Three glucose sensors were developed based on different technology platforms:
• a classical amperometric sensor using thick film technology;
• a fibre-optic fluorometric glucose sensor based on oxygen measurement, and
• a spectroscopic glucose sensor using mid-infrared spectroscopy
Amperometric glucose sensor
An enzymatic-electrochemical sensor was developed for continuous glucose monitoring based on a novel miniaturized planar sensor flow-through cell arrangement. The sensor, which was manufactured using polymer thick film technology, features four electrodes serving as the amperometric detection unit and for measuring conductivity.
The enzyme is immobilized by a hydrogel matrix forming a self-adhesive layer at the surface of the working electrode. The electrode surfaces and the enzyme immobilisate are protected against interfering bio-compounds from the body fluid by a biocompatible selective diffusion barrier (molecular weight 160 kDa). The polymer coating improved the linearity range of the glucose sensor to 20 mM and the drift during long-term glucose measurements in serum solution was limited to +0.23 % per hour. The response time of the sensor is about 3 min and the running-in period is less than 10 minutes.
A plastic foil was micro-structured by hot embossing and glued to the sensor to create a disposable flow-through cell. The patented inner geometry of the flow-through cell, featuring a gap close to the indicator window and surrounded by channels for the sample flow, prevents air bubbles from forming or entering the sensitive area of the sensor. This arrangement provides a continuous flow of small volumes of sample to the sensitive area by capillary forces and convection. The volume of the measuring chamber is about 0.3 µl (Fig.1).
Successful in-vivo measurements were carried out using this arrangement in several workshops with healthy volunteers, diabetic patients and critically ill patients after major cardiothoracic surgery.
Left side: Fig.1: Disposable glucose sensor flow-through cell arrangement for continuous glucose measurement. Right side: Fig.2: Exemplary temporal glucose profile measured in a type 1 diabetes volunteer at the Clinical Research Center of the Medical University Graz, Austria
Fibre-optic glucose sensor
The fibre-optic sensor system uses the enzyme-based oxidation of glucose, in combination with an optical oxygen sensor as transducer. A fibre-optic dual sensor setup was integrated into a flow-through cell. One sensor measures oxygen sensor only, while the second oxygen sensor is covered with an enzyme (glucose oxidase, GOD) layer. The difference in oxygen concentration between these two sensors is linear up to 15 mM of glucose. The major advantage of this approach is the excellent selectivity of the oxygen optode transducer.
The sensors have already been tested in both clinical studies and intensive care units with very promising results. Improvements of the transducers are currently underway, including the synthesis of new fluorophors with improved properties such as greater brightness and lower temperature dependence.
Fig.a: System for fibre-optic fluorometric glucose measurement using a standard subcutaneous microdialysis system
Spectroscopic glucose sensor
A mid-infrared sensor for continuous glucose monitoring in combination with a subcutaneous or a vascular body interface was developed to meet the demand for reagent-free assays. The sensor is designed to ensure the utmost reliability needed for the intensive care environment. Monitoring of human subcutaneous interstitial fluid or whole blood dialysate was realized by an ex-vivo arrangement. The body interface between the subject and the glucose sensing device was an implantable microdialysis catheter or an extra-corporeal whole blood microdialysis device, respectively. A fluidic system for intermittent sample transport to a flow-through micro-cell was developed for transporting the biofluid sample to a mini-spectrometer. The body fluids were dialysed using perfusates at flow rates of 1 or 5 µl/min, realized by push-pull operation of a mini-peristaltic pump. As a result of the osmotic exchange of biocompounds between the biofluid and the perfusate, low molecular mass components were harvested. The fully automated system was explored for its multi-component capability for glucose, urea, lactate, acetate, bicarbonate, as well as for pCO2 and pH of the biofluid buffer system. Features also include dialysis recovery rate measurements and air bubble detection and removal. The multi-component capability of the system allows acetate of the ELO-MEL perfusate to be used as a suitable marker for the simultaneous determination of the microdialysis recovery rate for accurate estimation of interstitial or whole blood concentrations.
The prototype was tested successfully for online monitoring of blood glucose in healthy and type 1 diabetic subjects including its combination with an insulin pump for closing the loop.
Fig.b: Left side: Glucose concentration profiles of a type 1 diabetic subject monitored continuously using a mid-IR spectrometer and a subcutaneously implanted microdialysis probe (A), and lactate concentrations in the dialysate (B). Right side: Glucose concentration profiles of a type 1 diabetic subject monitored spectrometrically using extra-corporeal whole blood dialysis (C) and dialysis recovery rates using a perfusate marker (D)
Fig.c: Clarke Error Grid plot of sensor predicted glucose and reference blood glucose values from eight type 1 diabetic volunteers using a subcutaneous body interface (A) and from five type 1 diabetic volunteers using a vascular interface (B).
Other metabolite Sensors
The parameters pH, oxygen (O2) and carbon dioxide (CO2) are essential in determining the physiological status of critically ill patients. The limited reliability of continuous intravascular measurements and drawbacks of intermittent blood sampling led the CLINICIP team to develop optical sensors for measuring these parameters in the adipose tissue. The continuous measurement of pH, O2 and CO2 is based on the withdrawal of interstitial fluid from adipose tissue by means of a microdialysis catheter. The interstitial fluid then flows through a microfluidic circuit formed by the catheter arranged in series with optical sensors. While the pH sensor was based on a sensing capillary, the O2 and CO2 sensors were implemented in two forms: a sensing glass capillary and a planar sensing membrane within a flow-through cell.
pH sensor: The pH sensor consists of a glass capillary carrying on its internal wall phenol red covalently bound to the glass. The sensing capillary is inserted in the microfluidic circuit containing the microdialysis catheter used for the extraction of interstitial fluid (Fig.3). The optoelectronic instrument developed for pH interrogation is shown in Fig.4. Two optical fibres are used to couple the sensing capillary with the unit in which a light-emitting diode (LED) and a photodetector are used as source and detector, respectively. The device is controlled by a software program, developed in LabVIEW environment, which acquires the signals and evaluates, displays and stores the data.
In-vivo studies have shown that the adipose tissue seems to be a reliable alternative site for in-vivo continuous monitoring of stress conditions.
Fig.3: Fluidic System for Microdialysis Measurements.
Fig.4: Photo of the System for pH detection
O2 sensor: The oxygen opto-chemical sensor is based on the principle of the luminescence quenching of a fluorophore by collision with oxygen.
The inner surface of the glass capillary is coated with a sensor cocktail of Pt(II)-TFPP dissolved in chloroform. The optoelectronic instrumentation basically consists of a light source (LED emitting at about 517 nm) to excite the luminescence signal, a reference light source and a detection module comprising a Si photodiode and optical filters.
The planar sensing membranes are created by coating a transparent substrate with an oxygen sensitive solution and connecting them to the optoelectronic unit via an optical fibre. The main advantage of this approach is that the optoelectronic instrumentation can be separated from the flow-through cell, thus allowing greater miniaturization than for the capillary based sensor. Its small size makes it more flexible for use in in-vivo experiments and field tests in hospitals.
Fig.d: Measurement system for planar sensing membranes: the flow-through cell has a size of 35x35x12 mm (width x length x heigth) and can allocate one transparent substrate coated with up to 2 sensing spots for two different parameters, e.g. O2 and CO2
CO2 sensor: CO2 measurements are based on the dual luminophore referencing (DLR) approach, an internal ratiometric method for converting the analyte-sensitive fluorescence intensity signal into the frequency-domain or time-domain information by co-immobilizing an inert long-lifetime reference luminophore and a short-lifetime indicator material which responds to CO2 induced pH changes in the membrane.
Measurements involve a sensor cocktail containing Ru(dpp)32+ nanoparticles doped in polyacrilonitrile (PAN) with negligible permeability towards oxygen (to avoid interference from oxygen) and the HPTS(TOA)4 ion pair. This mixture is deposited onto the inner surface of glass capillaries with a needle (capillary sensors) or applied to a solid support (planar sensor membranes). The optoelectronic units used for the interrogation of the CO2 sensing capillary and the planar sensing membrane are the same as described for O2 detection.
The planar sensor membranes are characterised by a better stability in time due to a better adhesion of the sensing film on the solid support and a better reproducibility in terms of realisation.
The three sensors were tested in-vivo.
© 2004-07 CLINICIP Consortium
This project was co-funded by the EU through the IST programme under FP6