Glycemic variability trial sets stage to study effects on diabetes complications
BOSTON — When added to basal insulin and metformin, exenatide and prandial insulin yielded similar HbA1c levels, but exenatide maintained more stable glucose levels, according to research presented here.
“Our ultimate goal would be to carry out a long-term effectiveness trial to determine if the impact of ‘glipulin’ (a glucagon-like peptide-1 receptor agonist added to insulin) would be important in both micro- and macrovascular outcomes,” Jeffrey L. Probstfield, MD, said during a presentation of data.
In a complex proof-of-concept trial dubbed FLAT SUGAR, Probstfield and Irl B. Hirsch, MD, both of the University of Washington School of Medicine, and colleagues screened 255 patients with type 2 diabetes requiring insulin and identified 146 with an HbA1c level between 7.5% and 8.5% to participate in the study run-in period. Of these, 122 patients were stabilized on basal-bolus insulin with metformin with HbA1c levels between 6.7% and 8% in the 2 weeks preceding baseline, and then were randomly assigned to continue that regimen (n = 50) or start basal insulin with metformin plus the GLP-1 agonist exenatide (glipulin; n = 52). On-trial HbA1c target was 6.7% to 8%.
Irl B. Hirsch
Participants used Holter monitoring and masked continuous glucose monitoring (CGM), and metabolic markers of cardiovascular risk were evaluated at baseline and weeks 11 through 13 and 24 through 26. Researchers compared changes in the coefficients of variation of glucose for the two groups from baseline to 26 weeks; 92 participants finished the study.
At randomization, both groups had an HbA1c of 7.9 + 0.3%. Coefficients of variation in CGM were 31.9 + 6.3 for the glipulin group vs. 30.3 + 6.2 for the basal-bolus insulin group. At 26 weeks, the groups maintained similar HbA1c levels (7.1 + 0.6% for the glipulin group vs. 7.2 + 0.6% for the basal-bolus insulin group), but differences were observed in mean changes in CGM coefficients of variation (–2.4 + 7.9 for the glipulin group vs. 0.4 + 5.5 for the basal-bolus insulin group; t test P = .047; rank sum P = .024).
The glipulin group saw improvement in mean amplitude of glycemic excursion (rank sum P = .049), but other markers remained unchanged. Neither group experienced any severe hypoglycemia. The glipulin group had a 4.76 kg reduction in mean weight, whereas the basal-bolus insulin group had a 0.69 kg weight gain (P < .001).
“We had a positive result for our primary endpoint that is significant reduction of coefficients of variability on CGM with nearly equivalent HbA1c levels at the final visit,” Hirsch said. “We showed no clear difference in hypoglycemia, although you saw a trend for reduction in hypoglycemia, and there was also a significant reduction in weight. We also saw a significant decrease in [alanine transaminase] and [serum amyloid A].”
Although all of the tools are available to reduce glycemic variability in practice, “what we don’t know is: What is the clinical impact of lowering the variability,” Hirsch told Endocrine Today.
“The fundamental take-home message is we can randomize patients with type 2 diabetes and keep the HbA1c the same but give them different degrees of glucose variability,” Hirsch said. “So now, hopefully, we can do a subsequent trial to see how important is glucose variability for the development of diabetes complications.” – by Jill Rollet
Hirsch IB, et al. Abstract 385-OR. Presented at: American Diabetes Association’s 75th Scientific Sessions; June 5-9, 2015; Boston.
Disclosure: Hirsch reports relationships with Abbott Diabetes Care, Novo Nordisk, Roche Diagnostics, Sanofi USA and Valeritas. Probstfield reports relationships with AstraZeneca, Bristol-Myers Squibb, Janssen and Sanofi.