|
||||||
|
|
|
|
|||||||||||
|
|||||||||||||
| ||||||||||||||||||||||||||||||||||||
Clifford Rosen, MD, has just written a commentary in the New England Journal of Medicine about the issues surrounding one of the year's most controversial medical stories, rosiglitazone.1 Dr. Rosen, who chaired the Food and Drug Administration (FDA) committee on rosiglitazone that met on July 30, writes that the regulatory pathway for new drug approvals for type 2 diabetes should require clinical outcomes (hard end points such as reduction of heart attack, stroke, or kidney failure), as opposed to the surrogate outcomes we now use. Currently, for a drug to get an indication for diabetes, it is required to significantly reduce blood glucose levels. The metric used is glycated hemoglobin (A1C), but traditionally the FDA also has looked at fasting blood glucose as the other end point.
|
According to Dr. Rosen, rosiglitazone seems to have won the battle but lost the war. For example, in the study, A Diabetes Outcome Progression Trial (ADOPT), rosiglitazone was superior to glyburide and metformin in durability (long-term effectiveness), but it caused more weight gain and edema compared to both agents and more cardiovascular events (including heart failure) than glyburide.2 ADOPT was a study with durability as the primary end point, yet cardiovascular outcomes also were noted and, in this case, it was not protective compared to glyburide.
Dr. Rosen argues that setting and using these types of cardiovascular end points for outcomes for new diabetes drugs would not create a precedent. For example, new agents for the treatment of osteoporosis now need to show improvements in fractures, not a fracture surrogate such as bone density.
So this brings us to an important crossroads. For a new drug, should we wait to have improvement in macrovascular (or microvascular) outcomes before approval? What would the implications be to our current system?
Let's think about this. In 1998, the U.K. Prospective Diabetes Study (UKPDS), a landmark analysis of type 2 diabetes control, showed that microvascular outcomes could be improved with insulin (available since 1922), sulfonylureas (available since the 1950s), and metformin (available in the U.S. since 1995, yet available in other parts of the world since the late 1950s).3
Many have noted—and the current American Diabetes Association (ADA) Consensus Treatment Algorithm is based on the premise—that these drugs, in addition to the thiazolidinediones (TZDs), are the only medications required because they have known outcomes and because potential dangers always lurk with new agents.4 (With the controversy about rosiglitazone this year, a new recommendation from ADA about TZDs is certainly possible in the future.) But as a medical community, do we want to prove that our agents to treat type 2 diabetes improve heart attack and stroke risk before approval?
I think not.
Since insulin's discovery, 76 years were required to prove that it improves outcomes in type 2 diabetes. It took approximately 40 years to demonstrate the same with sulfonylureas and metformin. The UKPDS suggested metformin may have a macrovascular benefit but, importantly, metformin is generally weight-neutral, while the other two drugs result in an increase of weight.5
We currently have no evidence alpha-glucosidase inhibitors, TZDs, GLP-1 receptor agonists, dipeptidyl peptidase-4 (DPP-4) inhibitors, and amylinomimetics improve macrovascular outcomes. What is clear, in general, is that meticulous control of hyperglycemia benefits microvascular end points more than macrovascular end points, yet we don't even have those outcomes with these agents. Was the FDA wrong to approve these drugs without more evidence? Are we wrong to prescribe them?
Of course not. What we do need, however, is more rigorous follow-up data regarding safety and long-term end points. Exactly how that would be accomplished remains to be determined. But whatever method is chosen, it must be much easier and cheaper than requiring the pharmaceutical industry, on its own, to perform studies that may take decades. Even for microvascular outcomes, the cost of doing a UKPDS-like study would make new drug development prohibitively expensive.
We also need to emphasize other end points, the most important of which is weight in this obese population. Weight neutrality, such as seen with metformin or sitagliptin (Januvia, Merck), or weight reduction, as noted with exenatide (Byetta, Amylin/Eli Lilly), certainly makes these agents more attractive than agents resulting in weight gain. Obviously, for any given patient, the agent of choice needs to be individualized; among our options, for example, insulin still seems to be underused in cases in which insulin secretion becomes severely compromised.
Perhaps what we truly require is a system in which each end point is evaluated based on its relative importance. Given our current choice of agents, safety should be the first consideration, followed by efficacy, weight change, and then possibly other surrogate end points such as lipoprotein levels and inflammatory markers. Ideally, we would ensure that an agent diminished the risk of retinopathy or myocardial infarction before it was approved, but that's not practical given the time it takes to develop these complications.
At first glance, asking government to assist us in determining long-term outcomes seems absurd. Yet that's how we've done it in the past. The Diabetes Control and Complications Trial (DCCT) was funded by the National Institutes of Health (NIH) and industry partners, and the UKPDS was funded by the U.K. and industry. (Interestingly, this British study also received funding from the American NIH.) Indeed, government support in collaboration with industry can be noted for many recent and ongoing diabetes studies, including Diabetes Reduction Assessment with Ramipril and Rosiglitazone Medication (DREAM), Diabetes Prevention Program (DPP), Diabetes Prevention Trial—Type 1 (DPT-1), Bypass Angioplasty Revascularization Investigation (BARI) 2 Diabetes, Veterans Affairs Diabetes Trial (VADT), and Action to Control Cardiovascular Risk in Diabetes (ACCORD), to name a few. My concern is that putting the full burden of evidence on industry will potentially reduce diabetes drug development, if not halt it entirely.
My other big concern with the current environment for new-drug evaluations involves comparators. For oral agents, I see little use for a placebo-controlled trial if the end point is A1C. Furthermore, for prandial insulin, the comparator needs to be a rapid-acting insulin analogue, not regular insulin. And I'm tired of seeing studies comparing new long-acting analogues with neutral protamine Hagedorn (NPH) insulin. The appropriate comparator is a long-acting analogue. For insulin, we need to do a better job of comparing new analogues with our existing insulin analogues, not standard insulins that have been shown to cause more hypoglycemia.6
The silver lining of the rosiglitazone controversy is that it is forcing us
to review our current system, which clearly is in need of reform. I am eager
to see what changes will result. 
Footnotes
|
References
2. Kahn SE, Haffner SM, Heise R, et al.: Glycemic durability of
rosiglitazone, metformin, or glyburide therapy. N Engl J
Med 355:2427–2443, 2006.
3. UKPDS Group: Intensive blood glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet 352: 837–853, 1998.[Medline]
4. Nathan DM, Buse JB, Davidson MB, et al.: Management of
hyperglycemia in type 2 diabetes: A consensus algorithm for the initiation and
adjustment of therapy: A consensus statement from the American Diabetes
Association and the European Association for the Study of Diabetes.
Diabetes Care 29:1963–1972, 2006.
5. UKPDS Group: Effect of intensive bloodglucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). Lancet 352:854–865, 1998.[Medline]
6. Hirsch IB: Insulin analogues. N Engl J Med 352: 174–183, 2005.
CiteULike 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
| ||||||||||||||||||||||||||||||||||||
|
||||||
|
| DOC News | Diabetes | Diabetes Care | Clinical Diabetes | Diabetes Spectrum |
