Abstract
Ethanol is the most widely used and abused drug. While blood is the preferred specimen for analysis, tissue specimens such as brain serve as alternative specimens for alcohol analysis in post-mortem cases where blood is unavailable or contaminated. A method was developed using headspace gas chromatography with flame ionization detection (HS-GC-FID) for the detection and quantification of ethanol, acetone, isopropanol, methanol and n-propanol in brain tissue specimens. Unfixed volatile-free brain tissue specimens were obtained from the Department of Pathology at Virginia Commonwealth University. Calibrators and controls were prepared from 4-fold diluted homogenates of these brain tissue specimens, and were analyzed using t-butanol as the internal standard. The chromatographic separation was performed with a Restek BAC2 column. A linear calibration was generated for all analytes (mean r2 > 0.9992) with the limits of detection and quantification of 100–110 mg/kg. Matrix effect from the brain tissue was determined by comparing the slopes of matrix prepared calibration curves with those of aqueous calibration curves; no significant differences were observed for ethanol, acetone, isopropanol, methanol and n-propanol. The bias and the CVs for all volatile controls were ≤10%. The method was also evaluated for carryover, selectivity, interferences, bench-top stability and freeze-thaw stability. The HS-GC-FID method was determined to be reliable and robust for the analysis of ethanol, acetone, isopropanol, methanol and n-propanol concentrations in brain tissue, effectively expanding the specimen options for post-mortem alcohol analysis.
Introduction
Ethanol is one of the most commonly found psychoactive substances in clinical and forensic toxicology, particularly in post-mortem analysis, as over-consumption of alcohol induces impairments that often lead to fatal accidents, violent crimes and suicides. A myriad of factors are important to the analysis and interpretation of post-mortem alcohol concentrations including ante-mortem trauma and life-saving treatment, interval between death and analysis, environmental factors, condition or decomposition of the body and types and location of the specimen collected (1). These factors can contribute to post-mortem production of ethanol and other volatile organic compounds (e.g. methanol, n-propanol, n-butanol, acetaldehyde) from microbial fermentation of natural substrates like glucose (2). Since the human body does not store glucose evenly within various fluids and tissues, post-mortem alcohol production is largely site-dependent and the choice of specimen for analysis is crucial to distinguish between post-mortem production and ante-mortem ingestion.
As with the analysis of other drugs of abuse, blood is the preferred specimen for alcohol analysis. Blood concentrations are used as indicators of intoxication in human performance testing such as driving under the influence or in post-mortem cases. Blood alcohol concentration (BAC) is used as the standard to which the alcohol concentrations in other biological specimens are compared (3–12). Alcohol analysis is generally performed either using direct injection gas chromatography (GC) or headspace gas chromatography (HS-GC) with flame ionization detection (FID). Of the two, headspace is preferred over direct injection, because HS-GC takes advantage of the volatility of alcohol and an extraction procedure is not necessary. Injection of the headspace gas also protects the column from non-volatile components in the matrix that would be present using direct injection. Determination of BAC has been extensively studied using both direct injection GC (13–15) and HS-GC (15, 16). Although blood is the preferred specimen, there are occasions when blood is unavailable or contaminated due to traumatic injuries or there is suspected post-mortem alcohol production. In such instances, alternative specimens are utilized. Urine is commonly submitted for post-mortem alcohol analysis as it generally has low risk of microbial contamination and low glucose concentration, reducing the likelihood of post-mortem alcohol production (1). Similarly, vitreous humor is used by virtue of its low glucose concentration and anatomic location, isolated from the spread of bacteria in the gut during decomposition (1). Both direct injection GC and HS-GC have been applied to urine and vitreous humor alcohol analysis (3, 4, 14, 17, 18). Tissue specimens, though less often, have also been analyzed for alcohols using both GC techniques. One study investigated post-mortem ethanol formation in unadulterated kidney and muscle tissue specimens using HS-GC (19). Skeletal muscle alcohol concentration has been determined from 1:4 diluted homogenate samples using direct injection GC; the muscle to blood alcohol ratio found was on average 0.94–1.48 (5). Liver tissue as a matrix has been analyzed using HS-GC, and the liver to blood alcohol ratio found was on average 0.47–0.85 (6).
Another alternative specimen for alcohol analysis is brain tissue. Where blood is easily lost due to ante-mortem trauma, the brain may be preserved due to its encasement in the protective skull. The isolated location and lack of glucose storage also make the brain an attractive specimen for analysis as it is less susceptible to post-mortem alcohol diffusion, such as from the stomach to heart blood (20) and post-mortem alcohol formation (2). Whereas there is concern over differences in alcohol concentration in post-mortem blood collected from various sites (2), the regional distribution of alcohol in the brain does not differ significantly (7). As a highly vascularized tissue with a rich blood supply, the brain displays rapid alcohol equilibrium with blood (8) and thus may be a good indicator of ante-mortem BAC. In cases where there is an acute subdural hemorrhage from head trauma and the victim survives hours before death, peripheral blood may become unreliable due to ongoing metabolism, yet ethanol concentration in brain tissues underneath the subdural hematoma may still reflect the concentration at the time of injury (9). Though the value of brain as an alternative specimen for post-mortem alcohol analysis cannot be overstated, unlike other specimens, there have been few studies performed to analyze alcohol concentrations in brain. Human brain alcohol concentrations, from aliquots of steam-distilled brain, have been found to correlate strongly with BAC (brain–blood ratio of 0.64–1.20) using direct injection GC (10). A brain–blood ratio of 0.80–1.50 (from 82% of the values within the study) was observed with homogenization, extraction and analysis via direct injection GC (11). Thus it seems that different extraction techniques for brain alcohol determination can substantially affect the measured ratios. There is only one published study that examined the reliability of the HS-GC method for the analysis of ethanol in brain tissue specimen. Distillation and direct tissue homogenate HS methods were compared and no significant difference was found. The presence of tissue matrix did not interfere with the determination of ethanol in brain. The brain–blood ratios obtained using both distillation and direct tissue homogenate HS methods were 0.97 and 1.32 for BACs of 0.1 and 0.2 g/dL, respectively (12).
In the present study, we developed a more comprehensive HS-GC-FID method for the detection and quantification of ethanol and other volatiles including acetone, isopropanol, methanol and n-propanol, in brain tissue specimens using t-butanol as the internal standard (ISTD). The method employs direct HS analysis of tissue homogenate with 4-fold dilution to help with matrix matching. The calibration model, limits of detection and quantification (LOD/LOQ), matrix effect, bias and precision, carryover, selectivity, interference and stability are presented.
Methods
Materials
Volatile-free human brain tissues were obtained from the Department of Pathology at Virginia Commonwealth University (Richmond, VA, USA). Two hundred proof (absolute grade) ethanol was obtained from The Warner-Graham Company (Cockeysville, MD, USA). Acetone (Optima grade), isopropanol (Optima grade), methanol (Optima grade), t-butanol (≥99.5%), acetonitrile (Optima grade), benzene (Spectrophotometric grade), 1-butanol (Optima grade), chloroform (Optima grade), ethyl acetate (Optima grade), n-heptane (HPLC grade), n-hexane (Optima grade), methylene chloride (Optima grade) and iso-octane (Optima grade) were purchased from Thermo Fisher Scientific, Inc. (Fair Lawn, NJ, USA). n-Propanol (Optima grade) was purchased from OmniSolv (Gibbstown, NJ, USA) and ethyl formate (97%) was obtained from Sigma-Aldrich (Milwaukee, WI, USA). Deionized (DI) water was obtained in-house.
Instrumental analysis
The headspace analysis was performed on a Varian 3900 gas chromatograph with an FID detector (Varian Associates, Inc., Walnut Creek, CA, USA) equipped with a Rtx®–BAC2 column measuring 30 m × 0.53 mm ID × 2.0 μm (Restek Corp., Bellefonte, PA, USA). The oven and detector temperatures were 40 and 235°C, respectively. The carrier gas was helium (16.8 mL/min) and the detector gas was helium makeup (25 mL/min), hydrogen (30 mL/min) and air (300 mL/min). Under these conditions, the retention times of methanol, ethanol, acetone, isopropanol, ISTD and n-propanol were 1.66, 2.16, 2.31, 2.53, 2.80 and 4.05 min, respectively (Figure 1). The autosampler was a Tekmar 7000 headspace autosampler with a Hydroguard MXT® 2 m × 0.53 mm ID (Restek Corp.) sample loop and a Restek guard column. The platen temperature was 80°C with 0.1 min platen equilibrium time, 3.5 min sample equilibration time, 0.2 min mixing time and 0.5 min stabilization time. The sample loop temperature was 160°C, fill time was 0.2 min, equilibrium time was 0.2 min and injection time was 0.3 min.
Figure 1.
The chromatographic separation of ethanol, acetone, isopropanol, methanol, n-propanol and t-butanol and the ISTD.
Method validation
The evaluation of the brain tissue assay was conducted over five separate days. The sample batches were analyzed as recommended for bioassay validation (21) for calibration model, LOD/LOQ, matrix effect, bias, precision, carryover, selectivity, interference and stability. Validation sample batches contained calibrators in duplicate, a volatile-free control with ISTD added, and a double negative control containing neither volatiles nor ISTD in aliquots of the brain tissue homogenate specimens. Quality control (QC) specimens were analyzed in triplicate on 4 days and for the intraday precision study, in replicates of six on a single day yielding a total of 18 data points for each of the following QC specimens containing ethanol, acetone, isopropanol, methanol and n-propanol: low control (LQC), target concentration of 300 mg/kg for n-propanol and 250 mg/kg for all other analytes; medium control (MQC), target concentration of 2,000 mg/kg for ethanol and 1,200 mg/kg for all other analytes; high control (HQC), target concentration of 4,000 mg/kg for ethanol and 2,500 mg/kg for all other analytes.
Sample preparation
Aqueous samples were prepared by diluting the volatiles, ethanol, acetone, isopropanol, methanol and n-propanol, with DI water and stored at 5°C. For brain samples, human brain tissue was weighed and diluted 1:4 with DI water and was then homogenized using a Dremel 985370 homogenizer (BioSpec Products, Inc., Bartlesville, OK, USA). Samples were then prepared using the homogenate with the same volatiles, as received for the controls and from 5% stock solutions for the calibrators, and stored at 5°C. The ISTD solution (1,580 mg/kg) was prepared by diluting t-butanol with DI water. The ISTD working solution (160 mg/kg) was prepared similarly from the stock solution. The ISTD solutions were stored at 5°C. 0.4 mL of calibrators, controls or brain homogenate and 0.9 mL of the ISTD working solution were introduced into a 22 mL autosampler vial. Vials were then capped and placed on the HS-GC-FID for analysis.
Calibration model
Calibration models were determined from six-point calibration curves with calibrators prepared in duplicate in homogenized brain tissue. Concentrations were 100, 500, 790, 1,580, 3,160 and 4,740 mg/kg for ethanol; 100, 190, 380, 760, 1,580 and 2,980 mg/kg for acetone; 100, 120, 380, 750, 1,570 and 3,020 mg/kg for isopropanol; 100, 130, 380, 760, 1,580 and 2,980 mg/kg for methanol and 110, 140, 400, 790, 1,580 and 3,010 mg/kg for n-propanol. A linear regression of the ratio of the peak area counts of the analyte and ISTD versus the ratio of the concentrations of the analyte and ISTD was used to construct the calibration curves for each analyte. The average coefficient of determination (r2) (n = 10) and the standardized residual plot, generated using R Studio statistical software, were used to check for outliers (outside ± 3 SD) and assess the fit of the chosen model for each analyte.
Limit of detection and limit of quantification
The LOD and LOQ were administratively set at and verified with the lowest calibrator prepared in triplicate in three lots of brain tissue homogenates. Bias and precision were calculated (n = 9) for each analyte and considered acceptable if the bias was within ±10% and the precision was ≤10%.
Matrix effect
Linear aqueous calibration curves were generated from six-point calibration curves with calibrators prepared in duplicate in DI water. Concentrations of the analytes were the same as with the matrix prepared calibrators. The average slope (n = 10) of the aqueous calibration curves were compared to those of the matrix prepared calibration curves for determination of differences, if any.
Bias and precision
Bias and precision were determined from the prepared volatile QC brain tissue samples. The bias was determined for each analyte at each concentration using the QC values (n = 15) over the five validation runs. The between-run and within-run % coefficient of variation (CV) for each analyte at each concentration were calculated (n = 3 in each group, n = 15 total) with the one-way ANOVA approach as outlined in the validation guideline (21), using R Studio statistical software. Bias was considered acceptable if the values were within ±10%, and % CV was considered acceptable if the values were ≤10%.
Carryover
The sample carryover was evaluated in each of the five validation batches by analyzing a volatile-free matrix control immediately following the analysis of the highest matrix prepared calibrator (4,740 mg/kg for ethanol, 2,980 mg/kg for acetone and methanol, 3,020 mg/kg for isopropanol and 3,010 mg/kg for n-propanol). A conclusion of no carryover was determined if the concentrations of the analytes were below the LOD in the negative control.
Selectivity
The selectivity of the assay was determined using 10 different lots each containing two brains diluted 1:4 with DI water and homogenized. Each individual lot was analyzed with and without the ISTD. A conclusion of no endogenous interference was determined if no peaks were detected that co-eluted with the analytes or the ISTD.
Interference
Common potential interferences of ethanol like acetone, isopropanol, methanol and n-propanol were included as analytes in the assay validation. The following compounds were also run as interferences: acetonitrile, benzene, 1-butanol, chloroform, ethyl acetate, ethyl formate, n-heptane, n-hexane, methylene chloride and iso-octane.
Stability
The stability of the analytes in brain tissues was determined under several specific conditions and different time intervals. The studies were performed using three of the volatile control specimens: LQC, MQC and HQC. The bench-top stability of the analytes in brain tissue at room temperature was assessed to evaluate the possible effects of specimen transportation and processing delay in the laboratory by having the QC specimens sit at room temperature for 72 h. The bench-top stability samples were analyzed in four replicates (n = 4). The freeze-thaw stability of the analytes in brain tissue was evaluated as brain tissue specimens are often stored frozen and then thawed for re-analysis. The stability of the analytes in brain tissue was determined for each of the three freeze-thaw cycles. The QC specimens were stored at −20°C for 24 h, allowed to thaw unassisted, and then analyzed. The process was repeated twice more for the 48 and 72 h cycles. All freeze-thaw stability samples were analyzed in triplicate (n = 3). The results of all stability samples were compared with those of the time zero responses taken from the first run of triplicate QC specimens (n = 3) in the bias and precision study. Analytes were considered stable under the conditions of the bench-top and freeze-thaw stability studies if the concentrations of the QC samples were within ±10% of the time zero average concentrations.
Results
The mean slope and the mean r2 of the matrix prepared calibration curves in the five batches indicated a good fit of the linear model for all analytes (Table I). The linear model was further confirmed for each analyte with a standardized residual plot generated from the pooled data. No outliers (>±3 SD) were detected for ethanol, isopropanol, methanol and n-propanol. One outlier data point was detected for acetone for the highest concentration ratio. The outlier was removed and found to have no significant impact, only a slight increase of 0.0004 on the mean slope and r2 of acetone. Replicates of the lowest calibrator in three lots of brain homogenates verified that the LOD/LOQ for ethanol, acetone, isopropanol and n-propanol was within ±10% of the target values. A deviation of 16% was observed for the LOD/LOQ of methanol. The mean slopes of the matrix prepared calibration curves were compared to those of the aqueous prepared calibration curves (Table I), and no significant differences were observed, indicating the absence of matrix effect. Accuracy/bias as well as between-run and within-run precisions (Table II) of the assay were determined not to exceed ±10% over the dynamic range of the assay for all analytes. A lack of carryover was confirmed as the analytes were not detected or present below the LOD in the negative control. The selectivity of the assay was also confirmed as no endogenous interferences were detected in any of the 10 brain homogenates analyzed. A study of potential exogenous interferences was carried out and the compounds and their elution times were as follows: acetonitrile (3.31 min), benzene (4.34 min), 1-butanol (>4.50 min), chloroform (4.23 min), ethyl acetate (3.60 min), ethyl formate (2.14 min), n-heptane (2.95 min), n-hexane (1.77 min), methylene chloride (2.42 min) and iso-octane (2.71 min). Ethyl formate was found to co-elute with ethanol (2.16 min), methylene chloride with acetone (2.31 min), and n-heptane and iso-octane with the ISTD (2.80 min). All analytes in the QC samples, except acetone in the LQC, were determined to be stable in brain tissue at room temperature for 72 h (Table III). Ethanol, acetone, isopropanol and n-propanol in all the QC samples were considered stable for the 24, 48 and 72 h freeze-thaw cycles (Figure 2). Methanol was considered stable in the MQC and HQC samples but not the LQC sample, as the concentration fell below 10% of the time zero mean concentration.
Table I.
Calibration models of ethanol, acetone, isopropanol, methanol and n-propanol in brain tissue and DI water
| Brain tissue | DI water | |||
|---|---|---|---|---|
| Volatile | Mean slope ± SDa | Mean r2 ± SDa | Mean slope ± SDa | Mean r2 ± SDa |
| Ethanol | 0.0208 ± 0.0004 | 0.9996 ± 0.0003 | 0.0228 ± 0.0006 | 0.9995 ± 0.0007 |
| Acetone | 0.0731 ± 0.0022 | 0.9992 ± 0.0002 | 0.0840 ± 0.0014 | 0.9996 ± 0.0003 |
| Isopropanol | 0.0423 ± 0.0008 | 0.9994 ± 0.0002 | 0.0477 ± 0.0009 | 0.9998 ± 0.0002 |
| Methanol | 0.0109 ± 0.0003 | 0.9992 ± 0.0004 | 0.0117 ± 0.0004 | 0.9995 ± 0.0005 |
| n-Propanol | 0.0362 ± 0.0007 | 0.9997 ± 0.0002 | 0.0398 ± 0.0009 | 0.9997 ± 0.0002 |
an = 10.
Table II.
Accuracy/bias and precision of ethanol, acetone, isopropanol, methanol and n-propanol in brain tissue determined with prepared QC specimens
| Volatile | Control | Mean concentration ± SD (mg/kg)a | Bias (%)a | Between-run CV (%)a | Within-run CV (%)a |
|---|---|---|---|---|---|
| Ethanol | LQC (250 mg/kg) | 274 ± 12 | 10 | 3 | 4 |
| MQC (2,000 mg/kg) | 1,988 ± 37 | 1 | 1 | 2 | |
| HQC (4,000 mg/kg) | 4,264 ± 49 | 7 | 1 | 1 | |
| Acetone | LQC (250 mg/kg) | 249 ± 7 | 0.2 | 2 | 2 |
| MQC (1,200 mg/kg) | 1,145 ± 28 | 5 | 1 | 2 | |
| HQC (2,500 mg/kg) | 2,537 ± 57 | 1 | 1 | 2 | |
| Isopropanol | LQC (250 mg/kg) | 244 ± 5 | 2 | 1 | 2 |
| MQC (1,200 mg/kg) | 1,218 ± 19 | 1 | 1 | 2 | |
| HQC (2,500 mg/kg) | 2,515 ± 28 | 1 | 1 | 1 | |
| Methanol | LQC (250 mg/kg) | 276 ± 18 | 10 | 4 | 2 |
| MQC (1,200 mg/kg) | 1,155 ± 29 | 4 | 1 | 2 | |
| HQC (2,500 mg/kg) | 2,533 ± 38 | 1 | 1 | 1 | |
| n-propanol | LQC (300 mg/kg) | 283 ± 6 | 6 | 1 | 2 |
| MQC (1,200 mg/kg) | 1,145 ± 20 | 5 | 1 | 2 | |
| HQC (2,500 mg/kg) | 2,415 ± 26 | 3 | 1 | 1 |
an = 15.
Table III.
Bench-top stability of ethanol, acetone, isopropanol, methanol and n-propanol in brain tissue determined with prepared QC specimens
| Volatile | Control | Time zero mean concentration ± SD (mg/kg)a | 72 h mean concentration ± SD (mg/kg)a | Bias (%) |
|---|---|---|---|---|
| Ethanol | LQC | 267 ± 1 | 248 ± 20 | 7 |
| MQC | 1,988 ± 9 | 1,903 ± 9 | 4 | |
| HQC | 4,261 ± 32 | 4,151 ± 118 | 3 | |
| Acetone | LQC | 252 ± 5 | 219 ± 8 | 13 |
| MQC | 1,130 ± 19 | 1,056 ± 16 | 7 | |
| HQC | 2,530 ± 74 | 2,406 ± 89 | 5 | |
| Isopropanol | LQC | 244 ± 4 | 226 ± 9 | 7 |
| MQC | 1,224 ± 7 | 1,172 ± 19 | 4 | |
| HQC | 2,525 ± 28 | 2,462 ± 77 | 3 | |
| Methanol | LQC | 310 ± 7 | 293 ± 51 | 5 |
| MQC | 1,188 ± 12 | 1,181 ± 75 | 1 | |
| HQC | 2,507 ± 20 | 2,600 ± 37 | 4 | |
| n-Propanol | LQC | 284 ± 3 | 272 ± 13 | 4 |
| MQC | 1,149 ± 8 | 1,102 ± 21 | 4 | |
| HQC | 2,423 ± 25 | 2,356 ± 74 | 3 |
an = 4.
Figure 2.
Results of the freeze-thaw stability of (A) LQC, (B) MQC and (C) HQC samples.
Discussion
There is a dearth of published analytical method for the analysis of ethanol and other volatile compounds in brain tissue using HS-GC-FID. The presented method used t-butanol as the ISTD and a simple dilution of tissue homogenates. The ISTD t-butanol was chosen because it is not a post-mortem artifact like n-propanol, and it is widely used (9, 12, 15, 19). The method demonstrated acceptable accuracy and reproducibility for the detection and quantification of ethanol, acetone, isopropanol, methanol and n-propanol. A linear calibration model was determined to be a good fit for all analytes according to evaluations of the r2 values, all means are 0.9992 or better, and of the standardized residual plots, all showing no outliers or no significant effect from outliers. A linear model for ethanol is consistent with that obtained by Bonventre et al. (12) using another HS-GC-FID method. The dynamic range evaluated for ethanol by Bonventre et al. (12) was from 500 to 4,000 mg/kg. In this study, the range was extended and the LOD/LOQ set administratively at 100 mg/kg for ethanol, acetone, isopropanol and methanol, and 110 mg/kg for n-propanol. The only LOD/LOQ bias over 10% was for methanol. This was likely due to interference from an unidentified compound (1.68 min; methanol 1.66 min) generated with the ISTD degrading over time. Bonventre et al. (12) reported no significant difference between results obtained from brain distillate and direct brain tissue homogenate using HS-GC, which suggested the matrix components of brain tissue did not interfere with ethanol determination. With the use of non-matrix-matched standards, however, the effect of the brain matrix must be evaluated to determine appropriateness. In this study, calibrators in aqueous medium were used to compare with calibrators prepared from 4-fold aqueous diluted brain tissue homogenates. No significant matrix effect was observed from comparison of the slopes generated. This was likely due to the dilution compensating for any potential differences in the liquid/air partition coefficients. The values for bias and precision were all acceptable, indicating accuracy and reproducibility of this method. Additionally, the assay was free from significant analyte carryover and free of significant interference from the brain tissue matrix. Interferences from some volatile organic compounds such as ethyl formate, methylene chloride, n-heptane and iso-octane were observed, but as these compounds are rarely encountered in significant quantity in post-mortem cases, they are unlikely to interfere with the determination of the analytes. Results from the stability studies indicate that samples should be analyzed before 72 h of storage at room temperature, as acetone can become unstable, and samples should be analyzed before 24 h of storage in the freezer, as methanol can become unstable with extended freeze-thaw time.
Conclusion
An HS-GC-FID method for the detection and quantification of ethanol, acetone, isopropanol, methanol and n-propanol in brain tissue was developed. A linear calibration model was found to be appropriate for all analytes and no significant matrix effect was observed when compared to aqueous calibration. The method was determined to be both reliable and robust in post-mortem alcohol analysis in brain tissue specimen.
Funding
This work was supported in part by the National Institute of Health (Grant P30DA033934).
References
- 1.Kugelberg F.C., Jones A.W. (2007) Interpreting results of ethanol analysis in postmortem specimens: a review of the literature. Forensic Science International, 165, 10–29. [DOI] [PubMed] [Google Scholar]
- 2.O'Neal C.L., Poklis A. (1996) Postmortem production of ethanol and factors that influence interpretation: a critical review. The American Journal of Forensic Medicine and Pathology: Official Publication of the National Association of Medical Examiners, 17, 8–20. [DOI] [PubMed] [Google Scholar]
- 3.Dural E., Yalçin S., Mergen G., Kayaalti Z., Kaya S., Aliyev V. et al. (2010) Simultaneous headspace-GC-FID analysis for methanol and ethanol in blood, saliva, and urine validation of method and comparison of specimens. LCGC North America, 28, 540–543. [Google Scholar]
- 4.Honey D., Caylor C., Luthi R., Kerrigan S. (2005) Comparative alcohol concentrations in blood and vitreous fluid with illustrative case studies. Journal of Analytical Toxicology, 29, 365–369. [DOI] [PubMed] [Google Scholar]
- 5.Garriott J.C. (1991) Skeletal muscle as an alternative specimen for alcohol and drug analysis. Journal of Forensic Sciences, 36, 60–69. [PubMed] [Google Scholar]
- 6.Jenkins A.J., Levine B.S., Smialek J.E. (1995) Distribution of ethanol in postmortem liver. Journal of Forensic Sciences, 40, 611–613. [PubMed] [Google Scholar]
- 7.Moore K.A., Kunsman G.W., Levine B.S., Herman M.M., Cervenak J., Hyde T.M. (1997) A comparison of ethanol concentrations in the occipital lobe and cerebellum. Forensic Science International, 86, 127–134. [DOI] [PubMed] [Google Scholar]
- 8.Harger R.N., Hulpie H.R., Lamb E.B. (1937) The speed with which various parts of the body reach equilibrium in the storage of ethyl alcohol. The Journal of Biological Chemistry, 120, 689–704. [Google Scholar]
- 9.Takahashi K., Ikeda N., Kudo K., Funayama M. (2001) Forensic significance of concentrations of ethanol in brain tissues following induced acute subdural hemorrhage. International Journal of Legal Medicine, 115, 1–5. [DOI] [PubMed] [Google Scholar]
- 10.Backer R.C., Pisano R.V., Sopher I.M. (1980) The comparison of alcohol concentrations in postmortem fluids and tissues. Journal of Forensic Sciences, 25, 327–331. [PubMed] [Google Scholar]
- 11.Budd R.D. (1983) Post mortem brain alcohol levels. Journal of Chromatography, 259, 353–355. [DOI] [PubMed] [Google Scholar]
- 12.Bonventre J., Valanju S., Bastos M. (1982) Evaluation of ethanol analysis on brain and liver by head-space gas chromatography. Forensic Science International, 19, 75–83. [DOI] [PubMed] [Google Scholar]
- 13.Jain N.C. (1971) Direct blood-injection method for gas chromatographic determination of alcohols and other volatile compounds. Clinical Chemistry, 17, 82–85. [PubMed] [Google Scholar]
- 14.Tangerman A. (1997) Highly sensitive gas chromatographic analysis of ethanol in whole blood, serum, urine, and fecal supernatants by the direct injection method. Clinical Chemistry, 43, 1003–1009. [PubMed] [Google Scholar]
- 15.O'Neal C.L., Wolf C.E., Levine B., Kunsman G., Poklis A. (1996) Gas chromatographic procedures for determination of ethanol in postmortem blood using t-butanol and methyl ethyl ketone as internal standards. Forensic Science International, 83, 31–38. [DOI] [PubMed] [Google Scholar]
- 16.Watts M.T., McDonald O.L. (1987) The effect of biological specimen type on the gas chromatographic headspace analysis of ethanol and other volatile compounds. American Journal of Clinical Pathology, 87, 79–85. [DOI] [PubMed] [Google Scholar]
- 17.Corrêa C.L., Pedroso R.C. (1997) Headspace gas chromatography with capillary column for urine alcohol determination. Journal of Chromatography. B, Biomedical Sciences and Applications, 704, 365–368. [DOI] [PubMed] [Google Scholar]
- 18.Pontes H., Guedes de Pinho P., Casal S., Carmo H., Santos A., Magalhães T. et al. (2009) GC determination of acetone, acetaldehyde, ethanol, and methanol in biological matrices and cell culture. Journal of Chromatographic Science, 47, 272–278. [DOI] [PubMed] [Google Scholar]
- 19.Lewis R.J., Johnson R.D., Angier M.K., Vu N.T. (2004) Ethanol formation in unadulterated postmortem tissues. Forensic Science International, 146, 17–24. [DOI] [PubMed] [Google Scholar]
- 20.Hebold G. (1958) Postmortem diffusion of alcohol through the gastric wall. Deutsche Zeitschrift fur die gesamte gerichtliche Medizin, 47, 619–24. [PubMed] [Google Scholar]
- 21.Scientific Working Group for Forensic Toxicology (2013) Report from the Scientific Working Group for Forensic Toxicology: Scientific Working Group for Forensic Toxicology (SWGTOX) standard practices for method validation in forensic toxicology. Journal of Analytical Toxicology, 37, 452–474. [DOI] [PubMed] [Google Scholar]