Recently we had the good fortune to chat with Dr. Eugene Fine, professor at the Albert Einstein College of Medicine and leader of the RECHARGE pilot study looking at how low carb diets may be helpful to cancer patients.
Low carb or ketogenic diets for cancer may not be news to many already practicing in the Functional Medicine space, but Dr. Fine highlighted some interesting points around the roles of insulin, ketone bodies, glucose, and ROS, as well as how low carb diets can integrate with conventional therapies.
We also discussed the variability in responsiveness to low carb diets in cancer therapy, including underlying metabolic dysregulations that may make tumor cells more resistant to the anti-cancer effects of ketone bodies.
[Dr. Fitzgerald] Can you give us a bit of background as to the work you’ve been doing in the area of low carbohydrate or ketogenic diets for cancer?
[Dr. Fine] Sure. This has to do both with the clinical pilot study in 10 patients that we’ve done but also cell culture work with my colleague Richard Feinman and that’s something that has been progressing in a very good direction. In that area we have demonstrated some very specific mechanisms, somewhat indirect evidence so far but we are also in the process of obtaining more direct evidence, about how ketone bodies inhibit both ATP generation and cell growth in cancer cell lines, while having no such effects on normal cells.
We think this could have some significance as far as cancers are concerned. It should be noted that many cancers have mutations that activate signaling pathways that lead to growth, proliferation, resistance to apoptosis, etc. These mutations cause constitutive activation of these pathways meaning that insulin inhibition via diet won’t slow down these cancerous behaviors. The importance of the ketosis aspect – the independent effect that insulin inhibition has on the liver, and the fact that ketosis also inhibits cancer cells, provides an alternative route by which the cancers may be inhibited.
There are drugs targeting cancers by inhibiting the downstream activity of insulin, which would appear to indicate that there would also be some value in inhibiting insulin. But that’s my overview as to why I think the diet has value, not just through generalized insulin signaling pathways within cancer cells.
In our pilot trial we had adults, who were fairly established in their ways, and asked them to shop for different foods to enable them to become ketotic. Considering this, the extent of insulin inhibition that we got in the human trial, was really not bad at all. As far as ketosis, anything over 0.5 mmolar ketone body concentration is considered ketosis, and participants averaged about 1 mmol among the group. One patient reached 5 mmolar.
Blood sugar fell, but not by a whole lot. I’m not sure about the exact role of inhibiting glucose in cancer. Most cancers that we’re talking about express the GLUT1 transporter, so the Michaelis constant – a biochemical measure of how avidly cells take up glucose – is relevant. Even on a ketogenic diet, we don’t really drop glucose levels that much. Say from 95 to 75, still within the normal range.
If I look up the Michaelis constant of GLUT1 in the literature, I find values all over the page in from 0.5 mmol up to 26 mmol, so I don’t really know what the number means, except that it’s usually quoted as 1 mmolar, or 18 mg/dl.. So it’s possible if certain cancer cells have a much higher Michaelis constant then perhaps reducing blood glucose even within the normal range of 3-5 mM can have a reduced supply effect. I don’t think anyone has proof that this is what’s going on.
But I must agree that I do think there’s some value in reducing glucose spikes after meals, which could then reduce the additional ROS spikes not just in cancer cells but in adipose tissue. Adipose responds to this with increase production of toxic adipokines and cytokines which circulate in the body and can have deleterious effects. So we don’t really know how blood glucose fits exactly into the picture mechanistically to cause cancer. To this point it remains a very interesting speculation.
[Dr. Fitzgerald] Did you do any calorie restriction as part of your trial?
[Dr. Fine] Calorie restriction is a little tangential to what I study, but I have looked at the work of Valter Longo. It’s not really clear how to interpret our work in contrast to his. In our own 10 patients, all of the patients had about a 35% calorie restriction from their baseline diet, which is exactly the extent of restriction that Longo and others have proposed is inhibitory for cancers. We didn’t see cancer inhibition from 35% reduced calories. We did see better outcomes from higher levels of ketosis, only.
[Dr. Fitzgerald] Did you have the participants reduce protein as part of the program, or just reduce carbohydrate and increase fat?
[Dr. Fine] No, we were trying to make the diet as simple for them as possible. We felt that one intervention alone was going to be difficult enough for adults with habitual food preferences. If we had been able to supply them with all the food then maybe we would have done that. If we were to do it again, we probably would. But I’m not really sure that existing data on protein supports that it is tumorigenic. It may be true in mice. But mouse metabolism is quite different from human. Observational data have been cited to show association of protein consumption with cancer incidence, but associations don’t imply causation; at least they don’t without a strong mechanism that can be assigned. I don’t think we have that.
Some groups advocate that high protein kicks you out of ketosis. I’m not sure that it does that. At least I haven’t seen the evidence that it does.
The other thing is that IGF-1, which is supposedly modulated by protein intake, is also modulated by insulin because insulin modulates the IGF binding proteins. So IGF-1 behavior in the face of protein uptake is more complex than we may realize.
Circling back to calorie restriction. Calorie restriction is extremely poorly defined. It depends on what your starting diet is. Most papers that I’ve found that employ calorie restriction are actually reducing carbohydrates.
The whole idea that calorie restriction is a separate mechanism is somewhat dubious to me anyway because I don’t understand the mechanism? There’s no calorie receptor. With carb restriction we have epidemiologic evidence of hyperinsulinemia, hyperglycemia and obesity all contributing to increased incidence of cancer, and all are specific insulin-related effects. And we have some limited evidence that carb restriction may have a role in treatment of cancer.
So, while calorie restriction may end up with a role, it’s not obvious to me right now how those data fit.
Fasting, however, may be a different situation. It certainly has insulin effects but it may have other effects as well – such as cell cycle arrest, recalibrating the cell cycle.
[Dr. Fitzgerald] There’s been some research in the past about the connection between calorie restriction and physiological changes such as lowering thyroid hormone. Do you think that there is any plausible connection there?
[Dr. Fine] If you look at thyroxine, that is T4, it is not actually reduced. What does happen if I’m not mistaken is that T3, the biologically more active form, is reduced. I’m not sure if that is clearly validated, but I would have to go and look further. I believe that the same effects are seen in ketogenic diets as well.
[Dr. Fitzgerald] One of the angles we are considering is whether it makes sense to ‘pulse’ a ketogenic diet for cancer patients to attempt to prevent cancer cells becoming keto-adapted themselves. What are your thoughts on that?
[Dr. Fine] There are lots of things that are plausible, but hard to say when we haven’t done research on this specifically. To me it makes some sense. That’s not the same thing as saying this is proven.
In our pilot study, 4 out of 10 patients did get worse. These were the ones that had the least ketosis, and in retrospect had evidence of prediabetes before beginning. So I don’t know if that is something that predisposes to resistance to ketogenic diet.
But the 4 people that had worst progression of disease were also the heaviest with BMIs 28 – 31. Three of these subjects had FBG of 100 or more, suggesting that these patients had evidence of pre-diabetes. So maybe that was why they had worse outcomes. And the lack of response to the ketogenic diet was possibly predicated on that as well. Because their baseline ketosis was higher.
In diabetes, baseline ketone bodies are higher. If you have insulin resistance the body could be starting to generate some ketone bodies, right. We can think of pre-diabetes or insulin resistance as ineffective insulin action. So tumors perhaps have already adapted to higher levels of basal ketones. So that makes sense to me, but I don’t feel that the data we’ve generated is strong enough to recommend it to anyone. Though perhaps if you had evidence of prediabetes you wouldn’t do it until evidence became stronger that treatment might work for you too.
[Dr. Fitzgerald] What kinds of cancer did the four non-responders (those who got worse) have?
[Dr. Fine] In our study we took participants with cancers that were demonstrably positive on PET scans. We didn’t pick by tissue type, rather by phenotype – i.e. all glucose-avid cancers. Two had breast cancer, three had colorectal. One ovarian, one fallopian tube cancer. The four non-responders had breast, colon, ovarian and lung cancers. So, even in this small group, the cancer tissue types were very similar.
What I envision, which is purely speculative, is that cancers might get keto-adapted after prolonged dietary ketosis. But this is possibly a rationale for pulsing, possibly in combination with standard therapies which usually involve rounds of 4-6 weeks. Meanwhile, eating protein can be hard while on chemotherapy due to nausea. So alternatively, diet might be pulsed in between chemotherapy rounds. Longo’s research has shown that fasting before or after chemo reduces chemo side effects, but this might also apply to how to use the ketogenic diet in between chemotherapy.
In our group of dietary responders, one patient had remission, four had their disease stabilized. But these, again, were adults buying and preparing their own foods and they had a pretty reasonable 1 mmol ketosis. So if they could have a pre-prepared ketogenic diet they could probably get up to 3-5 mmol. That might have a better outcome and I hope we get the chance to try this.
I don’t know that a ketogenic diet that by itself can cure cancer, but it may well improve cancer control and maybe it can be coupled with other kinds of therapies. The result could be that you could lower the dose of the toxic therapies. Lowering toxicity and improving efficacy could result in something more meaningful, extending life and with a higher quality of life. I think these are all reasonable goals.