Use of a xanthine oxidase inhibitor in autoimmune hepatitis

Abstract

A 62-year-old woman with type 1 autoimmune hepatitis (AIH) failed to sustain remission when steroids were withdrawn from a regimen of steroids and azathioprine (AZA). Thiopurine metabolites revealed elevated 6-MMP (6-methyl mercaptopurine) and low 6-TGN (6-thioguanine nucleotide) consistent with AZA-induced hepatotoxicity. Introducing the xanthine oxidase inhibitor allopurinol led to rapid normalization of alanine aminotransferase (ALT) and discontinuation of steroids. AIH, autoimmune hepatitis; ALT, alanine transaminase; AMA, antimitochondrial antibody; AZA, azathioprine; LKM1, liver-kidney microsome 1; 6-MMP, 6-methyl mercaptopurine; pANCA, perinuclear antineutrophil cytoplasmic antibody; RBC, red blood cell; SMA, smooth muscle antibody; TPMT, thiopurine methyltransferase; ULN, upper limit of normal. A 62-year-old woman was referred for evaluation of deranged liver function tests; alanine transferase ALT 150 IU/L (upper limit of normal [ULN] <30 IU/L) and weakly positive antinuclear antibody titer. Antimitochondrial antibody (AMA), smooth muscle antibody (SMA), perinuclear antineutrophil cytoplasmic antibody (pANCA), and liver-kidney microsome 1 (LKM1) were negative. Her immunoglobulin G (IgG) level was 25.8 g/L (range 6-16). She drank no alcohol. Negative serology excluded viral hepatitis and there was no hepatotoxic medication in her drug history. Liver biopsy revealed interface hepatitis with severe lymphoplasmacytic infiltration and mild fibrosis. The International Autoimmune Hepatitis Group (IAIHG) revised score1 was 17, suggesting a definite diagnosis of AIH (human leukocyte antigen [HLA]-DR genotype not tested). Forty milligrams prednisolone taper was commenced with a remission goal of ALT normalization as per American Association for the Study of Liver Diseases (AASLD) guidelines.2 Thiopurine methyltransferase (TPMT) level was normal (53 nmol/g, range 35-79). Subsequently she started azathioprine (AZA) 75 mg daily (1 mg/kg). Her ALT and IgG normalized 6 weeks later. Following prednisolone reduction to 10 mg, her ALT increased to 147 IU/L despite AZA compliance. Forty milligrams prednisolone was reinstituted. On tapering to 10 mg she developed persistently fluctuating ALT (40-80 IU/L) despite increased AZA to 150 mg (2 mg/kg). A rise of ALT to 133 IU/L required 40 mg prednisolone again. Repeat liver biopsy revealed steatosis but much reduced interface hepatitis. Thiopurine metabolite levels (Lennard method) revealed a 6-thioguanine (6-TGN-the active moiety) level of 100 pmol/8 × 108 red blood cells (RBC) (normal range 250-450) and a 6-methyl mercaptopurine (6-MMP) level of 5800 pmol/8 × 108 RBC, consistent with hypermethylation and preferential shunting to 6-MMP. Allopurinol 100 mg once daily was added and her AZA dose reduced by 75%. Subsequently, her ALT normalized within 4 weeks (Fig. 1). This combination corrected her metabolite levels (6-TGN 202 pmol/8 × 108 RBC and 6-MMP 196 pmol/8 × 108 RBC). Prednisolone was withdrawn after 3 months; her ALT remains 10-15 IU/L 12 months later. ALT fluctuation over time. AZA is essential in managing AIH, often obviating the need for long-term prednisolone. Between 3%-13% of inflammatory bowel disease (IBD) and AIH patients will develop AZA-induced hepatotoxicity,3-5 which may be difficult to distinguish from AIH nonresponse or relapse without liver biopsy. In this setting, measuring thiopurine metabolites can provide diagnostic guidance. The finding of increased 6-MMP >5700 pmol/8 × 108 RBC with low or normal 6-TGN (“shunting”) is associated with hepatotoxicity and other side effects including nausea, anorexia, and influenza-like symptoms in IBD patients.5 Our patient's persistently elevated ALT was consistent with AZA-induced hepatotoxicity (raised ALT secondary to high 6-MMP.) Repeat liver biopsy revealed prednisolone-induced steatosis with only mildly active AIH. High 6-MMP levels do not always cause hepatotoxicity, however. Allopurinol acts through inhibition of xanthine oxidase, producing preferential AZA breakdown by the TPMT enzymatic pathway resulting in higher 6-TGN and lower 6-MMP (Fig. 2), although the exact mechanism is not fully understood. The combination requires AZA dose reduction to prevent excess 6-TGN production. AZA metabolism and role of allopurinol. This is the first reported adult case of allopurinol co-therapy in AIH for the management of AZA-induced hepatotoxicity. This strategy allowed our patient to discontinue prednisolone and avoid alternative immunosuppression. A study into the utility of measuring thiopurine metabolites identified AZA-induced hepatotoxicity, although the investigators found no role for measuring levels to monitor treatment response.3 In conclusion, in AIH patients who develop persistently significant ALT elevation after steroid discontinuation from a prednisolone-AZA regimen we recommend measuring thiopurine metabolites to distinguish between AZA-induced hepatotoxicity, nonadherence, or relapse and avoid repeat liver biopsy. AZA-induced hepatotoxicity should be suspected in patients with elevated 6-MMP (regardless of 6-TGN levels). The addition of allopurinol with appropriate AZA dose reduction may correct AZA-induced hepatoxicity and induce remission.