bioanalytical method was validated to demonstrate the specificity, sensitivity,
linearity, recovery, accuracy, precision,
dilution integrity, carry-over effect and stability according to Guidance for Industry: Bioanalytical Method
Validation of USFDA7.
The specificity was investigated by analyzing
processed drug-free plasma from six individual rats and pooled drug-free rat plasma.
Specificity was established by the lack of peaks
at the retention time for the S012-1332 and internal standard.
The lower limit of detection (LOD) was the amount in
plasma after the sample cleanup corresponding to
three times the signal to noise ratio (S/N>3). The lower limit of
quantitation (LLOQ) is the concentration of the sample having ten times the
signal to noise ratio (S/N>10) and quantified with less than 20% variation
in precision and accuracy.
Linearity was evaluated at eight levels of
concentrations covering a range of 1-200 ng/mL estimated for 5 days. The
calibration curve was prepared by determining the best fit of ratio of
S012-1332 peak area and IS peak area versus concentration and fitted to the y =
mx + c using weighing factor (1/x2).
extraction recovery of S012-1332 was determined
by comparing the mean peak area of extracted samples spiked with known amounts
of S012-1332 with the mean peak area of analytical standards at corresponding
concentration. These experiments were performed at four concentration
levels (low-low, low, medium, and high) in five replicates.
Matrix effect is the typical
predicament when mass spectrometer is the
detector because of the competition for
ionization of analyte and co-eluent residual matrix component from plasma. To evaluate the matrix effects, drug-free
plasma samples were processed according to the extraction method described above
and then spiked with S012-1332 and IS at the final concentration. Then the
percentage ratio of the peak area of analyte spiked post-extraction to that of
the standard solution at equivalent concentration is calculated. If the
percentage peak area ratio is less than 85% or more than 115%, a matrix effect is implied.
The accuracy (% bias) and precision (% RSD) of this
analytical method were determined using QC samples (n = 5) in five
replicates of 1, 3, 75 and 150 ng/mL of S012-1332 for 5 days. The criteria for
acceptability of the data included accuracy (inter-day and intra-day) within ±15
% standard deviation (SD) from nominal values and precision (inter-day and
intra-day) of within ±15 % relative standard deviation (RSD), except for LLOQ,
where it should not exceed ±20% of accuracy and as well as precision. The precision was determined by one-way ANOVA
as within and between % RSD. The accuracy was
expressed as % bias.
integrity was evaluated by spiking the drug-free plasma with an analyte concentration
of about two times of ULOQ and diluted by ten- and twenty-fold with similar
matrix (n=6, each dilution factor). The diluted and undiluted matrix samples
were extracted and quantified using the calibration curve. Accuracy and
precision should be within ±15%.
Carry-over effect was assessed by injecting extracted drug-free plasma (n=6) samples
after calibration standard at the ULOQ. Carry-over in the blank sample should
not be greater than 20% of the LLOQ and 5% for the IS.
The autosampler stability of S012-1332 in
reconstituted processed plasma sample was
examined at 4°C for 24 h. The bench-top stability was evaluated at ambient temperature (24 ± 4°C)
for 8 h using QC samples in six
replicates. The long-term storage stability at -70 ± 10°C for over 30 days was also evaluated. Dry-residue stability was analyzed by storing extracted sample at -70
± 10°C for 72 h. The freeze-thaw
stability was determined after three freeze-thaw cycles (room temperature to -70 ±
10°C). Analytes were considered stable
when 80–120% of the initial concentration was
found for low and 85–115% for other concentrations.