- Email: [email protected]
Effect of a sulfonylurea (gliclazide) treatment on insulin sensitivity and glucose-mediated glucose disposal in patients with non-insulin-dependent diabetes mellitus (NIDDM)
Diabetes Research and Clinical Practice, 20 (1993) 147-154 0 1993 Elsevier Scientific Publishers Ireland Ltd. All rights reserved. 016%8227/93/$06.00
147
DIABET 00749
Effect of a sulfonylurea (gliclazide) treatment on insulin sensitivity and glucose-mediated glucose disposal in patients with non-insulin-dependent diabetes mellitus (NIDDM) Bernard0 L. Wajchenberg, Ana Tereza M.G. Santomauro
and Raul N. Porrelli
Endocrine Secretion and Laboratory of Medical Investigation (LIM No. 251. Hospital das Clinicas. Sao Paula. SP. Brazil
(Received 28 October 1992; revision accepted 13 January 1993)
Five male non-obese newly diagnosed NIDDM and 5 age-, sex- and body mass index (BMI) matched healthy controls without a family history of diabetes were submitted to a frequently sampled intravenous (i.v.) glucose tolerance test modified by exogenous insulin administration for estimation of insulin sensitivity (St) and glucose-mediated glucose disposal (So) with Bergman’s minimal model computer analysis of glucose kinetics. The tests were repeated after 3 months treatment with a second generation sulfonylurea, gliclazide, in the diabetics subjects. St and So were markedly reduced before gliclazide therapy in the diabetics in comparison to the paired controls. After gliclazide, despite significantly lower (almost normal) plasma glucose, normalization of glycosylated hemoglobin and increased fasting insulin levels, there was a slight but significant increase in St while So showed a further reduction, the improvement in glucose control being also associated to the significant increased first and 2nd phase insulin release for the first 20 min after glucose infusion.
Key words Non-insulin-dependent diabetes mellitus; Gliclazide; Modified intravenous glucose tolerance; Insulin sensitivity; Glucose-mediated glucose disposal
Introduction
Total glucose disposal in humans consists of insulin and non-insulin-mediated glucose disposal the latter occurring in both insulin-sensitive and non-insulin sensitive tissues (e.g. brain and blood Correspondence to: Bernard0 Leo Wajchenberg, M.D., Endocrine Service, Hospital das Clinicas, P.O. Box 8091, 01065-970, Sao Paulo, SP, Brazil.
cells). Recent data indicate that non-insulinmediated glucose uptake is the major pathway for glucose disposal in the postabsorptive state in both diabetic normal and non-insulin-dependent (NIDDM) subjects, accounting for 70% of glucose disposal [ 11. From the minimal model approach, Bergman [2] and Welch et al. [3] reported that both insulin sensitivity (Sr) and glucose-mediated glucose disposal (So) are reduced in NIDDM proposing that
148
the decline in So may be pathogenic for NIDDM, using the modification of the intravenous glucose tolerance test (IVGTT) by giving an insulin bolus 20 min after the glucose injection [3], since the measurement of insulin sensitivity relies on intact insulin secretion in response to glucose thus precluding examination of diabetic patients, We used the minimal model technique of Bergman to quantitate the Si and So on 5 male non-obese patients with NIDDM and 5 age-, sex- and BMImatched healthy control subjects with no family history of diabetes. Our study was designed to determine the effect of long-term (3 months) second generation sulfonylurea, gliclazide, treatment on the above mentioned parameters.
Patients and Methods Studies were performed in 5 male non-obese insulin-independent newly diagnosed diabetics and 5 male non-obese normal subjects, without family history of diabetes. Informed consent was obtained from all subjects and approval was granted by the Ethical Committee of Hospital das Clinicas. Every subject was submitted to a modified IVGTT performed with the individual in a supine position. After a 10-12-h overnight fast, three baseline blood samples were obtained at -30, - 15 and 0 min. Intravenous glucose load (0.3 g/kg, 50% dextrose) was administered over 1 min at t = 0 min. Subsequent samples were drawn at each 1 min from 1 to 12 min, each 2 min from 14 to 22 min, each 1 min from 23 to 27 min, each 10 min from 30 to 160 min and at 180 min. At 20 min, 0.08 U/kg of regular monocomponent pork insulin (‘Actrapid’, NOVO) in the diabetics and 0.04 U/kg in the normal subjects were given over a I-min period [3]. Samples were centrifuged and kept at -20°C until assayed. The test was repeated in the diabetic patients after 3 months of the second generation sulfonylurea, gliclazide, treatment in a dose of 80- 160 mg/day depending on the response of their fasting plasma glucose measured every 2 weeks. Before testing, all subjects were maintained on their usual diets with the exception of free sugar for the diabetic individuals.
Plasma glucose was measured in a Technicon Autoanalyzer by the ferricyanide method [4]. Plasma insulin was determined by a radioimmunoassay using a modification of the method of Desbuquois and Aurbach [5] on which a second antibody was added to the polyethylene glycol to separate free from insulin-bound antibody fractions and glycosylated hemoglobin was measured after elution from an ion-exchange column by modification of the method of Trivelli et al. [6]. The i.v. glucose tolerance was evaluated by three of the usual parameters, the first phase insulin release, using the first live minutes after glucose infusion, the insulin sensitivity index (St) and the glucose effectiveness index (So). The second phase insulin release was considered from IO-20 min since the first phase is completed within the first 10 min [7]. Insulin sensitivity index (Sr) is defined as the increase in fractional glucose disappearance per unit increase in plasma insulin, i.e. insulin action (independent of both glucose and insulin levels) and glucose effectiveness (So) as the ability of glucose per se to enhance its own disappearance independent of any increment in plasma insulin above the basal level [8]. Sr and So parameters were calculated using the minimal model analysis with the MINIMOD software program [8] in the modified IVGTT by the peripheral insulin bolus substituted for endogenous insulin release to measure insulin sensitivity in normal and diabetic subjects. All data were expressed as the mean f S.E.M. The statistical significance of differences between mean values was calculated using the MannWhitney U-test [9] which does not rely on the assumption of normal distribution of data.
Results The clinical and laboratory data of the diabetic subjects are indicated in Table 1. The patients matched for the normals (41.8 & 3.1 years) for age and BMI (23.5 f 0.8 kg/m’). The fasting plasma glucose levels were significantly elevated in the diabetic subjects as well as glycosylated hemoglobin which was above the normal range (6-8%).
Age (year)
40 44
46 46 56
46.4 2.2
1 2
3 4 5
Mean S.E.M. P
and laboratory
Clinical
Subjects
1
TABLE
67 80 61 70 69 69.4 3.0
65 79
61 69 68
68.4 3.0 NS
24.0 0.3 NS
23.8 23.9 25.0
22.7 24.4
Control
Glic. Control 15.5 16.0 13.0 19.4 14.5 15.7 1.1 0.01
23.4 24.1 23.8 24.2 25.4 24.3 0.3
Glucose
(Glic.) treatment
Glic.
and gliclazide
Control
before
BMI (kg/m*)
patients
Wt. (kg)
Fasting
data of the NlDDM
50.3 57.4 14.4 14.4 57.4 38.8 10.0 0.05
6.1 7.0
6.6 0.2
6.7 6.0 6.4
Control
Insulin
Glic.
(mmol/l)
Control 14.8 17.9 12.3 11.6 15.6 14.3 1.1 0.01
71.8 114.9 40.3 57.4 86.2 16.1 11.5
HbA,,
Glic.
(wmol/l)
(“%)
7.1 0.2
1.2 7.7 6.3 7.1 7.0
Glic.
50
-&
G
$ +-
- - ,&i& ......._.. ~ 1
151 TABLE
2
Insulin-modified
Before glic. After glic. Pre- vs. post-glic.
P
intravenous
glucose
tolerance
test
O-20 min
20- 180 min
Insulin
Insulin peak (~molil)
response
1st phase (pmol/l)
2nd phase
71.8 f 28.7 244.1 f 79.0
56.0 f 176.7 f
0.027
0.01
nadir
(mmol/l)
Time (mitt)
6476.4 + 3037.1 5629.1 f 2383.8
8.8 f 2.0 4.5 f 0.1
160 160
NS
NS
NS
(flmolfl) 16.5 17.2
After 3 months gliclazide, mean body weight and BMI did not change. Mean fasting plasma glucose and corresponding glycosylated hemoglobin fell significantly while mean plasma insulin levels increased with borderline significance after sulfonylurea. Response to i. v. glucose (O-20 min) (Table 2)
The diabetic patients either before or after gliclazide presented higher basal and peak glucose values, similar basal insulin and reduced insulin responses to i.v. glucose (Table 1 and Fig. 1) in comparison to the normal subjects. While the normal controls exhibited the peak of the first phase of insulin responses at 4 min of 947.8 + 122.1 pmol/l, insulin levels rose only to 71.8 f 28.7 pmol/l in the untreated diabetics and increasing significantly to 244.1 f 79.0 ~molll after treatment. The 2nd phase of insulin secretion attained a plateau (the last 5 samples from 12-20 min) at 82.6 f 28.7 in the normals while it was 56.0 f 16.5 in the diabetic patients before treatment and increasing significantly after gliclazide to 176.7 f 17.2 pmol/l. Response to i.v. insulin (20-180
Glucose
min) (Table 2}
In the normal subjects, the dose of insulin administered (0.04 U/kg) produced mean peak insulin levels of 1852.4 f 315.9 pmol/l without hypoglycemia. Injections of 0.08 U/kg used in the diabetics gave a reliable change in the rate of glucose concentration also without hypoglycemia after inducing a peak insulin of 6476.4 f 3037.1 pmol/l in the untreated patients not significantly
different from that observed in the normals due to the dispersion of the values obtained. In the NIDDM, after gliclazide, there were no significant differences in all parameters evaluated in comparison to the correspondent control test (still statistically different from that observed in the normal subjects). Mean S’ was significantly lower in the diabetics that in the control subjects (2.37 + 0.21 vs. 47.10 f 6.68 X 10-4.min-’ .p/mol-’ . I-‘; P = 0.001). After 3 months gliclazide S’ increased significantly when compared to baseline study (13.86 f 5.24 vs. 2.37 +Z0.21 x 10e4 min-I. pmol-‘.I’; P < 0.05) but still it was lower than that found in the normal controls (P = 0.003) (Fig. 2). Baseline mean So was significantly lower in the diabetics when compared to the control subjects (2.04 f 0.21 vs. 3.36 f 0.28 x 10-2.min-‘; P = 0.08) and fell further after gliclazide to 1.72 f 0.26 x lo-‘.min-’ (P = 0.01) (Fig. 2). Discussion With a bolus infusion of insulin and using the Bergman minimal model simultaneous estimation of S’ and So were readily obtained in normal and NIDDM by Bergman and coworkers [2,3] and by Finegood et al. [lo) in normal and IDDM. In accordance with previously published reports [2,3,1 l] defects in both S’ and So contribute to the marked decrease of glucose utilization in NIDDM. We have shown that with the use of a second generation sulfonylurea, gliclazide, the
I52 60 p=O.O06
p=o.o03 I
I
I
r
I p=O.O06
p=O.OOl
I
I
I
6
50
p=o.o1
5 I
I
.A 40
-
+
E 30
d
I 0 x
20,
H ul
IO
0
CONTROL
CONTROL
NIDDM BEFORE
3mo GLICLAZIDE
GLICLAZIDE
Fig. 2. Insulin sensitivity (S,) and glucose-mediated and after 3 months gliclazide therapy. Columns
NIDDM BEFORE
3mo
glucose disposal and bars indicate
greatly improved glycemic control was associated with a small statistically significant rise in Si. Similar observations by using euglycemic insulin clamps were made with all therapeutic interventions effective in reducing the fasting plasma glucose concentration (cited in Ref. 12). Since the primary abnormality inherited in NIDDM is probably insulin resistance if the resistance from glucose toxicity is superimposed on the resistance inherited in type 2 diabetes we would
(SG) of healthy control subjects and NIDDM patients before mean f S.E.M.. P-value analysis by Mann-Whitney U-test.
not expect that normalization of the plasma glucose concentration and removal of glucose toxicity to produce major improvement in insulin sensitivity as our data have been clearly demonstrated. On the other hand, the glucose effectiveness defect is considered pathogenic for NIDDM according to Bergman and coworkers [2,3] since it declines only with the development of glucose intolerance being normal in non-diabetic offsprings of patients with NIDDM, however, presenting in-
153
sulin insensitivity of both parameters evaluated by the modified IVGTT [ 131. We have observed a significant fall in So after 3 months gliclazide therapy at the time S, increased in all to values much lower, except in one patient, to that found in our matched normal subjects. The fall in So is suggestive that the reduction in glucose effectiveness once established cannot be corrected despite normalization of plasma glucose levels at least for the period of our study. Our observation that glucose effectiveness is not normalized after sulfonylurea treatment is similar to that of Finegood et al. [lo] who have found that cyclosporine-induced complete remission of type 1 diabetes did not improve So. The improvement of glucose control with significantly lower (almost normal) fasting plasma glucose after gliclazide therapy without major improvements in St and So for the period of treatment could suggest that the amelioration of plasma glucose levels is determined at least in part to the increased insulin secretion as evaluated by the first and 2nd phases of insulin release during the first 20 min after glucose infusion. The change in insulin secretion is probably related, partially at least, to reduction in blood glucose with subsequent reduction in glucose toxicity effects on the beta cell [12]. However, we have previously demonstrated a of insulin-mediated significant improvement peripheral sensitivity (glucose clearance) after 3 months gliclazide treatment in 9 NIDDM [14] which could suggest that a small change in Si assessed in the unphysiologic conditions of the present study could result in major change in glucose concentrations. Thus, it is difficult to determine whether the alterations in insulin secretion are indeed the major cause of the lower plasma glucose concentrations. The significance and mechanism underlying the reduced So in newly diagnosed NIDDM and IDDM and not normalized during diabetes control remains to be determined. Since So is glucose-induced glucose uptake at basal glucose levels according to the minimal model, the fact that hyperglycemia itself increases glucose uptake [15] would not seem to bear on the value of So. Finally, since our diabetics received the same
amount of insulin before and after treatment then it could not affect the estimation of So, as Finegood et al. [lo] demonstrated that So was greater in those subjects receiving higher insulin infusion compared to those receiving a lower insulin infusion rate. They postulated that the dependence on the dose of insulin could be associated with the difference in nadir plasma glucose subsequent to the insulin infusion. In our study, the mean nadir glucose level was lower although non-significantly after gliclazide.
Acknowledgements Supported by grants from Institute de Recherches Internationales Servier, Neuilly-Sur-Seine, France and Conselho National de Desenvolvimento Cientitico e Tecnologico (CNPq), No. 500112/90-B, Brasilia, Brazil. The authors wish to thank Dr. R.N. Bergman for his help in the development of the modified IVGTT in our Laboratory. References Baron, A.D., Kolterman, O.G., Bell, J.. Mandarino, J. and Olefsky, J.R. (1985) Rates of noninsulin-mediated glucose uptake are elevated in Type II diabetic subjects. J. Clin. Invest. 76, 1782-1788. Bergman, R.N. (1989) Toward physiological understanding of glucose tolerance. Minimal-model approach. Diabetes 38, 1512-1527. Welch, N.S., Gebhart, S.S.P., Bergman, R.N. and Phillips, L.S. (1990) Minimal-model analysis of intravenous glucose tolerance test-derived insulin sensitivity in diabetic subjects. J. Clin. Endocrinol. Metab. 71. 1508-1518. Hoffman, W. (1937) A rapid photoelectric method for the determination of glucose in urine and blood. J. Biol. Chem. 120, 51-55. Desbuquois, B. and Aurbach, G.D. (1981) Use of polyethylene glycol to separate free and antibody-bound peptide hormones in radioimmunoassay. J. Clin. Endocrinol. Metab. 33, 732-738. Trivelli. L.A., Ranney, H.N. and Lai, H.T. (1978) Hemoglobin components in patients with diabetes mellitus. New. Engl. J. Med. 284, 353-357. Ward, W.K., Beard, I.C.. Halter, J.B., Pfeifer, M.A. and Porte Jr., D. (1984) Pathophysiology of insulin secretion in non-insulin-dependent diabetes mellitus. Diabetes Care 7, 491-502.
154 8
Bergman, R.N. and Pacini, G. (1986) MINIMOD: A computer program to calculate insulin sensitivity and pancreatic responsitivity from frequently sampled intravenous glucose tolerance test. Comput. Meth. Prog. Biomed. 23, 11-122. 9 Kuo, S.S. (1972) Computer Aplications of Numerical Methods. Reading, Massachusetts. Addison Wesley Publishing Co., pp. 275-279. 10 Finegood, D.T., Hramiak. I.M. and Dupre, J. (1990) A modified protocol for estimation of insulin sensitivity with the minimal-model of glucose kinetics in patients with Insulin-dependent diabetes. J. Clin. Endocrinol. Metab. 70, 1538-1549. II Ward, G.M., Weber, K.M., Walters, I.M. et al. (1991) A rapid modified minimal model analysis of insulin sensitivity and glucose-mediated glucose disposal in Insulindependent diabetes. Metabolism 40, 4-9.
12 Rossetti, L., Giaccari, A. and DeFronzo. R.A. (1990) Glucose toxicity. Diabetes Care 13, 610-630. 13 Osei, K., Cottrell, D.A. and Orabella, M.M. (1991) Insulin sensitivity, glucose effectiveness and body fat distribution pattern on nondiabetic offsprings of patients with NIDDM. Diabetes Care 14, 890-896. 14 Wajchenberg, B.L., Malerbi, D.A.C., Giurno Filho, A., Giannela Neto, D., Cherem, J.J. and Lerario, A.C. (1987) Effect of gliclazide treatment on red blood cell insulin receptors, hepatic glucose production and peripheral glucose utilization in non-insulin-dependent diabetes mellitus. Bull. IDF 32, 16-21. I5 Capaldo, R.. Santoro, D., Riccardi, G., Perrotti, N. and Sacca, L. (1981) Direct evidence for a stimulatory effect of hyperglycemia per se on peripheral glucose disposal in type II diabetes. J. Clin. Invest. 77, 1285-1290.
147
DIABET 00749
Effect of a sulfonylurea (gliclazide) treatment on insulin sensitivity and glucose-mediated glucose disposal in patients with non-insulin-dependent diabetes mellitus (NIDDM) Bernard0 L. Wajchenberg, Ana Tereza M.G. Santomauro
and Raul N. Porrelli
Endocrine Secretion and Laboratory of Medical Investigation (LIM No. 251. Hospital das Clinicas. Sao Paula. SP. Brazil
(Received 28 October 1992; revision accepted 13 January 1993)
Five male non-obese newly diagnosed NIDDM and 5 age-, sex- and body mass index (BMI) matched healthy controls without a family history of diabetes were submitted to a frequently sampled intravenous (i.v.) glucose tolerance test modified by exogenous insulin administration for estimation of insulin sensitivity (St) and glucose-mediated glucose disposal (So) with Bergman’s minimal model computer analysis of glucose kinetics. The tests were repeated after 3 months treatment with a second generation sulfonylurea, gliclazide, in the diabetics subjects. St and So were markedly reduced before gliclazide therapy in the diabetics in comparison to the paired controls. After gliclazide, despite significantly lower (almost normal) plasma glucose, normalization of glycosylated hemoglobin and increased fasting insulin levels, there was a slight but significant increase in St while So showed a further reduction, the improvement in glucose control being also associated to the significant increased first and 2nd phase insulin release for the first 20 min after glucose infusion.
Key words Non-insulin-dependent diabetes mellitus; Gliclazide; Modified intravenous glucose tolerance; Insulin sensitivity; Glucose-mediated glucose disposal
Introduction
Total glucose disposal in humans consists of insulin and non-insulin-mediated glucose disposal the latter occurring in both insulin-sensitive and non-insulin sensitive tissues (e.g. brain and blood Correspondence to: Bernard0 Leo Wajchenberg, M.D., Endocrine Service, Hospital das Clinicas, P.O. Box 8091, 01065-970, Sao Paulo, SP, Brazil.
cells). Recent data indicate that non-insulinmediated glucose uptake is the major pathway for glucose disposal in the postabsorptive state in both diabetic normal and non-insulin-dependent (NIDDM) subjects, accounting for 70% of glucose disposal [ 11. From the minimal model approach, Bergman [2] and Welch et al. [3] reported that both insulin sensitivity (Sr) and glucose-mediated glucose disposal (So) are reduced in NIDDM proposing that
148
the decline in So may be pathogenic for NIDDM, using the modification of the intravenous glucose tolerance test (IVGTT) by giving an insulin bolus 20 min after the glucose injection [3], since the measurement of insulin sensitivity relies on intact insulin secretion in response to glucose thus precluding examination of diabetic patients, We used the minimal model technique of Bergman to quantitate the Si and So on 5 male non-obese patients with NIDDM and 5 age-, sex- and BMImatched healthy control subjects with no family history of diabetes. Our study was designed to determine the effect of long-term (3 months) second generation sulfonylurea, gliclazide, treatment on the above mentioned parameters.
Patients and Methods Studies were performed in 5 male non-obese insulin-independent newly diagnosed diabetics and 5 male non-obese normal subjects, without family history of diabetes. Informed consent was obtained from all subjects and approval was granted by the Ethical Committee of Hospital das Clinicas. Every subject was submitted to a modified IVGTT performed with the individual in a supine position. After a 10-12-h overnight fast, three baseline blood samples were obtained at -30, - 15 and 0 min. Intravenous glucose load (0.3 g/kg, 50% dextrose) was administered over 1 min at t = 0 min. Subsequent samples were drawn at each 1 min from 1 to 12 min, each 2 min from 14 to 22 min, each 1 min from 23 to 27 min, each 10 min from 30 to 160 min and at 180 min. At 20 min, 0.08 U/kg of regular monocomponent pork insulin (‘Actrapid’, NOVO) in the diabetics and 0.04 U/kg in the normal subjects were given over a I-min period [3]. Samples were centrifuged and kept at -20°C until assayed. The test was repeated in the diabetic patients after 3 months of the second generation sulfonylurea, gliclazide, treatment in a dose of 80- 160 mg/day depending on the response of their fasting plasma glucose measured every 2 weeks. Before testing, all subjects were maintained on their usual diets with the exception of free sugar for the diabetic individuals.
Plasma glucose was measured in a Technicon Autoanalyzer by the ferricyanide method [4]. Plasma insulin was determined by a radioimmunoassay using a modification of the method of Desbuquois and Aurbach [5] on which a second antibody was added to the polyethylene glycol to separate free from insulin-bound antibody fractions and glycosylated hemoglobin was measured after elution from an ion-exchange column by modification of the method of Trivelli et al. [6]. The i.v. glucose tolerance was evaluated by three of the usual parameters, the first phase insulin release, using the first live minutes after glucose infusion, the insulin sensitivity index (St) and the glucose effectiveness index (So). The second phase insulin release was considered from IO-20 min since the first phase is completed within the first 10 min [7]. Insulin sensitivity index (Sr) is defined as the increase in fractional glucose disappearance per unit increase in plasma insulin, i.e. insulin action (independent of both glucose and insulin levels) and glucose effectiveness (So) as the ability of glucose per se to enhance its own disappearance independent of any increment in plasma insulin above the basal level [8]. Sr and So parameters were calculated using the minimal model analysis with the MINIMOD software program [8] in the modified IVGTT by the peripheral insulin bolus substituted for endogenous insulin release to measure insulin sensitivity in normal and diabetic subjects. All data were expressed as the mean f S.E.M. The statistical significance of differences between mean values was calculated using the MannWhitney U-test [9] which does not rely on the assumption of normal distribution of data.
Results The clinical and laboratory data of the diabetic subjects are indicated in Table 1. The patients matched for the normals (41.8 & 3.1 years) for age and BMI (23.5 f 0.8 kg/m’). The fasting plasma glucose levels were significantly elevated in the diabetic subjects as well as glycosylated hemoglobin which was above the normal range (6-8%).
Age (year)
40 44
46 46 56
46.4 2.2
1 2
3 4 5
Mean S.E.M. P
and laboratory
Clinical
Subjects
1
TABLE
67 80 61 70 69 69.4 3.0
65 79
61 69 68
68.4 3.0 NS
24.0 0.3 NS
23.8 23.9 25.0
22.7 24.4
Control
Glic. Control 15.5 16.0 13.0 19.4 14.5 15.7 1.1 0.01
23.4 24.1 23.8 24.2 25.4 24.3 0.3
Glucose
(Glic.) treatment
Glic.
and gliclazide
Control
before
BMI (kg/m*)
patients
Wt. (kg)
Fasting
data of the NlDDM
50.3 57.4 14.4 14.4 57.4 38.8 10.0 0.05
6.1 7.0
6.6 0.2
6.7 6.0 6.4
Control
Insulin
Glic.
(mmol/l)
Control 14.8 17.9 12.3 11.6 15.6 14.3 1.1 0.01
71.8 114.9 40.3 57.4 86.2 16.1 11.5
HbA,,
Glic.
(wmol/l)
(“%)
7.1 0.2
1.2 7.7 6.3 7.1 7.0
Glic.
50
-&
G
$ +-
- - ,&i& ......._.. ~ 1
151 TABLE
2
Insulin-modified
Before glic. After glic. Pre- vs. post-glic.
P
intravenous
glucose
tolerance
test
O-20 min
20- 180 min
Insulin
Insulin peak (~molil)
response
1st phase (pmol/l)
2nd phase
71.8 f 28.7 244.1 f 79.0
56.0 f 176.7 f
0.027
0.01
nadir
(mmol/l)
Time (mitt)
6476.4 + 3037.1 5629.1 f 2383.8
8.8 f 2.0 4.5 f 0.1
160 160
NS
NS
NS
(flmolfl) 16.5 17.2
After 3 months gliclazide, mean body weight and BMI did not change. Mean fasting plasma glucose and corresponding glycosylated hemoglobin fell significantly while mean plasma insulin levels increased with borderline significance after sulfonylurea. Response to i. v. glucose (O-20 min) (Table 2)
The diabetic patients either before or after gliclazide presented higher basal and peak glucose values, similar basal insulin and reduced insulin responses to i.v. glucose (Table 1 and Fig. 1) in comparison to the normal subjects. While the normal controls exhibited the peak of the first phase of insulin responses at 4 min of 947.8 + 122.1 pmol/l, insulin levels rose only to 71.8 f 28.7 pmol/l in the untreated diabetics and increasing significantly to 244.1 f 79.0 ~molll after treatment. The 2nd phase of insulin secretion attained a plateau (the last 5 samples from 12-20 min) at 82.6 f 28.7 in the normals while it was 56.0 f 16.5 in the diabetic patients before treatment and increasing significantly after gliclazide to 176.7 f 17.2 pmol/l. Response to i.v. insulin (20-180
Glucose
min) (Table 2}
In the normal subjects, the dose of insulin administered (0.04 U/kg) produced mean peak insulin levels of 1852.4 f 315.9 pmol/l without hypoglycemia. Injections of 0.08 U/kg used in the diabetics gave a reliable change in the rate of glucose concentration also without hypoglycemia after inducing a peak insulin of 6476.4 f 3037.1 pmol/l in the untreated patients not significantly
different from that observed in the normals due to the dispersion of the values obtained. In the NIDDM, after gliclazide, there were no significant differences in all parameters evaluated in comparison to the correspondent control test (still statistically different from that observed in the normal subjects). Mean S’ was significantly lower in the diabetics that in the control subjects (2.37 + 0.21 vs. 47.10 f 6.68 X 10-4.min-’ .p/mol-’ . I-‘; P = 0.001). After 3 months gliclazide S’ increased significantly when compared to baseline study (13.86 f 5.24 vs. 2.37 +Z0.21 x 10e4 min-I. pmol-‘.I’; P < 0.05) but still it was lower than that found in the normal controls (P = 0.003) (Fig. 2). Baseline mean So was significantly lower in the diabetics when compared to the control subjects (2.04 f 0.21 vs. 3.36 f 0.28 x 10-2.min-‘; P = 0.08) and fell further after gliclazide to 1.72 f 0.26 x lo-‘.min-’ (P = 0.01) (Fig. 2). Discussion With a bolus infusion of insulin and using the Bergman minimal model simultaneous estimation of S’ and So were readily obtained in normal and NIDDM by Bergman and coworkers [2,3] and by Finegood et al. [lo) in normal and IDDM. In accordance with previously published reports [2,3,1 l] defects in both S’ and So contribute to the marked decrease of glucose utilization in NIDDM. We have shown that with the use of a second generation sulfonylurea, gliclazide, the
I52 60 p=O.O06
p=o.o03 I
I
I
r
I p=O.O06
p=O.OOl
I
I
I
6
50
p=o.o1
5 I
I
.A 40
-
+
E 30
d
I 0 x
20,
H ul
IO
0
CONTROL
CONTROL
NIDDM BEFORE
3mo GLICLAZIDE
GLICLAZIDE
Fig. 2. Insulin sensitivity (S,) and glucose-mediated and after 3 months gliclazide therapy. Columns
NIDDM BEFORE
3mo
glucose disposal and bars indicate
greatly improved glycemic control was associated with a small statistically significant rise in Si. Similar observations by using euglycemic insulin clamps were made with all therapeutic interventions effective in reducing the fasting plasma glucose concentration (cited in Ref. 12). Since the primary abnormality inherited in NIDDM is probably insulin resistance if the resistance from glucose toxicity is superimposed on the resistance inherited in type 2 diabetes we would
(SG) of healthy control subjects and NIDDM patients before mean f S.E.M.. P-value analysis by Mann-Whitney U-test.
not expect that normalization of the plasma glucose concentration and removal of glucose toxicity to produce major improvement in insulin sensitivity as our data have been clearly demonstrated. On the other hand, the glucose effectiveness defect is considered pathogenic for NIDDM according to Bergman and coworkers [2,3] since it declines only with the development of glucose intolerance being normal in non-diabetic offsprings of patients with NIDDM, however, presenting in-
153
sulin insensitivity of both parameters evaluated by the modified IVGTT [ 131. We have observed a significant fall in So after 3 months gliclazide therapy at the time S, increased in all to values much lower, except in one patient, to that found in our matched normal subjects. The fall in So is suggestive that the reduction in glucose effectiveness once established cannot be corrected despite normalization of plasma glucose levels at least for the period of our study. Our observation that glucose effectiveness is not normalized after sulfonylurea treatment is similar to that of Finegood et al. [lo] who have found that cyclosporine-induced complete remission of type 1 diabetes did not improve So. The improvement of glucose control with significantly lower (almost normal) fasting plasma glucose after gliclazide therapy without major improvements in St and So for the period of treatment could suggest that the amelioration of plasma glucose levels is determined at least in part to the increased insulin secretion as evaluated by the first and 2nd phases of insulin release during the first 20 min after glucose infusion. The change in insulin secretion is probably related, partially at least, to reduction in blood glucose with subsequent reduction in glucose toxicity effects on the beta cell [12]. However, we have previously demonstrated a of insulin-mediated significant improvement peripheral sensitivity (glucose clearance) after 3 months gliclazide treatment in 9 NIDDM [14] which could suggest that a small change in Si assessed in the unphysiologic conditions of the present study could result in major change in glucose concentrations. Thus, it is difficult to determine whether the alterations in insulin secretion are indeed the major cause of the lower plasma glucose concentrations. The significance and mechanism underlying the reduced So in newly diagnosed NIDDM and IDDM and not normalized during diabetes control remains to be determined. Since So is glucose-induced glucose uptake at basal glucose levels according to the minimal model, the fact that hyperglycemia itself increases glucose uptake [15] would not seem to bear on the value of So. Finally, since our diabetics received the same
amount of insulin before and after treatment then it could not affect the estimation of So, as Finegood et al. [lo] demonstrated that So was greater in those subjects receiving higher insulin infusion compared to those receiving a lower insulin infusion rate. They postulated that the dependence on the dose of insulin could be associated with the difference in nadir plasma glucose subsequent to the insulin infusion. In our study, the mean nadir glucose level was lower although non-significantly after gliclazide.
Acknowledgements Supported by grants from Institute de Recherches Internationales Servier, Neuilly-Sur-Seine, France and Conselho National de Desenvolvimento Cientitico e Tecnologico (CNPq), No. 500112/90-B, Brasilia, Brazil. The authors wish to thank Dr. R.N. Bergman for his help in the development of the modified IVGTT in our Laboratory. References Baron, A.D., Kolterman, O.G., Bell, J.. Mandarino, J. and Olefsky, J.R. (1985) Rates of noninsulin-mediated glucose uptake are elevated in Type II diabetic subjects. J. Clin. Invest. 76, 1782-1788. Bergman, R.N. (1989) Toward physiological understanding of glucose tolerance. Minimal-model approach. Diabetes 38, 1512-1527. Welch, N.S., Gebhart, S.S.P., Bergman, R.N. and Phillips, L.S. (1990) Minimal-model analysis of intravenous glucose tolerance test-derived insulin sensitivity in diabetic subjects. J. Clin. Endocrinol. Metab. 71. 1508-1518. Hoffman, W. (1937) A rapid photoelectric method for the determination of glucose in urine and blood. J. Biol. Chem. 120, 51-55. Desbuquois, B. and Aurbach, G.D. (1981) Use of polyethylene glycol to separate free and antibody-bound peptide hormones in radioimmunoassay. J. Clin. Endocrinol. Metab. 33, 732-738. Trivelli. L.A., Ranney, H.N. and Lai, H.T. (1978) Hemoglobin components in patients with diabetes mellitus. New. Engl. J. Med. 284, 353-357. Ward, W.K., Beard, I.C.. Halter, J.B., Pfeifer, M.A. and Porte Jr., D. (1984) Pathophysiology of insulin secretion in non-insulin-dependent diabetes mellitus. Diabetes Care 7, 491-502.
154 8
Bergman, R.N. and Pacini, G. (1986) MINIMOD: A computer program to calculate insulin sensitivity and pancreatic responsitivity from frequently sampled intravenous glucose tolerance test. Comput. Meth. Prog. Biomed. 23, 11-122. 9 Kuo, S.S. (1972) Computer Aplications of Numerical Methods. Reading, Massachusetts. Addison Wesley Publishing Co., pp. 275-279. 10 Finegood, D.T., Hramiak. I.M. and Dupre, J. (1990) A modified protocol for estimation of insulin sensitivity with the minimal-model of glucose kinetics in patients with Insulin-dependent diabetes. J. Clin. Endocrinol. Metab. 70, 1538-1549. II Ward, G.M., Weber, K.M., Walters, I.M. et al. (1991) A rapid modified minimal model analysis of insulin sensitivity and glucose-mediated glucose disposal in Insulindependent diabetes. Metabolism 40, 4-9.
12 Rossetti, L., Giaccari, A. and DeFronzo. R.A. (1990) Glucose toxicity. Diabetes Care 13, 610-630. 13 Osei, K., Cottrell, D.A. and Orabella, M.M. (1991) Insulin sensitivity, glucose effectiveness and body fat distribution pattern on nondiabetic offsprings of patients with NIDDM. Diabetes Care 14, 890-896. 14 Wajchenberg, B.L., Malerbi, D.A.C., Giurno Filho, A., Giannela Neto, D., Cherem, J.J. and Lerario, A.C. (1987) Effect of gliclazide treatment on red blood cell insulin receptors, hepatic glucose production and peripheral glucose utilization in non-insulin-dependent diabetes mellitus. Bull. IDF 32, 16-21. I5 Capaldo, R.. Santoro, D., Riccardi, G., Perrotti, N. and Sacca, L. (1981) Direct evidence for a stimulatory effect of hyperglycemia per se on peripheral glucose disposal in type II diabetes. J. Clin. Invest. 77, 1285-1290.