loirrnal o j lnternal Medicine 1991 : 229: 233-239

ADONIS 095468209 100041 2

Felodipine in the treatment of hypertensive type I1 diabetics : effect on glucose homeostasis T. KJELLSTROM, E. BLYCHERT* & F. LINDGARDE From the Departnient of Medicine. Malmci General Hospital, University of Lurid. Malmci. and the 'Depnrtment of Medicine. AB Hissle. Mcilndal. Sweden

Abstract. Kjellstrom T. Blychert E. Lindgarde F (Department of Medicine. Malmo General Hospital, University of Lund. Malmo. and Department of Medicine, AB Hassle, Molndal. Sweden). Felodipine in the treatment of hypertensive type I1 diabetics: effect on glucose homeostasis. journal oj Internal Medicine 1991 : 229: 233-239. The effect of felodipine on glucose tolerance was evaluated in 18 male type I1 diabetic patients treated with diet alone, who were hypertensive despite P-blocker treatment. The study was a double-bind cross-over comparison of placebo and felodipine in addition to p-blockade. Oral glucose tolerance tests were performed at randomization and at the end of each 4-week double-blind treatment period. The doses of felodipine given were 5 mg b.i.d. for 2 weeks followed by 10 rng b.i.d for a further 2 weeks. Blood pressure was significantly reduced during felodipine treatment. whereas heart rate remained unaltered. HbA,, and fasting insulin levels did not change during the treatment periods. Fasting and maximal blood glucose levels were not altered between any of the treatment periods. However. there was a small but statistically significant increase (median increase 4%)in the area under the glucose concentration vs. time curve after felodipine as compared to placebo treatment. This increase was not considered to be clinicully significant in the short term, but the finding merits further investigation in a rigorous long-term study. Keywords: Ca-antagonist. diabetes, felodipine. glucose tolerance.

Introduction Antihypertensive therapy must be carefully selected in patients with diabetes mellitus. Some drugs, for example diuretics, have been reported to impair glucose tolerance, at least during short-term treatment 11 , 21. Beta adrenoreceptor-blockers, particularly of the non-selective type, may mask the symptoms of hypoglycaemia in diabetic patients 13, 41. The calcium antagonists are a fairly new group of antihypertensive agents, and we have little knowledge of how, or if, they might affect glucose homeostasis. There have been reports of impaired [5, 61, unchanged [7] and even improved [8] glucose tolerance after treatment with calcium antagonists 191. Felodipine is a calcium antagonist belonging to the dihydropyridine group. It selectively reduces the contractile activity in the smooth muscles of the resistance vessels, thereby causing a pronounced decrease in systemic arterial blood pressure (BP) and

a transient baroreflex-mediated increase in heart rate (HR). This increase in HR is attenuated or even abolished when a 8-blocker is administered concomitantly. Felodipine, both alone and in combination with a 8-blocker, is an effective and welltolerated antihypertensive drug 1110, 111. The effect of felodipine on glucose tolerance has been investigated in short-term (4-8 weeks) studies, and felodipine does not appear to change the glucose tolerance in non-diabetic subjects [12, 131 or in diabetic patients treated with felodipine alone [14]. The aim of this study was to investigate the effect of felodipine on glucose tolerance in patients with type I1 diabetes mellitus, and hypertension treated with a 8-adrenergic blocker. Since the combination of felodipine and a 8-blocker is commonly used, it is of clinical interest to determine how/if the glucose tolerance is affected by this combination.

233

234

T. KJELLSTKOM et al.

Patients and methods Patients Eighteen male patients with type I1 diabetes and hypertension were included in this study. Their mean age ( f SD) was 58 +_ 9 (range 39-70) years. They had a diastolic BP > 9 5 mmHg despite treatment with individualized doses of a 8-blocker. The known duration of hypertension varied from less than 1 year to 16 years (mean value 7 f 5 years). The known mean duration of diabetes mellitus was 4 k 3 years (range 0.5-11 years), and during the study all patients were treated with diet alone for their diabetes. Five of the patients had previously been treated with glibenclamide and this treatment had recently been stopped due to unsatisfactory or undesirable effects. Glibenclamide was withdrawn at least 1 month before randomization. None of the patients were treated with insulin. Patients who had suffered myocardial infarction within the last 3 months, or who showed severely impaired liver or kidney function, were excluded from the study. The characteristics of the individual patients are shown in Table 1. Study design (Fig. 1 ) After a run-in period consisting of 4 weeks of treatment with placebo and individual doses of pblocker, all patients were randomized to double-blind treatment starting with either felodipine or placebo. The patients were instructed to maintain their diet and p-blocker treatment unchanged. Pelodipine or the corresponding placebo was given in a dose of 5 mg b.i.d. for 2 weeks, followed by a forced dose increase to 10 mg b.i.d. for a further 2 weeks. The dose was increased in all patients except one, who experienced dizziness on the 5 mg dose. After this 4-week double-blind period, there followed a single-blind wash-out period of 2 weeks. and the patients were then switched to the alternative double-blind treatment for another 4 weeks. The study was approved by the Ethical Committee of the University of Lund. All subjects gave written informed consent to participate in the study. Clinical and laboratory assessnients An oral glucose tolerance test (75 g glucose) was performed after an overnight fast before randomization. and 1-3 h after tablet intake at the end of

each double-blind period. Capillary blood samples for the analysis of glucose were taken 0, 30. 60, 90, 120 and 180 min after the glucose intake. Venous blood for insulin analysis was taken before glucose intake (time 0), and blood for the assay of C-peptide in plasma was taken at 0, 3 0 and 60 min. Blood glucose was assayed, using a hexokinase method [15]. The total insulin concentration in serum and the C-peptide concentration in plasma were determined using immunoassay kits from Novo Research Institute, Copenhagen, Denmark. Systolic and diastolic (Korotkoff phase V) BP and HR were measured after 5 min in the supine position and after 1 min of standing, 1-3 h after tablet intake. Blood pressure was measured in the right arm with a mercury sphygmomanometer, and HR was measured by pulse palpation. Any adverse effects were recorded at each visit after spontaneous reports, as well as in response to open questioning. Laboratory investigations including Hb. haematocrit, KBC, WBC, platelets, ESR, bilirubin, alkaline phosphatase, ASAT, ALAT, gamma-GT. Na+, K', creatinine, urate, triglycerides, total cholesterol, HbA,, and tests for haematuria, proteinuria and glucosuria were performed on entry to the study and at the end of each double-blind treatment period. Blood samples for the determination of felodipine were taken at the end of the double-blind periods, and were analysed using a gas chromatographic method [16]. Statistical niethods and cornputed variables

In order to evaluate the difference in effect between placebo and felodipine for variables related to glucose tolerance, the Wilcoxon matched-pairs signed-ranks test was used. Non-parametric confidence intervals (95%) were computed. The area under the glucose concentration vs. time curve (AUC) was calculated using the trapezoidal method. No carry-over effect of the treatments on blood glucose levels was found, as indicated by the Mann-Whitney test. The difference in BP and HK was calculated using Student's paired t-tests. No carry-over effect on blood pressure was found using Student's t-test for independent groups. The laboratory variables were tested for differences using the Wilcoxon matchedpairs signed-ranks test. A P-value of < 0.05 was considered to be statistically significant. Mean values and standard deviations are given for variables considered to be normally

FELODIPINE IN TYPE I1 DIABETES

235

Table 1. Patient characteristics at randomization Paticnt

no. 1 2 3 4 5 6 7 8 9

Age (years)

Body weight (kg)

12 13 14 15 16 17 18

Fasting b-glucose

Diabetes

Hypertension 3 16 1 5 8 3 6 10 13 14 14 4 2 10

88 86 75 90 74 104 72 76 84 82 109 80 80 76 91 81 105 97

64 70 65 60 62 61 39 60 64 54 49 65 47 40 59

10 11

Known duration (years)

61

57 66

2.3 7.0 4.0 5.0 8.7 3.0 1.o 1 .o

4.0 0.9 7.0 5.0 3.0 5.0 11.0 1.1 10.0 0.5

1

0 9 10

(rnmol I-')

/%blocker and daily dose

Oral diabetes therapy given previous Iy

5.7 14.2 8.4 10.2 10.0 7.6 8.3 8.4 17.4 6.5 7.5 13.1 6.9 11.8 13.0 17.2 8.9 5.8

Atenolol. 100 mg Metoprolol. 50 rng Metoprolol. 50 rng Metoprolol, 50 mg Atenolol. 100 mg Metoprolol. 50 mg Metoprolol. 200 rng Pindolol. 10 rng Metoprolol. 200 rng Metoprolol, 100 rng Metoprolol, 100 rng Propranolol. 160 nig Metoprolol, 50 mg Atenolol. 100 rng Atenolol. 50 mg Atenolol. 50 rng Metoprolol. 50 mg Atenolol. 100 rng

No No No Yes No No No No No No Yes Yes No No Yes Yes No No

P-blocker

1

e

Plocebo

Placebo

Placebo

Plocebo

10 mg b.i.d.

10 mg b.1.d. Felodipine

Plocebo

Felodipine

I

1

I

I

I

I

I

-4

0

2

4

6

8

10

Time (weeks)

Fig. 1. Study design.

distributed (BP and HK), and median values as well as the 25 and 75 percentiles (Q1 ; Q3) are given for those variables not considered to be normally distributed (i.e. glucose, insulin, C-peptide and laboratory variables).

Results Blood glucose, insulin and C-peptide concentrations and

HbAk The fasting blood glucose levels on felodipine treatment did not differ significantly from the values at I 0

randomization, or from those on placebo treatment. The maximal glucose concentrations after oral glucose load did not differ significantly between any of the treatment periods (Table 2). The AUC of blood glucose 0-180 min after the glucose load did not differ significantly after intake of felodipine or placebo, as compared to the values at randomization. However, there was a statistically significant (P < 0.05) increase in the AUC after felodipine treatment (median, upper and lower quartiles; 3158 (2362;3943)mmoll-'xmin); as compared to placebo values (2765 (2138;3866) mmol 1-' x min); see Fig. 2 and Table 2. Fourteen of I M B 229

236

T. K J E L L S T R O M et al.

Table 2. Fasting blood glucose, serum insulin and plasma C-peptide levels during placebo and felodipine treatment. The maximal glucose concentration, the area under the glucose concentration vs. time curve (AUC) and the incremental AUC after an oral glucose tolerance test. as well as the slopes from the plot C-peptide vs. blood glucose levels 0-30 and 0-60 min after the glucose load are also given Median change between treatments F-P

Confidence intervals F-P

Placebo treatment

Felodipine treatment

18

9.0 (6.9: 15.2)

10.1 ( 7 . 8 : 15.4)

0.3

-0.5:1 .O

18

18.4 (15.2:23.8)

20.4 (15.1 :24.5)

0.0

- 0 . 6 ; 1.5

AUC of glucose (mmol I-' x min)

18

2765 (2138: 3866)

3158 (2362: 3943)

77

24:278

Incremental AUC of glucose (mmol I-' x min)

18

1034 (835: 1194)

1155 (874:1312)

118

41;171

Fasting insulin (mIE I-')

18

17 ( 1 5: 26)

19 (14 ; 2 7)

0.0

-2.5:2.0

Fasting C-peptide (nmol I-')

16

0.44 (0.37 :0.85)

0.32 (0.21 :0.58)

0.2

-0.07: -0.31

C-peptide vs. b-glucose (slope 0-30 min)

16

0.020

0.035 (0.007:0.075)

0.01

-0.01 :0.03

(-0.011 :0.052)

C-peptide vs. b-glucose (slope 0-60 min)

15

0.023 ( - 0 . 0 0 2 : 0.080)

0.055 (0.007:0.066)

0.02

-0.01 : a 0 3

n Fasting blood glucose (mmol 1-l)

Maximal blood glucose

(mmolI-')

Data are presented as medians (Ql:QJ, median change between treatment periods, and as 9 5 % confidence intervals for the difference between felodipine and placebo.

25t

5t

0L

1

1

0

I

30

1

60

90

I

120

I80

Time ofter glucose IWd [min)

Fig. 2. Glucose concentrations during oral glucose tolerance

tests in diet-treated diabetic patients (n = 18) (median:Ql : Q 3 ) :

(0-0) = placebo, (@---a) = felodipine.

the 18 patients had a greater AUC of glucose during felodipine therapy as compared to placebo. and the order of treatments did not affect this result. The

confidence intervals showed that the 'true difference ' between the placebo and felodipine AUCs was, with 9 5 % probability, between 24 and 278 mmol I-' x min. The median increase in AUC on felodipine treatment was 4%. As expected, patients with high ( > 10 mmol I-') fasting glucose levels at randomization had higher AUC values of glucose during active treatment than these with lower initial glucose levels. However, the increase in AUC during felodipine treatment tended to be lower in those with higher initial glucose levels. The median increase in glucose from placebo values in patients with initial blood glucose concentrations of > 10 mmol I-' was 1% (n = 7), and 5 % in those with lower fasting blood glucose values (n = 11). Fasting insulin levels did not differ between treatments. There was no significant difference between the fasting C-peptide levels on felodipine treatment and those at randomization, whereas the fasting Cpeptide levels during placebo treatment were significantly higher (P c 0.05) than both the levels at randomization and those after felodipine treatment. The C-peptide values were plotted against the blood glucose concentrations 0, 30 and 6 0 min after the glucose load, and the slope was calculated for 0-

FELODIPINE I N TYPE I1 DIABETES

237

Miscellaneous

1 T T

T

Supine

Standing

Fig. 3. Mean (kSD) blood pressure and heart rate during treatment with felodipine (El) 10 mg b i d . or placebo ( 0 )in addition to 8-blockade.

The plasma concentration of felodipine was measured immediately before the glucose load, i.e. close to the expected plasma peak of felodipine. The mean concentration was 24.9 f 12.0 ntnol 1-' (range 6.252.0 nmol 1-l). There were no significant differences in body weight between the treatment periods. There were statistically significant increases in the activity of alkaline phosphatase, from 2.8(2.3 ; 3.2) to 3.0(2.5 ; 3.3) pkat I-' (P < 0.01), and in the leucocyte counts, from 6.2(5.2:7.9) to 6.9(5.9: 7.8) x 10' 1-' (P < 0.01) during the felodipine period as compared to the placebo values. In addition, the ESR values were significantly higher during felodipine treatment (P < 0.05), despite the fact that the median values were the same during both felodipine and placebo treatment, 8 ( 4 ;16) to 8(7:20) mm h-', (P < 0.05). Since these changes were small and within the reference ranges, they were not considered to be of clinical importance. There were no significant differences in serum cholesterol or serum triglyceride levels between the treatment periods. The median cholesterol concentrations were 6.1 1 and 6.19 mmol 1-' during the felodipine and placebo periods, respectively, and the corresponding triglyceride concentrations were 2.40 and 2.28 mmol I-'.

30 min and for 0-60 min. The two slopes were used as a measurement of the C-peptide response. There were no significant changes in the slopes either Cb30 or 0-60 min after glucose intake between any of the treatment periods. Insulin and C-peptide data are summarized in Table 2. There were no significant differences in HbA,, levels between any of the treatment periods. The median (and Q1 :Q3) values for the measurements in the placebo and felodipine periods were 7.0(6.1:9.3) and 7.2 (6.0:9.2)%,respectively.

Mild to moderate adverse experiences were reported by three patients during the double-blind placebo period, and for nine patients during the felodipine period. The most common adverse effects of felodipine treatment were peripheral oedema (reported by three patients), flushing (three patients) and palpitations (two patients). No serious adverse effects were reported, and none of the patients were withdrawn from the study.

Blood pressure arid heart rate

Discussion

The mean supine and standing BP at randomization were 1 6 6 f 2 1 / 9 9 f 5 mmHg and 1 5 7 f 1 9 / 9 9 + 8 mmHg, respectively. During felodipine treatment supine and standing BP were significantly lower than the corresponding placebo blood pressures (Fig. 3), but heart rate was not significantly altered.

The treatment of hypertension is considered to be even more important in diabetic patients than in non-diabetic subjects [ 171. The antihypertensive treatment should ideally be well tolerated and effective without any negative erects on the diabetic state. In this study we have demonstrated the efficacy of felodipine as a blood-pressure-lowering agent in diet-treated type I1 diabetics with unsatisfactory BP levels despite 8-blocker treatment.

Adverse experiences

10.2

238

T. KJELLSTROM et al.

We found no effect of felodipine on the diabetic state as measured by HbA,,. fasting blood glucose or fasting insulin levels. The fasting C-peptide levels were significantly higher during the placebo period as compared to both the felodipine period and the pretreatment values. We can offer no explanation for this finding, and it renders interpretation of the Cpeptide data (a chance finding?) difficult. In another placebo-controlled study on felodipine by our group (to be published), no effect on C-peptide levels or response was found. In the present investigation the serum cholesterol and triglyceride values remained unchanged. Thus, in the short term, we consider felodipine to be devoid of any clinically significant effect on diet-treated type I1 diabetes. This is in agreement with most previous studies on calcium antagonists and glucose homeostasis [9]. Using repeated oral glucose tolerance tests, a questionable method in patients with impaired glucose tolerance [18], we found a statistically significant (4%) increase in the AUC of glucose after felodipine treatment. Similar (although not statistically significant) results have been reported previously [14]. In our study the changes in the AUC of glucose do not appear to be influenced by the initial glucose levels. The glucose value measured after 120 min showed the greatest deviation from the corresponding placebo value. Since the sampling interval is greater between 120 and 180 min than for the other measurements, this value will affect the AUC more than any other value. A possible explanation of the high glucose levels late after the glucose load could be delayed absorption of glucose after felodipine intake. However, there are no data available to support this hypothesis. The AUC in nondiabetics is not, under comparable conditions, affected by felodipine [12, 131. The finding in our study on diabetics might also, in part, be due to a possible effect on the 'late' insulin release [19] induced by felodipine alone or in combination with 8-blockade. In the present short-term study the number of patients is too small to be conclusive for the questions raised. A study using intravenous glucose tolerance tests might help to elucidate these issues. All subjects studied had previously been clinically diagnosed as type I1 diabetics according to the WHO criteria, and they were representative of the outpatients at, or admitted to, the diabetic care unit. They were on diet treatment, which was not always very successful, and this complicates the interpretation of the results obtained during the glucose

tolerance tests. On the other hand, such variation is commonly observed in the clinic. Five of the 18 patients had previously been treated with glibenclamide. As can be seen from Table 1, some of these patients were not well controlled on diet alone. However poor glucose control was not uncommon in patients without previous antidiabetic therapy either. This study demonstrates that felodipine represents an effective antihypertensive treatment in type I1 diabetics, insufficiently treated by 8-blockade. In the short term (4 weeks) there was no clinically significant deterioration of the diabetic state. However, the observed statistically significant increase in the AUC of blood glucose after glucose loading merits further investigation. We would like to emphasixe the need for long-term studies of antihypertensive treatment on glucose tolerance in appropriate patient groups.

Acknowledgements This study was made possible by the skilled assistance of nurse Monica Sjoo-Boquist. Assistant Professor Per Fernlund kindly analysed the insulin and Cpeptide samples, and AB Hassle, Molndal. Sweden, supplied the drugs for the study.

References 1 Amery A. Bulpitt C. de Schaepdryver A e l a / . Glucose intolerance during diuretic therapy. Results of a trial by the

2

3

4 5

European Working Party on Hypertension in the Elderly. Lancet 1978: 1 : 681-3. Murphy MB. Lewis PJ. Kohner E. Schumar B. Dollery C1'. Glucose intolerance in hypertensive patients treated with diuretics: a fourteen-year follow-up. Lancet 1982: 2 : 1293-5. Lloyd-Mostyn RH. Oram S. Modification by propranolol of cardiovascular effects of induced hypoglycaemia. Imicet 1975 : 1 : 1213-5. Editorial. Beta-blockers for diabetics. lnncel 1977; 1 : 843. Giugliano I). Torella R. Cacciapuoti F e l el. Impairment of insulin secretion in man by nifedipine. Eur / Cliri Plinrrnacol

1980: 18: 395-8. 6 Charles S . Ketelslegers JM. Buysschaert M. Larnbert AE. Hyperglycaemic effect of nifedipine. B M / 198 1 : 283: 19-20. 7 Segrestaa ]M. Caulln C. Dahan R e l nl. Effect of diltiazem on plasma glucose, insulin and glucagon during an oral glucose tolerance test in healthy volunteers. Eitr / Cliri Plinrrnctcol 1984: 26: 481-3. 8 Andersson DEH. Rojdmark S. Improvement of glucose tolerance by verapamil in patients with non-insulin-dependent diabetes mellitus. Acta Med Scarid 1981 : 210: 27-33. 9 Trost. Bernhard N. Glucose metabolism and calcium antagonists. Horrn Metab Res 1990: 22: 48-56. 10 Campbell LM. Ross, JRM. Coves JR eL el. A dose-linding.

F E L O D I P I N E I N T Y P E I1 D I A B E T E S

11

12

13

14

15

placebo-controlled study of extended-release felodipine once daily in treatment of hypertension. / Cardiovasc Pharmacol 1989: 14: 869-73. Campbell LM. Cowen KJ. Cranfield FR et a/. Felodipine-ER once daily as monotherapy in hypertension. / Cardiovasc Pharmacol 1990: 1 5 : 569-73. Hedner T. Elmfeldt D, v o n Schenk H. Sjogren E. Smith U. Glucose tolerance in hypertensive patients during treatment with the calcium antagonist, felodipine. flr 1 Clin Pharmacol 1987: 24: 145-9. Kahman PL. Hulthen UL. Hokfelt B. Unchanged insulin secretion and glucose tolerance but increased insulin clearance during long-term calcium antagonism with felodipine in essential hypertension. Horm Metab Res 1987; 19: 426-9. Capewell S. Collier A. Matthews D et a/. A trial of the calcium antagonist felodipine in hypertensive type 2 diabetic patients. Diabetic Med 1 9 8 9 : 6 : 809-12. Carrol J J , Smith N. Babson A. A colorimetric serum glucose determination using hexokinase and glucose-6-phosphate dehydrogenase. fliochem Med Metab fliol 1970: 4: 171-80.

239

1 6 Ahnoff M. Determination of felodipine in plasma by capillary gas chromatography with electron capture detection. / Pharm fliomed Anal 1984 : 2 : 5 19-26. 1 7 Parving HH. Andersen AR. Smidt UM. Hommel E. Mathiesen ER. Svendsen PA. Meet of antihypertensive treatment on kidney function in diabetic nephropathy. flM/ 1987: 294: 1443-7. 18 Riccardi C. Vaccaro 0.Rivellese A, Pignalosa S. Tutino L. Mancini M. Reproducibility of the new diagnostic criteria for impaired glucose tolerance. Am / Epidemiol1985 : 121 :422-9. 19 Efendic S. Wajngot A, Cerasi E. LuR R. Insulin release, insulin sensitivity and glucose intolerance. Proc Natl Acad Sci USA 1 9 8 0 : 77: 7425-9. Received 9 April 1990, accepted 2 4 August 1990.

Correspondence: Thomas Kjellstrom. MI). Department of Medicine, Malmo General Hospital, University of Lund. S-214 01 Malmo. Sweden.