Determination of Thiocyanate in Plasma by Ion Chromatography and ...

Journal of Analytical Toxicology, Vol. 31, September 2007

Determination of Thiocyanate in Plasmaby Ion Chromatographyand Ultraviolet Detection Elodie Saussereau*, Jean-Pierre Goull~, and Christian Lacroix

Pharmacokinetic and Toxicology Laboratory, J. Monod Hospital, 1:-76083 Le Havre, France

I Abstract A specific and sensitive rapid high-pedormance liquid chromatographic method has been developed for determination of thiocyanate (SCN-) in plasma. This method is based on anion exchange chromatography after a simple ultrafiltration of plasma diluted in water. The detection of SCN- is carried out in ultraviolet (~. = 210 nm). The proposed method is linear in the range 1-30 mg/L. Intra-assayand interassay accuracy and precision were maintained within the designated limit (< 20%). The total recovery of SCN- varied between 97 and 103.9%. The method described should be useful for clinical medicine. Moreover, this method was applied to the analysis of SCN- plasma in deceased subjects, within the context of fire, and could be of interest in forensic science as a useful additional measurement tool for cyanide determination in blood.

Introduction

Thiocyanate (SCN-), normally present in low concentrations in body fluids, is the major metabolic product of cyanide (CN-). In fact, cyanide released gradually from its combination with ferric ion of the cytochrome oxidase, is metabolized to thiocyanate after reacting with a sulfur donor, such as thiosulfate. This reaction is catalyzed by the thiosulfate sulfurtransferase (rhodanese, EC 2.8.1.1), a mitochondrial hepatic and renal enzyme. Thiocyanate present in blood is partly bound to plasma albumin (1). The mean • SD thiocyanate concentration in plasma is estimated at 2.5 • 1.0 mg/L in healthy non-smoking subjects and not exposed to other cyanocontaining organic substances (1). When plasma concentration of thiocyanate exceeds 14.5 rag/L, SCN-is excreted in urine as the reabsorption in the tubules becomes saturated (2). The estimated plasma elimination half-life of SCN- is approximately 2.7 days in healthy subjects (3,4). The SCN- plasma concentration is considered to be a good indicator for cyano-containing organic substances exposure, as SCN- is inevitably * Author to whom correspondenceshould be addressed:Dr. ElodieSaussereau, Pharmacokineticand Toxicology Laboratory,J. Monod Hospital, F-76083 Le Havre cedex, France. [email protected].

high when blood CN- levels are increased. Determination of thiocyanate in plasma is therefore of great interest in monitoring cyanide exposure from tobacco smoke, the main source of cyanide in the environment (1,4-6), and from fire smoke (7,8), in associationwith cyanide measurement in blood, which has a very brief half-life (60 rain). However, SCN- measurements are of growing importance in other disorders (endemic goiter, tropical diabetes, ataxic neuropathy, infantile myxedema) associatedwith increases in plasma SCN- resulting from ingestion of cyanogenic glycosides contained in food, particularly some vegetables (1,4,6,9,10). Similarly, when sodium nitroprusside, a potent hypotensive agent used to treat a variety of cardiovascular conditions, is infused for days, or even weeks, SCN- concentrations in plasma increase to potentially toxic levels (3,11), and SCN- monitoring is advisable in order to verify the absence of accumulation of this metabolite. Different methods of measuring thiocyanate in body fluids (plasma, urine, and saliva) have been previously described (1,4,8-10,12-14). The chromatographic method described by Chinaka et al. (8) is sensitive and accurate. However,this ion chromatographic method for the simultaneous determination of cyanide and thiocyanate in blood, requires a step of derivatization with 2,3-naphtalenedialdehyde and taurine after extraction by adding water and methanol to blood, to produce a fluorescent product for CN- detection (15,16). The aim of this study was to developa simple, sensitive and accurate ion chromatographic (IC) method for the rapid determination of thiocyanate in plasma. This procedure employs an ultraviolet detection (~ = 210 nm). The proposed method was satisfactorily demonstrated for the analysis of SCN in plasma. This analytical procedure could be applied to plasma from deceased subjects potentially exposedto cyanideand permits rapid direct analysis after ultrafiltration of diluted plasma.

Materials and Methods Reagents

All reagents were prepared from analytical-grade chemicals (unless otherwise specified). Deionized, distilled water was used for all procedures. Potassium thiocyanate (reagent no-

Reproduction (photocopying) of editorial content of this journal is prohibited without publisher's permission.

383

Journal of Analytical Toxicology, Vol. 31, September 2007

123K0028) was obtained from Sigma-Aldrich (Steinheim, Germany). Potassium dihydrogen-phosphate (reagent no104871000) was obtained from Merck (Darmstadt, Germany). Orthophosphoric acid (reagent no-20261295) was obtained from Prolabo Chemicals (Fontenay-Sous-Bois, France). The stock standard solution of SCN- at I g/L was prepared with potassium thiocyanate in water; this solution is stable for at least 8 days at + 4~ Potassium thiocyanate working standards 1, 2.5, 5, 10, 20, and 30 mg/L were prepared daily by dilution of stock solution with serum Technicon Miles (reagent no05788372) obtained from Bayer (Tarrytown, NY). As a blank, we used serum Technicon Miles. Is apparatus and conditions The IC system (Hewlett-Packard) consisted of an HP-1050 pump, HP-1050 on-line degasser, HP-1050 autoinjector, and an HP-1040 M ultraviolet detector equipped with a diode array. Thiocyanate was detected at 210 nm. A Zorbax SAX C18 anionexchange column (250 x 4.6-mm i.d., 5-1Jm particle size), obtained from Agilent Technologies (Massy, France), was used as a separation column. Eluent was 10 mmol/L phosphate buffer; pH was adjusted to 3.5 with orthophosphoric acid. The column temperature was 50~ and the flow-rate was 1.5 mL/min. Sample injection volume was 20 I~L.

Procedure A 0.2-mL aliquot of blank, working standard, or plasma was

ii I

/

I~'

J II 11,~

cn,ln~

diluted with 0.2 mL of water in a 1.5-mL Eppendorf tube. After vortex mixing for 30 s, the mixture (0.4 mL) was transferred in a Microcon-Ultracel YM-10 filter obtained from Millipore (Bedford, MA). The mixture was centrifuged at 10,500 x g for 25 rain, and a 40-1JL aliquot of ultrafiltrate was injected into the IC system. Pretreatment of blood samples Blood samples from 10 deceased subjects, potentially exposed to cyanide during a fire, were obtained from the heart at autopsy and preserved in heparinized tubes. Plasma was separated by centrifugation at 2500 x g for 10 min and stored at -20~ until analysis. Thiocyanate in plasma is stable at least 6 months when stored at -20~ (1). Evaluation of validation parameters Limit of detection (LOD) is based on signal-to-noise ratio; LOD is defined as three times the standard deviation (SD) value added to the mean of 10 blank determinations. For the concentration defined as the lowest limit of quantification (LOQ), the percent deviation from the nominal concentration (measure of accuracy) and the relative standard deviation (measure of precision) both have to be less than 20%. Linearity was investigated within the 1-30 mg/L range at the calibration points of 1, 2.5, 5, 10, 20, and 30 mg/L. Intraday precision was calculated as the relative standard deviation (RSD) of the analysis of five parallel samples prepared under the same operative conditions within one day. It was performed at the six calibration points; RSD (%) = (SD/mean) z 100. Interday precision was determined after quantification analysis of five spiked samples of five different assays performed on five different days but under similar operating conditions, as the RSD of the analysis. It was performed at two concentrations, 7.3 and 12.3 mg/L. The accuracy was calculated as the percent of deviation between the nominal and the found concentration. It was tested on the six calibration points, in five parallel samples. Accuracy (%) = [(found - nominal)/nominal] x 100

12

The absolute recovery was calculated by comparing the peak areas obtained from standard working solutions ultrafiltrated with the peak areas obtained from standard solutions prepared in water. It was carried out at three concentrations, 2.5, 10, and 30 mg/L, in five parallel samples. The stability of ultrafiltrates at -20~ was tested in determining the SCN- concentration 9, 14, 22, 30, and 90 days after the first analysis. It was performed at two concentrations, 10 and 30 mg/L.

Results o

~

B

1o

,2

Figure 1. UItraviolet chromatograms of working standard solution at 10 mg/L (A), plasma of deceased subject in fire (case no. 4) (B), and analytical blank (C). 384

The ultraviolet chromatogram of a standard working solution 10 mg/L is shown in Figure 1A. The retention time of thiocyanate was 10.9 min. A typical chromatogram of SCN-plasma in a deceased subject, potentially exposed to cyanide during a fire, is presented in Figure 1B (case no. 4). Figure 1C shows the

Journal of Analytical Toxicology, Vol. 31, September 2007

ultraviolet chromatogram of analytical blank. Validation parameters of the described method are presented in Table I. The calibration graph is linear within the range 1-30 mg/L of thiocyanate. The r 2 value of the linear regression was higher than 0.999 (Figure 2). Concentrations greater than 30 mg/L were not tested. The limits of detection and quantification were 0.5 and 1.0 mg/L, respectively. Intra- and interday precision were lower than 20% for each concentration tested. Moreover, accuracy values were lower than 20% for the six concentrations analyzed. The absolute recoveries, carried out at 2.5, 10, and 30 mg/L, were 97.1,103.9, and 103.2%, respectively, with standard deviations within the range of 0.5-2.3%. The results of the stability study of ultrafiltrates at -20~ are shown in Table If. In Table III, the results of the determination of thiocyanate in plasma of 10 deceased subjects potentially exposed to cyanidewithin the context of fire are presented. Some samples were severely hemolyzed, but the effect of hemolysis on SCN- determination in plasma was not considered significant. Indeed, the correlation between theoretical and measured concentrations in a pool of heparin-hemolyzed plasma and surcharged in SCN- was excellent (r2 = 0.995). The CNblood concentrations from these deceased subjects are presented in Table III. Cyanide concentrations in blood were determined by IC with fluorimetric detection after derivatization with 2,3-naphtalenedialdehyde and taurine (16). No interference was observed when SCN- was determined at 210 nm in a pool of heparin plasma spiked with CN- high concentration (> 1000 pg/L).

curacy values were considered satisfactory. LOD and LOQ were broadly sufficient for the quantification of plasma SCN- in healthy subjects whose mean concentration is approximately estimated to 2.5 mg/L (1). This method should be useful in clinical medicine for monitoring patients during lengthy infusions with sodium nitroprusside, for example. Moreover, the proposed method could be of great interest in public health, especially in order to assess active and/or passive tobacco exposure. Lastly, the method described in this study could be applied for the determination of plasma SCN- in deceased subTable I. Validation Parameters of the IC Method for Plasma SCN- Determination

Level 1.00 (mg/L) Mean 0.98 (mg/L) S.D.* 0.11 (mg/L) R.S.D. 11.25 (%) Accuracy-2.00 (%)

Intraday Assay

Interday

(n = 5)

Assay (n = 5)

2.50

5.00 10.00

30.00

7,30 12.30

2.56

5.22 10.42 19.95 30.12

7.47 12.48

0.16

0.24

0.28

0.39

0.38

0,30 0.41

6.38

4.50

2.65

1.96

1.25

3.98 3.28

2.40

4.40

4.20

-0.25

0.40

4.32 1.46

20.00

* Abbreviations: S.D., standard deviation and R.S.D., relative standard deviation.

~

~llll Rlsql(l~ ' 74'i

Discussion

Thiocyanate, which is normally present in low concentrations in body fluids, is the major metabolic product of cyanide. Therefore, individuals chronically exposed to HCN have increased SCN- concentrations in their plasma. The determination of SCN- in plasma is subsequently of great interest for monitoring cyanide intoxication. Chinaka et al. (8) have previously reported an IC method with ultraviolet detection for the SCN- measurement in blood. These authors describe a Amount (nigh) method which affords the advantage of not requiring a comFigure2. Calibration curve: thiocyanate content in plasma in mg SCN-/L plex pre-analyticalphase. Moreover,in the method described by plasma versus the UV peak area (~. = 210 nm). Chinaka et al. (8), the dosage is performed after extraction and a derivatization step with 2,3-naphtalenedialdehyde and taurine in order to detect Table II. Results of Stability Study of Ultrafiltrates at -20~ CN- by fluorescence; cyanide and thiocyanate were determined simultaneously. In contrast, FoundConcentration*and PercentDeviationt with our method, the preanalytical phase was Day 9 Day 14 Day 22 Day 30 Day 90 reduced to a simple ultrafiltration of the diluted plasma. The results of recovery confirm Initial Level FC PD FC PD FC PD FC PD FC PD the validity of this sample pretreatment. As re(mg/L) (mg/L) (%) (mg/L) (%) (mg/t) (%) (mg/L) (%) (rag/L)(%) gards the performance of the described method, the regression line, which permits lO 10.2 2.0 9.7 -3.0 10.5 5.0 10,1 1.0 10.2 2.0 linking SCN-concentrations with the areas of 30 30.2 0.7 30.5 1.7 29.8 -0.7 29.0 -3.3 28.9 -3.7 analytical peaks, was perfectly linear within * Abbreviations: FC, found concentration and PD, percent deviation. the range of 1-30 rag/L, with an excellent cort Percent deviation = [(found value - initial value)/initial value)] x 100. relation coefficient.All of the detailed and ac385

Journal of Analytical Toxicology,Vol. 31, September2007

Table III. Results of Analyses of SCN- and CN- in 10 Plasma or Blood Samples from Deceased Subjects in Fires CaseNo.

Thiocyanate(mg/t)

1 2 3 4 5 6 7 8 9 10

5.83 1.03 1.48 1.98 9.20 1.02 8.47 7.18 13.78 7.35

Cyanide(pg/t) 565 763 382 510 79 246 1448 1092 1799 1789

phase, but simply an ultrafiltration of diluted plasma. The present method, easy to develop in the laboratory, should be of great interest particularly in monitoring cyanide exposure from tobacco smoke, fire smoke, ingested cyano-containing food, and cyanogenic glucosides. Moreover, the proposed method may also be useful for rapidly confirming a cyanhydric poisoning in a deceased subject, especially in a forensic context where death could be due to cyanide exposition. The stability of ultrafiltrates at -20~ allowing controlling first results at least three months after initial determination is very useful because of the low volume of forensic samples, especially in fire victims. However, for this last specific application, the validation of the method described here warrants further investigation.

Acknowledgment jects when cyanide intoxication is suspected, particularly within a context of fire. In fact, thiocyanate concentration is, in most cases, high when CN- blood levels are increased. As regards the stability of ultrafiltrates, the results obtained emphasize the fact that filtrated samples can be maintained at least three months at -20~ in order to control and confirm initial results. This control possibility is of major interest within the forensic framework. The presented method is simple, rapid, accurate, and sensitive. Furthermore, it seems to be particularly adapted to forensic evaluation when a death probably due to cyanide intoxication is suspected. In fact, an increase of plasma SCN- is favorable to a cyanhydric poisoning. Therefore, thiocyanate determination in plasma could be considered a very important additional measurement for CN- blood concentration because of the very short CN- half-life and its poor stability. In contrast, this chromatographic method can be applied to hemolyzed samples, which is frequently the case in forensic specimens. In fact, the purity of peaks is not altered by hemolysis. However, as demonstrated, the results obtained in 10 deceased fire victims, a correlation between cyanide and thiocyanate in blood from these subjects is not always obvious, and the interpretation of concentrations can be difficult (9). This is probably because most fire victims die in or soon after the fire, leaving little or no time for the cyanide to be metabolized in SCN- before death. Moreover, the concentrations of these compounds in the blood of fire victims are primarily affected by environmental factors as tobacco consumption. Finally, in rare cases where there is no increased concentration of thiocyanate in the plasma of a deceased subject but only a slightly increased CN- blood concentration, a bacterial cyanogenesis in human postmortem specimens should be suspected

(17).

Conclusions The analytical method described in this study permits the rapid and accurate determination of thiocyanate in plasma by IC and ultraviolet detection. This chromatographic method affords the advantage of not requiring a complex preanalytical 386

The authors thank Richard Medeiros, Rouen University Hospital Medical Editor, for the valuable editing of this manuscript.

References 1. P. Lundquist, B. Kagedal, and L. Nilsson. An improved method for determination of thiocyanate in plasma and urine. Eur. J. Clin. Chem. Clin. Biochem. 33:343-349 (1995). 2. C.F. Funderburk and L. Van Middlesworth. The effect of thiocyanate concentration on thiocyanate distribution and excretion. Proc. Soc. Exp. Biol. Med. 136:1249-1252 (1971). 3. V. Schultz. Clinical pharmacokinetics of nitroprusside, cyanide, thiosulphate and thiocyanate. Clin. Pharmacokinet. 9:239-251 (1984). 4. P. Degiampietro, E. Peheim, D. Drew, H. Graf, and J.P.Colombo. Determination of thiocyanate in plasma and saliva without deproteinisation and its validation as a smoking parameter. J. Clin. Chem. Clin. Biochem. 25:711-717 (1987). 5. M.J. Gardner, T.L. McCarthy, and W.J. Jusko. Relationship of serum thiocyanate concentrations to smoking characteristics. J. Toxicol. Environ. Health 14:393-406 (1984). 6. C. Charlier, T. Gougnard, D. Lamiable, P. Levillain, and G. Plomteux. Cyanides and thiocyanates in a hospital toxicology laboratory. Ann. Toxicol. Anal. 12" 131-136 (2000). 7. F.J. Baud, P. 13arriot, V. Toffis, 13. Riou, E. Vicaut, Y. Lecarpentier, R. Bourdon, A. Astier, and C. Bismuth. Elevated blood cyanide concentrations in victims of smoke inhalation. N. Engl. J. Med. 325:1761-1766 (1991 ). 8. S. Chinaka, N. Takayama, Y. Michigami, and K. Ueda. Simultaneous determination of cyanide and thiocyanate in blood by ion chromatography with fluorescence and ultraviolet detection. J. Chromatogr. B 713:353-359 (1998). 9. M.R. Haque and J.H. 13radbury. Simple method for determination of thiocyanate in urine. Clin. Chem. 45:1459-1464 (1999). 10. C.J. Vesey and C.J. Kirk. Two automated methods for measuring plasma thiocyanate compared. Clin. Chem. 31:270-274 (1985). 11. C.J. Vesey and P.V. Cole. Blood cyanide and thiocyanate concentrations produced by long-term therapy with sodium nitroprusside. Br. J. Anaesthesiol. 57:148-155 (1985). 12. S. Tanabe, M. Kitahara, M. Nawata, and K. Kanawabe. Determination of oxidizable inorganic anions by high-performance liquid chromatography with fluorescence detection and application to the determination of salivary nitrite and thiocyanate and serum thiocyanate. J. Chromatogr. 424:29-38 (1988). 13. S.H. Chen, Z.Y. Yang, H.L. Wu, H.S. Kou, and S.J. Lin. Determi-

Journal of AnalyticalToxicology,Vol. 31, September2007 nation of thiocyanate anion by high-performance liquid chromatography with fluorimetric detection. J. Anal. Toxicol. 20' 38-42 (1996). 14. C.J. Vesey,H. McAIlister, and R.M. Langford. A safer method for the measurement of plasma thiocyanate. J. Anal. Toxicol. 23: 134-136 (1999). 15. A. Sano, S. Takezawa, and S. Takitani. High performance liquid chromatography determination of cyanide in urine by pre-column fluorescence derivatization. Biomed. Chromatogr. 3:209-212 (1989). 16. A. Sano, N. Takimoto, and S. Takitani. High performance liquid

chromatographic determination of cyanide in human red blood cells by pre-column fluorescence derivatization. J. Chromatogr. 582" 131-135 (1992). 17. R.J.Lokan, R.A. James, and R.B. Dymock. Apparent post-mortem production of high levels of cyanide in blood. J. Forensic Sci. Soc. 27" 253-259 (1987).

Manuscript received March 27, 2007; revision received April 23, 2007.

387