A Comparison of Basal Insulin Delivery: Continuous subcutaneous insulin infusion versus glargine

Continuous subcutaneous insulin infusion (CSII) is believed to more closely mimic pancreatic function and therefore to create more stability in blood glucose than multiple daily insulin injections (MDIs). One of the unique features of CSII is the ability to preprogram changes in basal insulin delivery. This feature is especially useful during the night when insulin pharmacokinetics and the dawn phenomenon may change basal insulin requirements (1). Glargine is an insulin analog designed to mimic endogenous insulin secretion patterns and has been proposed as a longer-acting, peakless insulin that can be administered once a day, usually at bedtime (2). In the present study, glargine was compared, during nighttime hours, with preprogrammed rates of lispro delivery in patients using CSII.

A sample of 19 patients, already scheduled for continuous glucose monitoring system (CGMS; Medtronic Mini-Med, Northridge, CA) evaluation, were included in the study. CGMS data were compared between 11 subjects on CSII (lispro) and 8 subjects on MDIs (lipro and glargine) and were evaluated during the overnight period. Age (46.4 ± 10.5 vs. 40.4 ± 10.0 years), duration of diabetes (9.6 ± 5.3 vs. 15.5 ± 8.7 years), and HbA_1c (7.3 ± 0.6 vs. 7.1 ± 1.0%) were similar between the groups.

Significant differences in nighttime glucose control were identified as a function of type of therapy. Subjects treated with glargine spent significantly more time outside target sensor glucose ranges (70–200 mg/dl) than subjects treated with CSII (50.7 vs. 20.9%, P = 0.04). Specifically, subjects treated with glargine experienced a threefold increase in time exposed to glucose values <70 mg/dl when compared with subjects treated with CSII (34.7 vs. 12.8%, P = 0.03). Subjects treated with glargine spent twice the amount of time exposed to glucose values >200 mg/dl (16.0 vs. 8.1%, P = NS). The average hourly basal insulin rate was significantly greater in subjects treated with glargine than subjects treated with CSII (0.8 ± 0.4 vs. 0.5 ± 0.3 units · kg⁻¹ · h⁻¹, P = 0.002). The mean number of nocturnal basal rate settings used by the CSII group was 2.4 ± 0.7; changes in basal rate are not possible with glargine.

These results highlight the challenges in developing a basal insulin that works as effectively as CSII therapy. Statistically equivalent baseline HbA_1c between the two groups reflected a balance between the time spent with glucose values >200 mg/dl and <70 mg/dl in the glargine group and time spent within target sensor glucose values in the CSII group. Lepore et al. (3) report that “an ideal basal insulin candidate is a peakless, long-lasting preparation that mimics the flat interprandial insulin secretion of nondiabetic subjects, with reproducible subcutaneous absorption.” According to the ideal basal insulin requirements suggested by Lepore et al., glargine appears to be superior to other intermittent, or long-acting, insulins such as NPH, lente, and ultralente. However, our results suggest that while glargine may most closely meet the criteria set by Lepore et al., the ideal basal insulin must also be delivered at a variable rate to satisfy changing daily insulin requirements. CSII is the only current means of achieving this variable rate.