The effect of exogenous glucagon-like peptide-1 on the glycaemic response to small intestinal nutrient in the critically ill: a randomised double-blind placebo-controlled cross over study

Seven critically ill adult patients (four males, three females, age range 28 to 76 years), predicted to remain mechanically ventilated for more than 48 hours, were studied. Exclusion criteria included pregnancy, pre-existing diabetes, contraindication to enteral feeding or post-pyloric catheter insertion, and previous surgery on the oesophagus, stomach or duodenum.

The study was approved by the Human Ethics Committee of the Royal Adelaide Hospital and performed according to the Australian National Health and Medical Research Committee guidelines for the conduct of research involving unconscious persons. Written, informed consent was obtained from the next of kin.

Patients were studied on two consecutive days, in which they received intravenous GLP-1 or placebo in a randomised, double-blind fashion. Twelve hours prior to the study, a naso-duodenal feeding catheter was inserted and confirmed via abdominal x-ray. On each study day enteral feeding was ceased at least four hours prior to the commencement of the study. Exogenous insulin (Actrapid, Novo-Nordisk, Copenhagen, Denmark) infusion was ceased at least two hours before the commencement of the study drug. Patient weight was provided by a relative and/or estimated by a dietician. Randomisation and reconstitution of synthetic GLP-1-(7–36) amide (Merck Biosciences, Melbourne, Australia), as a solution in 4% albumin, was performed by the Royal Adelaide Hospital Pharmacy department. The study investigators were blinded to each infusion, which were concealed in a glass bottle covered by black plastic. Both GLP-1 (1.2 pmol/kg/min) and placebo (4% albumin) were infused at a rate of 1 ml/min via a central venous catheter for 270 minutes (i.e. t = 0 to 270 min). At t = 30 minute Ensure^® (Abbott, Melbourne, Australia), a mixed nutrient liquid (64% carbohydrate, 1 kcal/ml), was delivered continuously into the small intestine at a rate of 1.5 ml/min for four hours (i.e. t = 30 to 270 min). An arterial blood sample was obtained every 15 minutes for measurement of blood glucose and at timed intervals for measurements of plasma insulin, GLP-1 and glucagon concentrations.

Blood glucose was measured at the bedside using a portable glucometer (Medisense Optimum, Abbott, Melbourne, Australia). Blood was collected, separated by centrifugation and the resulting plasma was stored at -70° until assayed for hormone concentrations. Plasma insulin was measured by ELISA (Diagnostics Systems Laboratories, Webster, Texas, USA) with an inter-assay coefficient of variation (CV) of 6.2%. Total plasma GLP-1 (GLPIT-36HK Linco Research, St. Charles, Missouri, USA) and glucagon (Siemens Medical Solution Diagnostics, Berkeley, California, USA) concentrations were measured by radioimmunoassay, with a CV of 9.2% and 12%, respectively.

Data are presented as mean ± standard error of the mean. Area under the curve (AUC) was calculated using the trapezoidal rule. The number of patients required to establish a glucose lowering effect of GLP-1 was based on power calculations derived from our previous work [8]. The ratios of insulin/glucose were calculated, as described previously [10]. Statistical analyses were performed using SPSS (Version 15.0, Chicago, Illinois, USA). Distribution and sensitivity analysis, using nonparametric analyses, allowed parametric testing of data. The difference between intervention and placebo was assessed using paired samples; Student's paired t-test and repeated measures analysis of variance. Data were evaluated for potential carry over effect. The null hypothesis was rejected at the 0.05 significance.

Blood glucose concentrations are shown in Figure 1. There was no difference in baseline blood glucose concentration prior to each infusion (t = 0 min GLP-1 7.5 ± 0.4 mmol/l vs. placebo 7.6 ± 0.6 mmol/l; P = not significant (NS)). Prior to the commencement of the small intestinal nutrient infusion (t = 30 min) GLP-1 had no effect on blood glucose. On both days, there was an increase in blood glucose concentration in response to intra-duodenal nutrient infusion (t = 180 min GLP-1 9.3 ± 0.6 mmol/l; placebo 12.2 ± 0.9 mmol/l; P < 0.01 for both). GLP-1 markedly attenuated the rise in blood glucose (e.g. t = 60 min GLP-1 7.5 ± -0.5 mmol vs. placebo 9.5 ± -0.8 mmol; P < 0.01), peak blood glucose (GLP-1 10.1 ± 0.7 mmol/l vs. placebo 12.7 ± 1.0 mmol/l; P < 0.01) and decreased the overall glycaemic response (AUC_{30–270 min} GLP-1 2077 ± 144 mmol/l min vs. placebo 2568 ± 208 mmol/l min; P = 0.02).

Plasma insulin and insulin:glucose ratio is shown in Figures 2a and 2b. There was no difference in plasma insulin concentration at baseline. Plasma insulin increased in response to intra-duodenal nutrient (e.g. t = 270 min GLP-1 79 ± 21 mU/l and placebo 61 ± 17 mU/l; P < 0.03 compared with fasting concentration for both days). At 270 minutes, the insulin/glucose ratio was greater with GLP-1 (GLP-1 9.1 ± 2.7 vs. placebo 5.8 ± 1.8; P = 0.02); however, there was no difference in absolute plasma insulin concentrations throughout the entire study period (AUC_{0–270 min} GLP-1 16,203 ± 5193 mU/l min vs. placebo 14,434 ± 4561 mU/l min; P = NS).

Plasma GLP-1 concentrations are shown in Figure 2c. Fasting plasma GLP-1 concentrations were similar between groups (t = 0 min GLP-1 36 ± 10 pmol/l vs. placebo 38 ± 11 pmol/l; P = NS). Exogenous GLP-1 markedly increased plasma GLP-1 concentration within 30 minutes (t = 30 min GLP-1 124 ± 15 pmol/l vs. placebo 43 ± 9 pmol/l; P < 0.01) and throughout the infusion period (AUC_{0–270 min} GLP-1 31,659 ± 4203 pmol/l min) vs. placebo 10,399 ± 2508 pmol/l min; P < 0.01). During GLP-1 infusion, steady state concentrations were achieved after 30 minutes (t = 30 min plasma GLP-1 124 ± 15 pmol/l vs. t = 90 min plasma GLP-1 131 ± 13 pmol/l; P = NS).

Plasma glucagon concentrations are shown in Figure 2d. Fasting glucagon concentrations were similar between study days (t = 0 min GLP-1 106 ± 14 pmol/ml vs. placebo 102 ± 13 pmol/ml; P = NS). There was a decrease in plasma glucagon from baseline during GLP-1 infusion (t = 30 min GLP-1 90 ± 12 pmol/ml vs. placebo 104 ± 10 pmol/ml; P < 0.01), which was non-sustained (AUC_{0–270 minutes} GLP-1 22,786 ± 6040 pmol/l min vs. placebo 25,830 ± 2644 pmol/l min); P = NS).