Cancer in the Cold

 BY DEREK LOWE

Here's an idea that you probably hadn't thought of! Let's take it through the conceptual steps. It's already known that many tumors (especially solid ones) rely more on glycolysis for their energy needs than normal tissues do. That makes sense, because glycolysis is an anaerobic pathway to make ATP (as opposed to the usual Krebs-cycle way, the citric acid cycle and oxidative phosphorylation), and solid tumors are often comparatively oxygen-poor once you get past their outer layers. This is often known as the Warburg effect, for Otto Warburg who first noticed the phenomenon (and who picked up a Nobel for it in 1931). And like any notable differences between tumor cells and regular ones, there have been attempts to exploit this metabolic change for cancer therapy, but that's been difficult to realize. It can be tricky to throw a wrench into glycolysis without messing up energy homeostasis in general, for one thing, and some of the more indirect attempts have not had clinical success. As that last linked paper outlines, even the reasons behind the Warburg effect are not completely clear - the low-oxygen story makes sense, sure, but there are tumors that don't seem to be oxygen-deprived that also show it.

Now let's switch over to another longtime medical hypothesis, in a different field. Type II diabetes is a growing problem all around the world, and its hallmark is the development of insulin resistance and thus inappropirately high circulating glucose levels. There are a lot of ways to address the disease therapeutically. Fixing the underlying insulin resistance would be nice, for starters, but we don't actually understand the biochemical mechanisms behind it enough to do that directly yet. Metformin is probably the closest thing, and it has several other beneficial effects as well, but we don't really understand its mechanism(s) of action either. After that, lowering glucose levels through other means isn't a bad option, and that's why you see things like SGLT2 inhibitors, which prevent re-absorption of glucose in the kidneys.

One idea along those lines that's been looked at for many years is the possibility of activating "brown fat" (or at least a brown-fat-like phenotype) in adipose tissue. Brown fat takes up large quantities of glucose and converts it to body heat via "non-shivering thermogenesis". There's a unique "uncoupling protein" involved that alters the mitochondrial membane, and this allows glucose to get oxidized (with the release of heat) without really producing much ATP. It's just burned off, basically, which would be a nifty process to engage in people with high glucose levels for starters, and could even help with weight loss if it were sufficiently active (and if you could do that without causing people to walk around sweating constantly, which remains to be seen).

That link will take you to some previous attempts to rev up brown fat oxidative pathways pharmacologically, which so far have not worked out. But we know how to get them going the old-fashioned way: sit in a cold environment. Thermogenesis, both shivering and non-shivering, will kick in automatically. What the paper linked to in the first sentence of this post shows is that putting tumor-bearing animals into these cold conditions slows the growth of their tumors. The effects on glucose metabolism brought on by non-shivering thermogenesis seem to land right in the middle of the Warburg effect pathways. It's kind of an obvious idea when you come at it this way, but a lot of things look obvious after someone else has tried them!

The authors demonstrate substantial inhibition of tumor growth in mice living at 4C versus more normal temperatures, and this effect holds across a whole range of tumor types. This happens in xenograft models as well as animals that are genetically prone to spontaneously form tumors of their own. This wasn't just an effect of being directly exposed to the cold on the skin surface, either - hepatocarcinomas deep inside the mice responded the same way. The brown adipose tissue in these mice was definitely activated, and the white adipose tissue showed a "browning" phenotype as well, but it appears that the presence of the tumors themselves did not affect these processes - rather, it was the other way around. And surgically removing the brown adipose tissue almost completely abolished the tumor growth effects, as did genetic deletion of that key uncoupling protein, UCP1. This looks like a pretty solid story.

What's more, the paper demonstrates (in a single patient with Hodgkin's lymphoma) that exposure to even mildly cool temperatures for several days (22C) both activating brown fat tissue and caused significantly decreased glucose uptake in the lymphoma tissue itself (as measured by radiolabled glucose uptake in PET imaging - this Warburg-increased glucose uptake is a common method for detecting such tumors). The experiments with mice showed that 22C exposure did not seem to have an effect in them (you had to go colder), but the two species do have different metabolic rates and different activation of brown fat). I'm still surprised that this temperature worked in this human subject, but it's quite encouraging that you might not have to go to 4C!

As the authors say, "This therapeutic approach is simple, cost-effective and feasible in almost all hospitals and even at home, and is most likely omnipresent for all cancer types" and they note that these effects are equipotent to those seen for a great many approved chemotherapies. That slower-growth effect will need to be demonstrated in human patients, of course, but everything does seem to point in that direction. You can fill in the experimental possibilities as well as I can: does going below 22C help even more? What levels of exposure to cold are effective? Can this effect be combined with other chemotherapies for even greater efficacy? I would guess that we're going to see these ideas and more put to the test rather quickly, and one can only hope for good results.

https://www.science.org/content/blog-post/cancer-cold