The stability of cyanide in human biological samples. A systematic review, meta-analysis and determination of cyanide (GC-QqQ-MS/MS) in an authentic casework 7 years after fatal intoxication

Abstract A 30 year old man was found with no signs of life in front of the house. The cyanide concentration in blood and urine was determined five years after the man’s death. What is more, a stability study was conducted for 730 days in an authentic casework blood sample. Sample preparation procedure included precipitation with methanol:water mixture, solid phase extraction (SPE) and derivatization with the use of PFB-Br (pentafluorobenzyl bromide). The sample was analyzed using GC-QqQ-MS/MS (gas chromatopraphy coupled with tandem mass spectrometry) isotope dilution method. Separation was done using a SH-RXI-5MS column (30 m x 0.25 mm, 0.25 µm). Detection of PFB-CN and PFB-13CN was achieved using a triple-quadrupole mass spectrometer with an electron ionization (EI) ion source in multiple reaction monitoring (MRM) mode. After 5 years from the man’s death, cyanide concentration was: 1900 ng/mL in blood and 500 ng/mL in urine. Stability study performed in an authentic blood sample 6 and 7 years after the man’s death revealed cyanide concentrations of 1898.2 ng/mL and 1618.7 ng/mL, respectively. While spectrophotometric and colorimetric methods recorded both decrease and increase in cyanide concentration over time, newer chromatographic methods mainly indicate a decrease. The studies presented in this paper seem to confirm this trend. However, in order to interpretate the results of cyanide concentration in biological material reliably, more research is still necessary.


Introduction
The global crisis related with introduction and distribution of new psychoactive substances (NPS), as well as numerous fatal intoxications as a result of overdosing of designer drugs, pose a huge challenge to forensic toxicology in terms of rapid detection and identification of novel organic structures.On the other hand, the increasing number of reports regarding the use of old inorganic poisons for suicidal purposes also seems alarming.In the era of a widespread internet access, the availability of such substances for purchase by individuals is surprisingly large.Cyanide compounds can be found not only on darkweb in the form of so-called suicide kits (Le Garff et al. 2016) but also on popular websites and online marketplaces such as AmazonV R or eBayV R (Cantrell 2005;James et al. 2020).From early 2000s the number of suicide attempts with the use of azide (Wachełko et al. 2021a), nitrate and nitrite (McCann et al. 2021) as well as cyanide (Wachełko et al. 2021b) salts has increased dramatically and became more popular than in previous decades.For this reason, careful monitoring of these anions' concentration in postmortem biological samples seems to be necessary nowadays.
In the case of other inorganic anions, (e.g.azide) their stability in biological material is described similarly among different studies (Le Blanc-Louvry et al. 2012;Wachełko et al. 2021a).However, in the case of cyanide stability there are huge discrepancies.The results of investigations carried out between 1938 and 2022 (Gettler and Baine 1938;Curry et al. 1967;Ballantyne et al. 1973Ballantyne et al. , 1974;;Ballantyne 1977;Egekeze and Oehme 1979;Lokan et al. 1987;Chikasue et al. 1988;Bright et al. 1990;Mateus et al. 2005;Bhandari et al. 2012;Desharnais et al. 2012;Bhandari et al. 2014a;Osak et al. 2021) indicate the possibility of a decrease in cyanide postmortem levels, the stability of these anions over time, as well as an increase of their concentration even to toxic and lethal concentrations in only 2 weeks.This fact can cause difficulties in the interpretation of the toxicological findings and in the evaluation of whether there has been a fatal intoxication with the use of cyanide compounds.
This paper aims to apply a gas chromatography-triple quadrupole-tandem mass spectrometry (GC-QqQ-MS/MS) isotope dilution method for quantification of cyanide ions (CN -) in authentic blood and urine samples 5 years after autopsy.A meta-analysis of papers published to date on cyanide stability was carried out.Also a stability study was conducted for 730 days in an authentic casework blood sample, in order to investigate if the cyanide concentration in postmortem specimens decreases, increases or is stable in time.

Case history
A 30 year old man was found with no signs of life by his parents in the garden in front of the house.An emergency physician confirmed the man's death.As parents stated, the man had informed them before his death that he no longer wanted to live.On the grass next to the deceased laid an empty bottle with the label 'potassium cyanide' and a bottle with the liquid substance used as diluent.According to the information provided by the victim's mother, he was under permanent psychiatric treatment for 5 years.The man had been diagnosed with schizophrenia.External and internal examination of the corpse did not reveal any traumas or lesions that could result in death (autopsy found only acute hemorrhagic gastritis and red colored gastric content).According to the autopsy report, the death was due to respiratory arrest secondary to cyanide poisoning.The participation of third parties in the death was excluded.For further toxicological analysis biological fluids (blood and urine) were collected.The preservative agent used in the test tube with blood was a mix of EDTA with NaF.A urine sample was collected into a test tube without any preservation agent.The investigation revealed that the man had an e-mail correspondence with a corporation that distributed chemicals to individuals (including such hazardous substances as potassium cyanide).A few days before his death, the man had bought 1 gram of potassium cyanide over the Internet.Initially, another laboratory tested biological material and revealed sertraline in blood in a concentration of 0.18 mg/mL.The cyanide concentration in blood was stated to be <0.2 mg/mL which the authors believe was the lowest limit of quantification of the method used.However, we do not possess information about which spectroscopic method was used nor about the sample preparation procedure, as well as the derivatization method that was used in order to determine cyanide ions.Given that the toxicological results obtained did not concur with the circumstances of death and the autopsy report, it was decided to obtain a second opinion.Five years after the death, a blood sample (approximately 1 mL) and a small amount of urine were sent to our laboratory for toxicological analysis for the presence of cyanide and drugs.Unfortunately, the amount of urine sample was insufficient for conducting a stability examination, therefore only blood sample was analyzed.

Sample preparation
A 200 mL of biological sample was transferred into 2-mL Eppendorf tube (polypropylene; Sarstedt, N€ umbrecht, Germany), and mixed with 20 mL of IS solution ( 13 CN -, 5000 ng/mL of isotope labeled cyanide ions).Next, precipitation with 500 mL ice cold mixture of methanol and water (v/v 4:1) was carried out.The samples were centrifuged at 13 500 rpm at 4 � C for 10 min in order to remove proteins.The supernatant collected from blood sample was additionally cleaned by removing phospholipids with the use of Phree TM tubes.Next, to both urine and blood samples, 300 mL of saturated sodium tetraborate (borax) water solution (pH 9) and 100 mL of acetonic PFB-Br solution (100 mM) were added and the tube was vortex mixed for 30s.Derivatization was carried out in 65 � C for 55 min.Next, liquid-liquid extraction (LLE) of derivatization product (pentafluorobenzyl cyanide) with 300 mL of n-hexane was carried out for 10 min by vortex mixing.The 50 mL of organic solution was then transferred into glass inserts of autosampler vials and analyzed by GC-QqQ-MS/MS.The summary of sample preparation procedure is presented in the Figure 1.

Instrumentation
Analyses were performed using a gas chromatograph (GC, Shimadzu QP2010; Kyoto, Japan) operated with an autosampler AOC-20s and autoinjector Milan,Italy).The separation was done using a SH-RXI-5MS column (30 m x 0.25 mm i.d., 0.25 mm film thickness; Shimadzu, Bellefonte, Pennsylvania, USA).The column temperature was initially held at 60 � C for 2 min, increased at 20 � C/min to 180 � C (held there for 1.0 min) and eventually temperature increased at 40 � C/min to 300 � C (and held there for 2.0 min).Helium (purity 6.0, Messer, Gumpoldskirchen, Austria) was used as a carrier gas at a flow-rate of 1.1 mL/min.The syringe size was 10 mL.The injection volume was 1.0 lL.A splitless injection mode was applied with sampling time of 1.0 min.The injector temperature was 260 � C. The syringe was set to auto cleaning by pre-injecting ethyl acetate and acetonitrile.The total run time was 14.0 min.
Detection of PFB-CN and PFB-13 CN was achieved using a triple-quadrupole mass spectrometer (QqQ, Shimadzu TQ8040, Kyoto, Japan).The spectrometer was equipped with an electron impact (EI) ion source; determination of the pentafluorobenzyl derivatives was carried out in the multiple reaction monitoring (MRM) mode.The following MS parameters were fixed: ion source temperature, 200 � C; interface temperature, 280 � C; electron ionization energy was 70 eV; the detector voltage was set at 1.6 kV.The following MRM transitions were chosen for quantitative analysis: PFB-CN: 207 !157.0, 188.0, 181.0 m/z and PFB-13 CN: 208.0 !158.0, 189.0, 181.0.A summary of precursor and product ions, collision energies, loop time, and retention time for each compound is presented in the Supplementary Material together with details regarding method development and validation.

Toxicological findings and stability of cyanide in authentic blood sample
Analysis performed 5 years after the biological sample was collected with the use of the developed GC-EI-QqQ-MS/MS method revealed cyanide concentration: 1900 ng/mL in blood and 500 ng/mL in urine, which are in a range of toxic concentrations found in fatal intoxication cases (cyanide concentration range 1.1-53 mg/mL for blood samples and 0.5-1.1 mg/mL for urine samples (Baselt 2017)).Multiple reaction monitoring (MRM) chromatograms and product ion scan spectra of cyanide and IS in examined biological samples are presented in the Supplementary Material.Other toxicological analysis performed with the use of UHPLC-QqQ-MS/MS found presence of three neuroleptic and antidepressant drugs: zuclopenthixol, sertraline and paliperidone at concentrations of: 17.6 ng/mL, 29.5 ng/mL, 3.0 ng/mL in blood and 3.6 ng/ mL, 21.8 ng/mL, 51.8 ng/mL in urine, respectively.Stability study performed in an authentic blood sample after another 365 days (6 years after the man's death) and 730 days (7 years after the man's death) revealed cyanide concentrations of 1898.2 ng/mL and 1618.7 ng/mL, respectively.

Meta-analysis: stability of cyanide in biological samples
A meta-analysis based on articles published over the past 100 years that have examined cyanide stability was conducted.A systematic review of the world literature concerning aforementioned topic was performed with the use of PubMed (using the search terms: 'Cyanides'[Mesh] AND 'stability' AND 'biological material' and 'Cyanides'[Mesh] AND 'stability' AND 'blood') and Google Scholar (using the search terms: 'Cyanide stability in biological material').Sixteen articles related to the biological material studied in this paper have passed verification criteria.In the mentioned literature it was shown that it is possible for cyanide concentration to both increase and decrease, as well as to remain relatively stable over time.However, there are some correlations that have not been considered so far, which can be a valuable source of information both in terms of interpreting the results of toxicological studies and designing stability studies in the future.A summary of the results of the conducted stability studies of cyanide in biological material can be found in the Table 1.In 1967, Curry et al. (1967) published the first scientific report suggesting the possibility of postmortem cyanide formation, the concentration of which in biological material can reach the order of toxic concentrations as early as several days after death.In the study, a colorimetric method of cyanide analysis was used.The article described three cases: (1) a blood sample where initial cyanide content was 0.0 mg/mL, while after one month the cyanide concentration was 5 mg/ mL; (2) a blood sample which initial cyanide concentration was about 3 mg/mL while after 3 months the concentration rose to 40 mg/mL; (3) the case of a woman who was found dead after an alleged soap and chlorxyphenol intranasal injection, whose femoral blood tested 14 days after death contained 1.0 mg/mL of cyanide, while after 23 days the concentration rose to 1.2 mg/mL.In the case of the first two blood samples, they were stored in a refrigerator temperature (4 � C), while in the third case, the storage conditions of the biological material were not specified.

Increase of cyanide concentration
Precisely 20 years later, Lokan et al. (1987) published a study that seemed to confirm the thesis raised by Curry et al. (1967).The authors also examined postmortem biological materials (blood, liver, and stomach content) from three deceased with the use of colorimetric method.In case (1), the blood sample (preserved with 0.1% sodium fluoride) and stomach content (collected without any preservation agent) were stored at 4 � C for 20 days before analysis; in case (2), the blood sample (collected without any preservation agent) and the liver were stored at −18 � C for 49 days; in case (3), the blood sample (preserved with 1% sodium fluoride), and the blood sample and stomach content (collected without any preservation agent) were stored at −18 � C for 41 days.Samples were not tested immediately after preservation, but after the aforementioned time period, cyanide concentrations were as follows, case (1) blood 22 mg/mL, liver 0.6 mg/g, stomach content 0.1 mg (no unit of weight or volume of a material available), case (2) blood 110 mg/mL, liver 6 mg/g; case (3) blood preserved with fluoride 0.0 mg/mL, blood without any preservation agent 150 mg/mL and stomach content 0.1 mg (no unit of weight or volume of a material available).Based on the results, the authors concluded that cyanide is formed in vitro because: (a) a toxic concentration of cyanide was detected in the blood in case 1 (0.1% sodium fluoride) while in the stomach content the concentration was lower than in other papers described so far; (b) in case 3, no cyanide was detected in the blood (1% sodium fluoride), while in the blood without any preservation agent the concentration was 150 mg/mL, moreover, in the stomach content the concentration was also low; (c) no person was suspected of possessing or being in contact with cyanide before death (this was claimed by third parties in testimonies) (Lokan et al. 1987).
Both papers have been cited over the years by other authors as a confirmation of the possibility of in vitro cyanide formation from 0.0 mg/mL to concentrations considered toxic or lethal.However, it is worth highlighting a few significant facts related to the conducted studies mentioned above.Firstly, in the study by Lokan et al. (1987), cyanide concentrations were not determined immediately after an autopsy, making it decidedly difficult to determine whether it is possible for cyanide concentrations to rise to the range of toxic levels.Secondly, both papers contain information about the undetectable cyanide odor during an autopsy as well as during toxicological analysis of biological material.It is noteworthy that the ability to smell cyanide is not equal among the entire population and may correlate strongly with genetic or environmental factors (Kirk and Stenhouse 1953;Khaja et al. 1993).Therefore, the absence of a cyanide odor should not provide any evidence of the lack of the ions' presence in biological material.Moreover, the concentrations determined in the liver in the two cases described by Lokan et al. (1987) (0.6 and 6 mg/g) are consistent with the range of liver concentrations reported to date during fatal intoxications, which range from 0.7 to 23 mg/g (Baselt 2017;Wachełko et al. 2021b).Furthermore, the low concentration of cyanide in the gastric content may be due to the strong acidic pH of these samples and therefore the degradation of this anion over time (20 days in case 1 and 41 days in case 3) may have occurred naturally as is the case with e.g.azide, which is undetectable at 4 � C already after about 20 days (Wachełko et al. 2021a).Both anions, CN -and N 3 -, form gaseous products, HCN and HN 3 , respectively, as a result of an interaction with acid, which can significantly complicate the analysis of these anions after extended time since death.
In the study by Curry et al. (1967), it is worth considering the methodology of the study.The method of detecting cyanide was based on a procedure from 1947 (Gettler and Goldbaum 1947), which involved estimating the intensity of the color of a previously prepared test paper (prussian blue paper test which was based on releasing HCN from the biological sample onto the filer paper soaked in in situ prepared iron hydroxide) and comparing it with reference papers at CN -concentrations in the range of 1-5 mg/mL.However, there are two factors that can bias the obtained result.The first is that the described technique also allows detection of the presence of nitrogen in organic compounds due to the formation of HCN.Endogenously, blood contains many organic compounds rich in nitrogen, while as a result of a putrefaction they are released in even greater amounts, which can potentially affect the result (as the reaction detects HCN non-selectively in the sample with no possibility of knowing its source).In addition, in an extensive research conducted by Gettler and Goldbaum (1947), it was shown that the coloration degree of the paper is not directly proportional to the cyanide concentration.Thus, it is not possible to accurately estimate whether the obtained result of the colorimetric test corresponds to the actual concentration TOXICOLOGY MECHANISMS AND METHODS of CN -in the material.Hence, there are two doubts regarding the results obtained by Curry et al. (1967): (1) it is not entirely clear whether the authors would have been able to distinguish between 1.0 mg/mL cyanide and 1.2 mg/mL cyanide (as was the case in case number 3, which they described); (2) reference (standard) papers were prepared in the concentration range of 1-5 mg/mL, so it is doubtful to provide an estimation of a concentration of 40 mg/mL (case 2) for which the analysts did not prepare a reference material for.In addition, in case 3 described by Curry et al. (1967) (as in the cases described by Lokan et al. (1987)), there is no information regarding testing the cyanide concentration immediately after the autopsy was performed and the samples were delivered to the toxicological laboratory, which impedes the possibility of estimating the cyanide increase over time under the specified temperature conditions.However, despite the doubts detailed above, both papers are still cited to this day as an evidence of the possibility of cyanide formation in vitro up to a range of toxic or lethal concentrations.This effectively complicates the interpretation of the results of analyses conducted to investigate the possibility of cyanide poisoning after a long period of time after a corpse has been found and/or samples have been secured for toxicological analysis.
After reviewing papers that used instrumental analytical methods, cyanide concentration can indeed increase over time, however this happens in two circumstances: when the cyanide concentration is endogenous (at an initial level of 0.03 mg/mL) and within about first 7 days of stability studies for toxic concentrations (at an initial level of 2.0 mg/mL (Egekeze and Oehme 1979) or 3.31 mg/mL (Chikasue et al. 1988)).Considering the low endogenous concentrations, the increase on a percentage scale seems to be a tremendous þ300% (Ballantyne and Changes 1976), þ233% or þ100% (Ballantyne 1977) however, considering the absolute value given in mg the increase is þ0.09, þ0.07 or þ0.03 mg, respectively, so that the final concentrations after both 6 weeks or 12 weeks still do not represent the concentrations detected during a fatal intoxication.In contrast, in other publications, where the initial cyanide concentrations were within the range of toxic concentrations (2-4 mg/mL) and increased over time, the studies were conducted over a period of 7-10 days, confirming the results of the 1930s study obtained by Gettler and Baine (1938).However, the authors also emphasized that although a slight increase in cyanide concentration was indeed observed over the course of a week, it began to gradually decrease after that time, normalizing (reaching equilibrium).In the case of the previously mentioned (Egekeze and Oehme 1979;Chikasue et al. 1988) studies, the experiment was discontinued after a period of 7-10 days, making it impossible to estimate what the further result and the final outcome regarding the stability of cyanide in biological material would be.On the other hand, in the most recent stability study conducted by Osak et al. (2021) using GC-MS/MS, the cyanide concentration decreased within 7 days.However, it is worth noting the fact that the aforementioned work was performed without the use of chromatographic separation of the tested substances,  Nd: information was not detailed in the article; a: the maximum permissible error in forensic toxicology is 20%, for this reason a higher value was chosen to indicate the scientific significance of the examination results; b: average value; c: value estimated from the graph, precise data was not specified; d: venous blood cyanide concentration was 14.5 mg/ml however this blood sample was not examined during stability study; e: cyanide concentration in venous blood was not provided; f: only provided information was that cyanide was stable for 2 days at -80, -20 and 4 � C; g: only provided information was that cyanide was stable for 2 days at -80 and -20 � C; h: percentage values were not given separately for low and high concentration levels.
resulting in low selectivity and, consequently, a higher probability of interfering factors affecting the determination of cyanide concentration.

Decrease of cyanide concentration
Among the studies published to date, the vast majority indicate a decrease in cyanide concentration over time.The very first stability study noted a 10% decrease in cyanide concentration in decaying organs over 28 days (Gettler and Baine 1938).When analyzing Table 1, a certain correlation is apparent between the degree of decrease in cyanide concentration in biological material related to the time of the study and the initial cyanide concentration in the sample.The largest decrease in concentration in proportion to the initial concentration is observed in a short period of time (60 min (Ballantyne et al. 1973)).However, when comparing the decrease in CN -concentration for samples containing an initial cyanide concentration of 15.9 mg/mL (test conducted for 24 h (Ballantyne et al. 1973)) and 5.7 mg/mL (test conducted for 3 weeks (Ballantyne et al. 1974)), it is apparent that a greater amount of cyanide was degraded in the sample with a higher initial concentration, even though the test was conducted for a much shorter period of time.In addition, blood samples subjected to an experiment conducted for the same period of time under the same temperature conditions but with different initial cyanide concentrations showed a similar pattern.In a study conducted by Ballantyne (1976) at 4 � C for 12 weeks, the decrease in cyanide was directly proportional to the initial concentration (in samples with CN -concentrations of 1.0 and 2.3 mg/mL cyanide decreased by 0.3 mg, whereas in a sample with an initial concentration of 3.7 mg/mL the concentration decreased by 0.8 mg, while the largest decrease was observed for a sample with an initial concentration of 6.5 mg/mL for which the cyanide loss value was 1.9 mg).The same correlation was observed for 20 � C (although with a proportionally larger loss), while at −20 � C the cyanide loss values were more similar to each other (in the range of 0.1 to 0.4 mg for low and high concentrations).A similar relationship was observed by Bright et al. (1990) who studied samples over a 6-month period.The corresponding decrease for 4 � C was 0.04, 1.9 and 3.75 mg for initial concentrations of 0.5 mg/mL, 2.5 mg/mL and 5.0 mg/mL, respectively.This trend was also observed for samples stored at −20 � C. The tendency for cyanide to decrease over time is supported by studies conducted with more selective methods than UV-Vis spectrophotometry, such as HPLC with fluorescence detection (Mateus et al. 2005), GC-MS (Bhandari et al. 2012) and HPLC-MS/MS (Bhandari et al. 2014a).However, in the case of the last two papers, detailed data on the final value of cyanide is missing which makes it impossible to compare the observed results with previous research works.The only information provided by the authors was that cyanide was stable for 2 days at 4 � C, −20 � C and −80 � C (Bhandari et al. 2012(Bhandari et al. , 2014a)).In contrast, a study by Desharnais et al. (2012) indicated a very slight increase in cyanide concentration over time for both low (0.5 mg/mL) and high (50 mg/mL) concentrations.However, the authors themselves emphasize that the obtained results were within the acceptable error of the method in forensic toxicology (±20%), and therefore it is impossible to determine whether the values achieved during the 15-day study have any statistical significance.
The stability study conducted in this article on an authentic blood sample was the longest experiment to date (730 days).However, due to the insufficient amount of biological material, a continuous study at systematic intervals was not possible to plot a curve indicating what changes the cyanide concentration undergoes over time.After one year, the cyanide concentration dropped slightly, whereas two years after the blood was received for toxicological testing, the cyanide concentration dropped by 0.3 mg, which corresponds to 15.8%.It is impossible to determine whether the result can be considered as a proof of a decrease in concentration over time, since the value obtained lies within the limit of acceptable error of the method of determination.The results were most definitely also influenced by the changes that the biological material underwent during storage (e.g.change in sample moisture content, pH, density).Taking into account all the gathered data and the meta-analysis performed, it seems impossible that cyanide, determined 5 years after death, could have grown to a range of toxic concentrations.To date, no instrumental analysis method has been published that would support such a possibility.Moreover, instrumental studies indicate that cyanide concentrations are either stable or slightly decrease (Ballantyne et al. 1973(Ballantyne et al. , 1974;;Ballantyne 1976;Bright et al. 1990;Mateus et al. 2005;Bhandari et al. 2012Bhandari et al. , 2014a)), so it seems most probable that the initial cyanide concentration (at the time of death) was higher.However due to the long period between the autopsy and our toxicological examination for cyanide content in the sample it is impossible to confirm this hypothesis with certainty or to estimate what the possible percentage of degradation of this anion in the biological material may have been.Despite these concerns, this casework should open up the topic for more extensive and standardized cyanide stability studies conducted over a longer period of time.In the meta-analysis presented here, the only factors that were comparable were the initial concentration, the time of the study, and the storage conditions.However, it is worth noting that factors such as the type of tube and preservative agent (e.g.sodium fluoride (McAllister et al. 2011)), volume of a biological material, type of a biological material (urine, serum, whole blood) or multiple freeze-thaw cycles (Osak et al. 2021) may affect the stability studies and obtained results.

Interferences in colorimetric tests for cyanides determination
One of the most commonly used approaches to date in the analysis of cyanide stability has been spectrophotometric methods, based on obtaining a colored chemical compound which final color intensity correlates with the concentration of cyanide in the analyzed sample.However, these methods, although simple, fast and inexpensive, carry a number of serious limitations that must be taken into account.The three most commonly used methods to date, along with their limitations and interfering factors, are described in the Table 2.Many of the methods mentioned are based on the K€ onig reaction (K€ onig 1904) and introduce various modifications regarding the reagents used, apparatus or analysis conditions.For example, Epstein (1947) proposed a method using chloramine T to form a cyanide derivative, which then, by reacting with a pyridine-pyrazolone mixture, leads to the formation of a colored compound which concentration can be examined with a spectrophotometer at 630 nm.Another spectrophotometric method used by Lundquist et al. (Lundquist et al. 1985) involves a reaction using NaClO to form a cyanide derivative, followed by a reaction with a barbituric acid-pyridine mixture and determination at 580 nm.It is also possible to combine both of the aforementioned methods -to carry out the K€ onig reaction and finally form a colored compound by reacting with a barbituric acid-pyridine mixture.
However, colorimetric reactions are not free from interferences.In the case of methods using the K€ onig reaction, an excess of chloramine T leads to a decrease in the intensity of the compound's color, and consequently a decrease in its absorbance (Epstein 1947;Chien et al. 1980), which translates into an underestimation of the cyanide concentration in the material.In the case of both methods, using either pyridinepyrazolone or barbituric acid-pyridine, the thiocyanates in the material can falsely increase the cyanide concentration, as they also undergo the described reactions (Epstein 1947;Chien et al. 1980), although Epstein (1947) suggests that the thiocyanate reaction is much slower and requires a catalyst in the form of iron (III) ions.Iron ions present in the samples analyzed (which may come, for example, from the degradation of hemoglobin (Panter 1994)) can also accelerate the reaction of cyanides with chloramine T, although not causing interference, but increasing the possibility of evaporation of a part of the analyte, which is possible if pyridine-pyrazolone is not added at the right time (Epstein 1947).The presence of proteins in the tested material, such as albumin, also affects the final cyanide concentration result, as they can react with the amino acids' sulfhydryl groups to produce forms of cyanide that are difficult to dissociate, thus lowering the final result (Baar 1966;Lundquist et al. 1985)).A similar problem was described by Baar (1966), where heavy metals present in the tested material formed complexes with cyanides that were difficult to dissociate.In the case of brabituric acid-pyridine reactions, bromine, sodium nitroprusside (which is still used today for the specific treatment of hypertension (Holme and Sharman 2022)) and other hexacyanoferrates can also be interfering agents (Lundquist et al. 1985;Tanaka et al. 1988).Moreover, in the Epstein method (Epstein 1947) and when using barbituric acidpyridine, pH affects the color of the final product (Epstein 1947) and its absorbance (Chien et al. 1980).Moreover, the absorbance value is affected by the concentration of the reagents (Sharma and Thibert 1985) and the buffer used (Chien et al. 1980), as well as the reaction time leading to the colored product (Chien et al. 1980).
Another commonly used spectrophotometric method is the prussian blue reaction, utilizing test paper previously soaked in FeSO 4 and NaOH solutions (Gettler and Goldbaum 1947).The reaction leads to the formation of a blue stain on the test paper, the intensity of which can then be later analyzed.However, as mentioned earlier, this is a very subjective method, which does not make it possible to distinguish cyanide concentrations that differ by a small amount from each other.In addition, a number of factors affect the accuracy of this method.The biggest bias is in the preparation of the test paper, as during soaking in a FeSO 4 solution and then air-drying it is impossible to predict what amount of iron (II) will oxidize to iron (III), which is the required substrate for the synthesis of prussian blue (Dixon et al. 1958).In the case of an iron (III) ions insufficiency, only a portion of the cyanide ions present in the material will be involved in the reaction to form prussian blue, which will significantly reduce the real concentration.Moreover, after drying, the test paper is soaked again, this time with a NaOH solution.When test paper is air-dried, the sodium hydroxide can absorb carbon dioxide from the air, leading to the formation of chemical compounds that can potentially interfere with the final result (Dixon et al. 1958).Sulfides (which can derive from protein degradation (Florence 1980)) and sulfates (IV) present in the sample mask the resulting blue color by forming precipitates with iron (Hyde 1975).

Biomarkers of cyanide exposure: an ideal tool for verifying the results of cyanide quantification in authentic caseworks?
Uncertainty of transformations and changes regarding cyanide concentration in biological material has led to potential cyanide metabolites analysis.A major metabolite is thiocyanate, to which about 80% of cyanide is transformed in the presence of a sulfide donor (Mitchell et al. 2013), although it is found not to be the best cyanide intoxication marker, because it is believed to be formed endogenously as well as delivered with food (Ru_ zycka et al. 2017).Another important metabolite (to which 15% of cyanides are transformed) is 2amino-2-thiazoline-4-carboxylic acid (ATCA) which is formed during reaction of cyanide with L-cysteine and exists in two tautomeric forms depending on the pH level (Bhattacharya et al. 2014).Numerous studies show that ATCA is stable in biological material (Logue et al. 2009(Logue et al. , 2014;;Giebułtowicz et al. 2016;Ru_ zycka et al. 2017) and for this reason, is often pointed out as a possible cyanide exposure marker, however Mitchell et al. (2013) conducted a study on pigs where ATCA was found to be eliminated even faster than cyanide ions (the elimination rate constants for ATCA and cyanides were 0.0499 and 0.0258, respectively).Correlation between cyanide concentration and ATCA concentration in biological specimens was a topic of studies regarding animals (Mitchell et al. 2013;Bhandari et al. 2014b) and humans (Vinnakota et al. 2012;Ru_ zycka et al. 2017).Animal studies results, although being able to broaden the knowledge of pharmacokinetics of ATCA, cannot be directly transferred to humans.Vinnakota et al. (Vinnakota et al. 2012) presented correlation between After 20 min, when the blue dye is created, the sample can be analyzed spectrophotometrically.
Formation of the dye: Excess of chloramine T results in bleaching of the dye.Iron salts hasten the reaction between cyanide ions and chloramine T.Although they do not interfere, faster termination of the reaction might increase the possibility to cyanides evaporation.
Thiocyanates will also react in this reaction giving blue dye, although the reaction is much slower.Sulphydryl groups react with cyanides leading to decreasing the final concentration.
In pH above 8, in vitro, cyanides might react with proteins forming thiocyanates.Heavy metals form complexes with cyanides which do not dissociate easily.
pH affects the color of the dye: pH below 7 results in red coloration and pH above 9 results in orange coloration.
Delay in addition of pyridinepyrazolone reagent can result in underestimation of the result because of possible cyanide evaporation.(Epstein 1947;Baar 1966) Pyridine-barbituric acid method H 2 SO 4 is added to the sample and gaseous HCN is extracted to NaOH by aeration.Then, acetic acid is added, 5 mM NaClO, and after 1 min, barbituric acid-pyridine mix is added.After 15 min, blue dye is formed and can be measured spectrophotometrically.
This reaction might also be performed as described in 'Epstein method' section above, with the use of K€ onig reaction, except for the last part, where barbituric acid-pyridine is added instead of pyridine-pyrazolone.  2017) performed an extensive study regarding post mortem material from fire victims.In almost all studied materials (blood, brain, lung, heart, liver, kidney, spleen) increase in cyanide concentration was related with increase in ATCA concentration.Nevertheless, authors point out that ATCA is not an appropriate marker for cyanide intoxication when taking into consideration sublethal levels of cyanide.Authors performing studies regarding potential ATCA usage as a cyanide intoxication marker all agree that further studies are required, especially regarding distribution, exertion, pharmacokinetics in humans and relation between cyanide and ATCA concentrations (Logue et al. 2009(Logue et al. , 2010;;Vinnakota et al. 2012;Bhandari et al. 2014b).In this study, although much time has passed from collection of the sample to the analysis in our laboratory, we did not perform ATCA determination, due to the limited data regarding cyanide and this metabolite concentrations correlations available, as well as due to the small samples volume that was collected, which we did not want to waste on unproven methods.

Conclusions
The growing trend of suicidal poisonings using inorganic compounds such as cyanides imposes the need to develop sensitive, selective and reliable methods for determining these anions in biological material, such as presented in this paper GC-QqQ-MS/MS method.Since there are often cases in which a long period of time passes between collection of a material for testing and toxicological analysis, it seems necessary to define changes in cyanide concentrations in post mortem biological material over time.This will significantly increase the reliability of the results and expert reports provided during the proceedings.

The past and the future of cyanide stability studies
As mentioned in this paper, over the years, different analytical techniques have been used to determine cyanide concentrations in biological material, as well as different and inconsistent results have been obtained (Figure 2).While spectrophotometric and colorimetric methods used in the 1930s-90s recorded both decrease and increase in cyanide concentrations over time, newer chromatographic methods that began to be used from the 2000s mainly indicate a decrease in cyanide concentrations over time.This trend also seems to be confirmed by the studies presented in this paper.
An analysis of the available literature shows that in order to properly interpretate the cyanide concentration results in biological material, further research and consideration of the following issues are necessary.Highly selective instrumental methods such as chromatographic ones coupled with mass spectrometry should be used for analysis.A number of factors affecting the final result such as added preservatives, pH, temperature, etc. should be taken into account during testing.In addition, future tests of cyanide stability in biological material, both short and long-term ones using different types of biological matrices, would add value to the present knowledge.Last but not least, it may be necessary to determine the transformation products of cyanides in biological material and to determine the correlation between the analyzed ions and degradation products.It also seems advisable to look for new markers of cyanide poisoning and products of its metabolism in the body.In the authors' opinion, only by considering these goals and challenges will it be possible to properly interpretate the cyanide concentration results in post mortem biological material.
bioethics committee (consent no.KB-184/2023).Biological fluids (blood and urine) collection from decedent was made by judicial authorities, and the samples were sent to our institute for toxicological analysis at their request.This article does not contain any studies with living human participants or animals performed by any of the authors.

Figure 1 .
Figure 1.Biological sample preparation procedure in details.

Table 1 .
Summary of cyanide stability study conducted with the use of instrumental analytical methods.

Table 2 .
The summary of interference factors in spectrophotometric methods used in quantification of cyanides in stability studies.
cyanide and ATCA concentrations in smokers and nonsmokers with results being that smoking increases both cyanide and ATCA concentrations in plasma, saliva, urine and erythrocytes.Ru_ zycka et al. ( Formation of the dye: Cyanide ions react with albumin present in blood resulting in non-detectable cyanide form.Sodium nitroprusside presence in blood results in falsely higher cyanide values.Bromine reacts with cyanide and thus, interfere resulting in cyanogen bromide.Hexacyanoferrate (II) and hexacyanoferrate (III) release cyanide in pH 8.1 resulting in higher cyanide values.Thiocyanates give a positive error in cyanide values determined.Excess of chloramine T results in lowering absorbance values.6 ] 3 Sulfides present in a sample result in a formation of FeS, which is visible on the test paper as a black spot.Saturation of the test paper with high concentration of sulfides reduces the retention of cyanide and thus, Prussian blue might not be visible.Sulfites present in a sample mask the Prussian blue intensity even to a greater extent than sulfides.