Proficiency Survey-Based Evaluation of Clinical Total and Allergen-Specific IgE Assay Performance. Robert G. Hamilton, PhD, D (ABMLI)

Transcription

1 Proficiency Survey-Based Evaluation of Clinical Total and Allergen-Specific IgE Assay Performance N Context. The diagnostic algorithm for human allergic disease involves confirmation of sensitization by detection of allergen-specific immunoglobulin E (IgE) antibody in individuals suspected of having allergic disease because of a history of allergic symptoms after known allergen exposure. Previous studies showed wide disparity among clinically reported allergen-specific IgE levels from different serologic assays. Objective. To validate the relative analytic performance (sensitivity, interassay reproducibility, linearity/ parallelism, intermethod agreement) of clinically used total and allergen-specific IgE assays by using College of American Pathologists Diagnostic Allergy SE Proficiency Survey data. Design. Data from 2 SE survey cycles were used to assess relative analytic performance of the ImmunoCAP (Phadia), Immulite (Siemens Healthcare-Diagnostics), and HYTEC 288 (HYCOR-Agilent Technologies) total and allergen-specific IgE assays. In each cycle, 2 recalcified plasma pools from atopic donors were diluted twice with IgE-negative serum and evaluated in approximately 200 The diagnosis of human allergic disease begins with a clinical history. When the history identifies an association between allergic symptoms and a known allergen exposure, the diagnostic algorithm recommends confirming sensitization by monitoring the patient for allergen-specific immunoglobulin E (IgE) antibody by a skin test or serologic assay. Historically, dichotomous IgE antibody test results in the skin or blood have been used to enhance (positive IgE) or diminish (negative IgE) the probability of allergic disease. The presence of IgE antibody also has been used to confirm atopic status and verify the specificity of offending allergen triggers. 1,2 Serologic assays used in the quantification of total and allergen-specific IgE have continued to experience marked technical improvements. Recent advances have involved automation, which has led to improved precision and reduced turnaround times and to the use of nonisotopic labels that have improved sensitivity with new substrates Accepted for publication October 28, From the Departments of Pathology and Medicine, Johns Hopkins University School of Medicine, Baltimore, Maryland. The author has no relevant financial interest in the products or companies described in this article. Reprints: Robert G. Hamilton, PhD, D (ABMLI), Departments of Pathology and Medicine, Johns Hopkins University School of Medicine, Room 1A20, Johns Hopkins Asthma and Allergy Center, 5501 Hopkins Bayview Cir, Baltimore, Maryland ( rhamilt2@jhmi.edu). Robert G. Hamilton, PhD, D (ABMLI) federally certified clinical laboratories for total IgE and IgE antibody to 5 allergen specificities. Statistical analysis evaluated analytic sensitivity, linearity, reproducibility, and intermethod agreement. Results. Interlaboratory intramethod, intermethod, and interdilution agreement of all 6 clinically used total serum IgE assays were excellent, with coefficients of variation (CVs) below 15%. Interlaboratory intramethod, and interdilution agreement of 3 clinically used allergenspecific IgE assays were also excellent with CVs below 15%. However, intermethod CVs identified between-assay disagreement greater than 20% in 80% of allergen-specific IgE measurements. Allergen reagents and patients immune response heterogeneity are suggested probable causes. Conclusions. Clinical total and allergen-specific IgE assays display excellent analytic sensitivity, precision, reproducibility, and linearity. Marked variability in quantitative estimates of allergen-specific IgE from clinically used automated immunoassays is a concern that may be ameliorated with component allergen use. (Arch Pathol Lab Med. 2010;134: ) and reduced nonspecific binding. These advances in recent years have enhanced the sensitivity and accuracy of IgE antibody assays. 3 As a result, investigators are increasingly using the magnitude of the IgE antibody responses to ingested or inhaled allergens to assess the patients relative probability for true allergic reactivity. 4 7 However, published IgE antibody levels, reported as 95% confidence interval cutpoints generated by 1 IgE antibody assay for defining positive allergic reactivity, cannot be used with IgE antibody results generated with a different IgE antibody assay. Many assay designs and companies have come and gone in the IgE antibody field since development of the original 1968 radioallergosorbent test. 8,9 Immunoassays from Phadia (ImmunoCAP, Uppsala, Sweden), Siemens Healthcare Diagnostics (Immulite, Newark, Delaware), and HYCOR-Agilent Technologies (HYTEC 288, Garden Grove, California) have remained the only methods in clinical laboratory use to assess patients serum for IgE antibody in North America. Newer microarray-based chip assays with component allergens (immuno solid-phase allergen chip or ISAC; VBC Genomics, Vienna, Austria and Phadia) and the point-of-care lateral flow assay (ImmunoCAP Rapid, Phadia) are still considered emerging technologies for allergen-specific IgE antibody detection that have not yet made mainstream testing in North American laboratories. 10,11 A critical review of IgE anti- Arch Pathol Lab Med Vol 134, July 2010 IgE Antibody Assay Performance Hamilton 975

2 body assay accuracy of the 3 established IgE antibody assays was reported in 2007, by using limited IgE antipeanut and engineered humanized IgE chimeric antibody data that were measured in single laboratories. This study 12 concluded that the available clinical assays report highly variable quantitative levels of IgE antibody. From the time of this report, the technical improvements involving automation and enhanced reagent quality have encouraged reexamination of this conclusion, especially by using more impartial data involving a broader number of allergen specificities and a larger number of participating laboratories. The current study was therefore performed to reexamine current physician concerns 12 about the lack of agreement among the allergen-specific IgE antibody levels that are measured with different IgE antibody assays used in North American clinical laboratories. In this report, analytic sensitivity, interlaboratory reproducibility, parallelism/linearity, and intermethod agreement, as an indicator of relative accuracy of clinically used total and allergen-specific IgE antibody assays, are examined using College of American Pathologists (CAP) Diagnostic Allergy Proficiency Survey data. MATERIALS AND METHODS Proficiency Challenge Specimens During each of 2 cycles of the CAP SE survey (cycle C-2008 and cycle A-2009), two containers (of approximately 2 liters) of recalcified plasma were collected at a US Food and Drug Administration (FDA) licensed blood bank from atopic donors who had given informed consent. Since clinically used IgE antibody assays have a reported working range from 0.1 to 100 ku A /L (kilounits per liter, where A represents the amount of allergen-specific IgE antibody in the blood), atopic donors were selected to achieve levels of IgE antibody to many aeroallergen and food-allergen specificities that fell below 30 ku A /L in an undiluted specimen. The rationale for this decision was to challenge assay methods and laboratories using specimens that contained IgE antibody levels that fell into the middle to lower working range (1 to 30 ku A /L) of their assays, where variation and linearity would be expected to be most problematic. The serum pools were diluted twice with a serum diluent containing a total IgE concentration of less than 5 ku/l (kilounits per liter). A 2.75-fold dilution factor was used to produce 3 nominal concentrations at 1.000, 0.364, and Two-milliliter aliquots were prepared, labeled, and sent in coded form (eg, CAP SE ) to approximately 200 clinical laboratories certified by Clinical Laboratory Improvement Amendments of 1988 that used 1 of 6 FDA-cleared total serum IgE assays and 1 of 3 FDA-cleared allergen-specific IgE assays. Each laboratory measured total serum IgE and IgE antibody to 5 allergen specificities by using their chosen assay system, and in a manner consistent with testing unknown clinical laboratory specimens. Total IgE and the 5 allergen-specific IgE antibody data for the 6 specimens in both cycles were recorded on a participant s report form and electronically returned to the CAP headquarters (Northfield, Illinois), where they were statistically analyzed to determine mean, median, and variance levels of total and allergen-specific IgE antibody for each assay. Data Analysis Collated laboratory data were subdivided at the CAP headquarters by assay method, analyte, and allergen specificity. Results were reported in tabulated format in the participant s summary report. From the total serum IgE data, an empirically derived dilution factor for each challenge specimen was computed for each assay. This was compared to the actual dilution factor used to prepare the specimens. The total IgE based dilution factors were used to evaluate the allergen-specific IgE assay data for linearity (parallelism) by using dilutionrecovery analysis as defined in appendix B of the Clinical Laboratory Standards Institute IgE antibody IL/A-20 Guideline. 13 Interdilution coefficients of variation were plotted as a function of reported IgE antibody concentration to create a precision profile to assess the linearity of the 3 IgE antibody assays. Data from the 3 IgE antibody assays were also compared by regression analysis and by a Bland-Altman Plot that directly compares differences among the allergen-specific IgE antibody assay results. Intramethod and interlaboratory variation was computed as an indicator of reproducibility of the assays between laboratories. Sensitivity was evaluated with data from select specimens that were diluted into the low working range of the IgE antibody assays between the current 0.1 ku A /L and the previous FDA-accepted minimum detectable dose of 0.35 ku A /L. RESULTS Total Serum IgE Table 1 presents the compiled median total serum IgE data in ku/l as reported by laboratories in the CAP s SE survey by using 6 FDA-cleared assays. Data are separated for each of the 2 specimens in each of the 2 survey cycles (2008 cycle C and 2009 cycle A) at 3 dilutions. Total serum IgE levels in the undiluted specimens ranged from approximately ku/l to ku/l (median 6 1 SD). As a measure of linearity, the intermethod coefficients of variation (CVs) for total IgE measurements across the clinically used assays ranged from 4.2% to 11.5%. These values were judged as excellent by using a consensus-derived linearity benchmark for ideal assays of CVs below 15%. 12 Figure 1 displays the tight agreement between the 6 clinically used total serum IgE assays. From the intermethod median total serum IgE data, dilution factors for each of the sets of specimens were computed (see entries set boldface in Table 1). The dilution factors from the measured total IgE data were and (mean 6 1 SD) for the first and second dilutions, respectively. The empirically determined dilution factors were comparable, within 1.1 SD of the expected 2.75 and 7.56 dilution targets that were actually used in preparing the specimens. Across the 6 assay methods, the computed dilution factor variances, expressed as CV, were 2.3% and 5.4%; these were considered excellent by the consensus-based target of 15% for interdilution coefficients of variation, which demonstrate internal assay parallelism (linearity). 12 Allergen-Specific IgE Table 2 lists the 16 allergen specificities that were assessed in the study. Dog dander, short ragweed, and Timothy grass, are common aeroallergen specificities, that were repeated in 2 assay cycles with different specimens to assess assay robustness with different allergen-specificity distributions of IgE antibody. Figure 2 displays the actual median allergen-specific IgE data (in ku A /L) for indoor aeroallergen and food (Figure 2, A), outdoor aeroallergen (Figure 2, B), and grass aeroallergen (Figure 2, C), as reported in the CAP survey for the 3 assay methods. Absolute levels of specific IgE antibody, as measured by different assays, agreed within 10% variance for only a small number of allergen specificities (eg, catspecific IgE [Phadia versus HYCOR] and cow s milk specific IgE [Siemens versus Phadia]); Figure 2, A). For most of the measurements, there were marked differences observed among the reported quantitative IgE antibody results from the 3 assays (Figure 2, A through C). 976 Arch Pathol Lab Med Vol 134, July 2010 IgE Antibody Assay Performance Hamilton

3 Table 1. Total Serum Immunoglobulin E (IgE) Levels in College of American Pathologists Dilution Analysis Study a Nephelometer, Siemens EIA, HYCOR IgE Concentration, ku/l ImmunoCAP, Phadia ADVIA Centaur, Siemens Immulite 2000, Siemens Immulite 2500, Siemens Intermethod Mean (SD) [IgE], ku/l CV % Dilution Factor Cycle C of 2008 (n = 7) (n = 8) (n = 119) (n = 14) (n = 43) (n = 10) (n = 6) C (20.8) 8.2 Undiluted C (7.3) C (1.7) C (38.6) 11.5 Undiluted C (8.3) C (3.3) Cycle A of 2009 (n = 5) (n = 8) (n = 134) (n = 12) (n = 44) (n = 21) (n = 6) A (11.3) 6.0 Undiluted A (3.2) A (1.4) A (21.9) 8.6 Undiluted A (6.6) A (1.6) Abbreviation: CV, coefficient of variation. a Reported median total serum IgE levels are presented in kilounits per liter (ku/l) for the 6 clinically used total serum IgE assays. Data are provided for 2 sets of specimens analyzed in the College of American Pathologists Diagnostic Allergy SE Proficiency Survey cycle C of 2008 (eg, C-11-08) and cycle A of 2009 (eg, A-01-09). The Immulite 2000 and 2500 are separated in the table because they are considered separate assay methods in the survey. The CVs between methods are excellent as judged by using a benchmark below 15%. Differences among IgE antibody assay results appeared to be patient/serum dependent (eg, dog- and short ragweed specific IgE results; Figure 2, A and B, respectively), indicating the importance of IgE antibody heterogeneity between subjects on intermethod agreement. Linearity Figure 3 presents the interdilution precision profile (CV versus dose) plot, which indicates that more than 90% of all interdilution coefficients of variation were below 15%. These data confirm that all 3 of the allergen-specific IgE antibody assays demonstrated acceptable linearity (parallelism) between their total serum IgE calibration curve and dilution of the test sera. Linearity is 1 critical property that confirms the quantitative nature of an IgE antibody assay. The correlation plot in Figure 4 compares the IgE antibody levels generated by the ImmunoCAP (comparator, x-axis) with those obtained with Immulite and HYTEC 288 IgE (on the y-axis). The Immulite regression line indicates a trend toward higher estimates of IgE antibody as compared to those of the ImmunoCAP, at IgE antibody levels above 1 ku A /L. The HYTEC 288 regression line indicates a trend toward lower estimates of IgE antibody levels as compared to those of the ImmunoCAP, at IgE antibody levels below 1 ku A /L. Figure 5 presents a Bland-Altman Plot (difference plot) that evaluates the agreement between the 3 IgE antibody assays for the clinically important lower working range from 0.1 to 18 ku A /L. Eighty percent of the IgE antibody results produced by the Immulite and HYTEC 288 system differed from those of the ImmunoCAP by more than 20%, which is the accepted deviation for results that are similar between 2 assays. The data in Figure 5 support the conclusion that the 3 IgE antibody assays continue to measure different populations of IgE antibody in any given serum. COMMENT In recent years, allergen-specific IgE antibody measurements have taken on new meaning in their ability to predict whether an individual with a suspected food allergy to potent allergenic foods, such as chicken egg, cow s milk, peanut, and fish, will fail a double-blind placebo-controlled food challenge with a 95% probability To date, however, the reported IgE anti-food values leading to the generation of predictive cutpoints have been exclusively measured by using 1 of 3 clinically available IgE antibody assays, namely, the ImmunoCAP. Past studies that have examined IgE assay performance have continued to show that despite the fact that various assays report results in the same IgE antibody units (ku A /L), they measure different populations of IgE antibody Thus, published predictive IgE antibody levels generated with one assay are not transferrable to measured IgE antibody levels generated with a different assay. The current thought is that the allergen reagent in each assay varies substantially in its composition because most allergen specificities have multiple major allergens, and allergens used in the different assays are manufactured from different source materials. Recent advances in allergen characterization, consensus guidelines, critiques from proficiency surveys, and general market forces have contributed to IgE assay harmonization. These factors have fostered the impression that the reported differences in quantitative IgE antibody levels between assays have narrowed. The current study was undertaken to examine this hypothesis and to comprehensively study the extent of parallelism displayed by the 3 clinically used IgE antibody assays in North America. Data from 2 CAP Diagnostic Allergy Survey cycles were used to examine the overall agreement in reported total and allergen-specific IgE antibody levels among the 3 assays. In contrast to the 2007 study, the term accuracy is intentionally not used in this report to describe how well a particular IgE antibody assay performs in comparison to another. This is because there exists no absolute IgE antibody standard from which one can judge the correctness or true accuracy of any of the available IgE antibody assays. Rather, comparability of the IgE antibody Arch Pathol Lab Med Vol 134, July 2010 IgE Antibody Assay Performance Hamilton 977

4 Figure 1. Total serum immunoglobulin (IgE) levels are presented as measured by 6 US Food and Drug Administration (FDA) cleared assays in the C-2008 and A-2009 cycles of the Diagnostic Allergy SE Survey of the College of American Pathologists. Each specimen was diluted twice, approximately 2.75-fold from the previous dilution. Good interassay agreement and linearity were observed with interdilution coefficients of variation below 15%. results from each assay are discussed to address the hypothesis that technology advances and automation have led to improved harmonization among assays, such that they can measure the same population of IgE antibody in any given human serum. Specimens in the 2 SE survey cycles, which spanned a 4-month interval, were specifically prepared to study the extent of linearity that exists between the total IgE calibration curve Table 2. Allergen Specificities Evaluated Allergen Code Allergen Common Name Allergen Scientific Name M6 Alternaria alternata Alternaria alternata M3 Aspergillus fumigatus Aspergillus fumigatus G2 Bermuda grass Cynodon dactylon E1 Cat dander Felis domesticus F2 Cow s milk Bos domesticus D2 Dust mite (farinae) Dermatophagoides farinae D1 Dust mite (pteronyssinus) Dermatophagoides pteronyssinus E5 Dog dander a Canis domesticus W9 English plantain (ribwort) Plantago lanceolata T7 White oak tree Quercus alba G3 Orchard grass (cocksfoot) Dactylis glomerata F13 Peanut a Arachis hypogaea W1 Short (common) ragweed a Ambrosia elatior T3 Silver birch Betula verrucosa G6 Timothy grass a Phleum pratense I3 Yellow jacket venom Vespula vulgaris a Four allergen specificities (dog dander, peanut, short ragweed, and Timothy grass) were repeated twice in the study to examine interspecimen variability. and IgE antibody portion of the assay. Linearity was targeted as a performance parameter because it is one clear mark of a quantitative assay. The data in Figure 3 confirm excellent linearity, with interdilution coefficients of variation below 15% for all 3 assays. Data (not repeated herein) in the SE-2008 cycle C and SE-2009 cycle A participants summary reports also confirm excellent within-assay reproducibility, with coefficients of variation between laboratories that are below 15% for more than 90% of all quantitative IgE measurements performed in these 2 SE survey cycles. Thus, automation appears to have had a positive impact in securing both intramethod interlaboratory reproducibility and linearity of the 3 clinically used allergen-specific IgE assays. Unfortunately, automation cannot address the concern of assay variability that points to the heterogeneity of the allergen-containing assay reagent. Differences in the composition of the allergen-containing reagents of the same specificity between assays is believed to be the primary contributor to the differences that are observed in final reported allergen-specific IgE antibody levels between assay methods. The use of a common total IgE calibration system, traceable to the same World Health Organization 75/502 IgE standard, does not appear to resolve this problem of variability. In the 2007 critique, 12 the authors presented a logarithmic Bland-Altman plot that examined differences between the 3 assays by using a single specificity (peanut) and engineered humanized IgE antibodies, which may not mimic the heterogeneity of naturally occurring human IgE antibody responses. Moreover, their measurements were performed in a single 978 Arch Pathol Lab Med Vol 134, July 2010 IgE Antibody Assay Performance Hamilton

5 Figure 2. A, Indoor-specific and food-allergen specific IgE levels (in ku A /L), as measured by 3 US Food and Drug Administration (FDA)-cleared autoanalyzers in the C-2008 and A-2009 cycles of the Diagnostic Allergy SE Survey of the College of American Pathologists. Each specimen was diluted approximately 2.75-fold from the previous dilution. Good interassay agreement and linearity were observed, with interdilution coefficient of variation below 15%, as illustrated in the precision profile in Figure 3. B, Outdoor aeroallergen specific IgE levels, as measured by 3 FDA-cleared autoanalyzers in the C-2008 and A-2009 cycles of the Diagnostic Allergy SE Survey of the College of American Pathologists. Each specimen was diluted approximately 2.75-fold from the previous dilution. Good interassay agreement and linearity were observed, with interdilution coefficient of variation below 15%, as illustrated in the precision profile in Figure 3. C, Grass aeroallergen specific IgE levels, as measured by 3 FDA-cleared autoanalyzers in the SE C-2008 and SE A-2009 cycles of the Diagnostic Allergy SE Survey of the College of American Pathologists. Each specimen was diluted approximately 2.75-fold from the previous dilution. Good interassay agreement and linearity were observed, with interdilution coefficient of variation below 15%, as illustrated in the precision profile in Figure 3. Arch Pathol Lab Med Vol 134, July 2010 IgE Antibody Assay Performance Hamilton 979

6 Figure 3. To assess the quantitative nature of each allergen-specific immunoglobulin E (IgE) assay, the interdilution coefficient of variation was computed after each dilution was corrected for its appropriate dilution factor, as determined by the total serum IgE levels. All 3 assays demonstrate good parallelism, indicating that the specimens dilute out in parallel with the total IgE calibration curve. The occasional interdilution coefficient of variation above 15% occurred with all 3 assays when the IgE antibody level was low. laboratory for each method, which raised questions of possible systematic bias. 20 An analogous logarithmic Bland-Altman plot is presented in Figure 5 in the current study that uses data from 200 laboratories and 16 allergen specificities. It shows essentially no improvement in interassay agreement, with the same relative variable distributions seen in the 2007 study. More than 80% of the IgE antibody measurements were discordant between assays, by greater than 20%. These data provide no evidence that agreement among the absolute reported IgE antibody levels from the 3 IgE antibody assays has moved closer. These interassay differences in the reported IgE antibody levels could be platform related, but they are more probably a result of the heterogeneity of the allergen component in the different assays. This is suggested by the observation that sera from different patients allergic to dog and ragweed produce different rank order levels of reactivity to dog and ragweed allergens on the same autoanalyzer platform (Figure 2, A and B). One approach to possibly resolving this issue of allergen and immune response heterogeneity is the use of allergen components, such as the 103 individual native and recombinant allergenic molecules that are currently used in the immuno solid-phase allergen chip (ISAC) microarray IgE antibody assay. 10 For instance, by using the principal group 1 allergen from short ragweed (Amb a 1) instead of a crude short ragweed allergen extract as the assay reagent, all 3 assays should be able to measure a comparable population of Amb a 1 specific IgE antibodies in any patient s blood circulation. 13 However, even if this strategy were adopted, heterogeneity in the epitope specificity of the human IgE antibody response between patients may preclude the quantitative measurement of identical populations of IgE antibody. This is because different isoforms of a given allergen component have been adopted as reagents by different assay manufacturers because of patent constraints. Another difficulty with this strategy of allergen component use is the need to be comprehensive with regard to all the allergen components in a given allergen specificity. For instance, using Amb a 1 as the allergen reagent will cause a less plentiful allergen, such as Amb a 5, to be missed in ragweed-allergic patients who produce an exclusive IgE antibody response. Combinations of all the major allergenic components may partially resolve this concern, if they are known and available; however, this raises the issue of differentially detecting cross-reactive IgE antibody and generating IgE antibody results that are difficult to interpret with sera from multisensitized patients. This particular problem was observed when Cor a 1 the group 1 hazelnut allergen that is cross-reactive with Bet v 1 from birch pollen was added to the hazelnut reagent used in 1 of the clinical assays. Use of the Cor a 1 enriched reagent caused this assay to detect hazelnut-specific IgE in serum from individuals who were predominantly or exclusively sensitized to birch pollen. 21 On a positive note, the CAP survey data in the recent 2 years have confirmed that the 3 principal IgE antibody assays, in general, agree well with respect to their qualitative (positive or negative) results when the IgE antibody levels in the challenge specimens are clearly present (eg,.1 ku A /L). Since allergists often use the qualitative presence of IgE antibody to support their clinical history in making diagnostic decisions about patients with allergies, this bodes well for the general clinical utility of IgE antibody measurements generated by the 3 assays. The problem comes when quantitative data are needed to draw a conclusion about the relative probability of passing a challenge test involving a potentially life-threatening allergic response to a food or venom. Then, quantitative differences among the 3 assays becomes critical, and predictive quantitative IgE antibody levels generated with 1 assay cannot be used to make decisions when the measurement is done with a different IgE antibody assay. While the excellent agreement among the 6 clinically used total serum IgE assays has not been questioned in 980 Arch Pathol Lab Med Vol 134, July 2010 IgE Antibody Assay Performance Hamilton

7 Figure 4. Correlation plot of immunoglobulin E (IgE) antibody results generated by the ImmunoCAP (Phadia) (comparator), Immulite (Siemens Healthcare Diagnostics), and HY- TEC 288 (HYCOR-Agilent Technologies) IgE antibody assays. The Immulite regression line indicates a trend toward higher estimates of IgE antibody levels as compared to those of the ImmunoCAP, at IgE antibody levels above 1 ku A /L. The HYTEC 288 regression line indicates a trend toward lower estimates of IgE antibody levels when compared to those of the ImmunoCAP, at IgE antibody levels below 1 ku A /L. Figure 5. Bland-Altman plot (difference plot) evaluates the agreement between the 3 IgE antibody assays for the clinically important, lower working range from 0.1 to 18 ku A / L of allergen-specific IgE. The dashed lines indicate 620%, which is accepted deviation for results that are similar between 2 assays. recent years, the data in the current study support this conclusion. Table 1 and Figure 1 demonstrate excellent intermethod coefficients of variation, below 15%. Moreover, the total IgE data provided a useful independent confirmation of the dilution scheme that was used in the current study to prepare the specimens used in these 2 CAP SE survey cycles. The only major concern today with total IgE measurements is the increasingly widespread use of omalizumab (anti-human IgE; trade name, Xolair) as a therapeutic treatment for drug-resistant asthma. The presence of therapeutic levels of omalizumab in sera that are submitted for either total or allergen-specific IgE determinations will interfere in several of the clinically used IgE assays. 22 Thus, knowledge of omalizumab administration for any patient providing a serum for analysis of IgE antibody is important for proper method selection and final interpretation of the results. In conclusion, harmonization of different IgE antibody assay technologies has involved the use of a common calibration scheme traceable to a single IgE standard, certain adjustments to achieve comparable analytic sensitivities as low as 0.1 ku A /L, and the systematic removal of poorly documented allergen specificities. These improvements have enhanced the performance of current IgE antibody autoanalyzers in terms of their linearity and reproducibility (within-laboratory, between-laboratory, and between-dilution). However, intermethod agreement remains a continued concern that will be difficult to resolve as long as heterogeneous extract-based allergen reagents are used to detect the complex humoral (IgE) immune responses of patients with allergies. Expanded use of allergen components as reagents is one future direction that can help resolve this last significant analytic performance concern for clinically used IgE antibody assays. Arch Pathol Lab Med Vol 134, July 2010 IgE Antibody Assay Performance Hamilton 981

8 References 1. Bernstein IL, Li JT, Bernstein DI, et al. Allergy diagnostic testing: an updated practice parameter. Ann Allergy Asthma Immunol. 2008;100(3 suppl 3):S1 S Hamilton RG. Assessment of human allergic diseases. In: Rich RR, Fleisher TA, Shearer WT, Schroeder HW, Frew AJ, Weyand CM, eds. Clinical Immunology: Principles and Practice. 3rd ed. London, United Kingdom: Mosby, Elsevier; 2008: Hamilton RG, Saito H. IgE antibody concentration, specific activity, clonality, and affinity measures from future diagnostic confirmatory tests. J Allergy Clin Immunol. 2008;122(2): Garcia-Ara C, Boyano-Martinez T, Diaz-Pena M, Martin-Estaban M. Specific IgE levels in the diagnosis of immediate hypersensitivity to cow s milk in the infant. J Allergy Clin Immunol. 2001;107(1): Sampson HA. Utility of food-specific IgE concentrations in predicting symptomatic food allergy. J Allergy Clin Immunol. 2001;107(5): Simpson A, Soderstrom L, Ahlstedt S, Murray C, Woodcock A, Custovic A. IgE antibody quantification and the probability of wheeze in preschool children. J Allergy Clin Immunol. 2005;116(4): Pastorello EA, Incorvaia C, Ortolani C, et al. Studies on the relationship between the level of specific IgE antibody and the clinical expression of allergy I: definition of levels distinguishing patients with symptomatic from patients with asymptomatic allergy to common aeroallergens. J Allergy Clin Immunol. 1995; 96(5, pt 1): Wide L, Bennich H, Johansson SGO. Diagnosis by an in vitro test for allergen specific IgE antibodies. Lancet. 1967;2(7526): Hamilton RG. The science behind the discovery of IgE. J Allergy Clin. Immunol. 2005;115(3): Hiller R, Laffer S, Harwanegg C, et al. Microarrayed allergen molecules: diagnostic gatekeepers for allergy treatment. FASEB J. 2002;16(3): Diaz-Vazquez C, Torregrosa-Bertet MJ, Carvajal-Urueña I, et al. Accuracy of ImmunoCAPH Rapid in the diagnosis of allergic sensitization in children between 1 and 14 years with recurrent wheezing: the IReNE study. Pediatr Allergy Immunol. 2009;20(6) Wood RA, Segall N, Ahlstedt S, Williams PB. Accuracy of IgE antibody laboratory results. Ann Allergy Asthma Immunol. 2007;99(1): Matsson P, Hamilton RG, Esch RE, et al. Analytical Performance Characteristics and Clinical Utility of Immunological Assays for Human Immunoglobulin E (IgE) Antibody of Defined Allergen Specificities: Approved Guideline. 2nd ed. Wayne, PA: Clinical Laboratory Standards Institute 1/LA20- A2; vol 29, no Sampson HA, Ho DG. Relationship between food-specific IgE concentrations and the risk of positive food challenges in children and adolescents. J Allergy Clin Immunol. 1997;100(4): Garcia-Ara MC, Boyano-Martinez T, Diaz-Pena JM, Martin-Muñoz F, Martin-Esteban M. Cow s milk-specific immunoglobulin E levels as predictors of clinical reactivity in the follow-up of the cow s milk allergy infants. Clin Exp Allergy. 2004;34(6): Celik-Bilgili S, Mehl A, Verstege A, et al. The predictive value of specific immunoglobulin E levels in serum for the outcome of oral food challenges. Clin Exp Allergy. 2005;35(3): van der Gugten AC, den Otter M, Meijer Y, Pasmans SG, Knulst AC, Hoekstra MO. Usefulness of specific IgE levels in predicting cow s milk allergy. J Allergy Clin Immunol. 2008;121(2): Perry TT, Matsui EC, Conover-Walker MK, Wood RA. The relationship of allergen-specific IgE levels and oral food challenge outcome. J Allergy Clin Immunol. 2004;114(1): Komata T, Söderström L, Borres MP, Tachimoto H, Ebisawa M. The predictive relationship of food-specific IgE concentrations to challenge outcomes for egg and milk varies by patient age. J Allergy Clin Immunol. 2007;119(5): Hovanec-Burns D. Accuracy of IgE antibody laboratory results. Ann Allergy Asthma Immunol. 2008;100(2): Sicherer SH, Dhillon G, Laughery KA, Hamilton RG, Wood RA. Caution: the Phadia hazelnut ImmunoCAP (f17) has been supplemented with recombinant Cor a 1 and now detects Bet v 1-specific IgE, which leads to elevated values for persons with birch pollen allergy. J Allergy Clin Immunol. 2008;122(2): Hamilton RG. Accuracy of Food and Drug Administration-cleared IgE antibody assays in the presence of anti-ige (omalizumab). J Allergy Clin Immunol. 2006;117(4): Arch Pathol Lab Med Vol 134, July 2010 IgE Antibody Assay Performance Hamilton