(T1330-06-36) Duplicate-Well vs Single-Well Analysis in Ligand Binding Assays (LBAs): A Retrospective Comparison of Method Validations and Nonclinical Toxicokinetic (TK) Sample Analyses Data

Purpose: Traditionally, ligand binding assays (LBAs) have employed duplicate-well analysis to monitor the reproducibility of results. However, with the recent advances in LBA techniques, such as the use of automation, the feasibility of single-well analysis in LBAs has gained significant interest in bioanalytical community due to its potential advantages over duplicate-well analysis in reducing study time and resources. Single-well LBAs can also substantially reduce reagent and matrix consumption, and better conform with the principle of 3Rs (replacement, reduction, and refinement). Here we report a retrospective comparison of duplicate-well vs single-well method validation and nonclinical toxicokinetic (TK) sample analysis data generated using 4 different LBAs.

Methods: We retrospectively evaluated validation data (accuracy and precision) from 2 enzyme-linked immunosorbent assay (ELISA) and 2 Electrochemiluminescence (ECL) assays using duplicate-well and single-well analyses. Single-well data were taken from Well 1 (Replicate 1) for the comparative analyses. For each assay, intra‑ and inter‑assay precision and accuracy were evaluated in replicates of 3 (2 wells vs 1 well per replicate) from at least 6 separate occasions performed at 5 concentrations [lower limit of quantification (LLOQ), low QC, medium QC, high QC, and upper limit of quantification (ULOQ)]. The standard curves were assayed in duplicate wells for all analyses. We also retrospectively evaluated sample analysis data for two of the LBAs (a total of 181 concentration data) by comparing duplicate-well and single-well incurred sample reanalysis (ISR) data against the original results assayed in duplicate wells. The ISR results were statistically compared using Lin’s concordance correlation coefficient (CCC) which measures agreement between two data sets.

Results: The validation accuracy and precision data were found to be equivalent, regardless of evaluation method (duplicate-well vs single-well). Inter-assay precision and accuracy were within acceptance criteria for all 4 LBAs in both duplicate and single-well analyses. For duplicate-well and single-well analyses, the intra-assay accuracy (relative error, %RE) ranged from -16% to 32% and -16% to 41%, respectively. The inter-assay accuracy ranged from -11% to 16% for both duplicate and single-well analyses. For duplicate-well and single-well analyses, the intra-assay precision (coefficient of variation, %CV) ranged from 0% to 32% and 0% to 29%, respectively. The inter-assay precision ranged from 4% to 20% for duplicate-well analysis and 4% to 25% for single-well analysis (25% at the ULOQ). During sample analyses, the percentage of samples meeting ISR criteria (ISR results within ±30% of original results) ranged from 95% to 99% in duplicate analysis. When single-well ISR results were evaluated, the same 95%-99% samples were observed to meet the ISR criteria. Both duplicate and single-well ISR results were found to be in agreement with the original sample results as evident from the determined concordance correlation coefficient (CCC) range of 0.9600 to 0.9675, and 0.9599 to 0.9676, respectively. Additionally, the CCC between duplicate and single-well ISR results ranged from 0.9996 to 0.9998 showing no statistical difference between duplicate vs single-well analysis.

Conclusion: The findings indicate the feasibility of single-well analysis in ligand-binding assays for TK assessments, however, the decision of analytical strategy should be based on method development and validation data.

References: 1. Feasibility of Singlet Analysis for Ligand Binding Assays: a Retrospective Examination of Data Generated Using the Gyrolab Platform. Tracey H Clark, Phillip D Yates, Allison Given Chunyk, Alison P Joyce, Aidong Wu, Petar Pop-Damkov, Yiqun Zhang, Elizabeth A Dreher, Laurie A Tylaska, Jo-Ann A Wentland, Kathleen B Pelletier, Lindsay E King, Chad A Ray. AAPS J. 2016 Sep;18(5):1300-1308.
2. Feasibility of singlicate-based analysis in bridging ADA assay on Meso-Scale Discovery platform: comparison with duplicate analysis. Zhihua Jiang , John Kamerud , Zhiping You, Soma Basak, Elena Seletskaia, Gregory S Steeno, Boris Gorovits. Bioanalysis. 2021 Jul;13(14):1123-1134.