Scientific review

Comparison of Three Latex Agglutination Kits and a Commercial EIA for the Detection ofCryptococcal AntigenJ.E. Bestrom, D.J. Jespersen, and N.L. WengenackMayo Clinic, Rochester, MN 55905

Evaluation of the Clostridium difficile Immunocard® Toxin A and B TestK. J. Brown, D. J. Jespersen, M. F. Jones, S. K. Schneider, T.F. Smith, Ph.D. and J.E. Rosenblatt, M.D.Mayo Clinic, Rochester, MN 55905

Comparison of Focus Technologies, Inc. and Panbio, Inc.Elisa Assays for Detection of Antibodies to West Nile VirusDJ Jespersen, MF Jones, EM Beito, TF SmithDivision of Clinical Microbiology, Mayo Clinic College of Medicine, Rochester, MN 55905

Comparative Evaluation of the Triturus® Analyzer and the Eti-Max® 3000 for Performance ofEnzyme Immunoassays on Viral Hepatitis MarkersD. Granger, D. Jespersen, T. Rasmussen, P.S. Mitchell, J.D.C. YaoMayo Clinic, Rochester, MN 55905

An Evaluation Of Three Serologic Commercial EIA Kits for The LaboratoryDiagnosis of Mycoplasma Pneumoniae InfectionsD. J. Jespersen, J. A. Harring, M. F. Jones, T.F. Smith, Ph.D.Mayo Clinic, Rochester, MN 55905

Validity of Enzyme Immunoassay (Eia) Serologic Results for The Laboratory Diagnosis ofEpstein-barr Virus (Ebv) Infections.Julie A. Helgason, Deborah J. Jespersen, Mary F. Jones, Elaine M. Beito, and Thomas F. Smith.Mayo Clinic and Foundation, Rochester, MN 55905

Scientific review

J.E. Bestrom, D.J. Jespersen, and N.L. Wengenack Mayo Clinic, Rochester, MN

AbstractLatex agglutination tests have been used for >40 years for the rapid, sensitive,and specific identification of cryptococcal capsular antigen. However no comparisonof latex agglutination kits has been reported in the last 10 years.We thereforecompared the performance of the Meridian CALAS®,Wampole Crypto-LA®, andMurex Cryptococcus Test kits for the detection of cryptococcal antigen from serumand CSF. In addition, the Meridian EIA for cryptococcal antigen was adapted tothe Triturus autoanalyzer and compared to the Meridian latex kit.Fifty archivedantigen-positive and 50 archived antigen negative sera and CSF specimens wereexamined using the 3 latex kits and the EIA. The manufacturer’s instructions werefollowed for each kit, including the use of pronase to remove potential nonspecificinterference with the Meridian and Murex kits. The percent concordance of the100 specimens for each pair of kits was as follows: Meridian and Wampole latex= 96%,Meridian and Murex latex = 93%,Wampole andMurex latex = 93%,Meridian latex and Meridian EIA =98%. The endpoint titers were also compared betweenthe 3 latex kits. Agreement between the titers was definedas ±1 two-fold dilution and was found to occur 60% and48% of the time respectively for the Wampole and Murexkits relative to the Meridian kit. Endpoint titer agreementbetween the Wampole and Murex kits was 48%. Therewere no false positive results among the 50 negativespecimens tested by any of the latex kits or the EIA.Weconclude that the three latex kits as available today havesimilar performance characteristics based on concordancevalues ≥ 93%. The Meridian latex kit endpoint was feltto be the easiest to read. The endpoint titers from thevarious kits should not be compared when followingindividual patients since the absolute titer value variedwidely between the kits. The Meridian EIA performed aswell as the latex kit and has the added advantage ofpermitting automated specimen processing in high-volumelaboratories.

BackgroundC. neoformans is an opportunistic fungal pathogenthat causes systemic disease in immunocompromisedpersons (eg., those with AIDS, diabetes, malignancies).Detection of C. neoformans antigen in serum or CSFis often more sensitive and rapid than culture of theorganism. Latex agglutination tests have been widelyused for the detection of cryptococcal antigen for thelast 40 years but no comparison of commercials kitshas been reported for almost a decade 1-2. The objectof this study was to compare the performance of threecommercially-available latex agglutination kits and acommercial EIA kit for the detection of cryptococcalantigen.

Materials And MethodsLatex agglutination kits evaluated

• CALAS® Cryptococcal Antigen Latex Agglutination System (Meridian Bioscience,Inc., Cincinnati, OH)

• Murex Cryptooccus Test (Remel, Lenexa, KS)

• Crypto-LA® Test (Wampole Laboratories, Cranbury, NJ) EIA kit evaluated

• Meridian Premier® Cryptococcal Antigen

A side-by-side comparison of the 3 latex agglutination kits and the EIA kit wasconducted using 50 archived, previously positive and 50 archived, previouslynegative serum and CSF specimens. The archived specimens had been testedinitially using the Meridian CALAS latex kit prior to archiving. All latex agglutinationtests were performed according to the manufacturer’s package insert and theendpoint was read by two experienced technologists. Any endpoint titer, including1:1, was considered positive for cryptococcal antigen. The EIA kit testing wasperformed according to the manufacturer’s instructions using the automatedTriturus® EIA Analyzer (Grifols-Quest, Miami, FL).

Comparison of Three Latex Agglutination Kits and a Commercial EIA for theDetection of Cryptococcal Antigen

Triturus

Conclusions• No false positive results were found among the 50 negative specimens.

• The three C. neoformans latex agglutination kits as available today have similar performance characteristics based on concordances of ≥ s93%.

• The Meridian latex kit endpoint was felt to be easiest to read by the technologists performing the testing.

• The endpoint titers from the latex kits should not be compared when followingindividual patients since the absolute titer varies widely between the kits.

• The Meridian Premier® EIA kit performed as well as the latex kits and has the added advantage of permitting automated specimen processing using the Triturus analyzer for highvolume laboratories.

References1. Tanner TC, Meinstein MP, Fedorciw B, Joho KL,Thorpe JJ, Reller LB. 1994.

Comparison of Commercial Kits for Detection of Cryptococcal Antigen. J ClinMicrobiol. 32:1680-4.

2. Jaye DL,Waites KB, Parker B, Bragg SL, Moser SA. 1998. Comparison of two rapid latex agglutination tests for detection of cryptococcalcapsular polysaccharide. 109:634-41.

1 1:128 1:16 1:256 POS 51 NEG NEG NEG NEG2 1:256 1:8 1:8 POS 52 NEG NEG NEG NEG3 1:2048 1:1024 1:2048 POS 53 NEG NEG NEG NEG4 1:1024 1:256 1:256 POS 54 NEG NEG NEG NEG5 1:2048 1:256 1:256 POS 55 NEG NEG NEG NEG6 >1:8192 >1:8192 1:512 POS 56 NEG NEG NEG NEG7 1:128 1:32 1:128 POS 57 NEG NEG NEG NEG8 1:512 1:512 1:1024 POS 58 NEG NEG NEG NEG9 1:8 1:2 1:8 POS 59 NEG NEG NEG NEG

10 1:16 1:2 1:4 POS 60 NEG NEG NEG NEG11 >1:8192 >1:8192 >1:8192 POS 61 NEG NEG NEG NEG12 1:16 1:8 1:4 POS 62 NEG NEG NEG NEG13 1:512 1:128 1:512 POS 63 NEG NEG NEG NEG14 1:1024 1:256 1:256 POS 64 NEG NEG NEG NEG15 1:2 NEG 1:2 POS 65 NEG NEG NEG NEG16 1:1024 1:512 1:32 POS 66 NEG NEG NEG NEG17 1:2 1:1 NEG POS 67 NEG NEG NEG NEG18 1:2048 1:2048 1:2048 POS 68 NEG NEG NEG NEG19 1:16 1:2 NEG POS 69 NEG NEG NEG NEG20 1:128 1:8 1:256 POS 70 NEG NEG NEG NEG21 NEG NEG 1:256 POS 71 NEG NEG NEG NEG22 1:2048 1:512 1:1024 POS 72 NEG NEG NEG NEG23 1:4 NEG NEG NEG 73 NEG NEG NEG NEG24 1:1024 1:512 1:2048 POS 74 NEG NEG NEG NEG25 1:1024 1:512 1:2048 POS 75 NEG NEG NEG NEG26 1:16 1:4 NEG GZ 76 NEG NEG NEG NEG27 1:64 1:32 1:128 POS 77 NEG NEG NEG NEG28 1:512 1:512 1:1024 POS 78 NEG NEG NEG NEG29 1:512 1:128 1:32 POS 79 NEG NEG NEG NEG30 1:32 1:16 1:8 POS 80 NEG NEG NEG NEG31 1:8192 1:4096 1:2048 POS 81 NEG NEG NEG NEG32 1:16 1:16 1:16 POS 82 NEG NEG NEG NEG33 1:128 1:64 1:64 POS 83 NEG NEG NEG NEG34 >1:8192 >1:8192 >1:8192 POS 84 NEG NEG NEG NEG35 1:512 1:256 1:2048 POS 85 NEG NEG NEG NEG36 >1:8192 >1:8192 1:4096 POS 86 NEG NEG NEG NEG37 1:128 1:128 1:2048 POS 87 NEG NEG NEG NEG38 1:1024 1:1024 1:4096 POS 88 NEG NEG NEG NEG39 1:8 1:4 NEG POS 89 NEG NEG NEG NEG40 1:512 1:512 1:2048 POS 90 NEG NEG NEG NEG41 1:8 1:4 1:8 POS 91 NEG NEG NEG NEG42 1:16 NEG 1:32 POS 92 NEG NEG NEG NEG43 NEG NEG NEG NEG 93 NEG NEG NEG NEG44 1:512 1:256 1:1024 POS 94 NEG NEG NEG NEG45 1:1024 1:512 1:8192 POS 95 NEG NEG NEG NEG46 1:1024 1:4096 1:4096 POS 96 NEG NEG NEG NEG47 1:16 NEG NEG POS 97 NEG NEG NEG NEG48 1:512 1:512 1:2048 POS 98 NEG NEG NEG NEG49 1:1024 1:2048 1:4096 POS 99 NEG NEG NEG NEG50 1:4096 1:1024 1:8192 POS 100 NEG NEG NEG NEG

Table 1.C. neoformans Latex Agglutination Endpoint Titers and EIA results

Concordance of Endpoint Titers for Three Commercial C. neoformans Latex Agglutination Kits (Meridian, Wampole,

and Murex) and the Meridian Premier®EIA Test

MeridianLatexTiter

WampoleLatexTiter

SampleNumber

Murex LatexTiter

MeridianEIA

ResultSamplenumber

MeridianLatexTiter

Wampole LatexTiter

MurexLatexTiter

MeridianEIA

Result

Meridian Latex

Wampole Latex Positive NegativePositive 44 0Negative 4 52

Meridian Latex

Murex Latex Positive NegativePositive 42 1Negative 6 51

Wampole Latex

Murex Latex Positive NegativePositive 40 3Negative 4 53

Meridian Latex

Meridian EIA Positive NegativePositive 46 1Negative 1 51

% concordance = 98%n=99 because 1 EIA result was "equivocal" and was not repeated

In order to determine concordance for the latex agglutinationtests, any titer including undiluted (1:1) was considered positive

Table 3.

Enpoint Titer Agreement for the Latex Agglutination Kits

Kits Endpoint Agreement(±1 two-fold dilution)

Wampole and Meridian 60%Murex and Meridian 48%Wampole and Murex 48%

Table 2.

Results

Evaluation of the Clostridium difficile Immunocard® Toxin A and B Test

K. J. Brown, D. J. Jespersen, M. F. Jones, S. K. Schneider, T.F. Smith, Ph.D. and J.E. Rosenblatt, M.D.Mayo Clinic, Rochester, MN

Sensitivity = 100%Sensitivity = 98%

Premier® Toxin A and B+ -

Immunocard® + 59 3

Toxin A and B - 0 148

Sensitivity = 91%Sensitivity = 100%

Premier® and Immunocard+ -

Clearview® + 57 0

C. Diff A - 5 148

AbstractWe compared the Immunocard® Toxin A and B assay with thePremier® Toxin A and B assay (both manufactured by MeridianBioscience, Inc, Cincinnati, Ohio) for the detection of C.diff ici letoxins A and B. Addit ionally, al l samples were tested with theClearview® C. Diff A® (Wampole Laboratories, Princeton, NJ) todetermine how many positive samples were likely to contain Toxin Bsince this test detects only Toxin A. 210 consecutive stool samplessubmitted to the laboratory for C.difficile toxin testing were equallyaliquotted, coded, and tested according to the product inserts.Samples were tested by the Premier® kit on the Triturus (GrifolsUSA, Miami, FL) automated EIA analyzer. Both the Immunocard®

and Clearview® assays are horizontal-flow EIAs and were testedmanually. For the detection of Toxin A, discrepant results wereresolved by agreement of two out of three different method results(extended gold standard). The Immunocard® had a sensitivity of100% and a specificity of 98%. There were three additional positivesamples (p=0.25) with the Immunocard® that were confirmed by theClearview® test indicating that the Immunocard® was slightly moresensitive than the Premier® assay for detection of Toxin A. Since theClearview® Assay detects only Toxin A and 5 of 62 (8%) positivesamples were negative only by Clearview®, this suggests that thesesamples contained Toxin B. The Immunocard® and Premier® assaysappear to be comparable for the detection of C.difficile toxins A andB. The Immunocard® can potential ly improve turn around t ime(analytical time of 15 minutes vs. 60 minutes) and allows for on-demandsingle sample testing.

The Premier® Toxin A and B assay was programmed on the Triturus accordingto the product insert .

Immunocard®

ResultsFor the detection of Toxin A, discrepant results were resolved by agreement oftwo out of three different method results (extended gold standard). TheImmunocard® had a sensitivity of 100% and a specificity of 98%. There werethree additional positive samples (p=0.25) with the Immunocard® that wereconfirmed by the Clearview® test indicating that the Immunocard® was slightlymore sensitive than the Premier® assay for detection of Toxin A. Since theClearview® Assay detects only Toxin A and 5 of 62 (8%) positive samples werenegative only by Clearview®, this suggests that these samples contained ToxinB. The Immunocard® and Premier® assays appear to be comparable for thedetection of C. difficile toxins A and B.

Conclusions

• The Immunocard® and Premier® assays appear to be comparable fordetecting C.difficile Toxins A and B.

• The Immunocard® would improve analytical turn around time(15 minutes vs 60 minutes).

• Use of the Immunocard® would allow for single sample, STATtesting, ensuring rapid institution of proper treatment andisolation precautions.

• 8% of the C.difficile positive specimens were not detected usingthe Clearview® C. Diff A kit, suggesting that these samplescontained Toxin B.

TriturusThe Triturus is an open platform, fully automated, multi-test, multibatch analyzer.

ObjectivesToxigenic C.difficile produces two distinct toxins. Toxin A is anenterotoxin and Toxin B is a cytotoxin: These toxins are responsiblefor a variety of diseases including antibiotic associated diarrhea, andpseudomembranous colitis. Our goal in this study was to comparethe Immunocard® Toxin A and B assay with the Premier® Toxin A andB EIA assay. Additionally, we tested all of the samples with theClearview® C. Diff Kit to determine how many of the positive sampleswere likely to contain Toxin B since this assay only detects Toxin A.

Comparison of Focus Technologies, Inc. and Panbio, Inc.Elisa Assays for Detection of Antibodies to West Nile Virus

DJ Jespersen, MF Jones, EM Beito, TF SmithDivision of Clinical Microbiology, Mayo Clinic College of Medicine, Rochester, MN 55905

Abstract

Laboratory diagnosis of West Nile Virus [WNV] is best achieved by demonstrationof specific IgG and IgM class antibodies in serum or CSF specimens; amplifiedtarget nucleic acid of the virus by PCR has been demonstrable only in approximately55% of CSF and 10% of serum specimens from known WNV infected patients.Two test methods for detection of antibodies to WNV were compared using apanel of 100 sera from human cases; 75 specimens were from patients withsigns and symptoms compatible with WNV infection. These antibodies werespecific for this virus based on results of the plaque reduction neutralization test[PRNT] performed by the Centers for Disease Control [CDC] or the CaliforniaState Department of Health. All 75 PRNT positives in the panel had IgM classantibody. FOCUS Technologies, Inc. [Cypress, California] and PanBio,Inc. [Columbia, Maryland] have developed Flavivirus (WNV) ELISA assays fordetecting both IgG and IgM antibodies in human serum. These assays inconjunction with exposure history and local epidemiology have been used asan indication of early WNV infections. Both assays were also formatted andtested on the Triturus (Grifols-Quest), an automated open platform ELISA analyzer.IgM antibody is generally detected by the eighth day of illness while IgG antibodyis usually found within three weeks post infection. Of the 75 PRNT positive sera,IgG antibodies were detected by the FOCUS test in 31 [41%] of thesamples while the PanBio assay detected IgG in 38 [51%] of the samples. Bothassays detected IgG antibodies in the same 31 specimens while 7specimens were positive exclusively with the PanBio test. Not all samples thatcontained IgM antibody were positive for IgG antibody detection. Using bothmanual and automated formats, all sera with IgM antibodies to WNV [n=75]were detected with the FOCUS assay [sensitivity 100%] whereas the PanBioassay failed to detect 4 [5%] of the positive IgM samples [sensitivity 95%]. Thesetwo FDA approved kits were sensitive [95% - 100%] and specific, technicallysimple, and both assays can be formatted on the Triturus with 100% agreementbetween manual and automated testing necessary for testing high volumes.

Objective

Focus Technologies, Inc. and PanBio, Inc. both have FDA approved ELISAS forWNV. Our goal in this study was to compare the IgM results from both kits withsamples that were known to be PRNT positive. Additionally, we wanted to seeif the Focus kit performed as well on an automated system (Triturus) as it didwith the manual method.

Triturus

The Triturus is an open, fully automated EIA analyzer

ResultsFor the detection of Toxin A, discrepant results were resolved by agreement oftwo out of three different method results (extended gold standard). TheImmunocard® had a sensitivity of 100% and a specificity of 98%. There werethree additional positive samples (p=0.25) with the Immunocard® that wereconfirmed by the Clearview® test indicating that the Immunocard® was slightlymore sensitive than the Premier® assay for detection of Toxin A. Since theClearview® Assay detects only Toxin A and 5 of 62 (8%) positive samples werenegative only by Clearview®, this suggests that these samples contained ToxinB. The Immunocard® and Premier® assays appear to be comparable for thedetection of C. difficile toxins A and B.

Conclusions• Both kits provided high sensitivity and specificity (95% -100%)

• Both assays can be formatted on the Triturus

• While both assays are technically simple, the PanBio assay does have shorterincubation steps than the Focus test

Results for IgM Antibodies

Sensitivity = 100%Sensitivity = 100%

FOCUS (manual) + 75 0

- 0 25

CDC + -

Sensitivity = 100%Sensitivity = 100%

FOCUS (Automated Triturus) + 75 0

- 0 25

Sensitivity = 95%Sensitivity = 100%

PanBio (manual) + 71 0

- 4 25

CDC + -

CDC + -

Comparative Evaluation of the Triturus® Analyzer and the Eti-Max® 3000 forPerformance of Enzyme Immunoassays on Viral Hepatitis Markers

D. Granger, D. Jespersen, T. Rasmussen, P.S. Mitchell, J.D.C. Yao Mayo Clinic, Rochester, MN

AbstractVery few automated commercial systems are available for the processing andtesting of clinical specimens for serologic markers of viral hepatitis that are donecommonly in diagnostic laboratories. An accurate and labor-saving process forsuch an operation is essential for an efficient laboratory to produce test resultswith rapid turn-around time. In this study, we compared the performancecharacteristics, specimen throughput, and material costs of two automatedsystems, Triturus Analyzer (Triturus; Grifols-Quest Inc.,Miami, FL) and ETI-MAX3000 (ETI-MAX; DiaSorin Inc., Stillwater, MN), on seven enzyme immunoassays(EIAs) for viral hepatitis markers. For each EIA, 50 known-positive, 50 known-negative, and 100 prospectively collected clinical serum specimens were testedon both the Triturus and the ETI-MAX. Specimens yielding discrepant resultswere retested in duplicate by each system, with the final result of each specimenbeing the best two of three results. Each system was programmed to processand test these specimens according to the assay manufacturer’s instructions.The following EIAs manufactured by DiaSorin Inc. were compared: anti-HAVTotal antibody, anti-HAV IgM antibody, HBsAg, anti-HBs antibody, anti-HBc IgMantibody, HBeAg, and anti-HBe antibody. Total time duration (including hands-on and hands-off times) for each EIA was compared between the two systemsin processing and testing 48 specimens. Direct material costs for consumableswere also compared. When compared to ETI-MAX, the diagnostic sensitivity /specificity and difference in total time duration for each EIA onTriturus were as follows:

Assay Sensitivity (%) Specificity (%) Total time difference (hrs)

Anti-HAV Total 97.5 100 -1.4Anti-HAV IgM 100 100 -1.6HBsAg 100 99.3 -1.2Anti-HBs 100 100 -1.4Anti-HBc IgM 100 100 -0.9HBeAg 100 98.5 -0.9Anti-HBe 98.7 100 -0.9

Significant labor savings were noted for all EIAs performed on the Triturus withhands on times ranging from 0.2 to 0.3 hr for each assay, compared to 1.2 to1.3 hr on ETIMAX. Based on current specimen test volumes in our laboratory,use of the Triturus for these seven EIAs would result in direct material costsavings of $17,500 annually. Although the performance characteristics of thetwo automated systems are comparable, the Triturus has several significantadvantages over the ETI-MAX. In addition to the shorter hands-on times andlabor savings, Triturus offered user-friendly instrument software, improved manualsample tube barcode reader, simple loading process of specimens for continuousbatch testing, minimal time delay between batch testing, dual sampling probeswith non-disposable tips, and capability of detecting clots in specimens.

ETI-MAX 3000ETI-MAX 3000 is a fully automated walk-away microtiter plate analyzer.

ObjectiveThis study evaluated seven DiaSorin hepatitis EIAs on both the Triturus and theETIMAX 3000. Accuracy of test results, labor savings, test turn-around-time,and specimen throughput were compared between the two systems.

Results

TriturusThe Triturus is an open, fully-automated, multi-test, multi-batch analyzer.

ConclusionsTriturus demonstrated the following advantages over ETI-MAX 3000:

• Labor savings due to decreased hands-on time• Higher specimen throughput due to greater flexibility in continuous

batch testing• More efficient software• Easier sample loading process• Large cost savings on consumables (disposable tips are not necessary)• Improved manual barcode reader• Dual testing probes (shorter test turn-around time)

Anti-HAV Total

Triturus +–

Sensitivity = 97.5%Specificity = 100%Total time to processand analyze 48 specimens:ETI-MAX: 5.3 hrTriturus: 3.9 hr

ETI-MAX+ –78 02 120

Anti-HBsAg

Triturus +–

Sensitivity = 100%Specificity = 99.3%Total time to processand analyze 48 specimens:ETI-MAX: 5.3 hrTriturus: 4.1 hr

ETI-MAX+ –54 10 145

Anti-HBc IgM

Triturus +–

Sensitivity = 100%Specificity = 100%Total time to processand analyze 48 specimens:ETI-MAX: 5.7 hrTriturus: 4.8 hr

ETI-MAX+ –53 00 147

Anti-HBe

Triturus +–

Sensitivity = 98.7%Specificity = 100%Total time to processand analyze 48 specimens:ETI-MAX: 5.1 hrTriturus: 3.4 hr

ETI-MAX+ –75 01 124

Anti-IgM

Triturus +–

Sensitivity = 100%Specificity = 100%Total time to processand analyze 48 specimens:ETI-MAX: 5.7 hrTriturus: 4.1 hr

ETI-MAX+ –53 00 147

Anti-HBs

Triturus +–

Sensitivity = 100%Specificity = 100%Total time to processand analyze 48 specimens:ETI-MAX: 5.3 hrTriturus: 3.9 hr

ETI-MAX+ –98 00 102

Anti-HBeAg

Triturus +–

Sensitivity = 100%Specificity = 98.5%Total time to processand analyze 48 specimens:ETI-MAX: 5.1 hrTriturus: 4.2 hr

ETI-MAX+ –65 20 133

An Evaluation Of Three Serologic Commercial EIA Kits For The LaboratoryDiagnosis Of Mycoplasma Pneumoniae Infections

D. J. Jespersen, J. A. Harring, M. F. Jones, T.F. Smith, Ph.D. Mayo Clinic, Rochester, MN

MethodsAll three of the EIAs were programmed on the Triturus and performedaccording to the manufacturer’s product inserts for each kit.

Conclusions• There was a significant variability between commercially available

FDA-approved kits.• All three kits were easily adapted to an automated system.• The Wampole kit provided the shorted testing time.• The Wampole IgG kit provided the best sensitivity and specificity.• The Savyon and Wampole IgM kits yielded similar sensitivities

and specificities [86% and 91%, and 100% and 82%, respectively].• Collectively correlation of EIA with IFA was poor

Meridian IgM Immunocard• 26 of 29 [90%] IFA negative, Wampole IgM EIA positive samples

tested positive with the Meridian IgM Immunocard• 10 of 10 [100%] IFA negative, Diamedix IgM EIA positive samples

tested positive with the Meridian IgM Immunocard• 10 of 13 [77%] IFA negative, Savyon IgM EIA positive samples

tested positive with the Meridian IgM Immunocard

ResultsWampole IgG and IgMThe Wampole EIA is a qualitative system for determining IgG and IgM Antibodiesto M. pneumoniae using plates coated with partially purified inactivated antigen.This assay takes approximately one hour to perform on the Triturus.

Savyon IgG and IgMThe Savyon EIA is a qualitative assay used for determining IgG and IgM antibodiesto M. pneumoniae using plates coated with a purified fraction of M. pneumoniaemembrane proteins containing the P1 protein. This assay takes approximately 2hours and 40 minutes on the Triturus.

Diamedix IgG and IgMThe Diamedix EIA is a Qualitative Assay used for determining IgGand IgM antibodies to M. pneumoniae using plates coated with purifieddetergent – treated M. pneumoniae extracts as antigens. This assay takesapproximately 1 hour and 40 minutes to perform on the Triturus.

ObjectiveLaboratory diagnosis of M. pneumoniae is often established throughserological testing since cultivating the organism is slow and difficult.Many commercial assays are available for the serologic evaluation ofinfection due to M. pneumoniae. Our goal was to determine which ofthree FDA approved IgG and IgM EIA kits performed best on the Triturus, anautomated EIA analyzer, when compared to IFA.

Triturus

AbstractMycoplasma pneumoniae is a common cause of primary atypical pneumoniain both adults and children. An estimated 2 million cases with 100,000pneumonia-related hospitalizations occur in the United States annually. BecauseM. pneumoniae is fastidious and grows very slowly, a definitive diagnosis isoften based on the results of serological testing. Although indirectimmunofluorescence (IFA) testing has historically been considered the “goldstandard,” many laboratories perform an enzyme immunoassay (EIA) for thediagnosis of these infections. Automated EIA analyzers are often used tofacilitate serologic testing and eliminate the subjectivity involved in processingand interpreting results. Our goal was to compare commercially-available EIAkits to the IFA standard method for the laboratory diagnosis of M. pneumoniaewhen tested by the Triturus instrument (Grifols, USA) (Miami, FL.) Two-hundredsamples representing different patterns of serological results by IFA werecoded and analyzed by three commercial kits for both IgG and IgM classantibodies. Kits from Wampole Laboratories (Princeton, NJ), Diamedix (Miami,FL), and Savyon (Afhdod, Israel) were tested. Discordant IgM results wereresolved using the Meridian Immunocard (Cincinnati, Ohio). Sensitivities andspecificities for IgG were 52% and 98% with the Diamedix assay, 70% and94% with the Savyon assay, and 93% and 96% with the Wampole kit,respectively. IgM results were similar with 69% and 94% with theDiamedix kit, 86% and 92% with the Savyon assay and 100% and 82% withthe Wampole assay. IgM samples that were negative by IFA but positive byEIA were retested with the Immunocard and 77% of Savyon, 100% of Diamedix,and 90% of the Wampole results retested as positive. Although serology maybe helpful, these results indicate a significant variability between FDA approvedkits. We recommend that the diagnosis of M. pneumoniae infection not bebased on serology alone, but should be correlated with clinical findings.

The Triturus is an open, Fully-automated, Multi-test,Multi-Batch analyzer that can Accommodate any EIA

IgG ≥ 1:40IFA

+ –

+ 138 2– 11 49

sens = 93%spec = 96%

IgM ≥ 1:10IFA

+ –

+ 44 29– 0 127

sens =100%spec = 82%

IgG ≥ 1:40IFA

+ –

+ 102 3– 44 51

sens = 70%spec = 94%

IgM ≥ 1:10IFA

+ –

+ 36 13– 6 145

sens =86%spec = 92%

IgG ≥ 1:40IFA

+ –

+ 77 1– 72 50

sens = 52%spec = 98%

IgM ≥ 1:10IFA

+ –

+ 31 10– 14 145

sens =69%spec = 94%

VALIDITY OF ENZYME IMMUNOASSAY (EIA) SEROLOGIC RESULTS FOR THELABORATORY DIAGNOSIS OF EPSTEIN-BARR VIRUS (EBV) INFECTIONS.

Julie A. Helgason, Deborah J. Jespersen, Mary F. Jones, Elaine M. Beito, and Thomas F. Smith.Mayo Clinic and Foundation, Rochester, MN 55905

IntroductionEpstein-Barr virus (EBV) is the etiological agent of infectious mononucleosis (IM), Burkitt’slymphoma and nasopharyngeal carcinoma. The virus occurs worldwide and most peoplebecome infected with EBV in early childhood.Symptomatic infection, commonly in young adults, produces fever, sore throat and swollenlymph glands. IM (primary infection with EBV) is usually diagnosed in the laboratory byrapid latex tests based on detection of heterophile antibodies. Persistant, active EBVinfection can occur as primary or reactivated latent virus in immunocompromised hosts.These infections manifest as posttransplantation lymphomas and smooth muscle celltumors. These EBV associated conditions are diagnosed by detection of antibodies tospecific antigens of the virus. Although indirect fluorescent antibody (IFA) testing hashistorically been considered the gold standard and is most often used to detect antibodiesto EBV antigens, the IFA assays are labor-intensive and require the experience andsubjective judgment of an expert technologist. In addition, IFA measurements of serumspecimens may be misleading due to non-specific reactions for patients with autoimmunediseases. Serologic results of enzyme immunoassays (EIA) correlate with IFA measurementsand interpretation is standardized. The EIA procedure allows an objective determinationof antibody status to be made from a single dilution of.

Abstract

MethodBased on predominant use reported in proficiency test surveys, we selected fivemanufacturers kits to evaluate on the Triturus® continuous batch analyzer; Diamedix(Diamedix Corporation, Miami, FL), DiaSorin (DiaSorin Inc., Stillwater, MN), SeraQuest(SeraQuest, North Miami, FL), Wampole (Wampole Laboratories, Dist., Cranbury, NJ) andZeus (Zeus Scientific, Inc., Raritan, NJ). One hundred-fifty samples representing thedifferent patterns of serological results (VCA IgG,VCA IgM and EBNA IgG) were analyzedand compared with IFA. Discordant results with wide ranges of sensitivity and specificityfor EBV markers were obtained with all assay kits, although the SeraQuest assaydemonstrated the best overall performance. A well-defined panel of sera [Anti-EBV MixedTiter Performance Panel (PME201)] obtained from Boston Biomedica, Inc. [(BBI); Bridgewater,CT] was used to further compare the SeraQuest assay with the IFA method. Results ofanalysis of this panel demonstrated the same overall lack of sensitivity previously foundwith the other assays.

Triturus

Conclusions1. All current generation EIA assays were easily adapted to the Triturus® continuous

batch analyzer; however, our results indicated that all commercial EBV EIA kits produce variable results as compared to IFA serology.

2. Discordant results were obtained with all assay kits; however, the SeraQuest EIA kithad the best correlation compared with IFA results.

3. We recommend that the diagnosis of persistent EBV infection not be based on EIA results alone. Rather, interpretation of serologic results, together with clinical featuresand other laboratory findings, be collectively considered for the diagnosis of EBV infections.

Antibody Responses toEBV-specific Antigens

Diseases

VCA-IgG VCA-IgM Anti-EBNA-IgG

IFA

Sensitivity vs. IFA Specificity vs. IFA

VCA IgG 51 % 100 %Diamedix VCA IgM 94.9 % 75.8 %

EBNA IgG 79.5 % 97.1 %

Sensitivity vs. IFA Specificity vs. IFA

VCA IgG 89.8 % 95.7 %DiaSorin VCA IgM 90.5 % 54.2 %

EBNA IgG 71.1 % 92.9 %

Sensitivity vs. IFA Specificity vs. IFA

VCA IgG 82.6 % 100 %SeraQuest VCA IgM 100 % 55.7 %

EBNA IgG 73 % 97.1 %

Sensitivity vs. IFA Specificity vs. IFA

VCA IgG 55.1 % 100 %Wampole VCA IgM 76.3 % 33.8 %

EBNA IgG 81.6 % 95.7 %

Sensitivity vs. IFA Specificity vs. IFA

VCA IgG 82.9 % 95.2 %Zeus VCA IgM 92.7 % 75.2 %

EBNA IgG 74.7 % 94 %

PHASE I. BEST CORRELATION COMPARED TO IFA WAS OBTAINED WITHTHE SERAQUEST ASSAY.

Sensitivity vs. IFA Specificity vs. IFA

VCA IgG 83.3 % 100 %SeraQuest vs VCA IgM 77.8 % 100 %BBI Panel EBNA IgG 87.5 % 87.5 %

PHASE II. THE SERAQUEST ASSAY WAS THEN COMPARED TO IFA WITH A CASEDEFINED PANEL OF SERA OBTAINED FROM BBI.

Epstein-Barr Virus is a member of the herpesvirus group. EBV is the causative agent ofinfectious mononucleosis; however, this virus can produce persistent infections andneoplastic conditions in immunocompromised hosts. Although indirect fluorescent antibody(IFA) testing has historically been considered the gold standard, most laboratories performenzyme immunoassay for the diagnosis of these infections. Automated EIA analyzers areoften used to eliminate the subjectivity involved in processing and interpreting serologicalresults. Our goal was to determine which commercially-available EBV EIA kit had the bestperformance characteristics when tested by the Triturus® instrument (Diagnostic Grifols,S.A.). Based on predominant use reported in proficiency test surveys, we selected Diamedix(Diamedix Corporation, Miami, FL), DiaSorin (DiaSorin Inc., Stillwater, MN), SeraQuest,(SeraQuest, North Miami, FL), Wampole (Wampole Laboratories, Dist., Cranbury NJ) andZeus (Zeus Scientific, Inc., Raritan,NJ) EIA kits to compare with IFA. One hundred-fiftysamples that represented the different patterns of serological results (VCA, IgG, VCA IgM,EBNA IgG) were analyzed. Discordant results with wide ranges of sensitivity and specificityfor EBV markers were obtained with all assay kits, although the SeraQuest assaydemonstrated the best overall performance.

Commercial Supplier Sensitivity vs. IFA Specificity vs. IFA

SeraQuest VCA IgG 82.6% 100%VCA IgM 100% 55.7%

EBNA IgG 73% 97.1%

A well-defined panel of sera [Anti-EBV Mixed Titer Performance Panel (PME201)] obtained fromBoston Biomedica, Inc. [(BBI); Bridgewater, CT] was used to further compare the SeraQuest assaywith the IFA method. Results of this panel analysis demonstrated the same overall lack of sensitivitypreviously found with the other assays.

SeraQuest vs. BBI Panel Sensitivity vs. IFA Specificity vs. IFA

VCA IgG 83.3% 100%VCA IgM 77.8% 100%EBNA IgG 87.5% 87.5%

With the Triturus automated instrument for serologic testing, all evaluated kits performedsimilarly. All current generation EIA assays were easily adapted to the Triturus instrument;however, all commercial EBV EIA kits produced variable results as compared to IFA serology.We recommend that the diagnosis of EBV infections not be based on EIA results alone. Rather,interpretation of serologic results together with clinical features and other laboratory findings[hematological evaluation and other serological markers (e.g. rapid screening for infectious.

07/3

337

1050

7