Hepatocyte expression of the senescence marker p21 is linked to fibrosis and an adverse liver-related outcome in alcohol-related liver disease

Abstract

Background and aim: Alcohol-related liver disease (ALD) remains a leading cause of liver-related morbidity and mortality. Age, fibrosis stage, MELD score and continued alcohol consumption predict outcome in everyday clinical practice. In previous studies increased hepatocyte nuclear area and hepatocyte expression of p21, both markers of senescence, were associated with increased fibrosis stage and a poor outcome in non-alcohol-related fatty liver disease, while increased hepatocyte nuclear area was related to liver dysfunction in ALD cirrhosis. This study, therefore, investigated the pattern of hepatocyte cell cycle phase distribution and hepatocyte p21 expression in relation to outcome in ALD.

Methods: Liver sections from two cohorts were studied. The first comprised 42 patients across the full spectrum of ALD. The second cohort comprised 77 patients with ALD cirrhosis. Immunohistochemistry assessed hepatocyte expression of cell cycle phase markers and p21. Regenerating liver (n=12) and "normal" liver sections (n=5) served as positive and negative controls, respectively.

Results: In the first cohort there was little cell cycle progression beyond G1/S phase and increased hepatocyte p21 expression (p<0.0001), which correlated independently with fibrosis stage (p=0.005) and an adverse liver-related outcome (p=0.03). In the second cohort, both hepatocyte p21 expression (p<0.001) and MELD score (p=0.006) were associated independently with an adverse liver-related outcome; this association was stronger with hepatocyte p21 expression (AUROC 0.74; p=0.0002) than with MELD score (AUROC 0.59; p=0.13). Further, hepatocyte p21 expression co-localised with increased hepatic stellate cell activation.

Conclusions: The findings are consistent with impaired cell cycle progression beyond the G1/S phase in ALD. The striking independent associations between increased hepatocyte p21 expression and both fibrosis stage and an adverse liver-related outcome in both cohorts suggests hepatocyte senescence plays an important role in ALD. Measuring hepatocyte p21 expression is simple and cheap and in this series was a useful measure of long-term prognosis in ALD.

Figure 1. The distribution of cell cycle phase markers in the first cohort and controls.

Examples of the immunohistochemical staining of Mcm-2, cyclin A, PH3 and p21 in regenerating liver (positive control tissue), liver from a representative patient with ALD and normal liver (negative control tissue). Hepatocyte Mcm-2 expression was higher in ALD and regenerating liver (positive control) compared to normal liver (negative control). Cyclin A and PH3 expression were lower in ALD compared to regenerating liver. In contrast hepatocyte p21 expression was higher in ALD compared to regenerating liver.

Figure 2. Hepatocyte p21 expression in relation to fibrosis stage and laboratory indices.

Hepatocyte p21 expression demonstrated an association with fibrosis stage [2a] in the first cohort; the proportion of hepatocytes that expressed p21 increased with increasing fibrosis stage. Areas of increased α-SMA expression (a marker of activated hepatic stellate cells) were associated with higher hepatocyte p21 expression (2b) than those areas with less α-SMA expression, which were associated with lower hepatocyte p21 expression (2c) even within the same tissue. Expression of hepatocyte p21 correlated positively with prothrombin time in both cohorts (Figure 2d & 2e, respectively) and correlated negatively with serum albumin in the second cohort (Figure 2f).

Figure 3. Differential hepatocyte p21 expression in the second cohort.

Figure 3 shows fibrosis using reticulin staining (above) and hepatocyte p21 expression (below) in a patient with ALD cirrhosis in second cohort.

Figure 4. Hepatocyte p21 expression, MELD score and alcohol consumption at the time of liver biopsy in relation to an adverse liver-related outcome.

Kaplan-Meier analysis in 42 patients in the first cohort, divided according to hepatocyte p21 expression above the median [red line] or below the median [black line]) from the time of liver biopsy to outcome or censor point by log-rank test (p=0.03) [4a]. Kaplan-Meier analysis of 77 patients in the second cohort from the time of liver biopsy to outcome or censor point by log-rank test - divided according to p21 expression above the median [red line] or below the median [black line] (p<0.0001) [4b]; MELD score above the median [red line] or MELD score below the median [black line] (p=0.03) [4c]; alcohol consumption above the median [red line] or alcohol consumption below the median [black line] (p=0.03) [4d].

Figure 5. ROC curves comparing hepatocyte p21 expression alone, MELD score alone and hepatocyte p21 expression plus MELD score in predicting an adverse liver related outcome in patients with alcohol-related cirrhosis.

The area under the receiver operating characteristic (ROC) curves for hepatocyte p21 expression alone (0.74), MELD score alone (0.59) and hepatocyte p21 plus MELD score (0.76) are shown