NAFLD, the changing face of IBD

J. Olynyk,L. Thin,J. Kava,I. Lawrance

2012-10-01

Abstract:Background It is estimated that up to 5% of inflammatory bowel disease (IBD) patients have clinically significant liver disease due to multifactorial causes such as underlying Primary Sclerosing Cholangitis, pharmacotherapy, fatty liver disease or nodular regenerative hyperplasia. In recent years, transient elastography (TE), which uses the sonic detection of liver stiffness to predict hepatic fibrosis has increasingly replaced the need for a liver biopsy. It has been validated in patients with chronic hepatitis C as an accurate non-invasive predictor of advance fi brosis and cirrhosis. Aims Our aim was to evaluate the prevalence of clinically significant liver disease in IBD patients as defi ned by an increased liver stiffness measurement (LSM) using Transient Elastography (FibroScan®). Methods 110 random IBD patients and 55 non-IBD control patients (composed of patient relatives and hospital staff) had their LSM recorded. The median reading in kilopascals (kPa) of 10 successful acquisitions was recorded by a nurse trained in performing TE. Age, gender, body mass index (BMI), history of liver disease, presence of diabetes, number of cigarettes smoked and standard drinks of alcohol consumed per week, were recorded and entered as covariates in multivariate analysis. Results Of the 110 IBD patients, 71 (64.5%) had Crohn’s disease. There were no significant differences in mean age (43 ± 15 yrs vs. 41 ± 17 yrs), gender (males, 47% vs. 51%) and number of standard alcoholic drinks consumed per week (6 ± 9 vs. 4 ± 7) between IBD patients and non-IBD controls respectively (P > 0.05). IBD patients smoked significantly more cigarettes per week compared to non-IBD controls (15 ± 42 vs. 1 ± 6, P = 0.001). Four IBD patients had diabetes compared to none in the non-IBD group. Seven patients in the IBD group (6.4%) had an LSM reading of greater than 8 kPa compared to none in non-IBD control group. One patient, had an LSM of 14.9 kPa, normal liver function tests and evidence of portal hypertension on ultrasound. Although the overall LSM in the IBD group was higher, there was no statistical difference between the mean LSM reading of IBD patients (5.2 ± 2.5 kPa) and non-IBD controls (4.8 ± 1.2 kPa) on univariate (P = 0.22) and multivariate (P = 0.67) analysis. IBD patients had a significantly higher BMI compared to non-IBD controls (26 ± 4.6 vs. 24 ± 3.3, P = 0.04) and this was the only independent predictor of a high LSM reading on multivariate analysis (P < 0.0001). Conclusion IBD patients are more likely to have a higher BMI compared to non-IBD subjects and this is the main reason for their higher LSM reading. This implies that non-alcoholic fatty liver disease may be a significant cause of occult liver disease in IBD patients and attention should be paid to optimise metabolic risk factors such as minimising corticosteroid use.

Medicine