Development and Optimization of an Immunoassay

1 Selection of Blank Matrix Using Prototype Assay 

The analyte may be present in a variety of matrices such as plasma, serum, cerebrospinal fluid (CSF), urine, or other biological fluid, all of which contain a multitude of interfering factors that can impact the method s performance. As mentioned earlier in the chapter, LBA samples are not pretreated prior to analysis, thus calibrators and quality control (QC) samples should also be prepared in the study matrix to best mimic these samples and ensure accurate measurements. The ideal matrix (1) has low background signal in the assay (OD 0.1 for chromogenic end points), (2) has minimal or no analyte-like activity, (3) is devoid of interfering factors, and (4) demonstrates a response that is proportional to the concentration of the spiked analyte. Access to a prototype method using an assay buffer of defined composition as a reference helps to identify an appropriate matrix. 

A prototype method can be obtained from a pharmaceutical company s QC or discovery group, an academic institution, or some other source, although often the scientist may need to design and develop such a method. Checkerboard experiments [8,10] define the optimal concentration of coating material, detector antibody concentration, and approximate range of the calibration curve. A calibration curve generated in assay buffer serves as a reference to assess the suitability of blank matrix for spiking calibrators and QC samples. 

Initially, 15 20 individual blank matrix samples should be screened as described below. The samples are obtained in small aliquots (5-mL size) from a commercial 

Plate coating, the first step of an ELISA, is influenced by the concentration and nature of the coating molecule, the pH and ionic strength of the coating buffer, and the time and temperature of incubation [26]. A sandwich immunoassay requires an excess of coating agent while only a limited amount is used in a competitive ELISA. Parameters 

---

The Immunoassay submitted for evaluation must be "mature," because expensive evaluation studies cannot be undertaken on unoptimized methods. The developer must first optimize and evaluate the immunoassay in-house, preferably by individuals most familiar with the assay system. Parameters addressed should include bias, precision (repeatability and reproducibility at the detection levels of interest), and rate of false positive and false negative results. The Agency should be supplied the raw data documenting assay performance. When possible, this information should be submitted on an IBM compatible floppy disk. A review of the data must be conducted before undertaking the evaluation study to verify interpretations of the data made the developer. 

An important extension of our large validation studies involves the use of data bases from field studies in the development of improved statistical methods for a variety of problems in quantitative applications of immunoassays. These problems include the preparation and analysis of calibration curves, treatment of "outliers" and values below detection limits, and the optimization of resource allocation in the analytical procedure. This last area is a difficult one because of the multiple level nested designs frequently used in large studies such as ours (22.). We have developed collaborations with David Rocke and Davis Bunch (statisticians and numerical analysts at Davis) in order to address these problems within the context of working assays. Hopefully we also can address the mathematical basis of using multiple immunoassays as biochemical "tasters" to approach multianalyte situations. 

---