Impact of concomitant diabetes mellitus on liver fibrosis and steatosis in chronic hepatitis B

Abstract

Chronic hepatitis B (CHB) infection affects approximately 257 million people globally and causes various debilitating complications including cirrhosis, liver decompensation and hepatocellular carcinoma. On the other hand, diabetes mellitus is an up-surging disease burden worldwide. It is estimated that 8.2% of CHB patients have concomitant diabetes. Despite existing evidence suggesting increased risk of cirrhosis in diabetes patients, data on various diabetes parameters (e.g. diabetic duration, glycaemic control, diabetic complications) on CHB virology and risk of fibrosis/cirrhosis is scanty, and we hypothesised that a few of these diabetic factors might be able to serve as predictors of severe fibrosis/cirrhosis risk. Together with an increasing incidence of diabetes in the CHB population ages, this research gap has become more apparent. On the other hand, regular monitoring of liver fibrosis and steatosis is an essential part in managing CHB patients. Due to the limitations and impracticality of regular liver biopsy, various non-invasive techniques have been developed. Vibration-controlled transient elastography (VCTE) is a well validated technique recommended by the World Health Organisation as the first line investigations of CHB. Recently, multiparametric magnetic resonance (MMR) has been proposed to quantify liver fibrosis and steatosis. However, there was lack of data investigating MMR’s concordance with VCTE. In view of the research gaps, we conducted two studies. The first study (Chapter 3) studies the interaction between glycaemic control, CHB virologic profile and the extent of liver fibrosis in CHB patients with concomitant diabetes. The second study (Chapter 4) investigates the concordance between MMR and VCTE in the assessment of liver fibrosis and steatosis in CHB patients. All study individuals are recruited from the Liver Clinics of Queen Mary Hospital, Hong Kong. In the first study, 560 diabetic CHB patients were matched with non-diabetic counterparts in 1:1 ratio according to age, sex and CHB treatment status. Concomitant diabetes mellitus was shown to be independently associated with the development of severe fibrosis/cirrhosis in CHB. Nearly 30% diabetic CHB patients in our cohort had severe fibrosis/cirrhosis. Compared to non-diabetic ones, they had at least two times the risk of developing severe fibrosis/cirrhosis irrespective of antiviral treatment and HBV DNA levels. The presence of microvascular complications was a significant predictor of severe fibrosis/cirrhosis, whereas diabetes duration and glycaemic control did not show significant correlation with fibrosis. The use of statins among diabetics was found to significantly lower the risk of fibrosis and cirrhosis by half. In the second study, 166 CHB patients received both VCTE and MMR measurements. Liver steatosis quantification was highly concordant between VCTE and MMR. However, concordance between VCTE and MMR in liver fibrosis measurement was suboptimal, especially in patients with low level of liver steatosis. We suggested that further fine-tuning and validation of MMR technique was needed before its integration into complementing CHB fibrosis measurement. In conclusion, this dissertation provided important data on the extent and predictors of liver fibrosis in diabetic CHB patients. The suboptimal concordance between MMR and VCTE also prompted future validation studies.