2.5 Method validation
The method validation was carried out for specificity, linearity, accuracy, precision, carryover, matrix effect, recovery, stability, dilution study following FDA guidelines and Chinese Pharmacopoeia.21
Six different lots of blank plasma samples were selected for the specificity assessment. Double blank plasma sample, blank plasma sample spiked with imatinib at LLOQ (10 ng/mL), plasma sample of the patient after 22.3 h of the last imatinib administration were processed to investigate possible endogenous interference. The response of the interference in blank plasma sample should be less than 20% of 10 ng/mL imatinib and 5% of IS.
Linearity was determined by plotting the peak area ratio (imatinib/IS) response versus the plasma concentrations of imatinib over the range from 10 to 2000 ng/mL. The weighting factor (1/x2) of each calibration curve was used and should be >0.99. The accuracy (relative error, RE, %) of calibration standards should be 85%-115% and 80%-120% for LLOQ .
For inter- and intra- precision and accuracy studies, samples were prepared at 10, 30, 300, 1800 ng/mL with five replicates each analyzed in three consecutive runs. The deviation between the actual concentration and measured concentration was calculated to assess accuracy of imatinib, and it should be within ±15% for 30, 300, 1800 ng/mL and ±20% for LLOQ (10 ng/mL). The precision was expressed as relative standard deviation (RSD,%) and it should be <15% for 30, 300, 1800 ng/mL and <20% for LLOQ (10 ng/mL).
Carryover was assessed by injecting a blank plasma sample after the highest concentration calibrator (2000 ng/mL) in three independent runs. The response of the blank plasma sample should not >20% of the LLOQ and >5% of IS.
Recovery of imatinib were both determined at LQC, MQC and HQC concentration levels with five replicates each. The procedure was carried out according to the section of ”sample preparation”, and the peak area was recorded as A. In addition, blank plasma plasma was taken and prepared according to the section of ”sample preparation”. LQC, MQC and HQC solutions were added to the obtained supernatant, and the corresponding peak area was calculated as B. Recovery was assessed by the ratio of A to B. Matrix effect was calculated by the ratio of peak area obtained by adding LQC, MQC and HQC solutions to post-extracted blank plasma and peak area by added to water. Recovery and matrix effect of IS were also evaluated.
Three QC samples with 30, 300, 1800 ng/mL imatinib were taken, and five samples were measured in parallel for each concentration. The stability was investigated when untreated plasma samples were placed at room temperature for 4 h, repeated freeze-thaw cycles for 3 times, placed in -80℃ for 350 days and treated plasma samples at the autosampler for 24 h. The stability of the analyst in working solution was investigated for 14 days at -80℃ with the same concentration level.
Imatinib standard solution with the concentration greater than 2000 ng/mL was added to blank plasma. Then five replicates of this QC sample prepared in parallel were analyzed after being diluted 10 times with blank plasma to assess dilution stability.