Hepatic progenitor cells in acute and chronic liver disease: Clinical aspects

In humans, increasing evidence supports the existence and activation of hepatic progenitor cells (HPCs) in different liver diseases, which are all characterised by a variable degree of hepatocyte loss and damage with impaired regeneration of remaining hepatocytes and/or bile duct epithelial cells [1], [2], [3], [4], [5], [6], [7], [8], [9]. In chronic hepatitis, progenitor cell expansion is an important source of hepatocyte regeneration. The degree of activation and differentiation towards hepatocytes correlates with the degree of inflammatory activity and with the stage of the disease as well [10], [11], [12]. However, little is known in humans about the response of HPCs in advanced stages of liver failure, such as acute liver failure (ALF) and acute-on-chronic liver failure (ACLF).

Liver failure can occur in three different clinical conditions: ALF, decompensated cirrhosis and ACLF. ALF is characterised by acute massive necrosis of hepatocytes leading to development of jaundice, coagulopathy and hepatic encephalopathy. It can own to a large variety of causes. This rather uncommon condition is clinically relevant because of the high mortality and morbidity. Liver transplantation has significantly improved the survival [13], [14]. However, it is often difficult to determine both the need and the exact timing for transplantation as spontaneous recovery might be possible [15]. In general, the prognosis of ALF depends on the interaction of different factors: hepatic regeneration, hepatocellular failure, brain oedema (and encephalopathy) and the presence of multi-organ failure. Most of the current scores that assess the outcome of ALF contain almost only elements that are the expressions of liver impairment (MELD score is a typical example of that). However, an important factor that influences the outcome of ALF is the capacity of regeneration of the liver and this is probably the reason why aetiology and age are important predictors of outcome; this has been illustrated in the studies at King’s College Hospital and the Clichy criteria for acute liver transplantation [16], [17]. It has recently been shown in acetaminophen-induced liver failure that changes in α-fetoprotein levels, as a parameter of liver regeneration, might indeed improve our prognostic accuracy [18]. Therefore, we recently investigated the role of hepatocyte proliferation and of HPC activation and differentiation in the liver repopulation in this condition [19]. We examined liver specimens from 74 patients with acute or sub-acute liver impairment. Of the 74 patients, 32% survived without transplantation, 14% died without transplantation and 54% were transplanted. Our results show that a threshold of 50% loss of hepatocytes, associated with significant decrease in the proliferative activity of remaining mature hepatocytes, is needed for extensive hepatic progenitor cell activation. Such activation was a sign of disease severity and occurs early (within 1week) in the course of disease. However, development of intermediate hepatocytes, suggesting HPC differentiation towards mature hepatocytes, takes at least a week’s time. We found a positive correlation between histopathological parameters (percentage hepatocyte loss, number of proliferating hepatocytes and number of HPCs) and clinical parameters of liver impairment, such as MELD. Surviving patients had significantly less hepatocyte loss, less HPC activation and more mature hepatocyte proliferative activity compared with those who either died or were transplanted. Hepatocyte proliferative activity and degree of hepatocyte loss were the most important independent histopathological parameters in predicting outcome. The data reconfirm the value of HPCs in liver regeneration. However, the clinical diagnostic implications might be limited. The above-mentioned histological signs were mostly heterogeneously disturbed inside the liver and can be as such missed by needle puncture biopsy. Nevertheless, the data are encouraging enough for further validation.

A second condition of liver failure occurs in chronic hepatic decompensation in patients with end-stage cirrhosis as a result of the progression of the primary liver disease. It is irreversible when there is no control of the aetiologic factors. This condition is well characterised and the prognosis can be accurately assessed by the MELD score, which contains factors (INR, the international normalised ratio, and bilirubin), representative for the liver function and creatinine as a parameter of the degree of systemic haemodynamic instability induced by the extrahepatic manifestation of portal hypertension [20].

The third condition is the ACLF. It remains a poorly defined entity. It occurs in patients with previously well-compensated cirrhosis in whom an acute decompensation of liver function occurs due to a precipitating event such as sepsis, upper gastrointestinal bleeding, ischaemia and additional liver injury due to alcohol or hepato toxic drugs or hepatitis virus infection. This condition is mostly characterised by a progressively extensive increase in serum bilirubin in a patient with underlying cirrhosis. The haemodynamic derangements typical of cirrhosis are exaggerated [20], [21]. Current medical therapy involves the management of the precipitating event and treatment of complications until the liver eventually recovers, leaving us with no other treatment options than transplantation if these attempts fail. As liver failure in ACLF is often partially potentially reversible, considerable effort has been invested in the development of liver support systems. Currently, most of the experience is available for non-biological support systems [22]. A recent meta-analysis suggested that an artificial liver support reduces mortality in ACLF, whereas another systematic review was unfavourable [23], [24]. The results of ongoing randomised-controlled multicentre studies are eagerly awaited. Since ACLF is potentially reversible, we investigated the clinicopathological features corresponding to HPCs also in this situation and compared the results with our findings in patients with chronic hepatic decompensation. There was a substantial difference in outcome in the two groups of patients: the 1-month mortality was, respectively, 46% and 6% for ACLF and decompensated cirrhosis. In both groups of patients HPCs were strongly activated and a significant increase of intermediate hepatocytes was seen, in accordance with our previous observation in patients with cirrhosis [4]. In contrast, for patients with ALF the histopathological features were not directly correlated with prognosis. However, there was no difference between the two groups of patients, despite a significant difference in outcome. This suggests that, in cirrhosis, HPCs are already maximally activated. Intrahepatic shunts, which are present in case of cirrhosis, will further harm the process of regeneration.

In conclusion, we provide further evidence of the role of hepatic progenitor cells in the regeneration of the liver in human liver disease. Activation of HPCs is a sign of severe liver impairment if the normal regenerative capacity of the liver fails. In ALF, this is directly linked with outcome.