CLIN. CHEM. 39/9, 1899-1903 (1993)

CUNICAL CHEMISTRY, Vol. 39, No. 9, 1993 1899

Multianalyte Assay System Developed for Drugs of AbuseRobert G. Parsons,1 Robert Kowal, David LeBlond, Vincent T. Yue, Lawrence Neargarder, Laura Bond,Deyanira Garcia, Debra Slater, and Patrick Rogers

A simple, 10-mm immunoassay system has been devel-oped that simultaneously screens for five different classesof drugs of abuse in a urine sample. This system tests foramphetamines, cannabinoids, cocaine metabolites, opi-ates, and phencyclidmne, and each assay has a specificpreset cutoffconcentration. Accuracy is >99% for report-ing positive or negative results for samples with 200% or50%, respectively, of the cutoff concentrations of thedrugs. Tests of a panel of 96 compounds yielded onlythree cases of nonspecific reactMty (at a drug concentra-tion of 100 mg/L). Another panel of 12 compounds thatcould normally be found in urine samples was also eval-uated and no interferences were observed. Concordancewas >95% between this system and the comparableautomated immunoassays for detecting drugs of abuse.Greater than 98% of GC/MS-conflrmed positive samplesgave positive resultswith this assay system.

IndexingTerms: screening urine agglutinationinhibitionimmunoassay fluoresceinlabel

Many different immunoassay systems have been de-veloped to test for drugs of abuse in urine samples (e.g.,Abbott TDx”/ADf”, Syva EMITTh, Roche Abu-screentm). These systems fill a valuable role in the fieldof drug testing, allowing the performance of accurate,specific screening of urine samples in laboratory set-tings. Drug testing generally requires screening foreach of several different drugs in each sample. This hastraditionally made the process of drug testing quitelaborious, given the need to analyze every sample sep-arately by each required drug assay system with itsrespective calibrators or controls. Consequently, drugtesting has been most commonly performed in central-ized laboratories, where sufficient numbers of samplescould be obtained to perform these assays in large“batch” runs. In some instances, for medical or safety

reasons, there is a need to determine very quicklywhether urine samples do or do not contain drugs.These applications require testing systems that can beperformed on-site.

Several systems have been developed and marketedfor use as on-site drug screening systems (1-5). Some ofthese products can be used to screen for only one or twodrugs at a time, so that, to screen for all five drugsrecommended for screening by the National Institute onDrug Abuse (NIDA), the user must set up and run all

Diagnostics Division, Abbott Laboratories, North Chicago, IL60064.

1Mdress correspondence to this author at the above address,

Department 9ML, Bldg. Ri. Fax 708-937-1214.Received December 1,1992;accepted March 19,1993.

five individual assays.2 Other products offer several dif-ferent assays on one device; however, the accuracy andspecificity of one of these, a rapid screening test formultiple drugs with paper chromatography, has beenquestioned by several studies (1, 6, 7).

We have recently developed the Advisortm drugscreening system, a rapid, simple test system that si-multaneously screens for five different classes of abuseddrugs. This system combines the sensitivity and speci-ficity of immunoassay systems with a unique disposableassay vessel to permit multianalyte, on-site drugscreening.

Principle

The Advisor drug screening system is a multianalyteagglutination assay. Each assay utilizes three majorreagent components: (a) anti-fluorescein antibody-coated particles [blue-stained Duracytestm (fixed humanerythrocytes; Abbott Laboratories, N. Chicago, IL) (8,9)1, (b) anti-drug antibody, and (c) a drug-fluoresceinconjugate. As shown in Figure IA, these three compo-nents bind together to produce agglutinates in the ab-sence of any free drug analyte (negative sample). Freedrug added to these three components (Figure 1B) com-petitively inhibits the cross-linking of the partides andtherefore blocks agglutination.

The Advisor disposable component is a multicham-bored vessel that automatically distributes the liquidreagent (containing anti-fluorescein antibody-coatedparticles and anti-drug antibodies) into distinct assaychannels, each of which contains different dried re-agents (drug conjugates). In this manner each channelbecomes a specific reaction site for the drugs related tothe drug conjugate in that channel.

Materials and MethodsDisposableSystem

The Advisor reaction disc has been designed as aself-performing unit, which allows multiple agglutina-

tion assays to occur within a laminated test card (Figure2). A central well serves as the entrance site for themixture of test sample and reagents. Liquid flowing intothis well is diverted into seven radial channels thathave absorbent paper strips located at their outermostsections. These paper strips serve as blotters, maintain-ing a very slow flow of liquid through the channels overthe course of several minutes. The size (shape and thick-ness) of these paper strips has been optimized to control

2Nonstandard abbreviati . Nll)A, National Institute on Drug

Abuse; PCP, phencydidine; and GC/MS, gas chromatography/mass spectrometry.

A

FU

4,

2

D

By

RS -LGATE

D

DRUG

FIg. 1. Schematic presentationsof the reactionsbetween reagentsfrom the Advisor system: (A) negative reactionin the absence ofdrug above the cutoffconcentrations(agglutination);(B) positivereactionwith drug present(noagglutination)

REACT1ONCHANNa.S

Fig. 2. Advisorlaminatedtest cardBSA, bovine serum aibumin

#{149}S.WPI.E LOADINGZONE

1900 CLINICAL CHEMISTRY, Vol. 39, No. 9, 1993

the fluid flow through the reaction channel such thatstrong agglutination reactions are produced in <10 mm.After the paper strips become saturated with liquid, thefluid flow through the channels stops, and the reactionsare automatically halted and fixed. This eliminates theneed to read the assay results at a precise time.

Each channel has a unique reagent dried within itthat enables specific reactions to occur simultaneouslywithin each of the channels. The card contains fluores-cein-labeled analogs of amphetamines, cannabinoids,benzoylecgonine, morphine, and phencydidine (PCP) infive of the channels to provide for their respective as-says. An additional channel contains fluorescein-labeledbovine serum albumin as a real-time negative reaction

control. Another channel contains no dried reagents (areal-time positive reaction control).

The laminated test cards are housed in a molded plas-tic housing-, this keeps the cards flat and provides amixing well for sample and reagent.

Reagent SystemThe Advisor reagent contains blue-stained Duracytes

coated with anti-fluorescein antibody and a cocktail ofantisera to each of the drugs (amphetamines, cannab-inoids, cocaine metabolites, opiates, and PCP). Theamount of each antiserum is precisely set to produce thedesired cutoff between positive and negative reactionsfor that analyte.

The Duracytes were stained with Remazol BrilliantBlue R (Sigma Chemical Co., St. Louis, MO) at a finalconcentration of 6 giL. The stained Duracytes werecoated with affinity-purified rabbit anti-fluorescein an-tibody (75 mgfL) in the presence of chromic chloride (0.5g/L) in 0.1 mol/L sodium acetate buffer, pH 4.0.

The final antibody cocktail contains a 16 mLIL sus-pension of coated Duracytes plus all five antisera in aphosphate buffer (0.073 mol/L, pH 7.3) containing 0.5molIL NaCl, 15 milL fetal calf serum, 0.008 mollLEDTA, 0.08 g/L povidone, and 0.19 g/L gelatin. Theamphetamine antibody is a mouse monoclonal antibody;the others are all polyclonal sheep antisera.

Assay Procedure

To run an Advisor assay, place three drops (-200 pL)of the antibody cocktail into the mixing well in themolded plastic housing of the reaction disc. Add sample(20 L) to the antibody cocktail in the mixing well, mixthe combined solution of the sample plus antibody cock-tail, and transfer the mixture to the center of the lam-inated test card in the reaction disc. After 10 mm, com-pare the reaction patterns in each assay channel withthe reactions in the positive and negative control chan-nels. Negative reactions (no drug) have agglutinatedparticles, and positive reactions (drug present at a con-centration greater than the cutoff) have a uniform pat-tern of unaggregated partides.

Drug Standards and SamplesDrug standards for reproducibility testing were pre-

pared by dissolving drug analytes at specified concen-trations in drug-free human urine containing dimethylsulfoxide, 100 milL. Drug concentrations were verifiedby TDx analyses (Abbott Laboratories) with the respec-tive TDx reagent systems (Amphetamine/Methamphet-amine II, Cannabinoids, Cocaine Metabolite, Opiates,and Phencycidine II). The standards used for the fiveAdvisor assays were D-amphetamine, 11-nor-delta-9-

tetrahydrocannabinol-9-carboxylic acid, benzoylecgo-nine, morphine, and PCP.

Urine samples were collected and frozen at -20 #{176}C.Before analysis all samples were completely thawed andmixed. TDx analyses of the urine samples were per-formed with the same reagent kits mentioned above.Gas chromatography/mass spectrometry (GCIMS) anal-yses were performed at a commercial reference labora-tory (Aegis Analytical Laboratories Inc., Nashville,

Table 1. Results of Reproducibility Studies with theAdvisor Test System

Dg (and cuteflvalue, paJL)

Amphetamines

(300)

Cannabinolds

(50)

Cons,’150450600

2575

100100

300400150450600

12.537.550

No.poe.

467497

3497500

2495499

0500500

0498500

0498

5

50000

50020

10099.699.099.8

100100100100

99.6100

FIg. 3. Advisorreactions with multiconstituent controls having all fivedrugsat (A) 0.5 timesthecutoff value; (B) 2.0 timesthe cutoff valueAssay channels (starting with the uppermostchanneland proceeding clock-wis: negative control, amphetamines (AMP), cannabinolds (CAN), cocainemetaboifte (COC), opiates (OPI), phencyclidine (PCP), and positive control

CUNICAL CHEMISTRY,Vol. 39, No. 9, 1993 1901

TN). Cutoff values for these confirmatory assays were at300 gfL amphetamine for the amphetamine assay, 10p.g/L carboxytetrahydrocannabinol for the cannabinoidsassay, 100 p.gfL benzoylecgonine for the cocaine assay,300 jtg/L morphine and (or) codeine for the opiates as-say, and 25 g’L PCP for the PCP assay.

Results

Precision near the assay cutoffs. Agglutination reac-tions in channels of the Advisor reaction card producecharacteristic banded patterns (Figure 3), which caneasily be discriminated from channels without aggluti-nation. Because this technology uses a competitive as-

say format, negative assays (no drug present) have anagglutinated reaction product, whereas positive assays(drug present above the cutoff) show no agglutination.Each assay has a unique preset cutoff concentration,above which positive results will be obtained. This im-munoassay configuration provides a very steep transi-tion between positive and negative reactions. As seen inFigure 3, very strong agglutination reactions will occurfor each of the drug channels when sample containingone-half the cutoff concentration of drug is used. Con-versely, samples containing drug at twice the cutoffvalues yield smooth, nonagglutinated assay results. Ta-ble 1 shows the results obtained from 1500 tests withdrug standards. Accuracy was >99% for samples con-taining drug concentrations representing one-half thecutoff value for each assay (lowest concentration onchart for each assay). Also, positive results were ob-tained with >99% accuracy for samples containing drugat twice the cutoff with all five drug assays.

Cross-reactivity. A large panel of drugs or other po-tentially interfering compounds (Table 2) was evaluatedfor cross-reactivity with the Advisor assays. Samples ofeach drug were prepared in drug-free human urine atconcentrations of 100 mgIL or greater and aliquots wereassayed to determine undesirable cross-reactivity. Allbut three compounds showed no reactivity with any ofthe assays at the concentrations tested. Phenethyl-amine at 100 mg/L produced a positive reaction for theamphetamine assay, but no cross-reactivity was seen at

Cocaine metabolite(200)

Opiates(300)

Phencyclidine(25)

No.fig. S correct

499 99.8

33 93.43 99.4

497 99.43 99.4

a Concentrationsrepresent0.5, 1.5, and 2.0 tImes the cutoff values, respec-

tively, for each drug.

10 mg’L. Both meperidine and trihexylphenidyl pro-duced positive PCP reactions at 100 mg/L. Cross-reac-tivity with these compounds was also absent at 10 mgfLconcentrations.

Interferences. Several compounds that ordinarilycould be present in a urine specimen were evaluated fornonspecific interference in the Advisor assays. Com-pounds were dissolved in normal human urine at theconcentrations listed in Table 3. We then added to theseurine samples enough concentrated drug solution toyield final drug concentrations of 0.5 or 1.5 times thecutoff value and tested the samples with the Advisorassay system. No interferences were observed:i.e., sam-ples with 0.5 times the drug cutoff values gave negativeresults, and samples with 1.5 times the drug cutoff val-ues gave positive results.

Concordance of assay results with TDx and GC/MS. Apanel of 859 urine samples was analyzed by both Advi-sor and TDx systems. This urine sample population wasenriched with known positive samples for all five of thedrug classes tested by the Advisor system, to ensurethat a significant number of positives would be ana-lyzed. Cutoff values for the TDx assays were set to be thesame as the preset cutoff values for Advisor assays. Allsamples with positive results in either assay systemwere sent to a commercial reference laboratory forCC/MS confirmation. As the results show (Table 4), theAdvisor and the TDx assays have >99% concordance forall assays except amphetamine (95.6% concordance).3The Advisor amphetamine assay is specific for amphet-amine and it does not directly detect methamphet-amine, whereas the TDx assay detects both analytes. Ofthe 37 samples that were Advisor negativefTDx posi-tive, 36 were negative for amphetamines (<300 g/L) by

3Concordance (%) = [no. of Advisor-positive and TDx-positive(greater than or equal to cutoff value) samples + no. of Advisor-negative and TDx-negative (less than cutoff value)samplesj/totalno. of samples.

Table 2. Compounds Tested for Cross-Reactivity withthe Advisor Assaysa

Table 4. Concordance between Advisor and TDx InTestIng 859 Urine Samples

Advlsor+ Advisor-

TDX+ TDx- TDx+ TDx-

74 1 37 747

116 6 1 733

Drug

AmphetamineCannabinoidsCocaine

metabolitesOpiatesPCP

Concordance,S

95.6

99.2

105 1 2 750 99.794 8 0 756 99.196 4 2 753 99.3

AcetaminophenAcetylsallcycllc acidAiphaprodineAminopyrineArnitriptytineAmobarbitalAnipicillinArtileridineApomorphineAtropineBarbfturlc acidBenzocaineButabarbltalButorphanolCaffeine (1 g/L)Calcium hypochloriteChiordiazepoxideChloroqulneChlorpheniramlneChlorpromazlneClemastineClomipramineCyclizineCyclobenzaprlneDesipramineDextromethorphanDiazepamDibenzepin

DiphenhydramineDiphenyloxylateDoxepinDoxylamineEphedrlne (3 g/L)

Eplnephrine(lg/L)

ErythromycinEstilolFenfluramineFurosemideGentisic acidGlutethimideHydrochlorothiazideIbuprofen

(0.5 giL)ImipramineIsometheptenelsoxsuprineKetamine(1 g/L)LabetalolLevallorphanUdocalneLSDMeperidineMethadoneMethapyrileneMethaqualoneMethotilmeprazineMethoxyphenamineMethylphenidate

MethyprylonNaloxoneNaproxenNiacinamideNorethindroneNorpropoxypheneNortilptyllne

NylidrinOphenadrineOpipramolOxazepamPenicillin G (1 g/L)PhenethylarninePheniraminePhenmetrazinePhenobarbltal

PhenothiazinePhentazocinePhenylbutazonePhenyipropanolamine

(lg/L)ProcaIne (1 g/L)Promazine

PromethazineProproxyphenePropylhexadnneQuinineSecobarbitalSulindacTerfenadineTetracyclineTetrahydrozolineThioridazineTrlfluoperazineTnhexyphenidylTrlmethobenzamideTrimipramineTripelennamineVerapamilZomepirac

termined by GC/MS). Seventy-one of the 93 metham-phet.amine samples were positive for amphetamines inthe Advisor system. Figure 4 shows the concordancebetween the concentrations of amphetamine ormethamphetamine and Advisor assay reactivity. Of the22 samples for which the Advisor assays did not detectamphetamine, only 4 contained methamphetamine atconcentrations � 1 mg/L.

Table 5 displays the results of Advisor assays forGC/MS-eonflrmed positive samples. Of the GC/MS-pos-itive samples, >98% were positive by the respectiveAdvisor drug assays.

Discussion

The Advisor system provides a simple, rapid systemfor screening for five major classes of drugs of abuse.

‘Compounds were tested at 100 mcJLexceptwhere noted.LSD, lysergic acid diethylamide.

9.4

0

Table 3. Compounds Tested for Interferencewith theAdvisor Assays

0 0

eco

40010

-.

-

SS

CompoundAcetoneAscorbic acidBilirubinCreatinineEthanolGlucoseSodiumchlorideOxalic acidProteinRiboflavinLysod erythrocytesUrea

-j

‘I,

12O0

x0

0 400 eoo 1200 15)0 2000

AMPHETAMINES (j.ig/L)#{149}ADVISOR#EGAIM 0 ADViSORPO)0TTVE

Fig. 4. Concordance between amphetamine and methamphetamineconcentrations and Advisor reactivity in methamphetamine-contaln-ing clinical samples (n = 93)Quantitativeamphetamineand methamphetamineconcentrationswere deter-mined by GC/MS

Cons tested, o/L1015

0.015.0

102060

1.00.500.075

1.15 (hemoglobin)

63

CC/MS. CC/MS analysis of the total population of dis-cordant samples for the cannabinoids, cocaine, opiates,and PCP assays revealed that Advisor was correct for50% of the samples (12/24) and TDx was correct for theother 50% (12/24).

Ninety-three of the urine samples in this study con-tained detectableamounts of methamphetamine (as de-

Table 5. Detection of GC/MS-Conflrmabie PositiveSamples by Advisor

No.of positivesamplesDrug

AmphetamineCannabinoidsCocainemetabolitesOpiatesPCP

71

110

10597

92

S detectedby Advisor

98.6

98.299.0

100100

1902 CUNICAL CHEMISTRY,Vol.39, No.9, 1993

CUNICAL CHEMISTRY, Vol. 39, No. 9, 1993 1903

Although this system is a manual qualitative test, itcompares very well (>95% concordance with clinicalspecimens) with the quantitative assays on the TDxinstrument system. The reagent configurationused isa

modified form of hemagglutination inhibition assays.Hemagglutination assays specific for morphine werefirst described in 1971 (10, 11). These assays, whichutilized classical hemagglutination approaches with mi-crotiter plates, took 2 h to complete. Vanzetti et al. (12)demonstrated that the hemagglutination inhibitiontechnology provided a sensitivity similar to that of ra-dioimmunoassay. The Advisor disposable system per-mits hemagglutination inhibition assays to be run inmuch shorter times (10 miii) but maintaining low-con-centration sensitivities for PCP (25 g/L) and cannab-molds (50 g’L). The cutoff values for the amphetamine,cannabinoids, and cocaine assays (300, 50, and 200p.g/L) are all less than the current NIDA-recommendedvalues. Several reports have suggested the need toscreen urine samples with assays having lower cutoffvalues than the NIDA ones (13-16). Results of the 1990NIDA Consensus Conference also suggested that thescreening cutoff values for these assays be lowered (17).

The Advisor amphetamine assay is specific for am-phetamine; it does not detect methaunphetamine di-rectly. In our study, however, most of the methamphet-amine-containing samples produced positive reactionsfor the amphetamine assay. Because amphetamine is aprimary metabolite of methamphetamine, this systemdetects methamphetamine samples by reacting with theamphetamine that is metabolically generated. The Ad-visor amphetamine assay has a cutoff value of 300 zg/L,lower than the NIDA-recommended screening cutoffvalue of 1 mg/L. Because of this cutoff for the amphet-amine assay, 94% of the samples with 1 mgfL or more ofmethamphetamine (NIDA-recommended screening cut-off) were detected as amphetamine positive. Thus theAdvisor system can be used to detect the abuse of bothamphetamine and methamphetamine. Although itmight be feasible to add an additional assay specific formethamphetamine to this screening panel, we question

how beneficial this would be because the increase in thepercentage of positive samples detected would be verysmall.

The Advisor system provides an excellent tool for usein on-site drug testing. The test system is quick andsimple, and is not dependent on a laboratory environ-

ment or sophisticated laboratory equipment. Results ofthese studies support the use of the Advisor system as aprimary screening tool for drug testing programs. Sam-ples that produce positive results in the Advisor systemshould be confirmed by CC/MS or other confirmatoryprocedures, just the same as with other types of screen-ing tests.

References1. Bogema 5, Schwartz R, Godwm I. Evaluation of the KeystoneDiagnostics Quik Test”' using previously screened urine speci-mens. J Anal Toxicol 1988;12:272-3.2. Schwartz RH, Bogema S, Thorne MM. Evaluation of the EZ-SCREEN” enzyme immunoassay test for detection of cocaine andmarijuana metabolites in urine specimens. Pediatr Emerg Care1990;6:147-9.3. Schwartz RH, Bogema S, Thorne MM. Evaluation of a rapidlatex-particle alutination-inhibifion screening assay for cocainein urine. J Pediatr 1990;117:670-2.4. Armbruster DA, Krolak JM. Screening for drugsof abusewiththe Roche ONTRAK Assays. J Anal Toxicol 1992;16:172-5.5. Buechler KF, Moi 5, Noar B, McGrath D, Villela J, Clancy M,et al. Simultaneous detection of seven drugs of abuseby Triage’”panel for drugs of abuse. Clin Chem 1992;38:1678-84.6. Cone EJ, Menchen SL. Lack of validity of the KDI Quik Testdrug screen for detection of benzoylecgonine in urine. J AnalToxicol 1987;11:276-7.7. Wells H, Cone EJ. A discussion of the KDI Quick Test’” drugscreen. J Anal Toxicol 1988;12:111.8. Hirata AA, Boley WF. Hepatitis B virus antigen detection byreverse passive hemagglutination. Proc Soc Exp Biol Med 1973;143:761-3.9. Hirata AA, Brandriss MW. Passive hemagglutination proce-dures for protein and polysaccharide antigens using erythrocytesstabilized with aldehydes. J Ixnmunol 1968;100:641-6.10. MIer FL, Liu CT. Detection of morphine by hemagglutina-tion-inhibition. J Immunol 1971;106:1684-.5.11. Mler FL, Liu CT, Cathn DH. Immunological studies of heroinaddiction. I. Methodolo and application of a hemagglutination-inhibition test for detecting morphine. Clin Immunol Immunopa-thol 1972;1:53-68.12. Vanzetti G, Cassani M, Valente D. Detection of morphine inurine by hemagglutination inhibition, with use of lyophilizedreagents. Clin Chem 1983;29:1376-9.13. Smith DE, Gutgesell ME, Schwartz RH, Thorne MM, BogemaS. Federal guidelines for marijuana screening shouldhave lowercutoff levels. Arch PatholLabMed 1989;113:1299-300.14. Schwartz RH, Wifiette RE, Hayden GF, Bogema S, ThorneMM, Hicks J. Urinary cannabinoids in monitoring abstinence in adrug abuse treatment program. Arch PatholLab Med 1987;111:708-11.15. Mule SJ, Lomax P, Gross SJ. Active and realistic passivemarijuana exposure tested by three immunoassays and GC/MS inurine. J Anal Toxicol 1988;12:113-6.16. Hicks JM, Morales A, Soldin SJ. Drugs of abuse in a pediatricoutpatient population [Letter]. Clin Chem 1990;36:1256-7.17. Finkle BS, Blanks RV, Walsh JM, eds. Technical, scientificand procedural issues ofemployee drug testing, Consensus Report.DHHS PubL No. (ADM)90-1684. Washington, DC: US Govern-ment Printing Office, 1990.