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.