Immune checkpoint inhibitors (ICPIs) such as ipilimumab, a CTLA-4 inhibitor, and nivolumab, a PD-1 inhibitor, are more widely used to treat various types of cancers, but they can also be associated with immune-related adverse reactions (irAEs) including enterocolitis, hepatitis, or skin rash. Careful attribution of irAEs is important as it may impact the patient’s subsequent cancer therapy. The development of liver injury secondary to ICPI administration is infrequent but well-known and reported. The use of infliximab as an agent for steroid-refractory colitis secondary to ICPI has been reported and potentially effective. However, infliximab is also associated with hepatotoxicity. We describe a case with a complicated history that prompted careful analysis of the differential diagnoses for acute liver injury presented with new elevation in liver enzymes.

As ICPI therapies are being more frequently employed to treat many advanced cancers, hepatotoxicity is being recognized as one of the irAEs [1–13]. New elevation in liver transaminases during the ICPI treatment should prompt investigation of potential drug-induced liver injury. However, other co-existing conditions and/or infections that require additional medical treatments could also be associated with hepatotoxicity, the differential diagnosis toward the etiology may be confounding. This diagnostic challenge may be overcome by careful review and analysis of the history. Points of consideration include evaluation for acute viral hepatitides, autoimmune disease, alcohol-associated hepatitis, immune-related adverse event on the liver (especially given prior irAE associated with skin rash and colitis), and potentially infliximab-associated hepatitis.

Ipilimumab and nivolumab have been independently associated with the liver irAE [3–13]. The incidence of hepatoxicity for ipilimumab alone, for instance, was estimated to be 1.6–3.9% (ascertained by ALT elevations) in any grade of severity [7, 11]. Grade 3 or 4 hepatotoxicity has an estimated 3–7% incidence [3]. Recent data suggests an incidence of hepatotoxicity of as high as 11%, with a reported fatality rate of 0.2% [13]. Liver injury as a result of exposure to ipilimumab usually occur after 2–4 cycles or a median time of 3–9 weeks after initiation of the drug, with potential to cause elevations in liver tests to reflect hepatocellular or mixed cholestatic injury [3–5]. Nivolumab may also cause hepatotoxicity, with onset of injury 2–6 cycles or 1–3 month after initiation of treatment [6]. The incidence of ALT elevations from nivolumab alone is estimated to be up to 3.8% in any grade of severity, based on clinical trial data [7]. The pattern of liver injury from nivolumab resembles that from ipilimumab. Varied findings may result on abdominal imaging of the liver (CT or ultrasound) from normal findings to mild hepatomegaly, periportal edema, and low-attenuation of the liver parenchyma [14]. Combination therapy utilizing both a CTLA-4 inhibitor and PD-1 inhibitor have been applied towards advanced melanoma treatment and is also being explored in other cancers at the expense of higher risk for irAE [15, 16]. Recent data estimates hepatotoxicity from combined therapy to be about 13% with median time to onset of about 2.1 months [13]. Prior clinical trial data had suggested the overall incidence of ALT elevations in combined therapy to be as high as 17.6% [7]. Hepatocellular and/or cholestatic liver injury patterns may be seen [17]. There is no compelling relationship between the autoimmune antibodies and autoimmune liver disease [5, 17, 18].

Histopathological findings of ICPI-related hepatitis had been reported previously. Such features included: lymphohistiocytic inflammation with interface hepatitis, eosinophilic infiltration or acidophil bodies, granulomas, perisinusoidal fibrosis, periportal fibrosis, and/or biliary ductal injury [9, 14, 17, 33]. While there is no known pathognomonic histologic finding to distinguish ICPI-related hepatitis from histologic findings generally seen in drug-induced liver injury (DILI), granulomas, with or without fibrin deposits, and variable presence of plasma cells, continue to be characterized in more recent cases of ICPI-related hepatitis [17].

The current standard initial treatment for ICPI-related hepatitis is immunosuppressive therapy with high-dose corticosteroids such as methylprednisolone or prednisone [7, 8, 11, 16, 17]. In a recent study, few patients were documented to have achieved spontaneous remission without corticosteroid therapy [17].

Infliximab is an anti-TNF-α inhibitor that has been well-studied in treating inflammatory bowel disease. Published reports of its application in this context were reported since 2006 [19], and subsequent applications of its use to treat ipilimumab-induced colitis followed [20–24], where infliximab has demonstrated efficacy in providing prompt remission in diarrhea symptoms. Infliximab-associated liver injury remains a potential concern in such patients [25–30]. The incidence of infliximab-associated liver injury was estimated to be 0.68-0.83% in prior studies examining DILI [29, 30]. This adverse effect may present in several potential forms: hepatocellular injury; cholestatic liver injury, autoimmune-like, or reactivation of viral hepatitis B [26, 28, 31, 32]. Unless jaundice manifests, cases are generally asymptomatic. Liver enzymes are generally monitored periodically while on infliximab therapy.

Despite alcohol use and prior administration of three cycles of ipilimumab and nivolumab, this particular patient’s liver biochemical profile was unremarkable and stable 4 months after initiation of the immune agents. The first sign of ALT elevation occurred only 6 days after the first infusion of infliximab and had increased six-fold by 4 weeks. Hepatitis occurred despite being on prior steroid therapy prescribed for his treatment of colitis. A brief period of high-dose intravenous steroids for presumed ICPI-related hepatitis did not improve the liver enzymes elevation. Overall, ALT levels remained significantly elevated above 200 U/L until 8 weeks post-infusion of infliximab when ALT levels began to decline. There was a mixed cholestatic pattern of liver injury with peak alkaline phosphatase levels as high as 1852 U/L by 8 weeks post-infusion, and direct hyperbilirubinemia with total bilirubin as high as 20.7 mg/dL by 7 weeks post-infusion. This patient did not have hepatitis B infection nor develop any pertinent autoimmune markers. Although EBV and CMV DNA PCR were not specifically tested, the patient’s clinical condition did not eventually worsen in the absence of antiviral therapy, despite the new addition of immunosuppressive therapy.

Cases of ICPI-associated hepatitis, AIH, and other forms of DILI were compared in a previous histologic study that suggested expectation of less confluent necrosis compared to both AIH and DILI, less plasmacytosis compared to AIH, and less eosinophilic infiltration than in DILI, but no difference in portal tract inflammation [9]. The liver biopsy of this patient revealed inflammatory infiltrates attributed to acute hepatitis, whereas findings of steatosis and mild fibrosis may be related to subtle chronic liver disease in the setting of chronic alcohol use. Since the current liver histologic knowledge was insufficient to differentiate ICPI-associated liver injury from other forms of DILI, additional understanding of the pathophysiology of irAE in conjunction with the analysis of the clinical timeline led to a clinical judgment towards attribution of the patient’s liver injury to infliximab exposure.

Diarrhea is a well-known side effect of ICPIs that often requires the use of steroids to induce remission. In steroid-refractory cases, biologic agents such as infliximab have been increasingly employed with good therapeutic effect. Since future patients are likely to encounter such a clinical scenario, hepatotoxicity from infliximab must be carefully monitored. The differential of acute liver injury in this patient population will become more expansive. Therefore, careful evaluation of the pharmacologic chronology and prompt exclusion of other potential acute liver disease, including acute viral hepatitis and immune-related liver injury, are important. While liver histology itself may not necessarily distinguish the causative agent, it may help exclude other known forms of liver disease and assess the degree of liver injury.