Journal of Analytical Toxicology, Vol.
30, April 2006
[ TechnicalNote
A Novel LC-ESI-MS-MS Method for Sensitive
Quantification of Colchicine in Human Plasma:
Application to Two Case Reports
Emuri Abe1, Anne-Sophie Lemaire-Hurtel 2, Charlotte DuverneuiU, Isabelle Etting1, Emmanuelle Guillot 1,
Philippe de Mazancourt 1, and Jean-Claude Alvarez TM
ILaboratoirede Pharmacologie--Toxicologie,Centre Hospitalier UniversitaireRaymondPoincar6,AP-HP,
104 BoulevardR. Poincar6, 92380 Garches, Franceand 2Servicede PharmacologieClinique, CentreHospitalier
Universitaire Sud, 80000 Amiens, France
l Abstract ]
A novel method based upon liquid chromatography coupled to
ion trap mass spectrometry (MS) detection with electrospray
ionization interface has been developed for the identification and
quantification of colchicine in plasma or whole blood. Colchicine
was isolated from plasma using a liquid-liquid extraction with
dichloromethane at pH 8.0 and embutramide as an internal
standard, with satisfactory extraction recoveries. Solutes were
separated on a 3-pro C18 Uptisphere (Interchim) column (150 x
2.0-ram i.d.) using acetonitrile/2mM NH4COOH pH 3.8 buffer
(50:50, v/v) as the mobile phase with a flow-rate of 200 pL/min.
Data were collected either in full-scan MS mode at m/z100-450
or in full-scan MS-MS mode, selecting the ion m/z400.1 for
colchicine and m/z294.1 for embutramide. The most intense
daughter ion of colchicine (m/z 358.1) and embutramide
(m/z207.9) were used for quantification. Retention times were
2.40 and 4.25 rain for colchicine and embutramide, respectively.
Calibration curves were linear in the 0.50-50 ng/mL range. The
limits of detection and quantification were 0.05 ng/mL and
0.50 ng/mL, respectively. The intra- and interassay precisions were
< 14%, and the intra- and interassay accuracies were in the
97-105.8% range at either 2 or 20 ng/mL. A fatal case of
colchicine self-poisoning with a lethal blood concentration
of 60 ng/mL and nonfatal case with a plasma sample collected
very late (at least 36 h after the ingestion) are presented. The
described method enables the unambiguous identification and
quantification of colchicine with a very good sensitivity, using
only 1 mL of sample.
Introduction
Colchicine is a lipid-soluble alkaloid extracted from the seeds
and the corm of Colchicum autumnale. It is the medicine of
choice as a prophylactic agent against acute gouty arthritis (1)
* Author to whom correspondenceshould be addressed.
E-mail: jean-claude.alvarez@rpc.ap-hop-pairs.fr.
210 Reproduction(photocopying)of editorialcontentof thisjournal is prohibitedwithout publisher'spermission.
Downloaded from https://academic.oup.com/jat/article-abstract/30/3/210/719491 by guest on 07 January 2020
and in the treatment of acute attack (2). It is an antimitotic
drug, and its activity is mainly related to its ability to inhibit mi-
crotubule polymerization. Intoxications are rare but always se-
vere (3), generally associated with suicides (4,5), therapeutic
misadventures (6,7), ingestion of Colchicurn autumnale (8-10),
or homicides (11). The signs and symptoms of colchicine in-
toxication occur in three phases (12): (i) early complications,
within 24 h, with gastrointestinal symptoms (nausea, vom-
iting, diarrhea, abdominal pain, and anorexia), hypovolema,
electrolyte balance disorders, and peripheral leucocytosis; (ii)
followed the next days by bone marrow hypoplasia (profound
leucopenia and thrombocytopenia), cardiac arrhythmias and
cardiovascular collapse, respiratory distress, hypoxia, pul-
monary edema, oliguric renal failure, rhabdomyolysis, elec-
trolyte derangements, metabolic acidosis, mental state changes;
and (iii) one week after the intoxication occurred a rebond leu-
cocytosis and a transient alopecia. Intoxication can lead to
death (13), and ingested doses more than 0.8 mg/kg are con-
sidered to be invariably fatal (14).
Many analytical methods using colorimetry (15), fluorimetry
(16), gas chromatography-mass spectormetry (GC-MS) (17),
liquid chromatography (LC) coupled with either single-wave-
length UV (18-20) or diode-array detection (DAD) (4,5,21,22),
and LC coupled with ion spray MS (7,14) or sonicspray ioniza-
tion MS (10) have been previously described for quantification
in plasma, serum, whole blood, or tissues. No LC method cou-
pled with tandem MS for human specimens has been pub-
lished. However, sensitivity may be a major criterion for a
method devoted to the analysis of colchicine because this com-
pound is present in blood at very low concentration (23). In this
work, a new analytical method using only 1 mL of sample is de-
scribed, based on liquid-liquid extraction and LC-electrospray
ionization (ESI)-MS-MS for the identification and quantifica-
tion of colchicine. A case of fatal colchicine intoxication and a
non-fatal case with plasma sample collected at least 36 h after
ingestion are reported here to demonstrate the suitability and
the sensitivity of the method.
Journal of Analytical Toxicology, Vo[. 301 April 2006

Experimental Preparation of stock solutions

Primary stock solution of colchicine was prepared at a con-
centration of 1.0 g/L in methanol. Working solutions (1.0 mg/L
Reagents and 0.1 mg/L) were prepared by appropriate dilutions of stock
Colchicine (free base, Mw = 399.4) and the internal standard
solution in methanol. The stock and working solutions of em-
(IS) embutramide (freebase, Mw= 293.t) were purchased from
butramide (IS) were prepared in methanol at 1.0 and 0.1 rag/L,
Aventis (Paris, France) and Akzo Nobel (Munchen, Germany),
respectively. Because of the light-sensitivity of colchicine, all
respectively.Concentrated formic acid (HCOOH),ammonium
stock and working solutions were stored at -20~ in the dark
formate (NI-I4COOH),and potassium dihydrogeno-phosphate
for a maximum of three months and one month, respectively.
(KH2PO4)were purchased from Sigma-Aldrich(Paris, France).
Sodium hydroxidewas obtained from Prolabo (Paris, France).
HPLC-gradeacetonitrile, methanol, and dichloromethane were Preparation of calibration curves and quality
obtained from Riedel De Hai~n(Paris, France), Prolabo (Paris, control samples
France), and E. Merck (Darmstadt, Germany), respectively. Calibration curves were prepared by spiking drug-free plasma

Downloaded from https://academic.oup.com/jat/article-abstract/30/3/210/719491 by guest on 07 January 2020

The phosphate buffer (pH 8.4) was prepared in purified water (1 mL) with appropriate volumes of the previously mentioned
(AFS 200 water purification system, Millipore, Molsheim, colchicine working solutions to produce the calibration curve
France) using a 2M KH2PO4solution adjusted to the desired pH points equivalent to 0.5, 1, 5, 10, 25, and 50 ng/mL of
by appropriate addition of NaOH. colchicine. Quality control (QC) samples were prepared at a
Mobile phase buffer was prepared in purified water with a therapeutic level (2 ng/mL) and toxic level (20 ng/mL) by
2raM NH4COOHsolution, adjusted at pH 3.8 with HCOOH. spiking two 20-mL plasma samples in bulk with appropriate
The buffer was filtered through a 0.45-pm filter prior to use. volumes of colchicine working solution prepared from separate
Drug-free human plasma was supplied by the local blood bank weighting. Theywere then aliquoted and frozen. Six aliquots of
at the Etab]issementFran(;ais du Sang (Le Chesnay,France). each level were thawed on each day of analysis.

Sample preparation
400,1 25#271 RT: 3.08
100- AV: 1 NL: 2.46E6 T: + QC, calibration curve, blank plasmasamples,
95i ~= C ESI Full ms [
90- 100.00-450.00] and unknown samples were extracted using a
88i liquid-liquid extraction technique. One hun-
8o~ dred microliters of the internal standard
75i
7oi working solution (embutramide 0.1 mg/L), 250
89i pL of phosphate buffer (pH 8.4), and 2 mL of
551 the extracting solvent (dichloromethane) were
8oi addedto eachtube containing I mL of plasma.
45i The sampleswere then shaken for 15 min and
4oi
39i centrifuged at 3500 x g for 10 min. The or-
30- ganic layer was decanted into another tube
25i
20i 195.0
401,2
where it was evaporated to complete dryness
15- 181 1 331.9 i 410,4 under a nitrogen stream. Sampleswere recon-
I 4165
1oi 172.0 " 221.7 23,~.0 z61.e 20,8.0295.0 ~, 1.3859"9 3900 422,0
stituted with 60 pL of mobile phasebuffer (pH
0 358,1
r,h tI,, ITll~J..h
25#272 RT: 3,09
3.8)/acetonitrile (50:50,v/v), then vortex mixed
23"~
227
AV: 1 NL: 5,73E5 F: §
c ES] SRM ms2 for 30 s, and transferred to a microvial (250 pL,
400.20@36.00 [
110.00-405.00] 30 x 5 mm, ThermoFinnigan). Ten microliters
28~ B
were injected into the liquid chromatographic
181 332.0
system.
15-d
14t 326.1341.1 Equipment and chromatographic conditions
12-j
Chromatography was performed on
Thermo-Finnigan Surveyor high-performance
lO-t
388.1 liquid chromatography (HPLC) system
309.0
(Thermo Electron Corporation, Les Ulis,
France) with an autosampler injector, using a
3-pro C18 Uptisphere (Interchim) column
285.2
279,9
[ (150 x 2.1-mm i.d.) maintained at 30~ The
216,2 250,0 39
39~.9 devicewas completedwith a precolumn (C18,
15o.1 I I I , ,
100 150 200 250 300 350
I
400
= i I = I
450
3-pro, 4 x 2.0-mm i.d., Interchim). The elution
mJz
was achievedisocratically with a mobile phase
Figure 1. Full-scan (m/z 100-450) positive-ion mass spectrum of colchicine (A), and MS-MS spec- of 2raM NH4COOHpH 3.8 bufferlacetonitrile
trum of colchicine with its most intense product (m/z 358.1) used for quantification (B).
(50:50, v/v) at a flow rate of 200 pLlmin.

211
 Journal of Analytical Toxicology, Vol. 30, April 2006

MS and m/z 207.9 (quantification ion) for IS, monitored with a

The compounds were detected by a Thermo-Finnigan LCQ mass resolution of 1.0 amu.
Advantage ion trap MS (Thermo) equipped with an ESI source.
The Thermo-Finnigan Excalibur (v 1.3) software was used for Validation of the method
system control, data acquisition, and compound quantifica- The absolute recoveries were evaluated at two levels (5 and 25
tion. Nitrogen (Nitrox UHPLCMS 18, nitrogen generator, Dom- ng/mL) by comparing the peak areas of the extracted samples
nick Hunter, Villefranche sur Sa6ne, France) was employed as (n = 6) with those of extracted blank plasma samples spiked af-
sheath and auxiliary gas at a pressure of 40 and 10 arbitrary terwards with the same amount of compounds (n = 3).
units, respectively. The ESI source was set in the positive ion- In order to evaluate ionization suppression, the peak areas of
ization mode, and an ion-spray voltage of + 5.0 kVwas applied. extracted blank plasma samples (n = 3) and blank postmortem
The capillary temperature was set to 250~ under a voltage of whole blood (n = 3) spiked afterwards with colchicine and IS
+ 4 V. The system was tuned using a continuous 5 lJL/min in- (10 ng/mL each) were compared with those obtained by direct
fusion of a colchicine solution (10 rag/L) in mobile phase buffer injection of the same amount of compounds (n = 3). The ma-

Downloaded from https://academic.oup.com/jat/article-abstract/30/3/210/719491 by guest on 07 January 2020

(pH 3.8)/acetonitrile (50:50, v/v). The acquisition was performed trix effects were also tested by comparing the peak areas of ex-
either in full-scan mode atm/z 100-450 or in full-scan MS-MS tracted plasma samples (n = 6) containing 10 ng/mL of
mode. The protonated precursor molecular ions [MH]+ of colchicine and IS with those of extracted postmortem whole
colchicine (m/z 400.1) and IS (m/z 294.1) were trapped with a blood samples (n = 6) containing the same amount of com-
mass resolution of 1.0 amu and fragmented with an activation pounds.
time of 30 ms and a collision energy of 36% and 30%, respec- For the linearity study, six calibration curves were obtained
tively. The most intense daughter ions resulting from these in three days. Linearity was tested for the 0.5-50 ng/mL range
fragmentations were m/z 358.1 (used as quantification ion), m/z of concentrations, employing standard calibration curves of at
382.0, and m/z 341.1 (used as confirmation ions) for colchicine least six points. In addition, a blank and zero plasma samples
were also analyzed to confirm the absence of
RT: 0.00-6.01 SM: 7G
2.41 NL: 3.07E6
interferences. These two samples were not
100
400.1 A In/z= 399.7-400.7 F:
+ c ESI Full ms [
used to construct the calibration function.
100,00-450.00] MS
90
sg au2 Quantification was performed by calculating
60 the ratio between the peak area of colchicine
70
and the peak area of embutramide. The preci-
60
sion and accuracy of the method were carried
50
40
out over three days. Each day, one calibration
30
curve and six determinations of each QC level
20 were analyzed. The values obtained were ana-
1o
0 L lyzed using the analysis of variance, which
2,38 NL: 6,36E5 separated the intraday and interday assay stan-
358.1 m/z= 357.6-358.6 F:
1 ii B + c ESI SRM ms2 dard deviations and, consequently, the corre-
400,20@36.00 [
110.00405.001 MS sponding coefficients of variation (CV). The
s9 au2
intraday assay CV took into account the vari-
ability of the six replicates each day for three
days and the interday assay CV the variability
of the days of analysis. The accuracy was de-
termined by comparing the mean calculated
concentration with the spiked target concen-
tration of the QC samples. The limit of de-
4,25 NL: 9.76E5 tection (LOD) was defined as the lowest
207.9 m/z= 207.5-208.5 F:
C + c ESI SRM ms2 concentration of the analyte that can be de-
90
1001 it 994,10@30.00 [
100.00-300.00I MS
sg au2
tected with a signal-to-noise ratio greater than
80
70
3:1. The limit of quantification (LOQ) was de-
6O
fined as the lowest concentration of the com-
pound that can be measured with both an
4O accuracy of _+10% of the true value and a CV
3O <2O%.

~1
'1
20- i
1o j
Otl~-l.l(lllllT,;-i,lli-i~,,,il. i i - i l i l , i~ I ..~'] , , , li~-;i7-I , , , b } Pi t ' l i
Applications of the method
0.0 0.5 1.0 1,5 2.0 2.5 3.0 3,5 4.0 4.5 5.0 5.5 6.0 In the first case, a 40-year-old man was
Time (rain)
found dead at home. A box of Colchicine
Figure 2. Chromatogram of colchicine parent ion (m/z 400.1) (A), product ion (m/z 358.1) (B),
and of IS (embutramide) (m/z 207.9) (C) extracted from 1 mL of plasma half-diluted of our case
Houd~ Tablets| was present near the body. No
report with total concentration of 60 ng/mL of colchicine. abnormal external findings were noted during
the postmortem examination of the body. Au-

212
Journal of Analytical Toxicology, Vol. 30, April 2006

topsy was not performed, and sample of peripheric blood (7
mL) was collected and sent to the laboratory for toxicology.
GC-MS and HPLC-DAD screening were carried out. Psy-
chotropic drugs such as benzodiazepines, barbiturates, antide-
pressant, or neuroleptic drugs were screened with specific
methods using GC-MS, LC-MS-MS, and HPLC-DAD.The qual-
itative and quantitative analysis of drugs of abuse (opiates,
cocaine, cannabis, amphetamines, methadone, and buprenor-
phine) were carried out using GC-MS. Alcohol was measured
with GC coupled with a flame-ionization detector.
In the second case, a 70-year-old woman was hospitalized for
vomiting and profuse diarrhea. She had been treated with
colchicine (2 mg/day for 10 days). Four hours after admission,
Method validation
Results of the absolute recoveries determined for colchicine
ranged from 82% at 5 ng/mL to 93% at 25 ng/mL. Absolute an-
alytical recovery of internal standard was 78%. As shown in
Table I, no ionization suppression was observed either in the
plasma sample or in postmortem whole blood sample. No dif-
ference was observed for colchicine and IS peak areas or the area
ratio between the extracted blank plasma and extracted blank
postmortemwhole blood, both spiked afterwards with 10 ng/mL
of both compounds, and those obtained by direct injection of the
same amount of compounds. On the other hand, a matrix effect
on extraction recovery was observed in a postmortem whole
blood sample compared to the plasma sample. Extraction re-

Downloaded from https://academic.oup.com/jat/article-abstract/30/3/210/719491 by guest on 07 January 2020

she presented a metabolic acidosis and renal failure and ad- covery of colchicine was 24.8% lower in the postmortem whole
mitted to intensive care for heart failure. Three days later, a leu- blood sample than in the plasma sample, but the decreased ex-
copenia occurred. Onlya small blood sample (2.5 mL), collected traction recovery of the IS was similar (28%), with an identical
36 h after the ingestion of colchicine, was available for diagnosis area ratio (104% in whole blood sample compared to plasma).
of colchicine intoxication. Initial testing of colchicine at the The method exhibited a reliable linear response for the range
local hospital was performed by HPLC-DAD. of concentrations from 0.5 to 50 ng/mL. Results of six calibra-
tion curves are given in Table II, with the back-calculated con-
centrations, the accuracy, and precision of each calibration
point. The LOQ of the method is 0.5 ng/mL, and the LOD was
Results at 0.05 ng/mL. Accuracy and precision of the assay was mea-
sured by analyzing 36 QC samples. Data are presented in Table
Separation and specificity III. Intraday accuracies of the method for colchicine were 97.0%
Figure 1A presents the full-scan (m/z 100-450) positive-ion and 105.5%, and the intraday precisions were 7.6% and 6.3% at
mass spectrum of colchicine recorded from the continuous in- the concentrations of 2 and 20 ng/mL, respectively.Interday ac-
fusion of a 10.0 mg/L solution of the compound in mobile phase curacies were 102.9% and 105.8%, and the intraday precisions
buffer (pH 3.8)/acetonitrile (50:50, v/v). The spectrum shows a were 11.4% and 14.0% at the concentrations of 2 and 20 ng/mL,
major peak at m/z 400.1, corresponding to the protonated respectively.
colchicine [MH]+. MS-MS spectrum of colchicine is shown in
Figure lB. The most intense product ion observed is at m/z Clinical application of the method
358.1, which was used for quantification of colchicine. The In the first case, the GC-MS and HPLC-DAD screening and
ions at m/z 382.0 and m/z 341.1 were used as confirmation
ions. Table II. Linearity of the Method*
The colchicine parent ion (A),and the chromatograms of the
product ions of colchicine (B), and IS (C) extracted from 1 mL Theoretical Concentration (ng/mL)
of half-diluted plasma from the fatal case (with total colchicine
0.5 1 5 10 25 50
concentration of 60 ng/mL) are shown in Figure 2.
Linearity (n = 6)
Back calculated 0.52 1.05 5.0 9.2 25.7 49.8
Table I. Peak Area and Peak-Area Ratio of Colchicine concentration (mean)
(10 ng/mL) and IS (10 ng/mL) in Plasma, Whole CV (%) 4.1 10.2 7.4 8.6 1.2 1.0
Postmortem Blood, and Direct Injection (arbitrary units) Accuracy(%) 104.4 105.2 100 92.1 102.8 99.6

PeakArea of PeakArea * Six calibration curveswere obtained in threedays.Other factors: slope (mean
Colchicine of IS and CV) = 0.02 and 14.0; intercept(meanand CV) = 9.2E- 04 and 10.6;
0.997 < r< 0.9999; LOD = 0.05 ng/mL; and LOQ = 0.5 ng/mL.
10 ng/mL 10 ng/mt Area Ratio

Extracted blank 1.54 + 13% 13.3 + 12% 0.119 + 12%

plasma spiked
Table III. Intra- and Interassay Precision and Accuracy*
afterwards (n = 3)
Colcbicine Intra-assay Intra-assay Interassay Interassay
Extracted blank 1.48 + 10% 14.9 _+4% 0.100 + 14% (ng/mL) CV (%) Accuracy(%) CV (%) Accuracy(%)
whole post mortem
blood spiked 2 (n = 18) 7.6 97.0 11.4 102.9
afterwards (n = 3) 20 (n = 18) 6.3 105.5 14.0 105.8

Direct injection 1.62 + 6% 15.6 + 2% 0.103 +_5% * Eachday for three days,one calibration curve with six determinationsof the
(n = 3) two CQ levelswere measured.

213
 Journal of Analytical Toxicology, Vol. 30, April 2006

the specific screening for research of psychotropic drugs and
drugs of abuse were negatives. No alcohol was found in the
blood sample. The present method was applied to blood to high-
light the presence of colchicine. The drug was detected with a
blood concentration of 60 ng/mL.
In the second case, colchicine was not detected using
HPLC-DAD.A 1-mL plasma sample collected at least 36 h after
the ingestion of colchicinewas then analyzedwith LC-MS-MS.
The latest method highlighted the presence of colchicine at a
concentration of 4.5 ng/mL in the plasma sample.
Discussion
LOQ found in the literature. The low LOD (0.05 ng/mL) and low
LOQ make our method particularly convenient for colchicine
screeningand quantification in human plasma in clinical as well
as toxicologicalsituation. Thisvery good sensitivity is a conse-
quence of the MS-MS detection. In the LC-MS methods devel-
oped by Tracqui et al. (14) and Jones et al. (7), the authors
found an LOD at 7.0 n~mL with 4.0-mLsample and an LOD at
0.5 ng/mL with 1.0-mL sample, respectively,with similar ex-
traction procedure. In a more recent HPLC-DADmethod (21),
the LOD and LOQ for colchicine in whole blood were 2 and 4
ng/mL, respectively, using a 5.0-mL blood sample. As an ex-
ample of the very good sensitivity of the method, Figure 3
shows parent ion (A, m/z 400.1) and the product ion (B, m/z

Downloaded from https://academic.oup.com/jat/article-abstract/30/3/210/719491 by guest on 07 January 2020

358.1) chromatograms of colchicine extracted from spiked
drug-free plasma with 0.1 ng/mL. Colchicine is not detected in
Sensitivity may be a major criterion for a method devoted to the MS method (parent ion), but the product ion of colchicine
the analysisof colchicine because this compound is therapeutic could be detected at the good retention time with good sensi-
and toxic at particularly low concentrations in blood. Thera- tivity. In our second case, only 2.5 mL was availablefor toxico-
peutic plasma concentrations have been reported to be in the logical analysis. HPLC-DAD did not allow highlighting the
range of 0.3-2.4 ng/mL at steady-state with 1-mg daily dosing compound using 1 mL of plasma sample. Our method showed
(23). The LOQ of our method is 0.5 ng/mL, which is the lowest the presence of colchicine at a toxic concentration despite the
small sample collected at least 36 h after the
last ingestion of colchicine, confirming that it
RT: 0.00-6.O0 SM: 15G
2.09 NL: 3,73E4 was a colchicine intoxication. In the first case,
400.6 m/z= 399.7-406.7
100 A F: + c ESI Full ms [ the drug was detected with a whole blood con-
95- 100,00450.00] MS
9O 2,85
O,lext centration of 60 ng/mL, which is a lethal con-
399,9
85 centration according to the literature on fatal
BO
colchicine poisoning (7,21). The colchicine
finding provided a cause of death. Police in-
65 031
400.3 1.27 vestigation showed that it was a suicide.
60 400.;
400.2 1 78
A single-step liquid-liquid extraction with

,/
400.3
50
45 gg 4o1
dichloromethane was employed for reasons of
4 speed and convenience.As usual, chlorinated
solvents have widely proven to be the media of
30!
25 ' 4.82 400.1 choice for colchicine extraction (15,17,22). Re-
20 / oo a suits of the absolute recoveriesdetermined for
1
colchicine ranged from 82% at 5 ng/mL to
93% at 25 ng/mL. These results were identical
2.44
358.1
NL: 3.86E3
m/z= 357.5-358 5
to those obtained by Tracqui et al. (14) with a
io%
95~ B F: § c ESI SRM
ms2 400.20@36.00
similar extraction procedure. These absolute
[ 110,00-405 001
9o-] MS 0.1ext recoveries were evaluated by comparing the
85-~
8o=3
peak areas of the extracted samples with those
75-] of extracted blank plasma samples spiked af-
7o~
65~
terwards with the same amount of compounds,
8o~ in order to detach extraction efficiencyfrom
55-3 ionization suppression. However, no ioniza-
8o~
45-] tion suppression was observed in our method
4o-~ either in plasma or in postmortem whole
35~
30~ blood. The peak areas of colchicine and IS (or
25~ area ratio) were similar between extracted
28~ I
blank plasma spiked afterwardswith 10 ng/mL
15-~ l of both compounds, extracted blank post-
:1 0.O 0.5 1..0
f
1.5 2.0 2.5 3,0
Time (rain)
3.5 4.0 4.5 5.0 5.5 6.0
mortem whole blood spiked afterwards with
the same amount of compounds, and direct
injection of the same amount of compounds.
Figure 3. Chromatograms of colchicine parent ion (m/z 400.1) (A) and the product ion On the other hand, a matrix effectwas observed
(m/z358.1) (B) extracted from spiked drug-free plasma with 0.1 ng/mL. Colchicine is not detected
on extraction efficiencyin a postmortem whole
in MS method, but the product ion of colchicine is detected with a good sensitivity.
blood sample when compared to a plasma

214
Journal of Analytical Toxicology, Vol. 30, April 2006
sample. Extraction recovery of colchicine was 24.8% lower in a
postmortem whole blood sample than in a plasma sample. How-
ever, the similar decreased extraction recovery of the IS, with an
identical area ratio obtained between plasma and postmortem
whole blood, allowed the use of our method for the measure-
ment of colchicine concentration in postmortem whole blood.
Embutramide is a derived compound from gamma-hydroxy-
butyrate that possesses a strong narcotic effectand induces a deep
anesthesia by paralyzing the brain centers that control breathing
(24). It is found in Tanax~, a euthanasia solution commonly used
in veterinary medicine (25). Because a deuterated standard of
colchicine is not available on the market, and even if embu-
tramide is not a structural analogue of colchicine, we used this
compound as an IS because it is unusual to find it in patients (26),
it is well co-extracted, and it elutes with good resolution towards
colchicine. The retention times of colchicine and the IS were 2.40
and 4.25 rain, respectively, ensuring both a short analysis and ad-
equate resolution between colchicine and IS. The full-scan pos-
itive-ion mass spectrum of embutramide shows a major peak at
m/z 294.1, and the most intense product ion observed in the
MS-MS spectrum was m/z 207.9 (data not shown).
Conclusions
To our knowledge, the present method is the first described
for the analysis of colchicine in human biological samples by
means of LC-MS-MS. Owing to the single-step liquid-liquid ex-
traction, the use of 1.0 mL of sample, and MS-MS detection, the
method is simple, rapid, and highly specific and sensitive. The
good validation criteria results of the method allow its use in
colchicine detection and quantification, notably in small spec-
imens or in specimens collected long after intoxication.
References
1. E.K. Li. Gout: a review of its aetiology and treatment. Hong Kong
Med. J. 10:261-270 (2004).
2. R.L. Wortmann. Treatment of acute gouty arthritis: one physician's
approach and where this management stands relative to develop-
ments in the field. Curr. Rheumatol. Rep. 6" 235-239 (2004).
3. S.T. Milne and P.D. Meek. Fatal colchicine overdose: report of a
case and review of the literature. Am. J. Emerg. Med. 16" 603-608
(1998).
4. R Kintz, C. Jamey, A. Tracqui, and R Mangin. Colchicine poi-
soning: report of a fatal case and presentation of an HPLC proce-
dure for body fluid and tissue analyses.J. Anal. Toxicol. 21" 70-72
(1997).
5. B. Dehon, J.L. Chagnon, E. Vinner, J. Pommery, D. Mathieu, and
M. Lhermitte. Colchicine poisoning: report of a fatal case with body
fluid and post-mortem tissue analysis by HPLC. Biomed. Chro-
matogr. 13" 235-238 (1999).
6. R.A. Bonnel, M.L. Villalba, C.B. Karwoski, and J. Beitz. Deaths as-
sociated with inappropriate intravenous colchicine administra-
tion. J. Emerg. Med. 22" 385-387 (2002).
7. G.R. Jones, RP. Singer, and B. Bannach. Applications of LC-MS
analysis to a colchicine fatality. J. Anal. Toxicol. 26:365-369
(2002).
8. M. Klintschar, C. Beham-Schmidt, H. Radner, G. Henning, and
P. Roll. Colchicine poisoning by accidental ingestion of meadow
saffron (Colchicum autumnale): pathological and medicolegal as-
pects. Forensic Sci. Int. 106:191-200 (1999).
9. V.C. Danel, J.F.Wiart, G.A. Hardy, F.H. Vincent, and N.M. Houdret.
Self-poisoning with Colchicum autumnale L. flowers.
J. Toxicol. Clin. Toxicol. 39:409-411 (2001).
10. S. Sannohe, Y. Makino, T. Kita, N. Kuroda, and T. Shinozuka.
Colchicine poisoning resulting from accidental ingestion of
meadow saffron (Colchicum autumnale). J. Forensic Sci. 47:
1391-1396 (2002).
11. B. Weakley-Jones, J.E. Gerber, and G. Biggs. Colchicine poisoning:
case report of two homicides. Am. J. Forensic Med. Pathol. 22:
203-206 (2001).
12. M.J. Maxwell, P. Muthu, and P.E. Pritty. Accidental colchicine
overdose. A case report and literature review. Emerg. Med. J. 19:
Downloaded from https://academic.oup.com/jat/article-abstract/30/3/210/719491 by guest on 07 January 2020
265-267 (2002).
13. A. Folpini and R Furfori. Colchicine toxicity: clinical features and
treatment: massive overdose case report. Clin. ToxicoL 33' 71-77
(1995).
14. A. Tracqui, P. Kintz, B. Ludes, C. Roug~, H. Douibi, and P. Mangin.
High-performance liquid chromatography coupled to ion spray
mass spectrometry for the determination of colchicine at ppb
levels in human biofluids. J. Chromatogr. B 675:235-242 (1996).
15. W. Fleischmann, H.G. Price, and S.K. Fleischmann. Fate of in-
traperitoneally administered colchicine in the golden hamster.
Med. Pharmacol. Exp. Int. J. Exp. Med. 12:172-176 (1965).
16. R. Bourdon and M. Galliot. Determination of colchicine in bio-
logical fluids. Ann. BioL Clin. 34:393-401 (1976).
17. C.V. Clevenger, T.F.August, and L.M. Shaw. Colchicine poisoning:
report of a fatal case with body fluid analysis by GC/MS and
histopathologic examination of postmortem tissues. J. Anal. Tox-
icol. 15:151-154 (1991).
18. M. Lhermitte, J.L. Bernier, D. Mathieu, M. Mathieu-Nolf, F. Erb, and
P. Roussel.Colchicine quantitation by HPLC in human plasma and
urine. J. Chromatogr. B 342:416-423 (1985).
19. R.J. Ko, W.Y. Li, and R.T. Koda. Determination of the antimitotic
agents N-desacetylcolchicine, demecolcine and colchicine in
serum and urine. J. Chromatogr. 525:411-418 (1990).
20. P. Fernandez, A.M. Bermejo, M.J. Tabernero, M. Lopez-Rivadulla,
and A. Cruz. Determination of colchicine in biological fluids by re-
verse-phase HPLC. Variation of colchicine levels in rats. Forensic
Sci. Int. 59" 15-18 (1993).
21. M. Deveaux, N. Hubert, and C. Demarly. Colchicine poisoning:
case report of two suicides. Forensic ScL Int. 143:219-222 (2004).
22. I.M. Mclntyre, A.R. Ruszkiewicz, K. Crump, and O.H. Drummer.
Death following colchicine poisoning. J. Forensic Sci. 39:280-286
(1994).
23. H. Halkin, S. Dany, M. Greenwald, Y. Shnaps, and M. Tirosh.
Colchicine kinetics in patients with familial Mediterranean fever.
Clin. Pharmacol. Ther. 28" 82-87 (1980).
24. U. Cavaliere, C. Andreano, G. Raducci, C.Andreoni, and A. la-
covella. Intossicazione da T-61 (Tanax| Min. Anest. 48:861-864
(1982).
25. H. Eikmeier. Experiences with a new preparation for painless de-
struction of small animal (T61). Die Blauen Hefte. Tieraerztl. 5:
22-23 (1962).
26. E. Abe, M. Delamoye, B. Mathieu, M. Durigon, P. de Mazancourt,
C. Advenier, and J.C. Alvarez. A rapid and sensitive high-perfor-
mance liquid chromatography method for the determination of em-
butramide (a Tanax| or T-61 component) in human blood with
photodiode-array detection. J. Anal. Toxicol. 28" 118-121 (2004).
Manuscript received June 29, 2005;
revision received September 26, 2005.
215