Feasibility of Singlet Analysis for Ligand Binding Assays: a Retrospective Examination of Data Generated Using the Gyrolab Platform.

There are many sources of analytical variability in ligand binding assays (LBA). One strategy to reduce variability has been duplicate analyses. With recent advances in LBA technologies, it is conceivable that singlet analysis is possible. We retrospectively evaluated singlet analysis using Gyrolab data. Relative precision of duplicates compared to singlets was evaluated using 60 datasets from toxicokinetic (TK) or pharmacokinetic (PK) studies which contained over 23,000 replicate pairs composed of standards, quality control (QC), and animal samples measured with 23 different bioanalytical assays. The comparison was first done with standard curve and QCs followed by PK parameters (i.e., Cmax and AUC). Statistical analyses were performed on combined duplicate versus singlets using a concordance correlation coefficient (CCC), a measurement used to assess agreement. Variance component analyses were conducted on PK estimates to assess the relative analytical and biological variability. Overall, 97.5% of replicate pairs had a %CV of <11% and 50% of the results had a %CV of ≤1.38%. There was no observable bias in concentration comparing the first replicate with the second (CCC of 0.99746 and accuracy value of 1). The comparison of AUC and Cmax showed no observable difference between singlet and duplicate (CCC for AUC and Cmax >0.99999). Analysis of variance indicated an AUC inter-subject variability 35.3-fold greater than replicate variability and 8.5-fold greater for Cmax. Running replicates from the same sample will not significantly reduce variation or change PK parameters. These analyses indicated the majority of variance was inter-subject and supported the use of a singlet strategy.

Development and validation of an alpha fetoprotein immunoassay using Gyros technology.

Circulating alpha fetoprotein (AFP) is a diagnostic and prognostic biomarker for hepatocellular carcinoma (HCC) with potential utility as a pharmacodynamic endpoint in rodent tumor models. This application is limited, however, by low sample volumes, highlighting the need for sensitive, sample-sparing biomarker assay methods. In order to improve the utility of AFP as an oncology biomarker, we developed a method for AFP using the Gyrolab™, an automated microimmunoassay platform. Commercially available antibodies were screened to identify optimal combinations that were then used in a multi-factorial design of experiments (DOE) to optimize reaction conditions. Analytical validation included assessments of accuracy and precision (A&P), and dilutional linearity/hook effect, as well as reagent and sample stability. The method is reliable, with total error, a measure of accuracy and precision, less than 30% for all concentrations tested. AFP concentrations were measurable in diseased mice and undetectable in normal mice. Therefore, this novel, low volume AFP immunoassay is suitable for pre-clinical drug development, where its miniaturized format facilitates serial sampling in rodent models of cancer.

A strategy for improving comparability across sites for ligand binding assays measuring therapeutic proteins.

Outsourcing and multi-site testing has increased for ligand binding assays supporting protein therapeutic measurement. It is common to combine and compare data across studies with data from multiple bioanalytical sites. We designed a prospective study to determine the benefits of increasing control over the transfer process to improve ruggedness. The experiment involved the testing of 30 incurred samples at 3 stages with incremental laboratory harmonization in standard/quality controls and assay components: Stage I represented a transfer of a detailed protocol and critical reagents. Stage II, a single source of standards and quality controls were provided to each site. Stage III, standards and quality controls plus a ready-to-use kit were provided. The results indicated that all testing facilities failed agreement testing using the stage I procedure. The introduction of standards from a single source improved the agreement. The modification reduced variation by 33% compared to the stage I approach. There was no additional benefit when a packaged kit was provided. In conclusion, introduction of a single source of standards and quality controls reduced the inter-site component of variation and should allow for combinability of data.

Strategies to minimize variability and bias associated with manual pipetting in ligand binding assays to assure data quality of protein therapeutic quantification.

Bioanalytical laboratories require accurate and precise pipetting to assure reproducible and accurate results for reliable data. Two areas where pipetting differences among analysts lead to poor reproducibility are long term stability testing and sample dilution. The purpose of this paper is to illustrate the problems with manual pipetting, describe an automation strategy to mitigate risks associated with manual pipetting, and provide recommendations on a control strategy that properly monitors samples requiring dilutions. We determined differences among various manual pipetting techniques by analysts within a laboratory. To reduce variability in pipetting, a flexible modular liquid handling script was created on the Hamilton Microlab Star (HMS) to perform sample dilution, pre-treatment and plate loading. The script is capable of handling variable dilution factors. Additionally, two dilution controls were prepared and tested at concentrations of high and mid quality controls (QC). These same dilution controls were incorporated into both pre-study validation and in-study QCs to monitor dilution processing and assay performance. Variability of manual pipetting among 11 analysts was more negatively biased with increasing dilution. Forward and reverse pipetting delivering different volumes contributed to the discordance. The dilutional bias with manual pipetting was eliminated using the liquid handler. Total error of dilution controls was less than 20%. The in-study pass rate was 100%. Application of liquid handlers minimizes the variability and bias due to manual pipetting differences among analysts. The incorporation of dilution QCs serves a dual purpose to monitor the dilution process of the samples as well as the binding assay performance.

Application of multi-factorial design of experiments to successfully optimize immunoassays for robust measurements of therapeutic proteins.

Developing a process that generates robust immunoassays that can be used to support studies with tight timelines is a common challenge for bioanalytical laboratories. Design of experiments (DOEs) is a tool that has been used by many industries for the purpose of optimizing processes. The approach is capable of identifying critical factors and their interactions with a minimal number of experiments. The challenge for implementing this tool in the bioanalytical laboratory is to develop a user-friendly approach that scientists can understand and apply. We have successfully addressed these challenges by eliminating the screening design, introducing automation, and applying a simple mathematical approach for the output parameter. A modified central composite design (CCD) was applied to three ligand binding assays. The intra-plate factors selected were coating, detection antibody concentration, and streptavidin-HRP concentrations. The inter-plate factors included incubation times for each step. The objective was to maximize the logS/B (S/B) of the low standard to the blank. The maximum desirable conditions were determined using JMP 7.0. To verify the validity of the predictions, the logS/B prediction was compared against the observed logS/B during pre-study validation experiments. The three assays were optimized using the multi-factorial DOE. The total error for all three methods was less than 20% which indicated method robustness. DOE identified interactions in one of the methods. The model predictions for logS/B were within 25% of the observed pre-study validation values for all methods tested. The comparison between the CCD and hybrid screening design yielded comparable parameter estimates. The user-friendly design enables effective application of multi-factorial DOE to optimize ligand binding assays for therapeutic proteins. The approach allows for identification of interactions between factors, consistency in optimal parameter determination, and reduced method development time.

TGF-beta signalling-related markers in cancer patients with bone metastasis.

We measured transforming growth factor (TGF)-beta-dependent biomarkers in plasma and in peripheral blood mononuclear cells (PBMCs) to identify suitable pharmacodynamic markers for future clinical trials with TGF-beta inhibitors. Forty-nine patients with bone metastasis were enrolled in the study, including patients with breast (n=23) and prostate cancer (n=15). Plasma TGF-beta1 levels were elevated in more than half of the cancer patients (geometric mean 2.63 ng ml(-1)) and positively correlated with increased platelet factor 4 (PF4) levels, parathyroid-related protein (PTHrP), von Willebrand Factor (vWF) and interleukin (IL)-10. PBMC were stimulated ex vivo to determine the individual biological variability of an ex vivo assay measuring pSMAD expression. This assay performed sufficiently well to allow its future use in a clinical trial of a TGF-beta inhibitor.

Polyamine-mediated apoptosis of alveolar macrophages during Pneumocystis pneumonia.

The number of alveolar macrophages is decreased during Pneumocystis pneumonia (Pcp), partly because of activation of apoptosis in these cells. This apoptosis occurs in both rat and mouse models of Pcp. Bronchoalveolar lavage (BAL) fluids from Pneumocystis-infected animals were found to contain high levels of polyamines, including spermidine, N1-acetylspermine, and N1-acetylspermidine. These BAL fluids and exogenous polyamines were able to induce apoptosis in alveolar macrophages. Apoptosis of alveolar macrophages during infection, after incubation with BAL fluids from Pneumocystis-infected animals, or after incubation with polyamines was marked by an increase in intracellular reactive oxygen species, activation of caspases-3 and -9, DNA fragmentation, and leakage of mitochondrial cytochrome c into the cytoplasm. When polyamines were depleted from the BAL fluids of infected animals, the ability of these BAL fluids to induce apoptosis was lost. Interestingly, the apoptosis inducing activity of the polyamine-depleted BAL fluids was restored when polyamines were added back. The results of this study suggested that Pneumocystis infection results in accumulation of high levels of polyamines in the lung. These polyamines activate apoptosis of alveolar macrophages, perhaps because of the ROS that are produced during polyamine metabolism.

Suppression of alveolar macrophage apoptosis prolongs survival of rats and mice with pneumocystis pneumonia.

The number of alveolar macrophages is decreased in patients or animals with Pneumocystis pneumonia (Pcp). This loss of alveolar macrophages is in part due to apoptosis caused by Pneumocystis infection. The mechanism of apoptosis induction is unknown. Cell-free bronchoalveolar lavage fluids from Pneumocystis-infected rats or mice have the ability to induce apoptosis in normal alveolar macrophages. To characterize the mechanisms by which apoptosis proceeds in alveolar macrophages during Pcp, specific caspase inhibitors are tested for their ability to suppress the apoptosis. In vitro induction of apoptosis can be inhibited by the caspase-9 inhibitor (Z-LEHD-FMK) but not by the inhibitor to caspase-8 or -10. The caspase-9 inhibitor can also inhibit apoptosis of alveolar macrophages in vivo when it is intranasally instilled into dexamethasone-immunosuppressed, Pneumocystis-infected rats or L3T4 cell-depleted, Pneumocystis-infected mice. The number of alveolar macrophages rebounds in caspase-9 inhibitor-treated Pcp animals. Phagocytic activity of alveolar macrophages in treated animals is also recovered, and organism burden in these animals is reduced. Administration of caspase-9 inhibitor also clears the exudate that normally fills the alveoli during Pcp and decreases lung inflammation. Furthermore, caspase-9-treated Pcp animals survive for the entire 70-day period of the study, whereas nontreated Pcp animals die 40-60 days after initiation of infection. Depletion of recovered alveolar macrophages by intranasal administration of clodronate-containing liposomes in caspase-9 inhibitor-treated animals abrogates the effects of the inhibitor. Together, these results indicate that immunomodulation of the host response may be an alternative to current treatments for Pcp.

Optimization of analytical and pre-analytical variables associated with an ex vivo cytokine secretion assay.

PURPOSE: Measurements of cytokine release in whole blood after ex vivo stimulation are useful in drug development. The components contributing to variation within such assays have not been clearly defined. Therefore, we characterized the sources of variability within an ex vivo stimulation assay for TNF-alpha release. METHOD: Fresh whole blood or mononuclear cells from a cell preparation tube were added to silanized, screw-top tubes with a final concentration of 1 microg/mL lipopolysaccharide (LPS). Each tube was purged with 95% air/5%CO2 and incubated 4 or 6 h at 37 degrees C in a metabolic water bath. Plasma TNF-alpha was next measured in supernatants by immunoassay. Total method variability was assessed in 10 normal donors each drawn in the morning and afternoon over 3 days. Four additional samples were pre-treated with dexamethasone to investigate inhibition of TNF-alpha release. RESULTS: Our analysis indicated precise temperature control, the timing and duration of stimulation, and the surface properties of the stimulation vessel most significantly influenced assay performance. A comparison of multiple anticoagulants indicated that careful consideration should be taken in selecting the optimal anticoagulant. The estimated total assay CV for all anticoagulants tested was less than 33.81%. The analytical variability (stimulation and measurement) was less than 25.88% CV. The one exception was mononuclear cells collected in sodium heparin. The total variability estimate incorporated day-to-day, diurnal, inter-donor, tube-to-tube and immunoassay variability. Using our optimized conditions, TNF-alpha release was inhibited by dexamethasone with a mean IC50 of 33.3 +/- 4.6 nM. CONCLUSIONS: We have described an optimal set of conditions for collection, storage and processing of an ex vivo cytokine stimulation assay. These conditions were selected for operational feasibility, minimal imprecision and elimination of potential confounding factors. The end result is a more robust method that can be applied to clinical drug development.

Development, validation, and implementation of a multiplex immunoassay for the simultaneous determination of five cytokines in human serum.

Quantification of biomarkers can provide important information about the safety and efficacy of candidate drugs. Unfortunately, limited sample volume and excess costs often limit analysis of multiple biomarkers. We developed, optimized, validated, and implemented a multiplex immunoassay for simultaneous measurement of multiple circulating cytokines: IL-1beta, TNFalpha, IL-6, IL-8, and IL-10. Multiplex immuoassays were performed using the Luminex LabMAP instrument. Capture antibodies for each cytokine were covalently bound to distinct microsphere subsets distinguished by differing dye ratios. The concentration of each individual cytokine determined by measuring orange fluorescence produced by a complex of a biotinylated cytokine-specific antibody and streptavidin-phycoerythrin. The lower limit of quantification for all assays was 20 pg/mL with the exception of IL-8 which was 100 pg/mL. The inter-assay precision was less than 25%CV for all analytes at all control levels both pre-study and in-study. The percent recovery ranged from 83 to 108% pre-study and 90 to 125% in-study. In a linearity assessment, a 15,000 pg/mL multi-analyte control could be diluted 1:50 and maintain expected accuracy. We measured the cytokine concentrations in more than 2000 serum samples from patients with sepsis. Multiplex results for IL-6 were compared to a conventional commercially available ELISA kit. The degree of agreement between the two methods as measured by the concordance correlation coefficient was 84.5%. Multiplex results were 2.36-fold higher than ELISA values on the average. After adjusting for this mean difference, the 95% empirical limits of agreement for the ratio of individual sample values were 0.33, 2.65. This multiplex immunoassay provided simultaneous measurement of circulating cytokines using 80% less patient specimen compared to traditional approaches and at a significantly decreased cost. Efficient use of this platform requires process improvements to fully maximize the positive impact of multiplex assays in clinical drug development.