Study Finds High-Fat Diets Drive Breast Cancer Growth and Invasion

In a new study, Princeton researchers demonstrate that high-fat diets promote invasive growth in triple negative breast cancer using a 3D microfluidic tumor model. Their work, published in APL Bioengineering, shows how fatty acids and cholesterol trigger structural changes that make tumors more aggressive. 

The team grew hundreds of 3D tumors over several years and perfused them with human-like plasma mimicking various diets. Most nutrient conditions, including high insulin, glycerol, and ketones, produced compact tumors similar to baseline. High-fat conditions with fatty acids and cholesterol led to small, hollow appendages extending outward, a hallmark of invasion. 

“That’s where the name cancer comes from, crab-like,” said Celeste Nelson, the study’s principal investigator. “Aggressive cancers have these tendrils, and it’s the leading edges that end up invading into our normal tissues and making it into either a lymphatic or a blood vessel and escaping and metastasizing.”

High-fat tumors grew at similar rates to others but underwent structural remodeling. Cells migrated from the core to the edges, accompanied by elevated MMP1 gene expression, which breaks down collagen. MMP1 levels strongly correlated with invasive changes, suggesting high-fat diets enable environmental degradation that fosters aggressiveness, though causation remains unproven. 

Future work may test tumor growth under high-fat conditions with inhibited MMP1 activity. Nelson noted these findings apply specifically to triple negative breast cancer and may not extend to other types. The study opens paths to explore diet-prognosis links and therapy targets.

A ketogenic diet simulation surprised the team. Model tumors fed high-fat, low-carbohydrate nutrients showed no health benefits over baseline. “We were expecting a ketogenic diet to be protective,” Nelson said. “Yet we didn’t see that here. And it tells us a few possible things. One is that, for this particular type of cancer, maybe a ketogenic diet could be protective, but it operates through other cells that we don’t have in this particular model.”

The 3D microfluidic models balance realism and control. They replicate tumor geometry, tissue properties, and fluid chemistry better than 2D cultures, which use stiff simple media, or animal models, which add unmanageable complexity. This middle ground isolates diet effects effectively, revealing how ketogenic benefits may act through absent environmental factors. 

“Every tumor is an individual’s tumor,” Nelson said. “How do you know when you have enough different tumor models to represent the patient population? Maybe that’s not feasible.”