Assay precision profile

The specification development process is a data-driven activity that requires a validated analytical method. The levels of data needed include assay precision, replicate process results (process precision), and real-time stability profiles. A statistical analysis of these data is critical in setting a realistic specification. Most often, aggregation and fragmentation degradation mechanisms are common to protein and peptide therapeutics. Therefore, the SE-HPLC method provides a critical quality parameter that would need to be controlled by a specification limit. 

Figure 6.22 PWG assay dose precision profiles of multiplexed three-analyte immunoassay for two sets of experiments. Dose precisions correspond to standard deviations of analyte concentrations that were back-calculated using corresponding dose response curves. (From Pawlak, M. et al., Proteomics, 2, 383-393, 2002. With permission.)...

The response of an MIP-ILA is nonlinear and the variance is non-uniform so it would be desirable to establish the precision profiles by calculating the standard deviation, or the percentage of the coefficient of variation (%CV) vs concentration. This shows the interval of concentrations where precision is maximum (usually given as 20-80% inhibition). In immunoassays, and by extension in ILAs, this range is considered as the assay working range [45]. 

If an assay is being used in which it is customary to duplicate the measurement on each sample within the assay (e.g. immunoassay techniques), then precision profiles can be prepared from the duplicate measurements. This is done by calculating the random error for each pair of samples as a coefficient of variation, given by the expression... 

Another reason is that there is a high variation in assay precision close to zero dose, i.e., it usually drops asymptotically parallel with the y-axis reaching a minimum value around the middle of the calibration curve (see Fig. 9.5b). Thus, a considerable difference in CV in the absence of the analyte and at the MDC is expected. This is why the concept of functional sensitivity (FS) was introduced [14], defined as the lowest concentration of analyte for which the CV falls below 20%. FS is a parameter that is far more practical than AS and it is derived from the knowledge of the precision profile (PP), i.e., the changes in precision with the concentration of the analyte (Fig. 9.5b) [14]. 

The low detection limits of immunoassays depend on the typically high affinities of antibodies for haptens and antigens, as well as the detection limits of the labels used. The detection limit is usually defined as the concentration that yields a signal that is equal to the mean of the blank signal plus two or three standard deviations. This establishes the confidence range for the zero response. For this calculation, the bound/free versus concentration plot is used. While detection limits allow comparisons of different immunoassay methods at the lower concentration end, they say nothing about assay reliability for this reason, both detection limits and precision profiles should be compared. 

Precision varies with concentration because of this, precision should be evaluated at the low, middle, and high concentration regions of the standard curve, and should be evaluated in the different matrices that will be encountered in real assays. The precision profile (Section 16.4.3) is used to establish the working range of the assay. 

The relative uncertainty of measurements at or just exceeding the LoD may be large, and often a quantitative result is not reported. The lower limit for reporting quantitative results, the limit of quantitation (LoQ), relates to the total error being considered acceptable for an assay. From a precision profile for the assay and an evaluation of the bias in the low range, LoQ may be determined in relation to specifications of the method. For example, a laboratory may specify that the total error (e.g., expressed here as Bias -i- 2 SD) of an assay is lower than 45% (corresponding to a bias of 15% and a CV of 15%) of the measurement concentration. " In this case, the LoQ is the lowest assay value at which this specification is fulfilled. LoQ constitutes the lowest limit of the reportable range for quantitative results of an assay. 

Figure 1. Precision profile for C-peptide assay Broken lines indicate functional sensitivity (CV=10%) of the assay...

FIGURE 3.8 (a) Precision profile plate map. Four sets of 12 calibrator levels are assayed one predefined set is used to fit a calibration curve, and the remaining sets are treated as samples, (b) Precision profile plot. The %CV of calibrators as samples is plotted versus nominal concentration to give a precision profile. The concentrations at which a certain precision level is met, for example, 15% CV, define the preliminary lower and upper levels of quantification (LLOQ and ULOQ). 

The performance of an assay can be examined through profiling the precision measured for different sample (analyte) concentrations and conditions. Such assessment at any stage can be regarded as a precision profile. 

The precision profile of an assay showing nonuniform error. 

The precision of the assay for nonreduced samples was demonstrated by the evaluation of six independent sample preparations on a single day (repeatability) and the analysis of independent sample preparations on three separate days by two different analysts (intermediate precision). The RSD values for the migration time were 0.9%. The RSD values for peak area percent of the main peak and the minor peaks in the profile were 0.6 and 12.6%, respectively. The higher variability observed with the minor peaks was determined to be primarily related to the sample heating during preparation for the analysis. These results demonstrate that the use of uncoated fused-silica capillaries in combination with a sieving matrix can provide adequate precision and analyte recovery. 

A well-defined, precise, and validated method will help to determine the drug content accurately, whereas an assay method capable of detecting at low levels can help calculate drug losses during the manufacturing process. Different vendors can supply common APIs. The API characterization accompanied by information from the manufacturer on the synthetic route determines the impurities profile and the method used for the active assay. 

The quantitative abilities of the LC-TOF, although limited in the linear dynamic range, often allow measurement of metabolic profiles and PK profiles from the same samples. Zhang and co-workers [37] provide such an example. Quantitative results for a co-administration of five compounds to rats, with assay limits of quantitation (LOQ) between 1 and 5 ng/mL, and an upper limit of quantitation (ULQ) at 100 ng/mL were described. In this study, precision and accuracy were better than 20% for all five analytes, and comparable to triple-quadrupole quantitative data. More importantly, in addition to following the levels of dosed compound, the authors were able to identify several metabolites from the same full-scan data using the accurate mass capabilities ofthe LC/TOF-MS. 

---