The NEJM study randomized over 600 patient who were at high risk for developing diabetes to pioglitazone (Actos) or placebo and found that after about 2.5 years, 2.1% in the pioglitazone group and 7.6% in the placebo group progressed to diabetes (which is a relative 72% risk reduction). On the down side, patients on pioglitazone gained about 7 pounds more and about 7% more had edema.
To assess the value of using TZD's like Actos to prevent diabetes, it is important to understand what diabetes really is, how it is defined, and why TZD's might be a very important option despite the associated risks. Unlike type 1 diabetes which is mostly about the destruction of the beta cells in the pancreas and subsequent lack of insulin, type 2 diabetes is a long process that is about much more than blood sugar. Type 2 diabetes is a result of genetic factors that control how individuals utilize glucose, about age, and about obesity. These factors combine to cause much more than just the body's impaired ability to utilize sugar. There is increased inflammation, worsening of cholesterol, increased blood pressure and increased blood clotting, just to name a few. All of these factors, including elevated sugar, combine over years and years to increase the risk of microvascular complications (eye, kidney, nerves) and macrovascular complications (heart attack and stroke). The process occurs for many years, and it is estimated that at the time of diagnosis of diabetes, the process (let's call it metabolic syndrome) has been going on for about a decade and that about half of the pancreas' function is lost despite having relatively high levels of insulin. Men who meet the diagnostic criteria for metabolic syndrome have three times the risk of heart attack and stroke and women who meet this criteria have six times the risk.
When we call a diabetic a diabetic is quite arbitrary. The guidelines used to classify diabetes as a fasting glucose of 140. This was lowered several years ago to 126 recognizing that complications were occurring at sugar levels lower than 140, and the cut off of 140 was too high because it was delaying treatment. More recently, in 2010 the ADA recommended that diabetes be diagnosed at a hemoglobin A1c of 6.5% or greater, recognizing a fasting glucose of 126 missed many of the diabetics with impaired post-prandial glucose, thus leading to possible delays in therapy. In fact, based on this new criteria, some of the patients in the Actos study who were not classified as having diabetes, would now be called diabetics. It is not inconceivable in the near future, that the threshold for diagnosis and treatment of diabetes occurs at an even lower fasting glucose number, A1c, or some constellation of markers. In other words, whether or not we are really preventing diabetes or just delaying the time to reach an arbitrary threshold is more of semantics. The bottom line is that in patients with impaired fasting glucose, there is an underlying disease process leading to complications that ought to be addressed.
To make an informed decision, one has to review the three main studies that looked at drugs for the "prevention" of diabetes. The study above for Actos, was very similar in design and findings to the other TZD rosiglitazone or Avandia called the DREAM study. The other major study, called the Diabetes Prevention Program (or DPP) looked at metformin vs. diet and exercise vs. placebo. Interestingly, diet and exercise beat metformin for preventing diabetes. Why would this be? It has to do with the fact that diet and exercise reverses the process of insulin resistance whereas metformin merely lowers blood sugar. Metformin was not included in the rosi or pioglitazone studies, nor was there a diet and exercise arm, so it is hard to compare all three interventions (diet/exercise, metformin, TZD) in preventing diabetes complared to placebo. However, troglitazone, an early TZD which was pulled from the market for liver complications (Actos and Avandia have not shown these problems) was initially used in the Diabetes Prevention Program (DPP), but was stopped after less than a year. Interestingly, in that short time, troglitazone did better than both metformin and diet and exercise. In other words, when you address insulin resistance and metabolic syndrome either by diet and exercise or with a TZD, you prevent diabetes much better than with a drug like metformin that only addresses blood sugar or placebo. This is important because the Actos study not only showed decreased rates of diabetes, but also showed statistically significant decreases in blood pressure, plaque build up, and increases in HDL or good cholesterol.
This begs the question, if you can prevent diabetes (or treat the underlying process that has not yet met the arbitrary criteria to be called diabetes) with Actos or lifestyle modications, why would you choose a drug? The answer is that lifestyle modifications is always preferred, but often not practical or easy to do. For the lifestyle intervention group in the DPP, participants were instructed to limit their calories to 1200-1800, get 150 minutes of exercise a week, had 16 sessions of counselling, and access to nutritionists, personal trainers, and behavioral counsellors. Another way to look at this is that anyone can go on "The Biggest Loser" and lose weight if they are given that amount of support (and time to take off from work), but this is not always practical. Interestingly, much of this support is not covered by traditional health insurance (where medication is), and once could argue that this is where we ought to devote our precious health care dollars (topic for another post).
As a physician, I need to help my patient the best way I can. I need to be practical. For all patients, diet and exercise is clearly the first step and always encouraged at every step. However, practically speaking, this just doesn't work all the time. It takes a highly motivated patient with a lot of resources and support to do this. Thus, if I can use a medication that will help reverse the underlying disease process of insulin resistance and delay the diagnosis of overt diabetes (along with diet and exercise), then I believe I am ethically obligated to do so.
Finally, the big issue that has been brought up is side effects. Mainly weight gain and heart failure. First, the TZD's do not "cause" heart failure. What I mean by this is that Actos has no direct effect on the heart. What is does do is increase fluid retention. For patients whose hearts are not working that great (pre-heart failure), a little extra fluid can push them into heart failure. Though this is a serious risk, it is not that common and can be addressed by carefully monitoring patients before and after treatment. Secondly, weight gain is a real issue. There are two components of weight gain caused by TZD's. The first is the aforementioned fluid retention. The second has to do with reversing the underlying disease process. Patients with diabetes and metabolic syndrome do not utilize glucose correctly. This causes subsequent increases in insulin and eventual pancreas failure leading to the need (over years) of supplemental insulin. By not using sugar, weight is not gained. By correcting the process, the sugar goes to where it is supposed to, which leads to weight gain. However, it is not clear that this weight gain is necessarily "bad." Studies show that TZD treated patients shift their fat from the dangerous visceral abdominal fat (associated with high cardiovascular risk) to more centralized fat stores. In other words, though not proven by a randomized control trial, reversing the disease process (glucose, cholesterol, blood pressure, weight gain) is likely to prevent more adverse events (heart attack and stroke) then events caused by additional weight.
Bottom line: Obesity is an epidemic in our country and will soon be the single leading cause of preventable death in the US. Along with obesity comes diabetes, which takes years to develop and is defined by arbitrary criteria. This diease process (metabolic syndrome) is associated with it's own consequences. Diet and exercise leading to subsequent weight loss and improved cardiovascular health is clearly the best choice. However, for most patients for very practical reasons, this method is not successful. Until there are changes to our health care system and/or public health initiatives that make intense lifestyle modifications more reasonable, pharmacotherapy to prevent diabetes has an important role. Pioglitazone has demonstrated that it can effectively prevent diabetes in most patients with known but manageable side effects, and therefore should be considered a useful tool.