`
`RECENT ADVANCES IN CLINICAL BIOCHEMISTRY, C.P. PRICE,
`
`K.G.M.M. ALBERTI; CHAPTER 12, “PRINCIPLES AND PRACTICE
`
`OF DRY CHEMISTRY SYSTEMS,” R.L. STEINHAUSEN, C.P.
`
`PRICE
`
`Infopia Ex. 1023 pg. 1
`
`
`
`
`
`
`'
`
`Récén’ét AdvanCes in.
`.3 CLENECAL .
`'
`VELNCLLEfifiEgiflgL?
`
`
`EDITED BY
`
`0. P. PRICE
`‘K. G. M. M. ALBERTI
`
`NUMBER THREE
`
`DEVflan):DEV
`
`'
`CHURCHILL IIVINGSTONE
`EDINBURGH LONDON MELBOURNE AND NEW YORK 1985
`
`Infopia Ex. 1023 pg. 2
`
`
`
`
`
`
`
`
`w.,5F.—r
`
`CHURCHILL LIVINGSTON]!
`MedicnlDivisibn ofLongman Group Limited
`Distributed in the United States of America by Churchill
`Livingstone Inc. , 1560 Broadway, New York, NY
`10036, and by associated companies, branches and ‘
`representatives throughout the world.
`
`© L'ongman Group Limited 1985
`
`All rights reserved. No pdrt of this publication may be
`reproduced, stored in n retrieval system, or transmitted
`in any form or by any means , electronic, mechanical,
`«photocopying, recording or otherwise, without the prior
`permission of the publishers (Churchill livingStone,
`Robert Stevenson House, 1—3 Baxter’s Place, Leith
`Walk, Edinburgh E111 BAF).
`
`. First published 1985
`ISBN 0 443 02.797 8
`ISSN 0 143—6767
`
`British Library Cataloguingin Pttblication Data
`Recent advancesih clinical biochémlstty.—No. 3
`'
`1.,Blological chemistry 2. Chemistry, Medical
`and pharmaceutical
`612'015
`RS403 ‘
`
`Printed in Great Britain by The Pitman Press, Bath
`
`Infopia Ex: 1023 pg. 3
`
`
`
`
`
`
`
`
`Contents
`
`
`
`W,.-P_.,.._.-
`
`.Chemiluminescencein medical biochemistryAnthony K Campbell
`Ma7y R Hoh Ashok Patel
`
`The diagnostic potential ofnuclear magnetic resonanceBn'én D. Ross
`
`Monoclonal antibodies in clinical biochemistry Kennett};Siddle .
`
`Acute phase proteins—‘3 group of protective proteins Carl-BertilLaurell
`
`. The investigative value of bile acid measurements in man
`Iain W. Percy-Robb
`
`.
`
`. Hyponatraemia Geofl‘ey V. Gil-l Cecil 'T. G. Fleur
`
`Role ofopiate receptors in the. control of reproductive function
`Emilio del Pogo
`
`.’ Tubeless tests of small intestinal function M . F.‘ Laker K. Bartlett
`
`The biochemicai'inves ligation ”of respiratory disease A. G. Leitch
`
`10.
`
`ll.
`
`The biochemistiy of burninjury Peter GShdkespear3
`Edward} Coomb‘es Gifi‘ohdF Batstone
`
`Bedside monitoring: the role of electrodes P. Vadganuz
`K. G. M’ M Albem‘
`‘
`
`12.
`
`Principles and practice of thy chemistry systems R.’.L Steinhausen
`C P. Pfice
`
`‘ Index
`
`31
`
`63
`
`'103 .
`
`125
`
`149
`
`160"
`
`195
`
`221
`
`239 ,
`
`‘ 255
`
`273
`
`297 '
`
`
`
`‘.K._....._.._...._.....
`
`Infopia Ex. 1023 pg. 4.
`
`
`
`
`
`
`
`
`
`...._—...V~..,_’_._a._'..,...._._._:1..—--mw‘a—VM
`
`12 Principles and practice of dry chemistry
`systems
`
`R. L. Steinhatésen C. P. Price
`
`INTRODUCTION
`
`Laboratory investigations have become an integral part ofthe clinicians decision,
`making. process,
`in the hospital and~ the community onyitonment, both in the
`diagnosis and monitoring of therapy. Laboratory investigations have received greater
`recognition in recent years due, in large part, to developments'111 analytical techniques
`and our increasing knowledge of the pathophysiological changes associated with
`disease More recently the plethora ofdiagnostic procedures available to the clinician
`has focused attention on the relative merits, both economic and patient orientated, of
`. each test and this no doubt will consolidate the Value of laboratory investigations
`The develdpment of large automatic analysers has111 recent years been associated
`withthe philosophy of centralisation of laboratory facilities However, more recently
`the limitations of this philosophy have become apparent; there is now growing
`emphasis on the potential for undertaking analytical tests nearer to the patientwiu
`the hospital bed, at the outpatient clinic or in the patient’s home. It is envisaged that
`.this will promote a more intimate relationship betWeen medical examination, clinical
`history and laboratory diagnosis. Expressions such as
`‘bedside chemistry’ and
`real-time analysis’ have been adopted to describe such patient-orientated examination
`methods.
`~
`These new diagnostic procedures employ analytical principles that have already
`been used successfully'1n the clinical laborato1y f01 25 years in the form of the familiar
`urine and blood test strips; these new test formulations aie now commonly referred to
`. as ‘dry chemistry’ systems. Actually, the expression ‘dry chemistry’ is misleading:
`chemical reactions cannot proceed under absolutely dry conditions, and generally '
`need tobe mediated by a solvent. The expression ‘dry chemistry’, as currently
`understood, simply means that all the reagents and auxiliary substances necessary for
`his reaction are embedded iu- a paper or plastic matrix. in thei1 dry state,
`thus
`obviating the need to prepare reagent solutions In order to initiate the 1eaction,
`however, these new procedures still require a solvent in which the substance to be
`determined is dissolved,in the clinical laboratory the solvent will be the sample,
`' namely urine, serum, plasma 01 whole blood.
`Despite this fundamental error of definition, the term dry chemistry has become
`'widely established and provides a reasonably good way of describing the latest
`diagnostic systems as opposed to the conventional methods. In the past the results
`were evaluated purely visually by comparing the reaction colours that develop with a
`table of reference colours, more recently the use of reflectance photometry has
`enabled the production of quantitative results that are comparablewith those of the
`classical photometric methodswith respect to precision and accutacy (Drucker et al,
`r73
`
`'
`
`Infopia' Ex. 1023 pg. 5
`
`
`
`
`
`
`'7v~ey~m
`
`".1-
`
`.
`
`'
`
`..
`
`r‘
`
`. »
`\
`
`'.
`
`3‘ 274 enemas/mam CLINICAL BIOCHEMISTRY
`
`.
`
`I.
`
`9'
`
`
`
`" ”
`
`T983). In‘addition, the measurement range that is available without dilution is in .
`principle greater than in the familiar transmission photometry systems.
`>
`_
`‘Dry chemistry systemshave received even more attention since it became possible,
`as the result of comprehensivedevelopment activities, to use them for the'simp‘lc,‘
`‘ rapid determination of the entire rangeof classical biochemical parameters in seruin,
`plasma and- even whole blood; Furthermore, it has also been shown that homogenous
`iinmunoaSSay techniques can 'also be adapted to these dry chemistry systems. ‘
`'A limited number of reviews on developments in this field have appeared in the
`literature (Zipp, 1981; Walter, 1933).
`,
`‘
`.
`
`-
`
`(
`
`'
`
`4 -
`
`HISTORICAL DEVELOPMENT
`
`'
`
`.
`
`‘
`
`'
`
`The-'use of substances borne by reagent carriers already had a long history, even
`outside the field of clinical laboratory diagnostics, the “most striking example being the
`- use of litmus paper to determine the pHof aqueous solutions.
`'
`~
`‘.
`'
`a
`‘Among the first diagnostic aids to use dry reagents were the test strips'for
`'. determination of glucose in urine,-which have been in use since the 19503 and have '
`“assumed an important role in the diagnosis and management of diabetes mellitus
`(Clinistix®, Ames; SuGlulcotest®, Boehringer Mannheim)“
`.
`.
`.
`‘
`One by one,” similar analytical principles were used for. other clinically relevant
`.urine paramaters;
`the first visualisation of single erythrocytes on a test strip
`(Sangur—Test®, Boehrlnger Mannheim) in 1974 and the introduction of an enzymatic '
`. leukocyte assay in urine, also on a test strip (Cytur~Test®, Boehringer Mannheim) '
`were notable milestones in the development of the technique. Today as many, as nine"
`’parameters can be reliably assessed simultaneously simply by dipping the‘appropriate
`. emultiple test strip (N-Multistix SG®, Ames; Combur~9~Test®, Combur«9-Test®U,
`"Boehringer Mannheim; Rapignost® Total-Screen A, Behring; Medi~Test Combi 9®,
`Macherey & Nagel) into urine and reading it visually or by a multi-channel,
`instrument (Clinitek®, Amos; urotron®, Boehringet Mannhehng’Rapimat®, Behring).
`A comprehensive review on this development and the technical principles has been
`given by Kutter (1983).
`'
`r
`V
`.
`'
`The routine testing of urine has acquired a new perspective, particularly as a result
`of the introduction of the enzymatic leukocyte test strip. Microscopic evaluation of
`the urine sediment is no longer obligatory asa‘n initial test: only when one of the
`' parameters. on the'test strip is pathologically altered will a careful examination of' the .
`. sediment provide further information for thesubsequ‘ent diagnosis. Consistent use of ‘
`.the so—called‘test‘strip sieve, comprising the parameters nitrite, protein, leukocytes
`and erythrocytes, can drastically reduce the number of sediments that have tobe
`examined (Kutter, 1982; B‘onard ct al, l9,823 Briihl et a1, 1982).
`,
`-
`The urine density or specific gravity, which under. standardised conditions and with -
`‘ corresponding preparation of the patient, can also indicate the efficiency of the kidney
`functionj this investigation has recently become available as ‘a rapid test ona urine test
`strip (N—Multistix SG®, Amos), and evaluation data has been presented by Kutter 8: g
`Holtzmer (1982).
`‘
`‘
`In this case the ion concentration in the urine is relied upon for the. determination of
`"the urine density; however,
`in some cases this can lead. to deviations from the :
`refractometric determination; for example, discrepancies have been observed when
`
`lnfopia Ex; 1023
`
`pg. 6
`
`
`
`
`
`PRINCIPLES AND PRACTICE OF DRY CHEMISTRY SYSTEMS
`
`275
`
`1
`‘N -1 1 v 1 .
`u
`’.
`'¢'y'. 1'1 m 1 u 1‘"’y'x’fll'.’’‘n'1“"- v'-.'."’u'-''A‘J'1'o's‘.‘.‘.‘.'»'-'.‘.'.'. 1'.£2. ,
`
`’ paper
`
`'a'y''y'1h‘a'.31}.
`
`u1‘1'1'1"cl:”
`
`.. ‘ ‘
`
`.'.".v1:‘.‘1‘.'1'u.....11‘1’
`III
`IIIIIIII
`
`lodeteflmpregnaled mesh layer
`
`carrler toll ‘
`
`reagent paper
`1 absorbent
`
`Fig. 1211 I Cross‘sscti'on‘of the new blood test patch on Combur—9~Tcst® urine test stripsimanufacturer:
`Boehringe: Mannheim).
`
`increase in the Urine density have been due to an'increas‘e in the concentration of
`nomionic substances (Adams,: 1983).
`'
`. Many of the assay reactions exploited in urine strips are based on chemical reactions
`that involve oxidising or reducing agents. A substance that is known to interfere in
`inany assays is ascorbic acid, which may lead to depressed or even falsennegatlve
`. rcsults'11:1 the test fields for blood'1n the urine (Daae & Joell, 1983).
`Therefore, some manufacturers (Behring, Machercy ’& Nagel) haVe included a test
`for ascorbic acid in their multiple test strips, whiCh can, with a positive finding,"
`indicate decreased values with other parameters.
`. Animportant aim in the further development of dry chemistry 1eagent carriers,
`especially1n urinalysis, has been to eliminate spurious influences so as to improve
`always the reliability of the assay methods.
`.
`,
`.
`~
`As one of the consequences, ascorbic acid interference has now been ahnost‘
`completely eliminated from the blood test patch of some test strips (Sangur-Test®
`Comburr9~Test® Boehringer Mannheim) by incorporating an iodatc—irnpregnated
`mesh The iodate adhering to the mesh oxidises any ascorbic acid in the urine befo1e it
`can interferewith the actual test patchm which the assay reaction'takcs place (Fig. ‘
`ii 1) (Nagel et a1 1982)
`1
`1
`a .1
`\
`
`‘
`
`.
`
`1
`
`'
`
`.
`
`lnfopia‘Ex. 1023 pg. 7
`
`1
`I
`
`
`
`
`
`‘
`
`.
`
`f
`
`‘
`
`'276 ‘ scream ADVANCES IN CLINICAL BrocrtsMIsrRY
`Pi further challenge which faced the manufacturers of urine test strips-‘idha long' 7'
`time was to extend the simplicity of application of urine test strips to the biochemical
`analysis of blood, serum and plasma. Here,
`too, the. emphasis was primarily on '
`' glucose, since there were obvious potential benefits to the use of such test strips in the
`‘ management of diabetes, including the growing area of home monitoring by’ the
`patient. Test s'trlpsof this type for visual interpretation became available in the early.
`19605 (Dextrostix®; Ames; Haemo~Glul<otest®, Boehringer Mannheim) and became
`’ indispensable in many areas of clinical practice.
`.
`'
`.
`Following further development work; a blood glucose test strip with a two-colour
`test patch which permitted nearly quantitative blood glucose determinations over the
`entire clinically relevant range of 1.1—4410mmol/l (20—800 mg/d1) was introduced in
`31979, (Haemo-«Glui<otest® 204300, Boehringer Mannheim). Evaluation reports have
`~ been published by Frey et 2110979), Unger & Willms, (1980), Whifcn‘d et al, (1980): i
`A second two~patch test strip (Visidex®, Ames) has been introduced with compara»
`. his results being achieved (Clark et al, 1983).
`'
`~
`‘
`~
`'
`' Despite the fact that many diabetologists consider the results obtained .with the
`A visually indicating test strips to be sufiiciently precise'and reliable for the majority of
`areas of use, there has been a need expressed for instruments capable of producing
`'quantitative results; this has led to the development ofseveral reflectance photometefi, ‘1
`Most of these photometers still require manual calibration stepsl(Eyetone®,
`'Dextrometer®,, Glucometer®, Ames)——-additionally lot—specific scales (Reflomatg,
`lBochringer Mannheim) or code strips (Reflolux®, Boehringer Mannheim) to reach
`- results with improved precision and accuracy (Kattermann 8t Naroska, 19743.,
`Summer 8: Herbinger, 1975; sommer 8t Hohenwallneu, 1976; Kubilis. et 2113,1981;
`Walford et al, 1984).
`'
`The introduction of new, compact, nncroprocessor-controllcd instruments with
`automatic calibration (Reflocheck®, Boehringer Mannheim) produced. a decisive
`further simplification of the assay procedure for the user without any loss of precision
`‘ or-accuracy (Kattermann et a1, 1983;1(030hinsky er al, 1984).
`1
`In the course of the 19703 it became apparent that many users'of dry chemistry
`systems considered there to be a need for such systems to determine further.
`clinically—important parameters in serum, plasma and whole blood. However, the
`development of the requisite systems posed many problems: highly complex mixtures
`of substances, often including sensitive enzymes, had to be applied to carrier materials
`whilst maintaining reagent eficacy, and this required a great deal of. research.
`‘ Real-time analysis imposed the additional requirement that .the results be provided
`within. 2—3 mihutes without sample preparation, while of courSe maintaining the.
`precision and accuracytof conventional photometric methods.
`The development of modern dry chemistry reagent carriers has profited from
`technological advances in various areas such as the manufacture and processing of thin
`films, paper carriers and synthetic porous carrier materials, and not least the ’
`. development .of colour photography, which makes similar use of quantitative chemical
`reactions within a solid matrix.
`‘
`Manufacturers of test strips have developed precise techniques for cutting paper
`and foil, for laminating these support media, and for subjecting them to the
`appropriate pretreatment. These advances required a detailed'ltnowledge of .the
`response of. a very wide range of materials to body fluids and their influence on the _'
`
`’
`
`
`
`Infopia Ex. 1023 pg. 8
`
`
`
`PRINCIPLES AND essence or DRY CHEMISTRY SYSTEMS
`
`277,
`
`‘ substances contained in them. It has become clear that important properties of these
`materials such as layer thickness, fibre strength and absorption characteristics can
`now be controlled on a large scale, allowing the production oflarge batches of reagent
`carriers with consistent quality.
`The use of microprocessor technology has led to considerable advances in the
`simplification of the analytical procedure, so that the user generally only has to .
`perform a few steps to get a result and all manual calculations are eliminated, the
`reading being given in the desired concentration units In addition the microprocessor
`can facilitate many of the verification procedures necessary to achieve reproducible .,
`instrument—performance.
`The latest developments have shown that the dry reagent technology may be also
`extended to immunological systems. The prerequisite has been the invention of
`homogeneous immunoassay techniques.Eortexample, the substrate labelled fluoro-
`immunoassay (SLFIA) has been adapted to a reagent carrier for the measurement of
`serum theophylline (Walter, Greenquist 8: Howard, 1983). More recently,
`the
`prostetic group ligand homogeneous assay (PGLIA) described by Morris et a1, 1981
`has beenadapted to a dry reagent car1ier system (Tyhach, 1981)
`A homogeneous immunoassay for morphine has also been adapted to a reagent
`strip The. assay employs an enzyme channelling immunoassay technique With a
`glucose oxidase—horse radish peroxidase enzyme pair (Litman et 211,1983)
`'
`Since these assay techniques have only become available very recently, little
`performance data have been published up to now.
`In the following sections the basic principles of the newer dry Chemisuy systems for
`the analysis of classical blood parameters are discussed, includingsome theoretical
`aspects of reflectance photometry, and with the dry chemistry products that have so
`far been developedby the various research groups.
`<
`
`-
`
`BASIC PRINCIPLES
`
`~
`Construction of dry—reagent ca11ie1s'
`Practically all the existing systems for the determination of clinical biochemical
`parameters in blood, serum or plasma on the basis of dry~reagent carrieis consist in
`principle of three parts:the 1eagent part, the reflective part and the support part.
`Thus, a dry chemistry system may comprise just three basic layers, or may be
`supplemented by further layers soch as a spreading layer or a plasma separation layer.
`In Figure 12.2 the constructibn of dry-reagent cariiers, as offered by several
`manufacturers is shown. Generally, the suppo1t part consists of tough plastic material
`which may be'filled withinorganic pigments, particularly if'it is also intended to settle '
`as a reflective surface.
`The essential function of the reflectiVe pa1t is to reflect the incidentlight at the
`moment of measurement onto the detection system without any loss by absorbance,
`thus ensuring the high sensitivity of the measuring process in some cases even’ the'
`reagent support medium may serveas a reflective part, particularly when opaque foils
`or paper layers are used
`Normally, the reagent part contains, in its dry form, all the reagents and auxiliaiy
`‘ substances that are required f01 the specific detection reaction, they are reconstituted .
`by adding thefluid that1s to be tested Owing to the complexity ofthe reaction that183‘.
`
`I
`
`-
`
`‘
`
`‘
`
`i
`
`_
`
`‘
`
`lnfopia Ex. 1023 pgjs
`
`
`
`
`
`
`
`
`"..-.':—.—.u—,~~—
`
`‘~.'¢.
`
`4
`
`.
`
`.A.v
`.‘.
`.,
`
`.
`
`278
`
`RECENT ADVANCES IN CLINIGAI; BIOQEEMisqgfiy'“ ‘
`.
`‘
`.
`ééNMSample.‘
`.
`..
`
`'
`
`J
`
`..'_
`.'
`.‘-.
`
`
`
`
`
`
`Serum Sample
`
`b)
`
`
`
`
`
`
`
`
`Mount
`
`Spreading and reflective
`layer
`
`Reagent layer
`
`Indlculor layer ~
`
`Transparent support V
`
`
`
`Pmlanuva layer
`
`Transpatent prolecktva
`layer
`
`
`
`
`Hasma separation layer
`
`' ,
`9
`
`O Q Q'OV 4
`O‘Q‘Q’Q’o'o’
`3.92:0
`
`
`
`.‘
`SUPPOI"
`‘
`Plasma resewolr
`Fig. 12.2 Cross-sectionsof different dry chemistry reagent carriers. :1, Amos Semlyzer® system, b.
`Eastman Kodak Ektuchem® system, c. Boehringer Mannheim Reflotron® system.
`
`lnfopia Ex. 1023 pg. 10
`
`
`
`
`
`—~-—7|:x~'r.rr'v-
`
`‘PRINCIPLES AND PRACTICE or Dar CHEMISTRY SYSTEMS
`
`279
`
`to be carried, out, it is often necessary to embed the individual reagents and auxiliary
`substances in diiferent reagent layers in order to guarantee the requisite long lifetime
`of the reagentcarrier, e.g. by having a true‘reagent layer and arr-indicator layer. This;
`may be pardcularly‘important where individual reagent components are incompatible
`in solution (cg. enzymes and actiVators).
`
` ' msm. REAGENT LAYER
`
`MIDDLE REAGENV wen
`PHOXlMAL REAGENT LAYER”
`
`.
`
`_
`. .
`,
`-melmal Layer
`
`.
`
`Cellulose Fibre containing
`Gallon Exchange (nouns
`
`INDICATOR
`
`
`
`
`
`' W Polymorlu Sepumllon Layer
`
`
`
`A
`Dlalul Layer ’
`Cny_flW
`.
`
`J“
`
`Fig.‘12,3 Coatingpfiindividual cellulose fibres with dry chemistry reagents (Seralyzef‘E—BUN; Ames
`Company), Redmvn and modified from Zipp 8: Hornby.
`
`.
`
`.
`
`It may also be necessary to incorporate intermediate layers that isolate desired
`reaction products while blocking, degrading or masking undesired ones, as will be
`- shown later by the example of some urea tests (Fig. 12.6). Additional spreading layers
`may be added to ensure uniform penetration of the applied sample fluid into the
`reagent layer.(Shirey, 1983).
`‘
`,
`i
`- In some cases, ‘microvscale’ reaction layers are produced on individual fibres by '
`'seQuential coatidgof a' fibre matrix with different reagent solutions (Plischke, 1983) ~
`'(Fig; 12..s),'
`.
`.
`,
`‘
`.
`.
`.
`.The technically most advanced systems, to which even whole blood can be applied,
`' Contain a plasma separation layer in which erythrocytes are separated; the actuai'
`reaction then. takes plaée in plasma. The structural organisation that is’necessa‘ry. to ,_
`accommodate this entails spatial separation of the different parts (as shown in Fig.
`.
`.12,2);5'
`.
`‘2.
`.
`‘. ..'
`.
`.
`in the production process, the reagents are incorporated in their carrier by soaking
`the appropriate material in a reagentsolution or embedding the reagents in a synthetic
`film matrix.
`u":
`The-hidivif pal layersvare united by self—adhesive films, by laminating step by step ‘
` the casefofpaper layers, by combining all the layers. simultaneously.
`i\‘$~.
`5; Qfisim
`
`v
`
`‘
`
`Infopia Ex. 1023 pg. 11
`
`
`
`
`
`
`
`
`
`.280
`
`RECENT anvancssm CLINICAL Biochsmsrar'
`
`'
`
`.
`
`l,
`
`.1
`
`"13"}
`
`
`
`‘
`
`’
`
`..
`
`Reflectance spectroscopy
`.-
`As stated above, the reactions on dry chemistry reagent carriers are measured by
`means of reflectance spedtroscopy, which, depending on the application, permits not
`. "only endpoint determinations but also kinetic measurements with a varyingnur‘nher .
`of measurement points.
`,
`.
`.
`'
`'.
`‘
`1
`1
`.~
`Since reflectance spectroscopy is already yery. widelyused, cg. inthejbloocl glucose
`- measurement instruments of various manufacturers, and since it differs in certain
`respects from the absorption photometry normally being used‘in cliniéal’chefnistry,
`' the theoretical principles will now be briefly presented; a detailed presentation was
`published by Werner 81 Rittersdorf (1983):
`‘
`'.
`-
`f
`,
`.
`,
`When a light my enters a layer, the major effects that need to be considered are:
`Reflection (and, refraction) at the boundary surface
`‘
`5
`'
`'
`“
`Absorption within the layer, depending on the components it contains
`‘ ..
`;
`.
`a
`Scattering within the layer and‘at the boundary surface ‘
`-
`’
`~
`'
`From these three eflects it is possible to Workout the proportion of the lightthat
`re-emerges from the layer and the angle at which it emerges. All three‘efiec‘ts depend
`~ on wavelength, and the following discussion therefore refers to monochromatic light;
`In reflectance photometry, abasic term is the lightintensity‘IR that emerges from
`the medium in the direction opposite to the direction of incidence. IR is related to the .
`incident light intensity 10 by the reflecting power R:
`(a)
`~
`‘
`.
`7
`sax/Io
`where R lies between 0' and 1'. The reflecting power'is commonly expressed as the
`percentage reflecting power 100 R or %R:
`I
`‘
`(b)
`‘
`%R='(IR/IO)-1oo,. “
`,
`.
`_
`'1‘We limiting cases of reflection are distinguished!
`Specular reflection, i.e. reflection in the true sense as observed at shiny‘su'rfaces, in
`' which the angle of reflection is equal to the angle of incidence.
`.
`Difiuse reflection (usually called reflectance), which is the result of scattering within
`the layer and/or at irregular surfaces, The reflected light does not have a direction of ‘
`preference but is uniformly distributed over the half sphere.
`.
`Normally, in the course of a reaction, a characteristic specific colour change is
`produced on the reagent carrier and quantified by measuring the“var-iations in the
`reflectance at a defined wavelength, any components due to specular reflection being
`largely eliminated by the geometry of the measuring system.
`'
`Figure 12.4 shows such a measurement system—Ulhricht’s Sphere—Lin which
`directly reflected light is virtually incapable of reaching the detector.
`I
`'
`.
`There are various mathematical descriptions of the reflectance of lightscattering
`layers. The equations described by Kuhelka & Munlt (1931) have so far proved to he
`the most suitable.
`‘
`'
`'
`‘
`."r.
`‘._.
`"If it is assumed that: -
`y
`.
`the particles of a layer are small in comparison with the layer thickness
`the particles are uniform and are distributed statistically
`the layer is irradiated by diffuse light
`_
`there is no regular reflection
`
`.
`
`'
`
`‘
`
`.
`
`.
`
`.
`
`Infopia Ex. 1023 pg. 12
`
`
`
`281 .
`PRINCIPLES ant) emanation DRY CHEMISTRY sysrms
`with the further assumption that the layer is Opaque and of infinite thickness, so that
`>
`the bacltgtoniici is no longer visible, the Kubcil<a~Munl< function is obtained:
`'
`§~(1~R°°)2‘
`(Ci
`'-
`.-
`‘
`l
`s t
`ZRoo
`where K is the absorption coefficient of the substance to be determined, 8 the
`scattering coefficient and, R00 the (iiflr‘use reflectance of an opaque layer. If the
`integrating sphere
`
`'
`
`
`
`
`
`WW...“._......
`
` LED:
`
`‘
`D:
`'
`~
`DR:
`
`
`light-emitting diode
`
`detector
`retsmnce detector
`
`
`Fig. 12.4 i Uibticht’s sphere used as measuringsystem in BoehringerMannheim’s reflectance
`s
`photometers.
`scattering particles are present in excess, the scattering coefficient S can be tegardecl as
`constant and the concentration c of the substance to be determined is simply .
`proportional to the absorption coefi’icient K, so that a simple correlation .bethen ‘~
`,‘refle’ctance‘and concentration is Obtained:
`
`~
`.
`'
`'
`-
`'
`=‘5 ,(1 —'R<>o)2
`‘
`an
`2Roo
`c
`‘1
`(d) y ..
`‘ ¢=e(S’/en+<8/26u)'Rw«to/zest?“
`it)
`'
`Here the extinction coefiicient GR is comparable to, but—‘owing' to possible interac- ~
`fions between absorbing molecules and the reagent layer matrixv—not identical to the
`ET insolution that appears in the Beer—Lambert law.
`'
`'
`«
`In: generalthe'Kube'lka—Munk function in thisbasic .form cannot be fiscd i0"
`' evaluate reactions in reagent carriers because the underlying assumptions are not fully
`u
`,
`.
`4
`,
`satisfied.
`~
`.
`However, the relation between reflectance and concentration maybe expressed
`« .tnatliematically either by adding linear correction terms (Piischkc & Storz, 1982) on
`by iisixig any expedientnonviinearfunction that, like the Kubellca—Munk function (e),
`
`.
`
`’
`
`a
`
`'
`
`lnfopia Ex. 1023 pg. 13
`
`
`
`
`
`
`
`c
`
`
`
`'
`
`'
`
`.
`
`.
`
`t
`
`.
`{.2 a
`_.
`'
`1282‘ access sovaucnsIrvcunrcacsroounmsrnr
`is. a'combinatio‘n of 'a constant com‘ponentja linear -component- and ahyperbolic .
`component (Stilhler, 1983).
`‘
`'
`.
`~
`, Moreover, it is advisable to use functions that contain linear coefficientsysince only
`these functions can be unambiguouslysol‘ved by regression methods, cg: with the
`‘help of an electronic calculator.
`.
`One practical example is the relationship between reflectanCe and concentration:
`used in the'Reliot'ronGD-Systern":1
`'
`(f)
`‘
`cwA0+A1R+rual + Agar-s
`This equation contains an additiOnal correction term which,‘for n>0,1enablesr
`sigmoidal curves to be evaluated, whereas the original Kubell<a~Munk ‘function
`contains only a‘hyperboliccomponent. Equation (f) gives a hyperbolic curve forné 0.
`When the function curve is fixed to calibrate the system, the number of pairs of
`concentration-reflectance values that is used must be at least equal to the requisite
`number of coeificients. Moreover, it is advisable to measure a greater‘number ’01".
`~concentrations at the limits of the measurement range so as to enhance the reliability
`of the measurement method.
`.
`.
`‘ Transparent layers with an optically dense, scattering background as used in ‘
`dry-chemistry reagent carriers of some manufacturers (Shirey, 1983) represent special
`cases which will not receive further attention here. The theoretical aspects of this type
`A of system have been discussed by Williams & Clapper (1953).
`.‘
`.
`A further interesting aspect deserves attention.
`If the relationship between
`reflectance and concentration as given by the Rubella—Monk function is compared
`' with the relationship between transmittance and concentration as given; by the '
`Beer~Lambert law, it is found that the curvesare broadly similar.
`However, as the emerging light yield is generally greater when using transmission .
`photometry, it is essential to work. With diluted samples or with Very thin layers,
`whereas reflectance.measurements in the range in which the Kubelka—Muii’k function
`is valid are independent of layer thickness. Reflectance photometry therefore, in
`‘ principle, provides a means of performing measurements in undiluted sample
`material that are largely independent of dosing accurac .
`Moreover, the sensitivity of the transmittance curve decreases more rapidly than.
`the reflectance curve. This means that in principle a greater concentration range can
`,
`be covered by reflectance measurements than by transmittance measurements.
`Since the range where the concentration function base reasonable slope is smaller '
`in reflectance photometry than in transmittance photometry,'the requirements in
`terms of precision are generally higher for reflectance photometers. Hoyt/ever,
`experience with dry chemistry systems has shown,
`that, based‘ on the latest
`A development in the field of. electronic processing, the instruments work completely
`I
`satisfactorily. ,
`
`E l
`
`l"
`i}.
`
`Ii
`
`E
`
`,
`
`. TECHNICAL DETAILS AND PERFORMANCE DATA OF DRY .
`CHEMISTRY SYSTEMS
`The various research groups working on the development of dry chemistry systems ~
`for laboratory diagnostics have adopted differentrroutenghe results of three groups
`-1 Manufactured by Boebrlnget Mannheim
`
`lnfopia Ex. 1023 pg. 14
`
`
`
`,
`
`283
`PnINCIPLEs AND PRACTICE or our CHEMISTRY SYSTEMS
`'will be given by way‘ of example, including data on the analytical performance of the
`corresponding systems.
`~
`~
`V
`-
`-
`'
`,
`‘
`'
`‘
`lIn numerous publications, basically, it has been found that these systems largely
`meet the reci ‘ irements of the users and the specifications given by the manufacturers
`and are comparable in quality to the traditional transmission photometric methods,
`although it must be added that certain technical requirements must still be met by the
`operators, even though these requirements relate more to correct sampling than to the .
`’ actual performance and evaluation of the measurements.
`-
`
`~
`
`.
`.
`Seralyzer® system (Ames)
`The Seralyzer® system developed by the Ames company uses paper'strips on a ~
`supporting foil as reagent carrier. The determination is started by applying 30 pl of .
`diluted serum onto the reagent carrier‘and inserting the carrier into the Seralyzer® ’
`reflectance photometer on a reagent table, the temperature of the measurement
`chamber of the photometer being’ kept at 87"C. The reaction patch is irradiated with
`polychromatic light from a xenon flash-lamp. The diffusely reflected light is collected 1
`in a measurement system with spherical geometry, and after it has passed through an.
`interferencefilter it is measured and compared with a reference. The system is
`controlled by a microprocessor. By choosing suitable flash frequencies, endpoint
`measurements as well as kinetic determinations can be carried out. Each test requires
`a special plug-in module which contains the interference filter ‘for selecting the ,
`measurement wavelength and a memory containing the test-specific data.
`‘
`The instrument is calibrated by means of a two—point measurement 'with ‘low’ and
`‘high’calibrator solutions. The reflectance values are converted to'concentration~
`values via the Kubelka—Munk function since for all purposes the reaction layers can
`beregarded as infinitelythick.
`_
`‘
`,
`.
`The reaction times ‘for the individual parameters range from 30 seconds .to 4‘
`minutes, so that on average approximately 20 determinations can be carried out per
`hour.
`-
`‘
`.
`.
`~
`.
`’
`'
`.
`-
`Up't'o now, the following clinical parameters have been presented for the Seralyzer
`system: glucose, urea, uric acid, cholesterol, LDH, bilirubin, CK, creatinine, AST,
`‘ ALT, and triglycerides. Further parameters are being worked on intensively.
`‘.
`Sexieral publications describing the principles and applications of this system have
`appeared in the literature (Greyson, 1981;. Zipp, 1981).
`.
`Of the published results of clinical trials. currently available, (Karmen 8: Lent,
`1982; Burger et al, 1982; Thomas et al, 1982; Clarke 8: Broughton, 1983; Walter,
`1983),.those obtained by Thomas et a1 (1982) for. five of the available parameters will
`‘ be mentioned here (Tables 12.1, 12.2). The practicability ofthe system is Confirmed-if
`the relatively high coefficients of variation for bilirubin, extreme:concentrations of
`V urea-nitrogen and low concentrations of uric acldare taken into account.
`. With respect to inaccuracy the method comparison yielded correlation coefficients
`'. of equal to or better than-0.97 with the‘exception of urea nitrogen.
`‘
`'
`‘- .The first results with an. immunological test for theophyllin~(Rupchocl< e