Scientific Publications

Fred S Apple, Lyn M Steffan, Lesly A Pearce, MaryAnn M Murakami, Russell V Luepker

American Heart Association Scientific Sessions 2010

Anju Bhardwaj, W. Frank Peacock, Kiang-Teck J. Yeo, Alan Storrow, Quyhn Truong, Stephen Thomas, Kevin M. Curtis, Robert S. Foote, Hong K. Lee, Karen F. Miller, James L. Januzzi, Jr.

American Heart Association Scientific Sessions 2010

Joel Estis, Jenny Hsu, Quynh Anh Lu, Lynn Zieske, John Todd, Sara J. Agee. American Association of Cancer Research - Translational Cancer Medicine

Purpose: The MAPK and PI3K signaling cascades are critical pathways in malignant transformation, tumor progression and multi-drug resistance (MDR) of human cancers. Elucidation of the relative contributions of distinct kinases is made difficult due to many factors, including prevalence of multiple isoforms, post-translational modifications, synergistic activation and pathway cross-talk. Clinical investigations into the role of phosphorylation of specific kinases are further hindered by the lack of ultra-sensitive assay technology with requisite isoform- and phophorylation-state specificity. Improved assays are necessary to determine clinical relevance and potential diagnostic utility with a high degree of specificity.

Methods: Adjacent and primary normal breast cancer tissue lysates were analyzed using a novel panel of ten immunoassays for quantifying the total and phosphorylated forms of AKT1, GSK3β, JNK1, JNK2 and ERK2. The ultra-sensitive assay panel was developed for use with the Erenna® Immunoassay System (Singulex), which uses single molecule detection technology. Analytical sensitivity, inter-assay precision, linearity and specificity were determined and compared to corresponding ELISA assays. The novel assay panel was validated in HeLa and Jurkat cell lysates, region of linearity was determined, and phospho-specificity of each assay was confirmed. All assays were validated for use in human plasma, cell culture lysates, and human tissue specimens.

Results: Testing in matched breast cancer tissue lysates showed that all total kinase concentrations (tAKT1, tGSK3β, tERK2, tJNK1 and tJNK2) were found to be elevated in primary tumors compared to surrounding adjacent normal tissue, however concentrations of phosphorylated forms (pAKT1, pGSK3β, pERK2, pJNK1 and pJNK2) were not. All ten novel assays displayed exceptional linearity (R2=0.99) and precision (%CV 4.1 – 9.6), and were significantly more sensitive than the corresponding ELISA assay method based upon sample endpoint dilutions. The detection limits of the tAKT1, pAKT1, tGSK3β, pGSK3β, pERK2, tJNK2 and pJNK2 assays were all ≤0.3 pg/mL, while the detection limit of tERK2 was < 2pg/mL, tJNK1 was 4.5 pg/mL, and pJNK1 was 7 pg/mL. The LLoQ of each assay in the panel was sufficient for quantification from a minimum of approximately 200 cells, with exception of the total JNK1 assay, which required a minimum of approximately 800 cells in order to be quantifiable.

Conclusions: We show that significant elevations in total kinase concentrations of tAKT1, tGSK3β, tERK2, tJNK1 and tJNK2 can be quantified in primary tumors compared to surrounding adjacent tissue using a panel of isoform- and phosphorylation-specific single molecule immunoassays. We have further demonstrated that assays in this panel can quantify analytes from as little as 100 cells, and exhibit superior analytical performance over ELISA based methods. Thus, these novel assays greatly improve upon pre-existing ELISA technology that is limited by assay sensitivity, providing a new tool for clinical investigations of the MAPK and PI3K signaling cascades.

Walter E. Kelley, Christopher deFilippi, Show-Hong Duh, Mary Lou Gantzer, John Todd, Robert H. Christenson.AACC 2009

Objective: We compared the agreement (harmony) of cardiac troponin I (cTnI) measurements for six commercially available assays using samples from suspected acute coronary syndrome (ACS) patients with values near the 99th percentile of the normal reference population.

Relevance: Cardiac troponin is the preferred cardiac marker for myocardial infarction diagnosis and risk stratification; the consensus cutpoint according to the joint ESC/ACCF/AHA/WHF Task Force for the Redefinition of Myocardial Infarction is the 99th percentile of cTn concentration of a normal reference population as measured by each method. Harmonization of popular commercial assays in ACS samples at cTnI concentrations near the 99th percentile is uncertain.

Methodology: Appropriate plasma samples from 60 ACS patients were analyzed on the Stratus CS (Siemens), Advia-Centaur (Ultra assay, Siemens), Erenna (high sensitivity cTnI; Singulex), ACCESS II (AccuTnI; Beckman Coulter), Triage cTnI (Biosite) and i-Stat (Abbott) platforms. To determine harmonization, the results for each cTnI assay (y-axis) were compared by linear regression analysis to the median results of all platforms tested (x-axis; except Triage, see below).

Results: For the Triage assay, 52 of the 60 measurements yielded results of <0.05 ng/mL (non-numerical result); thus linear regression and harmony assessment was not possible. Triage results were excluded from calculation of the median. The results of the cTnI linear regression analysis for the other methods are displayed below.

S. Karimi, P. Simpson, K. Bui, E. Quan, L. Le, S. Agee, G. Baechler. Singulex, Inc., Alameda, CA,

Background: Measurement of cardiac Troponin-I (cTnI) has been designated the gold standard for diagnosis of acute myocardial infarction and as indicative of cardiac disease. Measurement of cTnI can be used for diagnosis of acute myocardial infarction and in the risk stratification of patients with non-ST segment elevation acute coronary syndromes, with respect to relative risk of mortality, myocardial infarction, or increased probability of ischemic events requiring urgent revascularization procedures. Recently a new ultra-sensitive assay for in vitro diagnostic measurement of cTnI has become available in the Singulex Clinical Laboratory (CLIA#:05D1092709, CLF#:338067). This assay utilizes the Erenna Immunoassay System, which is based on single-molecule detection.

Objective: Evaluate the analytical performance of the Erenna cTnI assay for the Singulex Clinical Laboratory.

Methods: Analytical accuracy and spike recovery were determined by preparing and testing spiked samples of EDTA plasma from a single donor. This plasma was spiked with cTnI standard from NIST (SRM 2921, Gaithersburg, MD) to generate a series of samples ranging from 5.2 - 52 pg/mL. Assay precision and lot-to-lot variability of assay reagents were assessed using two controls prepared by spiking two EDTA plasma samples to levels of 6 and 55 pg/mL, respectively, with a cTnI reference standard from HyTest (Turku, Finland). A reference range for the Erenna cTnI diagnostic assay was established using specimens from 153 apparently healthy blood donors (122 male, 31 female, average age 34.5 yrs, age range 19 - 63 yrs), and the 99th%, 10%CV and inter-quartile reference ranges were determined.

Results: Analytical sensitivity of the Erenna cTnI assay was 0.2 pg/mL (0.0002 ng/L) with a LLoQ of 1.0 pg/mL with good linearity (1 - 70 pg/mL, slope = 1.01, R2 = 0.999) and spike recovery (98-103%) of the HyTest cTnI material. Analytical accuracy, determined using NIST-spiked plasma, was >85% over a range of 6 pg/mL to 50 pg/mL (y = 0.9x - 0.3261, R2 = 0.9993) with average spike recovery of 89.1% (range 85.4-91.5%). Within- and between-run precision varied between 6-8%, with acceptable lot-to-lot variability (<15%) in testing of HyTest-spiked control plasma. Using donor plasma (n=153), the distribution of cTnI concentration was Gaussian, with a mean (+/-SD) of 1.97 (+/-1.85) pg/mL, and a calculated 99th percentile value of 6.28 pg/mL (<10%CV).

Conclusions: The Erenna cTnI assay shows excellent analytical performance in the Singulex Clinical Laboratory, with a 99th% cut-off value far below that of other clinically available in vitro diagnostic immunoassay systems. CLIA-regulated availability of this new, ultra-sensitive assay is an important step towards enabling early disease detection and clinical investigations into cardiac disease.

Robert Puskas, Douglas Held, Damon Sheets, Barbara Klein, Elizabeth Macy, Sara Agee, John Todd.
American Society of Clinical Oncology, Annual Meeting 2008; Abs No. 147.

Background: Growth in biomarkers as therapeutic targets and as surrogate markers for efficacy presents a need for increasingly sensitive immunoassays to expand biomarker applicability. Improved immunoassays will provide: (1) better evaluation and validation of new drug candidates, (2) better matching of patients to new therapies, (3) accelerated drug approval (4) earlier diagnosis of at-risk patients, and (5) a deeper understanding of cancer biology. Towards this end, Singulex® has developed two ultra-sensitive VEGF Immunoassays for human and mouse vascular endothelial growth factor (VEGF). Here we report the preliminary validation of these two novel assays.

Methods: Two novel assays were developed with the Erenna™ Immunoassay System for detecting VEGF: human (hVEGF) and mouse (mVEGF). Analytical sensitivity, cross-reactivity and precision were determined and compared to an ELISA based VEGF assay. Both the Singulex assay and ELISA assay were used to test a range of specimen types (plasma, cell lysates, conditioned media, and tissue specimens) from humans and mice. Preliminary assays with human plasma and tissue specimens were conducted to compare hVEGF levels between normal and breast cancer samples.

Results: The Singulex hVEGF assay had an LOD of 0.1 pg/mL, an LLOQ of 0.3 pg/mL, and 84-107% spike recovery; 90X more sensitive than the ELISA assay. Human VEGF concentrations were quantified in all specimens tested compared to the ELISA, which quantified VEGF in only 8% of plasma samples, but all of the cell lysate samples. The Singulex mVEGF assay had an LOD of 3.5 pg/mL, LLOQ of 5 pg/mL, and 68-111% spike recovery; 3X more sensitive than the ELISA assay. Cross-reactivity for the two assays was minimal for all specimen types tested, except for human plasma samples where the mVEGF assay demonstrated 80-100% CR.

Conclusions: We show that the Singulex hVEGF and mVEGF Immunoassays can detect VEGF at or below pg/mL levels, and can effectively quantify VEGF levels in plasma, cell lysates, conditioned media, and tissue samples from mice and humans. These novel assays are an important tool when used to assess tumor and normal breast cancer tissue and plasma.

V. I. Luzzi, P. Gilmore, K. Mills, C. Brunkan, S. Imai, R. Townsend, B. K. Klein, S. J. Agee, J. Todd

Background: While numerous new mouse models of diabetes have been developed to study the pathophysiology of this complex disease, very few efforts have been made to improve the methodology for the accurate measurement of insulin in mice. To robustly measure insulin concentrations in small volumes (≤10µL) of mouse plasma, sensitivity, accuracy and precision of immunoassays should definitely be improved. Although ultra-sensitive ELISA has been available that allows us to detect low concentrations of insulin in small volumes of mouse plasma, this method usually comes at the expense of accuracy and precision especially in the low range of insulin concentrations.

Objective: We report a comparative analysis of an ultra-sensitive immunoassay for mouse insulin based upon the single molecule counting technology of the Erenna™ System (Singulex) to two other assays based on the ultra-sensitive ELISA method.

Methods: Mouse insulin concentrations were determined by three different methods: the Erenna System (5µL plasma), the ALPCO mouse ultra-sensitive ELISA (5µL plasma), and the Linco/Millipore ELISA (10µL plasma). Samples were analyzed using a 384-well plate format with monoclonal capture and detection antibodies. Human insulin, which exhibits complete cross-reactivity in these methods, was used as the standard reference. For the Singulex vs. Linco/Millipore comparison, plasma was purchased from starved CD-1, Balb/c and C57BL/6 mice (Innovative Research). Mouse plasma samples (~100µL) for the Singulex vs. ALPCO comparison were provided by the Permutt Laboratory (WUSM).

Results: The Singulex Erenna and the ALPCO methods with 5µL mouse plasma resulted in analytical concentrations of insulin ranging from 503-2786 pg/mL vs. 936-2918 pg/mL, respectively. Linear regression of each data set showed that insulin concentrations determined by these two respective assays were highly correlated, but slight up-shifts of measured insulin concentrations were observed in the ALPCO method (y = 1.02x - 342, R2 = 0.936). Sensitivity, lower limit of quantification, and linearity were 12 pg/mL vs. 115 pg/mL, 20 pg/mL vs. 125 pg/mL, and 20-1280 pg/mL vs. 278-1280 pg/mL between the Erenna and the ALPCO methods, respectively. Within-assay precisions and measured insulin concentrations were 2-4% vs. 5-26% and 150-359 pg/mL vs. 119-326 pg/mL for the Erenna and the ALPCO methods, respectively. We also compared Singulex Erenna assays with 5µL plasma and Linco/Millipore assays with 10µL plasma. Measured insulin levels ranged 94-1961 pg/mL vs. 351-2002 pg/mL, respectively, and the measured values were highly correlated.

Conclusion: The comparison of the novel Singulex ErennaTM System and other ELISA methods clearly showed that the Erenna method had superior precision and accuracy with a smaller sample volume over currently available ELISA methods. Therefore, the single molecule detection technology of the Singulex ErennaTM System is able to meet much higher requirements of insulin measurement particularly when a small sample volume and low analyte concentration are evident, but a high level of precision and accuracy are required. Currently, we are attempting to apply this novel technology to measure insulin concentrations in several different mouse models.

A. H. Wu, A. Q. Lu, S. J. Agee, J. Todd, A. S. Jaffe

Background: Detection of cardiac injury by measurement of cardiac troponin (cTn) is essential for diagnosis of acute myocardial infarction (AMI) and provides information for risk stratification. Recent guidelines recommend use of the 99th percentile cutoff value with assay imprecision <10%, which most commercial assays do not meet. In addition, cTn values below the 99th% value provide diagnostic and prognostic information. Recently, a novel assay, the Erenna™ cTnI Immunoassay (Singulex) utilizing single photon fluorescence detection and paramagnetic microparticles (MPs) has been shown to have increased sensitivity and precision. However, the analytical specificity of this assay has not been shown.

Objective: Accordingly, we investigated the specificity of the Erenna MP-based cTnI Immunoassay for quantification of cTnI concentration in normal human subjects in matched serum and plasma (three different anticoagulants) samples.

Methods: We collected serum, EDTA, lithium heparin and citrate plasma from 20 random subjects free of overt cardiac disease and symptoms (13 female, 7 male, average 43y, range 23-64y). Concentrations of cTnI were determined in triplicate by the Erenna assay for each specimen type as an average (+/-SDEV). The average cTnI value was assessed across the 4 matched specimen types. To test for specificity, MPs were coated with either a cTnI capture monoclonal antibody (R&D Systems, Minneapolis MN), an unrelated monoclonal antibody directed against the beta-amyloid 42-mer peptide (αA-beta42, Covance) or left un-coated (blank MPs). Precision of the Erenna assay was determined by calculating average concentration and %CV of cTnI from two control sera over 14 assay runs performed over 4 days.

Results: cTnI was quantifiable in all serum and plasma specimens, and no significant differences in cTnI concentrations across samples were observed. No trends were observed in cTnI values related to age or sex. Average cTnI concentration across the 4 specimen types ranged from 0.84-22.37 ng/L, with an average of 3.76 (+/- 4.76) ng/L. Two donors provided significantly elevated average cTnI measurements (9.6 and 22.4 ng/L respectively). Taking the 22.4 ng/L sample as a statistical outlier, removal from the analysis yielded an adjusted average of 2.78 (+/- 1.93) pg/ml, similar to that reported previously. cTnI was undetectable (<0.1 ng/L) in 20/20 donor samples when blank MPs were used as the solid phase capture material and 16/20 samples when the MPs were coated with anti A-beta-42. The remaining 4 donors provided a 50%, 65%, 70%, and 82% reduction in cTnI concentration, respectively, and this non-specific binding accounted for 18-50% of the total cTnI signal. Precision of the Erenna assay using control sera were 7% CV at 8.3 ng/L and 10% CV at 2.2 ng/L.

Conclusion: Our data demonstrate that the Erenna MP-based cTnI Immunoassay has high specificity and can be used to reliably measure cTnI in apparently healthy human subjects. This high degree of specificity and high sensitivity should allow for further investigation of the clinical significance of mild increases in cTnI in patients with and without acute coronary disease.

Philippe Goix, Sara J. Agee, Robert Puskas, and John Todd.Innovations Corridor, Bio2008, San Diego, CA.

Many cardiovascular (cTnI), cancer (VEGF, PSA), inflammation (cytokines), and metabolic (insulin, GLP-1) biomarkers used for diagnostics cannot be measured in normal, healthy human subjects using currently available technologies due to extremely low circulating concentrations. However, in order to understand the progression of disease and evaluate effectiveness of therapeutics, it is essential to understand the baseline concentrations of these biomarkers in healthy subjects. To address this problem, Singulex has developed the Erenna™ MP-based Immunoassay System, which currently leads the next generation of molecular diagnostic technologies capable of quantifying biomarkers at the sub-picogram level. The Erenna technology utilizes single-photon fluorescence counting to provide a broad dynamic range and paramagnetic microparticles (MPs) for capture and detection of analyte from a complex biological sample, providing specificity, sensitivity and precision. Through an Erenna Technology Access Program (ETAP), Singulex has fostered key collaborations in both academia and the pharmaceutical industry that have demonstrated the utility and value of the Erenna Immunoassay System for applications defining clinically relevant biology around existing and emerging biomarkers. Cardiac troponin (cTnI) is considered the gold-standard biomarker for diagnosis of acute myocardial infarction (AMI) and cardiotoxicity. The Erenna System has been used to specifically quantify cTnI in serum from healthy human subjects, which is below the limit of detection for the previous generation of immunoassays. Additional studies have investigated the short- and long-term biological variation of cTnI, for use in clinical diagnosis of acute coronary syndrome (ACS) and chronic cardiac disease. The Erenna System can also detect low-level changes in circulating cTnI for subjects experiencing transient myocardial ischemia (TMI). There is also applicability for studies of cardiotoxicity in established animal model systems, and the Erenna System has been used to successfully measure the temporal variability of cTnI in rats. One feature of the Erenna System is the flexibility with which the assay can be customized to meet diverse diagnostic needs of investigators. The Erenna System has been successfully utilized to measure a diverse panel of important biomarkers for human disease. Baseline levels of cytokine concentrations from plasma of normal human subjects have been measured using the Erenna System, previously immeasurable by current assay technologies. Prostate specific antigen (PSA) has been quantified at exceptionally low concentrations both in male and female subjects, with promise for cancer diagnostics and therapeutics. The Erenna System has also been utilized to detect insulin levels in small volumes (≤10µL) of mouse plasma, allowing the next generation of mouse models for diabetes to be evaluated using molecular diagnostics. These collaborative efforts have shown that the Erenna System can be utilized to initiate studies requiring extreme sensitivity, and which were previously untenable. Such improvements in molecular detection technology are essential for developing biomarker baseline profiles for better diagnosis of disease, development of promising new diagnostic information and evaluation of therapeutic interventions.

Lu A, Agee SJ, Held D, Rieckenberg C, Freese B, Sheets D, Livingston R, Puskas R, Todd J.

Background: Cytokines are a complex family of secreted signaling molecules which trigger a variety of responses from the immune system in a combinatorial manner. Thus, the level of secretion of specific cytokines can have implications for a variety of pathologies, including inflammation, cancer, diabetes, and autoimmune disorders, to name a few. Interestingly, baseline concentrations from plasma of normal human subjects for many cytokines have yet to be defined by current assay technologies. Recently, a novel assay, the Erenna™ Immunoassay System (Singulex) utilizing single-photon fluorescence detection and paramagnetic microparticles (MPs) has been shown to have increased sensitivity and precision, making it a good candidate technology for such determinations.

Objective: We investigated the utility of the Erenna MP-based Immunoassay for quantification of a specific panel of cytokine concentrations in normal human plasma samples.

Methods: Human plasma was obtained from normal blood bank donors, with no apparent health concerns. Samples (100µl) were assayed for a panel of cytokines with the Erenna MP-based Immunoassay System (Singulex), which utilized paramagnetic microparticles as the solid phase format in combination with single-molecule counting. The assay panel consisted of: IL-1-beta, IL-6, IL-8, and IL-17 capture and detection antibodies in a sandwich assay format (R&D Systems). In the Erenna assay system, the immunoassay complex formed on the MP surface results in the release of fluorescently labeled detection antibody. The resulting solution is sipped into a 100µl flow capillary and photons are counted, via confocal microscopy, as they pass through a 2µm interrogation space. A large dynamic range of 4+ logs is obtained by combining single-molecule counting (low range) with photon counting (mid range) and total light measurements (high range). The sensitivity of each assay (LoD in pg/mL) and concentration range (pg/mL) for each cytokine in normal plasma was determined.

Results: Normal distributions were observed for each cytokine assayed in the panel: IL-1-beta (n=16), IL-6 (n=32), IL-8 (n=32), IL-17 (n=9). For IL-6 and IL-8, the concentration of each cytokine was quantifiable in all plasma specimens (32/32). For IL-1-beta, average cytokine concentration ranged from 0.05-0.25 pg/mL, with an assay LoD at 0.02 pg/mL (y = 771x + 375, R2=0.99). For IL-6, average cytokine concentration ranged from 0.2-26 pg/mL (Mean 2.3 pg/mL), with an assay LoD at 0.01 pg/mL (y = 1900x + 165, R2=0.99). For IL-8, average cytokine concentration ranged from 1.2-26 pg/mL (Mean 7.3 pg/mL). For IL-17, average cytokine concentration ranged from 0.2-0.7 pg/mL, with an assay LoD at 0.05 pg/mL (y = 0.86x + 5.4, R2=0.999) and LLoQ at 0.1 pg/mL.

Conclusion: Molecular detection technology has a broad range of applications, including clinical diagnostics and pharmaceutical development. The Erenna MP-based Immunoassays have low limits of detection, measuring a variety of low cytokine concentrations in normal human subjects. Such improvements in molecular detection technology are essential to understanding how deviations from normal cytokine secretion levels correspond to development of disease, and whether normal secretion patterns return after therapeutic intervention.

Alan H. Wu, Ann Q. Lu, Bob Freese, John A. Todd
American Association of Clinical Chemistry, Annual Meeting 2007; Abs No. A-46.

Background: The European Society/American College of Cardiology established cardiac troponin (cTnI) as the gold standard for diagnosis of acute myocardial infarction (AMI) and risk stratification for adverse cardiac events. We previously reported a highly sensitive cTnI assay and the use of this assay to define cTnI levels in normal subjects. This assay has been further developed to yield greater sensitivity and currently we report its analytical and preliminary clinical performance.

Methods: The Singulex cTnI assay was modified to use paramagnetic microparticles (MP) as the solid phase with a monoclonal capture antibody and a fluorescently-tagged affinity-purified goat detection antibody. After incubations and washing, the fluorescently-tagged antibody is chemically released from MPs and an aliquot is pumped into the Erenna™ Digital Molecule Counting (DMC) System. Individually-labeled antibodies are measured during capillary flow by setting the interrogation volume such that the emission of single fluorescent molecules is detected in a defined space following laser excitation. Total fluorescent signal is determined as a sum of the individual digital events. A 4-log dynamic range is obtained by analyzing digital events as well as total photons. The limit of detection the Singulex cTnI assay was determined by the mean +3 SD method. The normal range was determined on a population of 150 apparently healthy subjects. We also examined 56 serial samples from 17 patients who presented to the ED with a diagnosis of AMI in a preliminary study. Results were compared to the Centaur cTnI 1st (n=11) and 2nd (n=6) generation (ultra) assays. All had initial Centaur cTnI results that were <350 pg/mL (10%CV, 1st gen. Centaur), 6 were <100 pg/mL (99th percentile, 1st gen. Centaur), and 2 were <40 pg/mL (10% CV 2nd gen. Centaur). The cTnI concentration was positive on at least one subsequent serial samples from these patients on the Centaur, establishing the diagnosis of AMI.

Results: The analytical sensitivity of the Singulex assay was 0.2-0.3 pg/mL. The precision was <10% at 1.6, 12.5 and 50 pg/mL. The assay provided a linear response from 0.39 to >100 pg/mL (r2 = 0.99; y=1x+0.14). The reference population exhibited a normal distribution with a mean at 2 pg/mL (range 0.4-9 pg/mL). We established a preliminary cutpoint at 10 pg/mL, which is 50 fold higher than the analytical sensitivity. In the 9 cases that had initial Centaur cTnI value between 100 and 300 pg/mL, all were positive for Singulex with values ranging from 37-91 pg/ml (3.7-9 times >cut point). In the 8 cases that had initial Centaur cTnI value <100 pg/mL, 3 of 8 cases were Singulex positive. The Singulex assay was positive at least 1 sample earlier than the 10%CV cutoffs for either the 1st gen (10 of 11 patients) or 2nd gen Centaur (1 of 6 patients) assays.

Conclusion: We demonstrated a cTnI reference range between 0.4 and 9 pg/ml, with a mean of 2 pg/ml, some 10-fold higher than analytical sensitivity. Monitoring increases in cTnI levels above this reference range enabled detection of AMI several hours earlier than the Centaur cTnI assays.

Alan H.B. Wu, Noelle Fukushima, Robert Puskas, John Todd, Philippe Goix
Clinical Chemistry. 2006;52:2157-2159.
doi:10.1373/clinchem.2006.073163   
[PMID: 18061987] 

Background: The European and American Cardiology Societies (ESC/ACC) have established cardiac troponin (cTnI) as the gold standard for diagnosis of myocardial infarction (AMI) and risk stratification for adverse cardiac events. The ESC/ACC recommends a cTnI cutoff at the 99th% of the normal range with an assay imprecision (CV) of <10%. Currently none of the commercial assays can detect cTnI in healthy subjects with the requisite precision, therefore many have advocated a cTnI cutoff at the 10% CV value. We developed and evaluated a high-sensitivity cTnI assay and determined its analytical and preliminary clinical performance.

Methods: The Singulex cTnI assay utilizes a 384-well ELISA plate, and the ZeptX™ Digital Molecule Counting (DMC) System. This assay uses a monoclonal capture antibody and a fluorescently-tagged affinity-purified goat detecting antibody. After washing, the fluorescently- tagged antibody is chemically released into each well. An aliquot is pumped into the analyzer. Individually-labeled antibodies are measured during capillary flow by setting the interrogation volume such that the emission of only 1 fluorescent molecule is detected in a defined space following laser excitation. With each signal representing a digital event, this configuration enables extremely high analytical sensitivities. Total fluorescent signal is determined as a sum of the individual digital events. Each molecule counted is a positive data point with hundreds to thousands of DMC events/sample. The limit of detection the Singulex cTnI assay was determined by the mean +3 SD method. The normal range was determined on a population of 88 apparently healthy subjects. This assay was correlated to the Bayer Centaur on 130 samples from patients admitted with chest pain. We also examined 47 serial samples from 18 patients who presented to the ED with a diagnosis of AMI. In this latter group, all had initial Centaur cTnI results that were <0.35 ng/mL (10% CV cutpoint), and 12 were <0.1 ng/mL (99th%). The cTnI concentration was positive on all subsequent serial samples from these patients on the Centaur, establishing the diagnosis of AMI.

Results: The analytical sensitivity of the Singulex assay was 1 pg/mL. The precision was 10% at 4 and 12 pg/mL. The reference population exhibited a normal distribution. The 99th percentile was determined to be <7 pg/mL. The linear regression between Singulex (y) and Centaur (x) was: y=0.113x + 0.048, r=0.937. In the 3 cases that had initial Centaur cTnI value between 0.1 and 0.35 ng/mL, all were positive for Singulex. In the 12 cases that had initial Centaur cTnI value <0.1 ng/mL, 5 of 12 cases were positive. The prospective use of the Singulex assay would have detected 53% more AMI cases than the Centaur when the admission sample was tested.

Conclusion: The use of a highly sensitive and precise cTnI assay will enable detection of AMI earlier than with existing cTnI assays. A higher number of patients at risk for adverse cardiac events may be detected. The increased sensitivity was achieved by counting individual fluorescent emission events by the ZeptX analyzer.