Emerging Treatments and Technologies ORIGINAL ARTICLE

Effects of Sibutramine in Obese Female Subjects With Type 2 Diabetes and Poor Blood Glucose Control

1 2 ADNAN GOKCEL, MD NEDRET TANACI, MD lar complications, and 14% for myocar- 1 1 HATICE KARAKOSE, MD NESLIHAN BASCIL TUTUNCU, MD dial infarction (7,8). The subgroup 2 1 EDA MELEK ERTORER, MD NILGUN GUVENER, MD analysis of the simvastatin (4S) study showed that reducing the level of LDL cholesterol decreased cardiovascular mortality in diabetic patients (9). Recent studies on dexfenfluramine OBJECTIVE — In this study, we evaluated the efficacy of sibutramine in combination with and fluoxetine have revealed that weight hypoglycemic drugs in obese type 2 diabetic women whose glucose levels were poorly regulated. reduction with these agents improves glu- cose control and reduces HbA , BMI, and RESEARCH DESIGN AND METHODS — Female patients with type 2 diabetes, poorly 1c Ͼ (10–12). Sibutramine is controlled glucose levels, and HbA1c 8% were randomly assigned to one of two groups. In addition to their prescribed hypoglycemic agents (maximum doses of sulfonylureas and met- an anti- drug that induces satiety formin), one group (n ϭ 30) received a placebo twice daily for 6 months and the other (n ϭ 30) and thermogenesis (13). Administration received sibutramine 10 mg b.i.d. for the same period. of sibutramine has been shown to reduce weight gain, lower the levels of nonesteri- RESULTS — One patient in the sibutramine group was excluded during the study period fied fatty acids, decrease hyperinsulin- because of ; thus, a total of 29 data sets were analyzed for this group. In the placebo emia, and reduce insulin resistance (IR) in group, five patients had to be excluded because of low treatment efficacy, leaving a total of 25 obese diabetic ob/ob mice (14). In a recent who completed the study. Comparing the changes that occurred over 6 months in the sibutra- investigation, we found that sibutramine mine and placebo groups, the former showed significantly greater reductions in fasting blood decreased IR and BMI in nondieting obese glucose (P Ͻ 0.0001), second-hour postprandial blood glucose (P Ͻ 0.0001), insulin resistance Ͻ Ͻ Ͻ Ͻ women (15). (P 0.0001), waist circumference (P 0.0001), BMI (P 0.0001), HbA1c (P 0.0001), diastolic blood pressure, pulse rate, uric acid levels, and all elements of the lipid profile except In this study, we evaluated the toler- HDL cholesterol and apolipoprotein A1. ability and efficacy of sibutramine (10 mg b.i.d.) in combination with dietary re- CONCLUSIONS — The addition of sibutramine to oral hypoglycemic therapy resulted in strictions and oral hypoglycemic drugs in significant and improvement in metabolic parameters in this patient group. Sibutra- obese women with type 2 diabetes and mine is an effective adjunct to oral hypoglycemic therapy in obese women with type 2 diabetes. poor glucose level control. Outcome was assessed based on improved glycemic Diabetes Care 24:1957–1960, 2001 control and positive changes in dyslipide- mia and BMI.

ost patients with type 2 diabetes nately, diet restriction alone usually does RESEARCH DESIGN AND are obese, dyslipidemic, and insu- not lead to adequate weight loss (3,5,6). METHODS — This double-blinded M lin-resistant (1,2). In most cases, Previous studies have shown that rank-randomized placebo-controlled high doses of hypoglycemic drugs and tight glycemic control reduces the long- prospective study was conducted at the statins or fibrates are required, and it is term complications of the disease. The U.K. Baskent University Endocrinology and usually difficult to regulate metabolic pa- Prospective Diabetes Study (UKPDS) Metabolism Clinic, which is a referral- rameters. Modification of dietary habits showed that a 1% reduction in the aver- based clinic, and all of the women studied were referred because of poor glucose and subsequent weight loss can improve age HbA1c level was associated with a glycemic control, insulin level, and lipid 21% reduction in risk for any end point control. A total of 60 female patients with profile findings (3,4); however, unfortu- related to diabetes, 37% for microvascu- type 2 diabetes, poor glucose control, Ͼ Ͼ 2 HbA1c 8%, and BMI 30 kg/m who ●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●●● were taking maximum doses of met-

1 formin (2,550 mg/day) and sulfonylureas From the Division of Endocrinology and Metabolism, Department of Internal Medicine, Baskent University, (gliclazide [320 mg/day] or glipizide [20 Adana; and the 2Department of Internal Medicine, Baskent University, Adana, Turkey. Address correspondence and reprint requests to Adnan Gokcel, MD, Baskent University Faculty of mg/day]) were enrolled. None of the pa- Medicine, Department of Internal Medicine, Division of Endocrinology and Metabolism, Dadaloglu Mah. tients had previously been on other Serin Evler 39. Sok. No. 6, 01250, Yuregir, Adana, Turkey. E-mail: [email protected]. weight-loss drugs or intensive weight-loss Received for publication 1 May 2001 and accepted in revised form 22 June 2001. programs. We excluded patients with Abbreviations: apo, apolipoprotein; HOMA, homeostasis model assessment; IR, insulin resistance; lipo(a), lipoprotein(a). type 1 diabetes, obesity of endocrine ori- A table elsewhere in this issue shows conventional and Syste`me International (SI) units and conversion gin other than that associated with type 2 factors for many substances. diabetes, uncontrolled hypertension (sys-

DIABETES CARE, VOLUME 24, NUMBER 11, NOVEMBER 2001 1957 Sibutramine in poorly controlled type 2 diabetes tolic blood pressure Ͼ160 mmHg or dia- 8 h. To measure the postprandial glucose els fell significantly below baseline stolic blood pressure Ͼ100 mmHg), or level, a second blood sample was drawn (124.88 Ϯ 8.58 and 102.24 Ϯ 51.99 mg/ , and we also excluded pregnant 2 h after ingestion of 75 g glucose. Levels dl, respectively; P Ͻ 0.0001) in patients women and patients who were using of plasma glucose, total cholesterol, HDL treated with sibutramine, but they did not ␤-adrenergic blockers, , cholesterol, and triglycerides were deter- change significantly in the placebo group. Ϯ monoamine oxidase inhibitors, or any mined by the calorimetric method, using Mean HbA1c values dropped by 2.73 drug that might affect or body a Cobas Mira Plus autoanalyzer (Roche 0.01% (P Ͻ 0.0001), and IR calculated by weight. The subjects gave their informed Diagnostics, Mannheim, Germany). LDL HOMA fell by 7.09 Ϯ 0.81 (P Ͻ 0.0001) consent to participate. cholesterol and VLDL cholesterol levels in the sibutramine group, but these pa- Patients were randomly assigned to were calculated using the Friedwald for- rameters remained unchanged in the pla- either the sibutramine 10 mg b.i.d. group mula. Insulin was measured in an AXSYM cebo group. (n ϭ 30; mean age 46.93 Ϯ 1.62 years) or autoanalyzer (Abbott Laboratories, Ab- In the sibutramine-treated patients, the placebo group (n ϭ 30; mean age bott Park, IL), using the microparticle fasting and postprandial blood glucose, 49.28 Ϯ 1.34 years), and neither the phy- enzyme immunoassay method. Apoli- insulin, total cholesterol, LDL cholesterol, sician nor the patient knew the group to poprotein (apo) A1, apoB, and lipopro- VLDL cholesterol, triglycerides, lipo(a), which they were assigned. They received tein(a) [lipo(a)] were quantitated by the apoB, uric acid levels, IR, HbA1c, waist this treatment in addition to their pre- immunoturbidimetric method in a measurement, BMI, diastolic blood pres- scribed diet and hypoglycemic drugs. All Roche/Hitachi 912 autoanalyzer (Roche sure, and pulse rate all fell significantly patients were taking maximum doses of Diagnostics). IR was calculated using the over the 6 months of treatment. These sulfonylurea and throughout homeostasis model assessment (HOMA) patients’ HDL cholesterol, apoA1, and the study. of IR (formula: glucose ϫ insulin/22.5). systolic blood pressure did not change At baseline, a physician examined Statistical testing was done with Stu- significantly. In the placebo-treated pa- each individual and measured her weight, dent’s t test analysis and homogeneity of tients, only fasting and postprandial glu- height, and waist circumference. Each pa- variance checked by Levene’s test using cose, LDL cholesterol, and HbA1c tient was placed on a diet of 25 kcal/kg SPSS version 9.05 (SPSS, Chicago, IL). decreased significantly. Patient character- ideal body weight. Under the recom- All results were expressed as the means istics at baseline are shown in Table 1, and mended regimen, patients would intake Ϯ SEM, and P Ͻ 0.05 was considered the differences between the changes that ϳ 50% of the calories from carbohydrates, significant. occurred in the sibutramine- and place- 30% from lipids, and 20% from proteins. bo-treated groups from baseline to 6 They received a list of those foods that months are summarized in Table 2. were permitted and not permitted, along with recommended portions and possible RESULTS — In the sibutramine group, combinations. one patient developed hypertension and CONCLUSIONS — Type 2 diabetes Visits were scheduled before the med- was withdrawn from the study, 11 pa- is a worldwide, growing health problem ication was started and then monthly tients reported dry mouth, and 16 re- that affects Ͼ150 million people. Non- throughout 6 months of treatment. At ported constipation. In the placebo pharmacological treatment modalities, each visit, the patient underwent a phys- group, five patients had to be excluded such as weight loss and increased physical ical examination and had her blood pres- and switched to insulin therapy because activity, represent the basis of treatment sure, heart rate, weight, height, and waist of low treatment efficacy. Thus, a total of for obese patients with this disease. The circumference measured. Any problems 25 subjects in the placebo group com- efficacy of these approaches has been or adverse events were recorded as they pleted the study. The sulfonylurea and demonstrated by numerous studies and were detected by the physician or stated metformin doses remained unchanged was summarized in recent reviews by the patient. Weight and height were throughout the study period in both (16,17); however, adequate weight loss is measured with the individual in light groups because none of the patients had seldom attained with diet therapy alone clothes and without shoes. Waist circum- problems with hypoglycemia. (3,5,6). ference was measured with the patient Findings after 6 months of treatment In the present study, the addition of standing, and the measurement was taken showed that sibutramine produced 10 mg sibutramine b.i.d. to the treatment at the midpoint between the highest point greater weight reduction than placebo regimen led to significant weight loss and of the iliac crest and the lowest limit of the (9.61 Ϯ 1.37 kg loss vs. 0.91 Ϯ 0.53 kg improvement in metabolic parameters. costal margin at the mid-axillary line. gain, respectively; P Ͻ 0.0001). The mean However, the weight reduction observed Blood pressure and heart rate were re- reduction in BMI was significantly greater was less than that noted with the same corded in the morning as the mean of in the sibutramine group than in the pla- dose of sibutramine in nondiabetic pa- three resting readings after 5 min in the cebo group (3.92 Ϯ 0.54 kg/m2 decrease tients after the same period of treatment seated position. vs. 0.36 Ϯ 0.21 kg/m2 increase, respec- (13 vs. 9.6 kg, respectively) (A.G., H.K., Blood chemistry and electrocardiog- tively; P Ͻ 0.0001). Waist circumference E.M.E., N.T., N.B.T., N.G., unpublished raphy examinations were performed be- also decreased significantly more in the study). This confirms other authors’find- fore the beginning of the medication and sibutramine group (8.04 Ϯ 3.43 cm de- ings that obese diabetic patients achieve after 6 months of treatment. Blood sam- crease vs. 0.92 Ϯ 0.49 cm increase, re- less weight reduction than nondiabetic ples were collected in the morning after spectively; P Ͻ 0.0001). obese patients taking sibutramine each patient had been fasting for at least Fasting and postprandial glucose lev- (18,19). The reason for this is unknown,

1958 DIABETES CARE, VOLUME 24, NUMBER 11, NOVEMBER 2001 Gokcel and Associates

Table 1—Patient characteristics at baseline adipose tissue and associated metabolic variables (22). We found that weight loss Characteristic Sibutramine Placebo with sibutramine produced a significant positive change in the lipid profile of Duration of diabetes (years) 8.45 Ϯ 0.55 8.30 Ϯ 0.59 obese patients with type 2 diabetes. Total ϩ Glipizide metformin users 17 14 cholesterol, LDL cholesterol, VLDL cho- ϩ Gliclazide metformin users 12 11 lesterol, triglycerides, lipo(a), and apoB Ϯ Ϯ Fasting blood glucose (mg/dl) 251.45 4.83 249.80 4.47 decreased significantly, although there Postprandial blood glucose (mg/dl) 269.79 Ϯ 11.90 277.08 Ϯ 4.80 ␮ Ϯ Ϯ were no significant changes in HDL cho- Insulin level ( U/ml) 17.85 1.15 18.90 1.01 lesterol and apoA1 levels. Because reduc- HOMA IR 11.08 Ϯ 0.73 11.67 Ϯ 0.68 Ϯ Ϯ tion of LDL cholesterol has been shown to HbA1c (%) 9.96 0.01 9.76 0.01 Ϯ Ϯ be associated with reduced cardiovascu- Total cholesterol (mg/dl) 218.69 8.87 216.44 5.40 lar morbidity and mortality in diabetic pa- HDL cholesterol (mg/dl) 45.76 Ϯ 1.75 45.92 Ϯ 1.56 tients (9), our results indicate that LDL cholesterol (mg/dl) 132.86 Ϯ 6.64 137.60 Ϯ 6.09 sibutramine is of significant potential VLDL cholesterol (mg/dl) 39.93 Ϯ 3.52 37.48 Ϯ 2.91 benefit relative to cardiovascular risk in Triglycerides (mg/dl) 200.10 Ϯ 17.66 188.04 Ϯ 14.56 Lipo(a) (mg/dl) 20.22 Ϯ 4.31 19.99 Ϯ 4.01 obese women with type 2 diabetes. apoA1 (mg/dl) 138.57 Ϯ 4.29 139.42 Ϯ 4.33 Previous studies have noted small apoB (mg/dl) 105.30 Ϯ 4.15 108.43 Ϯ 4.28 rises in blood pressure and modest in- Uric acid (mg/dl) 5.72 Ϯ 0.24 5.69 Ϯ 0.25 creases in heart rate in sibutramine- Body weight (kg) 95.57 Ϯ 3.41 95.48 Ϯ 2.84 treated patients (23,24). These effects of BMI (kg/m2) 39.30 Ϯ 1.36 37.40 Ϯ 0.99 the drug result from its mechanism of ac- Waist (cm) 109.59 Ϯ 1.46 107.84 Ϯ 2.51 tion as a and Systolic blood pressure (mmHg) 129.07 Ϯ 4.82 130.03 Ϯ 3.89 inhibitor (13). In the present in- Diastolic blood pressure (mmHg) 86.67 Ϯ 2.50 87.11 Ϯ 2.54 vestigation, patients who took sibutra- Heart rate (beats per min) 79.78 Ϯ 1.70 78.95 Ϯ 1.32 mine showed significant reductions in Data are means Ϯ SEM or n. diastolic blood pressure and heart rate. It is noteworthy that only one patient was withdrawn as a result of hypertension. but some investigators believe it is caused in patients with either type of the disease. In summary, our results show that the addition of sibutramine to oral hypogly- by sulfonylurea therapy, which is known Every 1% reduction in HbA1c decreased to promote weight gain (19). We should cardiac complications from 9 to 40%, de- cemic therapy leads to significant weight also consider the likelihood of excessive pending on the population and type of loss and improved metabolic parameters weight gain with insulin therapy had diabetes (7,8,20,21). in this patient group. Sibutramine was these patients been put on insulin, which Our sibutramine-treated patients also well tolerated by our study population. In would have been the next step toward im- showed a greater reduction in waist cir- our opinion, this drug is an effective and proving metabolic control. cumference, which is reportedly corre- safe adjunct to oral hypoglycemic therapy Our results showed statistically sig- lated with the level of abdominal visceral in obese women with type 2 diabetes. nificant improvement in glycemic control in the sibutramine-treated patients. Fast- ing and postprandial blood glucose levels, Table 2—Comparison of the mean changes that occurred from baseline to 6 months in the IR, and HbA1c decreased to a greater de- sibutramine- and placebo-treated groups gree in this group than in the placebo group. These significant differences are in Characteristic Sibutramine Placebo P all likelihood related to the weight loss that occurred in the sibutramine-treated Fasting blood glucose (mg/dl) 124.88 Ϯ 8.58 15.76 Ϯ 3.89 Ͻ0.0001 Ϯ Ϯ Ͻ individuals. HbA1c also decreased signif- Postprandial blood glucose (mg/dl) 102.24 51.99 32.88 6.13 0.0001 icantly in the placebo group, and the rea- Insulin level (␮U/ml) 5.66 Ϯ 0.97 0.68 Ϯ 0.43 Ͻ0.0001 son for this could be the “Hawthorne HOMA IR 7.09 Ϯ 0.81 0.31 Ϯ 0.28 Ͻ0.0001 Ϯ Ϯ Ͻ effect,” which is described as follows: pa- HbA1c (%) 2.73 0.01 0.53 0.01 0.0001 tients enrolled in any trial invariably show Total cholesterol (mg/dl) 28.08 Ϯ 4.93 7.68 Ϯ 5.04 Ͻ0.008 an improvement in glycemic control be- HDL cholesterol (mg/dl) 0.97 Ϯ 1.58 0.01 Ϯ 1.10 Ͼ0.05 cause of the greater attention paid to them LDL cholesterol (mg/dl) 20.92 Ϯ 4.66 13.32 Ϯ 3.94 Ͼ0.05 and because they may also want to show VLDL cholesterol (mg/dl) 8.68 Ϯ 2.58 0.01 Ϯ 2.58 Ͻ0.02 their interest in the study. Recent epide- Triglycerides (mg/dl) 46.76 Ϯ 13.09 0.36 Ϯ 7.26 Ͻ0.08 miological studies and prospective thera- Body weight (kg) 9.61 Ϯ 1.37 0.91 Ϯ 0.53* Ͻ0.0001 peutic trials have revealed that glycemic BMI (kg/m2) 3.92 Ϯ 0.54 0.36 Ϯ 0.21* Ͻ0.0001 control reduces the microvascular and Waist (cm) 8.04 Ϯ 3.43 0.92 Ϯ 0.49* Ͻ0.0001 macrovascular complications of diabetes Data are means Ϯ SEM. *Increases (all others are decreases).

DIABETES CARE, VOLUME 24, NUMBER 11, NOVEMBER 2001 1959 Sibutramine in poorly controlled type 2 diabetes

statin on coronary artery diseases in dia- proves glycaemic control and other met- References betic patients with coronary artery disease abolic parameters in obese patients with 1. Maggio CA, Pi-Sunyer FX: The prevention (Abstract). Diabetes 44 (Suppl. 1):35A, type 2 diabetes mellitus. Diab Obes Metab and treatment of obesity: application to 1995 2:175–187, 2000 type 2 diabetes. Diabetes Care 20:1744– 10. Willey KA, Molyneaux LM, Yue DK: 19. Finer N, Bloom SR, Frost GS, Banks LM, 1766, 1997 Obese patients with type 2 diabetes Griffiths J: Sibutramine is effective for 2. Albu J, Konnarides C, Pi-Sunyer X: poorly controlled by insulin and met- Weight control: metabolic and cardiovas- weight loss and diabetic control in obesity formin: effects of adjunctive dexfenflura- with type 2 diabetes: a randomised, dou- cular effects. Diabetes Rev 3:335–347, mine therapy on glycaemic control. 1995 ble-blind, placebo-controlled study. Diab Diabet Med 11:701–704, 1994 Obes Metab 2:105–112, 2000 3. Wing RR, Koeske R, Epstein LH, Nowalk 11. O’Kane M, Wiles PG, Wales JK: Fluox- 20. The Diabetes Control and Complications MP, Gooding W, Becker D: Long-term ef- etine in the treatment of obese type 2 di- Trial Research Group: The effect of inten- fects of modest weight loss in type II dia- abetic patients. Diabet Med 11:105–110, betic patients. Arch Intern Med 147:1749– 1994 sive treatment of diabetes on the develop- 1753, 1987 12. Daubresse JC, Kolanowski J, Krzentowski ment and progression of long-term 4. Lean ME, Powrie JK, Anderson AS, Garth- G, Kutnowski M, Scheen A, Van Gaal L: complications in insulin-dependent dia- waite PH: Obesity, weight loss and prog- Usefulness of fluoxetine in obese non-in- betes mellitus. N Engl J Med 329:977–986, nosis in type 2 diabetes. Diabetes Med sulin-dependent diabetics: a multicenter 1993 7:228–233, 1990 study. Obes Res 4:391–396, 1996 21. Klein R: Hyperglycemia and microvascu- 5. Close EJ, Wiles PG, Lockton JA, Walms- 13. McNeely W, Goa KL: Sibutramine: a re- lar and macrovascular disease in diabetes ley D, Oldham J, Wales JK: Diabetic diets view of its contribution to the manage- (Review). Diabetes Care 18:258–268, and nutritional recommendations: what ment of obesity. Drugs 56:1093–1124, 1995 happens in real life? Diabet Med 9:181– 1998 22. Pouliot MC, Despres JP, Lemieux S, 188, 1992 14. Day C, Bailey CJ: Effect of antiobesity Moorjani S, Bouchard C, Tremblay A, 6. Manning RM, Jung RT, Leese GP, Newton agent sibutramine in obese-diabetic ob/ob Nadeau A, Lupien PJ: Waist circumfer- RW: The comparison of four weight re- mice. Int J Obes Relat Metab Disord 22: ence and abdominal sagittal diameter: duction strategies aimed at overweight di- 619–623, 1998 best simple anthropometric indexes of ab- abetic patients. Diabet Med 12:409–415, 15. Gokcel A, Gumurdulu Y, Karakose H, Ka- dominal visceral adipose tissue accumu- 1995 rademir BM, Anarat R: Effects of sibutra- lation and related cardiovascular risk in 7. UK Prospective Diabetes Study (UKPDS) mine on leptin levels in nondieting obese men and women. Am J Cardiol 73:460– Group: Effect of intensive blood glucose women (Abstract). Int J Obes Relat Metab 468, 1994 control with metformin on complications Disord 24 (Suppl. 1):S171, 2000 23. Apfelbaum M, Vague P, Ziegler O, Hano- in overweight patients with type 2 diabe- 16. Kelley DE: Effects of weight loss on glu- tin C, Thomas F, Leutenegger E: Long- tes (UKPDS 34). Lancet 352:854–865, cose homeostasis in NIDDM. Diabetes Rev 1998 3:366–377, 1995 term maintenance of weight loss after 8. Stratton IM, Adler AI, Neil HAW, Mat- 17. Schneider SH, Morgado A: Effects of a very-low-calorie diet: a randomized thews DR, Manley SE, Cull CA, Hadden fitness and physical training on carbohy- blinded trial of the efficacy and tolerabil- D, Turner RC, Holman RR: Association of drate metabolism and associated cardio- ity of sibutramine. Am J Med 106:179– glycemia with macrovascular complica- vascular risk factors in patients with 184, 1999 tions of type 2 diabetes (UKPDS 35): diabetes. Diabetes Rev 3:378–407, 1995 24. Hanotin C, Thomas F, Jones SP, Leu- prospective observational study. BMJ 18. Fujioka K, Seaton TB, Rowe E, Jelinek CA, tenegger E, Drouin P: Efficacy and tolera- 321:405–412, 2000 Raskin P, Lebovitz HE, Weinstein SP, bility of sibutramine in obese patients: a 9. Pyo¨ra¨la¨ K, Pedersen TR, Kjekshus J: The Sibutramine/Diabetes Clinical Study dose-ranging study. Int J Obes Relat Metab effect of cholesterol lowering with simva- Group: Weight loss with sibutramine im- Disord 22:32–38, 1998

1960 DIABETES CARE, VOLUME 24, NUMBER 11, NOVEMBER 2001