A selfie from the Tripping over the Truth Conference 2017, prior to my book launch in 2018 How to Starve Cancer. Professor Tom Seyfried, Travis Christofferson and me after a heated discussion about what is actually ‘doable’ in the clinic and gaps I had spotted, - you see we are still smiling!
It is an elegant model.
Glucose is the door most people understand. Many cancers are highly glucose avid. Lowering glucose and insulin signalling through carbohydrate restriction, fasting, calorie restriction, or a carefully supervised ketogenic approach may place real metabolic pressure on those tumours.
But glucose is not the only door.
Many cancers can also use glutamine: an abundant amino acid that supports carbon metabolism, nitrogen handling, redox balance, nucleotide synthesis, and survival under metabolic stress. The Seyfried argument is that fatty acids and ketones cannot replace glucose or glutamine as fermentable fuels. Close both doors, and a cancer cell may be unable to sustain itself metabolically.
That is the theory.
The question we have to ask, particularly when discussing pancreatic cancer, AML, and other metabolically flexible cancers, is not whether the theory is intelligent. It is. The question is whether we can reliably, safely, and selectively carry it out in real people. The glucose door is only partly closable
We can reduce blood glucose. We can improve insulin sensitivity. We can use carbohydrate restriction, fasting windows, exercise, sleep, and in some settings therapeutic nutritional ketosis.
But we cannot reduce glucose to zero.
The body must maintain blood glucose for tissues with an obligatory or substantial glucose requirement, and it can make glucose through gluconeogenesis. Stress hormones, infection, steroid medicines, chemotherapy, sleep disruption, inflammation, liver dysfunction, insulin resistance, and cachexia can all complicate glucose control.
This does not make glucose reduction pointless. Far from it. It means we should be accurate about what we are doing: applying pressure to a dominant fuel pathway, not literally starving every cancer cell of glucose.
That distinction matters.
The more difficult issue is glutamine.
It is easy to say “block glutamine.” It is much harder to do this in a patient without also damaging normal tissue. Glutamine is not merely a cancer nutrient. It is important to the gut, immune cells, bone marrow, wound healing, and normal stress responses.
That is why broad glutamine-antagonist strategies have historically encountered dose-limiting toxicity. The problem is not that glutamine is unimportant to cancer. The problem is that it is important to many healthy cells too.
Newer approaches are trying to solve this through more selective glutaminase inhibition, tumour-targeted prodrugs, rational drug combinations, and carefully timed pulses rather than continuous blockade. But that is precisely the point: we do not yet possess a simple, validated, universally safe way of “closing the glutamine door” in patients.
And in cancers such as AML, where the immune system and bone marrow are already under pressure, this becomes even more delicate. A strategy designed to metabolically weaken leukaemia must not simultaneously undermine normal haematopoiesis or the immune system needed to clear damaged malignant cells.
This is where the ketone debate becomes more nuanced.
A tumour that can up-regulate ketogenesis, ketolysis, fatty-acid oxidation, lipid-droplet metabolism, or fatty-acid transporters such as CD36 and members of the FATP family is showing metabolic adaptability. It is not proof that ketones are “bad”. Nor is it proof that dietary ketosis is irrelevant.
It is evidence that the tumour is learning how to survive under pressure.
When someone says that a ketone-utilising tumour cannot exploit ketones if glucose and glutamine have been closed off, that may be true in some experimental settings. But in clinical practice, those doors are rarely completely closed.
Glucose is lowered, not abolished. Glutamine blockade remains difficult, toxic, and imperfect. Tumours may use endogenous production, autophagy, stromal support, macropinocytosis, lipid stores, circulating fatty acids, altered redox pathways, or metabolic exchange with neighbouring cells.
Cancer does not need every door open. Sometimes it only needs one open door, at the right time, to survive the pressure we apply.
There is a second challenge that metabolic discussions too often glide past: the host.
Pancreatic cancer is a particular concern because patients may already have weight loss, malabsorption, pancreatic exocrine insufficiency, sarcopenia, systemic inflammation, and cachexia. An intervention that improves glucose control but accelerates loss of lean mass, worsens food tolerance, or reduces resilience to chemotherapy may not be a therapeutic success.
This is not an argument against metabolic therapy. It is an argument against treating metabolic pressure as though it exists separately from nutrition, muscle, immunity, quality of life, and the ability to complete effective treatment.
The patient’s metabolism is not simply the battlefield. It is also the reserve force.
For some people, a highly restrictive approach may be useful for a defined window, carefully measured and combined with treatment. For others, the same intervention may be mistimed, too aggressive, or clinically unsafe.
This is the central issue.
The question is not:
“Can cancer use glucose, glutamine, fat, or ketones?”
The question is:
“In this tumour, in this patient, at this point in treatment, which metabolic intervention amplifies the therapies being used without depleting the host?”
A ketogenic intervention plus chemotherapy is not evidence for ketosis alone. It is evidence for a combination.
A fasting pulse around therapy is not the same intervention as continuous calorie restriction.
Glutamine targeting may have a different meaning alongside a cytotoxic drug, radiation, hyperbaric oxygen, immunotherapy, venetoclax, a hypomethylating agent, or a ferroptosis-inducing strategy.
Timing is not a minor detail. It may be the treatment.
Seyfried’s model remains one of the most coherent and important frameworks in metabolic oncology. It correctly forces us to look beyond mutations alone and to ask what fuels a cancer cell needs to survive.
But “close glucose and glutamine, and cancer cannot use fat” is not yet a clinically completed sentence.
It is a powerful hypothesis. It is a research programme. It is a rationale for intelligently designed combination trials.
It is not yet a guarantee that a high-ketone or high-fat systemic environment is safe for every ketone-capable tumour, particularly where glutamine is not effectively controlled, the tumour has demonstrated lipid flexibility, or the patient is vulnerable to wasting.
That does not mean we should abandon metabolic therapy.
It means we should apply it with enough precision to distinguish:
Lowering glucose and insulin signalling from imposing chronic deep ketosis.
Short, timed fasting or ketogenic pulses from permanent restriction.
A tumour’s theoretical fuel options from its demonstrated in-vivo behaviour.
The metabolic needs of the tumour from the nutritional needs of the patient.
An elegant mechanism from a validated clinical protocol.
The future is not “keto versus anti-keto.”
The future is measuring the tumour, protecting the patient, combining therapies intelligently, and finding out which doors must be closed – and for how long – before a cancer truly has nowhere left to go.
To read my previous posts on keto and to read my ‘doable’ posts This is Part 1A and this is Part 1B. Enjoy!
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.