Fentanyl Postmortem Redistribution: Preliminary Findings Regarding the Relationship Among Femoral Blood and Liver and Heart Tissue Concentrations

Transcription

1 Fentanyl Postmortem Redistribution: Preliminary Findings Regarding the Relationship Among Femoral Blood and Liver and Heart Tissue Concentrations Kristin Luckenbill 1,2, Jonathan Thompson 3, Owen Middleton 3, Julie Kloss 1, and Fred Apple 1,2, * 1 Hennepin County Medical Center, 2 University of Minnesota School of Medicine, and 3 Hennepin County Medical Examiner s Office, Minneapolis, Minnesota Abstract Postmortem redistribution refers to the process of drugs diffusing from tissues into blood along a concentration gradient between death and time of specimen collection at autopsy. Anatomical site-to-site variation can exist for drug concentrations. The purpose of this study was twofold. First femoral blood, liver, and heart fentanyl concentrations were compared in medical examiner cases to assist in determining which specimen most appropriately should be used for interpretation. Nine fentanylpositive cases were identified by history of drug use over a 15-month period ( ). Femoral blood fentanyl concentrations (n = 9) ranged from 2.7 to 52.5 µg/l, liver fentanyl tissue (n = 9) ranged from 37.0 to 179 µg/kg, and heart fentanyl tissue (n = 3) ranged from 52.8 to 179 µg/kg. Liver tissue to femoral blood ratios ranged from 0.85 to 35.8, and heart tissue to femoral blood ratios ranged from 1.9 to 5.4. Second, utilizing a published compendium of multiple postmortem drugs, liver and heart tissues to femoral blood drug ratios were compared to known volumes of distribution, solubilities, and pk a. No significant relationships were observed. In conclusion, establishing a larger evidence-based database using liver fentanyl concentrations may be more optimal than blood concentrations for interpretation of postmortem fentanyl concentrations in medical examiner and coroner cases. Introduction Postmortem redistribution (PMR) is the process of drugs diffusing from solid organs into blood and surrounding tissues along a concentration gradient in the time between death and sample collection at autopsy. Many factors can influence postmortem redistribution, such as the characteristics and pharmacokinetics of drugs, incomplete drug distribution at the time of death, the interval between death and sampling, * Author to whom correspondence should be addressed: Dr. Apple, Hennepin County Medical Center, Clinical Labs P4, 701 Park Ave., Minneapolis, MN apple004@umn.edu. anatomical collection site, the amount of sample drawn, the effects of cell death, putrefaction, and body position and movement after death (1). These factors may cloud the ability to accurately interpret postmortem toxicology results as they pertain to cause of death. Historically, heart blood and femoral blood concentrations reported in the 1960s for barbituates showed differences in central versus femoral blood concentrations (2,3). These studies concluded that peripheral blood should be the specimen of choice for postmortem toxicology quantitation. Additional studies conducted with other drugs agreed with this observation and found that, in general, heart blood concentrations tended to be higher than those of femoral blood (2,3). The ratio of heart blood to femoral blood drug concentrations has been discussed as an indicator of postmortem redistribution (4), but this concept is not universally accepted. In general few studies have accurately addressed whether 1. there is a correlation between the heart blood to femoral blood drug concentrations at the time of death and 2. whether there is a relationship between the heart blood to femoral blood drug ratio and the postmortem interval (4). Furthermore, large series of cases are not available to accurately assess postmortem relationships between tissue (heart or liver) and blood (heart or peripheral) drug concentration and what changes occur postmortem. Fentanyl is a synthetic opioid used as a surgical anesthetic and for chronic pain control (5). The effects of fentanyl include analgesia, euphoria, tolerance, physiological dependence, and respiratory depression (6). Fentanyl is highly lipophilic, with an n-octanol/water coefficient of 860:1 (7). Fentanyl is also 80 86% protein-bound and has a moderate to large volume of distribution (Vd) of 3 8 L/kg (8,9). The majority of fentanyl accumulates in muscle and adipose tissue, and it is slowly released into blood (10). Because of the high potency and short duration of action of fentanyl, it has become a therapeutically misused and recreationally abused drug and is viewed as a suitable replacement for heroin (11,12). However, because low concentrations of fentanyl can cause respiratory depression, Reproduction (photocopying) of editorial content of this journal is prohibited without publisher s permission. 639

2 abuse of fentanyl can be dangerous (12). Because there is no standardized guideline for what type or types of specimen should be collected, medical examiner and coroner offices vary in what they collect, including femoral and heart blood, heart tissue, liver tissue, and/or vitreous humor fluid (as well as other tissues) to help establish whether drugs found during toxicological analysis were at therapeutic or toxic levels. The forensic toxicologist often assists in medicolegal death investigations through interpretation of drug concentrations in the deceased based on what blood or tissue results are provided. It was hypothesized by the authors that heart blood may not be the ideal specimen for quantitating postmortem fentanyl concentrations because of the potential of postmortem redistribution of fentanyl from both heart and liver tissues into heart blood, falsely increasing the heart blood fentanyl level. To partially address this concept, comparisons were made for fentanyl concentrations from femoral blood, and liver and heart tissue in nine fentanyl-related deaths and heart tissue and liver tissue to femoral blood ratios. In addition, reanalysis of a published compendium (4) of numerous drugs, heart blood to femoral blood ratios were compared to their corresponding volumes of distribution, water solubilities, and pk a s to investigate their potential relationships with PMR. Experimental Specimens Nine cases over a 15-month period in 2007 and 2008 from the Hennepin County Medical Examiner s Office where fentanyl was suspected as a potential contributor to the investigation by history were examined. At autopsy, femoral blood samples were collected in EDTA or sodium fluoride tubes, and liver (right lobe) tissue and heart (left ventricle) tissue were stored without preservative in plastic vials. Samples were stored at 4 C and analyzed by gas chromatography mass spectrometry (GC MS) within 24 to 72 h of collection. Materials Fentanyl standards and deuterated fentanyl internal standards were obtained from Cerillant (Roundrock, TX). Bond Elut solid-phase extraction columns were purchased from Varian (Harbor City, CA). Instrumentation A Hewlett-Packard (Palo Alto, CA) 5890 GC equipped with a 5972 mass selective detector was used for identification and quantitation of fentanyl. The MS was operated in the selected ion-monitoring mode (SIM), and the following ions were monitored: quantitating ion 245 and qualifier ions 146, 189 for fentanyl and quantitating ion 250 and qualifier ions 151, 194 for fentanyl-d 5. The GC was equipped with a 30-m DB-5 capillary column (Agilent Technologies, Palo Alto, CA) with helium as the carrier gas set at a flow rate of 1.8 ml/min. The oven temperature was programmed at 100 C initially, increased at a rate of 30 C/min to 270 C, and was held for 5.5 min. Procedures Whole blood and liver and heart tissue specimens were quantitated for fentanyl using a modification of the method of van Rooy et al. (13) by GC MS. A whole blood low control (4 µg/l) and a whole blood high control (8 µg/l) were analyzed in each run along with the cases. The extraction consisted of either 1 ml of whole blood or 1 ml of homogenate (5 g of tissue in 5 mls of water thoroughly homogenized with a blender) mixed with 10 µl fentanyl-d 5 internal standard (10 mg/l) and 3 ml of water, which was then added to this solution and vortex mixed. After sitting for 5 min, it was centrifuged at 1950 g for 15 min, and the pellet was discarded. To the supernatant, 3 ml of 100 mm phosphate buffer was added followed by ph adjustment to 6. The specimen was transferred onto a solid-phase extraction column preconditioned with methanol, water and 100 mm phosphate buffer. Following treatments with water, 100 mm acetate buffer, and methanol, the column was eluted with methylene chloride, isopropanol, and ammonium hydroxide (78:20:2). The eluant was evaporated at C with nitrogen, reconstituted with 0.05 ml ethyl acetate, and transferred for analysis into the autosampler for injection on the GC MS. Standard curves were derived for each analysis using 2.0, 5.0, 10.0, and 20.0 µg/l whole blood calibrators and area ratios for unknowns were used to calculate the corresponding analyte concentration. Quantitation of fentanyl was based upon ratios of integrated ion areas to the corresponding deuterated internal standard. Analytes were identified based upon comparison of relative retention times (within 1% of retention time of calibrators) and ion ratios with the corresponding values of calibrators assayed in the same run. Ion ratios were calculated by dividing the area of the qualifier ion by the area of the quantitative ion and must be within ± 20% of calibrators. Limits of detection, quantitation, and linearity were experimentally found to be 2.0, 2.0, and 100 µg/l, respectively. Intra-assay precision was determined by assaying five replicates of fentanyl-fortified blood samples at 2.0 µg/l and was 4.9%. Interassay precision was determined to be 10.6% over a six-month period using the blood low control (N = 22). For this observational study, mean, median, and range of fentanyl concentrations and ratios were determined. Cases were further analyzed for findings in which fentanyl was the sole cause of death, or where fentanyl was part of a mixed drug overdose. The mean and range of fentanyl concentrations were compared among the subpopulations and with respect to the cause and manner of death. In the second part of the study, reanalysis of cardiac blood to peripheral blood ratios from the Dalpe-Scott et al. (4) paper were compared to known volumes of distribution, water solubility, and pk a s obtained from the Merck Index to assist in better understanding the relationship between these pharmacodynamic drug parameters and PMR (14). If there was a range of volumes of distribution, an average was calculated. Cardiac blood and peripheral blood ratios were then plotted against volumes of distribution, water solubilities, and pk a s, and correlations were calculated using MS Excel. 640

3 Results The nine medical examiner cases examined are described in Table I. Femoral blood, liver and heart tissue fentanyl concentrations, the route and dose of fentanyl, other drugs detected, route of drug administration, the cause and manner of death and postmortem interval are described. Two of the cases, 1 and 2, were determined to be solely fentanyl overdoses, five cases, 3 7, as mixed drug overdoses, and the remaining two cases, 8 and 9, as natural cause deaths. Other drugs identified in mixed drug toxicities were ethanol (3 cases), cocaine (2 cases), tramadol, benzodiazepines, and other opiates. Femoral blood concentrations ranged from 2.7 to 52.5 µg/l. The mean and median femoral blood fentanyl concentrations were 17.4 (SD 14.9) µg/l and 12.3 µg/l, respectively. The overdose cases attributed solely to fentanyl had the first (52.5 µg/l; case 1) and third (26.4 µg/l; case 2) highest femoral blood concentrations, and a natural death had the lowest fentanyl concentration of 2.7 µg/l (case 9). Liver concentrations (n = 9) ranged from 37.0 to 179 µg/kg. Mean and median liver concentrations were 83.1 (SD 50.1) and 87.6 µg/kg, respectively. The highest liver concentrations (179 µg/kg; case 7 and 161 µg/kg; case 2) occurred in a mixed drug overdose and a solely fentanyl overdose. Interestingly, the highest liver concentration had the second lowest femoral blood concentration (5.0 µg/l) and the shortest postmortem interval of 10 h. The lowest liver concentration (37.0 µg/kg; case 8) occurred in a death due to natural causes, and had one of the lower femoral blood concentrations of fentanyl (8.3 µg/l). Fentanyl concentrations in the heart tissue were 52.8 (case 3), 142 (case 2) and 179 (case 1) µg/kg; mean 124 (SD 53.2) µg/kg. The highest heart concentration (179 µg/kg) occurred in a solely fentanyl overdose case that also had the highest femoral blood concentration (52.5 µg/l) and one of the shorter postmortem intervals of 12 h. The lowest heart fentanyl concentration (52.8 µg/kg) was from a mixed drug overdose case with the second highest femoral blood concentration (27.7 µg/l) and the longest postmortem interval of 39 h. The ratios of liver to femoral blood ranged from 0.85 to 35.8, with a mean ratio of 9.6. The highest liver to femoral blood ratio was in a mixed drug overdose case with the highest liver concentration (179 µg/kg; case 7) and the second lowest femoral blood concentration (5.0 µg/l), and the lowest liver to femoral blood ratio occurred in a solely fentanyl overdose case with the highest heart concentration (179 µg/kg; case 1) and the highest femoral blood concentration (52.5 µg/l). There was poor correlation between liver to femoral blood ratios and postmortem intervals: r = The ratios of heart tissue to femoral blood ranged from 1.9 to 5.4, with a mean of 3.6. The highest heart to femoral blood ratio (5.4; case 2) occurred in a fentanyl overdose case with the Table I. Fentanyl Concentrations in Femoral Blood (n = 9) and Liver (n = 9) and Heart (n = 3) Tissues in Nine Fentanyl- Related Medical Examiner Cases Fentanyl Concentrations Case Femoral Liver Heart Other Drugs Route of Drug Cause of Manner of Postmortem No. blood (µg/l) (µg/kg) (µg/kg) Detected Administration Death Death Interval (h) IV (75 µg/h Fentanyl Accident 12 opened wrapper overdose found at scene) µg/h and Fentanyl Accident µg/h transdermal overdose patches (5 total) Ethanol, Transdermal patch; Mixed drug Accident 39 Tramadol dose unknown overdose Ethanol, History of oral Mixed drug Accident 28 Benzodiazepines administration overdose Cocaine Unknown Mixed drug Accident 17 overdose Opiates 2 75 µg/h Mixed drug Accident 16 transdermal patches overdose Ethanol, Transdermal patch; Mixed drug Accident 10 cocaine dose unknown overdose µg/h Arthero-sclerotic Natural 21 transdermal patch heart disease Unknown (surgery, Natural Natural 29 probably IV) 641

4 second highest liver concentration of 161 µg/kg and third highest femoral blood fentanyl concentration (26.4 µg/l). The lowest heart to femoral blood fentanyl concentration of 1.9 (case 3) occurred in a mixed drug overdose with the longest postmortem interval of 39 h and the second highest femoral blood fentanyl concentration of 27.7 µg/l. There was poor correlation between heart to femoral blood ratios and postmortem intervals: r = Five of the cases (2,3,6,7,8) involved the use of transdermal patches, including one fentanyl overdose, three of the mixed drug overdoses, and one of the natural deaths. The femoral blood concentrations from these cases ranged from 5.0 to 27.7 µg/l, and the liver tissue concentrations ranged from 37.0 to 179 µg/kg. One of the natural cause deaths, case 9, was most likely administered fentanyl intravenously, with a femoral blood concentration of 2.7 µg/l, and a liver tissue concentration of 53.0 µg/kg. One of the mixed drug overdoses, case 4, had a history of oral fentanyl abuse, and the femoral blood concentration was 12.6 µg/l, with a liver tissue concentration of 48.0 µg/kg. The fentanyl overdose case with the highest femoral blood concentration, 52.5 µg/l (case 1), and a liver tissue concentration of 45.0 µg/kg, was injecting contents from transdermal patches. Heart blood to femoral blood drug ratios obtained from the large drug database from the Dalpe-Scott et al. (4) paper were compared to the Vds, solubilities, and pk a s of drugs obtained from the Merck Index to better understand the role of these parameters regarding PMR. Figure 1A shows there was very poor correlation of the ratio compared to Vd (r = 0.08). When two outlying data points were eliminated, the correlation improved, but only marginally to Correlations were also poor for the heart blood to femoral blood ratios versus solubility in water, r = 0.24 (Figure 1B) and versus pk a, r = 0.34 (Figure 1C). Discussion Figure 1. Comparison of published (4) heart blood: femoral blood ratios to the volume of distribution for 87 drugs (2 outlying points removed) (A); solubility for 15 drugs (B); and pk a for 21 drugs (C). The current study s preliminary data demonstrated the wide variability in fentanyl concentrations across femoral blood and liver and heart tissues, independent of the cause and manner of death. Anderson and Muto (6) reported in their conclusion that fentanyl liver concentrations less than 31.0 µg/kg seemed to represent therapeutic levels, and concentrations above 69.0 µg/kg seemed to represent overdose situations. In the current study, none of the cases had liver concentrations below 31.0 µg/kg, four had concentrations above 69.0 µg/kg, and five had concentrations between 31.0 and 69.0 µg/kg. Of the four cases that were above 69.0 µg/kg, three were mixed drug overdoses, and one was a solely fentanyl overdose. The two cases signed out as deaths due to natural causes had fentanyl liver concentrations of 37.0 and 53.0 µg/kg. The highest liver concentration, 179 µg/kg (case 7), occurred in the case with the second lowest femoral blood concentration (5.0 µg/l) and shortest postmortem interval of 10 h, indicative of acute toxicity and possibly no postmortem redistribution. One could also consider an accumulation effect in the liver due to possible medical problems affecting the liver. However, the second shortest postmortem interval of 12 h had the highest femoral blood concentration (52.5 µg/l; case 1), and heart concentration (179 µg/kg), and also a lower liver concentration (45.0 µg/kg), suggestive of possible postmortem redistribution. These two contradictory cases demonstrate how problematic the interpretation of postmortem blood fentanyl concentrations can be since postmortem drug concentrations may vary substantially between individuals and are reliant upon numerous variables (15). Unlike many other drugs, femoral blood may not be the ideal sample for the interpretation of postmortem fentanyl concentrations. Fentanyl, a highly lipophilic and protein-bound drug, with a moderate to high volume of distribution, might be better interpreted from liver tissue as suggested by Anderson and Muto (6) and as has been routine practice in the current authors lab for 20 years for tricyclic antidepressants (16). 642

5 The authors have used the reanalysis of the Dalpe-Scott et al. (4) data, even though fentanyl was not part of the data, to underline their hypothesis that volume of distribution and solubility of drugs cannot always be used to assess the likelihood that PMR occurs. Based on the comparisons from this published compendium of heart blood to femoral blood drug ratios (4) to corresponding volumes of distribution, water solubilities and pk a s (Figure 1), poor correlations were demonstrated. Past studies have speculated that basic drugs with large volumes of distribution tend to have larger heart blood to femoral blood ratios; data not supported by the observations of the current study based on the reanalysis of Dalpe-Scott et al. (4) data. Further, some basic drugs and acidic drugs with large heart blood to femoral blood ratios have small volumes of distribution (4). Some forensic scientists, as opposed to postmortem toxicologists, believe that the heart blood to femoral blood ratios are an indicator of whether postmortem redistribution occurred and use a high ratio as a measure of PMR. The authors of the current paper do not agree with this rationale, as multiple mechanisms including variable postmortem redistribution and diffusion both centrally and peripherally may occur, as well as incomplete drug distribution at around the time of death in a very recent drug exposure. By comparing the heart to femoral blood ratios to water solubility as an indicator of drug lipophilicity, the authors hypothesized that drugs with a higher heart blood to femoral blood ratio might also tend to have lower water solubilities and vice versa, because lipophilic drugs might take longer to redistribute following death. The findings based on the additional analysis of drug data from this one study in the literature (4), as shown in the current study, did not demonstrate any predictable relationships between heart blood and femoral drug blood ratios and the volumes of distribution, solubilities, and pk a s of numerous drugs. Therefore, extrapolation of these findings to fentanyl supports the authors hypothesis that heart blood to femoral blood ratios need to be intrepreted with caution regarding PMR. Further, one needs to consider that, in general, lipophilic drugs can also move readily through cellular membranes, confounding the lack of relationships observed in postmortem blood analyses, including fentanyl. In conclusion, the preliminary findings presented in this small, observational study demonstrate a wide variability in fentanyl concentrations across femoral blood and liver and heart tissues. It demonstrates the difficulty in predicting the impact of multiple variables, such as volume of distribution, water solubility, and pk a on postmortem redistribution. It is apparent that additional studies are needed, including a larger series of cases, with comprehensive data from multiple blood sites (heart and femoral) compared to tissue sites to best determine an appropriate specimen type to use for cause of death interpretation to assist medical examiners and coroners in their case reviews. References 1. M.C. Yarema and C.E. Becker. Key concepts in postmortem drug redistribution. Clin. Toxicol. 43: (2005). 2. A.S. Curry and I. Sunshine. The liver:blood ratio in cases of barbiturate poisoning. Toxicol. Appl. Pharmacol. 2: (1960). 3. D.J. Gee, R.A. Dailey, M.A. Green, and L.A. Perkins. Postmortem diagnosis of barbiturate poisoning. In Forensic Toxicology. Proceedings of a symposium held at the chemical defense establishment, Porton Down, B. Ballantyne, Ed. John Wright and Sons, Bristol, U.K., 1972, pp M. Dalpe-Scott, M. Degouffe, D. Garbutt, and M. Drost. A comparison of drug concentrations in postmortem cardiac and peripheral blood in 320 cases. Can. Soc. Forensic Sci. J. 28: (1995). 5. A. Poklis. Fentanyl: a review for clinical and analytical toxicologists. J. Toxicol. Clin. Toxicol. 33: (1995). 6. D.T. Anderson and J.J. Muto. Duragesic transdermal patch: postmortem tissue distribution of fentanyl in 25 cases. J. Anal. Toxicol. 24: (2000). 7. K.A. Calis, D.R. Kohler, and D.M. Corso. Transdermally administered fentanyl for pain management. Clin. Pharm. 11: (1992). 8. S.A. Hargrave. The estimation of binding of 3H-fentanyl to plasma proteins. Br. J. Anaesth. 51: (1979). 9. J.R. Halliburton. The pharmacokinetics of fentanyl, sufetanil and alfentanil: a comparative review. AANA J. 56: (1988). 10. R.J. Hudson, I.R. Thomson, J.E. Cannon, R.M. Friesen and R.C. Meatherall. Pharmacokinetics of fentanyl in patients undergoing abdominal aortic surgery. Anesthesiology 64: (1986). 11. J.G. Thompson, A.M. Baker, A.H. Bracey, J. Seningen, J.S. Kloss, A.Q. Strobl, and F.S. Apple. Fentanyl concentrations in 23 postmortem cases from the Hennepin County Medical Examiner s Office. J. Forensic Sci. 52: (2007). 12. J.J. Kuhlman, R. McCaulley, T.J. Valouch, and G.S. Behonick. Fentanyl use, misuse and abuse: a summary of 23 postmortem cases. J. Anal. Toxicol. 27: (2003). 13. H.H. van Rooy, N.P. Vermeulen, and J.B. Bovill. The assay of fentanyl and its metabolites in plasma of patients using gas chromatography with alkali flame ionisation detection and gas chromatography mass spectrometry. J. Chromatogr. 223: (1981). 14. The Merck Index. Current contents, March C.S. Crandall, S. Kerrigan, R.L. Aguero, J. LaValley, and P.E. McKinney. The influence of collection site and methods on postmortem morphine concentrations in a porcine model. J. Anal. Toxicol. 30: (2006). 16. F.S. Apple and C.M. Bandt. Liver and blood postmortem tricyclic antidepressant concentrations. Am. J. Clin. Pathol. 89: (1988). 643