Development and Validation of an Isoform Independent Monoclonal Antibody-Based ELISA for Measurement of Lipoprotein(a)

INTRODUCTIONLipoprotein(a) [Lp(a)] is an apo B-containing lipoprotein that is a genetic, independent risk factor for cardiovascular disease and aortic stenosis (A test in context: Lipoprotein(a): diagnosis, prognosis, controversies, and emerging therapies.). Lp(a) is similar in composition to LDL but additionally characterized by the presence of a carbohydrate-rich protein termed apolipoprotein(a) [apo(a)] that is covalently linked by a disulfide bond to the single molecule of apoB-100. The presence of apo(a) imparts specific pathophysiological and metabolic characteristics to Lp(a) rendering it significantly different from LDL (Schmidt K. Noureen A. Kronenberg F. Utermann G. Structure, function, and genetics of lipoprotein(a).).Apo(a) is formed by repeated kringle (K) structures (KIV), a single copy of KV, and an inactive protease domain, all possessing a high amino acid sequence homology with the corresponding structure of plasminogen. In apo(a), KIV is formed by 10 different subtypes (KIV1 to KIV10) each present as a single copy with the exception of KIV2 which is present in a highly variable number in different individuals ranging from 1 to >40 copies of identical repeats. The repeats are due to copy number variations in the LPA gene and therefore, individuals may inherit highly different apo(a) molecular weight ranging from ∼300 to 800 kDa. The variation in the number of KIV2 gives origin to the >40 apo(a) isoforms circulating in human plasma of different individuals and is primarily responsible for the size heterogeneity of Lp(a). The concentration of Lp(a) is also highly heterogeneous, varying >1000 fold within the population, and to a major extent is genetically controlled and inversely related to the copy number variation in the LPA gene (Schmidt K. Noureen A. Kronenberg F. Utermann G. Structure, function, and genetics of lipoprotein(a).).The large size heterogeneity of apo(a) has been a major challenge to the immunochemical measurement of Lp(a) because the variable number of repeated KIV2 motifs results in a variable number of antigenic epitopes in the samples to be analyzed. Consequently, plasma levels of Lp(a) will be overestimated or underestimated in test samples when the number of KIV2 is higher or smaller than those present in the assay calibrator (Marcovina S.M. Albers J.J. Lipoprotein (a) measurements for clinical application.).Ideally, use of monoclonal antibodies directed to a single antigenic site on apo(a) will be able to solve the impact of the variable number of KIV2. However, the high homology of the apo(a) kringles (∼75 – 94%) (McLean J.W. Tomlinson J.E. Kuang W.J. Eaton D.L. Chen E.Y. Fless G.M. Scanu A.M. Lawn R.M. cDNA sequence of human apolipoprotein(a) is homologous to plasminogen.), has proven to be highly challenging in developing monoclonal antibodies binding to a unique epitope not present in KIV2.In 1995, Marcovina et al. described the production of a monoclonal antibody (a-40) directed to a unique epitope located in KIV9 of apo(a) and its use as a detecting antibody in the development of an enzyme-linked immunoassay (ELISA) demonstrated to accurately measure Lp(a) without the impact of the apo(a) size polymorphism in the samples (Marcovina S.M. Albers J.J. Gabel B. Koschinsky M.L. Gaur V.P. Effect of the number of apolipoprotein(a) kringle 4 domains on immunochemical measurements of lipoprotein(a).). Because this ELISA does not measure the variable mass of Lp(a) but the number of circulating particles, the Lp(a) concentration is expressed in nmol/L. However, while this ELISA has been extensively used in research and for assay standardization as a “gold standard”, the assay has never been made available outside of the University of Washington.A monoclonal antibody (LPA-KIV9), also directed to KIV9, was recently generated at the University of California San Diego and extensively evaluated as previously reported (Gonen A. Yang X. Yeang C. Alekseeva E. Koschinsky M. Witztum J.L. Boffa M. Tsimikas S. Generation and characterization of LPA-KIV9, a murine monoclonal antibody binding a single site on apolipoprotein (a).). The aim of the current study was to develop a new sandwich Lp(a) ELISA modeled on the approach used by Marcovina et al. (Marcovina S.M. Albers J.J. Gabel B. Koschinsky M.L. Gaur V.P. Effect of the number of apolipoprotein(a) kringle 4 domains on immunochemical measurements of lipoprotein(a).) and based on two monoclonal antibodies LPA-KIV9 (Gonen A. Yang X. Yeang C. Alekseeva E. Koschinsky M. Witztum J.L. Boffa M. Tsimikas S. Generation and characterization of LPA-KIV9, a murine monoclonal antibody binding a single site on apolipoprotein (a).) and LPA4 (Tsimikas S. Lau H.K. Han K.R. Shortal B. Miller E.R. Segev A. Curtiss L.K. Witztum J.L. Strauss B.H. Percutaneous coronary intervention results in acute increases in oxidized phospholipids and lipoprotein(a): short-term and long-term immunologic responses to oxidized low-density lipoprotein.). To demonstrate its performance characteristics, we report here the extensive validation of this assay.MATERIALS and METHODSAntibodiesThe generation and characterization of murine IgG monoclonal antibodies (MAB) LPA4 and LPA-KIV9 were previously described (Tsimikas S. Lau H.K. Han K.R. Shortal B. Miller E.R. Segev A. Curtiss L.K. Witztum J.L. Strauss B.H. Percutaneous coronary intervention results in acute increases in oxidized phospholipids and lipoprotein(a): short-term and long-term immunologic responses to oxidized low-density lipoprotein.). Briefly, LPA4 was generated by immunizing mice with the 14-amino acid peptide TRNYCRNPDAEIRP present on apo(a) KIV5, KIV7 and KIV8. However, the partial sequence NYCRNPDA is also present on KIV2 and appears to be immunologically dominant as LPA4 strongly interacts with KIV2 (Tsimikas S. Fazio S. Viney N.J. Xia S. Witztum J.L. Marcovina S.M. Relationship of lipoprotein(a) molar concentrations and mass according to lipoprotein(a) thresholds and apolipoprotein(a) isoform size.). As previously described, MAB LPA-KIV9 was generated by immunizing mice with a truncated recombinant apo(a) formed by 8 kringle KIV repeats containing one copy of KIV1, one copy of KIV2, a fusion of KIV3 and KIV5, individual KIV6 to KIV10 , KV, and the protease domain (Gonen A. Yang X. Yeang C. Alekseeva E. Koschinsky M. Witztum J.L. Boffa M. Tsimikas S. Generation and characterization of LPA-KIV9, a murine monoclonal antibody binding a single site on apolipoprotein (a)., Schneider M. Witztum J.L. Young S.G. Ludwig E.H. Miller E.R. Tsimikas S. Curtiss L.K. Marcovina S.M. Taylor J.M. Lawn R.M. Innerarity T.L. Pitas R.E. High-level lipoprotein [a] expression in transgenic mice: evidence for oxidized phospholipids in lipoprotein [a] but not in low density lipoproteins.). As reported, LPA-KIV9 has been shown to bind to sequence 4076LETPTVV4082 on KIV9, which is present only once on apo(a) (Gonen A. Yang X. Yeang C. Alekseeva E. Koschinsky M. Witztum J.L. Boffa M. Tsimikas S. Generation and characterization of LPA-KIV9, a murine monoclonal antibody binding a single site on apolipoprotein (a).).Double-Antibody Sandwich ELISAThe development of the ELISA method and the validation studies were performed at Medpace Reference Laboratories Cincinnati facility. This sandwich ELISA is based on the use of LPA4 as the capture antibody and the use of LPA-KIV9 as the detection antibody (Figure 1). Based on the LETPTVV epitope specificity of LPA-KIV9, this assay is not expected to be affected by the size variation of apo(a) in Lp(a) samples.Figure thumbnail gr1

Figure 1Methodology of the University of Washington and the UCSD ELISAs. The principle of both these sandwich ELISAs is the capture of Lp(a) with monoclonal antibodies that bind to KIV2 and detection antibodies that bind only once to KIV9.

Initial attempts to directly bind LPA4 to 96-well plates (MaxiSorp, Thermo Fisher) were not successful and therefore, LPA4 was biotinylated, and streptavidin coated plates were used. Ninety-six-well microplates pre-coated with streptavidin (Pierce® Streptavidin High Binding Capacity Coated 96-Well Plates, Thermo Fisher) and biotinylated LPA4 (Biotinylation Kit / Biotin Conjugation Kit (Fast, Type B) - Lightning-Link®, Abcam) (100 μl at 2 μg/ml) was added to the plates in a saturating amount and incubated for 2 hours in a shaker at 300 rpm at room temperature. The WHO/IFCC 1st Reference Material for Lp(a) Immunoassays, SRM-2B, with an assigned value of 107 nmol/L was used as the assay calibrator (Dati F. Tate J.R. Marcovina S.M. Steinmetz A. C. International Federation of Clinical, M. Laboratory, and I. W. G. f. L. A. Standardization
First WHO/IFCC Iinternational reference reagent for lipoprotein(a) for immunoassay--Lp(a) SRM 2B.). After reconstitution of the lyophilized reference material, a stock solution was prepared by diluting the material with dilution buffer to a concentration of 4.38 nmol/L. The 8-point standard curve was then prepared by diluting the stock solution x10, x12, x16, x20, x30, x40, x80 and x160 to yield solutions of 0.438, 0.365, 0.274, 0.219, 0.146, 0.110, 0.055 and 0.027 nmol/L, respectively.

After washing and blocking of non-specific sites with 3% BSA, the diluted standards and test samples diluted at 1:400 were added and incubated in the dark for 1 hour at 28°C with shaking at 1000 rpm. To increase the precision of the measurements, selected samples, based on their relative Lp(a) concentrations, were re-analyzed at optimal dilutions (1:10 to 1:3200) so that all the optical densities (OD) are confined in the middle and most stable part of the standard curve. The Lp(a) particles present in the standards and in the test samples are bound to the immobilized LPA4 and then detected with HRP-LPA-KIV9. The addition of a chromogenic substrate (OPD Substrate Tablets [o-phenylenediamine dihydrochloride] with 7 μl hydrogen peroxide, Sigma) allows for quantitative determination of HRP activity that is proportional to the amount of captured Lp(a). The plates were read on a Versa max microplate reader (Molecular Devices, San Jose, CA) at a wavelength of 490 nm with background subtraction at 630 nm. The Lp(a) concentrations are reported in nmol/L.

Assay ValidationThe Lp(a) ELISA validation follows the guidelines of the Clinical Laboratory Improvement Amendments (CLIA) and the College of American Pathologists (CAP). Method performance characteristics assessed during validation were following the subsections of CLIA regulation 42CFR493.1253. Acceptability of results throughout the validation were based on the recommendations of the FDA 2018 Bioanalytical Method validation Guidance for ligand binding assays as well as on the concept of Total Allowable Error (TAE) established using the Westgard Desirable Biological Variation Database (Schneider F. Maurer C. Friedberg R.C. International Organization for Standardization (ISO) 15189.) specifications for imprecision (10.4%) and bias (6.9%).Repeatability (within-run imprecision)

Three levels of in-house prepared quality control (QC) pools were analyzed in 22 replicates in a single plate. QC Low, QC Medium, and QC High were diluted before analysis x50, x400, and x1600, respectively. The replicates were obtained from 11 different dilutions. The repeatability was acceptable if the CV for each level QC was <10%.

Reproducibility (total imprecision)

The same QC samples used in the repeatability testing were analyzed in duplicate at the beginning and end of every plate for 8 days. The assay reproducibility was deemed acceptable if the calculated CV was <15.0%.

Analytical Measuring Range

The analytical measuring range (AMR), or reportable range of the assay, is the range of analyte concentrations with which the assay can provide measured quantity values that are directly proportional to the value of the measurand in the samples. The AMR was assessed by determining the accuracy and precision throughout the range of the standard curve from 0.027 nmol/L to 0.438 nmol/L. Results were evaluated to determine the Lower Limit of Quantification (LLOQ), Limit of Detection (LOD), and Upper Limit of Quantification (ULOQ) of the assay.

Linearity

Independent dilutions of the assay calibrator were prepared and each of the 8 solutions was analyzed in duplicate as an unknown against the calibration curve for 5 different times over 4 different days. Response values were back calculated, and interpolated concentrations were determined. The mean SD, %CV, bias, and %bias from the nominal concentration were calculated from the results of each standard solution.

Limit of Quantitation

Testing for the low limit of quantification (LLOQ) and upper limit of quantification (ULOQ) was incorporated into the linearity step by evaluating the performance of the assay at the limits of the analytical measuring range.

Spiking and Recovery

Spike and recovery experiments were performed to identify the systemic error that may arise from the interaction between other matrix components and Lp(a).

The concentration of Lp(a) in five serum samples from individual donors was determined before (baseline) and after spiking with two different levels of the WHO/IFCC Reference Material with the assigned value of 107 nmol/L (Dati F. Tate J.R. Marcovina S.M. Steinmetz A. C. International Federation of Clinical, M. Laboratory, and I. W. G. f. L. A. Standardization
First WHO/IFCC Iinternational reference reagent for lipoprotein(a) for immunoassay--Lp(a) SRM 2B.). The ratio of spike volume to total sample volume was 1:10 (1 part Lp(a) control material to 9 parts serum) and 1:20 (1 part Lp(a) control material to 19 parts serum). Samples were spiked after the appropriate dilution of all specimens. Each sample was run in duplicate, and the mean, SD, %CV, bias, and %bias were calculated. The recovery was deemed acceptable if the obtained mean %bias was within ± 20.0% of the baseline concentration (i.e., recovery of 80.0% - 120.0%).Dilutability

Dilutability experiments were performed to confirm compatibility between the sample diluent and the matrix. The dilutability was performed using five serum samples with known concentration of Lp(a) ranging from 3.2 nmol/L to 343 nmol/L. Multiple dilutions of each sample were performed for their values to fall in the range of the standard curve. Each sample dilution was analyzed four times, results corrected for the dilution, and the mean, SD, %CV, bias, and %bias were calculated. Results were acceptable if the recovery of each measured concentration of the diluted samples, after correcting for the dilution, was between 80% and 120%.

Matrix correlation (Serum versus Plasma)

Serum and K2EDTA-plasma samples were collected from 38 volunteer individuals and analyzed in parallel. The two matrixes were considered equivalent if no statistically significant bias was observed. If observed, the bias was considered not clinically significant if it was <15%. Lithium-heparin plasma was not investigated.

AccuracyIn the absence of a reference method for Lp(a), and following the recommendations of the Joint Committee for Traceability (JTLM) and the International Organization for Standardization (Schneider F. Maurer C. Friedberg R.C. International Organization for Standardization (ISO) 15189.,

International Organization for Standardization ISO 17511:2020. In vitro diagnostic medical devices – Requirements for establishing metrological traceability of values assigned to calibrators, trueness quality control materials and human samples. Available from: https//www.iso.org/standard/69984.html.

, Armbruster D. Miller R.R. The Joint Committee for Traceability in Laboratory Medicine (JCTLM): a global approach to promote the standardisation of clinical laboratory test results.), the accuracy of the method was evaluated by comparison of results obtained by the LPA4/LPA-KIV9 ELISA with those obtained by the monoclonal antibody-based ELISA developed by Marcovina et al. at the University of Washington, using capture monoclonal antibody a-6 and detection antibody a-40 (Marcovina S.M. Albers J.J. Gabel B. Koschinsky M.L. Gaur V.P. Effect of the number of apolipoprotein(a) kringle 4 domains on immunochemical measurements of lipoprotein(a).). This method, demonstrated to measure apo(a) and apoB-100 in Lp(a) on equimolar basis and therefore, unaffected by the apo(a) size polymorphism, is considered the “gold standard” method for Lp(a) measurement (Marcovina S.M. Albers J.J. Lipoprotein (a) measurements for clinical application.). Additionally, the results of the LPA4/LPA-KIV9 ELISA were compared to those obtained by a targeted liquid chromatography tandem mass spectrometry (LC-MS/MS) candidate reference method for Lp(a) (Marcovina S.M. Clouet-Foraison N. Koschinsky M.L. Lowenthal M.S. Orquillas A. Boffa M.B. Hoofnagle A.N. Vaisar T. Development of an LC-MS/MS Proposed Candidate Reference Method for the Standardization of Analytical Methods to Measure Lipoprotein(a).). This LC-MS/MS method was rendered independent from the size polymorphism of Lp(a) by selecting three specific quantification peptides not present in the KIV2 region of apo(a). Accuracy of the measurements were achieved by using a high-purity 14 KIV recombinant apo(a) as a primary reference material to calibrate the assay in SI units.The evaluation was performed on a set of 64 fresh-frozen samples from individual donors that were previously tested at the University of Washington by the Lp(a) ELISA and by the LC-MS/MS method. Analyses of the 64 samples, spanning a good range of Lp(a) levels and apo(a) isoform size, were performed by the LPA4/LPA-KIV9 ELISA in duplicate over three different days. The methods were considered equivalent if no statistically significant bias was observed. If observed, the bias was considered not clinically significant if it was Armbruster D. Miller R.R. The Joint Committee for Traceability in Laboratory Medicine (JCTLM): a global approach to promote the standardisation of clinical laboratory test results.) with each isoform designated by the relative number of KIV.Determination of the Peptide Epitope of Monoclonal Antibody a-40A peptide library array spanning KIV9, consisting of 100 overlapping, 15 amino acid peptides differing by only one amino acid residue was synthesized on cellulose membranes (JPT, Germany). The membrane was incubated with monoclonal antibody a-40 (1 μg/ml, overnight at 4°C) and antibody binding detected with an anti-mouse IgG antibody conjugated with horseradish peroxidase and SuperSignal West Dura HRP chemiluminescent substrate (Thermo Scientific), as previously described (Gonen A. Yang X. Yeang C. Alekseeva E. Koschinsky M. Witztum J.L. Boffa M. Tsimikas S. Generation and characterization of LPA-KIV9, a murine monoclonal antibody binding a single site on apolipoprotein (a).).

To further evaluate whether monoclonal antibodies a-40 and LPA-KIV9 react with the same or different epitopes on KIV9, a competition experiment was performed with a modified version of the LPA4/LPA-KIV9 assay. In brief, LPA4 was plated overnight in microtiter well plates at 5 ug/ml. A plasma sample with a previously determined Lp(a) value of 170 nmol/L was used and added to the well at 1:400 dilution. Biotin-labeled LPA-KIV9 was then added at 1 μg/ml to bind Lp(a). Plates were washed three times, incubated with alkaline phosphatase-conjugated NeutrAvidin (Thermo Scientific, 1:40,000 dilution) for 60 min at room temperature, washed, and incubated with Lumi-Phos-530 (Lumigen Inc., Southfield, Michigan 1:1 dilution in water) for 75 min at room temperature (25 ml/well) and luminescence measured (BioTek Instruments, Inc, Winooski, VT). Results were displayed as relative light units (RLU) per 100 milliseconds. The competition assay was performed by co-incubating the diluted plasma sample with 1, 5 and 10 μg/ml of a-40. These experiments were performed at UCSD.

StatisticsSoftMax Pro 7.1 GxP software was used to analyze microplate data generated from the Versa max microplate reader. Microsoft® Office Excel and a statistical software package Analyse-It® (version 3.90.7) were used for validation and statistical calculations. The mean values of the assay comparison were analyzed using Deming Regression and Bland Altman (Statistical methods for assessing agreement between two methods of clinical measurement.) plots to determine bias. Linearity was determined to evaluate the statistical significance of second and third order polynomials in the fit between the expected, theoretical concentration and the observed concentration. If second or third order polynomials significantly improved the fit at p = 0.05, the deviation from linearity was unacceptable if it was greater than 15.0% (25.0% at LLOQ and ULOQ).DISCUSSIONThis study describes the development and validation of a new, isoform-independent ELISA to measure Lp(a) in plasma and serum. The LPA4/LPA-KIV9 ELISA was shown to be highly sensitive and linear to a broad range of Lp(a) concentrations from as low as 0.27 nmol/L to as high as 1402 nmol/L and to not be affected by the size polymorphism of apo(a). It was further shown to have no statistically or clinically significant bias relative to both the “gold standard” ELISA (Marcovina S.M. Albers J.J. Gabel B. Koschinsky M.L. Gaur V.P. Effect of the number of apolipoprotein(a) kringle 4 domains on immunochemical measurements of lipoprotein(a).) or the LC-MS/MS candidate reference method (Marcovina S.M. Clouet-Foraison N. Koschinsky M.L. Lowenthal M.S. Orquillas A. Boffa M.B. Hoofnagle A.N. Vaisar T. Development of an LC-MS/MS Proposed Candidate Reference Method for the Standardization of Analytical Methods to Measure Lipoprotein(a).) and therefore, the three methods are considered equivalent.The development of the LPA4/LPA-KIV9 ELISA was modeled a priori after the well-validated a-6/a-40 ELISA which is the only Lp(a) method validated to be isoform independent (Marcovina S.M. Albers J.J. Gabel B. Koschinsky M.L. Gaur V.P. Effect of the number of apolipoprotein(a) kringle 4 domains on immunochemical measurements of lipoprotein(a).). Interestingly, we have demonstrated that LPA-KIV9 and a-40 both bind to the same epitope LETPTVV on KIV9 of apo(a). KIV9 contains 114 amino acids (including inter-kringle regions), but it is highly homologous to all other kringles, except at amino acids 89-114 where modest variability exists allowing for the generation of monoclonal antibodies (McLean J.W. Tomlinson J.E. Kuang W.J. Eaton D.L. Chen E.Y. Fless G.M. Scanu A.M. Lawn R.M. cDNA sequence of human apolipoprotein(a) is homologous to plasminogen.). Despite the fact that the detecting monoclonal antibodies used in the two ELISA methods are both directed to the same epitope in KIV9, a higher than expected bias was observed in some of the samples. In addition to possible differences in affinity, one possible explanation may relate to the fact that a-40 was generated using native, purified Lp(a) as antigen (Marcovina S.M. Albers J.J. Gabel B. Koschinsky M.L. Gaur V.P. Effect of the number of apolipoprotein(a) kringle 4 domains on immunochemical measurements of lipoprotein(a).), whereas LPA-KIV9 was generated using a truncated apo(a) recombinant construct (Gonen A. Yang X. Yeang C. Alekseeva E. Koschinsky M. Witztum J.L. Boffa M. Tsimikas S. Generation and characterization of LPA-KIV9, a murine monoclonal antibody binding a single site on apolipoprotein (a).) and therefore, it may be more susceptible to conformational changes. Furthermore, the two antibodies are expected to have different sequences at the variable region as they were generated by independent processes. However, the 2.7% bias observed between Lp(a) measurements obtained by the two ELISAs was not statistically or clinically significant and therefore, the two methods are considered equivalent. Additionally, the apo(a) Kringle IV number only accounts for 0.2% of the bias variation, confirming that the LPA-KIV9 ELISA does not appear to be affected by the size variability of apo(a). Likewise, no statistically or clinically significant bias nor apo(a) isoform bias were observed when results were compared with those obtained by the LC-MS/MS method despite the profound differences in the principle as well as assay calibration between the two methods.It is anticipated that the determination of Lp(a) values will continue to expand with at least 7 guidelines now advocating testing of Lp(a) in moderate to high-risk individuals (Tsimikas S. Stroes E.S.G. The dedicated "Lp(a) clinic": A concept whose time has arrived?.). The European and Canadian guidelines have further recommended every adult person to have Lp(a) measured at least once in their lifetime (Mach F. Baigent C. Catapano A.L. Koskinas K.C. Casula M. Badimon L. Chapman M.J. De Backer G.G. Delgado V. Ference B.A. Graham I.M. Halliday A. Landmesser U. Mihaylova B. Pedersen T.R. Riccardi G. Richter D.J. Sabatine M.S. Taskinen M.R. Tokgozoglu L. Wiklund O. Group E.S.C.S.D. 2019 ESC/EAS Guidelines for the management of dyslipidaemias: lipid modification to reduce cardiovascular risk., Pearson G.J. Thanassoulis G. Anderson T.J. Barry A.R. Couture P. Dayan N. Francis G.A. Genest J. Gregoire J. Grover S.A. Gupta M. Hegele R.A. Lau D. Leiter L.A. Leung A.A. Lonn E. Mancini G.B.J. Manjoo P. McPherson R. Ngui D. Piche M.E. Poirier P. Sievenpiper J. Stone J. Ward R. Wray W. 2021 Canadian Cardiovascular Society Guidelines for the Management of Dyslipidemia for the Prevention of Cardiovascular Disease in Adults.). In clinical laboratories, analyses of Lp(a) continue to be performed by immunoassays using polyclonal antibodies against apo(a), which may bind throughout the apo(a) protein, including the variable number of the identical copies of KIV2. These assays have the major limitation of over- or under-estimating Lp(a) values depending on the size of apo(a) in the test samples relative to the assay calibrators. More immunocomplexes are formed between the polyclonal antibodies and the larger apo(a) isoforms, which are generally found in subjects with low plasma Lp(a) and thus the low values tend to be overestimated. In contrast less immunocomplexes are formed with smaller isoforms, which are generally found in subjects with high plasma Lp(a) and thus the high values tend to be underestimated (Marcovina S.M. Albers J.J. Lipoprotein (a) measurements for clinical application., Tsimikas S. Fazio S. Viney N.J. Xia S. Witztum J.L. Marcovina S.M. Relationship of lipoprotein(a) molar concentrations and mass according to lipoprotein(a) thresholds and apolipoprotein(a) isoform size.). Although optimizing the assay calibration can minimize isoform bias as demonstrated by the Denka Seiken assay (Marcovina S.M. Albers J.J. Lipoprotein (a) measurements for clinical application.), which is considered the least isoform dependent method, none of the commercially available assays are truly isoform-independent.A recent study in 2020 (Scharnagl H. Stojakovic T. Dieplinger B. Dieplinger H. Erhart G. Kostner G.M. Herrmann M. Marz W. Grammer T.B. Comparison of lipoprotein (a) serum concentrations measured by six commercially available immunoassays.) evaluated Lp(a) levels measured in the same samples by 6 commercially available assays showing that most bivariate correlation coefficients were greater than 0.90. However, compared to the WHO/IFCC reference material, the results of the different assays diverged from the target values by range of −8% to +22% in a concentration dependent manner. The authors concluded that current immunological assays biases differed significantly across the clinically relevant concentration ranges in a non-linear manner not entirely dependent on apo(a) phenotypes. Another study in 2021 from a large referral laboratory (ARUP, Salt Lake City, UT) measured Lp(a) with five commercially available assays showing significant variations when comparing results to an all-method average (Performance evaluation of five lipoprotein(a) immunoassays on the Roche cobas c501 chemistry analyzer.). The main limitation of both study designs is that the comparison of results was performed using Lp(a) values obtained by methods calibrated either in mg/dL of total Lp(a) mass or in nmol/L. However, both studies confirm the need of standardization of Lp(a) measurements, the traceability of the calibrators to a common reference material, and the expression of Lp(a) values in molar units.Currently different societies hav

留言 (0)

沒有登入
gif