Bioanalytical Method Validations of Three Alpha1-Antitrypsin Measurement Methods Required for Clinical Sample Analysis

Abstract

Background/Objectives: The quality of clinical studies is largely determined by the bioanalytical methods used for testing study samples. Rigorous assay validation following defined criteria, for example, the European Medicines Agency guideline for bioanalytical method validation, is a prerequisite for such assays. Alpha1-antitrypsin (AAT) measurement, i.e., the specific measurement of AAT protein and its associated elastase-inhibitory activity, is an integral part of assay panels for clinical studies addressing AAT deficiency. Specifically, AAT must be measured in the matrix of citrated human plasma as well as in diluted solutions with high salt concentrations obtained through bronchoalveolar lavage (BAL). Sensitive and selective measurement methods are required, as BAL has a low level of AAT. Methods: We present the validation data obtained for three AAT measurement methods. Two of them, nephelometry and the enzyme-linked immunosorbent assay, which clearly differ in their sensitivity, provide AAT protein concentrations. The third is the highly sensitive, newly developed elastase complex formation immunosorbent assay that specifically measures the inhibitory activity of AAT against its pivotal target, protease neutrophil elastase. Using samples with relevant AAT concentrations, we addressed the assays' characteristics: accuracy, precision, linearity, selectivity, specificity, limit of quantification and short-term analyte stability Results: Overall, the three methods demonstrated low total errors, a combined measure reflecting accuracy and precision, even at low analyte concentrations of less than 0.5 µg/mL; adequate linearity over the required assay range; and acceptable selectivity and specificity. Furthermore, the short-time stability of the analyte was also demonstrated. Conclusions: All three AAT measurement methods met the acceptance criteria defined by the guidelines on bioanalytical assay validation, qualifying these methods for clinical sample analysis.

Keywords: assay accuracy; assay precision; bioanalytical method validation; clinical sample testing; α1-anitrypsin; α1-proteinase inhibitor.

Figure 1

Mean calibration curve of the ECFISA AAT activity measurement. The mean (n = 59) six-point, log–log calibration ranging from 6 to 192 ng/mL functional active AAT/mL is shown. Error bars indicate the relative standard deviations of the means (red circles). The inserts provide mean quality characteristics including the slope, correlation coefficient and RTE of the calibrations curve and the agreement of the back-fitted assay calibrators with their nominal concentration, respectively.

Figure 2

Mean calibration curve of the nephelometric AAT assay. The mean six-point calibration curve (n = 12) ranging from 9.69 to 310 µg AAT/mL is shown. Red circles indicate the mean signals. The inserts provide the mean signals obtained for the six assay calibrators and their RSDs and the agreement of the back-fitted assay calibrators with their nominal concentration, respectively.

Figure 3

Mean calibration curve of the AAT ELISA. The mean (n = 29) five-point, log–log calibration ranging from 1.75 to 28 ng/mL AAT/mL is shown. Error bars indicate the relative standard deviations of the means (red circles). The inserts provide mean quality characteristics including the slope, correlation coefficient and RTE of the calibrations curve and the agreement of the back-fitted assay calibrators with their nominal concentration, respectively.

Figure 4

Recovery of AAT activity and AAT protein in samples with low protein content and citrated plasma. The recovery of AAT is shown as a percentage of the expected AAT concentration. (a,b) show the data obtained for the AAT activity measurement in BAL-mimicking samples with a low protein content and in citrated plasma, respectively, while (c,d) provide the recoveries of AAT protein, determined with the nephelometric method and the AAT ELISA.

Figure 5

Results of the precision analysis. ND stands for not done, because intra-run precision was determined using a bracketing approach, if the analysis procedure included the measurement of a serial dilution series. The results of the precision analysis are expressed as the RSDs obtained for the means of six independent tests, carried out in one run (=intra-run precision) or six runs (=intermediate precision). (a,b) show the precision data of the AAT activity measurement in BAL-mimicking samples and citrated plasma, while (c,d) show the precision data of the AAT protein measurement with the nephelometric method and the ELISA, respectively, carried out for citrated plasma and BAL-mimicking samples.

Figure 6

Results of the linearity analysis. The linear regression curves comparing the AAT concentrations as found with the nominal values are shown. The coefficient of determination R² determined for these linear regression curves is also given. (a,b) show the curves for the AAT activity measurement in BAL-mimicking samples and citrated plasma, respectively; (c,d) show the regression curves obtained for the AAT protein measurement in citrated plasma and BAL-mimicking samples, respectively.