Acarbose in ambulatory treatment of non-insulin-dependent diabetes mellitus associated to imminent sulfonylurea failure: A randomised-multicentric trial in primary health-care

To assess the efficacy and safety of acarbose as an adjunct to high sulfonylurea (SU) doses in patients with imminent SU failure, a randomised, multicentric, 6 month double-blind, parallel and placebo-controlled trial was performed in primary healthcare. Entry criteria were: NIDDM patients in concomitant dietary follow-up, age > 40 year-old, more than 3 years of diagnosed diabetes, baseline HbA(1c) levels between 8-12% (N: 4-6%), stable body mass index < 35 kg m(-2) and glibenclamide daily dose > 10 mg. After 1 month placebo run-in period all patients were randomly allocated into two groups of treatment (acarbose 100 mg t.i.d. vs placebo). HbA(1c) levels, the main efficacy variable, lipid profile, fasting and postprandial blood glucose levels were performed and adverse events were also recorded. A total number of 65 patients were randomised, 36 in acarbose and 29 in a placebo group. No statistical differences were found on age (60.2/61.7 year-old), BMI (28.7/27.4 kg m(-2)), glibenclamide dose (14.5/14.0 mg/day) and baseline HbA(1c) (9.0/8.8%). Acarbose-treated patients significantly reduced HbA(1c) levels (9.0/7.9 vs 8.8/8.5%; P < 0.01), based upon a marked decrease, but statistically not significant, in mean postprandial plasma glucose levels (11.9/9.6 vs 12.4/11.1 mmol l(-1)). No significant differences between fasting plasma glucose and lipid profile were detected. A total of 31 patients (47.7%) reported adverse events, 20 (55.5%) and 11 (37.9%) in acarbose and placebo treatment group respectively. Relationship with drug was estimated as possible or probable in 16 (44.4%) of acarbose-treated patients. None of them were excluded from study participation due to insulin requirement. Only seven patients (10.7%), six with acarbose (16.6%) and one with placebo (3.8%), withdrew the study because of the adverse events. Thus, acarbose seems to be a useful option in order to improve HbA(1c) levels in non-insulin-dependent diabetes mellitus with imminent sulfonylurea failure. (C) 1997 Elsevier Science Ireland Ltd.