Journal of Analytical Toxicology 2012;36:429 433 doi:10.

1093/jat/bks045 Advance Access publication May 11, 2012

Article

Comparison of Relative Distribution of Ketamine and Norketamine in Decomposed Skeletal Tissues Following Single and Repeated Exposures
James H. Watterson*, Joseph P. Donohue and Caroline C. Betit Department of Forensic Science, Laurentian University, 935 Ramsey Lake Rd., Sudbury, Ontario, Canada
*Author to whom correspondence should be addressed. Email: jwatterson@laurentian.ca

Bone was analyzed for ketamine and norketamine to examine whether different patterns of drug exposure could be discriminated. Rats received (intraperitoneally) one 75 mg/kg dose (Acute-1 and Acute-2 groups), three 25-mg/kg doses 1 hour apart (Repeated group), or nine single daily ketamine doses of 75 mg/kg followed by a 24-h washout period (Chronic group). Following euthanasia, all animals decomposed to skeleton outdoors. Ground samples of recovered bone underwent methanolic extraction and analysis by gas chromatography mass spectrometry after solid-phase extraction. Drug levels (mass normalized response ratios) were compared across bone types and exposure pattern. Bone type signicantly inuenced drug level for the Acute-1 and Repeated dose groups, and the drug/metabolite level ratio (DMLR) for the Acute-1 group. Mean ketamine and norketamine level and DMLR varied by up to 8fold, 7-fold and 3-fold, respectively, in the Acute-1 group, and by up to 24-fold, 5-fold and 10-fold, respectively, in the Repeated group. Drug level and DMLR differed signicantly between the Acute-1 and Repeated groups for most bone types. In the Chronic group, only 1/16 and 4/16 samples were positive for ketamine and norketamine, respectively. All Acute-2 samples were positive for ketamine and norketamine. The Acute-2 and Chronic groups differed signicantly in ketamine and norketamine levels, and DMLR.

drug exposure (e.g., acute, high-dose exposure versus repeated exposures at a lower dose), while an absolute drug level measurement may yield a limited amount of information. Here we expand upon the work to date by comparing the relative distribution of KET and NKET in skeletal tissues of rats that underwent different patterns of KET exposure. In the rst part of the study, one group of rats was acutely exposed to an anesthetic dose of KET and then euthanized shortly after exposure. A second group of rats received a smaller dose (3-fold), which was repeated every hour (roughly 2 t1/2 (16)) for a total of three doses. These animals were then euthanized after another 1-hour delay period, in an attempt to facilitate some accumulation of the metabolite. The purpose of this study was to determine whether the different circumstances of exposure could be differentiated by measured levels of KET and NKET, and by the ratio of levels of KET and NKET. In the second part of the study, we report on a comparative analysis of skeletal tissues of rats exposed to acute doses of KET and those of rats that underwent a chronic daily (nine days) exposure to single doses of KET, followed by an extended 24-h washout period, which was undertaken to examine whether these patterns of exposure may be discriminated through toxicological analysis of skeletal remains.

Introduction Despite an increase in the number of studies that have reported detection of various drugs in skeletal tissues following various postmortem and dosing conditions (1 14), interpretation of such measurements remains very complex, because large variability in levels of a given drug and its metabolite(s) has been observed within a given bone (e.g., epiphyseal versus diaphyseal femoral bone (2)) and between bones in a given body (12). Notwithstanding these complexities, some data suggest that drug levels in bone may be dose-dependent (2, 11). Furthermore, recent data have suggested that the sitedependence of assay response may be much lower for the ratio of responses for drug and metabolite(s) than for the absolute responses to any of those compounds alone (1, 12, 14). For example, we published a recent study examining the relative distribution of ketamine (KET) and norketamine (NKET) in skeletal tissues of rats that underwent various degrees of decomposition (1). In that work, mean ketamine and norketamine levels across a range of different bone types varied by as much as 23- and 18-fold, respectively, while the mean ratio of KET to NKET in a given bone type varied by only 5-fold across the tissues examined. Consequently, we proposed that the relative level of parent drug and metabolite(s) might be useful in differentiating between different circumstances of

Methods Chemicals Methanol used in drug extraction was high-performance liquid chromatography (HPLC)-grade and purchased from EMD Chemicals (Gibbstown, NJ) Drug standards (Cerilliant; Round Rock, TX) were obtained as 1 mg/mL methanolic solutions and diluted as required. All other chemicals were reagent grade and were obtained from EMD Chemicals.

Animals and drug administration: acute versus repeated dosingshort term Male Wistar rats (n 10) were divided into two groups. One group (Acute-1; n 5) was acutely exposed to KET (75 mg/kg, i.p.) and euthanized by CO2 asphyxiation within 20 min of drug exposure. Another group (Repeated; n 5) was given three doses of 25 mg/kg ketamine (i.p.), 1 hour apart and euthanized by CO2 asphyxiation within 1 h of the nal drug administration. Doses were chosen because they are nonfatal anesthetic doses in rats. Four drug-free control animals were used as negative controls. After euthanasia, tibiae were excised to examine fresh skeletal tissue levels. All other remains were left to decompose outdoors in Northern Ontario (Sudbury) during midsummer (July August), with full exposure to sunlight and other weather features ( precipitation) for

# The Author [2012]. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com

two weeks, with average daily temperatures ranging from 17 248C. Bones, including vertebrae, pelvi, femora, ulnae and rib were collected from all animals for analysis.

Animals and drug administration: acute versus chronic dosing The purpose of this portion of the work was to determine whether chronic KET exposure would result in detectable KET or NKET after a signicant post-exposure delay. Acute-2 rats (n 8) were dosed with 75 mg/kg KET (i.p.) and euthanized by CO2 asphyxiation 15 min after administration. Chronic rats (n 8) were dosed once per day with 75 mg/kg KET (i.p.) on nine consecutive days, and were euthanized on the 10th day by CO2 asphyxiation, roughly 24 hours after the last dose. The remains of both groups, as well as four drug-free animals, were permitted to decompose, and bones were recovered as outlined previously.

Expression of drug levels As discussed in previous papers (1, 9, 12), accurate measurement of analyte recovery from solid bone tissue of exposed animals is not possible using conventional analytical techniques such as GC MS, because bone is not a homogeneous matrix to which drug standard may be added and completely distributed throughout. Consequently, we maintain the convention that we have used in those previous studies in expressing drug level in a given tissue as the RR, normalized for the mass of tissue

Sample preparation and analysis Bones were separated according to type (e.g., vertebra versus pelvis) and source animal (e.g., Acute Exposure, Animal 1). Bones underwent washing, pulverization and passive methanolic extraction, as outlined previously (1). Bone extracts were recovered and underwent further preparation by solid-phase extraction, using D4-ketamine (D4KET) (200 ng) as an internal standard. Prepared bone extracts then underwent analysis by gas chromatography mass spectrometry (GC MS) (1) in the selected ion monitoring (SIM) mode, where the monitored ions were m/z 180, 167, 152 (KET; bold ion used for quantitative analysis), m/z 166, 165 and 131 (NKET) and m/z 184, 213 and 156 (D4KET). The response ratio (RR) is dened as the ratio of peak area for the quantitation ion (m/z 180 or 166) to that of the ion with m/z 184 corresponding to D4KET. The assay was linear (R2 . 0.99) with respect to both analytes over the concentration range of 10 2,000 ng/mL, with precision fmeasured as the coefcient of variation [CV(%)] in replicate analyses [ni 3, within run] of matrix-matched spiked bone extracts, prepared as described previously [1]g less than 20% over the concentration range of 10 2,000 ng/mL. Limits of detection for both KET and NKET were approximately 5 ng/mL.

Statistical analysis Statistical analysis of data was done using the Kruskall Wallis test, a non-parametric analysis of variance, using StatPlus 2009 software (AnalystSoft Inc., www.analystsoft.com). The Kruskall Wallis test was utilized to compare the variability in extractions of a given bone type to the variability in analytical response across all tissues assayed, to determine whether any signicant tissue-dependent distribution existed. This nonparametric test was chosen because it does not presume that measured values within a group are normally distributed. The Mann Whitney U-test, a non-parametric alternative to the t-test, was used to compare measured values (RR/m for KET and NKET, or RRKET/RRNKET) between drug exposure treatments (i.e., acute versus repeated dosing).
430 Watterson et al.

Figure 1. Summary of data for acute versus repeated dose exposure (Acute-1 versus Repeated), according to bone type. Mean ( + SD) mass-normalized response ratios (RR/m) for KET in various bone types (A); Mean ( + SD) mass-normalized response ratios (RR/m) for NKET in various bone types (B); Mean ( + SD) ratio of levels of KET and NKET (RRKET/RRNKET) in various bone types (C).

sampled (i.e., RR/m). This quantity is proportional to drug concentration (1, 9, 12) in an extract and provides a means to compare relative drug levels between different experimental treatment conditions, such as the dosing regimens used here, so that the analytical consequences of those treatments may be understood. All statistical comparisons of drug or metabolite levels were made using RR/m values. For perspective, RR/m values are also converted to estimates of concentration, but such values may have limited comparative value relative to measurements made in other laboratories.

Table II Summary of RR/m Data for NKET in Various Bone Types NKET (RR/m) Acute-1 mean (SD) Acute-1 range Repeated mean (SD) Repeated range FEM 7.5 (5.8) 3.6 17 1.8 (0.4) 1.2 1.9 VRT 13.2 (12.8) 2.8 35 2.7 (0.7) 1.8 3.5 PEL 2.0 (0.3) 1.7 2.6 1.3 (0.6) 0.9 2.3 HUM 4.3 (1.9) 2.9 7.5 0.9 (0.3) 0.7 1.4 RIB 2.6 (1.3) 2.2 5.9 1.8 (0.5) 1.1 2.3 ULNA 2.6 (1.9) 1.6 6.0 0.5 (0.2) 0.38 0.78

Table III Summary of RRKET/RRNKET Data in Various Bone Types

Results Acute versus repeated doses: short term The mean levels (RR/m) of KET and NKET for the various tissues are summarized in Figures 1A and 1B, and the ratios of levels in each tissue (i.e., RRKET/RRNKET) are summarized in Figure 1C. The data are also summarized in Tables I III. Tissue type was again found to be a primary effect with respect to levels of KET for both treatments (KruskallWallis test, P , 0.05). Despite large variability in bone KET levels within an animal and between animals, comparison of the KET levels between dosing conditions (acute versus repeated) showed signicant differences (MannWhitney, P , 0.05) for all bone types (e.g., vertebrae). Levels of KET corresponded to roughly 2 21 and 2 8 mg/g in bone samples from the Acute-1 and Repeated groups, respectively. Similarly, tissue type was found to be a primary effect with respect to NKET level for both treatments (KruskallWallis test, P , 0.05). For all bone types except pelvic bone, comparison of the NKET levels between dosing conditions (acute versus repeated) showed signicant differences (Mann Whitney, P , 0.05). Levels of NKET corresponded to roughly 0.2 10 and 0.3 2 mg/g in bone samples from the Acute-1 and Repeated groups, respectively. Examination of the ratio of KET and NKET levels (i.e., RRKET/RRNKET) in each sample showed that the ratio varied by factors similar to those observed with KET levels. Tissue type was found to be a primary effect with respect to RRKET/RRNKET for the repeated-dose group, but not for the acute dose group (Kruskall Wallis test, P , 0.05). Comparison of the RRKET/RRNKET values between dosing conditions (acute versus repeated) showed signicant differences in all bone types with the exception of femoral bone (Mann Whitney, P , 0.05).

RRKET/ RRNKET Acute-1 mean (SD) Acute-1 range Repeated mean (SD) Repeated range

FEM 2.7 (2.1)

VRT 1.4 (1.6)

PEL 1.8 (1.7)

HUM 1.4 (0.9)

RIB 1.2 (1.2)

ULNA 0.94 (0.63)

0.75 6.1 1.8 (1.7)

0.25 4.1 0.20 (0.07)

0.65 4.9 0.41 (0.22)

0.58 2.8 0.51 (0.17)

0.22 3.3 0.18 (0.07)

0.26 1.9 0.28 (0.15)

0.65 4.8

0.13 0.31

0.21 0.79

0.33 0.67

0.11 0.28

0.099 0.51

Data (RR/m and RRKET/RRNKET) from all bone types were also pooled and compared between the different patterns of exposure (Acute-1 versus Repeated) to determine whether statistical differences were maintained. For both KET and NKET, the pooled RR/m values differed signicantly between the acute and repeated dose groups (Mann Whitney, P , 0.05). Similarly, the pooled RRKET/RRNKET values differed signicantly between the acute and repeated dose groups (Mann Whitney, P , 0.05). The pooled data for the various bone types are summarized in Figure 2.

Table I Summary of RR/m Data for KET in Various Bone Types KET (RR/m) Acute-1 mean (SD) Acute-1 range Repeated mean (SD) Repeated range FEM 16.6 (9.4) VRT 8.6 (2.6) PEL 3.7 (3.5) HUM 5.4 (2.3) RIB 3.7 (2.5) ULNA 2.2 (1.4)

3.2 24 3.5 (4.1)

5.2 12 0.51 (0.11)

1.1 9.9 0.51 (0.30)

1.9 8.1 0.43 (0.09)

0.6 7.5 0.31 (0.05)

0.5 3.8 0.15 (0.08)

1.2 11

0.39 0.61

0.25 0.90

0.30 0.53

0.22 0.35

0.04 0.26

Acute versus chronic (long-term) doses Only pelvi and vertebrae were analyzed in this portion of the work, because an earlier study (17) showed very little evidence of KET or NKET in fresh bone marrow. Because marrow is known to signicantly concentrate drugs (2, 3, 5, 11), we chose to restrict analysis to those bones that have been demonstrated to routinely contain the largest drug levels in this animal model (1, 5, 9). Of the vertebral bone samples analyzed, KET and NKET were detected in all samples from the Acute-2 group. In those samples, KET levels corresponded to roughly 7 25 mg/g, while NKET levels corresponded to roughly 514 mg/g. In the Chronic group, only NKET was detected in 3/8 samples. In those samples, NKET levels corresponded to approximately 50 70 ng/g. Of the pelvic bone samples analyzed, KET and NKET were detected in all samples from the Acute-2 group, with KET and NKET levels corresponding to roughly 640 and 216 mg/g, respectively. In samples from the Chronic group, NKET was detected in 3/8 samples and KET was detected in one sample. In those samples, NKET levels corresponded to approximately 2090 ng/g, while the detected KET level corresponded to approximately 50 ng/g. Comparison of the RR/m values for KET and NKET between dosing conditions (Acute-2 versus Chronic) showed signicant differences in both pelvic and vertebral bone samples (MannWhitney, P , 0.05).

Comparison of Relative Distribution of Ketamine and Norketamine in Decomposed Skeletal Tissues Following Single and Repeated Exposures 431

The ratio of levels of KET to those of NKET (RRKET/RRNKET) ranged from 1.2 2 7.0 in vertebral bone in the Acute-2 group. In pelvic bone, RRKET/RRNKET values ranged from 0.5 2 18 in the ACUTE-2 group. In the only pelvic bone sample where KET was detected in the Chronic group, the RRKET/RRNKET value was 1.5. Discussion This work is the rst of a group of studies whose purpose is to examine whether different patterns of drug exposure (i.e., fatal acute overdose versus chronic use) may be discriminated through toxicological analysis of decomposed skeletal remains. In this study, bone tissues derived from rats that had undergone different patterns of KET exposure were assayed for KETand its primary metabolite, NKET. The background to this work suggests that interpretation of toxicological analysis of bone is complex. Experiments involving controlled drug administration to rats suggest that a dosedependent response may be observed in analysis of bone tissue (1, 5, 11), but the relationships between dose and measured drug level may be skewed by decomposition (5, 15), and signicant inter-animal and intra-animal variation in observed bone drug levels may limit the ability to discriminate between toxic and non-toxic exposures. In our most recent work (1, 12, 14), we observed substantial variation in drug and metabolite levels between different bone types following acute drug exposure. Perhaps more interestingly, we observed a lower variability in the ratio of levels of drug to those of drug metabolite in those same samples, suggesting that such ratios may provide a means to reduce the observed variability that may obscure our ability to derive any interpretive value from toxicological measurements in bone tissue. The data reported in this work suggest that the pattern of KET exposure may be discriminated based on measured levels of KET and NKET within a given bone type, as well as upon the ratio of levels of KET to those of NKET (i.e., RRKET/RRNKET). This was observed for both experimental sets: the comparison of acute versus chronic dosing (i.e., longer term), followed by an extended washout period, as well as the comparison of a larger acute dose versus the smaller dose, repeated three times over a shorter time interval. In the former case, the data support the notion that drug and metabolite have a minimal likelihood of substantial accumulation in skeletal tissues and is consistent with our earlier work examining the role of the time interval between acute drug exposure and death (3). In an earlier study (17) examining the marrow from animals from the acute versus chronic exposure study, there was no drug detected in the femoral or tibial marrow of any animals from the chronic group. We hypothesize that the very low levels of NKET and, in one case, KET, detected in some vertebral and pelvic bone samples from the chronic dose group were, in part, the result of bone absorption of residual drug and metabolite from decomposition of the soft visceral tissues. Although the data from the study involving acute versus repeated exposures over a short term suggested that the measured levels of KET and NKET were more discriminatory with respect to the pattern of exposure, we feel that the use of the ratio of levels of parent drug and metabolite(s) should be considered as the parameter of choice in forensic toxicological

Figure 2. Summary of pooled data for acute versus repeated doses exposure (Acute-1 versus Repeated). Box-and-whisker plots for pooled bone samples (i.e., all bone types), for mass-normalized response ratios (RR/m) for KET (A); mass-normalized response ratios (RR/m) for NKET (B); ratio of levels of KET and NKET (RRKET/RRNKET) in a given sample (C).

432 Watterson et al.

analysis of bone tissue. In the absence of standardized sample preparation methodologies of solid bone, there remains the potential for signicant variability in analyte recovery from the solid bone matrix, depending on how the bone is prepared for drug extraction. In cases where bone fragments are incubated in solvent to effect drug extraction, the surface area of the bone available for solvent contact inuences the degree to which drugs and metabolites will diffuse into the extraction solvent (i.e., drug recovery). To some extent, use of a ratio of levels of parent drug to those of metabolite(s) as a toxicological measure (as opposed to absolute drug levels) can offset the variation in drug recovery from one extraction to the next. Thus, this approach may prove more useful in death investigations and in comparison of data between different laboratories. Furthermore, a variety of data from our laboratory (1, 12, 14) has suggested that the variability in drug level between different bones within a given body may be substantial and complicate the toxicological interpretation of a given measurement. In comparison, the ratio of levels of parent drug to those of metabolite(s) has generally been less variable, and may provide a more precise indicator of the pattern of drug exposure in a particular case.

References
1. Watterson, J.H., Donohue, J.P. (2011) Relative distribution of ketamine and norketamine in skeletal tissues following various periods of decomposition. Journal of Analytical Toxicology, 35, 452458. 2. Vandenboer, T.C., Grummett, S.A., Watterson, J.H. (2008) Utility of immunoassay in drug screening in skeletal tissues: Sampling considerations in detection of ketamine exposure in femoral bone and bone marrow following acute administration using ELISA. Journal of Forensic Sciences, 53, 1474 1482. 3. Watterson, J.H., Vandenboer, T.C. (2008) Effects of tissue type and the dose-death interval on the detection of acute ketamine exposure in bone and marrow with solid-phase extraction and ELISA with liquid chromatography-tandem mass spectrometry conrmation. Journal of Analytical Toxicology, 3, 631 639. 4. McGrath, K.K., Jenkins, A.J. (2009) Detection of drugs of forensic importance in postmortem bone. American Journal of Forensic Medicine and Pathology, 30, 40 44. 5. Lafreniere, N.M., Watterson, J.H. (2009) Detection of acute fentanyl exposure in fresh and decomposed skeletal tissues. Forensic Science International, 185, 100106. 6. Lafreniere, N.M., Watterson, J.H. (2010) Detection of acute fentanyl exposure in fresh and decomposed skeletal tissues Part II: The effect of dose-death interval. Forensic Science International, 94, 60 66. 7. Guillot, E., de Mazancourt, P., Durigon, M., Alvarez, J.C. (2007) Morphine and 6-acetylmorphine concentrations in blood, brain, spinal cord, bone marrow and bone after lethal acute or chronic diacetylmorphine administration to mice. Forensic Science International, 166, 139144. 8. Raikos, N., Tsoukali, H., Njau, S.N. (2001) Determination of opiates in postmortem bone and bone marrow. Forensic Science International, 123, 140141. 9. Watterson, J.H., Desrosiers, N.A. (2011) Microwave-assisted extraction in the study of the effect of dose-death interval on meperidine detection in skeletal tissue. Forensic Science International, 207, 40 45. 10. Gorczynski, L.Y., Melbye, F.J. (2001) Detection of benzodiazepines in different tissues, including bone, using a quantitative ELISA assay. Journal of Forensic Sciences, 46, 916 918. 11. Watterson, J.H., Botman, J.E. (2009) Detection of acute diazepam exposure in bone and bone marrow: Inuence of tissue type and the dose-death interval on sensitivity of detection by ELISA and liquid chromatography tandem mass spectrometry conrmation. Journal of Forensic Sciences, 54, 708 714. 12. Watterson, J.H., Desrosiers, N.A., Betit, C.C., Dean, D.E., Wyman, J.F. (2010) Relative distribution of drugs in decomposed skeletal tissues. Journal of Analytical Toxicology, 34, 511515. 13. Horak, E.L., Jenkins, A.J. (2005) Postmortem tissue distribution of olanzapine and citalopram in a drug intoxication. Journal of Forensic Sciences, 50, 1 3. 14. Desrosiers, N.A., Watterson, J.H., Dean, D.E., Wyman, J.F. (2012) Detection of amitriptyline, citalopram, and metabolites in porcine bones following extended outdoor decomposition. Journal of Forensic Sciences, 57(2), 544 549. 15. Watterson, J.H., Desrosiers, N.A. (2010) The effects of burial on drug detection in skeletal tissues. Drug Testing and Analysis, 2, 346 356. 16. Williams, M.L., Mager, D.E., Parenteau, H. et al. (2004) Effects of protein calorie malnutrition on the pharmacokinetics of ketamine in rats. Drug Metabolism and Disposition, 32, 786 793. 17. Betit, C.C., Watterson, J.H. (2011) Comparison of ketamine and norketamine levels in bone marrow following acute and chronic ketamine exposure. Joint SOFT-TIAFT Conference, San Francisco, CA.

Conclusions The results presented here suggest that, in some cases, acute and repeated exposures to KET may be discriminated on the basis of the levels of KET and NKET in bone, but perhaps more importantly based on the ratio of levels of KET to those of NKET. This work is the rst in a series of studies to be undertaken in our laboratory investigating the utility of toxicological measurements in decomposed skeletal remains for discrimination of different patterns of drug exposure. These studies are a critical rst step in clarifying the utility of bone tissue as a viable matrix in postmortem toxicology. Clearly, the number and variety of drugs administered, and the complexity of exposure patterns used, must be expanded to generate a database of valuable forensic information, because the observed drug and metabolite levels, as well as the ratios of drug and metabolite levels, may be expected to vary as a function of dose, time between exposure and death, postmortem environment, and the chemical properties of the drug and metabolite(s) in question. Until a large breadth of such studies is conducted and the approach may be subjected to an appropriate level of scrutiny through a validation exercise, drug detection in human skeletal tissues from casework should be interpreted in a qualitative fashion.

Acknowledgments The authors wish to acknowledge the Natural Sciences and Engineering Research Council of Canada and the Laurentian University Research Fund for nancial support of this work.

Comparison of Relative Distribution of Ketamine and Norketamine in Decomposed Skeletal Tissues Following Single and Repeated Exposures 433