Journal of Clinical Pharmacy and Therapeutics (1991) 16,161-166. ADONIS 026947279100023P
A COMPARISON OF THE METABOLIC EFFECTS OF FLOSEQUINAN AND PROPRANOLOL IN PATIENTS WITH NON-INSULIN-DEPENDENT DIABETES MELLITUS H. M. Lewis, M. J. Kendall, A. D. Wright*, J. R. Bratty? and S. Maxwell Department of Medicine, Medical School, Edgbaston, *Department of Diabetic Medicine, The General Hospital, Steelhouse Lane, Birmingham and ?Boots Pharmaceuticals Research Department, Nottingham, U.K.
SUMMARY
T h e effects of steady-state flosequinan, a new peripheral vasodilator, and propranolol on glucose tolerance and plasma lipids in 22 non-insulindependent diabetics were investigated in a randomized double-blind placebo-controlled, three-way crossover trial. Flosequinan produced no impairment of glucose tolerance compared with placebo. Propranolol produced significant increases in fasting plasma glucose ( P < 0.01) and increases in the area under the glucose tolerance curve ( P < 0.05) compared to placebo. N o significant effects on cholesterol levels were seen on either treatment but triglyceride levels were significantly elevated on propranolol compared with placebo ( P < 0.01). These data suggest that flosequinan, used in therapeutic dosage, has no adverse metabolic effects on the non-insulin-dependent diabetic and this may be an advantage for a drug used in the treatment of hypertension or congestive heart failure. INTRODUCTION It is now established that hypertension (l), hypercholesterolaemia ( 2 ) and smoking (3) are the three most important factors in the premature development of coronary artery disease. I n addition, impaired glucose tolerance (4), hypertriglyceridaemia (5) and obesity (6) have also been shown to be important risk factors. Amongst these factors some coexist more frequently in the diabetic patient than in the general population (7). In particular, diabetes mellitus is associated with hypercholesterolaemia, hypertension and obesity as well as elevated levels of plasma glucose (8). As the known coronary risk factors are probably additive in combination (9) the increased incidence of cardiovascular events in this condition is not surprising (10). T h e treatment of hypertension in diabetic patients is complicated by the desire to avoid further adverse metabolic effects that might exacerbate the cardiovascular risk profile (1 1). Thiazide diuretics and P-blockers may both impair glucose tolerance (12) and induce unfavourable alterations in blood lipids (13). 8-Blockers may also decrease Correspondence: H.M. Lewis, Department of Medicine, Medical School, Edgbasron, Birmingham
B15 2TY, U.K.
161
162
H. M . Lewis et al.
blood flow in peripheral arteries and impair responses to hypoglycaemia (1 4). T h e need to find agents that do not have adverse metabolic effects or the potential to worsen the vascular complications of diabetes is self evident. Flosequinan (7-flU01-0-1-methylsulphinyl-4-quinolone)is a new peripheral vasodilator with balanced effects on arteries and veins which has an active metabolite with a long half-life which makes it suitable for once daily administration (15). It produces a prolonged reduction in blood pressure in healthy volunteers (16) and patients with mild untreated essential hypertension (17). This agent also shows considerable promise as a vasodilator in congestive heart failure where it induces sustained haemodynamic improvement and benefits in exercise tolerance (18-20). In a previous study flosequinan was shown to have no adverse effects on glucose tolerance in healthy volunteers (21). In this study we assess the metabolic effects of therapeutic doses of flosequinan in normotensive patients with non-insulin-dependent diabetes mellitus. Flosequinan has been compared with both an inactive placebo and propranolol, a non-selective P-blocker with known effects on plasma glucose and lipids, which was given as a positive control in this group of patients. METHODS
Patient selection Patients who were entered into the trial had non-insulin-dependent diabetes controlled on diet, glibenclamide alone or a combination of metformin and a sulphonylurea. Subjects were included whose diabetes was judged stable over the previous 6 months and who had a fasting glucose concentration of less than 12 mmol/l. Adult male and female patients, aged 35-75 years were included after giving oral and written concent. T h e study was approved by the Central Birmingham Health District Ethical Committee. Patients in whom P-blockers were contraindicated were excluded from the study. T h e patients were judged healthy apart from their diabetes mellitus by a medical examination including electrocardiogram, routine haematology and biochemistry. Their assessment also included measurement of glycosylated haemoglobin (HbA,), plasma renin activity (PRA) and fasting plasma lipids. Venous blood for measurement of PRA was collected into chilled lithium-heparin tubes, spun immediately at 4°C and the plasma stored at -70°C until assay.
Study design T h e study was a randomized, double-blind, three-way, crossover trial with three 4week treatment periods with a 2-week washout period between each treatment. T h e treatment regimens were as follows. (a) Placebo one tablet twice daily for 28 days and one tablet on the study morning. (b) Flosequinan, 50 mg twice daily for 28 days and 50 mg on the study morning. (c) Propranolol, 80 mg twice daily for 28 days and propranolol 80 mg on the study morning. T h e patients attended the Clinical Investigation Unit on day 29 of each treatment period having fasted from 22.00-hours the previous evening. T h e study medication was given at - 60 min together with the oral hypoglycaemic agent if appropriate. A cannula
Metabolic effects of jlosequinan and propranolol
163
was inserted into a suitable forearm vein and kept patent with 0.9", saline flush. T h e subjects rested for 30 min following which blood pressure was recorded using an automatic recorder (Dinamap, Critikon) and venous blood was taken for routine haematology and biochemistry, fasting lipids, PRA and HbA, measurement. Precisely 1 h after the study medication was given (time 0) a glucose tolerance test was performed by giving 75 g glucose monohydrate in 200 ml water. Three baseline venous blood samples for glucose estimation were taken at 5-min intervals before the glucose load and at 30-min intervals for 3 h thereafter. T h e blood pressure was recorded on completion of the glucose tolerance test. A visual analogue scale was used to assess well being on each study day. Assays Plasma glucose concentration was measured using a glucose hexokinase method (BCL). Plasma renin activity was measured by radioimmunoassay using the method described by O'Donnell et a1.(22). T h e HbA, level was determined by electroendosmosis (Corning Medical). Statistical analysis An analysis of variance was carried out with factors for treatment period, study and volunteer on fasting plasma glucose concentration, area under the glucose tolerance curve, plasma lipid measurements, plasma renin activity and glycosylated haemoglobin. Fasting plasma glucose concentration was the average of the three values taken before administration of the glucose load. T h e area under the glucose tolerance curve was calculated using the trapezoidal rule. An analysis of covariance was carried out on maximum glucose concentration using the same factors as before but including the fasting plasma glucose concentration as covariate. Comparisons between pairs of treatment were performed on adjusted means using the Newman-Keuls multiple comparison procedure. All analyses were carried out using procedures in SAS (Statistical Analysis System). Probability values of less than 5O1, were accepted as statistically significant.
RESULTS Patients
Of the 22 patients, 19 were male and three were female [mean ( & SD) age 59f- 10.4 years, range 35-74 years; mean weight ( fSD) 77.7 f 16.9 kg, range 47-102 kg]. Eleven patients were controlled on diet alone and the remaining 11 were controlled on either glibenclamide alone (six patients) or a combination of metformin and a sulphonylurea. Additional medication was nitrazepam (one), indomethacin (one) and paracetamol (one). Measurements Plasma glucose. Comparison between treatments for fasting plasma glucose concentration showed no statistical difference between flosequinan and placebo, however, a significant increase in fasting plasma glucose concentration was seen on propranolol compared with placebo ( P< 0.01) and with flosequinan ( P
A COMPARISON OF THE METABOLIC EFFECTS OF FLOSEQUINAN AND PROPRANOLOL IN PATIENTS WITH NON-INSULIN-DEPENDENT DIABETES MELLITUS H. M. Lewis, M. J. Kendall, A. D. Wright*, J. R. Bratty? and S. Maxwell Department of Medicine, Medical School, Edgbaston, *Department of Diabetic Medicine, The General Hospital, Steelhouse Lane, Birmingham and ?Boots Pharmaceuticals Research Department, Nottingham, U.K.
SUMMARY
T h e effects of steady-state flosequinan, a new peripheral vasodilator, and propranolol on glucose tolerance and plasma lipids in 22 non-insulindependent diabetics were investigated in a randomized double-blind placebo-controlled, three-way crossover trial. Flosequinan produced no impairment of glucose tolerance compared with placebo. Propranolol produced significant increases in fasting plasma glucose ( P < 0.01) and increases in the area under the glucose tolerance curve ( P < 0.05) compared to placebo. N o significant effects on cholesterol levels were seen on either treatment but triglyceride levels were significantly elevated on propranolol compared with placebo ( P < 0.01). These data suggest that flosequinan, used in therapeutic dosage, has no adverse metabolic effects on the non-insulin-dependent diabetic and this may be an advantage for a drug used in the treatment of hypertension or congestive heart failure. INTRODUCTION It is now established that hypertension (l), hypercholesterolaemia ( 2 ) and smoking (3) are the three most important factors in the premature development of coronary artery disease. I n addition, impaired glucose tolerance (4), hypertriglyceridaemia (5) and obesity (6) have also been shown to be important risk factors. Amongst these factors some coexist more frequently in the diabetic patient than in the general population (7). In particular, diabetes mellitus is associated with hypercholesterolaemia, hypertension and obesity as well as elevated levels of plasma glucose (8). As the known coronary risk factors are probably additive in combination (9) the increased incidence of cardiovascular events in this condition is not surprising (10). T h e treatment of hypertension in diabetic patients is complicated by the desire to avoid further adverse metabolic effects that might exacerbate the cardiovascular risk profile (1 1). Thiazide diuretics and P-blockers may both impair glucose tolerance (12) and induce unfavourable alterations in blood lipids (13). 8-Blockers may also decrease Correspondence: H.M. Lewis, Department of Medicine, Medical School, Edgbasron, Birmingham
B15 2TY, U.K.
161
162
H. M . Lewis et al.
blood flow in peripheral arteries and impair responses to hypoglycaemia (1 4). T h e need to find agents that do not have adverse metabolic effects or the potential to worsen the vascular complications of diabetes is self evident. Flosequinan (7-flU01-0-1-methylsulphinyl-4-quinolone)is a new peripheral vasodilator with balanced effects on arteries and veins which has an active metabolite with a long half-life which makes it suitable for once daily administration (15). It produces a prolonged reduction in blood pressure in healthy volunteers (16) and patients with mild untreated essential hypertension (17). This agent also shows considerable promise as a vasodilator in congestive heart failure where it induces sustained haemodynamic improvement and benefits in exercise tolerance (18-20). In a previous study flosequinan was shown to have no adverse effects on glucose tolerance in healthy volunteers (21). In this study we assess the metabolic effects of therapeutic doses of flosequinan in normotensive patients with non-insulin-dependent diabetes mellitus. Flosequinan has been compared with both an inactive placebo and propranolol, a non-selective P-blocker with known effects on plasma glucose and lipids, which was given as a positive control in this group of patients. METHODS
Patient selection Patients who were entered into the trial had non-insulin-dependent diabetes controlled on diet, glibenclamide alone or a combination of metformin and a sulphonylurea. Subjects were included whose diabetes was judged stable over the previous 6 months and who had a fasting glucose concentration of less than 12 mmol/l. Adult male and female patients, aged 35-75 years were included after giving oral and written concent. T h e study was approved by the Central Birmingham Health District Ethical Committee. Patients in whom P-blockers were contraindicated were excluded from the study. T h e patients were judged healthy apart from their diabetes mellitus by a medical examination including electrocardiogram, routine haematology and biochemistry. Their assessment also included measurement of glycosylated haemoglobin (HbA,), plasma renin activity (PRA) and fasting plasma lipids. Venous blood for measurement of PRA was collected into chilled lithium-heparin tubes, spun immediately at 4°C and the plasma stored at -70°C until assay.
Study design T h e study was a randomized, double-blind, three-way, crossover trial with three 4week treatment periods with a 2-week washout period between each treatment. T h e treatment regimens were as follows. (a) Placebo one tablet twice daily for 28 days and one tablet on the study morning. (b) Flosequinan, 50 mg twice daily for 28 days and 50 mg on the study morning. (c) Propranolol, 80 mg twice daily for 28 days and propranolol 80 mg on the study morning. T h e patients attended the Clinical Investigation Unit on day 29 of each treatment period having fasted from 22.00-hours the previous evening. T h e study medication was given at - 60 min together with the oral hypoglycaemic agent if appropriate. A cannula
Metabolic effects of jlosequinan and propranolol
163
was inserted into a suitable forearm vein and kept patent with 0.9", saline flush. T h e subjects rested for 30 min following which blood pressure was recorded using an automatic recorder (Dinamap, Critikon) and venous blood was taken for routine haematology and biochemistry, fasting lipids, PRA and HbA, measurement. Precisely 1 h after the study medication was given (time 0) a glucose tolerance test was performed by giving 75 g glucose monohydrate in 200 ml water. Three baseline venous blood samples for glucose estimation were taken at 5-min intervals before the glucose load and at 30-min intervals for 3 h thereafter. T h e blood pressure was recorded on completion of the glucose tolerance test. A visual analogue scale was used to assess well being on each study day. Assays Plasma glucose concentration was measured using a glucose hexokinase method (BCL). Plasma renin activity was measured by radioimmunoassay using the method described by O'Donnell et a1.(22). T h e HbA, level was determined by electroendosmosis (Corning Medical). Statistical analysis An analysis of variance was carried out with factors for treatment period, study and volunteer on fasting plasma glucose concentration, area under the glucose tolerance curve, plasma lipid measurements, plasma renin activity and glycosylated haemoglobin. Fasting plasma glucose concentration was the average of the three values taken before administration of the glucose load. T h e area under the glucose tolerance curve was calculated using the trapezoidal rule. An analysis of covariance was carried out on maximum glucose concentration using the same factors as before but including the fasting plasma glucose concentration as covariate. Comparisons between pairs of treatment were performed on adjusted means using the Newman-Keuls multiple comparison procedure. All analyses were carried out using procedures in SAS (Statistical Analysis System). Probability values of less than 5O1, were accepted as statistically significant.
RESULTS Patients
Of the 22 patients, 19 were male and three were female [mean ( & SD) age 59f- 10.4 years, range 35-74 years; mean weight ( fSD) 77.7 f 16.9 kg, range 47-102 kg]. Eleven patients were controlled on diet alone and the remaining 11 were controlled on either glibenclamide alone (six patients) or a combination of metformin and a sulphonylurea. Additional medication was nitrazepam (one), indomethacin (one) and paracetamol (one). Measurements Plasma glucose. Comparison between treatments for fasting plasma glucose concentration showed no statistical difference between flosequinan and placebo, however, a significant increase in fasting plasma glucose concentration was seen on propranolol compared with placebo ( P< 0.01) and with flosequinan ( P
