This feature is central to the effective diagnosis of non-alcoholic fatty liver disease (NAFLD) and non-alcoholic steatohepatitis (NASH) and remains an essential component of staging scores, including the NAFLD activity score (NAS) and the steatosis, activity and fibrosis (SAF) scoring system. Despite its clear histological importance, assessing hepatocytic ballooning in day-to-day, ‘real life’ healthcare settings remains challenging.
This article aimed to investigate the mechanisms underlying hepatocytic ballooning and its associations with clinical phenotypes.
At the cellular and molecular levels, hepatocytic ballooning is caused by increased endoplasmic reticulum stress, rearrangements of the intermediate filament cytoskeleton, the appearance of Mallory-Denk bodies, and the activation of the sonic Hedgehog (SHh) pathway. SHh’s secretion by dying hepatocytes directly activates hepatic stellate cells, linked to liver fibrosis. The ballooning observed in NAFLD patients is often 1.5-2 times greater than normal size and indicates greater liver disease severity and associated risks of intra- and extra-hepatic complications. Recent research suggests that patients with hepatocytic ballooning are more likely to have overweight/obesity, dyslipidemia, abnormal glucose tolerance, insulin resistance, and elevated serum biomarkers indicating necroinflammation. Importantly, however, the identification of hepatocytic ballooning is hindered by liver biopsy preparation, digital imaging issues, and significant inter-observer variability. Artificial intelligence and machine learning tools represent potential remedies to these challenges but remain to be extensively validated experimentally.