Diabetes drug could help tackle rare form of leukaemia

Cambridge scientists have found that a common diabetes drug might help prevent the blood cancer called acute myeloid leukaemia or AML. The study - which has been published in Nature - found that metformin slows the growth of cells that carry a key mutation linked to the cancer. This mutation is present in some bone marrow cells for years before the cancer manifests; it seems to confer a growth advantage over healthy cells, meaning affected cells slowly increase in number in the bone marrow before some of them eventually become cancerous and trigger AML. Metformin robs these pre-cancerous cells of their growth advantage, reducing the risk that they’ll turn nasty. Brian Huntly is head of the department of haematology at the University of Cambridge and one of the team behind the study…

Brian - So we've looked at the most common mutation that's associated with this pre-malignant condition. This is a condition called clonal hematopoiesis. It's very prevalent, it increases with age and people are at an increased risk of developing acute myeloid leukaemia. But not all of them develop acute myeloid leukaemia. In fact, the majority do not. The most common mutation in around about half of them is a mutation in a gene which gives cells their identity. We have looked at identifying vulnerabilities that cells that carry this mutation have, which are not present within normal cells, and many of them affect energy production within the cell. And these cells actually have an increased ability versus normal cells, and they increase in number within the bone marrow, which is the factory that produces the blood cells, and they come to be more dominant. And this is the process that leads towards the development of acute myeloid leukaemia. So if we can identify specific vulnerabilities within these cells, we can try and halt this process of expansion, and we can hopefully avert or delay the development of acute myeloid leukaemia.

Chris - And having spotted some of these vulnerabilities, have you got a way to block this?

Brian - Yes, we do. The abnormalities that we found in terms of the genes that cause the vulnerabilities are common in metabolic disorders of humans, things like diabetes mellitus, for example. And we have shown experimentally that we can prevent this expansion process with the use of a drug called metformin, which is the most common drug used for type 2 diabetes, and is probably used in hundreds of millions of patients worldwide with a pretty acceptable safety record. We're now thinking about a step change. We're thinking about treating people before they become patients.

Chris - This is a bit like treating high blood pressure in some respects, isn't it? You're giving people a pill. They haven't got a disease yet. They've just got high blood pressure. It hasn't caused a problem, but we know it will one day. So we give them an antihypertensive, bring down the blood pressure. We know we're reducing the risk of the consequences of high blood pressure. You're saying, I can give you a drug for a disease you haven't got yet, and it will reduce the risk of disease manifesting, and that's what metformin is doing?

Brian - That's what we think. That's what we hope. The experimental evidence shows that. We've backed that up by looking at patient registries where individuals who were taking metformin were at a significantly lower risk of developing this condition. So our experiments and the human evidence suggest that metformin will retard this process.

Chris - How is it working? What's the metformin doing that stops the disease manifesting?

Brian - We think it's targeting the energy production within the cells, and metformin does a number of things, but it targets this protein complex called Complex 1, which is involved in generating energy. But I think it's more complicated than this. Metformin does other things. It causes a lot of intracellular signaling differences, and the suggestion is that this pathway is implicated also.

Chris - Now, obviously, there are limits to what you can learn from mouse experiments, but when you do this in mice, what sort of a difference does the metformin make?

Brian - So basically, in the experiments, we look at the ability of blood stem cells to repopulate the blood system in animals who've been given lethal irradiation. Normally, the cells that carry this mutation have a significant advantage. This is almost completely reversed by the metformin treatment.

Chris - Most patients who come to see you, though, don't come and see you when they've just got some mutations but no disease yet. You pick up most people when they've already got the blood cancer, the AML problem. Does this mean, then, that now we need a screening programme, really, to try and pick this up, or is there a way we can get at the right people soon enough for them to benefit from what you've discovered and be put on metformin early enough to change the course of their disease?

Brian - So you identify a really important problem, and I'll go back to your analogy of high blood pressure. We don't know that people have high blood pressure just by looking at them. It either has to be some sort of test. Now, currently, the way that we test for this mutation is by complicated genetic screening. That is really not cost effective within the population. We often pick these patients up serendipitously. They have blood tests for either another cancer. They have an abnormality of blood counts which is transient. We pick them up. We're trying to develop low-cost screening tools to try and identify patients somewhat earlier. We do know that these disorders are more prevalent in some populations, patients with metabolic disorders, for example, patients with chronic inflammatory disorders. So we may be able to identify some patient populations in which screening would be manageable and likely to give a return on significant amounts of investment. But yes, we do have to come up, I think, with sensible screening strategies to identify the patients.