Biosensors for intracorporeal measurements–problems

and strategies

George S. Wilson, Gerard Reachf, Daniel R. Thevenot

To cite this version:

George S. Wilson, Gerard Reachf, Daniel R. Thevenot. Biosensors for intracorporealmeasurements–problems and strategies. Biosensors, Elsevier, 1991, 19 (1), pp.9-11. <hal-01179876>

HAL Id: hal-01179876

https://hal.archives-ouvertes.fr/hal-01179876

Submitted on 27 Jul 2015

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinee au depot et a la diffusion de documentsscientifiques de niveau recherche, publies ou non,emanant des etablissements d’enseignement et derecherche francais ou etrangers, des laboratoirespublics ou prives.

^ °)-l\ (l^MBiosensors

better understand the normal and pathological path¬ways of several metabolites in medical care and willfind many applications in the study of their altera¬tion or restoration to normal values.

1. Mascini, M. & Palleschi, (j. (1989) SelectiveElectrode Rev. 11, 191-264

2. Collison, M. E. Ik Meyeroff, M. E. (1990) Anal. Chem.62, 425A-437A

3. Matthews, I ). R., I lolman, R. R., Bown, E., Watson, A.,Steemson, J., I lughes, S. & Scott, I). ( 1987) Lancet i,778-779.

4. Updike, S. J. & I licks, G. P. (1967) Nature (London)214, 986-988

5. Scheller, F. W., Pfeifer, I)., Schubert, R, Renneberg,R. & Kirstein, I). (1987) in Biosensors: Fundamentalsand Application (Turner, A. P. R, Karube, I. &Wilson, G. S., eds.), pp. 315-346, Oxford UniversityPress

6. Turner, A. P. F. & Pickup, J. C. (1985) Biosensors 1,85-115

7. Fogt, H. J., I )odd, L. M. & Clemens, A. 11. (1978) Clin.Chem. 24, 1366-1372

8. Mascini, M., Fortunati, S., Moscone, I)., Palleschi, G.,Massi-Benedetti, M. is: Fabietti, P. (1985) Clin. Chem.31,451-453

9. Mascini, M., Mazzei, R, Moscone, 1 )., Calabrese, G. &Massi-Benedetti, M. ( 1987) Clin. Chem. 33, 591-593

10. Albisser, A. M. (1986) in Advanced Models for theTherapy of Insulin-dependent Diabetics (Brunetti, P.& Waldhausl, W. K., eds.), vol. 37, pp. 165-169,Serono Symposium, Raven Press, NY.

1 1. Palleschi, G., Mascini, M., Bernardi, L„ Bombardieri,G. & De Luca, A. M. (1989) Anal. Lett. 22,1209-1220

12. Del Prato, S., Ferrarini, R & Defronzo, R. A. (1986)in Methods in I )iabetes Research (Clarke, I,., Larner,J. & Pohl, S. L., eds.), vol. II, pp. 52-68, John Wiley,New York

13. Palleschi, G., Mascini, M., Bernardi, L. & Zeppilli, P.(1990) Med. Biol. Eng. Comput. 28, B25-B28

14. Karlsson, J., Nordesio, L. ()., Jjorfeldt, L. is: Saltin, B.(1972) J. Appl. Physiol. 33, 199-203

15. Karlsson, J. is: Jacobs, I. (1982) Int. Sports Med. 3,190-201

16. Wassermann, I\. & Mcllroy, M. B. (1964) Am. J.Cardiol. 14, 844-852

17. Mader, A., Leisen, 11., Heck, II., Philippi, II., Rost,R. J., Schurch, P. & Hollman, W. (1976) SportarztSportmed. 4, 80-1 12

18. Wahren, J. ( 1979) I )iabetes 28, 82-8819. Ungerstedt, 1. (1984) in Measurement of Neuro¬

transmitter Release in vivo (Marsden, C. A., ed.), pp.81-105, John Wiley, New York

Received 22 August 1990

Biosensors for intracorporeal measurements: problems and strategiesGeorge S. Wilson,* Gérard Reachf and Daniel R. Thévenoti

■'Department of Chemistry, University of Kansas, Lawrence, KS 66045, U.S.A., fService de Diabétologie,Hôtel-Dieu, I Place du Paris de Notre Dame, 75004 Paris, France and -Laboratoire de Bioélectrochimie et Analysedu Milieu (L.A.B.A.M.), U.F.R. de Sciences et de Technologie, Université Paris Xll-Val de Marne, Avenue du Général

de Gaulle, 94010 Créteil Cédex, France

IntroductionElectrochemical biosensors and, in particular,enzyme-based sensors have found wide applica¬tions in the measurement of specific species in com¬plex media such as biological, industrial andenvironmental samples. Although relatively limitedin number, analytical equipment using such bio¬sensors has been developed in the I '.S.A., Kuropeand Japan and are commercially available for suchapplications. Since the first publication bv Clark &Lyons | I |, one of the main challenges for such bio¬sensors has been, and still is, their implantation invivo, either for continuously monitoring metabolitesor drugs, especially in intensive care units, or forcontrolling artificial organs, such as insulin pumpsused by diabetic patients |2|, or haemodialysis

Abbreviation used: ( i( )x, glucose oxidase.

units. For more than a quarter of a century a largenumber of publications, review, books, workshopsand symposia have been devoted to this topic. Seenfrom outside, no apparent success or improvementhas been obtained, since no operating implantablebiosensor is presently available. This report willattempt to present some of the real improvementsobtained and the various strategies recentlydeveloped, mainly illustrated with European ex¬amples. Indeed, a biomedical engineering Europeanconcerted action (BME-COMAC) has been estab¬lished since January 1989 on 'Chemical Sensors forin vivo Monitoring', under the leadership of A. P. F.Turner.

Biosensors: principlesBiosensors mav be variously defined but the nameis usually restricted to chemical sensors monitoring

1991

Biochemical Society Transactions

species using biological molecules or reactions fortheir selective molecular recognition procedure.Thus pi I or oxygen sensors, when implanted invivo, are preferably called 'bioprobes1 instead of'biosensors' [3].

All biosensors are analytical devices in whicha molecular recognition system is closely associatedwith or integrated to a system transforming thischemical information into an electric signal [4]: (1)the sensor selectivity is based on a biologicalmolecular recognition system: immobilized orretained enzymes, antibodies or membrane recep¬tors, or biologically integrated systems (plant oranimal tissues, micro-organisms); today mostimplanted biosensors use immobilized enzymes; (2)a physico-chemical transducer or detector monitorsthe molecular recognition system: it may be electro¬chemical (potentiometric, amperometric, coulo-metric, ionic conductivity, surface charge-field effecttransistor), thermal or calorimetric (enthalpic) ormass specific (piezoelectric crystal); we will restrictthis paper to the first type of detectors, i.e. electro¬chemical, since they are the most widely used, and,even more specifically, to amperometric enzyme-based sensors.

Operating properties of biosensorsAlthough the detailed and accurate modelling ofbiosensors is not always available, their behaviour isgenerally understood, their rate-limiting steps con¬trolled and their operating parameters well defined.Three types of operating parameters are of import¬ance when these biosensors are used for clinicalanalysis in vitro: ( 1 ) analytical parameters charac¬terize their patterns as analytical tools (backgroundsignal, sensitivity, linear range, response time, preci¬sion, selectivity, sensor life-time and sensor/samplesize); (2) signal-controlling parameters may beeither physical (local hydrodynamics, membranepermeability, temperature), chemical (pll, buffercapacity, ionic force, cofactor concentration level) orbiological (concentration level of molecular recogni¬tion species) or, finally, the sample compositionitself, i.e. the level of interferents or inhibitors formolecular recognition or transducer reactions; and(3) sensor management methodologies include thecalibration procedure but also the evaluation ofthe above-mentioned analytical parameters.

As these operating parameters may be easilycontrolled when experiments are made in vitro,such devices have proven reasonably reliable. Suchevaluation in vitro enables the selection of sensorspresenting characteristics suitable for each applica¬

tion in vivo. Nevertheless, conditions of such evalu¬ation in vitro have to be defined in order to

approach actual environmental conditions in vivo.Evaluation in vivo is even more complex

since, besides the choice of animal model, site andmethod of sensor implantation, procedures have tobe found for modifying the metabolite level in sucha way that analytical performance of these sensorscan be determined accurately during extendedperiods of operation. These operating parametersmay not be directly measurable (for example, back¬ground signal, calibration, selectivity, influence oflocal hydrodynamic conditions, etc.) and specificdifficulties may be encountered (e.g. miniaturization,maximum of linear range, site of implantation,clotting on outer membrane, inflammatory andimmune reactions to the implant, sterilizationprocedure, etc.).

Strategies recently developed for invivo glucose sensorsThe most widely studied type of implanted bio¬sensor is definitely the glucose one. It is based uponthe /3-n-glucose oxidation by oxygen in thepresence of /3-n-glucose oxidase (GOx). Threemajor strategies have been developed and tested onshort-term animal or human experiments [5]: (1)cathodic detection of oxygen depletion by GOx inthe presence of glucose, using a specially designedelectrode for restricting oxygen partial pressuredependence of the response in blood vessels(Gough et al., San Diego, U.S.A.); (2) anodic detec¬tion of hydrogen peroxide produced by GOx in thepresence of glucose: after the pioneer work ofShichiri [6], several groups have developed similarstrategies (Ege, Copenhagen, DK; Fischer et al.,Karlsburg, G.D.R.; Koudelka et al., Neuchatel,Switzerland; Pfeiffer et al., Ulm, F.R.G.; Reach et al.,Paris; Vadgama et al., Manchester, U.K.); and (3)anodic detection of GOx reduced by glucose, usingferrocene-type mediators (Pickup et al., London,U.K.).

These strategies may be discussed togetherwith specific problems for in vivo glucose sensors:(1) miniaturization of sensors using needle-typegeometries suggested by Shichiri et al. [6, 7]; (2)deposition of active enzyme layers, using the />-benzoquinone covalent immobilization procedureor glutaraldehyde reticulation [7, 8]; (3) choice ofthe site of implantation, i.e. vascular or subcuta¬neous [9] ; (4) calibration procedure, i.e. determina¬tion of background signal and sensitivity in vivo 15,7]; and (5) biocompatibility assessments.

Volume 19

Biosensors

These examples demonstrate the importanceof a multidisciplinary approach of in vivo chemicalsensors and of a tight collaboration betweenphysico-chemists (involved in, for example, analy¬sis, electrochemistry, polymer science) and clini¬cians (in the fields of diabetology, surgery andbiomaterial science) to solve the numerous prob¬lems and difficulties encountered. They also showthat significant improvements have been obtainedallowing short-term in vivo implantation, but thatvery difficult problems have arisen for controllinginteractions between biosensor and tissue, i.e.modifications of the sensor by the tissue (e.g.clotting of the external membrane or layer) as wellas modifications of the tissue by the sensor (e.g.toxicity, inflammatory and immune reactions), bothreactions being usually described as biocompati-bility of the implant.

This work has been supported by the Caisse Nationalede l'Assurance Maladie des Travailleurs Salariés (France)(grant CNAMTS-INSBRM no. 85.3.54.8.E), FrenchMinistry of Foreign Affairs, National Institutes of I lealth

(U.S.A.) (grant DK 30718). Their financial help is grate¬fully acknowledged.1. Clark, I,. C. & Lyons, C. (1962) Ann. N.Y. Acad. Sci.

102, 29-452. Thévenot, D. R. (1982) Diabetes Care 5, 184-1893. Turner, A. P. F. (1987) in Biosensors: Fundamentals

and Applications (Turner, A. P. F., Karube, I. &Wilson, G. S., eds.), pp. v-vii, Oxford University Press,Oxford

4. Thévenot, D. R. (1989) Spectra 2000 17, 55-575. Velho, G., Froguel, Ph., Thévenot, 1). R. & Reach, G.

(1988) Diabetes, Nutr. Metabolism Clin. Exp. 1,227-233

6. Shichiri, M., Yamasaki, Y., Kawamori, R., I lakui, N. &Abe, H. (1982) Lancet ii, 1129-1131

7. Sternberg, R., Barrau, M.-B., Gangiotti, L., Bindra,D. S., Wilson, G. S., Velho, G., Reach, G. & Thévenot,I). R. (1989) Biosensors 4, 27-40

8. Sternberg, R., Bindra, D., Wilson, G. S. & Thévenot,1). R. (1988) Anal. Chem. 60, 2781-2786

9. Tallagrand, T., Sternberg, R., Reach, G. & Thévenot,I). R. (1988) Horm. Metab. Res. Suppl. 20, 13-16

Received 22 August 1990

Problems of clinical data interpretationPankaj Vadgama, Mohamed Desai, Zarah Koochaki and Paul Treloar

Department of Medicine (Clinical Biochemistry), University of Manchester Diabetes Research Centre,Hope Hospital, Eccles Old Road, Salford M6 8HD, U.K.

Biosensors offer the prospect of decentralized and,where necessary, continuous monitoring of a widerange of biochemically important analytes. Theyconstitute a means of formatting a biological recog¬nition molecule in an immobilized, reusable formand, more particularly, one which is characterizedby structural simplicity and reagent economy aswell as by the functional convenience resulting fromclose juxtaposition to the transducer element.Though much work has gone into the charac¬terization of the biolayer and its transducer inter¬face, comparatively little attention has been paid tothe external interface between the device and thesurrounding biomatrix. The consequent lack ofdetailed knowledge on this aspect has been a signifi¬cant impediment to optimum clinical use and,therefore, ultimately to reliable data interpretation.

Interactions of the biosensor and its bio-environment can be subdivided into those due to: (i)solution variables; (ii) responses from colloid- andcell-containing fluids; and (iii) a concerted tissue

response able to fundamentally change the nature ofa biosensor implantation site. These constitute ahierarchy of increasingly complex problems which,from initial contact, distort generated signals fromeven the most sophisticated and well-fabricateddevice. In the main, the final common pathway offailure is a progressive loss in signal, the decaykinetics of which are too poorly understood andirreproducible to be either modelled accurately orto serve as a firm basis for correcting signal drift inan individual case. Also, superimposed upon thismonotonie decay may be fluctuations in signal sizeover time-spans varying from minutes (Fig. 1) tohours |1|; these can be difficult to separate intoanalytical noise and true biological variation.Specific delineation of the latter could potentiate thediagnostic value of continuous monitoring bio¬sensors, as there are undoubted changes in the vari¬ability shown by individual biochemical parametersin different disease states |2|. Discussion here willinclude in vivo (),- and ion-selective electrodes,

1991