Selected Issues in Immunoassays: How to Evaluate Selected Issues in Immunoassays: How to Evaluate Problems and support Clinically Relevant DecisionsProblems and support Clinically Relevant Decisions

Workshop presented at the Canadian Laboratory Workshop presented at the Canadian Laboratory Medicine Congress, Medicine Congress,

Westin Westin HarbourHarbour Castle Hotel, Toronto, ON Castle Hotel, Toronto, ON Sunday Jun 10, 2007 (13:30 Sunday Jun 10, 2007 (13:30 –– 17:00)17:00)

Dr Godfrey Moses, Laboratory Director, GDML, London Dr Godfrey Moses, Laboratory Director, GDML, London & &

Dr. Patrick St.Louis, Scientific Director, MDS Dr. Patrick St.Louis, Scientific Director, MDS PharmaPharma

Generally Issues are Related toGenerally Issues are Related toAssay/MethodAssay/Method–– RIA RIA vsvs Others (ELISA, EIA, CLIA, FPIA, IRMA, TIA)Others (ELISA, EIA, CLIA, FPIA, IRMA, TIA)–– RIARIA…… Extraction (Indirect) Extraction (Indirect) vsvs NonNon--Extraction (Direct)Extraction (Direct)

»» CortisolCortisol, Vitamin D, Testosterone, Progesterone, etc., Vitamin D, Testosterone, Progesterone, etc.»» Well DocumentedWell Documented…… ClinClin ChemChem 49:1381 49:1381 –– 1395 (2003); 1395 (2003);

EQAS & CAP Summary ReportsEQAS & CAP Summary Reports»» Method of choice now is HPLC/Tandem Mass SpectrometryMethod of choice now is HPLC/Tandem Mass Spectrometry

Generally Issues Generally Issues contdcontd 11……AnalyteAnalyte–– Free & BioFree & Bio--available Testosterone; Free & available Testosterone; Free & ComplexedComplexed PSA; Beta PSA; Beta

hCGhCG’’ss; ; DigoxinDigoxin & & DigoxinDigoxin--Like SubstancesLike Substances–– Well DocumentedWell Documented……. Dr P Y Wong; . Dr P Y Wong; ManufManuf. Kit Inserts; . Kit Inserts; ClinClin

BiochemBiochemEndogenous & Exogenous Endogenous & Exogenous InterferentsInterferents–– Drugs (e.g. Drugs (e.g. DigoxinDigoxin & anti& anti--aldosteronealdosterone/hypertensive agents)/hypertensive agents)–– ImmunoglobulinsImmunoglobulins (e.g. (e.g. MacroprolactinMacroprolactin; Macro; Macro--AlkAlk PhosPhos))–– Monoclonal Proteins Monoclonal Proteins -- TIATIA’’ss

OthersOthers–– AntiseraAntisera Specificity & Source/Type; Polyclonal Specificity & Source/Type; Polyclonal vsvs MonclonalMonclonal; ;

Single Single vsvs Double Antibody; Human; Animals Double Antibody; Human; Animals

Generally Issues Generally Issues contdcontd 22……

–– Good reviews on Good reviews on analyteanalyte related issues related issues

»» Dr P Y Wong. Irritating Interference and Erroneous Errors in Dr P Y Wong. Irritating Interference and Erroneous Errors in Hormone immunoassaysHormone immunoassays…… a view from the laboratory . a view from the laboratory . Presented at the OSCC Annual Scientific Meeting , Hamilton Presented at the OSCC Annual Scientific Meeting , Hamilton Chamber of Commerce, Nov 2003.Chamber of Commerce, Nov 2003.

»» Dr L A Cole. Immunoassay of human Dr L A Cole. Immunoassay of human chorionicchorionic gonadotropingonadotropin, , its free subunits, and metabolites. its free subunits, and metabolites. ClinClin ChemChem 43: 2233 43: 2233 –– 2243 2243 (1997)(1997)

RIA vs Others – Anti-ds DNARIA vs. EIA A

(n = 83)

RIANormal (Up to 3.6 KU/L)

RIAAbnormal (> 4.1 KU/L)

EIA ANormal (< 25 IU/ml)

51 11

EIA A Abnormal (> 35 IU/ml)

5 16

Overall Agreement (%): 81False Positive Rate (%): 9False Negative Rate (%): 41

RIA vs Others – Anti-ds DNARIA vs. EIA B

(N = 62)

RIANormal (Up to 3.6 KU/L)

RIAAbnormal (> 4.1 KU/L)

EIA BNormal (< 25 IU/ml)

42 10

EIA BAbnormal (> 35 IU/ml)

2 8

Overall Agreement (%): 81False Positive Rate (%): 5False Negative Rate (%): 56

EIA A vs. EIA B(n = 57)

EIA ANormal (< 25 IU/ml)

EIA AAbnormal (> 35 IU/ml)

EIA BNormal (< 25 IU/ml)

43 4

EIA BAbnormal (> 35 IU/ml)

3 7

Overall Agreement (%): 88False Positive Rate (%): 7False Negative Rate (%): 36

RIA vs Others – Anti-ds DNA

ANA IFA Pattern

- (Homogenous) + (Homogenous)

RIA ─ 17 8

+ 12 9

EIA A ─ 22 9

+ 6 6

EIA B ─ 15 5

+ 3 1

RIA EIA A EIA B

% Agreement (overall) 57 65 67

False Negative Rate (%)

41 21 17

False Positive Rate (%) 47 60 83

Overall Assessment; Anti-ds DNA: EIA/RIA vs IFA (Gold STD)

AnalyteAnalyte Related Related –– Beta Beta hCGhCG

Antibody HeteroAntibody Hetero--specificity to different specificity to different hCGhCG species:species:–– Intact Beta Intact Beta hCGhCG assayassay–– Free Free hCGhCG BetaBeta--subunit assaysubunit assay–– Total Total hCGhCG assayassay–– Total Total hCGhCG + Beta assay+ Beta assay

False positive False positive hCGhCG results leading to results leading to unnecessary surgery and chemotherapy and unnecessary surgery and chemotherapy and needless occurrences of diabetes and comaneedless occurrences of diabetes and coma

AnalyteAnalyte Related Related –– A case of AntiA case of Anti--analyteanalyte antibodies antibodies ( PY Wong, OSCC Nov 2003)( PY Wong, OSCC Nov 2003)

84 y old female with clinically diagnosed hypothyroidism84 y old female with clinically diagnosed hypothyroidismInitial Lab Result: TT4 >500 nmol/L(RI:60 Initial Lab Result: TT4 >500 nmol/L(RI:60 –– 150)150)Further Further TFTTFT’’ss::

–– TT4 = 840 (after methanol Extraction = 140)TT4 = 840 (after methanol Extraction = 140)–– TT3 = 50 (after methanol extraction = 20; RI 1.2 TT3 = 50 (after methanol extraction = 20; RI 1.2 –– 2.8)2.8)–– FT4 (analog) = 124 FT4 (analog) = 124 nmolnmol/L (RI: 10 /L (RI: 10 –– 25)25)–– FT4 (dialysis FT4 (dialysis EquEqu) = 10 (RI: 12 ) = 10 (RI: 12 –– 42)42)–– TSH = 18 (0.3 TSH = 18 (0.3 –– 5.0)5.0)

Clinical ProtocolClinical Protocol–– If TSH, FT4, TT4, TT3 all elevated, refer patient for MRI of theIf TSH, FT4, TT4, TT3 all elevated, refer patient for MRI of the brain to brain to

rule out pituitary adenoma rule out pituitary adenoma –– MRI result: No evidence of Pit Adenoma MRI result: No evidence of Pit Adenoma

Appropriate Diagnosis: Hypothyroidism (high TSH) with Appropriate Diagnosis: Hypothyroidism (high TSH) with artifactualartifactual elevated elevated thyroid hormone concentrations due to antithyroid hormone concentrations due to anti--T4 & antiT4 & anti--T3 antibodies. Even T3 antibodies. Even after methanol extraction TT3 &TT4 remained significantly elevatafter methanol extraction TT3 &TT4 remained significantly elevateded.

AnalyteAnalyte Related Related ––TestosteroneTestosterone

Immunoassays (Immunoassays (espesp direct) not suitable for assaying direct) not suitable for assaying TestoTesto in children in children & woman; even those with solvent extraction may variable results& woman; even those with solvent extraction may variable results ––HPLC & Mass Spec method of choiceHPLC & Mass Spec method of choiceTaiebTaieb et al et al ClinClin ChemChem 2003; 49(8): 1381 2003; 49(8): 1381 –– 13911391–– 10 Immunoassays 10 Immunoassays vsvs ID/GCMS for total ID/GCMS for total TestoTesto–– 7/10 had higher values than GCMS (up to 46% higher) in Females7/10 had higher values than GCMS (up to 46% higher) in Females–– 4/10 had lower values than GCMS (up to 46% lower in Males4/10 had lower values than GCMS (up to 46% lower in Males–– Magnitude of Mean Differences (Bias) were also different in maleMagnitude of Mean Differences (Bias) were also different in maless–– & females& females–– 0/10 was sufficiently reliable for use in Children & women0/10 was sufficiently reliable for use in Children & women–– 7/10 had r > 0.95 but only in men; r < 0.95 for females in all 17/10 had r > 0.95 but only in men; r < 0.95 for females in all 10.0.

Same for Free Same for Free TestoTesto -- Yet several labs in Ontario are currently using Yet several labs in Ontario are currently using the DPC RIA assaythe DPC RIA assayCalculated Free Calculated Free TestoTesto & BAT & BAT –– Produce comparable, if not superior, results to the measured Produce comparable, if not superior, results to the measured

parametersparameters–– Issue is how to implement the new & more accurate alternatives iIssue is how to implement the new & more accurate alternatives into nto

the routine laboratory that relies heavily on automation the routine laboratory that relies heavily on automation

Endogenous/Exogenous Endogenous/Exogenous InterferentsInterferents –– Drug Effects on Drug Effects on DigoxinDigoxin AssaysAssays

DigoxinDigoxin Assays: Frequent, Substantial, and Potentially Dangerous Assays: Frequent, Substantial, and Potentially Dangerous Interference by Interference by SpironolactoneSpironolactone, , CanrenoneCanrenone, and Other Steroids, and Other Steroids–– Werner Werner SteimerSteimer, Christine Muller, and Barbara , Christine Muller, and Barbara EberEber. Drug . Drug MonitMonit & &

ToxicolToxicol 48: 507 48: 507 ––516 (2002)516 (2002)–– 9 assays (Major Manufacturers 9 assays (Major Manufacturers –– Abbot, Dade Abbot, Dade BehringBehring, Roche & Ortho), Roche & Ortho)–– EMIT, MEIA, FPIA, TIA, ECLIAEMIT, MEIA, FPIA, TIA, ECLIA–– DPCDPC’’ss CLIA Assays were CLIA Assays were Not Not studiedstudied

ConclusionConclusion–– VitrosVitros, TIA & EMIT: Minimal or No Interference, TIA & EMIT: Minimal or No Interference–– TDxTDx, ACA & ECLIA: Falsely Increased , ACA & ECLIA: Falsely Increased DigoxinDigoxin–– Dimension, Dimension, AxSYMAxSYM & & ImxImx: Falsely Reduced Results: Falsely Reduced Results

FalseFalse--Negative & FalseNegative & False--Positive Effects were observed at relatively Positive Effects were observed at relatively low drug concentrationlow drug concentration

Endogenous/Exogenous Endogenous/Exogenous InterferentsInterferents –– ImmunoglobinsImmunoglobins

MacroprolactinMacroprolactin ((ProlactinProlactin--IgIg Complex)Complex)–– Falsely Increased Falsely Increased ProlactinProlactin Levels; Suggesting Levels; Suggesting ProlactinomaProlactinoma–– MPRL; not biological activeMPRL; not biological active–– Most, but not all, commonly used methods are susceptibleMost, but not all, commonly used methods are susceptible–– Need to be delineated from true Need to be delineated from true prolactinemiaprolactinemia–– If unidentified, can lead to unnecessary clinical workIf unidentified, can lead to unnecessary clinical work--upup

Common ApproachesCommon Approaches–– Screen with existing methods & send elevated Screen with existing methods & send elevated PRLPRL’’ss & Ref out to other & Ref out to other

testing site testing site –– Issue is TAT & difference in methodology for PRL Issue is TAT & difference in methodology for PRL –– Removal of Removal of macroPRLmacroPRL on elevated PRL & repeat testing on elevated PRL & repeat testing

»» UltracentrifugationUltracentrifugation»» PEG precipitationPEG precipitation»» Gel Filtration Gel Filtration chromatogaphychromatogaphy ((GFcGFc; gold std but not amenable to run use); gold std but not amenable to run use)

–– Issues are effectiveness,easeIssues are effectiveness,ease--ofof--use and cost of the separation techniques, use and cost of the separation techniques, as well as, appropriateness of cutas well as, appropriateness of cut--off points and interpretation of resultsoff points and interpretation of results

–– A Case of A Case of MacroPRLMacroPRL method comparison (GDML) method comparison (GDML)

Evaluation of inEvaluation of in--house PEG Precipitation Assay house PEG Precipitation Assay –– Abbott ArchitectAbbott Architect

ANALYTICAL EVALUATION OF MACROPROLACTIN ON ABBOTT ARCHITECT i2000 SR and i2000 - Moses, Vandenberghe,Waite, and Sumner. Method: Randomly selected male and female samples (n=100) with prolactinabove the reference range (18-224 ug/L) were analyzed for prolactin following 1:2 dilution with Architect Multi-Assay Diluent and with 250 g/L PEG 6000 in same Architect diluent. Two methods of sample preparation were compared: A) precipitation in conical centrifugation tubes, ultracentrifugation at 3,750g and transfer of supernatant to cups with B) preparation in small polypropylene tubes (10x70 mm), centrifugation at 4,140g and direct sampling of supernatant by the by the Architect.

Results: Prolactin in PEG reagent was < 0.6ug/L. Correlation between straight and diluted prolactin was Y=1.011X -2.1 (r=0.997). Following PEG precipitation the % recovery of prolactin was: * 55 patients had >60% prolactinrecovery suggesting monomeric form, * 28 patients had <40% recovery consistent with macroprolactin and * 17 patients had between 40-60% recovery representing a mixture. Correlation between the 2 methods was Y=1.071X –0.90 (r=0.988) and interpretation of prolactin recovery after PEG precipitation the same.Conclusion: Prolactin and macroprolactin measurements are unaffected by dilution and PEG precipitation. Both sample precipitation techniques are acceptable for routine use in the primary care setting. The relative frequency of macroprolactin in our randomly selected group is consistent with those previously reported.

Performance Comparison of the Roche E170 PEG Performance Comparison of the Roche E170 PEG and Nonand Non--PEG Assays PEG Assays vsvs InIn--house PEP Architect house PEP Architect

AssayAssay

With Respect to With Respect to MacroPRLMacroPRL (in spite of slightly different RI), we (in spite of slightly different RI), we obtained 100% agreement between Beckman obtained 100% agreement between Beckman DxiDxi and the Architect and the Architect PEG protocol for actual patients (n=25) with PRL > Architect URLPEG protocol for actual patients (n=25) with PRL > Architect URL (1 (1 pt was in the gray zone by Architect & no pt was in the gray zone by Architect & no MacroPRLMacroPRL present for the present for the DxiDxi))For For monmericmonmeric (total (total PrlPrl), all elevated by Architect (n=28) were also ), all elevated by Architect (n=28) were also elevated by Roche original method but only 23 were elevated by televated by Roche original method but only 23 were elevated by the he new Roche new Roche mehtodmehtod (others were elevated for males but normal for (others were elevated for males but normal for females)females)Of the 28 elevated patients, Of the 28 elevated patients, MacroPRLMacroPRL was detected by the original was detected by the original Roche method (PEG) in 5. (19 %; consistent with previously reporRoche method (PEG) in 5. (19 %; consistent with previously reported ted values of 19 values of 19 –– 22 )22 )With use of appropriate RI, the new Roche appears to meet the With use of appropriate RI, the new Roche appears to meet the manufacturermanufacturer’’s claim of no interference from s claim of no interference from MacroPRLMacroPRL

Issues: Issues: MacroPRLMacroPRL –– CutCut--pointspoints

All All PRLPRL’’ss > ULR> ULRAbbott Abbott BeckmanBeckman RocheRoche

–– Males: Males: < 18 < 18 2 < 18 < 18 2 –– 2020–– Females Females < 26 < 25 3 < 26 < 25 3 –– 3030–– F (F (PregPreg) ) 10 10 –– 300300 10 10 –– 300 10 300 10 –– 350350

Macro PRLMacro PRL–– > 60 %, MPRL > 60 %, MPRL NOTNOT Present (Supernatant)Present (Supernatant)–– 40 40 –– 60, % Mixture of Macro, 60, % Mixture of Macro, OligoOligo & N& N--PRL PRL –– < 40 %, MPRL Present (Supernatant)< 40 %, MPRL Present (Supernatant)–– Opposite if the MPRL is measured on PPT (e.g. Opposite if the MPRL is measured on PPT (e.g. BeckamnBeckamn DX1 DX1

assay)assay)

Endogenous/Exogenous Endogenous/Exogenous InterferentsInterferents –– DigoxinDigoxin: The : The IssuesIssues……

What should Labs do if method is susceptibleWhat should Labs do if method is susceptible–– Change method (not always an option)Change method (not always an option)–– PrePre--treat samples to remove/minimize effects (not an options for treat samples to remove/minimize effects (not an options for

automated methods)automated methods)–– Wait for Supplier to correct the problem & add an interpretive Wait for Supplier to correct the problem & add an interpretive

comment to all patient reports (option taken by most labs)comment to all patient reports (option taken by most labs)Typical Comment ( Typical Comment ( e.g.GDMLe.g.GDML’’ss) )

““Very recent authoritative studies show that Very recent authoritative studies show that digoxindigoxin results by results by most commonly used methods including ours, can be used most commonly used methods including ours, can be used suppressed up to 40% by suppressed up to 40% by spironolactonespironolactone, , canrenonecanrenone, , hydrocortisone and hydrocortisone and prednisoloneprednisolone. We will inform you when we . We will inform you when we change to a method not affected by these drugschange to a method not affected by these drugs””

Question Question –– How do participants approach this issue in their Labs? How do participants approach this issue in their Labs?

Endogenous/Exogenous Endogenous/Exogenous InterferentsInterferents –– DigoxinDigoxin; The ; The IssuesIssues……

Waiting for Supplier creates other issuesWaiting for Supplier creates other issues–– New formulation yields clinically different results from old New formulation yields clinically different results from old

formulationformulation–– Over/Under correction of the initial problem or reformulation Over/Under correction of the initial problem or reformulation

introduced added errorintroduced added error–– Abbott Abbott AxSYMAxSYM Dig III yield 30% higher results than Dig IIDig III yield 30% higher results than Dig II–– Also Claims that drug effect is no longer an issue with Dig III Also Claims that drug effect is no longer an issue with Dig III

Abbott Dig III results did not correlate with the Roche TIA Abbott Dig III results did not correlate with the Roche TIA or DPC Assays, which have been shown not to affected by or DPC Assays, which have been shown not to affected by the drug the drug

n 38 (cases excluded: 5 due to missing val

Bias 0.28895% CI 0.214 to 0.362

95% limits of agreement 95% CILower -0.154 -0.276 to -0.032Upper 0.729 0.607 to 0.852

Zero bias

-0.2

0

0.2

0.4

0.6

0.8

0 2 4

Mean of all methods

Diff

eren

ce b

etwee

n m

etho

ds

Identity line A=B

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 2 4

Dig II Axsym

Dig

III A

xsym

0 5 10 15

Zero bias

-40

-20

0

20

40

60

80

100

120

140

0 2 4

Mean of all methods

Diff

eren

ce b

etwee

n m

etho

ds (%

)

Comparison studies Comparison studies –– AxSYMAxSYM III III vsvs IIII

n 35 (cases excluded: 8 due to missing val

Bias 0.25795% CI 0.117 to 0.396

95% limits of agreement 95% CILower -0.539 -0.769 to -0.309Upper 1.053 0.823 to 1.283

Zero bias

-1

-0.5

0

0.5

1

1.5

2

0 2 4

Mean of all methods

Diff

eren

ce b

etwee

n m

etho

ds

Identity line A=B

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

0 2 4

Integra Result

Dig

III A

xsym

0 10 20 30

Zero bias

-60

-40

-20

0

20

40

60

0 2 4

Mean of all methods

Diff

eren

ce b

etwee

n m

etho

ds (%

)

Comparison studies Comparison studies –– AxSYMAxSYM III III vsvs IntegraIntegra

Others Others -- AntiseraAntisera Specificity & Source/Type, ETCSpecificity & Source/Type, ETC

False Positives in Serological AssaysFalse Positives in Serological Assays–– IgMIgM due to RF & ANA Resultsdue to RF & ANA Results–– FcFc FragementsFragements ogog IgGIgG Antibodies used with Enzyme Antibodies used with Enzyme

conjugates (way around this is to use conjugates (way around this is to use FabFab fragments)fragments)–– AntiAnti--ToxoplasmaToxoplasma IgGIgG & & IgMIgM

»» Major problem is Major problem is IgMIgM specific test lacks specificityspecific test lacks specificity»» A Specific A Specific IgGIgG test is done in conjunction with test is done in conjunction with IgMIgM»» Results are interpreted in a binary mannerResults are interpreted in a binary manner

–– AntiAnti--Rubella AntibodyRubella Antibody»» Similar problems but Immunoassays have improved Similar problems but Immunoassays have improved specifityspecifity

Others Others -- AntiseraAntisera Specificity & Source/Type, ETCSpecificity & Source/Type, ETC

HAMA (Human antiHAMA (Human anti--Mouse AntibodyMouse AntibodyHA (HA (HetereophilicHetereophilic Antibodies)Antibodies)ManufacturerManufacturer’’s disclaimers (Motherhood s disclaimers (Motherhood statements)statements)Limitations to the proceduresLimitations to the proceduresCrossCross--reactivity informationreactivity information–– If negative effect, how does crossIf negative effect, how does cross--reactivity reactivity

determined?determined?–– Not the same as Interference testingNot the same as Interference testing–– Not often quantifiable.Not often quantifiable.

ConclusionsConclusionsPY Wong classified such issues asPY Wong classified such issues as……–– Exogenous (carryover, contaminants affecting label detection, Exogenous (carryover, contaminants affecting label detection,

additives in blood collection tubes & microadditives in blood collection tubes & micro--clots) clots) –– Endogenous (AntiEndogenous (Anti--analyteanalyte Antibodies, Human antiAntibodies, Human anti--mouse mouse

antibodies, antibodies, HeterophilicHeterophilic antibodies, crossantibodies, cross--reacting substance, high reacting substance, high dos hook effect & nondos hook effect & non--specific interference)specific interference)

–– All, or most of them, we encounter on an onAll, or most of them, we encounter on an on--going basis in our going basis in our dayday--toto--day activities in the clinical laboratoryday activities in the clinical laboratory

–– The presentation is a general The presentation is a general oveviewoveview of current issues in this wide of current issues in this wide & varied field. Hopefully, you will find the information useful.& varied field. Hopefully, you will find the information useful.

–– I note with much amazement, the last major breakthrough I note with much amazement, the last major breakthrough (milestone in the development) in Immunoassays was Kohler & (milestone in the development) in Immunoassays was Kohler & MilesteinMilestein’’ss monoclonal antibodies 1975 (32 years ago)!!!monoclonal antibodies 1975 (32 years ago)!!!

AcknowledgementsAcknowledgements

I would like to thank Gayle Waite, I would like to thank Gayle Waite, Manager, Chemistry GDML, for help in Manager, Chemistry GDML, for help in assembling the information for this talk; my assembling the information for this talk; my former former colleauguecolleaugue Dr Dr HildeHilde VandenbergheVandenberghefor her contribution on the for her contribution on the MacroprolactinMacroprolactinstudy and the supervisory staff in Chemistry study and the supervisory staff in Chemistry for their assistance in performing the for their assistance in performing the evaluations.evaluations.