Will HbA1c remain a valid surrogate endpoint?
The incidence of T2D is increasing globally and despite all clinical efforts cardiovascular disease remains the leading cause of mortality in this population. It seems therefore warrantable to expect that novel treatments for T2D should not only demonstrate efficacy and safety in achieving blood glucose lowering but even more so to clarify the benefit and risk regarding CV safety and outcomes.
Implication of recent CVOT results in diabetes.
While HbA1c is well accepted as a surrogate for assessment of diabetic treatment efficacy (and reduction of microvascular complications on the long-term), there is, however, lack of clear evidence for the causal relationship between glycemic control and the risk of future CV events. It has thus been debated whether glycemic-targeted surrogate endpoints (i.e. HbA1c lowering) hold true as valid proxies for clinically meaningful outcomes. This debate was further fueled by significant changes in drug regulatory assessment of CV risk and benefits of novel glucose-lowering therapies following the rosiglitazone controversy. The 2008 guidelines by the FDA now request hard clinical outcome studies to exclude detrimental cardiovascular effects and to establish CV safety of novel anti-diabetes drugs.
With several studies following this guidance another debate emerged whether such attempts should solely focus on CV safety or rather aim to also establish potential CV effects. The EMPA-REG OUTCOME® is an example for the latter. This study was designed as an outcome trial with sufficient power to establish both, safety and CV efficacy. The study demonstrated that empagliflozin significantly reduced the primary endpoint of 3-MACE, which was largely driven by a reduction in risk of CV death. Notably in this trial, CV effects of emapgliflozin were largely independent of glucose lowering- and HbA1c changes.
Given the complexity of the micro- and macrovascular disease biology in T2D, a sole focus on an HbA1c surrogate may not adequately capture the holistic risk profile relevant for long-term clinical outcomes. While a glucose-lowering intervention may positively impact one biomarker, it may negatively affect another component of the complex (cardio)vascular pathophysiology. An attempt to overcome future uncertainness in this regard may be to recommend that any anti-diabetes agent needs to establish a convincing cardiovascular benefit.
Dr. Maximilian von Eynatten, Boehringer Ingelheim Pharma GmbH & Co. KG, Germany