Therapeutic comparison of metformin and sulfonylurea, alone and in various combinations. A double-blind controlled...

To assess and compare the therapeutic efficacy and safety of metformin (M) and sulfonylurea (glyburide, G), alone and in various combinations, in patients with non-insulin-dependent diabetes mellitus (NIDDM).

Of 165 patients (fasting blood glucose [FBG] > or = 6.7 mmol/l) initially treated with diet alone, 144 (FBG still > or = 6.7 mmol/l) were randomized to double-blind, double-dummy controlled treatment with M, G, or primary combination therapy (MG). The dose was titrated, with FBG < 6.7 mmol/l as target, using, at most, six dose levels. The first three dose levels comprised increasing single-drug therapy (M or G) or primary combination at increasing but low dosage (MGL), and the second three levels were composed of various high-dose combinations, i.e., add-on therapy (M/G or G/M) and primary combination escalated to high dosage (MGH). Medication was maintained for 6 months after completed dose titration.

The FBG target was achieved in 9% of patients after diet alone. Single-drug therapy was insufficient in 36% and MGL in 25% (NS) of the randomized patients. There was further improvement in glucose control by the high-dose combinations. Mean FBG +/- SE was reduced (P = 0.001) from 9.1 +/- 0.4 to 7.0 +/- 0.2 mmol/l in those maintained on single-drug treatment or low-dose primary combination. Those treated with different high-dose combinations had a large mean FBG reduction, from 13.3 +/- 0.8 to 7.8 +/- 0.6 mmol/l. HbA1c levels showed corresponding reductions, and glycemic levels rose after drug discontinuation. Fasting C-peptide rose during treatment with G and MGL but not with M, while fasting insulin was not significantly changed. Meal-stimulated C-peptide and insulin levels were unchanged by M but increased by G and, to a lesser extent, by MGL. There were no significant insulin or C-peptide differences between the different high-dose combinations (M/G, G/M, and MGH). Body weight did not change following treatment with M or combination but increased by 2.8 +/- 0.7 kg following G alone. Blood pressure was unchanged. Overall effects on plasma lipids were small, with no significant differences between groups. Drug safety was satisfactory, even if the reporting of (usually modest) adverse events was high; the profile, but not the frequency, differed between groups.

Dose-effect titrated treatment with either metformin or glyburide promotes equal degrees of glycemic control. The former, but not the latter, is able to achieve this control without increasing body weight or hyperinsulinemia. Near-normal glycemia can be obtained by a combination of metformin and sulfonylurea, even in advanced NIDDM.