Journal of Analytical Toxicology, Vol.

32, October 2008

Stability of Cyanide in Cadavers and in Postmortem

Stored Tissue Specimens: A Review
J.L. McAllister* and R.J. Roby
Combustion Science and Engineering, Inc., Columbia, Maryland

Barry Levine
State of Maryland, Office of the Chief Medical Examiner, Baltimore, Maryland

David Purser
Hartford Environmental Research, Hatfield, United Kingdom

Abstract
The toxicological and postmortem analysis of fire victims blood
and tissue can disclose the type and quantity of toxic species, such
as carbon monoxide or hydrogen cyanide, that they inhaled prior
to death. For fire cases, these toxicological data can reveal
objective data about the nature and circumstances of a fire, and
thus assist both the Medical Examiner and the Fire Investigator in
their investigations. Assigning a level of significance to cyanide
concentrations found in the blood and tissue of fire victims is
often hampered by the fact that cyanide is inherently unstable in
cadavers and in stored tissue samples. Numerous researchers have
provided insight into and characterized the stability of cyanide in
the body and in collected biological specimens. Based on studies
by these researchers, the rate of transformation of cyanide in
blood and tissue specimens is dependent on the initial cyanide
concentration in the sample at time of death, the length of time
that a sample remains in the cadaver, the length of time that a
sample remains in storage, and the preservation (e.g., addition of
sodium fluoride to sample) and storage conditions (e.g.,
temperature) of the sample.

Introduction
The measurement and interpretation of cyanide (and to a
lesser extent thiocyanate) in blood and other tissues is obviously important in cases of suspected poisoning by cyanide or
cyanogenic compounds such as in certain foodstuffs, but a
major consideration by the authors of this paper is in relation
to the role of hydrogen cyanide (HCN) in injuries and deaths
of fire victims. Many fuels involved in fires have a significant
nitrogen content, and studies of experimental fires have shown
* Author to whom correspondence should be addressed. Combustion Science &
Engineering, Inc., 8940 Old Annapolis Road, Suite L, Columbia, MD 21045.
E-mail: jmcallister@csefire.com.

612

high concentrations of hydrogen cyanide in fire effluent

plumes, although these concentrations are very dependent
upon the exact fuel mix and the combustion conditions in the
fire (1,2). Experimental and modeling studies have shown that
exposure to HCN concentrations exceeding approximately 100
ppm can cause rapid incapacitation (loss of consciousness) of
exposed subjects, and very high blood cyanide concentrations
have been obtained from fresh blood samples taken at fire
scenes (3). Thus, HCN is considered likely to be a critical limiting factor in the time available for escape from some categories of fires and hence a major precipitating cause of death
(1,46). However, the ultimate cause of death in fire smoke victims often appears to correlate more closely with blood carboxyhemoglobin (COHb) concentrations than with those of
blood cyanide. A complication in interpretation of effects of
carbon monoxide (CO) and HCN in fires is that, whereas blood
COHb is relatively stable in tissues and samples postmortem,
cyanide concentrations can be relatively unstable. Although it
is relatively straightforward to establish COHb concentrations
and extent of CO exposure for subjects at the time of exposure,
it is more difficult to establish blood cyanide concentrations
and the extent of HCN exposure.
The toxicology of a fire victims blood can provide important
objective data about the nature and circumstances of a fire, and
thus can assist both the Medical Examiner and the Fire Investigator. A knowledge of the relative exposure to CO and HCN
can assist in establishing how a subject was overcome during
an incident and the likely causes of death. The blood toxicology can also help to establish what fuels or burning items
were involved during the critical stages of an incident when
subjects were affected, as well as the combustion conditions
during the fire. Thus, even though most fuels produce CO
during fires, HCN is produced only by nitrogen-containing
fuels (e.g., from polyurethane foams and acrylic fabrics in upholstered furniture) (1). Also, the yields of CO and HCN are
very dependent upon the combustion conditions, with open,

Reproduction (photocopying) of editorial content of this journal is prohibited without publishers permission.

Journal of Analytical Toxicology, Vol. 32, October 2008

well-ventilated combustion conditions producing low yields

and poorly ventilated (vitiated) combustion conditions producing high yields. Information on the blood cyanide and
COHb concentrations in exposed subjects can therefore be
used to assist investigators in developing hypotheses related to
the cause and manner of death and cause and origin of the fire.
Additionally, a toxicological screen which is positive for cyanide
indicates that a nitrogen-containing fuel may have been on fire
prior to the victims death; however, cyanide stability comes
into question when interpreting the results.
Another compound relevant to cyanide poisoning is thiocyanate, which is the major product of cyanide metabolism
by rhodonase. Blood thiocyanate levels can be elevated in
acute cases of cyanide poisoning when death has been delayed (enabling time for enzymatic conversion), but there is
often a large thiocyanate metabolic pool in a subject, which
fluctuates widely with variations in dietary intake of preformed thiocyanate and plant cyanogenic glycosides (7). The
chronic intake of cyanide resulting from low concentration
HCN inhalation by tobacco smokers also results in high
blood thiocyanate levels. In a study of fire victims and other
subjects, Anderson et al. (8), found elevated thiocyanate
levels in non-fatal, non-smoker fire casualties, at similar
concentrations to smoker controls, but somewhat higher
levels in non-fatal, smoker fire casualties. Thiocyanate concentrations in fire fatalities (smokers and non-smokers)
were relatively low, presumably because the victims died
before there was time for significant enzymatic conversion
of cyanide to thiocyanate. Based upon these findings, thiocyanate measurements likely have a limited relevance in
acute cyanide poisoning, although this paper does address
the influence of thiocyanate on the stability of cyanide in
stored blood samples.
The level of inhaled cyanide detected in the body after death
and in samples collected from the cadaver can fluctuate over
time. A number of researchers have attempted to capture the
variable characteristics of cyanide in postmortem blood and
tissue samples. The dominant and most consistently reported
changes in cyanide concentration consist of decreases with
time in whole cadavers and in postmortem specimens of blood
and other tissues. For example, Curry (9) found in a case of
death by inhalation of cyanide vapor that the blood level in a
sample taken at the moment of death was 3.5 mg/L (site not reported), and samples taken at autopsy the next day were 1.0
mg/L (femoral) and 0.5 mg/L (carotid), an average decrease of
approximately 79%.
Ballantyne et al. (1012) did extensive studies to quantify the
rate of transformation of cyanide in cadavers and postmortem
specimens. Several mechanisms were identified for the postmortem transformation (decrease) of cyanide, including conversion to thiocyanate, hydrolysis to ammonium formate, and
reactions with aldehydes and polysulphides in postmortem
tissues (13). These mechanisms were considered to readily
occur in dead animals and in most tissues, but less so in blood,
which may account for the slower decrease of cyanide concentrations in blood samples during storage compared to other
tissues. However, decreases of cyanide in serum appear to
occur more readily than in whole blood or red cells.

An important parameter in relation to rate and extent of decrease is temperature, with greater and more rapid decreases
occurring at higher temperatures. For fire victims, tissues
may be subjected to heating above body temperature for significant periods before specimens can be obtained. In some
localities, a cadaver may then remain at ambient temperature
for periods of hours to several days before autopsy. Similarly,
specimens such as blood samples may subsequently be stored
at ambient temperature for a period of time, then refrigerated
(4C) or frozen (20C) for periods of up to months before
analysis.
Another complicating factor is the possibility of formation of
cyanide in tissue during storage prior to analysis, which may
somewhat offset transformation losses or even result in an increase in cyanide concentrations under certain circumstances.
In this context, it is important to distinguish between the relatively high blood cyanide concentrations occurring in cases
involving cyanide poisoning and the low, background level
concentrations occurring naturally and in subjects such as
smokers, in those on diets high in cyanogens, and in patients
receiving cyanogenic drugs. According to Ballantyne (13) and
Vesey and Wilson (14), a small amount of cyanide can be
formed in samples that were initially negative for cyanide,
which is most likely due to conversion of plasma thiocyanate
(SCN) to cyanide on acidification of whole blood in the presence of free hemoglobin. This conversion may represent a significant artifact during the study of such conditions, but it
would be considered less relevant at levels of lethal significance. However, on rare occasions in some other studies
(15,16), considerably greater increases in cyanide concentration have been reported to occur following storage of initially
cyanide-negative blood.
This paper will review the published literature in an attempt
to establish a foundation for the proper interpretation of postmortem cyanide concentration changes.

Decreases in Cyanide Concentration

Arguably, effects on whole blood are the most important
consideration in relation to cyanide poisoning, both in the cadaver and in the stored sample removed from the cadaver.
However, rapid decreases in cyanide concentrations in serum
and plasma were also of interest, as illustrated by experimental
Table I. Summary of Results from Rate of Cyanide
Transformation Testing by Ballantyne et al. (11)

Specimen

Mean Percentage of
Original CN Concentration
Detectable after 1 h at
Ambient Temperature

Human urine
Human blood serum
Rabbit blood serum
Rabbit whole blood

73%
39%
33%
78%

Mean Percentage
of SCN Appreciation
after 1 h at
Ambient Temperature
30%
33%
26%
Not measured

613

Journal of Analytical Toxicology, Vol. 32, October 2008

work of Ballantyne et al. (10). Furthermore, the Ballantyne experiment revealed one possible mechanism of decrease in
whole blood via transformation of cyanide to thiocyanate. In
one of their studies, Ballantyne et al. (10) measured the
changes of cyanide concentrations in human serum and urine,
and in rabbit serum and whole blood, all spiked with potassium
cyanide. SCN, a metabolite of cyanide, was also measured in
the urine and serum specimens. To quantify the significance of
the thiocyanate development in the spiked samples, Ballantyne
and co-workers (10) determined the thiocyanate concentration
in control samples and found that human urine and serum and
rabbit serum were at concentrations of 2.49 0.81 mg/L, 5.80
1.74 mg/L, and 0.64 0.65 mg/L, respectively. Although
the authors did not report the storage temperatures of the
samples, all samples are assumed to have been stored at ambient room temperature. Potassium cyanide was added to each
substrate for a starting concentration of 1.2 mg CN/L. The
concentration of cyanide in the samples was tested at increments of 4, 10, 20, 30, 40, 50, and 60 min after addition of the
potassium cyanide. Table I depicts a summary of the results
from the Ballantyne et al. (10) testing. It is notable that, al-

Figure 1. Summary of results from rate of cyanide transformation testing

by Ballantyne et al. (11). (All samples removed within 30 min after
death.)

Figure 2. Summary of results from rate of cyanide transformation testing

by Ballantyne et al. (11). (Samples removed from rabbits at indicated
times after death.)

614

though there were significant decreases in cyanide content, especially in serum, there were increases in thiocyanate.
Based on this research, the authors concluded that the most
rapid decrease in cyanide concentration in human and rabbit
serum occurs in the first 20 min after the sample was spiked,
but that only a small amount of the lost cyanide was measured
as SCN in the serum specimens. Specifically, only approximately 26% of the rabbit and human serum cyanide was recovered as thiocyanate. The rate of depreciation of cyanide in
the spiked rabbit blood was less significant than in serum;
however, the most significant decrease of cyanide in the blood
occurred within the first 20 min, as was also seen in serum.
The rate of depreciation of cyanide in urine was also less than
in serum, and the deviation in measurements of cyanide urine
concentration was notable, ranging from 49% to 88%.
The authors further explored the transformation of cyanide
in whole blood removed from rabbits injected with 8 mg
CN/kg of body weight. Blood was drawn from the animals 4
min after death, and the cyanide concentration of the blood
was then measured at various time increments up to 24 h
after death. Although the authors did not report the storage
temperatures of the samples, all samples are assumed to have
been stored at ambient room temperature. At 4 min, the
cyanide blood concentration was approximately 15.9 mg
CN/L. After 1 h, only 84% of this concentration was present in
the collected blood. After a 24-h period, a mean of 67% of the
initial concentration was present in the blood. Therefore, the
same trend was found in the blood recovered from the rabbits
injected with cyanide as in the spiked samples.
Ballantyne et al. (11) further expanded their study in a second
publication that compared the rate of transformation of cyanide
in blood and tissue removed from rabbits immediately after
death with the rate of transformation of cyanide in blood and
tissue removed from rabbits at various times after death. The
specimens measured included blood, lung, brain, kidney, and
liver. The rabbits were killed by injection of 8 mg CN/kg of body
weight. Within 30 min of death, specimens were removed from
the rabbits. Three sets of specimens were immediately analyzed, followed by analysis of samples at 1, 3, 7, 14, and 21
days. All samples were stored between 10C and 15C. Three additional sets of specimens were removed at 30 min and analyzed, followed by the removal and analysis of samples at 1, 3,
7, 14, and 21 days after death. The rabbit carcasses were stored
at 10C and 15C in sealed polyethylene bags during the test period. Figure 1 depicts the results from measurements of samples removed from the animals immediately after death and
tested at various time increments. No detectable levels of
cyanide were found in the liver at 1 day, the kidney at 3 days,
and the brain and lungs at 14 days. The cyanide concentration
in the blood at 14 days was approximately 38% of the original
value measured immediately after death and increased by 8%
from 14 days to 21 days of storage. A slight increase in the
concentration of cyanide in the lung tissue was also documented between days 1 and 3. Figure 2 depicts the results from
testing of organs and blood samples removed from the animals at various times after death. No detectable levels of cyanide
were found in the liver and kidney at 3 days and the brain and
lung at 14 days. The cyanide concentration in the blood at 21

Journal of Analytical Toxicology, Vol. 32, October 2008

days was only 0.4% of the original value measured immediately

after death. No increases in concentration were noted over time
in this portion of the study. In general, the recovered quantity
of cyanide in these samples was less than that recovered in the
samples removed from the animal immediately after death and
tested at various time increments.
Based on this research, Ballantyne et al. (11) concluded that
toxicologically significant quantities of cyanide were still detectable in the blood stored up to 21 days after removal from
the rabbits; however, cyanide was almost undetectable in the
blood removed from the rabbit carcasses at 21 days and then
tested. The researchers attributed the highest concentrations
of cyanide in blood to the high affinity of erythrocytes for the
cyanide radical, noting that Barr (17) found that approximately 70% of cyanide in whole blood is bound to hemoglobin.
Ballantyne et al. (11) attributed the more rapid depreciation of
cyanide in the samples removed from the animals at various
times to autolysis and putrefactive effects not present in the
samples in stored, sealed containers.
In a third study, Ballantyne (18) investigated the effects of
storage temperature on cyanide-spiked human blood. He monitored the changes in cyanide levels in blood stored at 20C,
4C, and 20C for up to 12 weeks. Ballantyne also monitored
cyanide levels in normal, non-spiked, human blood samples.
The blank samples were found to have an average cyanide concentration of 0.03 mg CN/L, and the testing showed that
cyanide production was most notable in these samples when
they were stored at 20C. Concentrations of cyanide in the
normal samples increased 200% to 367% of the original concentration, with the most notable changes occurring in the
first few days of storage. Although these increases in the negative samples were large on a percentage basis, the resultant
cyanide concentrations, ranging from 0.09 mg CN/L to 0.14
mg CN/L, were still not toxicologically significant.
In the cyanide-spiked blood samples, concentrations of 1.02,
2.33, 3.72, and 6.52 mg CN/L were assessed for changes.
During the first four days of testing, cyanide levels decreased
for all sample concentrations and storage conditions, with the
most significant changes occurring in the 6.52 mg CN/L concentrations stored at 20C and 4C. In general, changes in
cyanide were less marked in the 1.02 and 2.33 mg CN/L concentrations and most marked in the 6.52 mg CN/L concentration. Additionally, changes in cyanide in spiked blood
samples were less significant at 20C than at refrigerated and
room temperatures. The 20C storage temperature was also
the only condition under which cyanide concentrations in the
samples returned to or exceeded their initial concentration
after the 12 week test period. A maximum decrease of 60% was
seen in the 1.02 mg CN/L concentration at 20C.
A study published by Bright et al. (19) investigated the depreciation of cyanide in spiked blood samples ranging in concentration from 0.1 to 5.0 mg CN/L. Samples were stored at
ambient conditions for up to 12 min and stored at refrigerated
(4C) and frozen (20C) conditions for up to 6 months. Unlike
in the other studies presented, researchers in this study used
blood from a blood bank. This blood contained additives such
as sodium citrate, dextrose, hydrated citric acid, sodium dihydrogen phosphate, and adenine. The blood was spiked with

potassium cyanide and stored in containers treated with

sodium fluoride and sodium oxalate.
The Bright et al. (19) study, like the Ballantyne et al.
(10,11,13) research discussed, showed a dependency between
cyanide concentration and depreciation in samples. Bright et al.
(19) concluded that blood containing the highest concentration
of cyanide (5.0 mg CN/L) depreciated more quickly under all
storage conditions than those samples at lower concentrations.
A summary of the test results is presented in Table II. Most notably, after 8 weeks of storage at 4C, the cyanide concentration
in the spiked blood decreased from 5.0 to 3.5 mg CN/L. Based
on a graphical representation in the Bright report, the same
concentration stored at 20C decreased slightly more than at
4C. By comparison, a cyanide concentration of 2.5 mg CN/L
over an 8 week period decreased to 2.08 mg CN/L at 4C and
2.07 mg CN/L at 20C. Interestingly, although blood stored at
ambient temperature at concentrations of 2.5 mg CN/L and 5.0
mg CN/L experienced significant depreciation, blood concentrations of 0.1 mg CN/L and 0.5 mg CN/L remained stable.
This same phenomenon was not seen under refrigerated or
frozen conditions, where the cyanide levels continued to slowly
decrease over time. Researchers ultimately concluded that
storage of samples at 20C was most conducive to longer term
blood cyanide stability. This same conclusion was supported by
the Ballantyne studies, which found that samples stored at
20C experienced the least variability over time.
The basic pattern emerging from this body of work is that
significant continuous decreases in blood cyanide concentration occur during storage. The greatest and most rapid decreases occur in cadavers and in blood samples containing
high cyanide concentrations, especially when stored at ambient temperature. Whole blood cyanide appears to decrease
less under refrigeration, while freezing is recommended to
provide the maximum stability.
A further consideration is that blood in cadavers may be
subjected to elevated temperatures. Moriya et al. (20) also
Table II. Cyanide Depreciation in Samples Stored for 8
Weeks at Different Temperatures

Sample

Storage
Temperature

Storage
Time

1
2
3
4

4C
20C
4C
20C

8 weeks
8 weeks
8 weeks
8 weeks

Initial Cyanide
% Loss of
Concentration (mg/L) Cyanide
5.0
5.0
2.5
2.5

30%
31%
17%
17%

Table III. Percent Loss of Cyanide from Groups A, B1,

and B2 at 24 and 72 h

Sample
Time

Group A
High MetHb
%Loss of Cyanide

24 h
72 h

85
90

Group B1
Group B2
Low MetHb
Low MetHb
% Loss of Cyanide % Loss of Cyanide
47
63

45
100

615

Journal of Analytical Toxicology, Vol. 32, October 2008

studied the changes in cyanide concentrations in the body

over time under various storage conditions. In addition, the
study investigated the effects of methemoglobin (MetHb) concentration on binding of cyanide in the blood, noting that
hemoglobin is converted into MetHb (the main cyanide
binding component in the blood) by oxides of nitrogen produced during fires. The Moriya et al. (20) study analyzed
cyanide concentrations in 10 fire victims and in rabbits injected
with potassium cyanide. In the fire victim cases, blood samples
were analyzed 1 h after autopsy or 48 h after storage at 4C.
Only one measurement of the blood sample was taken; therefore, transformation rates were not analyzed. However, the
tests did show that the highest quantities of measured cyanide
came from the pulmonary vein in 7 of 7 cases where blood was
measured from that site.
The animal experiments did provide data related to the rate
of transformation of cyanide. In these experiments, rabbits
were anesthetized via intramuscular injection with sodium
pentobarbital and intravenously injected with 0.33 mL/kg of
3% sodium nitrite (Group A) or saline (Group B), followed by
1 mL/kg intramuscular injection of 2% potassium cyanide.
Blood was drawn after the injection of the sodium nitrite or
saline to measure MetHb in the blood. The rabbit hearts were
removed after death and placed under various temperature
conditions. The rabbit hearts from Group A were left at 46C
for 1 h, at 20C to 25C for 23 h, and at 4C for the remaining
48 h (cumulative 72-h test duration). Three rabbit hearts from
Group B were left under the same conditions as those in Group
A, and three hearts were left at 46C for 1 h and 20C to 25C
for the remaining 71 h (cumulative 72-h test duration). The
rabbit hearts that were kept under the same conditions as
Group A were designated as Group B1. The hearts that were
kept under different storage conditions were designated as
Group B2. Blood samples were taken from the Group A and
Group B hearts at 1, 6, 24, 48, and 72 h of storage. The results
showed that average MetHb concentrations were 6.9% (Group
A with sodium nitrite) and 0.8% (Group B with saline) before
the injection of potassium cyanide. The results of testing are
shown in Table III and Figure 3.
The researchers found that 85%, 47%, and 45% of the
cyanide disappeared within 24 hours of administration of the
potassium cyanide in Groups A, B1 (same storage conditions as
Group A and low MetHb), and B2 (different storage conditions
and low MetHb), respectively. Although Group B2 was kept
under different storage conditions over the entire 72-h period,
during the first 24 h of testing, the temperature conditions of
Group B2 were the same as Groups A and B1 (46C for 1 h and
20C to 25C for 23 h). Therefore, in the first 24 h of testing, the
cyanide concentrations in Groups B1 and B2 (both low MethHb)
should have, and did, provide comparable results. After a 72-h
storage period, the cyanide concentration in Group B2 was almost nil, and only 10% and 37% of the original concentrations were present in Group A and B1, respectively. Ultimately,
Moriya et al. (20) concluded that more than 40% and nearly
100% of the original blood cyanide concentrations were gone
after 24 h and 3 days, respectively. It should be noted that the
initial concentration of cyanide in Group A samples was 47.4
mg/L and in Group B samples was 3.56 mg/L. The animals

616

were administered the same dose of cyanide, and Moriya and coworkers (20) noted that Group A rabbits survived longer (16
min) than Group B rabbits (6 min). This change in survival time
was presumably due to the protective reaction of some of the
cyanide with methemoglobin to form cyanomethemoglobin.
Although the author does not comment on the cause of the
higher concentrations of cyanide in Group A, the pretreatment
of the Group A rabbits with sodium nitrite may have allowed a
large proportion of cyanide to be sequestered into the blood and
bound to MetHb, while more cyanide in the Group B rabbits was
distributed into the tissue and other body fluid compartments.
Moriya commented that the rate of disappearance of cyanide
was greater in the blood containing higher concentration of
cyanide (Group A) than in Group B. Ballantyne and Bright
showed a strong relationship between cyanide concentration
and % depreciation; therefore, it is likely that this relationship
may have played a role in concentration changes. Because the
Group A rabbits had higher MetHb and higher cyanide concentrations, the independent role of each variable in the
changes in cyanide concentrations in the sample is unclear.
Based on the Moriya et al. (20) results, it appears that in the
high MetHb concentration group, there was increased cyanide
depreciation in samples stored under elevated temperature and
room temperature conditions. However, under refrigerated
conditions, the subsequent percentage depreciations of the
high MetHb samples were similar to those with low MetHb
(Figure 3), representing approximately 30% of that remaining
after 4 h. Additionally, when samples were stored at room temperature for a significant period of time (e.g., 3 days), almost all,
if not all, of the cyanide disappeared, as was seen in Group B2.
In a study by Narayanaswami et al. (21), human stomach
tissue samples were spiked with cyanide and stored at room
temperature for 1, 4, 8, 10, 15, and 20 days. The tissues were
maintained at a pH of 6.5 and spiked with 1 mL of sodium
cyanide ranging in concentration from 50 to 150 mg/L. An
additional set of samples maintained at a pH of 8.0 was spiked
with 1 mL of 100 mg/L sodium cyanide. Researchers found that
75% to 80% of the cyanide in the tissue at the 6.5 pH was lost
in the first week, after which point no significant changes were
seen. Sample initial concentration did not appear to have a significant effect on percent decrease in cyanide concentration.

Figure 3. Loss of cyanide from Groups A, B1, and B2 at 24 and 72 h.

Journal of Analytical Toxicology, Vol. 32, October 2008

The stomach tissue samples at the 8.0 pH, however, showed no

significant change throughout the duration of the 20 day
testing period.
Like Moriya and co-workers (20), Seto et al. (22) also investigated the effects of above-ambient temperatures on cyanide
stability in blood samples. In their study, Seto et al. (22) spiked
0.82 mol hemoglobin (Hb) with 200 ng of potassium cyanide
or 0.2 mL of blood with 100 ng of potassium cyanide (representing 0.5 mg CN/L). The samples were heated for 1 h at a
constant temperature of 25C, 50C, 63C, 75C, or 90C. After
heating, 0.5 mL of the sample was injected into a headspace gas
chromatograph (HS-GC) equipped with a nitrogen-phosphorus
detector, and the quantity of free cyanide in the sample was
measured. In order to determine the amount of bound cyanide
in the sample, approximately 0.2 mL of 50% phosphoric acid
was added to the remaining sample. Subsequently, the sample
was injected into the HS-GC and then measured. The author defined free cyanide as that which was detected in the blood or
hemoglobin without the addition of phosphoric acid, and defined bound cyanide as that which was only detected upon
treatment of the sample with phosphoric acid. At temperatures
below 63C, nearly 100% of the added cyanide was recovered in
the bound form in the hemoglobin. Above 63C, approximately
half of the cyanide was in the free form in the hemoglobin. The
quantity of bound cyanide in the hemoglobin decreased with increasing temperature. For example, hemoglobin at 75C had approximately 60% bound cyanide, and at 90C, the bound
cyanide was reduced to 40%. Additionally, the total cyanide
(i.e., bound plus free) in the hemoglobin also decreased with increasing temperature, resulting in a total cyanide concentration
in the hemoglobin of only 60% of the original at 90C compared
to 100% at 73C. When blood was heated at temperatures below
63C, similar trends to those found for hemoglobin were observed. Approximately 100% of the cyanide in the blood was recovered in the bound form. The bound and total cyanide also
decreased with increasing temperature in the blood. At 75C,
approximately 30% of the cyanide was bound, and at 90C only
5% of the cyanide was bound. The total cyanide at 75C was less
than 35%, and at 90C, the total cyanide in the blood was less
than 5%.

free cyanide was only produced in the samples at temperatures greater than 50C. The highest total cyanide (bound plus
free) was found to be produced in both hemoglobin and blood
at 75C. A total cyanide content of approximately 8 ng was
produced in hemoglobin, and 12 ng was produced in blood at
75C. Seto and co-workers (22) concluded that the maximum
heat-induced cyanide production, which was less than 0.050
mg/L in blood, did not exceed that which was toxicologically
significant and, therefore, would have a negligible influence on
confirmation of cyanide poisoning in fire victims.
In a study published by Chikasue et al. (23), blood and tissue
samples from five cyanide poisoning cases (non-fire related)
were collected and measured at the time of autopsy. The samples were again measured for cyanide concentrations between
1 day and 3 weeks after collection and storage at refrigerated
(4C) and frozen (20C) conditions. The times from death to
autopsy were 8 h, 13 h, 16 h, 13 h, and 1820 h for Cases 15.
The summaries of test results for Cases 1 and 2 are presented
in Figures 4 and 5, respectively. The abbreviations LHB and
RHB in Figures 4 and 5 represent the left and right heart
blood. The researchers found that the ratio of blood cyanide
concentrations after storage to that before storage ranged from
0.71 to 1.46 under refrigerated and frozen storage conditions.

Figure 4. Case 1 changes in cyanide concentration over time in various

samples. The abbreviations LHB and RHB represent the left and right
heart blood.

Increases in Cyanide Concentration

Although decreases in cyanide concentration are the main
finding from stability studies, particularly when initial tissue
concentrations are high, increases in concentration have been
reported in a number of studies, especially those involving
low (near background) concentrations. Moreover, other studies
have reported relatively large increases in certain situations,
possibly due to microbial action in poorly preserved samples.
In addition to the spiked cyanide studies, Seto et al. (22) also
measured the production of cyanide in negative hemoglobin
and blood samples heated at 25C, 50C, 63C, 75C, and 90C
to quantify the rate of endogenous production and determine
the extent to which recovered cyanide in spiked samples was related to endogenous production. The research showed that

Figure 5. Case 2 changes in cyanide concentration over time in various

samples. The abbreviations LHB and RHB represent the left and right
heart blood.

617

Journal of Analytical Toxicology, Vol. 32, October 2008

A larger range of ratios from 0.2 to 8.8 was measured for liver,
kidney, brain, and stomach content cyanide concentrations
under refrigerated and frozen conditions. Cases 1 and 2 are the
only cases for which full data was presented. In general, the researchers found increases in cyanide concentrations in Cases
1 and 2 after 20 days of freezing in all samples, with the exception of the stomach contents from Case 2, which showed a
slight decrease. In Case 1, after 24 days of refrigeration, an
overall increase was found in the right (+ 31%) and left
(+ 20%) heart blood, kidney (+ 87%), and brain (+ 60%) samples, and a decrease was found in the stomach contents (37%).
No change was seen in the liver sample. In Case 2, after 24 days
of refrigeration, an overall increase was measured in the right
(+ 26%) and left (+ 29%) heart blood, and decreases were measured in the liver (78%), kidney (53%), brain (43%), and
stomach contents (54%).
The research of Curry et al. (15) and Lokan et al. (16) focused
on the development of cyanide in negative samples stored
under various temperature conditions. In a publication by
Curry et al. (15), research on cyanide development in postmortem negative samples arose from a case in which a positive
cyanide sample was found where no cyanide exposure was possible. Curry et al. (15) found that serum and red cell samples
showed no cyanide production during a two-week period at
room temperature, with the exception of one sample that measured 0.25 mg/L cyanide. However, in transfusion blood samples, stored in test tubes closed with cotton wool plugs at room
temperature for two months, high cyanide concentrations
were found on the order of 100 mg/L. In one case stored at 4C
for 3 months, cyanide levels in the sample apparently increased
from 3 to 40 mg/L. Another sample, initially containing no detectable cyanide, showed a level of 5 mg/L after 2.5 months of
storage at 4C, but this level then decreased to 0.1 mg/L after
6 additional months of storage at 4C. The authors found no
cyanide in a sample preserved with 1% sodium fluoride, a
compound which the authors stated will completely inhibit
cyanide production. Nevertheless, the authors concluded that,
although it was not a constant phenomenon, cyanide production could occur in bacteriologically sterile stored blood. The
significance of this work is difficult to fully evaluate, because
this research was carried out under variable conditions using
a semi-qualitative method of analysis.
In contrast, Ballantyne (24), in a series of detailed studies of
fresh pooled blood, found a slow transformation in the cyanide
concentration when the samples were stored at room temperature and 4C for up to 15 weeks, with increases in negative
samples seen only in those samples stored at 20C. The increases were small and only reached a maximum peak cyanide
level of 0.20 mg/L. Ballantyne (13) and Vesey et al. (14) considered this increase to be due to conversion of approximately
3% of plasma thiocyanate to cyanide. Egekeze and Oehme (25)
also reported only minor changes over 7 days at 4C and 25C.
Lokan et al. (16) developed an interest in negative sample
cyanide production, because they identified three postmortem
samples that had high positive cyanide concentrations, but
where no sources of victim cyanide exposure could be identified. However, they pointed out that these three unusual cases
should be set against a total 331 cases routinely screened for

618

cyanide, the remainder of which showed no significant levels of

cyanide. Lokan and co-workers (16) studied the blood, liver,
and stomach contents from these cases for cyanide concentrations. In Case 1, the autopsy was conducted 12 h after death,
and the samples were preserved with 0.1% sodium fluoride and
stored at 4C for 20 days before testing. In Case 2, the autopsy
was conducted on the morning of death, and the samples were
stored for 49 days at 18C prior to thawing and then maintained at 4C during screening. In the last case, the autopsy was
conducted 46 h after death and blood samples (both unpreserved and preserved in 1% sodium fluoride) were stored for 41
days at 18C prior to thawing and then maintained at 4C
during screening.
In Case 1, the preserved blood, which contained 0.1% of
sodium fluoride, was found to have 22 mg/L of cyanide. The
liver and stomach contents, which had no preservative, contained 0.6 mg/kg and 0.1 mg of cyanide, respectively. In Case 2,
no preservative was present in the blood, and the blood cyanide
concentration was measured at 110 mg/L. The liver concentration was 6 mg/kg. Stomach contents were not measured. In
the last case (Case 3), blood was preserved with 1.0% sodium
fluoride, and no cyanide was detected after 41 days of storage.
However, the unpreserved blood sample from this case had a
concentration of 150 mg/L, and the stomach contents had a
concentration of 0.1 mg of cyanide. No liver concentration
was reported in this case.
Lokan et al. (16) could not explain their findings for these
three cases, but considered that microorganisms (such as Pseudomonas aeruginosa, a cyanide-producing species) may have
been responsible for the cyanide production. They agreed with
Curry et al. (15) that 1% sodium fluoride apparently inhibited
production of cyanide in the blood sample from Case 3; however, the low concentration of sodium fluoride (0.1%) had no
apparent effect in Case 1. The Case 3 sample also demonstrated
that there was no cyanide present in the body at the time of
autopsy. Lokan et al. also considered the possible role of
the freezing and thawing cycle to the high levels of cyanide
that were found in the blood of Cases 2 and 3 in comparison to
Case 1.
Karhunen et al. (26) reported a case of high cyanide levels in
a victim who was stabbed to death and then incinerated two
days postmortem. The victim was found to have a low COHb
level of 4%, consistent with his smoking habit, and a high
blood cyanide level of 10 mg/L. Based on a description of the
condition of the body provided in the Karhunen et al. (26)
paper, the fire consumed and exposed the victims thoracic
cavity. Karhunen et al. (26) ultimately concluded that the high
cyanide level in the victims blood was caused by diffusion of
cyanide produced by the fire into the open thoracic cavity.
Karhunen et al. (26) did not identify the material that produced
cyanide during the combustion event, nor did they discuss
the potential contribution of endogenous cyanide production
from the bacteriological mechanism identified by Curry et al.
(15). Because Curry et al. (15) and Lokan et al. (16) have shown
that cyanide levels in negative samples can exceed the concentration found in this incident (10 mg/L), the overall effects
of the fire on the victims blood cyanide concentration are
questionable.

Journal of Analytical Toxicology, Vol. 32, October 2008

In addition to the studies described, all of which followed

changes in blood cyanide concentrations using serial measurements, a number of studies have involved single blood
cyanide determinations in fire victims. These studies have included blood samples taken from cadavers after a range of
times (and from some fire survivors) and stored under refrigerated and frozen conditions for varying periods of hours to
months before analysis. Baud et al. (3) examined 109 cases of
survivors and decedents, Anderson et al. (27) examined 127 fatalities, Clark et al. (28) examined 53 fire survivors, and Levin
et al. (29) examined a large number of samples from a single
fire incident. Levin et al. found small increases in cyanide concentration on freezing and re-thawing samples, similar to
those reported by Ballantyne and co-workers (10,11). However, none of these studies, which involved a large numbers of
cases, reported exceptionally high cyanide concentrations similar to those obtained by Curry et al. (15) or Lokan et al. (16).
Thus, the exceptionally high cyanide concentrations reported
by Curry et al. (15) and Lokan et al. (16) appear to be outliers
when viewed in the context of all the other fire death cases involving blood cyanide.

Conclusions
If appropriate procedures are followed to preserve samples,
fire victim cyanide and COHb levels can provide investigators
with objective data for use in hypotheses development. Fuels
containing nitrogen as well as carbon produce high concentrations of both CO and HCN under vitiated combustion conditions occurring in many enclosure fires (such as many fires in
buildings or transport vehicles), whereas smoldering or nonflaming fire produce mainly CO and rapidly growing well-ventilated fires produce low yields of both CO and HCN. At
concentrations typical of vitiated fires, HCN can produce incapacitation significantly more rapidly than CO. For these reasons, the levels of both cyanide and COHB in the blood of fire
victims can provide important information relating to the materials involved during critical phases of fire development and
(especially when used in conjunction with fire experiments or
fire modeling), can contribute to evaluating the time course and
causes of incapacitation and death in fatal fire incidents.
As an example of how postmortem blood cyanide information can be used in a fire investigation, consider a case where
an investigator arrives on a fire scene and finds fire damage
confined to a living room, with a fire victim dead in an upstairs
bedroom. The fire was determined to have started in an overstuffed cotton chair and eventually spread to adjacent
polyurethane-treated wood paneling. The fire investigators
needed to determine whether the victim died early in the fire
when only the chair was involved, or later in the fire during full
room involvement. Although the victim in this case clearly
died of CO poisoning with a COHb level greater than 70%,
the victims blood was found to have nearly 2.0 g/mL of
cyanide The presence of significant cyanide in the victims
blood allowed the fire investigators to determine that the fire
spread to adjacent polyurethane-treated wood paneling. There-

fore, the fact that the victim did not die until late in the fire,
after involvement of the wood paneling, was determined using
blood cyanide toxicological data.
The material presented in this paper shows that the rate of
transformation of cyanide in blood and tissue specimens is dependent on four criteria: 1. initial sample concentration at time
of death; 2. length of time that sample remains in the cadaver
after death; 3. length of time that sample remains in storage before analysis; and 4. conditions of sample preservation and
storage (i.e., temperature, pH, sodium fluoride addition).
The basic findings from experimental work carried out under
laboratory conditions are that when cyanide concentrations of
near-lethal significance (i.e., approximately 3 mg/L) are present
in a cadaver at death or in postmortem blood or tissue samples,
a subsequent decrease in concentration occurs over periods of
hours to months. The greatest rate of cyanide decrease occurs
in cadavers, and to a slightly lesser extent in blood samples, at
ambient or elevated temperatures. Serum and plasma cyanide
levels decrease rapidly (within 20 minutes) when stored at ambient temperatures. The rate of cyanide decrease in whole blood
is slower: it can be significant at 4C, but is small when samples
are frozen at 15C to 20C. Under conditions where cyanide
decreases occur, these decreases tend to be greater with higher
initial blood cyanide concentrations. When blood is frozen, a
small increase tends to occur upon thawing (approximately 0.2
mg/L), which is considered to be due to the formation of small
amounts of cyanide resulting from acidification of plasma or
serum thiocyanate in the presence of hemoglobin.
Most studies, as discussed here, show a decrease in cyanide
concentration in postmortem cadaver, blood, and tissue samples. However, there are a limited number of reports from
studies of cyanide-negative transfusion blood and a small
number of cases considered not to have involved cyanide exposure [around 1% of routine cases examined by Lokan et al.
(16)] in which very high concentrations of cyanide have been
measured in blood samples stored under room temperature, refrigerated, or frozen conditions for periods of a month or more.
In several of these cases, measured cyanide levels were more
than 30 times the lethal concentration. No obvious reasons
have been identified for the very high levels of cyanide formation in these samples, although microbiological contamination
is a possible consideration. The authors (15,16) issue a caveat
that considerable care should be exercised in attempting to interpret high blood cyanide concentrations in unpreserved
blood specimens in isolation. The authors reporting these findings have found them not to occur in blood samples preserved
with 1% sodium fluoride; therefore, it is recommended that at
least one blood sample should be preserved with 1% sodium
fluoride at postmortem collection. The lack of reports of similar exceptionally high blood levels in several major blood
cyanide studies of fire fatalities and survivors indicates that
such instances are relatively uncommon.
The main recommendations from this review are that samples of blood and other tissues should be taken as soon as possible after death and analyzed for cyanide as soon as possible
after collection. Storage of samples at 4C provided relatively
stable results; however, samples are best preserved in deepfreezer (20C). Additionally, the time intervals between death,

619

Journal of Analytical Toxicology, Vol. 32, October 2008

sampling, and analysis should be recorded to allow for better

interpretation of toxicological results.
In general, repeatable (non-case study) tests that assessed increases of cyanide in samples showed that cyanide increases
were an order of magnitude less than those levels typically
found in lethal cyanide exposures (3 mg/L). Therefore, when
levels of cyanide in the blood are within lethal ranges and
samples are collected and analyzed within 24 h of death, artificial increases in cyanide can be considered negligible. If,
however, the sample cyanide concentrations are an order of
magnitude less than the lethal range, these levels should be
compared against the COHb blood level and the known facts of
the case regarding fuel loads and their relative burning characteristics prior to data interpretation. Average decreases of
sample cyanide concentrations within typical time periods of
autopsy and toxicological sample analysis (within 24 h of death)
showed an approximate maximum decrease of 60% in blood in
repeatable (non-case study) tests. Therefore, investigators
should consider the possibility of this decrease in their analysis
and bracket their resultant conclusions accordingly. The interpretation and application of toxicological data in fire cause
and origin determination are not simple processes. The authors
caution that every situation should be reviewed independently
and weighed against all the known facts of the case when developing and testing hypotheses. This paper intends to provide
investigators with a comprehensive look at the data regarding
cyanide stability so that this information can then be applied as
appropriate to specific cases.

9.
10.

11.
12.
13.
14.
15.

16.

17.
18.
19.

20.

21.

22.
23.

References
1. D.A. Purser. Toxicity assessment of combustion products. In The
SFPE Handbook of Fire Protection Engineering, 3rd ed. National
Fire Protection Association, Quincy, MA, 2002, pp 2/832/171.
2. D.A. Purser. Toxic product yield and hazard assessment for fully
enclosed design fires involving fire retarded materials. Polymer
Int. 49: 12321255 (2000).
3. F.J. Baud, P. Barriot, V. Toffis, B. Riou, E. Vicaut, Y. Lecarpentier,
R. Bourdon, A. Astier, and C. Bismuth. Elevated blood cyanide
concentrations in victims of smoke inhalation. New Engl. J. Med.
325: 17611766 (1991).
4. D.A. Purser and W.D. Woolley. Biological effects of combustion
atmospheres. J. Fire Sci. 1: 118144 (1982).
5. D.A. Purser, P. Grimshaw, and K.R. Berrill. Intoxication by cyanide
in fires: a study in monkeys using polyacrylonitrile. Arch. Environ.
Hlth. 39: 394400 (1984).
6. D.A. Purser. Behavioural impairment in smoke environments.
Toxicology 115: 2540 (1996).
7. B. Ballantyne. The forensic diagnosis of acute cyanide poisoning.
In Forensic Toxicology. Wright, Bristol, U.K., 1974, pp 99112.
8. R.A. Anderson, I. Thomson, and W.A. Harland. The importance

620

24.

25.

26.

27.

28.

29.

of cyanide and organic nitriles in fire fatalities. Fire Mater. 3:

9199 (1979).
A.S. Curry. Cyanide poisoning. Acta Pharmacol. Toxicol.
(Copenh.) 20: 291294 (1963).
B. Ballantyne, J. Bright, and P. Williams. An experimental assessment of decreases in measurable cyanide levels in biological
fluids. J. Forensic Sci. Soc. 13: 111117 (1973).
B. Ballantyne, J. Bright, and P. Williams. The post-mortem rate of
transformation of cyanide. Forensic Sci. 3: 7176 (1974).
B. Ballantyne and T. Marrs. Clinical and Experimental Toxicology
of Cyanides. Wright, Bristol, U.K., 1987.
B. Ballantyne. Artifacts in the definition of toxicity by cyanides
and cyanogens. Fundam. Appl. Toxicol. 3: 400408 (1983).
C.J. Vesey and J. Wilson. Red cell cyanide. J. Pharm. Pharmacol.
30: 2026 (1978).
A.S. Curry, D.E. Price, and E.R. Rutter. The production of cyanide
in post mortem material. Acta Pharmacol. Toxicol. 25: 339344
(1967).
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: 253259 (1987).
S. Barr. The micro determination of cyanide: Its application to the
analysis of whole blood. Analyst 91: 268272 (1966).
B. Ballantyne. Changes in blood cyanide as a function of storage
time and temperature. J. Forensic Sci. Soc. 16: 305310 (1976).
J.E. Bright, R.H. Inns, N.J. Tuckwell, and T.C. Marrs. The effect of
storage upon cyanide in blood samples. Hum. Exp. Toxicol. 9:
125129 (1990).
F. Moriya and Y. Hashimoto. Potential for error when assessing
blood cyanide concentrations in fire victims. J. Forensic Sci. 46:
14211425 (2001).
K. Narayanaswami, S. Jand, and R.S. Kotangale. Changes in
cyanide as a function of storage time in autopsy material by UV
Spectrometry. Ind. J. Med. Res. 73: 291295 (1981).
Y. Seto. Stability and spontaneous production of blood cyanide
during heating. J. Forensic Sci. 41: 465468 (1996).
F. Chikasue, M. Yashiki, T. Kojima, and T. Miyazaki. Cyanide
distribution in five fatal cyanide poisonings and the effect of
storage conditions on cyanide concentrations in tissues. Forensic
Sci. Int. 38: 173183 (1988).
B. Ballantyne. In vitro production of cyanide in normal human
blood and the influence of thiocyanate and storage temperature.
Clin. Toxicol. 11: 173193 (1977).
J.O. Egekeze and F.W. Oehme. Direct potentiometric method
for the determination of cyanide in biological materials. J. Anal.
Toxicol. 3: 119124 (1979).
P.J. Karhunen, I. Lukkari, and E. Vuori. High cyanide level in a
homicide victim burned after death: Evidence of post-mortem diffusion. Forensic Sci. Int. 49: 179183 (1991).
C.J. Anderson and W.A. Harland. Fire deaths in the Glasgow
area: III, The role of hydrogen cyanide. Med. Sci. Law 22: 3539
(1982).
C.J. Clark, D. Campbell, and W.H. Reid. Blood carboxyhaemoglobin and cyanide levels in fire survivors. Lancet 20:
13321336 (1981).
B.C. Levin, P.R. Rechani, J.L. Gurman, F. Landron, H.M. Clark,
M.F. Yoklavich, J.R. Rodriguez, L. Droz, F. Mattos de Cabrera, and
S. Kaye. Analysis of carboxyhemoglobin and cyanide in blood
from victims of the Dupont Plaza Hotel fire in Puerto Rico.
J. Forensic Sci. 35: 151168 (1990).