Infliximab-induced autoimmune hepatitis
- 1 Internal Medicine, Walter Reed National Military Medical Center, Bethesda, Maryland, USA
- 2 Pathology, Walter Reed National Military Medical Center, Bethesda, Maryland, USA
- 3 Internal Medicine, Naval Medical Center San Diego, San Diego, California, USA
- 4 Gastroenterology, Naval Medical Center Portsmouth, Portsmouth, Virginia, USA
- Correspondence to Dr Alexander Jenkins; arj3030@gmail.com
Abstract
Autoimmune hepatitis (AIH) is an inflammatory disorder of the liver with a wide spectrum of disease presentation, from asymptomatic elevations in liver-associated enzymes to acute liver failure. AIH is classically associated with elevated immunoglobulins and autoantibodies, although approximately 20% of patients with features of AIH lack circulating antibodies. Recently, tumour necrosis factor alpha inhibitors have been implicated in several cases of drug-induced AIH which impact treatment regimens for patients with inflammatory bowel disease (IBD). We present a case of infliximab-induced seronegative AIH responding to budesonide therapy with successful alteration of IBD treatment regimen to vedolizumab.
Background
Determining the aetiology of elevated transaminases in patients with inflammatory bowel disease (IBD) can be challenging given the frequent overlap of IBD and hepatobiliary diseases such as primary sclerosing cholangitis (PSC) and autoimmune hepatitis (AIH). This is further complicated by the fact that several immunosuppressive medications frequently prescribed for patients with IBD can cause both direct and indirect liver damage via drug-induced hepatotoxicity as well as less well understood immune-mediated mechanisms. Anti-tumour necrosis factor alpha (TNFα) drugs are one of the more commonly implicated medications in immune-mediated liver injury and recent literature suggests a substantial association with drug-induced AIH and this particular class of drugs.
Of further interest is the topic of alternative therapeutic options for patients with IBD who have a significant liver injury while on an anti-TNFα drug, as infliximab (IFX) is the preferred therapy for severe IBD, and deciding whether or not to discontinue a drug that has provided disease control can be a daunting task. The following case demonstrates an atypical presentation of suspected IFX-induced AIH as well as insight on alternative agents for continued management of moderate to severe IBD when anti-TNFα medications are no longer an option.
Case presentation
Our patient is a 31-year-old man with a medical history significant for severe, pancolonic ulcerative colitis (UC) diagnosed the previous year in clinical remission on IFX. Prior to this, he was treated with mesalamine, however, he suffered from a bout of pancreatitis shortly after initiation, so he was transitioned to a 5-week course of prednisone. The patient had difficulty tolerating prednisone due to insomnia and mood swings, so he was transitioned to budesonide which was subsequently tapered while adalimumab was initiated. While on adalimumab, he continued to have multiple watery bowel movements daily and subsequently developed Clostridium difficile colitis for which he was treated with oral vancomycin. At this point, another 5-week prednisone taper was initiated and adalimumab was discontinued in favour of IFX monotherapy. Though the patient had previously been prescribed another anti-TNFα with inadequate response, IFX was chosen due to patient preference (no need for self-injection or additional steroids while waiting for another biologic to take effect) as well as lack of evidence that the patient had ever reached stable therapeutic serum levels of adalimumab. All liver-associated enzymes (LAEs) were normal prior to initiation of therapy but progressively increased starting around the third month of active treatment with IFX (table 1). Additionally, the patient noted increased fatigue, joint pain and abdominal pain despite normal thyroid function tests and stable haemoglobin.
LAE trends over time
| AST (U/L) | ALT (U/L) | Significant events | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ALP, Alkaline phosphatase ; ALT, Alanine aminotransferase ; AST, Aspartate aminotransferase ; IFX, infliximab; LAE, liver-associated enzyme. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 31 January 2019 | 20 | 32 | Baseline labs prior to IFX initiation | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 02 April 2019 | 24 | 48 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 16 September 2019 | 65 | 130 | Last dose of IFX given | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 01 October 2019 | 78 | 181 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 01 November 2019 | 57 | 122 | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 05 December 2019 | 170 | 424 | Vedolizumab initiated | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 10 December 2019 | 168 | 440 | Budesonide and azathioprine initiated. ALP now elevated to 137 for first time. | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 26 December 2019 | 46 | 110 | Response to high dose oral budesonide | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 19 February 2020 | 30 | 34 | LAEs normalise with ongoing course of budesonide and azathioprine | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Investigations
Serologic evaluation including serum immunoglobulins, smooth muscle antibody (SMA), antimitochondrial antibody, liver kidney microsomal antibody, antinuclear antibody (ANA), alpha-1 antitrypsin and tissue transglutaminase antibody was unremarkable. IgA, IgG and IgM were within normal limits. Chronic and acute viral hepatitis serologies were also negative as was evaluation for haemochromatosis and Wilson disease. Differential diagnosis at this time included drug-induced liver injury (DILI) and IFX-induced AIH.
Due to the patient’s persistently increasing LAEs and no obvious cause, transcutaneous liver biopsy was obtained. Biopsy demonstrated moderate portal inflammation and focal interface hepatitis composed predominantly of lymphocytes with scattered plasma cells and eosinophils (figures 1 and 2). Apoptotic hepatocytes and zone three inflammation were also identified. These findings were histologically consistent with early AIH or DILI. PSC was initially also considered, as overlap with AIH can occur, however, the decidedly hepatocellular pattern of LAEs and no biopsy findings to suggest bile duct pathology makes this a highly unlikely aetiology.
H&E stain; 100× view of liver biopsy demonstrating mild to moderate portal and parenchymal inflammatory activity with focal interface hepatitis (arrow).
H&E; 400x view of the same biopsy featuring inflammatory cells, predominantly lymphocytes with scattered plasma cells (arrow), eosinophils (arrowhead) and macrophages.
Treatment
IFX was discontinued due to concern for hepatotoxicity and the patient was started on the humanised anti α4β7 integrin monoclonal antibody, vedolizumab, for continued management of his UC. His liver injury continued to worsen, with alanine aminotransferase and aspartate aminotransferase peaking at 440 U/L and 168 U/L as well as new elevation in alkaline phosphatase to 137 U/L nearly 3 months after his last IFX infusion. At this point, drug-induced seronegative AIH was deemed the most likely diagnosis given the lack of improvement with IFX cessation and liver biopsy results. 9 mg/day oral budesonide was initiated instead of prednisone due to patient preference and subsequently down-titrated to 6 mg and then 3 mg per day over the ensuing 3 months with the addition of azathioprine.
Outcome and follow-up
After transition to vedolizumab and initiation of steroids and azathioprine, the patient’s symptoms and LAEs improved rapidly. The patient’s LAEs normalised within 2 months of starting budesonide and over the ensuing 3 months, steroids were completely tapered off in favour of 100 mg of azathioprine per day. The patient remains asymptomatic with normal LAEs on vedolizumab infusions every 8 weeks with a plan in place to taper him off of azathioprine and continue to monitor LAEs monthly for the foreseeable future.
Discussion
Since obtaining U.S. Food and Drug Administration approval for use in the treatment of IBD in 1998, IFX has changed the way in which Crohn’s disease and UC are managed.1 In 2004, IFX manufacturers issued a warning that the drug could cause acute liver failure, jaundice, hepatitis, cholestasis as well as rare cases of AIH.1 The specific pathophysiology for how TNFα blocking drugs (anti-TNFα) cause liver injury is not completely understood, however, it has been proposed that hepatotoxicity can occur in several ways including direct chemical toxicity, generation of autoantibodies, cholestasis and reactivation of viral hepatitis B.2
It is often difficult to differentiate drug-induced AIH and drug-induced AIH-like liver injury from one another. While the effects of a DILI can persist for several months following withdrawal of the offending agent, serologic evidence of liver injury typically improves within 1 month of drug discontinuation and virtually always within 3 months.3 Additionally, there seems to be a shorter latency period after initiation of IFX for DILI compared with anti-TNFα induced AIH.4 This was not the case with this patient, as his LAEs continued to rise nearly 3 months after his last dose of IFX. In addition, liver biopsy findings in this case were compatible with AIH, although not pathognomonic. Typical AIH features significant interface hepatitis and portal fibrosis on biopsy,5 however, drug-induced AIH can feature less severe structural damage with only mild to moderate interface hepatitis.3 6 The rapid response to immune suppressive therapy with budesonide followed by the immunomodulator azathioprine augments the diagnosis of drug-induced AIH.
Seronegative AIH is uncommon, especially in men, however, lack of autoantibodies does not rule out AIH, as ANA and SMA may be expressed sporadically throughout the disease course of AIH or may be seen only in more advanced disease.7 Biopsy in this patient suggested early AIH without fibrosis. Based on the revised original score for AIH, our patient received a score of 9 points, corresponding to ‘possible AIH’, however, a score of 10 in this scoring system is designated as ‘probable AIH’, and part of what makes this such a unique case is the lack of autoantibodies. This case reflects that drug-induced AIH can still occur without positive autoantibodies and normal serum immunoglobulins and should still be considered as a potential diagnosis in a patient with undifferentiated liver injury on IFX.
This appears to be the first documented case in which a patient with IFX-induced AIH was transitioned to vedolizumab. The patient has had no further evidence of elevated LAEs and his UC symptoms have been well controlled for the last 5 months, well past tapering of steroids which could be suppressing his IBD symptoms. This suggests an alternative approach to therapy in patients who are unable to continue anti-TNFα therapy as a result of hepatotoxicity, regardless of aetiology. There are cases in which patients who had either a DILI or drug-induced AIH on one anti-TNFα drug were ultimately able to tolerate another medication within the same class,1 6 8 however, the use of vedolizumab offers a potentially safer treatment option as the mechanism of liver injury resulting from anti-TNFα therapy remains unclear.
It has been suggested that IFX-induced AIH is precipitated when an unknown drug metabolite becomes antigenic when bound to hepatocellular proteins, generating autoantibodies that in turn target molecularly similar hepatic antigens.9 While this model remains unproven, it offers some explanation as to why IFX can both induce and treat AIH as well as why there are multiple reported cases of patients tolerating adalimumab following liver injury on IFX.
Ultimately, definitive diagnosis of IFX-induced AIH is challenging, as there is considerable overlap with drug-induced AIH-like liver injury, however, we believe that the details of this case provide adequate evidence to support this diagnosis. This case suggests that vedolizumab offers another safe and effective therapeutic option for patients with severe IBD who develop IFX-induced AIH, which is valuable in the setting of a disease with pathophysiology that is poorly understood.
Patient’s perspective
I first experienced symptoms related to my ulcerative colitis (UC) during the Spring of 2018. My symptoms included diarrhoea, loose watery stools (sometimes 15 or more a day), unpredictable bowel movements, painful gas, fatigue and bloody stools. I experienced a lot of difficulties convincing my healthcare providers that my symptoms were something beyond gastroenteritis. It took many months and four different healthcare providers to finally obtain a referral to a gastroenterologist. I was eventually diagnosed with UC with pancolonic involvement during May 2018 and I was prescribed mesalamine and prednisone. While on prednisone, I became extremely irritable, suffered from insomnia, restlessness and significant weight gain. Although prednisone effectively controlled my UC symptoms, I felt that the side effects were just as bad as my original symptoms.
Unfortunately, after only 2 months treatment with mesalamine and no change to my original symptoms, I developed low back pain, nausea, fatigue, abdominal pain and fever. I was admitted to the hospital in August 2018 where I was diagnosed with pancreatitis. While admitted, it was determined that my condition was the caused by mesalamine, so my treatment was changed to adalimumab and budesonide. Although I had less side effects while on budesonide, it did not treat my original symptoms as effectively as prednisone.
After several months on adalimumab, I experienced no improvement in symptoms and continued to experience fatigue, bloody stools and unpredictable bowel movements. I was diagnosed with C. diff in November 2018 and adalimumab was stopped at that time. In January 2019, I was started on infliximab. My symptoms significantly subsided and I started to experience normal bowel movements for the first time since the Spring of 2018. I was doing much better until November 2019 when I began to experience significant joint pain and pain in my upper right abdomen. I was told by my doctors that infliximab had triggered autoimmune hepatitis so it was then decided to change my treatment once more to vedolizumab in November 2019.
Because of my long history of difficulties with treatment for my UC, I was extremely reluctant to make the transition to vedolizumab. My journey with UC and autoimmune hepatitis has been a long and difficult one. The healthcare providers I originally visited tended to downplay my symptoms and attributed them to more common diagnoses. As a result, a lot of time elapsed until I was finally correctly diagnosed and offered treatment that was right for me. Overall, my latest treatment on vedolizumab has been a positive experience, my symptoms remain under control and I have not noticed any major side effects.
Learning points
-
Autoimmune hepatitis (AIH) features a wide range of presentations and can still occur even if a patient has no detectable autoantibodies.
-
Recognition of drug-induced seronegative AIH can be challenging and biopsy results are often necessary to make the diagnosis.
-
Vedolizumab provides a reasonable alternative therapy to patients with inflammatory bowel disease who develop AIH while on an anti-TNFα drug.
Acknowledgments
Dr Jeannie Muir, MD, FASCP, MT (ASCP), Walter Reed National Military Medical Center. Dr Manish Singla, MD, FACG, Walter Reed National Military Medical Center.
Footnotes
-
Contributors AJ, internal medicine resident, interviewed the patient in this case, collected data and wrote most of the manuscript. Approved final manuscript prior to submission. AA, pathology resident, helped interpret biopsy findings, formatted the two histology images and provided captions. Approved final manuscript prior to submission. KH, internal medicine resident, helped write and edit the manuscript. Approved final manuscript prior to submission. BS, staff gastroenterologist, was heavily involved in interpreting the relevant data of this case, directly involved in patient care and provided edits to the manuscript. Approved final manuscript prior to submission. Dr Dawn Torres, senior staff gastroenterologist, was heavily involved in interpreting the relevant data of this case, directly involved in patient care and provided edits to the manuscript. Approved final manuscript prior to submission.
-
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
-
Competing interests None declared.
-
Provenance and peer review Not commissioned; externally peer reviewed.
- © BMJ Publishing Group Limited 2021. No commercial re-use. See rights and permissions. Published by BMJ.
References
Use of this content is subject to our disclaimer