- Email: [email protected]
Stability of cocaine in formalin solution and fixed tissues
Forensic Science International 193 (2009) 79–83
Contents lists available at ScienceDirect
Forensic Science International journal homepage: www.elsevier.com/locate/forsciint
Stability of cocaine in formalin solution and fixed tissues Guido Viel, Alessandro Nalesso, Giovanni Cecchetto, Massimo Montisci, Santo Davide Ferrara * Section of Legal Medicine - Forensic Toxicology and Antidoping Unit, Department of Environmental Medicine and Public Health, Via Falloppio 50, 35121 Padova, Italy
A R T I C L E I N F O
A B S T R A C T
Article history: Received 22 May 2009 Received in revised form 9 September 2009 Accepted 13 September 2009 Available online 20 October 2009
Embalming and formalin fixation are common, and yet they can create problems for the forensic scientist if a drug has been the cause of death and if the only available specimens to be analyzed are formalin-fixed tissues. Previous studies have demonstrated that during fixation xenobiotics are extracted into formalin according to tissue and fixing solution characteristics. In some cases formalin can react with the analyte resulting in the production of new chemical entities. Regarding cocaine and its metabolites, Cingolani et al. have reported that formalin-fixation extracts benzoylecgonine (BE) from tissues and that BE is stable in the fixing solution. However, the stability and kinetic properties of cocaine remain so far unexplored. Our data show that in buffered formalin (pH 7.4) cocaine is hydrolyzed to BE in agreement with a pseudo first-order reaction kinetic (half-life time 7 days), whereas in unbuffered formalin (pH 3.5) it is relatively stable over a period of 30 days. The analysis of brain and liver samples at different fixation times indicates that during fixation an extraction process occurs for both analytes and that the extraction is more efficient in the liver than in the brain, probably because of a greater lipophilicity of the brain tissue. In conclusion, our study demonstrates that formalin-fixed tissues and their fixing solutions can be used for cocaine analysis only if a short time period has passed since the fixation beginning. The rapid extraction process of cocaine into formalin and the concomitant hydrolysis to BE occurring in buffered formalin may prevent the identification of cocaine in both tissues and formalin solution already at 15–30 days after fixation. Moreover, the unpredictable extraction rate of both analytes, along with the hydrolysis of cocaine into BE significantly affects tissue concentrations, thus complicating the interpretation of quantitative results. ß 2009 Elsevier Ireland Ltd. All rights reserved.
Keywords: Forensic toxicology Cocaine Benzoylecgonine Formalin Fixation process
1. Introduction In the second half of the 1990s, a global trend of escalating cocaine use was observed across a number of different countries. Within the European Union, lifetime experience among 15–34year-olds ranges from 0.7% to 12.7%, with the highest levels being found in the United Kingdom [1]. Although use trends are increasing, data regarding the hazards of cocaine are limited. Acute deaths in which cocaine is present without opiates seem to be infrequent in Europe. Despite the limitations of the available information, and considering that many cocaine-related deaths may pass unreported being scheduled as cardiovascular deaths (due to a lack of toxicological analyses), cocaine seems to play a determinant role in between 8% and 12% of drug-related deaths in Germany, Spain, France and Hungary [2]. Especially in non-forensic autopsies, if there are no suspects of recent or chronic drug exposure, body fluids and/or tissues for toxicological investigations (to be preserved at 20 8C) are
* Corresponding author. Tel.: +39 049 8272225; fax: +39 049 663155. E-mail address: [email protected] (S.D. Ferrara). 0379-0738/$ – see front matter ß 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.forsciint.2009.09.010
generally not collected. On the contrary, in most clinical and forensic post-mortem examinations, specimens of various organs are gathered and fixed in buffered formalin for histological examination. In these cases, the only possible way to perform subsequent toxicological analyses is by exhuming the body (impossible in cases of cremation) or by analyzing the formalinfixed tissues collected at autopsy. Considering this premise and the fact that embalming and cremation procedures are ever more widely diffused, it is necessary to widen our knowledge as to stability and kinetic properties of drugs and drugs of abuse in formalin-fixed tissues [3–5]. Previous studies have clarified that during fixation xenobiotics are extracted into formalin according to the hydrophility index of the substance and of the tissue under fixation [6,7]. In some cases formalin can react with the substance in a time, pH and formalin concentration dependent manner resulting in the production of new chemical entities [8–11]. This has been demonstrated for metamphetamines [5,12] and several antidepressants (fluoxetine, sertraline, nortriptiline, desipramine) [10,13,14]. Regarding cocaine and its metabolites, Cingolani et al. [15] have reported that formalin-fixation extracts benzoylecgonine (BE) from tissues and that BE is quantifiable in formalin for months after
80
G. Viel et al. / Forensic Science International 193 (2009) 79–83
Fig. 1. (A) Brain tissue specimens used for the fixation experiments; (B) liver tissue specimens used for the fixation experiments; (C) small (1) and large (2) formalin-fixed liver fragments belonging to a forensic casework (analytical results reported in Table 1).
the fixation. However, the stability and kinetic properties of cocaine remain so far unexplored. Our study was undertaken to provide the forensic community with information regarding the stability of cocaine in formalin-fixed tissues and formalin solutions under various formaldehyde concentrations and pH conditions. 2. Materials and methods 2.1. Solvents and reagents All reagents and solvents were of analytical grade. Dichloromethane, isopropanol, ammonium hydroxide, and hydrochloridric acid were purchased from Sigma–Aldrich (Milano, Italy). Formalin 40% was obtained from Sigma–Aldrich (Milano, Italy), N-methyl-N-trifluoroacetamide with 1% trimethylchlorosilane (MSTFA—1% TMSC) from Pierce (Rockford, IL, USA). Methanol, potassium dihydrogen phosphate and potassium hydroxide were provided by Carlo Erba Reagents (Milano, Italy). Deionized water was produced with a Milli-Q system (Millipore, Bedford, MA, USA). Standard solutions of cocaine and benzoylecgonine (1 mg/mL) were purchased from Cerilliants (Round Rock, TX, USA), while scopolamine (internal standard) was obtained from Promochem-LGC (Teddington, UK). 2.2. Solutions Potassium phosphate buffers (pH 7.4, 10 mM) were prepared by dissolving 1.36 g of potassium dihydrogen phosphate (Carlo Erba reagents, Milano, Italy) in 1 L of deionized water and by adjusting the pH with 0.1 M potassium hydroxide (Carlo Erba reagents, Milano, Italy). Diluted formalin solutions (4% of formaldehyde) at pH 3.5 (unbuffered) and pH 7.4 (buffered) were prepared from the stock formalin solution (40% of formaldehyde) by dilution, respectively, with water and potassium phosphate buffer (pH 7.4, 10 mM). Working solutions of cocaine, BE and scopolamine (IS) at the final concentration of 100 mg/mL were obtained by dilution of the stock solutions with methanol. 2.3. Stability of cocaine in buffered (pH 7.4) and unbuffered formalin solutions (pH 3.5) Cocaine–formalin mixtures (2 mg/mL) at pH 3.5 and 7.4 were obtained by fortification of unbuffered formalin (10%, pH 3.5) and buffered formalin (10%, pH 7.4) with cocaine working solutions. Formalin–cocaine mixtures were stored at room temperature for 4 weeks. Samples were collected and analyzed by GC–MS for cocaine and BE immediately after preparation, and on days 1, 2, 5, 10, 15 and 30. Each experiment was conducted in triplicate. 2.4. Stability of cocaine in formalin-fixed tissues Post-mortem brain and liver tissues collected from a cocaine-related death (brain: cocaine 1.05 mg/g, BE 0.4 mg/g; liver: cocaine 0.75 mg/g, BE 1.2 mg/g; quantification performed according to the method described in Sections 2.5 and 2.6) were used for the following experiment. Five cubic-shaped specimens (volume 3.4 cm3; Fig. 1A and B) of each tissue were fixed in 100 mL of 10% buffered formalin (pH 7.4) and in 100 mL of unbuffered formalin (pH 3.5). Tissue specimens were recovered from formalin at different times (days 2, 5, 10, 15 and 30) along with the fixing solution (5 mL). GC–MS analysis for cocaine and BE was performed on both formalin and tissue samples according to the method described in the following sections. 2.5. Extraction procedures Brain and liver samples (2 g), collected from the central portion of the fixed specimens, were added of potassium phosphate buffer pH 7.4 (6 mL), and
scopolamine working solution (final concentration 2 mg/g). They were homogenized, centrifuged and filtrated. The clear liquid solution was purified by solid phase extraction (SPE) with Bond Elut Certify SPE columns (VARIAN, Harbor City, CA, USA). First, the columns were activated with methanol (2 mL) and potassium phosphate buffer pH 7.4 (2 mL), than samples were loaded and drawn through. Columns were washed with deionized water (3 mL), hydrochloric acid 0.1 M (3 mL) and methanol (3 mL), and then allowed to dry at full vacuum (5 min). Analytes were recovered by elution with a dichloromethane/isopropanol/ammonium hydroxide mixture (78:20:2 v/v; 2 mL). The eluates were gently evaporated to dryness under a stream of nitrogen, derivatized with MSTFA 1% TMSC (50 mL, 75 8C for 15 min), and analyzed by GC–MS. Formalin solutions (0.5 mL), diluted with potassium phosphate buffer pH 7.4 (1.5 mL), were added of scopolamine working solution (final concentration 2 mg/ mL). They were centrifuged, and purified by solid phase extraction (SPE) with the same procedure described for tissue samples. 2.6. GC–MS analysis Analysis was performed on an Agilent 6890N GC system (Agilent Technologies, Palo Alto, CA, USA) equipped with an Agilent 5973N single quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA). The MS was tuned with perfluorotributylamine (PFTBA) using the autotune function as per manufacturer’s recommendation. Data analysis was performed with a HP ChemStation software. Chromatographic separation was performed with an ULTRA-1 fused-silica capillary column (12 m 200 mm i.d. 0.33 mm film thickness) (Agilent Technologies, Palo Alto, CA, USA). GC–MS analysis was performed by injection of derivatized samples (1 mL) in splitless mode (purge time 0.5 min; purge flow 40 mL/min). The inlet temperature was set at 250 8C and the helium carrier gas at a constant flow-rate of 0.6 mL/min. An initial isotherm of 75 8C was maintained for 0.5 min and ramped at 20 8C/min to a temperature of 200 8C. The second ramp rate was 25 8C/min to an isotherm of 300 8C, which was maintained for 5 min. The MSD transfer line temperature was set at 280 8C and that of the quadrupole and source at 150 8C and 230 8C, respectively. A solvent delay time of 5 min was set. All mass spectra were recorded at 70 eV (electron impact, positive mode) with an EM offset of +200 V. Analysis was performed in the selected-ion monitoring (SIM) mode with respect to three significant ions for each compound (cocaine m/z 82, 182, 303; BE m/z 82, 240, 361; scopolamine m/z 94, 154, 375; the underlined ions were selected as quantifiers). Quantification was performed by means of a five-point calibration curve prepared by spiking 10% formalin solutions, and drug-free formalin-fixed homogenized brain and liver tissues at the following final concentrations: 0.02, 0.05, 0.2, 0.5, 1 and 2 mg/mL. Good linear regressions were obtained (y = 0.185x + 1.32, R2 = 0.965 for formalin; y = 0.611x + 3.75, R2 = 0.972 for brain; y = 0.621x + 2.75, R2 = 0.982 for liver). Accuracy and inter-day precision (calculated as relative standard deviation and bias) were always better than 12%. According to validation results the lower limit of quantification (LLOQ) was determined to be 0.02 mg/mL. 2.7. Investigation on a forensic case At non-forensic autopsy, a 33-year-old man unexpectedly died at his home was found to be affected by a transmural infarction of the antero-lateral wall of the left ventricle. Three months after the death the trial, on the basis of novel circumstantial data indicating a possible drug-related fatality, charged a physician and a toxicologist of our Unit with examining the autopsy report and performing toxicological investigations. Because of the presence of several formalin-fixed tissues (in particular a large globular piece of liver along with a smaller liver fragment, Fig. 1C) collected by the pathologist, a preliminary GC–MS screening on formalin solution was performed before exhuming the body. Only the presence of BE was identified, thus the central portion of each liver specimen and the formalin solution were analyzed according to the method described in Sections 2.5 and 2.6.
G. Viel et al. / Forensic Science International 193 (2009) 79–83
81
3. Results In unbuffered formalin (pH 3.5) cocaine proved to be stable over a period of 30 days (Fig. 2B), whereas in buffered formalin (pH 7.4) it was hydrolyzed to BE, with about 80% of cocaine being depleted in 15 days (Fig. 2A). Although there are two reactants in the hydrolysis process (cocaine and water), one of them (water) is in large excess so that, as the reaction progresses, only negligible amounts are consumed and its concentration can be considered to stay constant. As demonstrated by the linearity of the logarithmic plot shown in Fig. 3, the hydrolysis of cocaine occurring in buffered formalin (pH 7.4) can be considered pseudo first-order (y = 0.004x + 7.33). This means that the half-life (t1/2) of the reactant (cocaine) is independent of the starting concentration, being expressed by the following law: t 1=2 ¼
ln 2 0:693 ¼ ¼ 144:28 h 7 days k 0:004
Fig. 3. Logarithmic plot of cocaine concentrations versus time (hours). The linearity of the relationship (y = 0.004x + 7.330; R2 = 0.995) indicates that the hydrolysis reaction is well approximated by a pseudo first-order kinetics.
Therefore, regardless to the initial concentration, in 1 week half of the cocaine is converted to BE. Baselt showed that cocaine concentration in urine does not change over a period of 30 days if the sample is acidified to pH 5, and that setting pH at 8 causes a
significant hydrolysis to BE with a half-life of about 8 days [16]. Previous published data indicate that aqueous cocaine is stable within a pH range of 5 or below [17], with an ideal pH for maximum stability of 1.95 [18]. Our results, in substantial agreement with
Fig. 2. Decomposition of cocaine and formation of BE in buffered (A) and unbuffered formalin (B) over a period of 30 days.
Fig. 4. (A) Concentration of cocaine in brain and liver specimens fixed in buffered (pH 7.4) and unbuffered (pH 3.5) formalin for 30 days. (B) Concentration of cocaine in fixing solutions over 30 days.
82
G. Viel et al. / Forensic Science International 193 (2009) 79–83
Fig. 5. (A) Concentration of BE in brain and liver specimens fixed in buffered (pH 7.4) and unbuffered (pH 3.5) formalin for 30 days. (B) Concentration of BE in fixing solutions over 30 days.
these findings, suggest that formalin does not play an active role in the hydrolysis of cocaine, which is substantially stable in unbuffered formalin (pH 3.5) and unstable in buffered formalin (pH 7.4). The analysis of brain and liver samples at different fixation times (see Section 2.4 for details) showed that during fixation both cocaine and BE were rapidly extracted into formalin solution (Figs. 4 and 5). In the liver, the extraction process for both analytes was faster and more efficient than in the brain, probably because of a greater lipophilicity of the brain tissue (Figs. 4A and 5A). Unbuffered formalin (pH 3.5) demonstrated an overall enhanced extraction capability by comparison with buffered formalin and showed once again its stabilizing properties for cocaine. Indeed, in both liver and brain fixation experiments, the final cocaine concentration in non-buffered formalin was significantly higher than in buffered formalin (Figs. 4B and 5B). The results of the toxicological investigation performed on formalin solutions and formalin-fixed liver fragments belonging to a forensic casework (see Section 2.7 for details) are reported in Table 1. 4. Discussion Except in cases of massive cocaine exposure (couriers or ‘‘body packers’’), in which acute toxicity is dose-related and is primarily
characterized by sympathomimetic effects (tachycardia, hypertension and hyperthermia arrhythmias) [19], cocaine-related deaths mainly occur after a prolonged drug use. Long-term exposure induces a sequence of changes at molecular, cellular and tissue levels [20], many of which favor the occurrence of sudden cardiac death [21]. However, due to the absence of a linear dose–effect relationship and of a scientifically sound ‘‘lethal dose’’, the presence of cocaine in body fluids and/or tissues must be supported by clear histological signs of chronic drug use or abuse to conclude that cocaine played a role in causing the death [22]. Moreover, the interpretation of toxicological findings is complicated by various factors including route of administration, continued absorption after death, different metabolism in chronic/occasional users or in different body compartments, and ‘‘in vivo’’ or ‘‘in vitro’’ degradation of both parent drug and metabolites. Besides the quantification of cocaine and BE in blood, brain levels should always be determined, in order to evaluate and compare the different cocaine/BE ratios [21–24]. High brain concentrations of cocaine with low or absent metabolite levels indicate a recent exposure, whereas high levels of BE with low or null concentrations of cocaine suggest a relatively remote use [21,25]. These considerations are extremely important while interpreting a suspected drug-related fatality. Indeed, about 30% of cocaine-related deaths occur between 2 and 5 h from the assumption, whereas 65% occur within 5 h, thus presenting a high cocaine/BE ratio in the brain [26]. Our data demonstrate that formalin-fixed tissues (particularly the brain) and formalin solutions may be used as biological matrices for identifying suspected cocaine-related fatalities. However, the rapid and effective extraction process of both cocaine and BE occurring during formalin-fixation significantly modifies tissue concentrations, thus complicating the interpretation of quantitative results. In substantial agreement with the data reported by Cingolani et al. [15], we have found that after 30 days of fixation about 80% of cocaine and BE are extracted from the liver into formalin. Additionally, we have observed that in the brain the extraction efficiency is slightly reduced (60% of cocaine and BE are extracted after 30 days), probably because of a greater lipophilicity of the brain tissue that delays the penetration of the fixing solution into the core of the tissue fragment. Important parameters to be considered are obviously the shape and dimensions of the fixed specimens. Investigating on a forensic casework (see Section 2.7 for details on the case) we observed that, even after 3 months of formalin-fixation, cocaine was quantifiable in the central portion of a large globular piece of liver (volume 60 cm3, cocaine concentration 0.04 mg/g), being undetectable in the core portion of the smaller fragment (Fig. 1C and Table 1). The extraction rate of cocaine and BE from tissues seems, however, to be quite difficult to predict, depending not only on the morphological and metrical characteristics of the fixed specimens, but also on the hydrophilicity of the fixed tissues, and on the quantity, temperature and pH of the formalin solution [10,13]. Table 1 Investigation on a real forensic case. Concentrations of cocaine and BE in formalin solution, in the central portion of the formalin-fixed liver specimens (Fig. 1C), and in the liver tissue at exhumation (see Section 2.7 for details on the case). Specimen
Cocaine (mg/g–mg/mL)
BE (mg/g–mg/mL)
Formalin-fixed liver fragment (1); V 5 cm3 Formalin-fixed liver fragment (2); V 60 cm3 Formalin solution Liver tissue at exhumation
–
0.22
0.04
0.45
– 0.21
0.07 2.25
G. Viel et al. / Forensic Science International 193 (2009) 79–83
Additionally, the hydrolysis of cocaine into BE (half-life of about 7 days) observed in buffered formalin (the most widely used fixing solution) drastically affects cocaine/BE ratio further limiting the relevance of quantitative data interpretation. 5. Conclusions Formalin solutions and formalin-fixed tissues can be used for cocaine analysis only if a short time period has passed since the fixation beginning. The rapid extraction process of cocaine into formalin and the concomitant hydrolysis to BE may prevent the identification of cocaine in both tissues and buffered formalin solution already at 15–30 days after fixation. Moreover, the variable extraction rate of analytes into formalin, which depends on both tissue and fixation-fluid characteristics (such as shape, dimensions and lipophilicity of the fixed specimens, quantity and pH of the formalin solution), along with the hydrolysis of cocaine to BE, significantly affects tissue concentrations and cocaine/BE ratio, thus considerably complicating the interpretation of quantitative results. References [1] European Monitoring Centre for Drugs and Drug Addiction (EMCDDA), Annual Report 2008: The State of the Drugs Problem in Europe, 2009. Office for Official Publications of the European Communities, Luxembourg, Available at: http:// www.emcdda.europa.eu/publications/annual-report/2008. [2] F. Schifano, J. Corkery, Cocaine/crack cocaine consumption, treatment demand, seizures, related offences, prices, average purity levels and deaths in the UK (1990–2004), J. Psychopharmacol. 22 (2008) 71–79. [3] P.S. Tirumalai, D.M. Shakleya, P.M. Gannett, P.S. Callery, T.M. Bland, T.S. Tracy, Conversion of methamphetamine to N-methyl-methamphetamine in formalin solutions, J. Anal. Toxicol. 29 (2005) 48–53. [4] T. Takayasu, K. Saito, J. Nishigami, T. Ohshima, T. Nagano, Toxicological analysis of drugs and poisons in formalin-fixed organ tissues. 2. Volatile substances, Int. J. Legal Med. 107 (1994) 7–12. [5] D.M. Shakleya, J.C. Kraner, J.A. Kaplan, P.M. Gannett, P.S. Callery, Identification of N,N-dimethylamphetamine formed by methylation of methamphetamine in formalin-fixed liver tissue by multistage mass spectrometry, Forensic Sci. Int. 157 (2006) 87–92. [6] J. Nishigami, T. Takayasu, T. Ohshima, Toxicological analysis of the psychotropic drugs chlorpromazine and diazepam using chemically fixed organ tissues, Int. J. Legal Med. 107 (1995) 165–170.
83
[7] T.L. Kuo, Release of tissue paraquat into formalin solution during fixation, J. Forensic Sci. 35 (1990) 668–674. [8] P.M. Gannett, J.R. Daft, D. James, B. Rybeck, J.B. Knopp, T.S. Tracy, In vitro reaction of barbiturates with formaldehyde, J. Anal. Toxicol. 25 (2001) 443– 449. [9] T. Takayasu, K. Holterman, T. Ohshima, D.J. Pounder, Stability of tricyclic antidepressants in formalin solutions, J. Forensic Sci. 43 (1998) 1213–1219. [10] C.L. Winek, F.M. Esposito, D.P. Cinicola, The stability of several compounds in formalin fixed tissues and formalin–blood solutions, Forensic Sci. Int. 44 (1990) 159–168. [11] C.L. Winek, N.R. Zaveri, W.W. Wahba, The study of tricyclic antidepressants in formalin-fixed human liver and formalin solutions, Forensic Sci. Int. 61 (1993) 175–183. [12] T. Takayasu, T. Ohshima, J. Nishigami, T. Kondo, Z. Lin, M. Ohtsuji, T. Nagano, Toxicological analysis for drugs and poisons using the formalin-fixed organ tissues. 1. Methamphetamine, Nihon Hoigaku Zasshi 48 (1994) 33–37. [13] S. Ramagiri, S.K. Shukla, P.K. Sai Prakash, Stability study of fluoxetine in formalinfixed liver tissue, J. Anal. Toxicol. 30 (2006) 692–696. [14] R. Suma, P.K. Sai Prakash, Conversion of sertraline to N-methyl sertraline in embalming fluid: a forensic implication, J. Anal. Toxicol. 30 (2006) 395– 399. [15] M. Cingolani, M. Cippitelli, R. Froldi, V. Gambaro, G. Tassoni, Detection and quantitation analysis of cocaine and metabolites in fixed liver tissue and formalin solutions, J. Anal. Toxicol. 28 (2004) 16–19. [16] R.C. Baselt, Stability of cocaine in biological fluids, J. Chromatogr. 268 (1983) 502– 505. [17] E.R. Garrett, K. Seyda, Prediction of stability in pharmaceutical preparations. XX. Stability evaluation and bioanalysis of cocaine and benzoylecgonine by highperformance liquid chromatography, J. Pharm. Sci. 72 (1983) 258–271. [18] J.B. Murray, H. Al-Shoura, Stability of cocaine hydrochloride in aqueous solution [proceedings], J. Pharm. Pharmacol. 28 (Suppl.) (1976) 24P. [19] C. Klein, Y. Balash, L. Pollak, J. Hiss, M.J. Rabey, Body packer: cocaine intoxication, causing death, masked by concomitant administration of major tranquilizers, Eur. J. Neurol. 7 (2000) 555–558. [20] R.J. Henning, J. Cuevas, Cocaine activates calcium/calmodulin kinase II and causes cardiomyocyte hypertrophy, J. Cardiovasc. Pharmacol. 48 (2006) 802– 813. [21] E. Bertol, C. Trignano, M.G. Di Milia, M. Di Padua, F. Mari, Cocaine-related deaths: an enigma still under investigation, Forensic Sci. Int. 176 (2008) 121– 123. [22] A.J. Jenkins, B. Levine, J. Titus, J.E. Smialek, The interpretation of cocaine and benzoylecgonine concentrations in postmortem cases, Forensic Sci. Int. 101 (1999) 17–25. [23] B.S. Finkle, K.L. McCloskey, The forensic toxicology of cocaine (1971–1976), J. Forensic Sci. 23 (1978) 173–189. [24] A.W. Jones, A. Holmgren, F.C. Kugelberg, Concentrations of cocaine and its major metabolite benzoylecgonine in blood samples from apprehended drivers in Sweden, Forensic Sci. Int. 177 (2008) 133–139. [25] V.R. Spiehler, D. Reed, Brain concentrations of cocaine and benzoylecgonine in fatal cases, J. Forensic Sci. 30 (1985) 1003–1011. [26] S.B. Karch, Cocaine cardiovascular toxicity, South. Med. J. 98 (2005) 794–799.
Contents lists available at ScienceDirect
Forensic Science International journal homepage: www.elsevier.com/locate/forsciint
Stability of cocaine in formalin solution and fixed tissues Guido Viel, Alessandro Nalesso, Giovanni Cecchetto, Massimo Montisci, Santo Davide Ferrara * Section of Legal Medicine - Forensic Toxicology and Antidoping Unit, Department of Environmental Medicine and Public Health, Via Falloppio 50, 35121 Padova, Italy
A R T I C L E I N F O
A B S T R A C T
Article history: Received 22 May 2009 Received in revised form 9 September 2009 Accepted 13 September 2009 Available online 20 October 2009
Embalming and formalin fixation are common, and yet they can create problems for the forensic scientist if a drug has been the cause of death and if the only available specimens to be analyzed are formalin-fixed tissues. Previous studies have demonstrated that during fixation xenobiotics are extracted into formalin according to tissue and fixing solution characteristics. In some cases formalin can react with the analyte resulting in the production of new chemical entities. Regarding cocaine and its metabolites, Cingolani et al. have reported that formalin-fixation extracts benzoylecgonine (BE) from tissues and that BE is stable in the fixing solution. However, the stability and kinetic properties of cocaine remain so far unexplored. Our data show that in buffered formalin (pH 7.4) cocaine is hydrolyzed to BE in agreement with a pseudo first-order reaction kinetic (half-life time 7 days), whereas in unbuffered formalin (pH 3.5) it is relatively stable over a period of 30 days. The analysis of brain and liver samples at different fixation times indicates that during fixation an extraction process occurs for both analytes and that the extraction is more efficient in the liver than in the brain, probably because of a greater lipophilicity of the brain tissue. In conclusion, our study demonstrates that formalin-fixed tissues and their fixing solutions can be used for cocaine analysis only if a short time period has passed since the fixation beginning. The rapid extraction process of cocaine into formalin and the concomitant hydrolysis to BE occurring in buffered formalin may prevent the identification of cocaine in both tissues and formalin solution already at 15–30 days after fixation. Moreover, the unpredictable extraction rate of both analytes, along with the hydrolysis of cocaine into BE significantly affects tissue concentrations, thus complicating the interpretation of quantitative results. ß 2009 Elsevier Ireland Ltd. All rights reserved.
Keywords: Forensic toxicology Cocaine Benzoylecgonine Formalin Fixation process
1. Introduction In the second half of the 1990s, a global trend of escalating cocaine use was observed across a number of different countries. Within the European Union, lifetime experience among 15–34year-olds ranges from 0.7% to 12.7%, with the highest levels being found in the United Kingdom [1]. Although use trends are increasing, data regarding the hazards of cocaine are limited. Acute deaths in which cocaine is present without opiates seem to be infrequent in Europe. Despite the limitations of the available information, and considering that many cocaine-related deaths may pass unreported being scheduled as cardiovascular deaths (due to a lack of toxicological analyses), cocaine seems to play a determinant role in between 8% and 12% of drug-related deaths in Germany, Spain, France and Hungary [2]. Especially in non-forensic autopsies, if there are no suspects of recent or chronic drug exposure, body fluids and/or tissues for toxicological investigations (to be preserved at 20 8C) are
* Corresponding author. Tel.: +39 049 8272225; fax: +39 049 663155. E-mail address: [email protected] (S.D. Ferrara). 0379-0738/$ – see front matter ß 2009 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.forsciint.2009.09.010
generally not collected. On the contrary, in most clinical and forensic post-mortem examinations, specimens of various organs are gathered and fixed in buffered formalin for histological examination. In these cases, the only possible way to perform subsequent toxicological analyses is by exhuming the body (impossible in cases of cremation) or by analyzing the formalinfixed tissues collected at autopsy. Considering this premise and the fact that embalming and cremation procedures are ever more widely diffused, it is necessary to widen our knowledge as to stability and kinetic properties of drugs and drugs of abuse in formalin-fixed tissues [3–5]. Previous studies have clarified that during fixation xenobiotics are extracted into formalin according to the hydrophility index of the substance and of the tissue under fixation [6,7]. In some cases formalin can react with the substance in a time, pH and formalin concentration dependent manner resulting in the production of new chemical entities [8–11]. This has been demonstrated for metamphetamines [5,12] and several antidepressants (fluoxetine, sertraline, nortriptiline, desipramine) [10,13,14]. Regarding cocaine and its metabolites, Cingolani et al. [15] have reported that formalin-fixation extracts benzoylecgonine (BE) from tissues and that BE is quantifiable in formalin for months after
80
G. Viel et al. / Forensic Science International 193 (2009) 79–83
Fig. 1. (A) Brain tissue specimens used for the fixation experiments; (B) liver tissue specimens used for the fixation experiments; (C) small (1) and large (2) formalin-fixed liver fragments belonging to a forensic casework (analytical results reported in Table 1).
the fixation. However, the stability and kinetic properties of cocaine remain so far unexplored. Our study was undertaken to provide the forensic community with information regarding the stability of cocaine in formalin-fixed tissues and formalin solutions under various formaldehyde concentrations and pH conditions. 2. Materials and methods 2.1. Solvents and reagents All reagents and solvents were of analytical grade. Dichloromethane, isopropanol, ammonium hydroxide, and hydrochloridric acid were purchased from Sigma–Aldrich (Milano, Italy). Formalin 40% was obtained from Sigma–Aldrich (Milano, Italy), N-methyl-N-trifluoroacetamide with 1% trimethylchlorosilane (MSTFA—1% TMSC) from Pierce (Rockford, IL, USA). Methanol, potassium dihydrogen phosphate and potassium hydroxide were provided by Carlo Erba Reagents (Milano, Italy). Deionized water was produced with a Milli-Q system (Millipore, Bedford, MA, USA). Standard solutions of cocaine and benzoylecgonine (1 mg/mL) were purchased from Cerilliants (Round Rock, TX, USA), while scopolamine (internal standard) was obtained from Promochem-LGC (Teddington, UK). 2.2. Solutions Potassium phosphate buffers (pH 7.4, 10 mM) were prepared by dissolving 1.36 g of potassium dihydrogen phosphate (Carlo Erba reagents, Milano, Italy) in 1 L of deionized water and by adjusting the pH with 0.1 M potassium hydroxide (Carlo Erba reagents, Milano, Italy). Diluted formalin solutions (4% of formaldehyde) at pH 3.5 (unbuffered) and pH 7.4 (buffered) were prepared from the stock formalin solution (40% of formaldehyde) by dilution, respectively, with water and potassium phosphate buffer (pH 7.4, 10 mM). Working solutions of cocaine, BE and scopolamine (IS) at the final concentration of 100 mg/mL were obtained by dilution of the stock solutions with methanol. 2.3. Stability of cocaine in buffered (pH 7.4) and unbuffered formalin solutions (pH 3.5) Cocaine–formalin mixtures (2 mg/mL) at pH 3.5 and 7.4 were obtained by fortification of unbuffered formalin (10%, pH 3.5) and buffered formalin (10%, pH 7.4) with cocaine working solutions. Formalin–cocaine mixtures were stored at room temperature for 4 weeks. Samples were collected and analyzed by GC–MS for cocaine and BE immediately after preparation, and on days 1, 2, 5, 10, 15 and 30. Each experiment was conducted in triplicate. 2.4. Stability of cocaine in formalin-fixed tissues Post-mortem brain and liver tissues collected from a cocaine-related death (brain: cocaine 1.05 mg/g, BE 0.4 mg/g; liver: cocaine 0.75 mg/g, BE 1.2 mg/g; quantification performed according to the method described in Sections 2.5 and 2.6) were used for the following experiment. Five cubic-shaped specimens (volume 3.4 cm3; Fig. 1A and B) of each tissue were fixed in 100 mL of 10% buffered formalin (pH 7.4) and in 100 mL of unbuffered formalin (pH 3.5). Tissue specimens were recovered from formalin at different times (days 2, 5, 10, 15 and 30) along with the fixing solution (5 mL). GC–MS analysis for cocaine and BE was performed on both formalin and tissue samples according to the method described in the following sections. 2.5. Extraction procedures Brain and liver samples (2 g), collected from the central portion of the fixed specimens, were added of potassium phosphate buffer pH 7.4 (6 mL), and
scopolamine working solution (final concentration 2 mg/g). They were homogenized, centrifuged and filtrated. The clear liquid solution was purified by solid phase extraction (SPE) with Bond Elut Certify SPE columns (VARIAN, Harbor City, CA, USA). First, the columns were activated with methanol (2 mL) and potassium phosphate buffer pH 7.4 (2 mL), than samples were loaded and drawn through. Columns were washed with deionized water (3 mL), hydrochloric acid 0.1 M (3 mL) and methanol (3 mL), and then allowed to dry at full vacuum (5 min). Analytes were recovered by elution with a dichloromethane/isopropanol/ammonium hydroxide mixture (78:20:2 v/v; 2 mL). The eluates were gently evaporated to dryness under a stream of nitrogen, derivatized with MSTFA 1% TMSC (50 mL, 75 8C for 15 min), and analyzed by GC–MS. Formalin solutions (0.5 mL), diluted with potassium phosphate buffer pH 7.4 (1.5 mL), were added of scopolamine working solution (final concentration 2 mg/ mL). They were centrifuged, and purified by solid phase extraction (SPE) with the same procedure described for tissue samples. 2.6. GC–MS analysis Analysis was performed on an Agilent 6890N GC system (Agilent Technologies, Palo Alto, CA, USA) equipped with an Agilent 5973N single quadrupole mass spectrometer (Agilent Technologies, Palo Alto, CA, USA). The MS was tuned with perfluorotributylamine (PFTBA) using the autotune function as per manufacturer’s recommendation. Data analysis was performed with a HP ChemStation software. Chromatographic separation was performed with an ULTRA-1 fused-silica capillary column (12 m 200 mm i.d. 0.33 mm film thickness) (Agilent Technologies, Palo Alto, CA, USA). GC–MS analysis was performed by injection of derivatized samples (1 mL) in splitless mode (purge time 0.5 min; purge flow 40 mL/min). The inlet temperature was set at 250 8C and the helium carrier gas at a constant flow-rate of 0.6 mL/min. An initial isotherm of 75 8C was maintained for 0.5 min and ramped at 20 8C/min to a temperature of 200 8C. The second ramp rate was 25 8C/min to an isotherm of 300 8C, which was maintained for 5 min. The MSD transfer line temperature was set at 280 8C and that of the quadrupole and source at 150 8C and 230 8C, respectively. A solvent delay time of 5 min was set. All mass spectra were recorded at 70 eV (electron impact, positive mode) with an EM offset of +200 V. Analysis was performed in the selected-ion monitoring (SIM) mode with respect to three significant ions for each compound (cocaine m/z 82, 182, 303; BE m/z 82, 240, 361; scopolamine m/z 94, 154, 375; the underlined ions were selected as quantifiers). Quantification was performed by means of a five-point calibration curve prepared by spiking 10% formalin solutions, and drug-free formalin-fixed homogenized brain and liver tissues at the following final concentrations: 0.02, 0.05, 0.2, 0.5, 1 and 2 mg/mL. Good linear regressions were obtained (y = 0.185x + 1.32, R2 = 0.965 for formalin; y = 0.611x + 3.75, R2 = 0.972 for brain; y = 0.621x + 2.75, R2 = 0.982 for liver). Accuracy and inter-day precision (calculated as relative standard deviation and bias) were always better than 12%. According to validation results the lower limit of quantification (LLOQ) was determined to be 0.02 mg/mL. 2.7. Investigation on a forensic case At non-forensic autopsy, a 33-year-old man unexpectedly died at his home was found to be affected by a transmural infarction of the antero-lateral wall of the left ventricle. Three months after the death the trial, on the basis of novel circumstantial data indicating a possible drug-related fatality, charged a physician and a toxicologist of our Unit with examining the autopsy report and performing toxicological investigations. Because of the presence of several formalin-fixed tissues (in particular a large globular piece of liver along with a smaller liver fragment, Fig. 1C) collected by the pathologist, a preliminary GC–MS screening on formalin solution was performed before exhuming the body. Only the presence of BE was identified, thus the central portion of each liver specimen and the formalin solution were analyzed according to the method described in Sections 2.5 and 2.6.
G. Viel et al. / Forensic Science International 193 (2009) 79–83
81
3. Results In unbuffered formalin (pH 3.5) cocaine proved to be stable over a period of 30 days (Fig. 2B), whereas in buffered formalin (pH 7.4) it was hydrolyzed to BE, with about 80% of cocaine being depleted in 15 days (Fig. 2A). Although there are two reactants in the hydrolysis process (cocaine and water), one of them (water) is in large excess so that, as the reaction progresses, only negligible amounts are consumed and its concentration can be considered to stay constant. As demonstrated by the linearity of the logarithmic plot shown in Fig. 3, the hydrolysis of cocaine occurring in buffered formalin (pH 7.4) can be considered pseudo first-order (y = 0.004x + 7.33). This means that the half-life (t1/2) of the reactant (cocaine) is independent of the starting concentration, being expressed by the following law: t 1=2 ¼
ln 2 0:693 ¼ ¼ 144:28 h 7 days k 0:004
Fig. 3. Logarithmic plot of cocaine concentrations versus time (hours). The linearity of the relationship (y = 0.004x + 7.330; R2 = 0.995) indicates that the hydrolysis reaction is well approximated by a pseudo first-order kinetics.
Therefore, regardless to the initial concentration, in 1 week half of the cocaine is converted to BE. Baselt showed that cocaine concentration in urine does not change over a period of 30 days if the sample is acidified to pH 5, and that setting pH at 8 causes a
significant hydrolysis to BE with a half-life of about 8 days [16]. Previous published data indicate that aqueous cocaine is stable within a pH range of 5 or below [17], with an ideal pH for maximum stability of 1.95 [18]. Our results, in substantial agreement with
Fig. 2. Decomposition of cocaine and formation of BE in buffered (A) and unbuffered formalin (B) over a period of 30 days.
Fig. 4. (A) Concentration of cocaine in brain and liver specimens fixed in buffered (pH 7.4) and unbuffered (pH 3.5) formalin for 30 days. (B) Concentration of cocaine in fixing solutions over 30 days.
82
G. Viel et al. / Forensic Science International 193 (2009) 79–83
Fig. 5. (A) Concentration of BE in brain and liver specimens fixed in buffered (pH 7.4) and unbuffered (pH 3.5) formalin for 30 days. (B) Concentration of BE in fixing solutions over 30 days.
these findings, suggest that formalin does not play an active role in the hydrolysis of cocaine, which is substantially stable in unbuffered formalin (pH 3.5) and unstable in buffered formalin (pH 7.4). The analysis of brain and liver samples at different fixation times (see Section 2.4 for details) showed that during fixation both cocaine and BE were rapidly extracted into formalin solution (Figs. 4 and 5). In the liver, the extraction process for both analytes was faster and more efficient than in the brain, probably because of a greater lipophilicity of the brain tissue (Figs. 4A and 5A). Unbuffered formalin (pH 3.5) demonstrated an overall enhanced extraction capability by comparison with buffered formalin and showed once again its stabilizing properties for cocaine. Indeed, in both liver and brain fixation experiments, the final cocaine concentration in non-buffered formalin was significantly higher than in buffered formalin (Figs. 4B and 5B). The results of the toxicological investigation performed on formalin solutions and formalin-fixed liver fragments belonging to a forensic casework (see Section 2.7 for details) are reported in Table 1. 4. Discussion Except in cases of massive cocaine exposure (couriers or ‘‘body packers’’), in which acute toxicity is dose-related and is primarily
characterized by sympathomimetic effects (tachycardia, hypertension and hyperthermia arrhythmias) [19], cocaine-related deaths mainly occur after a prolonged drug use. Long-term exposure induces a sequence of changes at molecular, cellular and tissue levels [20], many of which favor the occurrence of sudden cardiac death [21]. However, due to the absence of a linear dose–effect relationship and of a scientifically sound ‘‘lethal dose’’, the presence of cocaine in body fluids and/or tissues must be supported by clear histological signs of chronic drug use or abuse to conclude that cocaine played a role in causing the death [22]. Moreover, the interpretation of toxicological findings is complicated by various factors including route of administration, continued absorption after death, different metabolism in chronic/occasional users or in different body compartments, and ‘‘in vivo’’ or ‘‘in vitro’’ degradation of both parent drug and metabolites. Besides the quantification of cocaine and BE in blood, brain levels should always be determined, in order to evaluate and compare the different cocaine/BE ratios [21–24]. High brain concentrations of cocaine with low or absent metabolite levels indicate a recent exposure, whereas high levels of BE with low or null concentrations of cocaine suggest a relatively remote use [21,25]. These considerations are extremely important while interpreting a suspected drug-related fatality. Indeed, about 30% of cocaine-related deaths occur between 2 and 5 h from the assumption, whereas 65% occur within 5 h, thus presenting a high cocaine/BE ratio in the brain [26]. Our data demonstrate that formalin-fixed tissues (particularly the brain) and formalin solutions may be used as biological matrices for identifying suspected cocaine-related fatalities. However, the rapid and effective extraction process of both cocaine and BE occurring during formalin-fixation significantly modifies tissue concentrations, thus complicating the interpretation of quantitative results. In substantial agreement with the data reported by Cingolani et al. [15], we have found that after 30 days of fixation about 80% of cocaine and BE are extracted from the liver into formalin. Additionally, we have observed that in the brain the extraction efficiency is slightly reduced (60% of cocaine and BE are extracted after 30 days), probably because of a greater lipophilicity of the brain tissue that delays the penetration of the fixing solution into the core of the tissue fragment. Important parameters to be considered are obviously the shape and dimensions of the fixed specimens. Investigating on a forensic casework (see Section 2.7 for details on the case) we observed that, even after 3 months of formalin-fixation, cocaine was quantifiable in the central portion of a large globular piece of liver (volume 60 cm3, cocaine concentration 0.04 mg/g), being undetectable in the core portion of the smaller fragment (Fig. 1C and Table 1). The extraction rate of cocaine and BE from tissues seems, however, to be quite difficult to predict, depending not only on the morphological and metrical characteristics of the fixed specimens, but also on the hydrophilicity of the fixed tissues, and on the quantity, temperature and pH of the formalin solution [10,13]. Table 1 Investigation on a real forensic case. Concentrations of cocaine and BE in formalin solution, in the central portion of the formalin-fixed liver specimens (Fig. 1C), and in the liver tissue at exhumation (see Section 2.7 for details on the case). Specimen
Cocaine (mg/g–mg/mL)
BE (mg/g–mg/mL)
Formalin-fixed liver fragment (1); V 5 cm3 Formalin-fixed liver fragment (2); V 60 cm3 Formalin solution Liver tissue at exhumation
–
0.22
0.04
0.45
– 0.21
0.07 2.25
G. Viel et al. / Forensic Science International 193 (2009) 79–83
Additionally, the hydrolysis of cocaine into BE (half-life of about 7 days) observed in buffered formalin (the most widely used fixing solution) drastically affects cocaine/BE ratio further limiting the relevance of quantitative data interpretation. 5. Conclusions Formalin solutions and formalin-fixed tissues can be used for cocaine analysis only if a short time period has passed since the fixation beginning. The rapid extraction process of cocaine into formalin and the concomitant hydrolysis to BE may prevent the identification of cocaine in both tissues and buffered formalin solution already at 15–30 days after fixation. Moreover, the variable extraction rate of analytes into formalin, which depends on both tissue and fixation-fluid characteristics (such as shape, dimensions and lipophilicity of the fixed specimens, quantity and pH of the formalin solution), along with the hydrolysis of cocaine to BE, significantly affects tissue concentrations and cocaine/BE ratio, thus considerably complicating the interpretation of quantitative results. References [1] European Monitoring Centre for Drugs and Drug Addiction (EMCDDA), Annual Report 2008: The State of the Drugs Problem in Europe, 2009. Office for Official Publications of the European Communities, Luxembourg, Available at: http:// www.emcdda.europa.eu/publications/annual-report/2008. [2] F. Schifano, J. Corkery, Cocaine/crack cocaine consumption, treatment demand, seizures, related offences, prices, average purity levels and deaths in the UK (1990–2004), J. Psychopharmacol. 22 (2008) 71–79. [3] P.S. Tirumalai, D.M. Shakleya, P.M. Gannett, P.S. Callery, T.M. Bland, T.S. Tracy, Conversion of methamphetamine to N-methyl-methamphetamine in formalin solutions, J. Anal. Toxicol. 29 (2005) 48–53. [4] T. Takayasu, K. Saito, J. Nishigami, T. Ohshima, T. Nagano, Toxicological analysis of drugs and poisons in formalin-fixed organ tissues. 2. Volatile substances, Int. J. Legal Med. 107 (1994) 7–12. [5] D.M. Shakleya, J.C. Kraner, J.A. Kaplan, P.M. Gannett, P.S. Callery, Identification of N,N-dimethylamphetamine formed by methylation of methamphetamine in formalin-fixed liver tissue by multistage mass spectrometry, Forensic Sci. Int. 157 (2006) 87–92. [6] J. Nishigami, T. Takayasu, T. Ohshima, Toxicological analysis of the psychotropic drugs chlorpromazine and diazepam using chemically fixed organ tissues, Int. J. Legal Med. 107 (1995) 165–170.
83
[7] T.L. Kuo, Release of tissue paraquat into formalin solution during fixation, J. Forensic Sci. 35 (1990) 668–674. [8] P.M. Gannett, J.R. Daft, D. James, B. Rybeck, J.B. Knopp, T.S. Tracy, In vitro reaction of barbiturates with formaldehyde, J. Anal. Toxicol. 25 (2001) 443– 449. [9] T. Takayasu, K. Holterman, T. Ohshima, D.J. Pounder, Stability of tricyclic antidepressants in formalin solutions, J. Forensic Sci. 43 (1998) 1213–1219. [10] C.L. Winek, F.M. Esposito, D.P. Cinicola, The stability of several compounds in formalin fixed tissues and formalin–blood solutions, Forensic Sci. Int. 44 (1990) 159–168. [11] C.L. Winek, N.R. Zaveri, W.W. Wahba, The study of tricyclic antidepressants in formalin-fixed human liver and formalin solutions, Forensic Sci. Int. 61 (1993) 175–183. [12] T. Takayasu, T. Ohshima, J. Nishigami, T. Kondo, Z. Lin, M. Ohtsuji, T. Nagano, Toxicological analysis for drugs and poisons using the formalin-fixed organ tissues. 1. Methamphetamine, Nihon Hoigaku Zasshi 48 (1994) 33–37. [13] S. Ramagiri, S.K. Shukla, P.K. Sai Prakash, Stability study of fluoxetine in formalinfixed liver tissue, J. Anal. Toxicol. 30 (2006) 692–696. [14] R. Suma, P.K. Sai Prakash, Conversion of sertraline to N-methyl sertraline in embalming fluid: a forensic implication, J. Anal. Toxicol. 30 (2006) 395– 399. [15] M. Cingolani, M. Cippitelli, R. Froldi, V. Gambaro, G. Tassoni, Detection and quantitation analysis of cocaine and metabolites in fixed liver tissue and formalin solutions, J. Anal. Toxicol. 28 (2004) 16–19. [16] R.C. Baselt, Stability of cocaine in biological fluids, J. Chromatogr. 268 (1983) 502– 505. [17] E.R. Garrett, K. Seyda, Prediction of stability in pharmaceutical preparations. XX. Stability evaluation and bioanalysis of cocaine and benzoylecgonine by highperformance liquid chromatography, J. Pharm. Sci. 72 (1983) 258–271. [18] J.B. Murray, H. Al-Shoura, Stability of cocaine hydrochloride in aqueous solution [proceedings], J. Pharm. Pharmacol. 28 (Suppl.) (1976) 24P. [19] C. Klein, Y. Balash, L. Pollak, J. Hiss, M.J. Rabey, Body packer: cocaine intoxication, causing death, masked by concomitant administration of major tranquilizers, Eur. J. Neurol. 7 (2000) 555–558. [20] R.J. Henning, J. Cuevas, Cocaine activates calcium/calmodulin kinase II and causes cardiomyocyte hypertrophy, J. Cardiovasc. Pharmacol. 48 (2006) 802– 813. [21] E. Bertol, C. Trignano, M.G. Di Milia, M. Di Padua, F. Mari, Cocaine-related deaths: an enigma still under investigation, Forensic Sci. Int. 176 (2008) 121– 123. [22] A.J. Jenkins, B. Levine, J. Titus, J.E. Smialek, The interpretation of cocaine and benzoylecgonine concentrations in postmortem cases, Forensic Sci. Int. 101 (1999) 17–25. [23] B.S. Finkle, K.L. McCloskey, The forensic toxicology of cocaine (1971–1976), J. Forensic Sci. 23 (1978) 173–189. [24] A.W. Jones, A. Holmgren, F.C. Kugelberg, Concentrations of cocaine and its major metabolite benzoylecgonine in blood samples from apprehended drivers in Sweden, Forensic Sci. Int. 177 (2008) 133–139. [25] V.R. Spiehler, D. Reed, Brain concentrations of cocaine and benzoylecgonine in fatal cases, J. Forensic Sci. 30 (1985) 1003–1011. [26] S.B. Karch, Cocaine cardiovascular toxicity, South. Med. J. 98 (2005) 794–799.