Introduction
Drug-induced liver injury (DILI) has been associated with over 1,100 drugs and herbal medications1 and can present with all forms of acute and chronic liver diseases.2,3 Most cases of DILI are self-limiting and resolve with prompt identification and cessation of the offending agent;2,3 however, chronic DILI, defined as persistent abnormalities in liver-associated enzymes at 3–6 months after drug withdrawal,3,4 is becoming increasingly recognized.5
Vedolizumab (VDZ) is an α4β7 integrin inhibitor recently approved for use in patients with moderate-to-severe Crohn’s disease and ulcerative colitis (UC).6 In general, VDZ is well tolerated; serious adverse events (SAEs) have been reported in 7–12% of patients in the pre-marketing phase III trials,7 but this range is similar to that of the SAEs experienced by the placebo groups (4–9%). Liver injury has rarely been reported with VDZ. Review of the aggregate pre-marketing data showed that among the 3,326 patients exposed to VDZ in phase III, controlled or open-label studies, < 2% experienced elevations in aminotransferases > 3× the upper limit of normal.6,7 Four cases of acute DILI, with or without cholestasis, have been reported and these resolved with drug discontinuation and/or administration of corticosteroids.6 Consequently, the drug-package insert states that VDZ should be discontinued in patients with jaundice or other evidence of significant liver injury.6 A recent post-marketing multicenter experience in 172 patients failed to find a single hepatotoxic event, and the authors concluded that VDZ is safe and well tolerated in clinical practice.8
We present herein the first case of post-marketing hepatotoxicity attributable to VDZ that presented with acute cholestasis and progressed to chronic liver injury despite prompt drug withdrawal and corticosteroid administration.
Case reports
A 23-year-old Caucasian male with pancolonic UC diagnosed at age 19 and concurrent primary sclerosing cholangitis (PSC) diagnosed at age 22 was evaluated at our institution in the Digestive Health outpatient clinic. His prior inflammatory bowel disease (IBD) therapy included 5-aminosalicylate medications, corticosteroids, azathioprine and biologic agents (infliximab and adalimumab) prescribed by his gastroenterologist following the initial diagnosis, but achieved no clinical or endoscopic remission. Examination at the time of his initial adult consultation did not reveal any stigmata of advanced liver disease. Pre-treatment liver biochemistries were notable for aspartate aminotransferase (AST) at 31 U/L, alanine aminotransferase (ALT) at 33 U/L, alkaline phosphatase at 648 U/L (61% heat stabile; normal range 25–40%), total bilirubin (TB) at 1.4 mg/dL and albumin at 2.9 g/dL. Gamma glutamyltransferase was 130 U/L (normal < 55 U/L). Other notable lab results included platelet count of 525 k/μL, international normalized ratio (INR) of 1.1, sedimentation rate of 24 mm/hr (normal 0–15 mm/hr) and C-reactive protein at 0.1 g/dL (normal < 0.5 g/dL).
After the patient’s initial consultation, he was initiated on VDZ and received three doses of 300 mg intravenous infusions at weeks 0, 2 and 6. After the third infusion, the patient developed insidious onset of jaundice and fatigue. Importantly, he had never experienced overt jaundice due to PSC previously. Laboratory tests at this time were remarkable for AST at 52 U/L, ALT at 27 U/L, alkaline phosphatase at 370 U/L, TB 13.3 mg/dL, and conjugated bilirubin 9.3 mg/dL. The R ratio was calculated [(ALT/upper limit of normal (ULN))/(alkaline phosphatase/ULN)] and found to be 1.1, indicating cholestatic liver injury (R ratio < 2 is diagnostic).
Magnetic resonance imaging of the abdomen, including magnetic resonance cholangiopancreatography scanning, did not reveal any dominant stricture and the findings were unchanged compared to the index images that were taken one year prior, at the time of the PSC diagnosis. A comprehensive hepatitis viral serology panel was unrevealing. The patient denied any alcohol or toxin exposure. Testing for autoimmune hepatitis and hemochromatosis screening was negative. The Roussel Uclaf Causality Assessment Method (commonly known as RUCAM), which uses seven different criteria to quantify the strength of association between a liver injury and the medication, implicated VDZ as the offending agent with a score of 7 (probable liver injury) (Fig. 1).
VDZ was subsequently discontinued after the third infusion. Six weeks after the patient’s last dose of VDZ, he experienced an increase in bowel frequency with associated bleeding and was started on prednisone. Lab results from several weeks after the corticosteroid administration showed marked improvement accompanying the continued corticosteroid use, with TB at 4.9 g/dL and alkaline phosphatase at 485 U/L. The patient’s jaundice and fatigue also improved. Corticosteroid therapy was continued for the IBD, and anti-TNF therapy with adalimumab was resumed 4 months after the last dose of VDZ; the patient experienced intervals of improvement in his bowel symptoms but not complete clinical remission.
Importantly, the patient’s cholestatic liver injury persisted 5 months following the drug withdrawal, meeting criteria for chronic DILI with TB at 1.7 g/dL and alkaline phosphatase at 556 U/L. Pre-DILI diagnosis lab results were compared to post-DILI lab results within a 6-week window in each period by using a paired t-test. TB levels were significantly different (pre-DILI: 2.3 ± 0.6 g/dL vs. post-DILI: 10.4 ± 4.2 g/dL, p = 0.037) while AST and ALT were not significantly different by the statistical testing (Table 1). Figure 2 shows the graphic trend in the patient’s liver-associated enzymes. Informed consent was obtained from the patient.
Table 1.Liver-associated enzymes in the course of vedolizumab treatment
| AST | ALT | Total bilirubin | ALP |
| First VDZ infusion | 31 | 33 | 1.8 | 461 |
| 8 days after VDZ initiation | 63 | 62 | 2.2 | 513 |
| 15 days after VDZ initiation/second infusion of VDZ | * | * | 2.9 | 460 |
| 41 days after VDZ initiation/VDZ discontinued | 52 | 27 | 13.3 | 359 |
| 35 days after VDZ discontinuation | 48 | 25 | 13 | 370 |
| 49 days after VDZ discontinuation/prednisone (10 mg daily) initiated | 55 | 16 | 7.4 | 509 |
| 62 days after VDZ discontinuation | 78 | 52 | 4.9 | 485 |
| 92 days after VDZ discontinuation/Humira started | 50 | 33 | 5.3 | 513 |
| 156 days after VDZ discontinuation | 55 | 77 | 1.7 | 556 |
Discussion
While pre-marketing data suggests that hepatotoxicity in the setting of VDZ use for IBD is a rare and self-limiting event, with or without cholestasis,6,7 we present the first post-marketing case of chronic DILI attributable to VDZ use. Chronic liver injury from medications is becoming increasingly recognized,6 with prevalence rates approaching 20% based upon data from the Drug-Induced Liver Injury Network (DILIN).3 Patients with cholestatic injury, such as our patient, appear to be at increased risk for progressing to chronic injury.3,10 Adherence to appropriate prescription regimens remains one of the most important aspects for preventing DILI in patients with underlying chronic liver disease.9 The drug package insert for VDZ states that elevations of liver-associated enzymes have been reported in patients receiving VDZ and that in the absence of biliary obstruction VDZ should be discontinued in patients with jaundice or other evidence of significant liver injury.6
Whether or not VDZ may worsen bile duct injury in the setting of PSC through an unknown mechanism remains to be determined. In general, PSC is not thought to place patients at increased risk for DILI.2 Rather, secondary sclerosing cholangitis typically occurs as a sequela of chronic drug-induced bile duct injury.10 A solitary case report of DILI in the setting of certirizine use in a patient with PSC has been reported.11 Based on the available evidence, in combination with our isolated case report, we cannot recommend avoiding VDZ in IBD patients with concomitant PSC; however, monitoring hepatic enzymes during VDZ therapy is warranted.
In general, treatment options for DILI are limited. Treatment is centered around prompt identification and removal of the offending drug.12 While corticosteroids were administered to several patients in the pre-marketing trials who experienced hepatotoxic events, the efficacy of the use of corticosteroids in treating DILI remains largely anecdotal.13 Immunosuppressive therapy may be considered on a case-by-case basis, where autoimmune-like DILI features, including the presence of auto-antibodies or liver histology supporting this diagnosis, are present and cessation of the drug does not lead to complete resolution of injury.5
In this case, corticosteroids were administered to address worsening IBD activity and not as treatment for DILI, but they may have hastened clinical improvement following the hepatotoxic event. On the other hand, corticosteroids and adalimumab have also been implicated in cholestatic liver injury. In the present case, the patient had previously been on these agents for his UC, without showing evidence of liver injury. In addition, the patient’s total bilirubin continued to improve after initiation of corticosteroids and adalimumab, which would argue against these therapies leading to persistent chronic liver injury.
While this case has several limitations, including an incomplete evaluation for other chronic liver diseases (namely, alpha-1 antitrypsin deficiency or Wilson’s disease) and a lack of liver biopsy as a confirmatory study, a high-index of suspicion for the development of hepatotoxicity in patients with IBD being treated with VDZ seems necessary based on our report, nonetheless.
Abbreviations
- DILI:
drug-induced liver injury
- VDZ:
Vedolizumab
- UC:
ulcerative colitis
- SAEs:
serious adverse events
- PSC:
primary sclerosing cholangitis
- IBD:
inflammatory bowel disease
- AST:
aminotransferase
- ALT:
alanine aminotransferase
- TB:
total bilirubin
- INR:
international normalized ratio
- ULN:
upper limit of normal
Declarations
Acknowledgement
The research reported in this publication was supported by the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health under award number T32DK007769. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
Conflict of interest
None
Authors’ contributions
Manuscript design, manuscript writing, final approval (JGS), manuscript writing, final approval (JW), manuscript design, final approval (BWB).