Antidiabetic medications and mortality reduction: a shift from surrogate to clinical endpoints

It was widely believed that antidiabetic medications should improve the macrovascular (such as mortality and myocardial infarction) as well as microvascular endpoints (such as retinopathy and nephropathy) by virtue of improved glycemic control and consequently a reduction of glycated haemoglobin (HbA1C). Therefore, the United States Food and Drug Administration (US FDA) initially considered the reduction of HbA1c as the basis of approval of antidiabetic medications. HbA1c is a surrogate endpoint where it acts as a marker for diabetes control and as for every other endpoint, it is important to demonstrate that it actually correlates with the clinical endpoints and patient safety.

As data accumulated it became clear that a reduction in HbA1c does not necessarily lead to a reduction in macrovascular endpoints such as mortality, albeit it does consistently improve the microvascular endpoints for the most part. ¹ A reduction in blood sugar does not necessarily equate to a reduction in cardiovascular outcomes, the most common cause of morbidity and mortality in diabetes. On the other hand, studies have also shown that effective monitoring and glycemic control of diabetic patients in the early course of the disease results in a significant decrease in mortality and cardiovascular risks. ²

The two trials, the Action to Control Cardio-vascular Risk in Diabetes (ACCORD) and the Action in Diabetes and Cardiovascular Disease: Preterax and Diamicron Modified Release Controlled Evaluation (ADVANCE) were meant to study this hypothesis, but their results came out to be unquestionably negative. ACCORD was prematurely terminated 17 months before the due date, by virtue of increased mortality from cardiovascular diseases and other causes,^3,4 and ADVANCE did not show any cardiovascular benefit to the intensively controlled group. ³ More troublesome was the fact that some drugs that have been known to reduce HbA1c actually increased macrovascular endpoints. For example, rosiglitazone used in the historical trials reduces HbA1c and improves lipid profiles, but has been paradoxically associated with an increased risk of myocardial infarctions and mortality. ⁵ In addition, both trials reported frequent occurrences of hypoglycemia in the intensively controlled groups.^3,4

However, not all is lost. Recently several medications that reduce HbA1c such as the sodium glucose co-transporter (SGLT2) inhibitors, empagliflozin (EMPA REG outcomes trial) and glucagon-like peptide (GLP)-1 agonists, liraglutide and semaglutide have improved mortality and cardiovascular risk factors in type 2 diabetic patients.^6,7 Furthermore, metformin has proven to have beneficial effects on cardiovascular diseases. ⁸ While the use of insulin is well tolerated, it does not seem to have a significant impact on macrovascular complications and mortality. ⁶ The treatment with dipeptidyl peptidase DPP-4 inhibitors saxagliptin, sitagliptin and alogliptin is questionable as they are linked to increased rates of hospitalizations due to cardiovascular complications. ⁶ The GLP-1 agonist, lixisenatide has revealed neutral effects on cardiovascular risks. ⁹

The US FDA has moved from surrogate endpoints and now mandates that these agents should at least demonstrate safety of macrovascular events, if not the reduction of mortality as the basis of approval for antidiabetic medications in the contemporary era. Clinicians should be educated for the importance of macrovascular endpoints while reading clinical trials and counselling patients instead of surrogate endpoints such as HbA1c. What remains a dilemma is the usage of hard endpoint data by the pharmaceutical companies in the development phases of antidiabetic medication, (specifically phase I and II trials), when no other data are available.