6.8 μg/dL, HCV genotype was 1, and quantitative HCV RNA viral load was 476,000 copies/mL (Roche Cobas Amplicor, Roche, Rotkreuz, Switerland). Liver biopsy revealed panlobular hepatitis and a lymphoid follicle in 1 of the portal tracts, consistent with severe recurrent hepatitis C infection (Fig. 1).
Antiviral therapy was commenced with an escalating regimen of peg-IFN (Pegasys, Roche, Nutley, NJ) and RBV. This initially consisted of 90 μg/week peg-IFN and 400 mg RBV per day for the first month, increasing in monthly increments to 180 μg/week peg-IFN and 800 mg RBV per day by the third month. Antiviral therapy was well tolerated with minimal clinical side effects, and HCV RNA became undetectable by polymerase chain reaction at week 12 of peg-IFN and RBV therapy. There was concomitant improvement in liver enzymes with normalization of ALT during this period.
Following 9 months of anti-HCV therapy, another elevation in liver enzymes was noted. On this occasion, liver enzymes were as follows: ALT, 303 IU/mL; aspartate aminotransferase, 186 IU/mL; alkaline phosphatase, 214 IU/mL; bilirubin, 0.8 mg/dL; tacrolimus level, 5.8mg/dL. At this time the patient was receiving 180 μg peg-IFN weekly and 800 mg RBV daily. HCV RNA remained undetectable by polymerase chain reaction, implying an alternative etiology for graft dysfunction. A repeat liver biopsy revealed an unexpected histological appearance consisting of a dense infiltration of plasma cells, bridging necrosis, and areas of parenchymal collapse in keeping with a diagnosis of immune-mediated hepatitis with severe activity. (Fig. 2 and 3). Portal fibrosis with fibrous septa was present. Typical histological features of acute cellular rejection were absent. Tests for autoantibodies, including antinuclear antibodies, anti-smooth-muscle antibodies, anti-liver-kidney microsomal antibodies, and antimitochondrial antibodies, were negative. However, total immunoglobulin (Ig) G was elevated at 45g/L (normal range, <17g/L), while IgM and IgA levels were normal.
A provisional diagnosis of immune-mediated hepatitis was made and therapy with 50 mg of azathioprine per day began. Corticosteroids were not administered due to clinical concerns of a potential relapse of HCV. Serum aminotransferases normalized over the following 4–6 weeks in response to azathioprine. Antiviral therapy (peg-IFN and RBV) was continued at this stage, given the virological response, the normalization of liver enzymes, and the wishes of the patient.
Four months after addition of azathioprine, the patient developed severe symptomatic anemia (hemoglobin, 7.6 g/dL) and required blood transfusion and erythropoietin therapy. This was attributed to the combined effects of RBV-induced hemolysis and bone marrow suppression due to peg-IFN and azathioprine. To optimize antiviral response azathioprine therapy was discontinued rather than antiviral dose reduction.
Six weeks later, another flare in serum aminotransferases developed: ALT, 475 IU/mL; aspartate aminotransferase, 287 IU/mL; alkaline phosphatase, 240 IU/mL; bilirubin, 1.1 mg/dL; and tacrolimus level, 6.8 mg/dL. Liver biopsy again showed features of severe immune-mediated hepatitis with bridging necrosis and collapse and significant necroinflammatory activity, similar to those seen in the prior biopsy. Autoimmune markers were repeated and smooth muscle antibody (2+/3+) had subsequently become positive. The other autoimmune markers remained negative. IgG was raised again, at 40 g/dL.
Once again, normalization of aminotransferases occurred following reintroduction of azathioprine. Hemoglobin concentrations were supported with erythropoietin therapy, allowing continuation of azathioprine and RBV. Figure 4 summarizes the time course of ALT elevation and interventions with azathioprine therapy. Anti-HCV combination therapy was continued for a total of 18 months with the goal of achieving permanent viral eradication. HCV viral load remained undetectable 6 months after discontinuation of peg-IFN and RBV combination therapy consistent with a sustained virological response. Protocol liver biopsy performed 12 months after completion of antiviral therapy revealed no significant histological inflammation (Fig. 5).
We report a case of immune-mediated hepatitis occurring in a patient treated with peg-IFN and RBV for recurrent HCV infection post–liver transplantation. HCV RNA was undetectable at the time of aminotransferase elevation, excluding recurrent HCV as the cause of graft dysfunction. The pattern of aminotransferase elevation, raised IgG concentrations, positive smooth-muscle antibody, histological features, and response to therapy used for autoimmune hepatitis with azathioprine are all consistent with a diagnosis of post-LT immune-mediated hepatitis.
In this case, treatment with azathioprine resulted in a prompt and dramatic response with normalization of biochemical abnormalities. Relapse occurred soon after discontinuation of azathioprine; however, it responded once again with reintroduction, further supporting this diagnosis. Applying the International Diagnostic Criteria for the Diagnosis of Autoimmune Hepatitis we would derive a score of 15 before treatment and 18 after treatment; generating an interpretation of “definite” autoimmune hepatitis.8
It is noteworthy that a sustained virological response was achieved by continuation of antiviral therapy despite the development of immune-mediated hepatitis.
The most common form of graft dysfunction post–liver transplantation is cellular rejection,9 while histological recurrent hepatitis C posttransplant remains a major issue. A higher rate of acute cellular rejection was recently demonstrated in 2 studies following IFN therapy for recurrent HCV infection post–liver transplantation.10, 11 However, changes of acute cellular rejection or chronic rejection were not demonstrated on liver biopsy. Instead, both biopsies revealed high necroinflammatory activity with abundant numbers of plasma cells, bridging necrosis, and parenchymal collapse—all features classically associated with autoimmune hepatitis12
HCV itself has a strong association with autoimmune disease independent of exposure to IFN.7 Thyroid dysfunction has been described as an extrahepatic manifestation of chronic HCV infection, with up to 10% of patients having thyroid autoantibodies.13, 14 The prevalence of antinuclear antibodies in patients with chronic HCV was 14% prior to IFN treatment and then increased to 35% following treatment in 1 study.15 This evidence suggests a synergistic effect of HCV and IFN in triggering autoimmunity.
Molecular mimicry between host autoantigens and HCV proteins is another possible explanation for the loss of tolerance and occurrence of autoimmune responses to self proteins.16 HCV has been implicated as a mimicry candidate in type 2 autoimmune hepatitis, because infection is associated with autoantibodies to CYP4502D6.17 In another study, Kammer et al. identified 2 sequences of CYP450 isoforms that showed homology with a 10-amino acid sequence of the HCV core and then used these sequences to induce autoreactive CD8+ cytotoxic T cells.18 Thus, IFN α therapy in the setting of chronic HCV infection may either unmask pre-existing autoimmune conditions or trigger autoimmunity.
IFN exerts a variety of effects on the immune system including modulation of immunoglobulin production and inhibition of T-suppressor-cell function.19 IFN induces natural killer cells and cytotoxic T lymphocytes, stimulates major histocompatability complex class 1 expression, and polarizes the adaptive immune response to Th1 (T helper cells).20 The most common IFN-related autoimmune phenomena are organ-specific.21 A predominance of Th1 activity has been demonstrated in organ-specific autoimmune states; it is thus postulated that this imbalance toward the Th1-mediated response induced by IFN is an important factor in IFN-induced autoimmunity.22
Our patient developed immune-mediated hepatitis causing a severe form of late graft dysfunction that occurred in the context of peg-IFN therapy. With increasing use of the pegylated forms of IFN for posttranplant HCV infection, this predominantly histological entity may be underrecognized. In the setting of recurrent HCV, immune-mediated hepatitis “induced” by antiviral therapy may be responsive to an increase in immunosuppression. Liver biopsy remains pivotal in the diagnosis, and differentiation of this condition as a correct diagnosis has important therapeutic implications.