**3.2 Instrumentation and operation conditions**

AB Sciex API3000 LC/MS/MS spectrometer system: HPLC pump, vacuum degasser, column oven (Series 200, Perkin Elmer), autosampler (HTS Pal, CTC Analytics) along with data handling system (Analyst Version 1.4.2, AB Sciex) and Laboratory Information Management System (Watson 7.0, Thermo Fisher Scientific) for operations and data analysis. Halo C18 LC column (75 2.1 mm, 2.7 μm, HiChrom) and PreFrit Filter guard column (0.5 μm, 6.35 1.57 mm, Anachem) were used. Column and autosampler temperatures were 60*°*C and 4*°*C, respectively. Mobile

phase A: MeOH:AA (100:0.2, *v/v*) and mobile phase B: H2O:AA (100:0.2, *v/v*). The gradient was 60% A (0 min), 80% A (5.0 min), 100% A (6.0 min), 60% A (6.5 min). Flow rate was 0.3 mL/min with injection volume of 10 μL and run time of 6.5 min. The solvent mixtures as needle wash were H2O:ACN/TFA (25:75:0.1, *v/v/v*) and H2O:ACN: FA:TFA (90:10:0.1:0.005, *v/v/v/v*). Sample diluent was MeOH:H2O:AA (50:50:0.2, *v/ v/v*). TurboIonSpray negative ionisation mode was applied to carry out the detection of mass spectrometry. Nitrogen (API3000) was used for nebulizing and drying, with 700*°*C of ion spray temperature and 4500 V of ion spray voltage. PH46 and IStd ions monitored were 381.10–135.10 (0.5) and 409.20–164.00 (0.5) dwell time 200 ms.

#### **3.3 Preparation of CSs and QC samples**

The PH46 and IStd primary stock solutions were prepared at 1 mg/mL in ACN: DMSO (50:50, *v/v*). The plasma/urine CSs at concentrations of 0.5, 1, 3.25, 12.5, 50, 150, 450 and 500 ng/mL) were made by serial dilutions of the primary stock in control matrix. For each batch of plasma/urine samples, 200 μL aliquots of the bulk CSs were extract (duplicate) resulting in different concentrations of PH6 within the linear concentration range of the assay. 15 ng/mL of IStd solution was also made from the primary IStd stock solution using a mixture of MeOH:CHCl3 (50:50, *v/v*). The bulk faeces CSs were prepared by serially diluting the stock solution in homogenised control faeces to give 10, 20, 48, 120, 280, 720, 1800 and 2000 ng/g faecal equivalent. For each batch of faeces samples, aliquots (1 mL, equivalent to 250 mg of faecal material) of the bulk CSs were extracted in duplicate to give a range of concentrations of PH46 within the linear range of the assay and a fixed concentration of IStd (60 ng/g faecal equivalent).

Plasma/urine QC samples were prepared in control matrix from the primary PH46 stock solution in similar manner as CSs, to give QC samples at 0.5 (LLOQ), 1.25 (low), 15 (medium) and 400 (high) ng/mL plasma/urine concentrations and aliquots (200 μL) of the bulk QC samples were extracted for analysis. For faeces QC samples, concentrations were made at 10 (LLOQ), 25 (low), 100 (medium) and 1600 (high) ng/mL faecal equivalent, and aliquots (1 mL, equivalent to 250 ng of faecal material) of the bulk QC samples were extracted for analysis. All sample solutions were stored in a freezer set to maintain a temperature of 20°C in dark when not use and brought to RT before analysis. To all weighing, a correction for batch specific purity and a correction for salt content (free acid (382.5)/salt (577.7)) were applied.

#### **3.4 Preparation of control matrix and sample extraction**

Control plasma (Lithium Heparin anticoagulant), urine and homogenised faeces (1:3 faecal/H2O slurry, *w/v* by blending until a smooth consistency was obtained), stored at 20°C when not in use, were removed from the freezer and allowed to thaw at RT. Control matrix (pooled) was vortex mixed and centrifuged (3500 rpm, 4°C & 5 min) prior to use. 200 μL (1 mL in the case of homogenised faeces, equivalent to 250 mg faeces) was added to a clean, dry screw cap glass tube (12 75 mm) for DB (control matrix only without PH46/IStd) and SB (control matrix with IStd only) samples. CSs, QC and test samples were removed from the freezer and allowed to thaw at RT and vortex mixed prior to aliquoting. 200 μL of aliquots (1 mL in the case of faeces) of each sample solution were added to clean glass tubes (12 75 mm). 50 μL (10 μL in the case of in the case of faeces) of IStd in MeOH:CHCl3 (50:50, *v/v*) at a working concentration of 0.06 μg/mL (1.5 μg/mL) was added to all samples except for *Bioactive Indanes: Development and Validation of a Bioanalytical Method of LC-MS/MS… DOI: http://dx.doi.org/10.5772/intechopen.112275*

the DB samples, where 50 μL (10 μL instead in the case of faeces) of MeOH:CHCl3 (50:50, *v/v*) was added. A further 2 mL of MeOH:CHCl3 mixture (50:50, *v/v*) (4 mL of CHCl3 in the case of faeces) was added to all samples and the tubes were sealed and vortex mixed thoroughly prior to centrifugation (3500 rpm, 5 min, 4°C). The supernatant from all samples (the lower CHCl3 layer in the case of faeces) was transferred to clean tubes, dried under nitrogen at 60°C and then reconstituted in 100 μL of MeOH:H2O:AA (50:50:0.2, *v/v/v*). The extracts were then vortex-mixed and transferred into plastic matrix tubes before being centrifuged for 5 min at 3500 rpm at 4°C.

#### **3.5 Validation procedures**

When the reference standard PH46A was weighed, a correction factor was made for the salt content. Therefore, all peak area measurement and determined concentrations reported in the study were for PH46 (the free acid form). Each batch of samples had matrix DB, matrix SB, CSs, QC and test sample extracts. Duplicate DB and SB samples and CSs at each level were extracted. The CSs which met the acceptance criteria were used to construct the calibration curve. Two replicates were injected: one at the start and one at the end of the run. Concentration order was used for CS injections in each section of the run. QC samples were made in control matrix at three concentrations: low, medium and high (n ≥ 2 at each level). Where n = 2 at each level were prepared, the samples were injected as follows: one low-QC and one medium-QC level samples were injected at the start of the run following the first set of CSs, one Low-QC and one High-QC level samples were injected at the end of the run prior to the second set, and the remaining Medium-QC and High-QC level samples were injected midway through the run. In the case of n > 2 at each level were prepared, extra QC samples were appropriately distributed through the run. In order to avoid potential assay carry-over, the injections of solvent or extracted matrix DB samples were made after the high concentration extracts before the injection of low concentration extract.

The acceptance criteria were: (i) at least 75% of the CSs must back-calculate to within 100 15% (plasma and urine assays) or 100 20% (faeces assay) of the nominal concentrations (100 20% at the LLOQ for plasma and urine assays or 100 25% at the LLOQ for faeces assay); (ii) at least 67% of the total number of QC samples in each batch had to be within 100 15% (plasma and urine) or 100 20% (faeces) of the nominal concentrations for the determined concentrations, including at least one QC sample at each concentration had to meet this criterion. Precision was calculated as CV of mean. Accuracy was calculated as mean determined concentration/nominal concentration.

#### *3.5.1 Determination of stability of stored spiking solutions*

The stability of the PH46 storing stock solutions in ACN:DMSO (50:50, *v/v*) in a refrigerator set to maintain 4°C was examined by comparing the peak area ratios of the fresh stock solutions with the stock solutions on 200 days storage. The preparation of all stock solutions were performed in the same way. The ULOQ solutions were prepared by diluting the stock solutions using an appropriate diluent. The stability of the IStd storing solutions in ACN:DMSO, (50:50, *v/v*) for the stock solution and MeOH:CHCl3 (50:50, *v/v*) for the working solution, in a refrigerator set to maintain 4° C was also studied by comparing the peak area ratios of the freshly prepared and diluted stock and working solutions with the previous stocks on 250 days storage and

previous working solutions on 192 days storage, respectively. PH46 and IStd were considered to be stable if the mean responses (MRs) for the stored solutions were within 100 10% of the MRs of the fresh solutions and ≤ 10% for the precision.

The stability of the PH46 storing stocks at ambient RT was also investigated by comparing the peak area ratios of an aliquot of the stock solution (4°C) against an aliquot of the same stock solution on storage at ambient RT for 24 h. Dilutions were made to both solutions to the ULOQ level with an appropriate diluent. PH46 was considered to be stable at RT when the MRs of the solutions stored at RT were within 100 10% of the MRs from the solutions at 4°C and ≤ 10% for the precision.

#### *3.5.2 Linearity and specificity*

For each batch, the calibration curve over 0.5–500 ng/mL (plasma and urine assays) or 10–2000 ng/g (faeces assay) was constructed from PH46:IStd area response ratios plotted against the nominal matrix concentrations of PH46 to determine the optimum regression parameters, using linear regression with 1/x<sup>2</sup> weighting factor. The matrix concentration of PH46 from each CSs was calculated from the corresponding curve. DB and SB samples were also extracted and analysed; but not included in the regression analysis. For the batch to be acceptable the determined matrix concentration of PH46 for each sample used to construct the calibration curve, had to be 100 15% (plasma and urine assays) or 100 20% (faeces assay) of the nominal matrix concentration (100 20% (plasma and urine assays) or 100 25% (faeces assay) of the nominal matrix concentration at LLOQ). At least 75% (including at least one LLOQ and one ULOQ samples) of the CSs had to meet the above criteria.

The specificity of the assay for PH46 and IStd was determined by extraction and analysis of one DB sample from six independent sources of control matrix, three SB samples in a single source of control matrix and three ULOQ samples (without IStd) in the same source of control matrix as for SBs. If interfering peaks at the retention time of PH46 and/or IStd were noted, these were deemed to be insignificant if the response at the retention time of PH46 in the DB and SB samples was ≤20% of the average analyte response in LLOQ CSs and if the response at the retention time of IStd in DB and ULOQ samples was ≤5% of the IStd response accepted in the calibration curve, including SBs.

#### *3.5.3 Assay recovery and matrix effect*

The matrix effects on the plasma assay were determined by extracting replicate (n = 6) samples of control matrix, from six individual sources with one of these sources also presented as haemolysed plasma and spiking appropriate volumes of PH46 and IStd solutions (prepared in MeOH:H2O:AA (50:50:0.2, *v/v/v*) following extraction. Three replicate matrix effect samples at each of the matrix equivalent concentrations (1.25 and 400 ng/mL) and IStd concentration (15 ng/mL) were generated. Spiking appropriate volumes (as above) of the PH46 and IStd solutions in the same ratio as the extract samples resulted in replicate (n = 3) non-extracted QC samples. The ratio of the MR of replicate samples spiked into matrix to the MR of the non-extracted samples was calculated to determine the MF for PH46 and IStd in each source of matrix. The value of (PH46-MF)/(IStd-MF) from the same source was calculated to determine the IStd-normalised MF. There was deemed to be no matrix effect in the samples when the precision of the IStd-normalised MF (calculated from the six sources) was ≤15% at each level.

*Bioactive Indanes: Development and Validation of a Bioanalytical Method of LC-MS/MS… DOI: http://dx.doi.org/10.5772/intechopen.112275*

The recovery of the plasma assay was investigated following the preparation, extraction and analysis of replicate (n = 3) matrix-effect samples, at low, medium and high levels, from a single matrix source. Replicate (n = 3) extracted samples were prepared following the details in Section 3.5.4 at the same concentration levels (low, medium and high) as the matrix-effect samples. The recoveries of PH46 and IStd were calculated as: MRs in the extracted samples/MRs in the matrix-effect samples.

#### *3.5.4 Intra-batch assay accuracy and precision*

The accuracy and precision of intra-batch assay were assessed using replicate (n = 6) QC samples prepared in control matrix at 0.5 (LLOQ), 1.25 (low), 15 (medium) and 400 (high) ng/mL for the plasma and urine assays, and 10 (LLOQ), 25 (low), 100 (medium) and 1600 (high) ng/g faecal equivalent for the faeces assay. The intra-batch accuracy expressed as the mean percentage determined concentration/ nominal concentration. For the plasma and urine assays, the acceptance criteria were expected to be within 100 15% at each concentration level (100 20% at LLOQ level) for all three occasions. For the faeces assay, the criteria for acceptance at each level were within 100 20% (100 25% at the LLOQ level). The intra-batch precision was determined by the CV of the mean determined concentration. The criteria for acceptance should be ≤15% at each level (≤20% at LLOQ level) for all three occasions for the plasma and urine assays or ≤ 20% (≤25% at the LLOQ level) for the faeces assay. The inter-batch assay accuracy and precision were determined on three occasions (on different days). The acceptance criteria at each level were that assay accuracy for all three occasions was within 100 15% (100 20% at the LLOQ level), and that the assay precision for all three occasions was ≤15% (≤20% at the LLOQ level).

#### *3.5.5 Assay carry-over and matrix dilution*

Carry-over was assessed by the preparation, extraction and analysis of replicate (n = 2) samples prepared in control matrix at concentrations of 0.5 ng/mL (assay LLOQ) and 500 ng/mL (assay ULOQ). Additional replicate (n = 2) DB samples were also prepared, extracted and analysed. The samples were analysed in the following sequence: ULOQ (x1), DB (x2), LLOQ (x1), ULOQ (x1), solvent samples (MeOH: H2O:AA, 50:50:0.2, *v/v/v*) (x2) and LLOQ (x1). The assay was deemed to have no carry-over for PH46 when the responses for PH46 in DB and solvent samples were ≤ 20% of the detector responses for PH46 in the next LLOQ samples. The assay was deemed to have no carry-over for IStd if the responses for IStd in DB and solvent samples were ≤ 5% of the detector responses for the IStd in the previous ULOQ samples. A bulk stock (4 mg/mL of PH46 in control plasma/urine or 16 mg/g faecal equivalents in control homogenised faeces) was prepared in control matrix. Replicate (n = 6) QC samples from this stock were diluted 10- and 100-fold with control matrix. Aliquots (200 μL for plasma/urine assays or 1 mL equivalent to 250 mg faecal material for faecal assay) (n = 6 of each dilution factor) were then extracted (nominal concentrations of 400 and 40 ng/mL respectively for plasma/urine assays or 1.6 mg/g and 160 ng/g faecal equivalent respectively for the faecal assay). The determined concentrations were corrected for the dilution factors. Matrix dilution was considered to be acceptable if the assay accuracy was within 100 15% for the plasma/urine assays or 100 20% for the faeces assay, and the assay precision was ≤15% for the urine assay and ≤ 20% for the faeces assay.

#### *3.5.6 Stability experiments*

For the following stability experiments, bulk QC samples were prepared at the appropriate low, high and dilution levels for each assay. A 10-fold dilution in control matrix was carried out on the dilution level QC before extraction and analysis.
