The mean RSS for the five-parameter fitted curve was < 0.001 (n = 26) which was significantly better than our acceptability criterion of RSS = 0.01 ( Fig. 2B). The error for the back-calculated values of the standards was within 30%, except for the lowest concentration (0.006 μg/mL). The CV was < 10% for concentrations above 0.011 μg/mL and the dynamic range of the assay was two orders of magnitude. To establish the LOB, blank samples were tested (negative control, 0 μg/mL) along with the standard CHIR-99021 nmr curve. The mean proportion value of the shifted area (immune complexes) over the total area

determined from the blanks was 0.011 ± 0.003 (n = 60). The LOB was thus calculated to be 0.015 (mean + 1.645 × SD) and the extrapolated PARP inhibitor ATI concentration from the standard curve was 0.006 μg/mL ( Table 1). To determine the LOD, the extrapolated value of the lowest standard concentration (0.006 μg/mL) was obtained as 0.014 ± 0.003 μg/mL (n = 26). The LOD was calculated from the LOB and the SD from the lowest concentration in the standard curve with < 30% error: LOD = LOB + 1.645 × SD(low concentration sample) which was 0.012 μg/mL. The LLOQ for

the ATI-HMSA assay was 0.011 μg/mL, which was determined by the interpolated concentrations of replicates of the low ATI concentration with CV < 30%. The ULOQ for the ATI-HMSA assay was 0.54 μg/mL, which was similarly determined by the interpolated concentrations of replicates of the high ATI concentration with CV < 20%. The effective serum concentrations corresponding to the LLOQ and the ULOQ for the ATI-HMSA were determined by multiplying

the concentration with the dilution factor (50), which corresponded to 0.56 μg/mL and 27 μg/mL, respectively. The performance characteristics of the IFX-HMSA standard curve in the concentration range of 0.03–3.75 μg/mL were similarly assessed over 38 experiments by multiple analysts using different instruments on different days (Table 2). The same methods were used to determine the LOB, LOD, stiripentol LLOQ, and ULOQ as described for the ATI-HMSA. The LOB, LOD, LLOQ, and ULOQ for the IFX-HMSA were 0.0027, 0.0074, 0.039, and 1.36 μg/mL, respectively. The effective IFX serum concentration for the LLOQ and ULOQ were 0.98 and 34 μg/mL (dilution factor = 25). To assess the precision and accuracy of the ATI-HMSA and the IFX-HMSA, two methods were used. First, we used the high, mid, and low QC samples in both assays to determine their recovery rate. As shown in Table 3, the ATI-HMSA intra-assay precision had a CV < 4% and the accuracy rate was < 12% error. The intra-assay precision and accuracy for the IFX-HMSA were < 6% and < 10% error, respectively (Table 4). Second, we tested the high, mid, and low control samples over different runs and instruments and by multiple analysts.