Biomedical subjects
Publications and source records attributed to Ludovic J Chassin.
Explore the source record for details and available documents.
OBJECTIVE: To evaluate a fully automated algorithm for the establishment of tight glycemic control in critically ill patients and to compare the results with different routine glucose management protocols of three intensive care units (ICUs) across Europe (Graz, Prague, and London). RESEARCH DESIGN AND METHODS: Sixty patients undergoing cardiac surgery (age 67 +/- 9 years, BMI 27.7 +/- 4.9 kg/m2, 17 women) with postsurgery blood glucose levels >120 mg/dl (6.7 mmol/l) were investigated in three different ICUs (20 per center). Patients were randomized to either blood glucose management (target range 80-110 mg/dl [4.4-6.1 mmol/l]) by the fully automated model predictive control (MPC) algorithm (n = 30, 10 per center) or implemented routine glucose management protocols (n = 30, 10 per center). In all patients, arterial glucose was measured hourly to describe the glucose profile until the end of the ICU stay but for a maximum period of 48 h. RESULTS: Compared with routine protocols, MPC treatment resulted in a significantly higher percentage of time within the target glycemic range (% median [min-max]: 52 [17-92] vs. 19 [0-71]) over 0-24 h (P < 0.01). Improved glycemic control with MPC treatment was confirmed in patients remaining in the ICU for 48 h (0-24 h: 50 [17-71] vs. 21 [4-67], P < 0.05, and 24-48 h: 65 [38-96] vs. 25 [8-79], P < 0.05, for MPC [n = 16] vs. routine protocol [n = 13], respectively). Two hypoglycemic events (<54 mg/dl [3.0 mmol/l]) were observed with routine protocol treatment. No hypoglycemic event occurred with MPC. CONCLUSIONS: The data suggest that the MPC algorithm is safe and effective in controlling glycemia in critically ill postsurgery patients.
Closed-loop control of the glucose concentration in type 1 diabetes has been the subject of extensive research over the last 3 decades. Building on the recent progress in continuous glucose sensing techniques, several prototypes of a closed-loop system have been developed. To complement existing measures of glucose control, we designed a grading system specifically designed to provide clinical assessment of closed-loop systems including that of glucose controllers. The system introduces six grades, A-F, describing the level of control and the therapeutic intervention during outside-meal and postprandial conditions. Grades A and B represent excellent and good glucose control, respectively, without the need for a corrective therapeutic action. Grade C represents suboptimal control with a recommendation for a corrective action. Grade D represents poor control requiring a corrective action. Grades E and F represent very poor and life-threatening control, respectively, with a need for an immediate corrective action or requiring external assistance. The outcome of grading is the quantification of time spent in each grade. The grading system is exemplified using data obtained with a model predictive controller within an in silico simulation environment. We conclude that the grading system provides suitable means to assess efficacy and safety of glucose controllers complementing existing measures of glucose control.
We investigated insulin lispro kinetics with bolus and continuous subcutaneous insulin infusion (CSII) modes of insulin delivery. Seven subjects with type-1 diabetes treated by CSII with insulin lispro have been studied during prandial and postprandial conditions over 12 hours. Eleven alternative models of insulin kinetics have been proposed implementing a number of putative characteristics. We assessed 1) the effect of insulin delivery mode, i.e., bolus or basal, on the insulin absorption rate, the effects of 2) insulin association state and 3) insulin dose on the rate of insulin absorption, 4) the remote insulin effect on its volume of distribution, 5) the effect of insulin dose on insulin disappearance, 6) the presence of insulin degradation at the injection site, and finally 7) the existence of two pathways, fast and slow, of insulin absorption. An iterative two-stage parameter estimation technique was used. Models were validated through assessing physiological feasibility of parameter estimates, posterior identifiability, and distribution of residuals. Based on the principle of parsimony, best model to fit our data combined the slow and fast absorption channels and included local insulin degradation. The model estimated that 67(53-82)% [mean (interquartile range)] of delivered insulin passed through the slow absorption channel [absorption rate 0.011(0.004-0.029) min(-1)] with the remaining 33% passed through the fast channel [absorption rate 0.021(0.011-0.040) min(-1)]. Local degradation rate was described as a saturable process with Michaelis-Menten characteristics [VMAX = 1.93(0.62 - 6.03) mU min(-1), KM = 62.6(62.6 - 62.6) mU]. Models representing the dependence of insulin absorption rate on insulin disappearance and the remote insulin effect on its volume of distribution could not be validated suggesting that these effects are not present or cannot be detected during physiological conditions.
OBJECTIVE: Adaptive systems to deliver medical treatment in humans are safety-critical systems and require particular care in both the testing and the evaluation phase, which are time-consuming, costly, and confounded by ethical issues. The objective of the present work is to develop a methodology to test glucose controllers of an artificial pancreas in a simulated (virtual) environment. MATERIAL AND METHODS: A virtual environment comprising a model of the carbohydrate metabolism and models of the insulin pump and the glucose sensor is employed to simulate individual glucose excursions in subjects with type 1 diabetes. The performance of the control algorithm within the virtual environment is evaluated by considering treatment and operational scenarios. RESULTS: The developed methodology includes two dimensions: testing in relation to specific life style conditions, i.e. fasting, post-prandial, and life style (metabolic) disturbances; and testing in relation to various operating conditions, i.e. expected operating conditions, adverse operating conditions, and system failure. We define safety and efficacy criteria and describe the measures to be taken prior to clinical testing. The use of the methodology is exemplified by tuning and evaluating a model predictive glucose controller being developed for a wearable artificial pancreas focused on fasting conditions. CONCLUSION: Our methodology to test glucose controllers in a virtual environment is instrumental in anticipating the results of real clinical tests for different physiological conditions and for different operating conditions. The thorough testing in the virtual environment reduces costs and speeds up the development process.
We investigated the dynamic relationship between interstitial glucose (IG) in the subcutaneous adipose tissue and plasma glucose (PG) during physiologic conditions in type 1 diabetes mellitus (T1DM). Nine subjects with T1DM (5/4 M/F; age, 33 +/- 13 years; body mass index, 26.6 +/- 4.3 kg/m(2); glycosylated hemoglobin [HbA(1c)], 8.6% +/- 0.9%; mean +/- SD) treated by continuous subcutaneous insulin infusion (CSII) with insulin lispro were studied over 12 hours after a standard meal (40 g carbohydrate [CHO]) and prandial insulin. IG was measured by open flow microperfusion. Nine compartment models were postulated to account for temporal variations in the IG/PG ratio. The models differed in the inclusion of physiologically motivated alterations of pathways entering/leaving the IG compartment in the adipose tissue. The best model included zero order (constant) glucose disposal from the interstitial fluid (ISF) and insulin-stimulated glucose transfer from plasma to the ISF. The former effect is expressed by a positive association between the IG/PG ratio and PG, eg, a decrease in PG from 9 to 3.3 mmol/L lowers the IG/PG ratio by 0.1. The latter effect results in the IG/PG ratio to be increased by 0.03 per 10 mU/L of plasma insulin. We were not able to detect the stimulatory effect of insulin on glucose disappearance from the ISF. In conclusion, we developed and quantified a model of IG kinetics in the adipose tissue applicable to physiologic conditions in subjects with T1DM.
A nonlinear model predictive controller has been developed to maintain normoglycemia in subjects with type 1 diabetes during fasting conditions such as during overnight fast. The controller employs a compartment model, which represents the glucoregulatory system and includes submodels representing absorption of subcutaneously administered short-acting insulin Lispro and gut absorption. The controller uses Bayesian parameter estimation to determine time-varying model parameters. Moving target trajectory facilitates slow, controlled normalization of elevated glucose levels and faster normalization of low glucose values. The predictive capabilities of the model have been evaluated using data from 15 clinical experiments in subjects with type 1 diabetes. The experiments employed intravenous glucose sampling (every 15 min) and subcutaneous infusion of insulin Lispro by insulin pump (modified also every 15 min). The model gave glucose predictions with a mean square error proportionally related to the prediction horizon with the value of 0.2 mmol L(-1) per 15 min. The assessment of clinical utility of model-based glucose predictions using Clarke error grid analysis gave 95% of values in zone A and the remaining 5% of values in zone B for glucose predictions up to 60 min (n = 1674). In conclusion, adaptive nonlinear model predictive control is promising for the control of glucose concentration during fasting conditions in subjects with type 1 diabetes.
The objective of the project Advanced Insulin Infusion using a Control Loop (ADICOL) was to develop a treatment system that continuously measures and controls the glucose concentration in subjects with type 1 diabetes. The modular concept of the ADICOL's extracorporeal artificial pancreas consisted of a minimally invasive subcutaneous glucose system, a handheld PocketPC computer, and an insulin pump (D-Tron, Disetronic, Burgdorf, Switzerland) delivering subcutaneously insulin lispro. The present paper describes a subset of ADICOL activities focusing on the development of a glucose controller for semi-closed-loop control, an in silico testing environment, clinical testing, and system integration. An incremental approach was adopted to evaluate experimentally a model predictive glucose controller. A feasibility study was followed by efficacy studies of increasing complexity. The ADICOL project demonstrated feasibility of a semi-closed-loop glucose control during fasting and fed conditions with a wearable, modular extracorporeal artificial pancreas.
We have separated the effect of insulin on glucose distribution/transport, glucose disposal, and endogenous production (EGP) during an intravenous glucose tolerance test (IVGTT) by use of a dual-tracer dilution methodology. Six healthy lean male subjects (age 33 +/- 3 yr, body mass index 22.7 +/- 0.6 kg/m(2)) underwent a 4-h IVGTT (0.3 g/kg glucose enriched with 3-6% D-[U-(13)C]glucose and 5-10% 3-O-methyl-D-glucose) preceded by a 2-h investigation under basal conditions (5 mg/kg of D-[U-(13)C]glucose and 8 mg/kg of 3-O-methyl-D-glucose). A new model described the kinetics of the two glucose tracers and native glucose with the use of a two-compartment structure for glucose and a one-compartment structure for insulin effects. Insulin sensitivities of distribution/transport, disposal, and EGP were similar (11.5 +/- 3.8 vs. 10.4 +/- 3.9 vs. 11.1 +/- 2.7 x 10(-2) ml small middle dot kg(-1) small middle dot min(-1) per mU/l; P = nonsignificant, ANOVA). When expressed in terms of ability to lower glucose concentration, stimulation of disposal and stimulation of distribution/transport accounted each independently for 25 and 30%, respectively, of the overall effect. Suppression of EGP was more effective (P < 0.01, ANOVA) and accounted for 50% of the overall effect. EGP was suppressed by 70% (52-82%) (95% confidence interval relative to basal) within 60 min of the IVGTT; glucose distribution/transport was least responsive to insulin and was maximally activated by 62% (34-96%) above basal at 80 min compared with maximum 279% (116-565%) activation of glucose disposal at 20 min. The deactivation of glucose distribution/transport was slower than that of glucose disposal and EGP (P < 0.02) with half-times of 207 (84-510), 12 (7-22), and 29 (16-54) min, respectively. The minimal-model insulin sensitivity was tightly correlated with and linearly related to sensitivity of EGP (r = 0.96, P < 0.005) and correlated positively but nonsignificantly with distribution/transport sensitivity (r = 0.73, P = 0.10) and disposal sensitivity (r = 0.55, P = 0.26). We conclude that, in healthy subjects during an IVGTT, the two peripheral insulin effects account jointly for approximately one-half of the overall insulin-stimulated glucose lowering, each effect contributing equally. Suppression of EGP matches the effect in the periphery.
We examined the ability of indices of insulin sensitivity and pancreatic beta-cell responsiveness to explain interindividual variability of clinical measures of glucose control in newly presenting type 2 diabetes. Subjects with newly presenting type 2 diabetes (n = 65; 53 males and 12 females; age, 54 +/- 1 yr; body mass index, 30.5 +/- 0.7 kg/m(2); mean +/- SE) underwent an insulin-modified iv glucose tolerance test to determine minimal model-derived insulin sensitivity (S(I)), glucose effectiveness, first-phase insulin secretion, and disposition index. Subjects also underwent a standard meal tolerance test (MTT) to measure fasting/basal (M(0)) and postprandial (M(I)) pancreatic beta-cell responsiveness. Stepwise linear regression used these indices to explain interindividual variability of fasting and postprandial plasma glucose and insulin concentrations and glycated hemoglobin (HbA(1C)). All measures of pancreatic beta-cell responsiveness (M(0), M(I), and first-phase insulin secretion) were negatively correlated with fasting plasma glucose (P < 0.01) and positively correlated with fasting plasma insulin (FPI) and insulin responses to MTT (P < 0.05). S(I) demonstrated negative correlation with FPI (P < 0.001) but failed to correlate with any glucose variable. M(I) followed by disposition index (composite index of insulin sensitivity and pancreatic beta-cell responsiveness) were most informative in explaining interindividual variability. It was possible to explain 70-80% interindividual variability of fasting plasma glucose, FPI, HbA(1C), and insulin responses to MTT, and only 25-40% interindividual variability of postprandial glucose. In conclusion, postprandial insulin deficiency is the most powerful explanatory factor of deteriorating glucose control in newly presenting type 2 diabetes. Indices of insulin sensitivity and pancreatic beta-cell responsiveness explain fasting glucose and HbA(1C) well but fail to explain postprandial glucose.