Determination of Direct Oral Anticoagulants in Forensic Toxicology

Oral anticoagulants have been used for more than 50 years as antithrombotic agents, particularly in the primary and secondary prophylaxis of thromboembolic disorders. Our study’s purpose is to synthetize and analyze the toxicological methods used to determine the oral anticoagulants, described in the literature, so we can identify the optimal methods, taking into consideration the diversity of forensic medicine workload. In the electronic databases, for the period 1990-2019, there were identified 56 articles that included the searched keywords. Of these, 28 were included in our study. We excluded the articles that were not lining up with the objective of our study, the update-type articles and the case studies. After analyzing the literature, it was observed that liquid chromatography-mass spectrometry is the ideal method for detecting low-dosed drugs, such as oral anticoagulants.


Experimental part Material and method
Our study's purpose is to synthetize and analyze the toxicological methods used to determine the oral anticoagulants, described in the literature, in order to identify the optimal methods, taking into consideration the diversity of forensic casuistry.
The articles were searched in the electronic databases platforms PUBMED, RESEARCHGATE and WEB OF SCIENCE, using keyword combinations in English, such as "legal medicine, forensic, toxicology, anticoagulant, oral, intoxication, death, method, chromatography, spectrometry". Studies containing data on toxicological methods for determining oral anticoagulants were selected.

Results and discussions
In the electronic databases, for the period 1990-2019, 56 articles were identified that corresponded to the above mentioned search keywords. Of these, 28 were included in our study. We excluded the articles that were not lining up with the objective of our study, the update-type articles and the case studies.

Choice of biosamples
The concentration of drugs is relatively high in the urine, so this is the sample of choice for comprehensive screening or for identifying unknown drugs, and more importantly, to identify the metabolites of these drugs [22][23][24]. Blood, plasma or serum is the sample of choice in the case of quantitative analysis, but, when the concentration of the drug is very high, this sample can also be used for screening. This can be advantageous, as frequently at the autopsy, it is the only sample that can be collected [25][26][27][28][29][30][31][32][33][34][35][36][37]. Toxicological analysis of hair shafts allows the detection of chronic drug use, but this sample is controversial due to external contamination [38,39]. In the literature, several studies have been identified that used tissues taken from the body as a sample, such as liver or brain [40][41][42][43].

Sample preparation
In the toxicological determination of oral anticoagulants it is very important to find a suitable preparation method for each type of sample. This may involve cleavage of conjugates, extraction and derivatization preceded or followed by clean-up steps. For LC-MS it is known that the method of preparation of the sample is the least complicated. Nonetheless, a more or less selective extraction is also important for LC-MS [44], especially in order to avoid ion suppression effects [44][45][46][47][48]. Cleavage of conjugates can be performed by fast acid hydrolysis or by gentle, but timeconsuming enzymatic hydrolysis [49]. Isolation can be performed by liquid-liquid extraction (LLE) at a pH at which the analyte is non-ionized or by solid-phase extraction (SPE) preceded or followed by clean-up steps. Depending on the type of the sample, for SPE, it is necessary to apply a pretreatment, for blood and tissues some steps of deproteinization and centrifugation are necessary, whereas for urine, a simple dilution or centrifugation is sufficient [50]. Most recent studies have introduced a new extraction method, called solid-phase microextraction (SPME), which is based on adsorption of the analyte on a stationary phase coating to fine rod of fused silica [51,52]. Derivatization steps are necessary, if relatively polar compounds containing carboxylic, hydroxy, primary or secondary amino groups are to be determined, and if electronegative moieties have to be introduced into the molecule for sensitive negative ion chemical ionization (NICI) detection [53].

Chromatographic and electrokinetic procedures
The techniques for toxicological determination of oral anticoagulants analyzed in the literature are summarized in Table 1. TLC Serum LLE UV scanner 200 ng/ml GC-MSgas chromatography-mass spectrometry, LCliquid chromatography; http://www.revistadechimie.ro TLCthin-layer chromatography; EX-MEextractive methylation; SPEsolid-phase extraction; LLEliquid-liquid extraction; EIelectron impact ionization; FLfluorescence detector / detection; DAD -(photo) diode array detector / detection; DDCI -2,6-dichloroquinone-4-chloroimide Linearity Baranowska et al. [22] evaluated the calibration curves at six concentration levels, and for each concentration triplicate injections were applied. In the study of Fourel et al. [25], five replicates of spiked drug-free dog plasma samples for four different concentrations were performed for a total of twenty extractions and the peak area was then used in conjunction with the calibration curve to determine the concentration in contaminated plasma samples. Di Rago et al. [27] evaluated the linearity based on comparisons of regression analysis with no weighting, 1/x and 1/x2. In the study of Gonzales et al. [29], for calibration purposes, the most suitable solution was chosen by analyzing the relative error values of the calibration standard concentrations and the linearity of the calibration curves. Denooz et al. [31] used the absolute beta-expectation tolerance limits, with acceptance limits at 20% expressed in the concentration units.

Precision and accuracy
Baranowska et al. [22] evaluated this criterion using standard mixture solution of the drugs under optimal conditions six times in one day for intra-day variation and three times a day on three consecutive days for the inter-day variation, expressed as the relative standard deviation. In the study of Fourel et al. [25], the intra-day accuracy was evaluated by performing five replicates of four spiked samples and the inter-day precision was evaluated three different days by performing two replicates of a spiked sample chosen in the middle of the concentration range. Di Rago et al. [27] evaluated the precision and accuracy of performing calculations using a full seven-point calibration curve and compared with one-point calibrations. To evaluate the accuracy and precision of the method, Gonzales et al. [29] analyzed the spiked samples with low, medium and high drug concentrations.

Specificity
To assess the specificity of the method, Baranowska et al. [22] used drug-free samples, and the retention times of endogenous compounds in the urine were compared with those of direct oral anticoagulants. Fourel et al. [25] tested the specificity for the 14 compounds with 20 blank plasmas, and there were no interferences to be found. Di Rago et al. [27] used post-mortem blood samples sent to the laboratory for toxicological analysis and whole blood samples collected from drug-free volunteers, in order to evaluate the method's specificity. Gonzales et al. [29] used only 6 blank plasma samples from different healthy volunteers.

Applicability of the method
Baranowska et al. [22] analyzed samples taken from hospitalized patients who were treated with direct oral anticoagulants. The concentrations of the drug ranged between 3.14-6.81 µg mL -1 . Di Rago et al. [27] analyzed 397 post-mortem blood samples and the cases that were above the highest calibrator were diluted appropriately to concentrations within the calibration range. Gonzales et al. [29] obtained plasma samples from 19 different patients under treatment with acenocoumarol between 1 and 20 hours after the oral intake of the drug. Besides studying quantitation and confirmation transitions, their ratio was measured and compared with the ratio obtained for a standard solution.

Conclusions
Systematic toxicological analysis covers the basic drugs, which are also the most important toxicants. However, it is necessary that drugs such as oral anticoagulants should be taken into consideration and be analyzed in forensic cases.
After analyzing the literature, it is observed that liquid chromatography-mass spectrometry is the ideal method to detect more polar, unstable or low-dosed drugs, such as oral anticoagulants. Moreover, high-performance liquid chromatography (HPLC) method with ultraviolet (UV) detection has been developed and validated for acenocoumarol, which is the antithrombotic of choice in Romania.