Cancer treatment is returning to a fundamental question: what tumors eat.
After decades focused on genetics, oncologists are re-examining tumor metabolism. Early in the 20th century Otto Warburg observed that many cancer cells consume large amounts of glucose and convert it to lactate even when oxygen is available. That metabolic signature, once overshadowed by genomic research, now appears central to how tumors adapt to therapy by rewiring fuel use.
Advocates propose a shift to tumor-informed metabolism: tailoring dietary and metabolic interventions to a tumor’s biology, the drugs a patient receives, and the host’s physiology. Rather than generic “cancer diets” that uniformly advise patients to cut sugar or carbs, clinicians would prescribe nutrition as part of a therapeutic regimen designed to complement specific pharmacologic mechanisms.
A typical scenario illustrates the concept. A woman in her 50s with hormone receptor–positive, HER2‑negative breast cancer carrying a PIK3CA mutation receives a PI3K inhibitor plus endocrine therapy. Initial response gives way to progression after months. Testing shows drug-induced rises in glucose and insulin, which can open metabolic escape routes for the tumor. Pairing the drug with a timed dietary plan—reducing refined carbohydrates and sweetened drinks, favoring slow-digesting carbohydrates, proteins and fats, and monitoring real-time metabolic biomarkers—can blunt insulin peaks, deepen response and extend remission. The intervention is intended to augment the drug, not replace it.
Tumor metabolism varies by tissue and context. Some cancers are vulnerable to depletion of specific amino acids: asparagine depletion is effective in acute lymphoblastic leukemia; serine and glycine can be critical for certain breast and colorectal cancers; methionine restriction alters one-carbon metabolism and can sensitize tumors to therapy. Tumors also respond differently depending on the organ microenvironment and concurrent therapies: a PI3K inhibitor can raise insulin, steroids can induce hyperglycemia, and both can change which fuels the tumor relies on.
Proponents argue that nutrition should be prescribed with the same precision and safety as medication: macronutrient and micronutrient targets, timing relative to dosing, monitoring for weight loss and muscle wasting, and safeguards for metabolic health. Early work pairing insulin-lowering or amino-acid–modifying diets with pathway-directed drugs has shown promising signals in preclinical models and small clinical studies, but generic or prolonged diets untied to a drug’s mechanism have failed or caused harm.
Siddhartha Mukherjee and colleagues at Faeth Therapeutics have designed prospective regimens that pair diets with PI3K/AKT/mTOR inhibitors in endometrial cancer and implement amino acid–restricted diets in rectal cancer, scripting plans around specific mechanisms and treatment windows. Mukherjee and other advocates call for more companies to develop rigorous nutrition–drug regimens and for sponsors and regulators to treat diet as a prespecified element of clinical protocols rather than an unmeasured background variable.
To integrate tumor-informed metabolism into standard care will require prospective, controlled trials anchored to drug mechanisms and objective endpoints: response rates, survival, preserved dose intensity and reduced toxicity. Clinical infrastructure must support multimodal regimens that shut down signaling without undermining immune response, precision nutrition delivered as therapy rather than lifestyle advice, and computational “metabolic operating systems” to model flux, predict resistance and test combinations in silico before patient deployment.
Equity is a key concern: metabolic interventions must be accessible, covered by payers, and adaptable to diverse cultural and socioeconomic contexts to avoid becoming concierge services available only to a few.
Metabolism is a mutable but actionable layer of tumor biology. When tightly integrated with pharmacology, nutrition can narrow metabolic escape routes, preserve treatment intensity and potentially extend benefit. Incorporating diet as a targeted co-therapy will require rigor, infrastructure and attention to access, but proponents argue it is a necessary expansion of what counts as treatment in modern oncology.
Siddhartha Mukherjee, M.D., D.Phil., is an oncologist and associate professor of medicine at Columbia University, a co-founder of Faeth Therapeutics and other biotechnology ventures, and the author of The Emperor of All Maladies, now available in a new edition.
Leave a Reply