PAF is a potent inflammatory lipid mediator that, by binding to a high-affinity G-protein–linked receptor, can be activated to exert diverse actions including stimulating secretion and aggregation of platelets, initiating neutrophil and macrophage chemotaxis, and inducing release of cytokines such as interleukins, tumor necrosis factor, and proteolytic enzymes, and thus may be involved in the development of circulation disorders and inflammation . PAF accumulation has been implicated in pathological processes and diseases including inflammation, endotoxin shock, acute pancreatitis, and cardiovascular disease. PAF is also involved in acute liver damage, cirrhosis, severe hepatitis, and ischemia-reperfusion–induced liver injury [19, 20]. As a major mediator of PAF inactivation, PAF-AH plays a crucial role in the regulation of serum PAF levels and in reducing PAF-induced damage [8, 21–23]. In experimental models of acetaminophen-induced liver injury in rats, PAF activities increased significantly between 24 and 32 h after acetaminophen administration, along with increases in other biochemical indexes (ALT, AST). The PAF-AH activity peaked between 72 and 96 h after acetaminophen treatment , indicating that PAF plays an important role in acetaminophen-induced liver injury and subsequent liver tissue repair, while PAF-AH can increase liver recovery and reduce liver damage [25, 26].
The pathological mechanism of hepatitis B is complex, and patients’ symptoms are often complicated by intestinal endotoxemia, the incidence of which can reach 80%–100% in severe hepatitis cases . Lipopolysaccharide (LPS) is the major chemical endotoxin capable of stimulating PAF and PAF-AH secretion from monocytes and macrophages. Kupffer cells, specialized macrophages located in the liver, account for 80%–90% of the total monocytes and macrophages in the body. Kupffer cells are the primary mediators of endotoxin clearance and detoxification in the liver. By injecting bacterial LPS into the rat mesenteric vein, Howard et al.  observed a 20-fold increase in the PAF-AH mRNA level in Kupffer cells and a 2-fold increase in serum PAF-AH activity after 24 h. Svetlov et al.  reported that cultured primary Kupffer cells could express more PAF-AH mRNA than hepatocytes and had a 20–25-fold higher PAF-AH secretion rate than hepatocytes, indicating that Kupffer cells may be the main source of PAF-AH during liver damage.
We found that the activity of circulating PAF-AH in patients diagnosed with hepatitis B was positively correlated with TBIL, TBA, ALT, AST, TG, and apoB, negatively correlated with ChE, HDL-c, and apoAI, and not correlated with Glu, BMI, Tch, and LDL-c. However, the PAF-AH activity in healthy controls was positively correlated with TBIL, ALT, TG, Tch, LDL-c, and apoB, negatively correlated with HDL-c, and not correlated with Glu, BMI, TBA, AST, and apoAI. The differences in correlations between hepatitis patients and controls, especially in the correlation of PAF-AH with lipids and bile acids, could be explained by a change in the source of PAF-AH during the development of hepatitis. PAF-AH can be classified into intracellular types I and II and the plasma type . Plasma PAF-AH exists in the blood and is predominantly produced by monocytes, macrophages, T lymphocytes, mast cells, and hepatocytes [30–32]. Under normal conditions, the main source of circulating PAF-AH is hematopoietic cells , and the main source of PAF-AH in bile juice is hepatocytes . However, during hepatitis bile excretion disorder, the retention of bile components such as bile acid can cause the retention of hepatocyte-secreted PAF-AH [33, 34]. This relationship could explain the association between serum PAF-AH activities and the TBIL and TBA levels in hepatitis. Another potential factor is the impact of liver damage on Kupffer cell PAF-AH secretion, leading to an increase in PAF-AH during hepatitis. Circulating PAF-AH mainly exists in complexes with lipoprotein particles . During hepatitis, however, liver damage causes dysfunction in the synthesis of cholesterol and other lipids , resulting in the altered correlation between PAF-AH and blood lipids we observed. These findings indicate that serum PAF-AH may be involved in oxidative stress and inflammation of the liver.
We also found that serum PAF-AH activities in patients with various stages of hepatitis B were significantly higher than those in healthy controls, and serum PAF-AH activity was significantly positively correlated with TBIL. The greatest elevation in serum PAF-AH was observed in patients diagnosed with CSHB, suggesting that the PAF-AH activity is involved in pathological liver damage, and that the detection of PAF-AH may serve as a surrogate marker for hepatic inflammatory activity, allowing disease progress and prognosis to be monitored. Similar to our study, the study of Ma et al. found that among CSHB patients, serum PAF activities were significantly higher in the death group than in the control group and were positively correlated with the prognosis of CSHB, and thus can be used as a prognostic factor in CSHB . However, Guerra et al. found that HCV-infected patients showed a significant decrease in PAF-AH activity, and PAF-AH activity recovered only in patients who cleared HCV after antiviral treatment . Most of the circulating HCV is associated with VLDL and LDL, and the circulating HCV lipoprotein complexes may reduce the amount of free LDL available for PAF-AH activity. Since HBV does not circulate in blood bound to LDL, HBV-infected patients did not show any reduction in PAF-AH activity [38, 39].
In this study, we show that serum PAF-AH may be used for predicting CSHB and 3-month mortality in HBV-infected patients. We evaluated the predictive power of PAF-AH and the MELD score by ROC curve analysis. The MELD score is a prospectively developed and validated scale for the severity of end-stage liver disease that uses the quantitative, objective values of serum TBIL, serum creatinine, and INR of prothrombin time to predict patient mortality in patients with advanced liver disease [16, 40]. Mao et al.’s study reported that the MELD score was related to the prognosis of patients with HBV-related acute-on-chronic liver failure . We found that although the AUC of PAF-AH was lower than the AUC of the MELD score for predicting CSHB, the AUC of PAF-AH combined with MELD score was higher than that of PAF-AH or the MELD score alone. This finding showed that combining PAF-AH with the MELD score further added to predictive power for CSHB. Multivariate logistic regression analysis showed that PAF-AH activity and the MELD score were independent factors predicting CSHB. We also found an association between PAF-AH and mortality in HBV-infected patients. The AUC of PAF-AH was lower than AUC of the MELD score for predicting mortality, and the AUC of PAF-AH combined with the MELD score was similar to that of the MELD score alone. Although this result shows a lower predictive power of PAF-AH than of the MELD score for mortality, PAF-AH activity is an easier and more convenient measurement than the MELD score. However, a potential limitation in our study was that we did not consider the influence of drugs taken before admission that may modulate PAF-AH activity.
In summary, the PAF-AH activities of patients with CSHB were significantly higher than those of patients with AHB, CHB, or LC, and higher PAF-AH activities were associated with a higher prevalence of CSHB. Spearman correlation showed that PAF-AH activity correlated positively with TBIL, TBA, ALT, AST, TG, and apoB and negatively with ChE, HDL-c, and apoAI in patients with hepatitis B. Moreover, serum PAF-AH may be used for predicting CSHB and mortality in patients with hepatitis B, and PAF-AH activity was an independent factor predicting CSHB.