Ligand-binding assays (LBAs) measure an analyte concentration via immunoreactivity of antibody, or binding partner, to the analyte of interest. When dealing with PK assays, reference materials are well-characterized because they are typically available in pure form and simple buffers. But in the world of biomarker LBAs, the calibrators are typically recombinant materials and therefore most often not identical to the endogenous form in the biological matrix being measured. In addition, most often it is hard to find a “clean matrix” to perform spike-recovery studies, and to confirm that the assay is suitable to accurately quantify the analyte of interest, like is the practice for PK assays.
Now, more and more companies are relying more on biomarkers to assess the efficacy, safety, and mechanism of action of drugs to make a “go” or “no go” decision—that means the data must be of the highest quality possible. I’ve discussed the concept of accuracy for biomarker analysis using LBAs quite a bit at various meetings over the past few years because we must find ways to address the unique characterization challenges of biomarkers as they continue to grow in importance.
The quality of biomarker data yielded by LBAs is influenced by various parameters and practices, such as reagents, kits, and assay validation practices. To ensure accuracy, you need to consider reference material and calibrator selection. Those that are WHO or NIBSC standards are best, but are not available for most early biomarker studies. You can also look to commercially available calibrators purified from the proper biological matrix, but these can be hard to find. That means the large majority of the time you must use the recombinant form as reference materials and calibrators.
This makes the concept of accuracy for biomarker analysis using LBA different than in other types of analyses. For example, on the accuracy scale from qualitative (the lowest level) to quasi quantitative, relative quantitative, and absolute quantitative (the highest level), PK assays can be well-characterized and therefore deal in absolute accuracy. In contrast, because most biomarkers LBAs must use research-only kits, it is acceptable for them to fall within the ‘relative quantitative accuracy’ or sometimes ‘quasi quantitative accuracy’ range.
It is important for your customers to understand upfront that the numbers generated do not represent absolute accuracy. However, in the biomarker world where we are looking at peaks and valleys—pre-dose vs. post-dose, etc.—the level of relative accuracy is for the most part sufficient for clinicians to make the decisions needed to move the drug process forward. Just keep the concept of “fit for purpose” and the intended use of the assay in mind next time you get into a long discussion with your colleague as to the level of accuracy required for a particular clinical study.
In a webinar that I recently hosted for Bioanalysis Zone, entitled “Considerations for Evaluation of Accuracy, Parallelism, and Reagent Lot-to-Lot Variability in Single and Multiplex Biomarker Ligand-Binding Assays”, I cover the concept of relative accuracy in biomarker LBAs in more depth, along with other key considerations for biomarker analysis. Click here to view the recording.
Also feel free to email me directly with any questions you may have regarding the evaluation of accuracy in your biomarker studies.