Philippe Henrivaux

Metabolic Alterations After a Two-hour Nocturnal Interruption of a Continuous Subcutaneous Insulin Infusion

In order to evaluate the metabolic consequences of a 2-h nocturnal interruption of continuous subcutaneous insulin infusion (CSII), seven insulin-dependent diabetic patients without residual insulin secretion were investigated. The changes in blood glucose, plasmafree insulin, glucagon, free fatty acids, and 3-hydroxybutyrate (3 OH-B) concentrations have been compared during two randomized tests carried out either during the normal functioning of a Mill-Hill pump from 10 p.m. to 8 a.m. (1.00 ± 0.06 U insulin/h, keeping adequate metabolic control) or during the same conditions but with a deliberate arrest of the pump between 11 p.m. and 1 a.m. Considering the value recorded at 11 p.m. asreference, interruption of the insulin infusion resulted in: (1) a rapid (already significant after 1 h) and sustained (maximal fall: –12.5 ± 2.5 mU/L at 3 a.m.) decrease in plasma free insulin; (2) a delayed (significant after 4 h) and linear rise in blood glucose (maximal increase: + 4.0 ± 1.3 mmol/L at 5 a.m.); (3) an early (significant at midnight) and prolonged rise in plasma free fatty acids (+ 387 ± 148 μumol/L at 3 a.m.); (4) a delayed (significant after 3 h) and sustained increase in plasma 3 OH-B (+ 347 ± 88 μumol/L at 3 a.m.); and (5) no significant changes in plasma glucagon. Thus, a 2-h interruption of CSII in resting nocturnal conditions is sufficient to induce significant, delayed, and sustained metabolic alterations in C-peptide-negative patients despite good baseline blood glucose control. Resetting the pump at its basal rate is insufficient to quicklyrestore adequate circulating insulin levels and effectively counteract the metabolic disturbances. The efficacy of a bolus insulin injection in these conditions should be evaluated.

Sandostatin, a new analogue of somatostatin, reduces the metabolic changes induced by the nocturnal interruption of continuous subcutaneous insulin infusion in Type 1 (insulin-dependent) diabetic patients

SummaryWith the aim of assessing a new somatostatin analogue to prevent the metabolic changes induced by a 6-h nocturnal arrest of an insulin pump, nine C-peptide negative Type 1 (insulin-dependent) diabetic patients were submitted blindly to two interruptions (from 23.00 to 05.00 hours) of their continuous s.c. insulin infusion, once after a single s.c. injection at 23.00 hours of 50 μg SMS 201-995 (Sandostatin, Sandoz) and once after 0.9% NaCl. Plasma SMS 201-995 levels peaked at 24.00 hours and then declined with an elimination half-life averaging 144±15 min. Plasma glucagon and growth hormone levels were significantly reduced after SMS 201-995 whereas the progressive fall in plasma-free insulin levels from 23.00 to 05.00 hours was unaffected. In the control test, blood glucose levels tended to decrease slightly from 23.00 to 02.00 hours and then increased markedly from 02.00 to 05.00 hours (+5.3±1.5mmol/l) while after SMS 201-995 they decreased significantly from 23.00 to 02.00 hours (−2.6±0.5 mmol/l), resulting in values below 3 mmol/l in seven subjects, but showed a secondary increase until 05.00 hours (+3.5±1.5 mmol vs 23.00h; p<0.05 vs 0.9% NaCl). While the rises in plasma non-esterified fatty acid and glycerol levels were not reduced by SMS 201-995, the increase in plasma 3-hydroxybutyrate levels, although similar from 23.00 to 02.00 hours, was significantly reduced from 02.00 to 05.00 hours (+77±20 vs+124±31 μmol·l−1·h−1p<0.005). Thus, SMS 201-995 significantly reduced the metabolic alterations due to a 6-h nocturnal interruption of a continuous s.c. insulin infusion but at the cost of a rather high risk of early hypoglycaemia.

Anti-Insulin Antibodies and Metabolic Deterioration After Interruption of Continuous Subcutaneous Insulin Infusion

Previous studies have suggested that high plasma levels of IgG anti-insulin antibodies were able to store insulin during insulin administration and to release it after insulin cessation (reviewed in ref. 1). Consequently, the blood glucose rise after interruption of an intravenous insulin infusion and the blood glucose recovery after insulin-induced hypoglycemia were significantly delayed and reduced in patients with type I, insulin-dependent diabetes mellitus (IDDM) with high plasma concentrations of IgG anti-insulin antibodies. We have previously reported that a 6-h nocturnal interruption of a continuous subcutaneous insulin infusion (CSII) induced a rapid and severe metabolic deterioration in IDDM patients without endogenous insulin secretion and with low plasma levels of IgG anti-insulin antibodies. During 4 yr of chronic treatment with CSII, three of the patients showed progressive, sustained, and unexplained increases in IgG antiinsulin antibody plasma levels, and this evolution was not reversed after switching from porcine insulin (Actrapid MC, Novo, Copenhagen, Denmark) to semisynthetic human insulin (Actrapid HM, Novo). Clinical characteristics were as follows at the beginning of CSII: two women and one man; aged 33, 51, and 59 yr; suffering from IDDM since 13, 18, and 6 yr of age; and weighing 106, 119, and 88% of ideal body weight, respectively. None showed a response of plasma C-peptide after glucagon. HLA-DR typing showed DR4/6, DR4/4, and DR6/7, respectively. CSII treatment used a Mill Hill HM 1001 pump with insulin (40 U/ml) diluted in saline (one patient) or in Novo diluting medium (two patients). Mean daily insulin requirements averaged 0.76 ± 0.14 U/ kg, with a basal insulin delivery rate of 1.1 ± 0.07 U/h. Each patient underwent a new 6-h nocturnal CSII interruption from 2300 to 0500 h with the same protocol as he or she had 4 yr previously. Blood glucose and plasma 3-hydroxybutyrate levels were measured every hour from 2200 to 0600 h in both tests. The only obvious difference between the two experimental conditions was the plasma concentration of IgG anti-insulin antibodies, measured by the method of Christiansen: low in the first test (0.31 ± 0.02 U/L) and high in the second test (7.42 ± 0.25 U/L). When high IgG anti-insulin antibody plasma levels were present, blood glucose remained stable during the first 3 h of CSII interruption (vs. only 2 h in the first test), and the 1200-

Prevention of metabolic alterations by insulin supplements administered either before or after 2-h nocturnal interruption of CSII.

To evaluate the efficacy of a bolus insulin injection to prevent the metabolic alterations induced by a 2-h nocturnal interruption of a continuous subcutaneous insulin infusion (CSII), nine type I (insulindependent) C-peptide-negative diabetic patients were studied from 2200 to 0800 h during two randomized tests. An insulin bolus (2.1 ± 0.2 U) was administered via the pump either at 2300 h, just before CSII interruption, or at 0100 h, after reactivating the pump at its usual basal rate (1.05 ± 0.11 U/ h). The insulin bolus at 2300 h induced a significant rise in plasma free-insulin levels at 2400 h (+ 6.9 ± 1.8 mU/L, P < .01), resulting in an early and marked fall in blood glucose concentrations between 2300 and 0100 h (−2 .7 ± 0.5 mM, P < .001), with hypoglycemic values in five patients. The insulin bolus at 0100 h counteracted the fall in plasma free-insulin levels observed between 2300 and 0100 h and significantly increased plasma insulin at 0200 h (+ 3.2 ± 0.8 mU/L, P < .01). Blood glucose concentrations that remained stable during the 2-h arrest of the pump fell significantly between 0100 and 0400 h (- 2.1 ± 0.5 mM, P < .005). This fall rate was significantly lower than that measured within the 3 h after the insulin bolus given before CSII interruption but significantly higher than that observed in a reference control group of patients whose pump was functioning normally throughout the night. Thus, an insulin bolus equivalent to the amount of insulin not infused during a 2-h CSII interruption is able to prevent the metabolic alterations resulting from the pump arrest. However, to reduce the risk of hypoglycemia, the insulin bolus should preferably be administered after rather than before the 2-h CSII interruption.

U-100 Insulin Gives Some Protection Against Metabolic Deterioration Due to CSII Interruption

We investigated the influence of insulin concentration within the insulin pump on the metabolic and plasma free-insulin changes induced by a 6-h nocturnal interruption of continuous subcutaneous insulin infusion (CSII) in five C-peptide-negative insulin-dependent diabetic patients with low circulating levels of anti-insulin antibodies. We compared the changes in blood glucose, plasma free fatty acids, 3-hydroxybutyrate, and free insulin during the interruption from 2300 to 0500 h of the Nordisk Infuser loaded with either U-100 or U-20 regular insulin. The decrease in plasma free-insulin levels was slower, resulting in a significantly delayed and smaller increase in blood glucose levels (2.4 ± 1.6 vs. 7.6 ± 2.9 mM, P < .025) when the pump contained U-100 instead of U-20 insulin. Although the increases in levels of plasma free fatty acids were similar in both tests, the rise in plasma 3-hydroxybutyrate levels tended to be reduced with U-100 insulin (414 ± 139 vs. 639 ± 67 μM, P < .10). Thus, our observations indicate that U-100 insulin gives some protection against the metabolic deterioration due to the interruption of CSII so that diabetic patients may be able to remain without the pump for longer periods with concentrated rather than diluted insulin.

Lack of Systematic Metabolic Alterations After a One-hour Interruption of Continuous Subcutaneous Insulin Infusion in Type I Diabetic Patients

Five male insulin-dependent diabetic patients (age range 46-68 yr) with chronic renal failure (serum creatinine range 1.97-6.66 mg/dl) were hospitalized in a metabolic ward for the entire duration of the study. They were given, in random order and for consecutive periods of 10 days, three isoenergetic diets composed exclusively of natural foods: a low-fiber diet with 50% carbohydrate (CHO), 38% fat, and 12% protein; a high-fiber diet, identical to the low-fiber diet but with an increased fiber content (65 versus 27 g/day); a low-protein; a high-fiber diet, identical to the low-fiber diet but with an increased fiber content (65 versus 27 g/day); a low-protein-low-CHO-low-fiber diet (9% protein, which is <50 g/ day, 40% CHO, 51% fat, 21 g fiber). Fiber was mainly of stant in the three diets, and the intake of dietary phosphate was always < 1 g/day. The intake of other major electrolytes did not vary with the three diets. The insulin dosage and all other experimental conditions were kept constant throughout the study. No other treatment for renal failure was used. The results obtained at the end of each dietary period are reported in Table 1. Our results show that a high-fiber diet improves blood glucose control in diabetic patients with uremia. This is almost exclusively due to its high fiber content, which is also able to reduce atherogenic lipoproteins (not measured in this study for the short period of observation). Moreover, the high-fiber diet has no deleterious effect on blood urea and nutritional status (body weight, plasma albumin and transferrin, plasma concentration of essential amino acids, and the valine/glycine ratio), which were not modified by each of the three diets. Conversely, intake of a high-fiber diet lowers serum creatinine levels, probably due to reduction of muscle protein catabolism, which is secondary to the improved blood glucose control. This observation is noteworthy since diabetes and chronic renal failure cooperate in aggravating the catabolic state, which, in turn, deteriorates the uremic condition. Our preliminary data justify more prolonged trials on a larger scale. If these results are confirmed, the high-fiber diet might become the diet of choice for diabetic patients with uremia. We thank Cilag Foundation for Therapeutic Research for financial support.

B-cell response to a standardized breakfast in end-stage renal failure.

SummaryTwelve uremic patients (U) on regular hemodialysis were submitted to a standardized test meal. In comparison with normal controls (C), U patients demonstrated a slight increase of HbA1 level and a definite elevation of fasting plasma C-peptide immunoreactivity. They showed glucose intolerance at 60 and 120 min. This was associated with an inappropriate insulin response as evidenced by a significantly lower insulin/glucose index at 60 min. U patients were tested again during a hemodialysis session in order to reduce the 60-min glucose intolerance. Six patients (U1) were selected because they exhibited mean fasting and 60-min glucose values similar to those of the controls. In these conditions, the insulin response at 60 min was significantly decreased in comparison to basal conditions and this could not be accounted for by a concomitant decrease of plasmaα-aminonitrogen values. It is concluded that, in uremic patients, glucose intolerance is associated with an inappropriately low B-cell response.