Cancer treatment should treat metabolism as a target as deliberately as it targets genes and proteins, argues Siddhartha Mukherjee.
Tumors rewire how they use nutrients, not just how they divide. Early 20th-century observations by Otto Warburg showed many cancers consume glucose heavily and convert it to lactate even when oxygen is present. That metabolic insight lost prominence as genetics rose to the fore, but metabolism remains central: tumors adapt to therapy by rerouting fuel use, and those shifts can enable resistance.
Mukherjee proposes “tumor-informed metabolism,” a precision approach that matches dietary and metabolic interventions to a patient’s tumor biology, the drug being used, and the patient’s physiology. Rather than generic “cancer diets” that advise all patients to cut sugar or adopt broad regimens, tumor-informed metabolism would be time-bound, measurable, and prescribed like a medication.
He illustrates the concept with a common clinical scenario. A woman in her 50s with hormone receptor–positive, HER2‑negative breast cancer and a PIK3CA mutation receives a PI3K inhibitor plus endocrine therapy. The drug initially controls disease, but over months the cancer progresses. Tests show elevated glucose and insulin—side effects of PI3K pathway inhibition that open a metabolic escape route. In that context, a dietary plan designed to blunt post-meal glucose and insulin spikes and timed around dosing can deepen and prolong the drug response, Mukherjee says. The benefit comes from combining drug and diet, not from either alone.
Metabolism interacts with tumor ecology and varies by site and context. Tumors in different organs rely on distinct fuels; therapies such as steroids or pathway inhibitors can alter host metabolism and thereby influence tumor behavior. Some cancers are vulnerable to depletion of specific nutrients: acute lymphoblastic leukemia responds to asparagine depletion, and dependency on serine, glycine, or methionine can affect sensitivity to treatment in breast, colorectal, and other cancers. These are mechanisms observed in cell biology and emerging clinical research.
Implementing tumor-informed metabolism would mean using food as information: prescribing macronutrient and micronutrient targets, timing meals to drug schedules, and monitoring metabolic biomarkers in real time. For some patients the goal may be to reduce insulin peaks during PI3K-pathway inhibition; for others it may be targeted amino acid restriction during chemotherapy or radiation. For patients at risk of wasting, the intervention may require precisely increased calories and protein to preserve strength and treatment tolerance.
Mukherjee and colleagues at Faeth Therapeutics say they have developed such regimens prospectively—pairing diets with PI3K/AKT/mTOR inhibitors in endometrial cancer and testing amino acid–restricted plans in rectal cancer—built around specific mechanisms and treatment windows. He urges other companies to design comparable, rigorously controlled nutrition–drug regimens and submit them to randomized trials. He also argues that sponsors and regulators should treat diet as a prespecified element of clinical protocols rather than an uncontrolled background variable.
The evidence base is early. Preclinical and some clinical studies show that pathway-directed drugs combined with insulin-lowering or amino acid–modifying diets can enhance efficacy, but results have been mixed when diets are generic, prolonged, or disconnected from drug mechanisms. Weight loss and weakness from poorly managed nutrition can harm patients and undermine therapy, underscoring the need for targeted, physician-directed interventions.
To make tumor-informed metabolism part of standard care, Mukherjee says, studies must be prospective and controlled, with objective endpoints such as response rates, survival, preserved dose intensity, and reduced toxicity. Clinical infrastructure is also required: multimodal regimens that shut down oncogenic signaling without collapsing immune responses, precision nutrition delivered as therapy, and computational “metabolic operating systems” to predict metabolic flux and resistance pathways in silico before patient implementation.
Skepticism is expected. Critics may argue that metabolic plasticity makes it too difficult to trap tumors or that metabolic contributions will be marginal. Mukherjee counters that oncology already succeeds with combination strategies that act through different mechanisms; even modest extensions of drug efficacy that preserve meaningful treatment cycles have real value to patients. He also warns that precision metabolic therapies must be equitable—covered and accessible, adapted to diverse cultures and kitchens rather than reserved for concierge care.
The argument reframes nutrition from lifestyle advice to a targeted adjuvant intended to close metabolic escape routes and enhance pharmacologic mechanisms. If integrated with drugs, sequencing, and ongoing supportive care, tumor-informed metabolism could add a new dimension to combination therapy: drug plus metabolism, rather than drug plus drug alone.
Siddhartha Mukherjee is a physician, researcher, and author. He has co-founded biotechnology and health care companies including Faeth Therapeutics and Manas AI, is an associate professor of medicine at Columbia University, and practices as an oncologist at Columbia University Medical Center. A new edition of his book The Emperor of All Maladies includes four new chapters.
Leave a Reply