Addition of Sulfonylurea to Insulin Treatment in Poorly Controlled Type II Diabetes

Abstract

This study examined the potential beneficial effects of the addition of a second-generation sulfonylurea to insulin therapy for poorly controlled type II diabetes. A randomized, double-blind, crossover experimental design was utilized in 16 type II diabetic patients for a period of eight months. Treatment with glyburide, 20 mg/d (plus insulin), compared with placebo (plus insulin) resulted in a significant reduction in mean basal glucose (232 +/- 12 vs 262 +/- 11 mg/dL [12.8 vs 14.4 mmol/L]) and hemoglobin A1C (10.2% +/- 0.5% vs 10.9% +/- 03%) concentrations. Concomitant with this change, basal C-peptide and free insulin values increased with glyburide therapy, but this pharmacological agent did not alter the ability of the patient's erythrocytes to bind insulin. We conclude that in type II diabetic subjects receiving more than 28 units of insulin per day, the addition of glyburide results in a marginal, but statistically significant improvement in basal glucose concentration, but not in glucose tolerance as assessed by integrated glucose concentration. Whether this small improvement in glycemia is worth the additional cost of sulfonylureas or the risk of drug side effects is not known.

Is Combination Sulfonylurea and Insulin Therapy Useful in NIDDM Patients?: A metaanalysis

“…One study (27,28) showed that urinary C-peptide before treatment (but not C-peptide response to sulfonylureas) predicted ability to discontinue insulin while maintaining euglycemia. In three studies, obesity predicted response (16,19,31), but in one study it did not (17). Degree of hyperglycemia (worse initial control) was positively correlated with response in two studies (16,31) but not in another (18).…”

“…As pointed out above, only 1 study achieved HbAj^ <8. Four studies measured posttreatment free insulins (16,17,24,26) and did not find significant differences between the groups. No studies measured portal insulins.…”

“…Some individuals have large responses to combination therapy, whereas others have none. This was dramatically illustrated by Schade et al (17) and Riddle et al (29). Clinically, it would be very useful to be able to choose combination therapy on the basis of known predictive patient characteristics.…”

“…(A list of excluded articles is available from the authors on request.) (14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32). Study size ranged from 9 to 64 subjects.…”

“…Another study used a fasting glucose of <7.77 mM (140 mg/dl) at three visits and a 25% reduction in total daily insulin dose at some time during the study (30). Others used any decrease over baseline (17,18). In two studies, instead of defining response at a particular level, correlations were performed between prognostic factors and outcomes (16,22).…”