MRI is a regular examination to measure the degree of fibrosis, portal hypertension and nodular in cirrhotic patients. In our study, we focused on the liver iron, fat content assessed by MRI, and try to figure out the prevalence of liver iron overload and NAFLD in chronic hepatitis B patients, risk factors that associated with iron overload, NAFLD were estimated too by comparing the common clinical characteristics, we further explored the potential of liver iron overload and NAFLD for identifying the severity of liver fibrosis. Our report indicated that iron overload was common in CHB with a prevalence of 35.9%, especially in those with older age, male and higher total bilirubin. The prevalence of NAFLD in CHB patients was 15%, especially in those with high platelet count and without portal hypertension. There seemed to be little correlation between LIC and liver fibrosis, with a slight diagnostic ability for cirrhosis with the AUC ranged from 0.5 to 0.7. The diagnostic ability of PDFF to distinguish cirrhosis from non-cirrhosis stages was slight to moderate when the AUC was between 0.6 to 0.8.
Martinelli et al. found that hepatic iron deposits in 48.7% of cases among 39 HBV patients measured by liver biopsy[25]. while Cynthia Ko et al. examined the prevalence of hepatic iron overload in 5224 patients undergoing liver transplantation, only 13.3% patients had liver iron overload in those with HBV infection[3]. Several studies had evaluated iron overload in hepatitis C, with an increased liver iron concentration in 10–36% patients[26, 27]. Ito et al. evaluated magnetic resonance images for diffuse hepatic iron deposition, 40% show positive in cirrhotic patients with hepatic cellular carcinoma[28]. The difference among those studies may due to the various criteria used in the definition of iron stores.
Ferritin is regarded as the primary tissue iron-storage protein in the liver, which is induced in iron overload disorders of various etiologies, resulting in increased hepatic and circulating ferritin levels[29]. Hyperferritinemia has been observed in chronic liver disease due to hepatitis C and alcohol[30–32], but its relationship with hepatic iron deposition in such situation is not known yet. In our study, serum ferritin elevated can be seen in more than a half patient with CHB, which had a strong correlation with liver iron concentration measured by MRI. Cristina Ripoll report 59% cirrhotic patients had increased serum ferritin, which was similar to our study[33]. But in the study of 81 CHB patients without cirrhosis, there was no significant difference between patients with or without hepatic iron deposit. What is more, The markers of liver failure such as bilirubin, INR and MELD score were observed significantly difference in the patients with elevated-ferritin level. A recently published study including 318 patients with decompensated cirrhosis divided patients in 3 groups according to ferritin and observed greater MELD and Child-Pugh score with increasing ferritin, similar to our observations. The marker of liver inflammation such as AST, ALT were also elevated in the high-level of ferritin group, but cannot seen in the subgroup of elevated LIC. It can be explained that the ferritin is induced by systemic inflammation, so that the situation of HBV replication and other inflammation could lead to the increased ferritin. However, the above liver enzymes showed no significant elevation in the patients with increased LIC. It suggested that when we use the serum ferritin level to evaluate the liver iron overload, the inflammation status should be taken into account.
Several physiological mechanisms, especially the reactive oxygen species (ROS) accumulation and damage could explain the iron overload in liver fibrosis, low to moderate levels of excess iron are sufficient to support the pathological progression[34]. Prior study indicated that iron can increase hepatitis B virus mRNA expression in HepG2 cells[35], which may contribute to sustenance of infection and inflammation, thereby potentiating fibrosis. The iron-related parameters aid in prediction, diagnosis, staging and prognosis of liver fibrosis, when used in combination with the routine markers of liver dysfunctionality. Metwally et al. found that increased hepatic iron deposition may be associated with more advanced hepatic fibrosis in patients with chronic hepatitis C virus infection[36]. Martinelli et al. also demonstrated that patients of CHB with liver iron deposits exhibited significantly higher scores for necroinflammatory activity and fibrosis than those without iron deposits. In addition, patients with moderate liver iron deposits had significantly higher histologic activity index (12.8 ± 3.2 and 7.3 ± 3.7, respectively, P = 0.001) and liver fibrosis (2.3 ± 0.8 and 1.5 ± 0.6, respectively, P = 0.02) scores compared with those with absent or mild liver iron deposits[25]. In our study, the LIC was not strongly positively associated with the severity of liver fibrosis, which was opposite to the above studies. The reason might due to the LIC could not reflect the whole iron overload status, and extrahepatic iron load can also lead to the progression of liver fibrosis. Further studies is needed to evaluate the value of combining iron deposition parameters and other noninvasive indexes in the prediction of liver fibrosis.
The prevalence of NAFLD in hepatitis B seemed less common. In our study, only 15.0% patients had liver steatosis measured by MRI-PDFF, while the most steatosis was mild. The global prevalence of NAFLD is currently estimated to be 24%, while 27% in Asia, and it is increasing year by year due to the change of life style[37]. The prevalence of NAFLD is approximately 14–67% in Asian individuals with CHB, similar to the data in western countries[38, 39]. In former studies, mounting evidence tends to support a potentially negative association between CHB and NAFLD in terms of hepatitis B serum markers, as well as onset of NAFLD. Some researchers demonstrated a significantly higher incidence of HBsAg clearance in HBeAg-seronegative CHB patients with hepatic steatosis than in those without, hepatic steatosis was further identified as an independent predictor (HR = 1.222) of spontaneous HBsAg seroclearance in patients with CHB (including both HBeAg seropositive and seronegative patients)[9]. On the other hand, there has been evidence that indicates the association of chronic HBV infection with reduction in either hyperlipidemia or NAFLD incidence, confirmed by our data. A cross-sectional study in 7695 Taiwanese adults showed that HBV-infected individuals exhibited lower risk of hypercholesterolemia (OR = 0.8), hypertriglyceridemia (OR = 0.7), and high low-density lipoprotein cholesterol level (OR = 0.8)[40]. A large cross-sectional study in Hong Kong found a significantly lower risk of NAFLD in HBsAg-positive subjects (adjusted OR = 0.42)[41]. Another cross-sectional study in Taiwan also found a negative association of HBV infection with NAFLD, especially in individuals with BMI > 22.4 kg/m2 or age > 50 years[42].
What’s more, higher PLT and no portal hypertension were predictive factors of liver steatosis in CHB cirrhotic patients. PDFF was lower in cirrhotic patients than those without cirrhosis. By the ROC analysis, PDFF show a slight to moderate diagnostic ability to distinguish cirrhosis from non-cirrhosis patients. It might get an conclusion that as the progress of liver fibrosis and portal hypertension, the prevalence of liver steatosis became decreased. But the above findings seemed to be opposite with other studies. A cohort study evaluated the evolution of liver fibrosis using liver stiffness measurement, along with hepatic steatosis by controlled attenuation parameter, in 459 HBeAg-negative patients across a 10-year interval. They found that hepatic steatosis was associated with fibrosis progression in CHB patients (OR = 7.799)[43]. A clinical study in Thailand identified steatohepatitis as an independent predictor of significant fibrosis (adjusting OR = 10) and advanced fibrosis (adjusting OR = 3.45)[44]. It can be explained by these reasons: firstly, our study was a cross-sectional study, which can not reveal the causal relationship; secondly, the included patients with severe fibrosis or cirrhosis shared a relatively high proportion, and some patients had been suffering from the ascites, variceal bleeding, malnutrition, resulting in change of liver fatty metabolism like steatosis attenuation or fatty loss. Finally, more studies are needed to offer the high-level evidence in terms of the correlation with NAFLD and CHB in the whole course of liver pathologic and immune progression.
Our study has some limitations. We didn’t get the pathologic data of live iron concentration, fat deposit and the HVPG as the gold standard for diagnosis of iron overload, steatosis, portal hypertension, and liver fibrosis stages, which may lead to the bias distortion between the gold standard and MRI measurement. Parts of the patients included were in decompensated stage and suffered from variceal bleeding, ascites, malnutrition and got some therapy like blood transfusion, iron therapy, nutrition therapy and non-selective β-blocker therapy, which may change the concentration of liver iron and fat, also the degree of portal hypertension. Since MRI is a non-invasive and convenient tool, we will further combine the liver biopsy and MRI measurement and comprehensively evaluate the role of MRI in identification of severity, complications, therapy response and progress in various causes of cirrhosis.
In conclusion, in CHB patients, iron overload was common and should be evaluated in those with older age, male and high TB level especially. Liver steatosis is less common and the steatosis was usually mild. Liver iron and fat measured by MRI might reflect the severity of liver fibrosis in CHB patients. NAFLD seems to be a protective factor of liver fibrosis. Further study should focus on the influence of iron overload and the therapy target, the effect of iron overload and NAFLD in different CHB stage and how to manage it.