Increased Risk of Liver Cirrhosis during Azathioprine Therapy for Crohn's Disease

Azathioprine is a cornerstone of the therapy of Crohn's disease. Unfortunately, infections and malignancies are relatively common adverse effects related to this drug; however, cirrhosis is exceptionally reported as a side effect. We report the case of a 49-year-old male patient with ileocolonic steno-penetrating Crohn's disease who developed hepatic cirrhosis while treated with azathioprine. After taking azathioprine for 3 years with regular follow-up, he developed pancytopenia, and liver cirrhosis was diagnosed with ultrasound, abdomen computed tomography scan, transient elastography, and liver biopsy. As all other causes of liver damage were excluded, azathioprine was believed to be the cause of liver injury and therefore was interrupted.

In this report, we present the case of a patient with Crohn's disease who developed hepatic cirrhosis while treated with AZA.

Case Report
A 49-year-old male Italian patient was diagnosed with ileocolonic steno-penetrating Crohn's disease in 1997. He was initially treated with oral steroids and achieved disease remission. He was then treated with mesalazine alone for many years. e patient was referred to our center in September 2014 for a disease relapse. He underwent a colonoscopy and a digestive MRI that showed wall thickening of the last 35 cm of the ileum and of the descending colon; thus, we proposed a step-up therapy with azathioprine 2 mg/kg/ day, which the patient started in October 2014. e patient regularly attended follow-up visits every three months, and routine blood tests revealed normal complete blood count, C-reactive protein, and pancreatic and hepatic enzymes until December 2017, when they gave the following results: pancytopenia (WBC 2200 cells/mm 3 , RBC 2500 cells/mm 3 , and PLT 108000 cells/mm 3 ), high levels of bilirubin (total 2.2 mg/ dL and direct 0.77 mg/dL), and normal levels of GOT, GPT, and cGT (respectively, 41, 26, and 59 U/L). e patient was diagnosed with liver cirrhosis with ultrasound, abdomen computed tomography scan, transient elastography, and a liver biopsy. Ultrasound showed signs of liver cirrhosis, like hypertrophy of the left lobe of the liver, ragged edges, and slightly increased liver echogenicity; portal vein Doppler ultrasound showed a portal diameter of 1 cm, hepatopetal portal venous circulation with a venal flow of 16 cm/sec, thin suprahepatic veins, and splenic enlargement (17 cm). e abdomen CT scan is shown in Figure 1. Transient elastography results were as follows: stiffness 21.6 kPa, IQR 3.4 kPa, and controlled attenuation parameter (CAP) 246 dB/m. e liver biopsy (Figure 2) showed signs of liver cirrhosis, excluding nodular regenerative hyperplasia, veno-occlusive hepatic disease, or a biliary tract disease. e gastroscopy showed congestive gastropathy. Research of possible viral liver damage (HAV, HBV, HCV, HEV, EBV, and CMV) came out negative. ere were no laboratory signs of dyslipidemia, insulin resistance, or diabetes mellitus. Laboratory tests for autoimmune liver diseases (research of antinuclear antibodies (ANA), antineutrophil cytoplasmic antibodies (ANCA), anti-soluble liver antigen antibodies (anti-SLA), anti-smooth muscle antibodies (ASMA), anti-SP100 antibodies, anti-liver kidney microsome antibodies (anti-LKM), antimitochondrial antibodies (AMA), antimitochondrial M2 antibodies (AMA-M2), anti-GP210 antibodies, and anti-liver cytosol type 1 antibodies (anti-LC1)) were all negative. Specific tests were performed to exclude iron and copper overload. As all other causes of liver damage were excluded, AZA therapy was believed to be the cause of liver injury and therefore was interrupted. Six months after discontinuation of AZA treatment, ultrasound and CT scans revealed similar results, whereas white cells and platelet count gradually normalized. e patient was Child-Pugh class A during the whole follow-up period. e patient was monitored every 6 months with abdomen ultrasound and serum alpha-fetoprotein. e patient is currently treated with mesalazine alone and regularly attends follow-up visits at our center.

Review of Literature.
ere are only two other reported cases of thiopurine-induced liver cirrhosis. Trabelsi et al. described the case of an ileocolonic steno-penetrating Crohn's disease patient who was diagnosed with liver cirrhosis during a surgical treatment for an intestinal stenosis. e patient had been treated with AZA therapy for four years, and no alterations of liver enzymes have been detected before [9]. In the second case, which was presented by De Boer et al., liver cirrhosis was discovered as further investigations were made for a new-onset thrombocytopenia, which arose after three years of therapy with azathioprine. In both cases, a liver biopsy was used to diagnose liver cirrhosis, so as to exclude noncirrhotic causes of portal hypertension (veno-occlusive disease, nodular regenerative hyperplasia, hepatic peliosis, etc.) [10].

Genetic Mutations of TPMT: Is ere a Predictive Test for
iopurine-Induced Liver Damage? Mutations of TPMT can lead to a modified ability to convert 6-MP in 6-MeMP. us, patients with modified TPMT activity, who are treated with   Case Reports in Gastrointestinal Medicine standard doses of thiopurine (AZA 2-2, 5 mg/kg/day and 6-MP 1-1, 5 mg/kg/day), have higher levels of 6-MeMP and therefore have an increased risk of thiopurine-induced toxicity [15]. Screening for variants in TPMT may allow the identification of patients with a high risk of thiopurineinduced toxicity, to whom drugs should be administered in reduced doses [16].

Conclusions
Liver cirrhosis is a rare but possible adverse event of AZA therapy for Crohn's disease; hence, we suggest adding the routine platelet count and upper abdomen US in Crohn's disease patients who receive thiopurine treatment. If thrombocytopenia arises during AZA therapy, further investigations should be performed. Furthermore, for preventive purposes, screening for variants in TPMT should be carried out in all patients before initiating thiopurine therapy, in order to adjust the dose of the drug, reduce the risk of complications, and avoid wasting a valid therapeutic option.

Conflicts of Interest
e authors declare that they have no conflicts of interest.