MAY 21, 2025, NEW YORK – A Ludwig Cancer Research study has uncovered a surprising link between diet, intestinal microbes and the efficacy of cancer therapy.
Led by Ludwig Princeton's Asael Roichman and Branch Director Joshua Rabinowitz , the study could help explain why drugs known as PI3 kinase (PI3K) inhibitors—which disrupt an abnormally activated biochemical signaling pathway that spurs cancer cell proliferation—haven't led to consistent, durable cancer control in patients with solid tumors.
"Many cancer drugs don't work equally well for all patients, and one emerging possibility is that diet plays a role in this variability," said Rabinowitz. "We found in this study that diet can indeed alter cancer treatment outcomes in preclinical models and can do so in an unexpected way, unrelated to its immediate nutritional effects. It turns out that certain small molecules in plant-based foods are transformed in mice by commensal gut bacteria into compounds that activate the liver to clear PI3K inhibitors more quickly, lowering the efficacy of the drug."
"While we focused in this study on PI3K inhibitors, the liver enzymes involved in clearing these drugs break down many others as well," added Roichman, a postdoc in the Rabinowitz lab and lead author of the study. "This suggests our findings could be of relevance to multiple classes of drugs used to treat cancer and other diseases."
The study, reported in the current issue of Cell, began with a surprising result obtained from an experiment examining the interplay between diet and cancer therapy from an entirely different perspective.
The Rabinowitz lab and others have shown that ketogenic diets dramatically enhance responses to cancer drugs in preclinical mouse models of cancer. Such diets are rich in fat and very low in carbohydrates like starch and sugar, and their enhancement of therapeutic responses was thought to be linked to their tendency to lower insulin and blood sugar levels. The researchers were therefore surprised when mice fed certain high-carbohydrate diets—which should raise blood sugar and spike insulin production—responded well to PI3K inhibitors in their experiment.
Digging into this unexpected result, they found that the keto diet's enhancement of responses to PI3K inhibitors had little to do with carbs, fat, blood sugar or insulin. Rather, experiments revealed that the key determinant was the molecular complexity of the diet—whether it was made of 'whole foods' versus highly processed ones.
The ketogenic food consumed by mice in preclinical studies is a highly processed formulation lacking the complex mix of plant-derived chemicals (phytochemicals), especially from legumes and soy, that are present in standard chow. It turned out that gut microbes break down phytochemicals, namely soyasaponins derived from soybeans, into molecules that induce the expression of a detoxifying liver enzyme, cytochrome P450.
Experiments revealed that elevated production of these hepatic enzymes in the chow-fed mice led to rapid clearance of PI3K inhibitors, reducing the anti-cancer efficacy of the regimen. In line with these findings, the researchers demonstrated that a high-carbohydrate but low-phytochemical diet—as well as antibiotics that suppressed the gut microbiome—enhanced PI3K inhibitor activity in the mice.
"These findings suggest that some plant-based diets, through their interactions with gut microbes, may lower cancer drug exposure by ramping up the body's drug clearance systems," said Roichman. "While the specific molecules that exert such an influence may differ in humans, our work highlights diet and the microbiome as key factors that can shape how cancer drugs behave in the body."
Further, the findings open opportunities to develop new strategies for cancer therapy that take into account such factors as a patient's diet, microbiome composition and recent use of antibiotics, which alter the ecosystem of commensal bacteria. Informed by additional research, such strategies could include analyses of patient microbiomes and the prescription of dietary changes and pharmaceutical interventions to modulate the metabolism of cancer therapies.
Aside from his post as Director of the Princeton Branch of the Ludwig Institute for Cancer Research, Rabinowitz is Professor in the Department of Chemistry & Lewis-Sigler Institute for Integrative Genomics and Member of the Rutgers Cancer Institute.
This study was supported by the Ludwig Institute for Cancer Research, Stand Up 2 Cancer, the U.S. National Institutes of Health and the New Jersey Commission on Cancer Research.
