WO2005007879A1

WO2005007879A1 - In vivo calibration method for subcutaneous amperometric glucose sensors and system therefor

Info

Publication number: WO2005007879A1
Application number: PCT/KR2004/001834
Authority: WO
Inventors: Hee-Chan Kim; Taek Dong Chung; Sejin Park; Ran-A Jeong; Jae Youn Hwang; Segyeong Joo; Sun Kil Kang; Won Yong Lee
Original assignee: Seoul National University Industry Foundation
Priority date: 2003-07-22
Filing date: 2004-07-22
Publication date: 2005-01-27
Also published as: KR100560183B1; KR20050011166A

Abstract

New two-point calibration method and system for the subcutaneous amperometric continuous glucose sensor is reported. The proposed method comprises a step of carrying out calibration by a background current, wherein the background current is obtained from direct measurement using a non-enzyme electrode without any external injection of glucose. The system used in the method comprises two electrodes in which one is an enzyme electrode and the other is a non-enzyme electrode having the same membrane composition and morphology with the enzyme electrode except that it contains no enzyme, wherein calibration for quantifying in vivo glucose concentration is performed by direct measurement of background current using the non-enzyme electrode. According to the method and the system, both of the convenience of the conventional one-point calibration and the accuracy of the conventional two-point calibration can be accomplishable.

Description

Description IN VIVO CALIBRATION METHOD FOR SUBCUTANEOUS AMPEROMETRIC GLUCOSE SENSORS AND SYSTEM THEREFOR Technical Field

[1] The present invention relates to a calibration method and system for quantifying blood glucose concentration. More specifically, the present invention relates to in vivo calibration method for the subcutaneous amperometric continuous glucose sensor and a system therefor. Background Art

[2] Many studies have been reported for developing implantable glucose sensors based on amperometric enzyme electrodes: V. Thome-Duret et al., 1996; V. Thome-Duret et al., 1998; Moatti-Sirat et al., 1992; W. K. Ward et al., 2002; J. Wang and X. Zhang, 2001; S. J. Updike et al., 2000. Based on recent reports (C. Choleau et al., 2002a), technologies have advanced to develop a glucose sensor implanted in the subcutaneous tissue of diabetic patients during up to 7 days. There remains, however, one unsolved crucial issue of the in vivo calibration of the sensor system. The calibration of the system is the procedure to determine the parameters describing the relationship between the sensor's current output and the corresponding blood glucose concentration. Two typical methods to determine these parameters are one-point calibration and two-point calibration on the basis of single blood glucose measurement and two measurements, respectively. The conventional two-point calibration procedure is to determine the sensor sensitivity (S) and the background current (I ) by obtaining o the responses at two different glucose concentrations. In spite of its theoretical superiority, the two-point calibration has the disadvantages of time-consuming procedure and its vulnerability to the measurement errors. Choleau et al reported that the two- point calibration for their clinical setup produced the negatively correlated S and I and o sometimes negative I (C. Choleau et al., 2002a). On the other hand, the one-point o calibration assumes that variation of I over time and interference by electroactive o species is negligible. So the ratio of measured current (l)/gluoose level (G) simply determines S. Although this is indeed a simple method and its feasibility was demonstrated in clinical applications (C. Choleau et al., 2002b), neglecting I leaves o standout points to be addressed. [3] W. K. Ward et al. reported that I from non-enzyme electrodes implanted gave the o result that is basically similar to the extrapolated I , demonstrating that the two-point o calibration has sufficient accuracy. But they also observed the time-related rise of I o from the direct measurements (W. K. Ward et al., 2000). Even not clearly elucidated until now, time-dependent variation of I is believed to result from various factors o including mechanical damage to the membrane, undesired oxidation of the elec- troactive species andor adhesion of various inorganic salts or proteins present in blood or interstitial fluid, during implantation. In addition, complicated factors from in vivo immune response or glucose regulation mechanism were shown to have an unexpected effect on the I of the implanted glucose sensor. This problem may be a serious source o of errors diminishing the credibility of the implantable amperometric sensors for continuous glucose monitoring as a dysglycemic alarm system. Therefore, even if the two-point calibration might be considered to reflect the real value as suggested by W. K. Ward et al., such a method suffers from the disadvantage that it should be frequently corrected in order to apply to human body. In addition, the two-point calibration is an inconvenient one and takes much longer measurement time, because it requires external injection of glucose or insulin after stabilization of glucose level at a predetermined time and condition. Therefore, new calibration method which satisfies both the convenience of the conventional one-point calibration and the accuracy of the conventional two-point calibration is demanded. Disclosure of Invention Technical Problem

[4] As mentioned in the prior art, the calibration by the background current is needed in order to continuously quantify the blood glucose concentration. According to the present invention, the background current is not obtained from extrapolation but from direct measurement using a non-enzyme electrode. Technical Solution

[5] Specifically, an object of the present invention is to provide a method for quantifying blood glucose concentration accompanying measurement of current output from an enzyme electrode, comprising a step of carrying out calibration by background current obtained from direct measurement using a non-enzyme electrode having the same membrane composition and morphology with the enzyme electrode but enzyme.

[6] Another object of the present invention is to provide a glucose sensor system comprising two electrodes in which one is an enzyme electrode and the other is a non- enzyme electrode having the same membrane composition and morphology with the enzyme electrode but enzyme, both of the enzyme and the non-enzyme electrodes are implantable to a subcutaneous tissue, and calibration for quantifying in vivo glucose concentration is performed by direct measurement of the background current using the non-enzyme electrode. According to the preferred embodiment of the present invention, the system was proven to be useful for continuous in vivo glucose monitoring, and did not require any external injection of glucose. [7] The enzyme electrode used in the present invention was obtained by forming an enzyme layer on a metallic wire, followed by forming a biooompatible outer membrane thereon which protects the enzyme layer. The non-enzyme electrode had the same membrane composition and morphology with the enzyme electrode except that it contains no enzyme. The two electrodes can be integrated into one body by MEMS (microelectromechanical system) technology, which solves the inconvenience caused by separate implantations of the two electrodes into the subcutaneous tissue. Further, contrary to the conventional two-point calibration, the system of the present invention directly measures the background current using the non-enzyme electrode such that blood sampling after meal and accompanying measurement of the glucose level is not required. The system of the present invention can be calibrated at any time, excluding an exceptional case in which an abrupt change of the glucose level caused by intense exercise or heavy food intake takes place, because artificial change for calibration of glucose level by additional meal or empty stomach is not required. Therefore, the method sufficiently relieves the strong burden of user's calibration work and enables frequent calibration, thereby substantially enhancing the stability of the implantable glucose sensor. Advantageous Effects

[8] The enzyme electrode used in the present invention was obtained by forming an enzyme layer on a metallic wire, followed by forming a biooompatible outer membrane thereon which protects the enzyme layer. The non-enzyme electrode had the same membrane composition and morphology with the enzyme electrode but enzyme. The two electrodes can be integrated into one body by MEMS (microelectromechanical system) technology, which solves the inconvenience caused by separate implantations of the two electrodes into the subcutaneous tissue. Further, contrary to the conventional two-point calibration, the system of the present invention directly measures the background current using the non-enzyme electrode such that blood sampling after meal and accompanying measurement of the glucose level is not required. The system of the present invention can be calibrated at any time, excluding an exceptional case in which an abrupt change of the glucose level caused by intense exercise or heavy food intake takes place, because artificial change for calibration of glucose level by additional meal or empty stomach is not required. Therefore, the method sufficiently relieves the strong burden of user's calibration work and enables frequent calibration, thereby substantially enhancing the stability of the implantable glucose sensor. Brief Description of the Drawings

[9] Fig. 1(a) shows in vivo experimental setup for measurement of glucose concentration in the subcutaneous tissue and blood vessel (ear artery).

[10] Fig 1(b) is a block diagram of the telemetry system for wireless data transmission.

[11] Fig. 2 are graphs showing results of in vivo evaluation of the glucose sensor: (a) Reference measurement by a commercial glυcosemeter; (b) current response of the sensor implanted in the ear artery; and (c) outputs of the two sensors implanted in the subcutaneous tissue where upper curve is for the normal sensor and lower one for the sensor without enzyme layer that was indicated by an arrow of I . o

[12] Fig. 3 is scanning electron microscopy (SEM) images showing morphological changes of the sensor surface: (a) before implantation; and (b) after explanation. The implantation site was subcutaneous tissue.

[13] Fig. 4 is a graph showing Error Grid Analysis (EGA) on the measurements in in vivo evaluation of the glucose sensor: 88% of the total values falls in zone A and 12% in zone B where D and ♦ correspond to measurements in artery and subcutaneous layer, respectively. The data for EGA were calculated from the current values that were sampled out from the curves such as Fig. 2(b) and 2(c) in five independent experiments.

[14] Fig. 5 is a graph showing step responses of current output of the glucose sensor to glucose concentrations of 3, 6, 9, 12, 15, 18, and 21 mM in PBS solution measured by the telemetry system according to the present invention. Mode for the Invention

[15] In the following, the present invention will be more fully illustrated referring to Examples.

[16] 1. Example

[17] 1.1. Fabrication of the enzyme electrodes and in vitro tests

[18] According to the present invention, the fabrication of the glucose sensors followed the procedure that had been described in our previous reports (Chung et al., 2001). In short, a pre-cleaned platinum wire is immersed in a phosphate buffer saline (PBS) solution containing glucose oxidase (GOx), glutaraldehyde, poly-L-lysine, and 1,3-diaminobenzene (1,3-PD). Electropolymenzation by potential cycling between 0.2 V and 1.0 V gives a thin enzyme layer whose thickness is self-controlled to be ca. 20 nm. Based on our recent experiences of using Kel-F (commercial name) film as the oxygen-rich outer membrane as well as the glucose diffusion-limiting barrier, outer layers were added onto the enzyme layer by dipping the electrode in 20% Kel-F oil, 30% perfluorinated tetrafluoroethylene (PTFE), 20% Kel-F oil and 1.5% Nafion sequentially for about 5 sec each. Then, the electrode was dried in a vacuum chamber (Lab-line, model 3606- ICE) at 30°C for 1 hour. All of the in vitro evaluations were performed in a cell containing 15 ml of PBS (pH 7.4, 37°C) with water-jacket. The temperature was carefully controlled by circulating water from a thermostatic bath (Ultra-Thermostat NB5, Colora) to the water-jacket. All chemical reagents were purchased from Aldrich and used as purchased.

[19] 1.2. Calibration in vivo

[20] The two-point calibration (C. Choleau et al., 2002a) is to transform the current observed (I) into the glucose level estimated (G) by following the typical procedure. In vivo sensitivity (S) is determined as the ratio of the increments in current to corresponding change in blood glucose concentration.

[21] S = (I - I ) / (G - G ) (l) 2 1 2 1

[22] In this invention, I was taken as the background current I that was directly 1 o measured from the non-enzyme electrode. So G was regarded as zero on the assumption that the non-enzyme electrode is perfectly insensitive to the glucose. And I was obtained from the stabilized initial current (I ) of the enzyme electrode implanted 2 l before glucose injection. The analysis in vitro with blood sampled prior to glucose injection gave G for G . Bath I and I are the current responses to the glucose levels in i 2 1 0 equilibrium. [23] S = (I - 1 ) / G (2)

[24] With the sensitivity and the background current, blood glucose levels, G, can be calculated from a simple relation as follows. [25] G = (I - 1 ) / S (3) o

[26] 1.3. Evaluation in vivo

[27] Five male New Zealand white rabbits weighing 3.5-4.0 kg were used for the purpose of implanting the glucose sensors as made. Two identical enzyme electrodes and one non-enzyme electrode with the same structure and composition but enzyme were implanted for every test. One of the two enzyme electrodes was inserted into the subcutaneous tissue and the other into ear artery of rabbit anesthetized with ketamine 40 mgAcg and xylazine 8 mg g. In order to measure the background current, the non- enzyme electrode was implanted 1 cm away from the enzyme electrode in the subcutaneous tissue. Two catheters were indwelled into both ear arteries for the sensor insertion on the left ear and blood sampling from the right (Fig. la). All blood samples TM were analyzed with a commercial glυcosemeter (Precision Plus , MediSense). When the current signals from the enzyme electrodes reached a steady state, a 50% glucose solution was injected through the third catheter into the right ear vein. The three electrodes were connected to multi-channel potentiostat (model cDAQ-1604, ELBIO Co., Seoul) and biased +0.4 V against the reference electrodes that were integrated in the same sensor bodies. The background current from the non-enzyme electrode was used to calibrate the responses from the enzyme electrode implanted in subcutaneous tissue. In vivo monitoring normally took about 3 h and the electrodes have been implanted in the living body for about 7 h in total. Once an in vivo monitoring was finished, the sensors were explanted from the subcutaneous tissue and blood vessel. And then an in vitro test was carried out to see any change in sensitivity before and after working in vivo. The images from the scanning electron microscope (SEM, model JSM 840- A) gave the information about what had happened to the surface morphology of the membrane surfaces.

[28] 1.4. Telemetry system

[29] The radiofrequency (RF) telemetry system used in this invention consists of two parts, i.e. transmitter and receiver. Fig. lb shows the block diagram illustrating this system. In transmitter 100, the current responses to the glucose oxidation measured by a potentiostat 101 are converted into voltage signals, which generate digital pulses whose frequency is proportional to the voltage by voltage-controlled oscillator 102 (VCO). The digital pulses modulate the carrier frequency of the RF transmitter module 103 (TXM-433-LC-R, LINX Co.) with on-off keying (OOK) and are transmitted into an air via an antenna 104. In a receiver 200, a RF receiver module 201 (RXM-433-LC-S, LINX Co.) receives the transmitted signal and demodulates the digital pulses. Phase locked loop 203 (PLL) locks these digital pulses and produces an output voltage that is proportional to the frequency of the input digital pulses. The output voltage of the PLL 203 is then converted into digital signal by a 12 bit analog- to-digital converter 204 (MCP3204, Microchip). Finally, the digital signal is recorded by a computer. The dimension of the transmitter 100 as made is 3 cm x 4 cm x 0.5 cm. The maximum distance between the transmitter 100 and the receiver 200 is about 120 m.

[30] 2. Results & Discussions [31] 2.1. In vivo signals from enzyme electrodes and background current

[32] Fig. 2 shows the current responses to a step increase of the blood glucose concentration, which were measured by a commercial gluoosemeter with the blood sampled from the ear artery and the implanted sensors developed in this invention. In this experiment, the blood glucose concentration of a rabbit was raised from 105 to 436 mg/dl by an intravenous injection of glucose solution (Fig. 2a). The current outputs from the enzyme electrodes rapidly increased upon the injection of glucose and then slowly decreased behind the peak. The outputs from the sensors in the ear artery (Fig. 2b) and in the subcutaneous tissue (Fig. 2c) were similar in shape to the reference concentration profile (Fig. 2a). The response of the enzyme electrode in the subcutaneous tissue had a peak with about 10 min time delay compared with the blood glucose level. It is believed that a push-pull mechanism for glucose kinetics makes significant effect as the plasma glucose concentration increases (V. Thome-Duret et al., 1996; B.Aussedat et al., 2000).

[33] On the contrary to the outputs from the enzyme electrodes, the background current (I ) from the non-enzyme electrode in the subcutaneous layer gave little response to the o changes in blood glucose concentration.

[34] In order to investigate the possible degradation of the sensor following implantation, the enzyme electrodes that had been implanted in the subcutaneous tissue and blood vessel were explanted and retested in vitro. Table 1 summarizes the results.

[35] Table 1

[36] ( Over the concentration range 1-12 mM. Background current is stabilized within 30 min. The statistics are from the 5 independent experiments.)

[37] As shown in Table 1, it is unambiguous that the background current as well as the sensitivity substantially increased after explantation compared with before implantation. It implies that some crevices appeared in the outer layer or even the membrane partially peeled off. Fig. 3 shows the morphological changes in the sensor surface by scanning electron microscope (SEM). The morphology of the membrane prior to implantation is clean and has little cracks as shown in Fig. 3a. On the contrary, many cracks are found on the surface morphology after explantation and some parts of the layer are seriously broken. Reportedly, Nafion has sufficient biccompatibility for implantable sensors (F. Moussy et al., 1993). But the sulfonate groups of Nafion coat hold calcium ions so that the film is vulnerable to calcification (T. I Valdes and F. Moussy, 1999). Therefore, calcification was accelerated by damages in even limited parts of the surface because significant amount of calcium ions were present in the outer layers, Nafion. The significance of background current is amplified as the implantation period is extended. W. K. Ward et al. used PTFE as a protecting layer instead of Nafion and observed a similar phenomenon that the background current kept on increasing over the implantation time (W. K. Ward et al., 2000). Consequently, more reliable monitoring of in vivo glucose level requires correction by the time- varying background current.

[38] 2.2. Calibration

[39] Since the one-point calibration method is based on the assumption of constant I , o undesirable errors can be brought about by any variation in I . Nevertheless, it is hardly o to say that the conventional two-point calibration is a successful alternative. As shown in Fig. 2, an external injection of glucose causes an instant rise in the current response of the sensor in artery while a significant time delay is present for the sensor in the subcutaneous layer. Because complex push-pull kinetics keeps working after the injection and thus the glucose concentration ceaselessly varies, uncertainty is always present in assuming that the peak current from the subcutaneous layer in Fig. 2(c) properly represents the blood glucose level. As a result, the sensitivity and extrapolated background current (intercept) from two-point calibration have a chance of being erroneous. [40] Fig. 4 shows the graph of Error Grid Analysis (EGA) (W. L. Clarke et al., 1987; D. A. Gough and E. L. IΞ tvinick, 1997) with the data obtained in this study. EGA is used to evaluate for the clinical accuracy of self-monitoring of blood glucose. Values falling within zone A are clinically accurate and zone B values differs substantially from the reference, but the appropriate action taken based on estimate would be benign or no treatment would be given. The sensitivity (S) and glucose concentration (G) were estimated by using equation (2) and (3), respectively. In the two equations, both I and o I were obtained in equilibrium without any external injection of glucose. It is believed that those currents represent the blood glucose level very accurately because any possible complex kinetics between blood and subcutaneous was minimized. All the points corresponding to the measurements from artery are within zone A. Most of the points from the subcutaneous layer fall in zone A except some points in zone B at higher concentration than 250 mg/dl. In brief, all of the values were existed within zone A and B. This calibration method is basically the two-point calibration. But it provides valuable advantages over the conventional two-point calibration method because of its use of single measurement only. The S and G values are expected to be more accurate than those obtained by external injection of glucose because a complex physiological kinetics may be excluded. Since the injection experiments in vivo usually start from normal glucose level, the two points are selected in a high concentration range. S is likely to be underestimated so that the small S makes G values overestimated according to equation (3), especially in hypoglycemia.

[41] 2.3. Telemetry system

[42] In vitro evaluation of the implantable glucose sensor and the telemetry system was performed in PBS solution. The telemetry system was calibrated at one point of 40nA current response. Fig. 5 represents the current response of the sensor measured by the telemetry system on adding 3, 6, 9, 12, 15, 18, and 21 mM of glucose. At first, a little noise exhibited but immediately the current response became stable. Consequently, we have evaluated a telemetric monitoring system for in vivo continuous glucose monitoring in order to mitigate the problems of motion artifact and will make more integrated sensor-telemetric system in further study.

[43] 3. Conclusions

[44] The present invention proposedsuggested a new in vivo two-point calibration method for subcutaneous amperometric glucose sensor and demonstrated its feasibility in relatively short-term animal implantations, successfully works in vivo. Possible sources of errors were substantially reduced by employing a non-enzyme electrode with the same membrane as the enzyme electrodes. Most of implantable glucose sensors that have been reported adopt one-point calibration method for its simplicity and convenience. But the result of this study shows that the signals from the enzyme electrodes should be corrected by simultaneous monitoring of the time- varying background current (I ). An additional non-enzyme electrode deserves to be implanted o as an extra device for the collection of more accurate and safer data. The addition of a non-enzyme electrode allows the calibration based on two-equilibrium points without any glucose injection just like one-point calibration. In clinical point of view, implanting two electrodes seems impractical. Integrating both enzyme and non- enzyme electrodes in a single sensor may increase the size of sensor not significantly, and also it will be hurdled by using microelectromechanical system (MEMS) technologies.

[45] As a consequence, the new two-point calibration method introduced in this invention improves the drawbacks of the conventional one -point and two-point calibration methods. blΞ th convenience of the one -point calibration and accuracy of the two-point calibration can bewere simultaneously achieved. A telemetry system with an effective data transfer method offered improved practicability of the chemical sensors.

[46] References

[47] Aussedat, B., Dupire-Angel, M., Gifford, R., Klein, J. C, Wilson, G. S., Reach, G.., 2000. Interstitial glucose concentration and glycemia: implications for continuous subcutaneous glucose monitoring. Am. J. Physiol. Endccrinol Metab. 278, E716-E728.

[48] Choleau, C, Klein, J. C, Reach, G., Aussedat, B., Demaria-Pesce, V., Wilson, G. S., Gifford, R., Ward, W. K., 2002a. Calibration of a subcutaneous amperometric glucose sensor Part 1. Effect of measurement uncertainties on the determination of sensor sensitivity and background current. Hosens. Hoelectron. 17, 641-646.

[49] Choleau, C, Klein, J. C, Reach, G., Aussedat, B., Demaria-Pesce, V., Wilson, G. S., Gifford, R., Ward, W. K., 2002b. Calibration of a subcutaneous amperometric glucose sensor implanted for 7 days in diabetic patients Part 2. Superiority of the one- point calibration method. Hosens. Hoelectron. 17, 647-654.

[50] Chung, T. D., Jeong, R-A., Kang, S. K., Kim, H. C, 2001. Reproducible Fabrication of Miniaturized Glucose Sensors: Preparation of Sensing Membranes for Continuous Monitoring. Hosens. Hoelectron. 16, 1079-1087.

[51] Clarke, W. L., Cox, D., Gonder-Frederick, L. A., Carter, W., Pohl, S. L., 1987. Evaluating clinical accuracy of systems for self-morning of blood glucose. Diabetes Care. 10(5), 622-628.

[52] Gough, D. A., IΞ tvinick, E. L., 1997. Reservations on the use of Error Grid Analysis for the validation of blood glucose assays. Diabetes Care. 20(6), 1034-1036.

[53] Kang, S. K., Jeong, R. A., Park, S., Chung, T. D., Kim, H. C, In vitro and short- term in vivo characteristics of a Kel-F thin film modified glucose sensor. Hosens. Hoelectron. submitted. [54] Matsumoto, T., Ohashi, A., Ito, N., Fujiwara, H., Matsumoto, T., 2001. A long-term lifetime amperometric glucose sensor with a peifluorccarbon polymer coating. Hosens. Hoelectron. 16, 271-276.

[55] Moatti-Sirat, D., Capron, F., Poitout, V., Reach, G.., Hndra, D. S., Zhang, Y., Wilson, G. S., Thevenot, D. R., 1992. Towards continuous glucose monitoring: in vivo evaluation of a miniaturized glucose sensor implanted for several days in rat subcutaneous tissue. Diabetologia. 35, 224-230.

[56] Moussy, F., Harrison, D. J., OBrien, A. W., Rajotte, R. V., 1993. Performance of Subcutaneously Implanted Needle-type glucose sensors employing a novel trilayer coating. Anal. Chem. 65, 2072-2077.

[57] Thome-Duret, V., Aussedat, B., Reach, G.., Gangnerau, M. N., Lemonnier, F., Klein, J. C, Zhang, Y., Hu, Y., Wilson, G. S., 1998. Continuous glucose monitoring in the free-moving rat. Metabolism. 47(7), 799-803.

[58] Thome-Duret, V., Reach, G.., Gangnerau, M. N., Lemonnier, F., Klein, J. C, Zhang, Y., Hu, Y., Wilson, G. S., 1996. Use of a subcutaneous glucose sensor to detect decrease in glucose concentration prior to observation in blood. Anal. Chem. 68, 3822-3826.

[59] Updike, S. J., Gilligan, B. J., Shults, M. C, Rhodes, R. K., 2000. A subcutaneous glucose sensor with improved longevity, dynamic range, and stability of calibration. Diabetes Care. 23(2), 208-214.

[60] Valdes, T. I, Moussy, F., 1999. A ferric chloride pre-treatment to prevent calcification of Nafion membrane used for implantable biosensors. Hosens. Hoelectron. 14, 579-585.

[61] Wang, J., Zhang, X., 2001. Needle-type dual microsensor for the simultaneous monitoring of glucose and insulin. Anal. Chem. 73, 844-847.

[62] Ward, W. K., Jansen, L. B., Anderson, E., Reach, G., Klein, J-C, Wilson, G. S., 2002. A new amperometric glucose microsensor: in vitro and short-term in vivo evaluation. Hosens. Hoelectron. 17, 181-189.

[63] Ward, W. K., Wood, M. D., Troupe, J. E., 2000. Rise in background current over time in a subcutaneous glucose sensor in the rabbit: relevance to calibration and accuracy. Hosens. Hoelectron. 15, 53-61.

Claims

[1] A glucose sensor system for quantifying blood glucose concentration, comprising two electrodes in which one is an enzyme electrode and the other is a non-enzyme electrode having the same membrane composition and morphology with the enzyme electrode but enzyme in which calibration for quantifying blood glucose concentration is performed by direct measurement of background current using the non-enzyme electrode.

[2] The glucose sensor system as set forth in claim 1, wherein both the enzyme electrode and the non-enzyme electrode are implanted in a subcutaneous tissue.

[3] The glucose sensor system as set forth in claim 1, being a telemetry system capable of wireless data communication.

[4] The glucose sensor system as set forth in claim 1, wherein the two electrodes are connected to a multi-channel potentiostat.

[5] The glucose sensor system as set forth in claim 1, wherein the two electrodes are integrated into a microelectromechanical system.

[6] A calibration method for quantifying blood glucose concentration which comprises a step of measuring background current, wherein the background current is directly measured by a non-enzyme electrode such that two-point calibration is achieved with a single measurement without any external injection of glucose.

[7] A method for quantifying blood glucose concentration accompanying measurement of current output from an enzyme electrode, wherein the method comprises a step of carrying out calibration by background current, the background current being obtained, without any external injection of glucose, from direct measurement using a non-enzyme electrode having the same membrane composition and morphology with the enzyme electrode but enzyme, and two-point calibration is achieved with a single measurement without any external injection of glucose.