Effect of Acarbose on. Carbohydrate and Lipid. Metabolism in NIDDM Patients. Poorly Controlled by Sulfonylureas. Gerald M. Reaven, MD. Claude K. Lardinois ...
Effect of Acarbose on Carbohydrate and Lipid Metabolism in NIDDM Patients Poorly Controlled by Sulfonylureas
The ability of acarbose to lower plasma glucose concentration was studied in 12 patients with noninsulin-dependent diabetes mellitus (NIDDM) who were poorly controlled by diet plus sulfonylurea drugs. Patients were studied before and 3 mo after the addition of acarbose to their treatment program, and a significant improvement in glycemic control was noted. Although the decrease in fasting plasma glucose concentration was modest (12.0 ± 0.8 to 10.8 ± 0.3 mM), average postprandial plasma glucose concentration decreased by 3.4 mM. When acarbose therapy was discontinued in 5 patients, plasma glucose levels rapidly returned toward pretreatment levels. In addition to the improvement in glycemia, acarbose treatment also led to a significant reduction in HbA1c (7.4 ± 0.2 to 6.4 ± 0.2%, P < 0.01) and triglyceride (2.4 ± 0.1 to 2.1 ± 0.1 mM, P < 0.01) concentrations. Neither the plasma insulin response to meals nor insulin-stimulated glucose uptake improved with acarbose therapy, consistent with the view that acarbose improves glycemic control by delaying glucose absorption. Considerable individual variation was noted in the response to acarbose, and the results in 4 patients were dramatic, with striking reductions in both fasting and postprandial glucose concentrations. The addition of acarbose to patients with NIDDM not well controlled by sulfonylureas appears to have significant clinical benefit. Diabetes Care 13 (Suppl. 3):32-36, 1990
lthough both weight loss and use of sulfonylurea drugs are effective in the treatment of patients with non-insulin-dependent diabetes mellitus (NIDDM), their use does not guarantee that glycemic control will be universally achieved. If therapeutic efficacy does not result from the use of diet plus sulfonylurea drugs, there are only a few alternative approaches available to the clinician. In this context,
A
32
Gerald M. Reaven, MD Claude K. Lardinois, MD Michael S. Greenfield, MD Herbert C. Schwartz, MD Hendrick ]. Vreman, PhD
the possibility that patients with NIDDM, poorly controlled by diet plus sulfonylurea drugs, might benefit from the administration of acarbose seemed worthy of evaluation. Acarbose is an intestinal a-glucosidase inhibitor with excellent affinity for sucrase, glucoamylase, and maltase that has been shown to reduce postprandial plasma glucose levels after carbohydrate ingestion in healthy volunteers (1,2). Based on these observations, an evaluation was undertaken to determine whether acarbose administration might be useful in the treatment of a particularly troublesome group of patients with NIDDM, i.e., overweight individuals with diabetes not well controlled by diet plus sulfonylurea therapy.
RESEARCH DESIGN AND METHODS Studies were conducted in 12 sulfonylurea-treated patients with NIDDM. None of the patients were considered to be under adequate glycemic control with diet plus maximal sulfonylurea treatment, and the mean ± SE fasting plasma glucose level before initiation of acarbose therapy was 12.0 ± 0.8 mM. Mean age of the study population was 62 yr (range 51-74 yr), and mean relative body weight was 122% (range 102-135%). Patients were admitted to the Stanford General Clinical Research Center (Stanford, CA) after informed consent had been obtained, and they were maintained on an isocaloric diet consisting of 20% protein, 40% carbo-
From the Department of Medicine, Stanford University School of Medicine, and Geriatric Research, Education and Clinical Center, Veterans Administration Medical Center, Palo Alto, California. Address correspondence and reprint requests to C M . Reaven, MD, GRECC (182-B), Veterans Administration Medical Center, 3801 Miranda Avenue, Palo Alto, CA 94304.
DIABETES CARE, VOL. 13, SUPPL. 3, AUGUST 1990
C M . REAVEN AND ASSOCIATES
hydrate, and 40% fat as percent of calories. The study did not include a placebo control group. Daily calorie intake was divided into portions of 20, 40, and 40% given at 0800, 1200, and 1800, respectively. This dietary program was followed throughout the study. Subjects were hospitalized for 7 days before the initiation of acarbose treatment. During this time, they were maintained on their usual sulfonylurea treatment program, and a series of baseline measurements were made. After the initial studies, acarbose was added to the treatment program at an initial dose of 50 mg with each meal. Patients were then discharged from the hospital and seen at weekly intervals, and the acarbose dosage was gradually increased to 150 mg at each meal. Most patients noted increased flatulence and softer stools at some time during the study, but the maximum dose of acarbose was reached in all patients without producing excessive gastrointestinal symptoms. All patients were admitted to the hospital for the second time ~3 mo after initiation of acarbose treatment. During this second hospitalization, all baseline measurements were repeated. Finally, five patients were studied for a third time ~6 wk after discontinuation of acarbose. Weight remained within 1 kg of initial value throughout the study in all patients. To measure fasting glucose and lipid concentrations, four blood samples were obtained at 0800 during each study period for determination of fasting plasma glucose (3), triglyceride (4), cholesterol (5), and high-density lipoprotein (HDL) cholesterol (6) concentrations. To better assess carbohydrate tolerance, the plasma glucose response to the 1200 meal was determined on four occasions, twice during each study period. This meal contained 40% of the day's total calories, consisted of 20% protein, 40% fat, and 40% carbohydrate, and was identical on every occasion. The morning dose of acarbose was given as usual the morning of the meal tolerance test. Blood was obtained at 1200 (before the 1200
meal), 1230, 1300, 1400, and 1500 for measurement of plasma glucose and insulin concentration (7). In vivo insulin action was estimated by the glucoseclamp technique. This approach has been described in detail in previous publications from our laboratory (8,9) and is based on determination of the amount of glucose utilized during 120 min of sustained hyperinsulinemia. Insulin was given initially as an intravenous bolus and then at a constant rate to achieve steady-state plasma insulin levels of —718 pM during the 120 min. Glucose was administered at a variable infusion rate in amounts that maintain basal plasma glucose concentration. The glucose infusion rate, when corrected for urinary glucose loss and residual hepatic glucose output (8,9), provides a direct estimate of the amount of glucose that must be metabolized (M) to maintain the basal plasma glucose concentration. By dividing M by the basal plasma glucose concentration, it is possible to obtain the glucose metabolic clearance rate (MCR), which permits comparison of insulin-stimulated glucose uptake rate of patients with varying basal glucose concentration (9). HbA1c was measured by modified high-performance liquid chromatography with a cation-exchange minicolumn, jacketed at 30 C and eluted at a pressure of 40 psi (10). Details of the sample preparation, dialysis to remove the labile pre-HbAlc, and technique for analysis of chromatograms were as previously reported (11). Data are expressed as means ± SE. The statistical significance of differences was estimated by Student's paired t test. RESULTS Mean ± SE fasting plasma glucose level, which was 12.0 ± 0.8 mM before therapy, fell modestly to 10.8 ± 0.3 mM after acarbose therapy. However,
B 300 |>250
150 125
2
$ 200 o o 3 150 O
100 75 50
CO
E 100 CO CO
Q.
25
50
Noon
I 8
1
Time
Noon
1
Time
FIG. 1. Mean ± SE plasma glucose (A) and insulin (6) concentrations under control conditions (•) and while receiving acarbose treatment (o) in 12 patients with non-insulin-dependent diabetes mellitus.
DIABETES CARE, VOL. 13, SUPPL. 3, AUGUST
1990
33
COMBINED SULFONYLUREA'ACARBOSE IN NIDDM
plasma glucose levels at 1200, before eating, and for 3 h after eating were significantly lower (P < 0.001) at 300 every time point in acarbose-treated patients, with a de250 crease in average plasma glucose concentration of 3.4 mM (Fig. 1/4). When plasma insulin concentrations were 200 measured before and after the addition of acarbose, it is clear that the improvement in glycemic control was 150 not associated with any increase in fasting or postpran- o 100 dial insulin concentration (Fig. IB). Although data in Fig. 1 illustrate the decrease in postprandial plasma glucose concentration that occurred with acarbose in the 12 patients as a group, they do not Noon 1 depict the great variability in the response of the indiTime vidual subjects that we noted. To do this, we selected the 4 patients with the "best" and the 4 patients with FIG. 3. Mean ± SE plasma glucose concentrations meathe "worst" plasma glucose response to acarbose. These sured before acarbose therapy (•), 6 wk after discontindata emphasize the fact that acarbose had little benefi- uing acarbose therapy (A), and while receiving drug (o) in cial effect in some patients but that it was capable of 5 patients. very effectively lowering glucose concentrations in oth-
Glue
mg/d
ers (Fig. 2). Indeed, in 4 subjects who were clearly treatment failures before the addition of acarbose, the addition of acarbose resulted in the achievement of 300 — quite acceptable glycemic control. To more fully evaluate the therapeutic effect of acar250 bose, we restudied five patients on a third occasion 6 wk after discontinuation of acarbose therapy. These re200 sults again document the ability of acarbose to reduce
