A new study led by the Shalek Lab at the Ragon Institute, published in Cell, shows that liver cells facing prolonged metabolic stress—like that seen in steatotic liver disease—activate cancer-associated programs long before any tumors appear. Beyond genetic mutations, these early cellular changes may also explain why some patients progress to liver cancer.

The team tracked how liver cells (hepatocytes) respond to high-fat diets and metabolic stress over time, using single-cell multi-omic technologies in mice and spatial -omics in human cohorts. They found that stressed hepatocytes turn up survival and developmental programs while dialing down their normal metabolic functions. Strikingly, these same shifts were present in human patients with steatotic liver disease and predicted worse survival in those who developed liver cancer.

Among the key regulators identified was HMGCS2, an enzyme critical for processing fats into ketone bodies. When the researchers knocked out HMGCS2 in mouse livers and combined this with a high-fat diet, hepatocytes showed accelerated stress responses and formed significantly more tumors after cancer-driving mutations were introduced. In human patients, lower HMGCS2 levels—even in liver tissue before tumorigenesis—predicted higher cancer risk up to 15 years later.

Using a new computational tool that they created called MATCHA, the team also identified transcription factors SOX4 and RELB as drivers of stress-induced reprogramming. When SOX4 was overexpressed in mouse livers, it pushed hepatocytes toward stressed, proliferative states—driving and accelerating happens naturally over months of dietary stress.

The findings suggest that the liver’s early coping mechanisms for metabolic overload inadvertently create fertile ground for cancer. Understanding these pre-cancer cellular states could open new windows for early intervention in patients with steatotic liver disease, which affects more than one-third of adults worldwide.