Forensic Science International 233 (2013) 304–311
Contents lists available at ScienceDirect
Forensic Science International journal homepage: www.elsevier.com/locate/forsciint
Direct determination of diazepam and its glucuronide metabolites in human whole blood by mElution solid-phase extraction and liquid chromatography–tandem mass spectrometry Rong Wang a,1, Xin Wang a,b,1, Chen Liang a, Chunfang Ni a, Lingjuan Xiong a,b, Yulan Rao c,*, Yurong Zhang a,** a
Shanghai Institute of Forensic Science, Shanghai Key Laboratory of Crime Scene Evidence, Shanghai 200083, China Shanghai Institute of Pharmaceutical Industry, State Key Laboratory of New Drug and Pharmaceutical Process, Shanghai 200040, China c Department of Forensic Medicine (Center of Forensic Science), School of Basic Medical Sciences, Fudan University, Shanghai 200032, China b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 8 June 2013 Received in revised form 25 September 2013 Accepted 5 October 2013 Available online 16 October 2013
A mElution solid-phase extraction (SPE) liquid chromatography–tandem mass spectrometry (LC–MS/ MS) method for simultaneous determination of diazepam, nordiazepam, oxazepam, oxazepam glucuronide, temazepam and temazepam glucuronide in human whole blood is presented. 200 mL of whole blood samples were loaded onto a Waters Oasis HLB 96-well mElution SPE plate using 75 mL of methanol as the elution solvent, and the eluents were injected into an Eclipse XDB C18 column. No hydrolysis, solvent transfer, evaporation or reconstitution was involved in the sample preparation procedures. Tandem mass spectrometric detection with Turbo Ion Spray was conducted via multiple reaction monitoring (MRM) under positive ionization mode. The method was validated and proved to be accurate (accuracy within 93–108%), precise (intra-day RSD < 9.9% and inter-day RSD < 7.2%) and sensitive with limits of detection (LOD) in the range of 0.05–0.25 ng/mL for all the compounds. Extraction recoveries were in the range of 31–80% for all the analytes. This method demonstrated to be reproducible and reliable. The applicability of the method was demonstrated by analysis of several forensic cases involving diazepam and its metabolites. ß 2013 Elsevier Ireland Ltd. All rights reserved.
Keywords: Diazepam Diazepam glucuronides 96-Well mElution plate solid-phase extraction Human whole blood LC–MS/MS
1. Introduction Diazepam is widely prescribed for anxity, sleep disorders, or convulsive attacks and also readily available [1]. It is frequently encountered in forensic cases due to misuse or abuse by drug addicts related to suicide attempts, road traffic offenses, drugfacilitated sexual assault or robbery [1–4]. This situation is often complicated by the fact that diazepam may induce coma so that the victims cannot describe what have happened, which would be the obstacle for crime investigation. What is more, drug interaction with diazepam probably potentiates certain side effects [5] which might be life-threatening. Thus, identification of diazepam ingestion is important, which would provide clues
* Corresponding author at: 138 Yi Xue Yuan Road, Shanghai 200032, China. Tel.: +86 21 54237403; fax: +86 21 54237404. ** Corresponding author at: 803 Zhong Shan Bei Yi Road, Shanghai 200083, China. Tel.: +86 21 22028901; fax: +86 21 65443131. E-mail addresses: [email protected] (Y. Rao), [email protected], [email protected] (Y. Zhang). 1 Rong Wang and Xin Wang contributed to the work equally and should be regarded as co-first authors. 0379-0738/$ – see front matter ß 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.forsciint.2013.10.004
to the police for case investigation and the result would be served as evidence at the court. It was estimated that the detection time of the glucuronide metabolites of diazepam (oxazepam glucuronide and temazepam glucuronide) might be longer than diazepam after ingestion of diazepam based on the fact that one literature [6] mentioned that the half-time of oxazepam glucuronide was longer than that of oxazepam in blood after administration of oxazepam. Meanwhile, information on diazepam/glucuronides ratios may be helpful to find out the dose [7] and the time elapsed after diazepam administration which will be conducive to judge crimes. One study [8] has demonstrated that interval time from doping to death in blood of fatally intoxicated heroin addicts could be concluded by the concentration and molar ratios (morphine-3-glucuronide (M3G)/morphine (M), morphine-6-glucuronide (M6G)/M, M3G/ M6G). Accordingly, approximate time from administration to testing point could be speculated if relationship between diazepam/glucuronides ratios and the time elapsed after diazepam administration could be summarized. Based on these facts, a conclusion could be made that, it is necessary to develop a method for the analysis of diazepam and its glucuronide metabolites in biological specimen.
R. Wang et al. / Forensic Science International 233 (2013) 304–311
Although urine is the preferred biological sample in the aspect of prolonging the detection time of drugs [9], concentrations of compounds in urine can be influenced significantly by individual difference or water consumption difference. Compared with urine, the concentrations of drugs and their metabolites in blood are directly related to actual impairment. The values are consistent and can be compared valuably with the levels reported in previous literatures. Meanwhile blood is a valuable specimen to reflect very recent intake of drugs [9], since certain drugs being tested often disappear from the blood fairly rapidly. Thus blood is generally considered to be the most useful sample for forensic investigation especially when urine is usually not available in autopsy. Compared with vast number of papers describing how to determine diazepam and its free metabolites in human blood [10–12], there are much less methods reported for the determination of oxazepam glucuronide and temazepam glucuronide. Only Simnk et al. [6] proposed a liquid chromatography–tandem mass spectrometry (LC–MS/MS) method for the simultaneous determination of oxazepam and oxazepam glucuronide in blood, serum and fluid after administration of oxazepam. Several literatures presented high performance liquid chromatography (HPLC) methods for the separation and quantification of diastereoisomeric glucuronides of oxazepam [13–15] or both diastereoisomers of oxazepam and temazepam glucuronides in plasma [16]. However, these papers have some limitations as their methods all involved enzymatic hydrolysis using b-glucuronidease, which have several critical problems [3,7,17]. The objective of this study was to determine diazepam, nordiazepam, oxazepam, oxazepam glucuronide, temazepam and temazepam glucuronide in human whole blood. The reason why whole blood was analyzed was in forensic cases, plasma or serum often could not be separated from whole blood as the blood has been hemolyzed already. Since plasma and serum is less complicated than whole blood, it is expected that, the method developed in this study using whole blood as the sample could also be applied to plasma or serum. Referring to our previous study [18], a direct LC–MS/MS method was reported for the analysis of diazepam, its free metabolites and its glucuronide metabolites in human urine without hydrolysis. In that study, individual SPE cartridges were utilized, while 96-well Oasis HLB mElution plate was chosen in this study. The postextraction solvent transfer, evaporation and reconstitution steps that are required in individual SPE cartridges are avoided due to the concentrating ability of the plate. This method reduced the labor cost and was more time-saving in sample preparation. Another difference of this study with the method we recently reported was the choosing of internal standards (IS). In both studies, isotope labeled IS were used to manage the influence of ion suppression or enhancement from endogenous components. In our previous study, diazepam-d5 was chosen as the IS for all compounds. While in this assay we chose diazepam-d5, oxazepam-d5, temazepam-d5 and oxazepam-d5 glucuronide as the IS for the corresponding non-labeled drugs. Stable isotope labeled nordiazepam and temazepam glucuronide were not commercial available for us, and instead estazepam-d5 was used as the IS for nordiazepam and temazepam glucuronide. Why none of diazepam-d5, oxazepam-d5, temazepam-d5 and oxazepam-d5 glucuronide was used as the IS to nordiazepam and temazepam glucuronide will be discussed in the Section 3. In this paper, a simple and rapid SPE method using small sample volumes (200 mL) with a low quantification limit ( 3). QC samples were prepared at the concentrations of 1, 50, 500 ng/mL for oxazepam glucuronide and temazepam glucuronide; 0.2, 10, 100 ng/mL for nordiazepam, temazepam and diazepam; and 0.4, 20, 200 ng/mL for oxazepam. Standard curves, freshly prepared with each batch of QC and authentic samples, were generated using a least-squares linear regression, with a 1/x2-weighting factor in Analyst 1.5 software. 2.6.3. Extraction recovery and matrix effect Matrix effect was assessed by the post-extraction addition method described by Matuszewski et al. [21]. It was calculated by comparing the peak areas obtained from standards spiked into the whole blood extracts after samples extraction (B) with mean peak areas of the standards at the same concentrations in the reconstitution solvent (A), and expressed as (B/A 100%). Extraction recovery was calculated by comparing the peak areas of QC samples (C) with the mean peak areas obtained from standards spiked into the whole blood extracts after samples extraction (B), and expressed as (C/B 100%). 2.6.4. Accuracy and precision Accuracy and precision were assayed at three concentrations in the linear dynamic range by preparing and analyzing five replicates on three different days. Intra-assay precision was evaluated by replicate analysis of the QC samples in one run (n = 5). Inter-assay precision was evaluated by replicate analysis of the QC samples in experiments performed on three different days. The precision was expressed as %relative standard deviation (%RSD) of measured concentrations. The accuracy of the assay was expressed by comparing the calculated concentrations of QC samples to their respective nominal values 100%.
2.7. Application of the method to authentic forensic samples of whole blood
3. Results and discussion 3.1. Optimization of sample preparation conditions Protein precipitation with methanol, acetone, acetonitrile and zinc sulfate were tested in this study. Though protein precipitation offers a fast sample preparation protocol with a minimum of manual labor, salts and endogenous compounds in whole blood, the results presented that matrix effect of the protein precipitation method was not good. The 96-well Oasis HLB mElution SPE plate consists of 2 mg of high-capacity SPE sorbent and allows loading of from 10 up to 375 mL of samples and the use of ultralow elution volume as little as 25 mL of elution solvent. Evaporation and reconstitution procedures are eliminated [22]. It was found in this study that, at least half an hour could be reduced by using 96-well mElution plate compared with individual SPE cartridges when one sample was analyzed. Much more time was saved when doing the method validation. It was estimated that, high throughput can be achieved for the application of pharmacokinetics in further study. The impact of different pHs of buffer solution was evaluated as shown in Fig. 1. We can see that when the pH of phosphate buffer was 3.0, peak areas of the analytes were the highest ones. Therefore, 200 mL of whole blood samples and 25 mL of IS were mixed with 400 mL of phosphate buffer (0.2 M, pH 3.0). Experiments were performed to optimize the washing solution to achieve the best clean-up results without loss of the analytes. Washing solvents with different percentage (0%, 2%, 5%) of methanol and different percentage (0%, 1%, 2%) of ammonia in water were evaluated. The results indicated that 200 mL of 5% methanol (with 2% ammonia) in water, followed by 200 mL of purified water was the best ones in terms of matrix effect. In order to optimize the composition of the elution solution, different combinations of methanol/H2O and acetonitrile/methanol/H2O were investigated (Fig. 2). Fig. 2 indicates that an elution
R. Wang et al. / Forensic Science International 233 (2013) 304–311
Fig. 1. Impacts of different pH of phosphate buffer on the peak area of the analytes.
307
and oxazepam-d5 glucuronide were used as the IS for the corresponding non-labeled drugs. Isotopically labeled nordiazepam and temazepam glucuronide were not commercial available for us. If one of diazepam-d5, oxazepam-d5, and temazpam-d5 was utilized as the IS, the concentration values of nordiazepam and temazepam glucuronide were supposed to be influenced a lot in the presence of high levels of diazepam and its free metabolites in blood. Oxazepam-d5 glucuronide was not used as the IS for nordiazepam and temazepam glucuronide either, since the concentration of oxazepam glucuronide might be quite high. It was reported to be equal to or just over oxazepam about 1 h after intake of oxazepam [6]. Thus estazolam-d5 was chosen as the IS for nordiazepam and temazepam glucuronide in this study. The choosing of IS verified that this method was reliable for the determination of diazepam and its metabolites. 3.3. Method validation
solution with 100% methanol achieved the minimum loss of analytes. Although the composition of acetonitrile in the elution solution improved the recovery of diazepam’s free metabolites, it was found to be not good for the recovery of glucuronide metabolites.
3.3.1. Specificity Chromatograms of blank human whole blood showed no endogenous peak co-eluted with analytes. Representative chromatograms of blank sample, spiked sample, and a sample from an authentic forensic case are shown in Fig. 3.
3.2. Choosing of IS
3.3.2. Linearity and sensitivity Statistics of the linearity and sensitivity of the method are shown in Table 2. Correlation coefficients were greater than 0.99 for all analytes, and the method was stable and repeatable (RSD < 0.1%). The linear range was 0.5–1000 ng/mL for oxazepam glucuronide and temazepam glucuronide. The linear range for oxazepam glucuronide was broader than that reported by Simnk et al. (5–1000 ng/mL) [6]. In the present paper, the LOQs were 0.5 ng/mL for oxazepam glucuronide and temazepam glucuronide and LODs were 0.25 ng/mL for oxazepam glucuronide and temazepam glucuronide, which was about ten times more sensitive than that reported by Simnk et al. [6]. The LOQs were 0.2 ng/mL for oxazepam, and 0.1 ng/mL for diazepam, nordiazepam and temazepam. Compared to most previous literatures [29–31], the LOQs obtained in this study for diazepam and its free metabolites in whole blood were much lower.
In previous studies, a labeled isotope of the analyte was usually added as the IS for the compensation of loses during extraction, variations in instrumental response and matrix effects. However, several literatures [23–28] have reported that the peak area of the labeled isotope IS decreases with the increasing of concentration of co-eluting non-labeled analyte in a calibration series, which is congruent with the phenomenon we observed. That means if the isotopically labeled IS of compound A was also used as the IS for compound B in the presence of co-eluting compound A at high concentration, quantification errors may arise from ionization suppression by compound A (for example, when diazepam-d5 was used as the IS for both diazepam and nordiazepam in the presence of diazepam at high concentration, quantification errors for nordiazepam may arise from ionization suppression produced by diazepam). From our previous data obtained from healthy volunteers, the concentration of diazepam in urine was low, thus it is appropriate and reliable to choose diazepam-d5 as the IS for other compounds in urine. However, it was supposed that, there were high levels of diazepam and its free metabolites in whole blood. In this study, diazepam-d5, oxazepam-d5, temazepam-d5
3.3.3. Extraction recovery and matrix effect Extraction recovery and matrix effect of all analytes at three concentrations are summarized in Table 3. The extraction recoveries were from 31% to 80% for all analytes which were acceptable because the method was sensitive enough. The matrix effects were in the range of 52–167%. Ion suppression and enhancement was consistent (RSD < 10%). 3.3.4. Accuracy and precision Results of intra- and inter-day accuracy and precision studies are presented in Table 4. Accuracy of the assay was in the range of 93–108% (intra-day) and in the range of 95–107% (inter-day). Relative standard deviations were from 0.5% to 9.9% (intra-day), and from 0.6% to 7.2% (inter-day), respectively. The intra-day and inter-day RSD and accuracy for all analytes were within the acceptable limit (
Contents lists available at ScienceDirect
Forensic Science International journal homepage: www.elsevier.com/locate/forsciint
Direct determination of diazepam and its glucuronide metabolites in human whole blood by mElution solid-phase extraction and liquid chromatography–tandem mass spectrometry Rong Wang a,1, Xin Wang a,b,1, Chen Liang a, Chunfang Ni a, Lingjuan Xiong a,b, Yulan Rao c,*, Yurong Zhang a,** a
Shanghai Institute of Forensic Science, Shanghai Key Laboratory of Crime Scene Evidence, Shanghai 200083, China Shanghai Institute of Pharmaceutical Industry, State Key Laboratory of New Drug and Pharmaceutical Process, Shanghai 200040, China c Department of Forensic Medicine (Center of Forensic Science), School of Basic Medical Sciences, Fudan University, Shanghai 200032, China b
A R T I C L E I N F O
A B S T R A C T
Article history: Received 8 June 2013 Received in revised form 25 September 2013 Accepted 5 October 2013 Available online 16 October 2013
A mElution solid-phase extraction (SPE) liquid chromatography–tandem mass spectrometry (LC–MS/ MS) method for simultaneous determination of diazepam, nordiazepam, oxazepam, oxazepam glucuronide, temazepam and temazepam glucuronide in human whole blood is presented. 200 mL of whole blood samples were loaded onto a Waters Oasis HLB 96-well mElution SPE plate using 75 mL of methanol as the elution solvent, and the eluents were injected into an Eclipse XDB C18 column. No hydrolysis, solvent transfer, evaporation or reconstitution was involved in the sample preparation procedures. Tandem mass spectrometric detection with Turbo Ion Spray was conducted via multiple reaction monitoring (MRM) under positive ionization mode. The method was validated and proved to be accurate (accuracy within 93–108%), precise (intra-day RSD < 9.9% and inter-day RSD < 7.2%) and sensitive with limits of detection (LOD) in the range of 0.05–0.25 ng/mL for all the compounds. Extraction recoveries were in the range of 31–80% for all the analytes. This method demonstrated to be reproducible and reliable. The applicability of the method was demonstrated by analysis of several forensic cases involving diazepam and its metabolites. ß 2013 Elsevier Ireland Ltd. All rights reserved.
Keywords: Diazepam Diazepam glucuronides 96-Well mElution plate solid-phase extraction Human whole blood LC–MS/MS
1. Introduction Diazepam is widely prescribed for anxity, sleep disorders, or convulsive attacks and also readily available [1]. It is frequently encountered in forensic cases due to misuse or abuse by drug addicts related to suicide attempts, road traffic offenses, drugfacilitated sexual assault or robbery [1–4]. This situation is often complicated by the fact that diazepam may induce coma so that the victims cannot describe what have happened, which would be the obstacle for crime investigation. What is more, drug interaction with diazepam probably potentiates certain side effects [5] which might be life-threatening. Thus, identification of diazepam ingestion is important, which would provide clues
* Corresponding author at: 138 Yi Xue Yuan Road, Shanghai 200032, China. Tel.: +86 21 54237403; fax: +86 21 54237404. ** Corresponding author at: 803 Zhong Shan Bei Yi Road, Shanghai 200083, China. Tel.: +86 21 22028901; fax: +86 21 65443131. E-mail addresses: [email protected] (Y. Rao), [email protected], [email protected] (Y. Zhang). 1 Rong Wang and Xin Wang contributed to the work equally and should be regarded as co-first authors. 0379-0738/$ – see front matter ß 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.forsciint.2013.10.004
to the police for case investigation and the result would be served as evidence at the court. It was estimated that the detection time of the glucuronide metabolites of diazepam (oxazepam glucuronide and temazepam glucuronide) might be longer than diazepam after ingestion of diazepam based on the fact that one literature [6] mentioned that the half-time of oxazepam glucuronide was longer than that of oxazepam in blood after administration of oxazepam. Meanwhile, information on diazepam/glucuronides ratios may be helpful to find out the dose [7] and the time elapsed after diazepam administration which will be conducive to judge crimes. One study [8] has demonstrated that interval time from doping to death in blood of fatally intoxicated heroin addicts could be concluded by the concentration and molar ratios (morphine-3-glucuronide (M3G)/morphine (M), morphine-6-glucuronide (M6G)/M, M3G/ M6G). Accordingly, approximate time from administration to testing point could be speculated if relationship between diazepam/glucuronides ratios and the time elapsed after diazepam administration could be summarized. Based on these facts, a conclusion could be made that, it is necessary to develop a method for the analysis of diazepam and its glucuronide metabolites in biological specimen.
R. Wang et al. / Forensic Science International 233 (2013) 304–311
Although urine is the preferred biological sample in the aspect of prolonging the detection time of drugs [9], concentrations of compounds in urine can be influenced significantly by individual difference or water consumption difference. Compared with urine, the concentrations of drugs and their metabolites in blood are directly related to actual impairment. The values are consistent and can be compared valuably with the levels reported in previous literatures. Meanwhile blood is a valuable specimen to reflect very recent intake of drugs [9], since certain drugs being tested often disappear from the blood fairly rapidly. Thus blood is generally considered to be the most useful sample for forensic investigation especially when urine is usually not available in autopsy. Compared with vast number of papers describing how to determine diazepam and its free metabolites in human blood [10–12], there are much less methods reported for the determination of oxazepam glucuronide and temazepam glucuronide. Only Simnk et al. [6] proposed a liquid chromatography–tandem mass spectrometry (LC–MS/MS) method for the simultaneous determination of oxazepam and oxazepam glucuronide in blood, serum and fluid after administration of oxazepam. Several literatures presented high performance liquid chromatography (HPLC) methods for the separation and quantification of diastereoisomeric glucuronides of oxazepam [13–15] or both diastereoisomers of oxazepam and temazepam glucuronides in plasma [16]. However, these papers have some limitations as their methods all involved enzymatic hydrolysis using b-glucuronidease, which have several critical problems [3,7,17]. The objective of this study was to determine diazepam, nordiazepam, oxazepam, oxazepam glucuronide, temazepam and temazepam glucuronide in human whole blood. The reason why whole blood was analyzed was in forensic cases, plasma or serum often could not be separated from whole blood as the blood has been hemolyzed already. Since plasma and serum is less complicated than whole blood, it is expected that, the method developed in this study using whole blood as the sample could also be applied to plasma or serum. Referring to our previous study [18], a direct LC–MS/MS method was reported for the analysis of diazepam, its free metabolites and its glucuronide metabolites in human urine without hydrolysis. In that study, individual SPE cartridges were utilized, while 96-well Oasis HLB mElution plate was chosen in this study. The postextraction solvent transfer, evaporation and reconstitution steps that are required in individual SPE cartridges are avoided due to the concentrating ability of the plate. This method reduced the labor cost and was more time-saving in sample preparation. Another difference of this study with the method we recently reported was the choosing of internal standards (IS). In both studies, isotope labeled IS were used to manage the influence of ion suppression or enhancement from endogenous components. In our previous study, diazepam-d5 was chosen as the IS for all compounds. While in this assay we chose diazepam-d5, oxazepam-d5, temazepam-d5 and oxazepam-d5 glucuronide as the IS for the corresponding non-labeled drugs. Stable isotope labeled nordiazepam and temazepam glucuronide were not commercial available for us, and instead estazepam-d5 was used as the IS for nordiazepam and temazepam glucuronide. Why none of diazepam-d5, oxazepam-d5, temazepam-d5 and oxazepam-d5 glucuronide was used as the IS to nordiazepam and temazepam glucuronide will be discussed in the Section 3. In this paper, a simple and rapid SPE method using small sample volumes (200 mL) with a low quantification limit ( 3). QC samples were prepared at the concentrations of 1, 50, 500 ng/mL for oxazepam glucuronide and temazepam glucuronide; 0.2, 10, 100 ng/mL for nordiazepam, temazepam and diazepam; and 0.4, 20, 200 ng/mL for oxazepam. Standard curves, freshly prepared with each batch of QC and authentic samples, were generated using a least-squares linear regression, with a 1/x2-weighting factor in Analyst 1.5 software. 2.6.3. Extraction recovery and matrix effect Matrix effect was assessed by the post-extraction addition method described by Matuszewski et al. [21]. It was calculated by comparing the peak areas obtained from standards spiked into the whole blood extracts after samples extraction (B) with mean peak areas of the standards at the same concentrations in the reconstitution solvent (A), and expressed as (B/A 100%). Extraction recovery was calculated by comparing the peak areas of QC samples (C) with the mean peak areas obtained from standards spiked into the whole blood extracts after samples extraction (B), and expressed as (C/B 100%). 2.6.4. Accuracy and precision Accuracy and precision were assayed at three concentrations in the linear dynamic range by preparing and analyzing five replicates on three different days. Intra-assay precision was evaluated by replicate analysis of the QC samples in one run (n = 5). Inter-assay precision was evaluated by replicate analysis of the QC samples in experiments performed on three different days. The precision was expressed as %relative standard deviation (%RSD) of measured concentrations. The accuracy of the assay was expressed by comparing the calculated concentrations of QC samples to their respective nominal values 100%.
2.7. Application of the method to authentic forensic samples of whole blood
3. Results and discussion 3.1. Optimization of sample preparation conditions Protein precipitation with methanol, acetone, acetonitrile and zinc sulfate were tested in this study. Though protein precipitation offers a fast sample preparation protocol with a minimum of manual labor, salts and endogenous compounds in whole blood, the results presented that matrix effect of the protein precipitation method was not good. The 96-well Oasis HLB mElution SPE plate consists of 2 mg of high-capacity SPE sorbent and allows loading of from 10 up to 375 mL of samples and the use of ultralow elution volume as little as 25 mL of elution solvent. Evaporation and reconstitution procedures are eliminated [22]. It was found in this study that, at least half an hour could be reduced by using 96-well mElution plate compared with individual SPE cartridges when one sample was analyzed. Much more time was saved when doing the method validation. It was estimated that, high throughput can be achieved for the application of pharmacokinetics in further study. The impact of different pHs of buffer solution was evaluated as shown in Fig. 1. We can see that when the pH of phosphate buffer was 3.0, peak areas of the analytes were the highest ones. Therefore, 200 mL of whole blood samples and 25 mL of IS were mixed with 400 mL of phosphate buffer (0.2 M, pH 3.0). Experiments were performed to optimize the washing solution to achieve the best clean-up results without loss of the analytes. Washing solvents with different percentage (0%, 2%, 5%) of methanol and different percentage (0%, 1%, 2%) of ammonia in water were evaluated. The results indicated that 200 mL of 5% methanol (with 2% ammonia) in water, followed by 200 mL of purified water was the best ones in terms of matrix effect. In order to optimize the composition of the elution solution, different combinations of methanol/H2O and acetonitrile/methanol/H2O were investigated (Fig. 2). Fig. 2 indicates that an elution
R. Wang et al. / Forensic Science International 233 (2013) 304–311
Fig. 1. Impacts of different pH of phosphate buffer on the peak area of the analytes.
307
and oxazepam-d5 glucuronide were used as the IS for the corresponding non-labeled drugs. Isotopically labeled nordiazepam and temazepam glucuronide were not commercial available for us. If one of diazepam-d5, oxazepam-d5, and temazpam-d5 was utilized as the IS, the concentration values of nordiazepam and temazepam glucuronide were supposed to be influenced a lot in the presence of high levels of diazepam and its free metabolites in blood. Oxazepam-d5 glucuronide was not used as the IS for nordiazepam and temazepam glucuronide either, since the concentration of oxazepam glucuronide might be quite high. It was reported to be equal to or just over oxazepam about 1 h after intake of oxazepam [6]. Thus estazolam-d5 was chosen as the IS for nordiazepam and temazepam glucuronide in this study. The choosing of IS verified that this method was reliable for the determination of diazepam and its metabolites. 3.3. Method validation
solution with 100% methanol achieved the minimum loss of analytes. Although the composition of acetonitrile in the elution solution improved the recovery of diazepam’s free metabolites, it was found to be not good for the recovery of glucuronide metabolites.
3.3.1. Specificity Chromatograms of blank human whole blood showed no endogenous peak co-eluted with analytes. Representative chromatograms of blank sample, spiked sample, and a sample from an authentic forensic case are shown in Fig. 3.
3.2. Choosing of IS
3.3.2. Linearity and sensitivity Statistics of the linearity and sensitivity of the method are shown in Table 2. Correlation coefficients were greater than 0.99 for all analytes, and the method was stable and repeatable (RSD < 0.1%). The linear range was 0.5–1000 ng/mL for oxazepam glucuronide and temazepam glucuronide. The linear range for oxazepam glucuronide was broader than that reported by Simnk et al. (5–1000 ng/mL) [6]. In the present paper, the LOQs were 0.5 ng/mL for oxazepam glucuronide and temazepam glucuronide and LODs were 0.25 ng/mL for oxazepam glucuronide and temazepam glucuronide, which was about ten times more sensitive than that reported by Simnk et al. [6]. The LOQs were 0.2 ng/mL for oxazepam, and 0.1 ng/mL for diazepam, nordiazepam and temazepam. Compared to most previous literatures [29–31], the LOQs obtained in this study for diazepam and its free metabolites in whole blood were much lower.
In previous studies, a labeled isotope of the analyte was usually added as the IS for the compensation of loses during extraction, variations in instrumental response and matrix effects. However, several literatures [23–28] have reported that the peak area of the labeled isotope IS decreases with the increasing of concentration of co-eluting non-labeled analyte in a calibration series, which is congruent with the phenomenon we observed. That means if the isotopically labeled IS of compound A was also used as the IS for compound B in the presence of co-eluting compound A at high concentration, quantification errors may arise from ionization suppression by compound A (for example, when diazepam-d5 was used as the IS for both diazepam and nordiazepam in the presence of diazepam at high concentration, quantification errors for nordiazepam may arise from ionization suppression produced by diazepam). From our previous data obtained from healthy volunteers, the concentration of diazepam in urine was low, thus it is appropriate and reliable to choose diazepam-d5 as the IS for other compounds in urine. However, it was supposed that, there were high levels of diazepam and its free metabolites in whole blood. In this study, diazepam-d5, oxazepam-d5, temazepam-d5
3.3.3. Extraction recovery and matrix effect Extraction recovery and matrix effect of all analytes at three concentrations are summarized in Table 3. The extraction recoveries were from 31% to 80% for all analytes which were acceptable because the method was sensitive enough. The matrix effects were in the range of 52–167%. Ion suppression and enhancement was consistent (RSD < 10%). 3.3.4. Accuracy and precision Results of intra- and inter-day accuracy and precision studies are presented in Table 4. Accuracy of the assay was in the range of 93–108% (intra-day) and in the range of 95–107% (inter-day). Relative standard deviations were from 0.5% to 9.9% (intra-day), and from 0.6% to 7.2% (inter-day), respectively. The intra-day and inter-day RSD and accuracy for all analytes were within the acceptable limit (
