ADA 2011: Assessment of Glycemic Variability and CD40 Ligand in the STAR 3 Study

Patients treated with sensor-augmented pump therapy may achieve lower glycemic variability than patients given multiple daily insulin injections.

A poster presented by John B. Welsh, MD, PhD, at the ADA 71st Annual Meeting in San Diego, showed how researchers in the STAR 3 study evaluated changes in A1C and soluble CD40 ligand by comparing multiple daily insulin injections (MDI) with sensor-augmented pump therapy (SAP), which provides patients with an alternative to daily insulin injections. CD40 ligand is a potential biomarker of oxidative stress, and studies have shown that it is elevated in type 1 diabetic neuropathy, acute insulin-induced hypoglycemia, diabetic patients with painless myocardial infarction, and type 1 diabetes and microangiopathy.

A total of 485 participants, ages 7-70, were assigned to either SAP (n=244) or MDI (n=241). The SAP group used the MiniMed Paradigm REAL-Time System with insulin aspart, and the MDI group used pens with insulin glargine and insulin aspart. Participants must have used MDI for at least 3 months prior and have had fewer than 2 severe hypoglycemic episodes in the year prior to enrollment. Baseline A1C ranged from 7.4% to 9.5%. Treatment groups were classified according to 12-month A1C levels: <6.5%, 6.5-7%, 7-<8%, and ≥8%. Week-long continuous glucose monitoring studies were conducted, and CD40 ligand was collected in blood samples at baseline and at 12 months. Coefficient of variations (CV) and standard deviations (SD) were used to summarize glycemic variability. In addition to measuring CD40 ligand in both groups, correlations were measured between CD40 ligand and BMI, A1C, hypoglycemic exposure, and glycemic variability.

After 52 weeks, results showed that mean sensor glucose values at A1C levels ≥6.5% were similar between the SAP and MDI groups. However, even at comparable A1C levels, participants on SAP therapy had lower glycemic variability than participants on MDI, with the greatest difference in both CV and SD observed in the <6.5% A1C group. The overall between-group difference was statistically significant, p<.01. With regard to CD40 ligand, mean levels rose over the course of the study by 50.43 pg/ml in the MDI group and fell by 32.92 pg/ml in the SAP groups. This difference was not significant, however. With regard to CD40 ligand, results showed no correlations with BMI, A1C, hypoglycemic exposure, or glycemic variability at 52 weeks.

“There are two take-home messages,” said Welsh. Despite prior studies showing CD40 ligand as a biomarker for diabetes, he is not convinced that it is a good biomarker for measuring sugar stability. “The other message is more reassuring,” he continued, “which is that even if A1C is under control, patients get a smoother ride of sugar with the pump compared to insulin injections.” That is, there is less variability in glucose. “Extreme spikes in glucose can be scary for patients.” Therefore the pump may offer incremental benefits beyond A1C reduction, he added.

Welsh described how the pump had two main components—one to deliver insulin and one to detect glucose levels. This prompted a question regarding whether pumps will ever be automated to deliver insulin when it detects unusual levels of glucose. He cited another poster presenting findings from a European study from one such pump. However, the pump is currently not available in the US.