Diabetes Research and Clinical Practice, 13 (1991) 69-76

0 1991 Elsevier Science Publishers

69

B.V. 016%8227/91/$03.50

DIABET 00521

Could intranasal insulin be useful in the treatment of non-insulin-dependent diabetes mellitus ? Renate Kimmerle, Section of Endocrinology

George Griffing *, Anthony McCall, Neil B. Ruderman, James C. Melby and Metabolism, Evans Memorial Department of Clinical Research, University Medical Center, Boston, Massachusetts,

Ed Stoltz and

University Hospital, Boston

U.S.A.

(Received 19 December 1990) (Revision accepted 2 April 1991)

Summary To evaluate whether intranasal insulin might be useful as a meal-adjunct in the treatment of NIDDM we compared plasma glucose and insulin responses to a mixed breakfast (9 kcal/kg, 50% carbohydrate) following either intranasal insulin (INI) or placebo in eleven patients with NIDDM. Five patients treated with subcutaneous insulin and in good to moderate glycemic control and six patients who were ‘failing’ on oral agents and in poor glycemic control were studied. In the patients usually on SC insulin, IN1 inhibited postprandial hyperglycemia. Lower doses (1 U/kg vs 1.5 U/kg b.w.) were needed to accomplish this in 2 patients with low fasting glucose (I 7.8 mmol/l) than in three patients with higher fasting glucose (10.5 f. 0.5 mmol/l). In the patients on oral agents who had marked fasting hyperglycemia (14.8 + 0.8 mmol/l) an only transient reduction (for 90 to 120 min) of postprandial hyperglycemia was achieved when IN1 (1 U/kg) was given in addition to po glyburide (10 mg) prior to the meal. Following placebo in the group previously treated with SC insulin, plasma free insulin levels increased maximally by 23 mu/l, 75 min after the meal. The group on oral agents had a comparable but later peak increment (at 180 min) indicative of an even greater impairment of endogenous insulin secretion in response to hyperglycemia. Following INI, the peak increment in plasma insulin occurred earlier (30 min after the meal) and was greater in all patients (55 2 18, 139 + 68,86 k 24 mU/l respectively for the prior SCinsulin therapy group at doses of 1.0 and 1.5 U/kg and for the oral agent group at 1.0 U/kg). In summary, intranasal insulin can reduce postprandial hyperglycemia in patients with NIDDM by increasing early postprandial insulin levels. It is more effective in those with less severe fasting hyperglycemia. Since elevated insulin levels and glucose lowering effect are transient, nasal insulin would need to be added to a regimen achieving fasting glucose control through diet, oral agents or long-acting subcutaneous insulin. Key words: Nasal drug therapy;

Insulin; NIDDM

Correspondence to: R. Kimmerle, Dept. ofNutrition and Metabolic Diseases, WHO Collaborating and Treatment, Heinrich Heine University, Moorenstr. 5, 4000 Diisseldorf, F.R.G. * G. Griffing is a Recipient of a New Investigators Award from the NIH.

Center for Diabetes Prevention

70

Introduction Nasal insulin in combination with an absorption promoting agent (surfactant) is rapidly absorbed across the nasal mucosa [ l-61. Peripheral insulin levels following nasal insulin more closely mimic the normal insulin response to a meal than does subcutaneous regular insulin [7]. In insulindependent diabetics nasal insulin can blunt postprandial hyperglycemia [5-91. Because of its rapid onset and brief duration of action intranasal insulin might also benefit patients with NIDDM because it should correct their delayed and relative insufficient insulin response to a meal while minimizing late hypoglycemia [ 10-131. However, because of the heterogeneity of type 2 patients as to the degree of insulin secretory capacity and insulin resistance, the effects of nasal insulin will be more variable than in type 1 patients [lo]. In the two previous studies on the meal-time use of nasal insulin in NIDDM it was shown that nasal insulin/glycocholate aerosol could inhibit postprandial hyperglycemia. However, in one study, data on the patients’ overall glycemic control are not given [ 131 and in the other study only a specific group oftype 2 patients, i.e. non-obese patients in moderate glycemic control were studied [ 151.

In the present study we examined the effectiveness of nasal insulin when given as one dose before the meal in addition to a sulfonylurea in patients who were in poor glycemic control on oral agent treatment. We chose to study these patients and additional patients who were in good to moderate glycemic control on subcutaneous insulin treatment, since in clinical practice such patients often desire the avoidance or reduction of insulin injections. The results of our studies give an indication in which type of patient and in which therapeutic setting nasal insulin might be useful as a meal adjunct in the treatment of NIDDM.

Methods PATIENT

POPULATION

Eleven patients were studied after obtaining their informed consent. Five patients had been on subcutaneous insulin treatment (see Table 1) and had a hemoglobin A, of 8.7 + 1.1 y0 and a fasting glucose level of 9.3 k 0.8 mmol/l. The average age was 49 k 4 yrs, duration of diabetes 6 f. 2 yrs and percent ideal body weight 123 t 9. The six

TABLE 1 Characteristics Patient

of NIDDM M/F

patients

Age

on subcutaneous Duration of diabetes

insulin treatment*

Weight (kg)

(yrs) 1 2 3 4 5 Mean SE

F M M F F

48 45 62 54 39 49 3.5

2 2 8 15 4 6.2 2.2

53 115 78 71 83 80 9

Percent ideal body weight

Fasting plasma glucose (mmol/l)

(%) **

Current treatment (units/day) breakfast-supper

104 157 120 104 130

7.1 7.8 9.8 11.8 10.0

8.0 6.3 8.0 7.7 13.3

NPH 12 NPH 15 UL 34/L 10 UL 16/L 18 NPH 30/R 10-R 10

9.3 0.8

8.7 1.1

123 8.5

UL denotes Ultralente, R Regular, L Lente, and NPH Neutral Protein Hagedorn * Laureth-9 at 1% was given as the surfactant in the nasal insulin aerosol. ** Glycosylated hemoglobin, range for non-diabetic controls, 4.9 to 7.6%.

Insulin.

HbA,

71

patients on oral agent treatment were referred to us because they were ‘failing’ on diet and maximal doses of a sulfonylurea agent and because they were reluctant to start subcutaneous insulin treatment (hemoglobin A, 11.2 f 0.5 %, fasting glucose level 14.8 & 0.8 mmol/l). They had a mean age of 61 k 4yrs, duration of diabetes of 7.3 f 2 yrs and were 120 + 3% of ideal body weight (see Table 2). All patients studied had NIDDM defined by the onset of diabetes after age 30, no history of ketoacidosis and the absence of insulin treatment within 1 year of diagnosis. Patients with serious underlying disease and SC insulin requirements of > 1.5 U/kg b.w. per day were excluded.

aerosol solution (U 300 to U 450) and by giving multiple sprays. Laureth-9 (polyoxyethylene-9 lauryl ether, PVO International, Boonton, N.J.) at 1% was used as the surfactant in the SC insulin treatment group. The oral agent treatment group received Tween 80/nonoxynol-9 (polyoxyethylene sorbitan monooleate/polyoxyethylene nonyl-phenyl ether, (Igepal CO-630, GAF)) (1 : 3) 0.5% as the surfactant [ 16-181. This surfactant blend (evaluated by us in normal volunteers) is similar to Laureth-9 1 y0 in terms of insulin absorption promoting potency and does not cause nasal irritation (unpublished results). Laureth-9 at 1% causes transient nasal stuffiness [6]. MEASUREMENTS

DELIVERY

DEVICE

AND MEDICATION

Doses of aerosolized insulin metered at 100 pi/puff (coefficient of variation + 10%) were delivered by a specially constructed non-vented mechanical pump valve (Pharmasol Corp., MA). The aerosol solution contained U-500 regular porcine insulin (Eli Lilly, Indianapolis), 0.01 M phosphate buffer, surfactant, and nitrous oxide as the propellant. Dosages of insulin were adjusted by varying the insulin concentrations in the

Plasma glucose was measured on a Yellow Springs glucose analyser (YSF 23 A). Plasma insulin was measured in duplicate in antibody coated assay tubes (Micromedic, Horsham, PA) on thawed samples which had been incubated for 2 h at room temperature after storage at - 20°C. Plasma free insulin (in the SC insulin treatment group) was measured following PEG precipitation [ 19,201. Intra-assay variation was 7.7% and interassay variation was 9.5 %. Hemoglobin

TABLE 2 Characteristics Patient

of NIDDM M/F

patients

Age (yrs)

on oral agent treatment* Duration of diabetes

Weight

Percent ideal body weight

Fasting plasma glucose (mmol/l)

(%)

Current treatment

64 14 84 91 74 98

123 128 118 116 108 130

13.3 15.2 14.2 13.8 18.6 13.8

11.4 12.3 10.2 9.8 13.0 10.5

Glyburide Glyburide Glyburide Glyburide Glyburide Glyburide

80 4.5

120 3

14.8 0.8

11.2 0.5

(kg)

(yrs) 1 2 3 4 5 6 Mean SE

F F M M M F

61 45 69 51 70 59 61 3.6

2 4 I 6 16 9 1.3 1.8

* Tween SO/Nonoxynol-9 (1 : 3) 0.5% was given as the surfactant. Glycosylated hemoglobin A,, range for non-diabetic controls, 4.9 to 7.6%.

HbA,

20 20 20 20 20 20

mg/d mg/d mg/d mg/d mg/d mg/d

72

A, was measured by ion exchange chromatography with the use of microcolumns (Isolab, Akron, OH) [21]. STUDY DESIGN Patients were admitted to the metabolic unit on the morning of the study after an overnight fast having omitted their insulin or oral agents for at least 14 h. While seated, an intravenous line for blood drawing was inserted in the forearm and kept open with 0.45% saline solution. After this the subjects ate a mixed breakfast (eggs, toast, cereal, milk and orange juice, 50% carbohydrate, 35% fat, 15% protein (9 kcal/kg)) over 20 min. On different mornings immediately before the meal they received, in random order, nasal insulin at 1 U/kg or placebo. Patients treated with SC insulin did not receive SC insulin during the meal studies. In this group, three patients with a 2-h postprandial glucose increment of > 2.2 mmol/l following 1 U/kg received nasal insulin at 1.5 U/kg (patients 3-5, Table 1). Patients previously treated with oral agents received glyburide 10 mg po 30 min before the meal. Blood samples for determination of plasma glucose and insulin were drawn over four h at frequent intervals from the indwelling intravenous line. DATA

EVALUATION

AND STATISTICAL

ANALYSIS

Results are reported as group means + S.E.M. or as individual values if groups were small. Analysis of variance adjusted for repeated measures followed by Dunnett’s t-test or for paired data Student’s test was used to assess statistical significance. A value of P < 0.05 was considered significant [ 221.

Results PRIOR

SUBCUTANEOUS

INSULIN

TREATMENT

GROUP

Glycemic response

As shown in Fig. 1, intranasal insulin (1 U/kg given with Laureth-9 1% as a surfactant) signifi-

” 0 1530

.

1’ 60

90

120

160

240

MINUTES

Fig. 1. Plasma glucose (upper panel) and free insulin levels (lower panels) in response

to a mixed meal (9 kcal/kg, 50% carbohydrate) in five NIDDM patients on prior SC insulin treatment (patients 1-5, Table 1) after administration of intranasal placebo (shaded area) and intranasal insulin (INI) 1.0 U/kg (filled circles). Three patients (patients 3-5, Table 1) with a 2-h postprandial plasma glucose increment of > 2.2 mmol/l received IN1 1.5 U/kg (open circles).

cantly reduced postprandial glucose excursions in the group previously treated with SC insulin. The 2-h increment in plasma glucose averaged 3.3 + 1.6 mmol/l vs 6.4 k 1.7 mmol/l following placebo. In two patients whose fasting plasma glucose were concentrations 57.8 mmol/l (patients 1 and 2, Table 1) nasal insulin caused a decrease of glucose of 0.6 and 0.5 mmol/l vs an increase of 3.0 and 3.1 mmol/l when placebo was administered. In three patients (patients 3-5, Table 1) with higher fasting glucose concentrations (10.5 2 0.5 mmol/l) only a minor improvement in postprandial hyperglycemia occurred at the 1 U/kg dose of nasal insulin. When these

73 patients received 1.5 U/kg before the meal however, the 2-h glucose declined by 0.9 k 1.4 mmol/l vs an increase of 8.3 + 2.1 mmol/l following placebo (P -C 0.01). Insulin response As shown in Fig. 1, plasma levels of free insulin after administration of placebo increased by a maximum of 23 + 15 mU/l above the fasting level of 22 + 7 mU/l at 60 to 90 min. In contrast, after nasal insulin administration, free insulin levels increased earlier (at 30 min) and to a greater degree (maximum increments were 55 + 18 and 139 + 68 mU/l respectively for 1 U/kg and 1.5 U/kg), (P < 0.05).

ORAL

AGENT

TREATMENT

GROUP

Intranasal insulin (1 U/kg) plus 10 mg glyburide reduced postprandial hyperglycemia compared to placebo plus glyburide during the first 90 min after the meal (Fig. 2), (P < 0.05). Insulin levels with placebo plus glyburide increased maximally by 27 k 12 mU/l above the fasting level of 18.5 & 5 mU/l at 180 min postprandially. In contrast, when intranasal insulin was administered, insulin levels peaked earlier (at 30 min after the meal) and were higher (86 + 24 mU/l above the fasting level). Insulin levels remained above those seen with placebo for 90 min. Between 120 and 240 min, both plasma insulin and glucose levels were similar with nasal insulin and placebo.

Discussion

01, 0

(

15

30

I

I

I

I

45

60

QO

120

I 180

I 240

MINUTES

Fig. 2. Plasma glucose (upper panel) and insulin levels (lower panel) in response to a mixed meal in six NIDDM patients on oral agent treatment (patients l-6, Table 2) after administration ofintranasal placebo/glyburide (shaded area) and intranasal insulin (INI) 1 U/kg/glyburide (tilled circles). The patients received glyburide 10 mg po 30 min before the meal.

These studies indicate that nasal insulin administered as one dose at the beginning of a test-meal can reduce postprandial hyperglycemia in patients with NIDDM. In patients with mild to moderate fasting hyperglycemia nasal insulin at 1.0 and 1.5 U/kg lowered postprandial glucose levels in an acceptable range (8-12 mmol/l). In patients with uncontrolled fasting glucose levels (14.8 mmol/l) postprandial glucose levels, although reduced as compared to placebo, remained unacceptably high. The administration of nasal insulin before the meal resulted in an early and high peak of insulin levels and thus corrected for the delayed and insufficient endogenous insulin response that followed the administration of placebo [ 11,141. In the patients on oral agents, also a second peak was observed reflecting their endogenous insulin response to still marked hyperglycemia [ 23-251. The glucose lowering effect of nasal insulin lasted only two h. Therefore, it is doubtful whether nasal insulin with meals would make a clinically significant contribution to overall 24-h glycemic control in patients with uncontrolled fasting hyperglycemia. In such patients, the duration of action of nasal insulin given before the

74

meal is too short to suppress endogenous glucose production in the postabsorptive state which can be effectively lowered only by improving ‘basal insulinisation and insulin sensitivity [ 131. Complete and prolonged inhibition of postprandial hyperglycemia with nasal insulin was achieved in patients with low fasting glucose levels and in those patients with moderately elevated fasting glucose levels in whom a higher dose (1.5 U/kg) of nasal insulin was used. In a study of nasal insulin in non-obese type 2 patients moderate fasting hyperglycemia exhibiting (10 mmol/l), Frauman et al. could achieve prolonged inhibition of postprandial hyperglycemia by giving nasal insulin before, 30 min and 60 min after the meal at a total dose of 90 units, a dose comparable to the 1 U/kg we used in our study patients [ 151. Although larger doses and repeated administration might render nasal insulin effective even in patients with high fasting glucose levels it may then become too expensive and lose its potential advantage as a meal-adjunct over SC regular insulin, i.e., its ease of administration [ 5,6]. Furthermore, insulin at high concentrations (to limit the volume delivered to the nose) tends to form large aggregates which can plug the delivery device and which might be poorly absorbed across the nasal mucosa. In summary, nasal insulin can reduce postprandial hyperglycemia in patients with NIDDM and moderate fasting hyperglycemia. Unless its bioavailability (I/I I to i/15 of iv insulin) [ 5,6,9] and duration of action are improved, its use in patients with high insulin requirements, for example in patients ‘failing’ on oral agents, does not seem practical. However, nasal insulin might prove useful as a meal-time adjunct in NIDDM patients in whom fasting hyperglycemia is already controlled through diet, oral hypoglycemic agents or SC insulin [ 13,23-261. Acknowledgements This work was supported by grants-in-aid l-ROl-HL-31-51, RR-533-NIH, 5-P-HL-18318,

5-ROI-AM12027, 5-P32-HL-07224, 401-AM28164, ADN-5-50-HL18318, NS-22213, and HL-26895 from the NIH. The authors wish to thank Mary Mulready, Kathy Mullin, Jackie Allen, and Monika Holbrook, for their expert assistance.

References 1 Lee, SW., Sciarra, J.J. (1976) Development of an aerosol dosage form containing insulin. J. Pharmacol. Sci. 65, 561-512. 2 Hirata, Y., Yokusuka, T., Kasahara, T., Kikchi, M. and Ooi, K. (1979) Nasal insulin administration in patients with diabetes. Symposium on Proinsulin, Insulin and C-Peptide, 1978 Tokushima, Japan. Amsterdam Excerpta Medica International Congress Series 468, 319-326. 3 Hirai, S., Yashiki, T. and Mima, H. (1981) Effects of surfactants on the nasal absorption of insulin in rats. Int. J. Pharm. 9, 165-172. 4 Hankiss, J. (1982) Insulin given intranasally induces hypoglycemia in normal and diabetic subjects. Br. Med. J. 284, 1707-1708. 5 Moses, A.C., Gordon, G.S., Cary, M.C. and Flier, J.S. (1983) Insulin administered intranasally as an insulin bile salt aerosol: effectiveness and reproducibility in normal and diabetic subjects. Diabetes 32, 1040-1047. 6 Salzman, R., Manson, J.E., Griffing, G.T., Kimmerle, R. et al. (1985) Intranasal aerosolized insulin: mixed meal studies and longterm use in type 1 diabetes. N. Engl. J. Med. 312, 1078-1084. 1 Binder, A.U., Lauritzen, T., Faber, 0. and Praming, S. (1984) Insulin pharmacokinetics. Diabetes Care 7, 188-199. 8 Pontiroli, A.E., Albaretto, M., Secchi, A., Dossi, G., Bosi, I., Pozza, G. (1982) Insulin given intranasally induces hypoglycemia in normal and diabetic subjects. Br. Med. J. 284, 303-306. 9 Fraumann, A.G., Cooper, M.E., Parsons, B.J., Jerums, G. and Louis, W.J. (1987) Longterm use ofintranasal insulin in insulin dependent diabetic patients. Diabetes Care 10, 573-578. 10 Pfeifer, M.A., Halter, J.B., Porte, D. (1981) Insulin secretion in diabetes mellitus. Am. J. Med. 70, 579-588. 11 Vague, P. and Moulin, J.P. (1982) The defective glucose sensitivity of the &cell in non-insulin-dependent diabetes; improvement after twenty hours of normoglycemia. Metabolism 31, 139-142. 12 Calles-Escandon, J., Jaspan, J. and Robbins, DC. (1989) Postprandial oscillatory patterns of blood glucose and insulin in NIDDM. Diabetes Care 12, 709-714.

75 13 Turner, R. and Holman, R.R. (1990) Insulin use in NIDDM. Rationale based on pathophysiology ofdisease. Diabetes Care 13, 101 l-1020. 14 El-Etr, M., Slama, G. and Desplanque, N. (1987) (Letter to the editor) Preprandial nasal insulin as adjuvant therapy in type 2 diabetics. Lancet 7, 1085-1086. 15 Fraumann, A.G., Jerums, G. and Louis, W.J. (1987) Effects of intranasal insulin in non-obese type 2 diabetics. Diabetes Res. Clin. Pratt. 3, 197-202. 16 Elworthy, P.H., Treon, J.F. (1967) Biology of nonionic surfactants. In: Schick, M., (Ed.), Nonionic Surfactants. Dekker, New York. 17 Helenius, A., McCaslin, D.R., Fries, E. and Tanford, C. (1979) Properties of detergents. Methods Enzymol. 56, 734-749. 18 Mukerjee, P., Mysels, K. (1971) Critical micelle concentration of aqueous surfactant systems. Washington, DC: US Government Printing Office. 19 Kuzuya, H., Blix, P.M., Howitz, D.C., Steiner, D.F. and Rubenstein, A.H. (1979) Determination of free and total insulin and C-peptide in insulin treated diabetics. Diabetes 26, 23-29. 20 Harming, I., Home, P.D. and Alberti, K.G.M.M. (1985) Measurement of free insulin concentrations: the influence of the timing of the extraction of insulin antibodies. Diabetologia 28. 83 l-835.

21 Abraham, EC., Huff, T.A., Cope, N.D., Wilson, J.B.Jr., Bransome, E.D. and Huisman, T.H.J. (1978) Determination ofthe glycosylated hemoglobin (HbA,) with a new microcolumn procedure. Diabetes 27, 93 l-937. 22 Snedecor, G.W. and Cochran, W.G. (1972) Statistical Methods Sixth Edit., Ames, Iowa State University Press. 23 DeFronzo, R.A., Ferrannini, E. and Koivisto, V. (1983) New concepts in the pathogenesis and treatment of noninsulin-dependent diabetes mellitus. In: G. Reaven (Ed.), Proceedings of a Symposium: The Role of Insulin Resistance in the Pathogenesis and Treatment of Non-insulindependent Diabetes Mellitus. Am. J. Med. 74, 52-81. 24 Tsalikian, E., Dunphy, T.W., Bohannon, N.V., Lorenzi, M., Gerich, J.E., Forsham, P.H., Kane, J.P. and Karam, J.H. (1977) The effect of chronic oral antidiabetic therapy on insulin and glucagon responses to a meal. Diabetes 26, 314-321. 25 Garvey, W.T., Revers, R.R., Kolterman, O.G., Rubenstein, A.H. and Olefsky, J.M. (1985) Modulation of insulin secretion by insulin and glucose in type 2 diabetes mellitus. J. Clin. Endocrinol. Metab. 60, 559-567. 26 U.K. Prospective study of therapies of maturity onset diabetes (1983) Effect of diet, sulfonylurea, insulin, or biguanide therapy on fasting plasma glucose and body weight over one year. Diabetologia 24, 404-4 11.

Could intranasal insulin be useful in the treatment of non-insulin-dependent diabetes mellitus?

To evaluate whether intranasal insulin might be useful as a meal-adjunct in the treatment of NIDDM we compared plasma glucose and insulin responses to...
619KB Sizes 0 Downloads 0 Views