DEATH FROM HEROIN OVERDOSE: FINDINGS FROM HAIR ANALYSIS

Summary

Background Morphine analysis of hair is used in forensic toxicology to
study the addiction history of heroin addicts. To clarify the features
underlying fatal heroin intake, we measured hair morphine content in a
group of deceased heroin addicts, to verify a possible correlation between
fatal heroin overdoses and the addiction behaviour of these individuals
before death.

Methods 91 deaths were attributed to heroin overdose in Verona, Italy, in
1993-96. We analysed the hair of 37 of these individuals, and of 37 active
heroin addicts, 37 former heroin users abstinent from the drug for several
months, and 20 individuals with no evidence of exposure to opioids. From
each individual, a hair sample of about 150 mg was analysed by RIA and
high-performance liquid chromatography, to measure the morphine content.

Findings The mean morphine content in the hair of the addicts who had died
was 1·15 ng/mg (SD 2·35 ng/mg; range 0-12·25 ng/mg) compared with 6·07
ng/mg (4·29; 1·15-17·0) in the active heroin addicts, 0·74 ng/mg (0·93;
0·10-3·32) in the abstinent former addicts, and values below the detection
limit in the non-exposed group. Hair morphine content among those who had
died was significantly lower than that in active heroin consumers
(p(0·0001), but not significantly different from that in the former addicts
(p0·978).

Interpretation Although our findings may be subject to selection bias,
since suitable hair samples were available for only 37 of the 91 addicts
who had died, these findings support the theory of high susceptibility to
opioid overdose after periods of intentional or unintentional abstinence,
due to loss of tolerance. Medical staff running detoxification programmes
should be aware of the risk inherent in relapse to heroin after a period of
abstinence. Moreover, occasional heroin use without a build-up of tolerance
could also give a high risk of overdose.

Lancet 1998; 351: 1923-25

Introduction

In Italy, according to official epidemiological data, heroin overdoses
account for about 1000 deaths per year.1 Despite the efforts of forensic
pathologists, clinical pathologists, and toxicologists, the mechanisms by
which heroin overdose leads to death are not yet clear. A major reason for
the lack of clarification is that blood samples taken from people who have
died from heroin overdose show great variation in the amounts of
biologically active metabolites of heroin present. Even in cases of acute
overdose, observed blood concentrations of morphine, the main active
metabolite of heroin, have ranged from 10 ng/mL to 4000 ng/mL.2,3 This
range hampers the definition of a clear threshold of lethal heroin intake.
The range can be partly ascribed to variable survival times after heroin
injection,4-6 which in most cases are unknown, and to the rapid
disappearance of heroin and its active metabolites from the blood
(half-life heroin, 9 min; 6-acetylmorphine, 38 min; morphine, 80 min2).
Thus, even though the blood concentrations fall, morphine bound to
receptors in the central nervous system may lead to death by respiratory
failure.7

Respiratory depression caused by opioids is the main physiological
explanation of fatal heroin overdoses, but other explanations include
metabolic variation in heroin tolerance,8,9 the toxicity of adulterants,10
pharmacological interactions with alcohol,11 and even allergic reaction to
components of heroin preparations.12 However, the relevance of these
factors in explaining the majority of deaths has not been statistically
proved. Moreover, investigation of the mechanisms of fatal heroin overdose
is hindered by gaps in individual case histories. Any information given by
relatives and friends concerning the medical history of the victim is
likely to be unreliable because of the addict's lifestyle and environment.

To address this issue, toxicological analysis of hair can be used in the
retrospective investigation of drug use and addiction. Head hair grows at
approximately 0·8-1·3 cm per month.13 Drugs can be detected in hair tissue
weeks or months after intake. Exogenous compounds are incorporated into
hair tissue at the root. They reach the growing hair matrix from capillary
blood surrounding the hair germination centre, from skin-gland
secretions,14 and, in some cases, from the external environment.15 The low
metabolic activity of the hair shaft, and the protection exerted by the
hair matrix components, contribute to the stability of the embedded
compounds. Although contamination of the hair by drugs present in the
environment,16 by hair bleaching, and by hair dyeing17 may affect the
accumulation of chemicals in the hair matrix, there is consensus about the
usefulness of hair analysis in the study of prevalence of drugs misuse.18

On these grounds, we used hair analysis, in addition to the usual forensic
tests of biological fluids, to study heroin-linked deaths in the province
of Verona, Italy. We aimed to verify a possible correlation between these
deaths, and the drug use of the individual in the months before death.

Methods

From among 91 heroin-related deaths between 1993 and 1996, we selected 37
individuals (29 men, eight women, aged 18-34) for hair analysis (group D).
Criteria for selection were availability of hair, state of decomposition,
lack of contamination of hair (blood, vomit, etc), lack of cosmetic
treatments, availability of devices to sample hair during necropsy, and
collaboration of the necropsy technicians. All the cases underwent our
routine pathological and toxicological analysis. Urine was screened for
opioids, benzoylecgonine (cocaine metabolite), amphetamines,
benzodiazepines, methadone, barbiturates, cannabinoids, and alcohol. Blood
was tested for free (unconjugated) morphine, cocaine, and alcohol.
Threshold positive concentrations of toxic agents in urine were as
suggested by the US Department of Health and Human Services' Division of
Workplace Drug Testing (300 ng/mL morphine or codeine/mL for opioids).

A "positive" control group of 37 heroin addicts was also studied (group
A1). The group consisted of 30 men and seven women, aged 18-32, who had
just entered a detoxification programme at a local medical centre for drug
addictions. Admission to the programme was based on clinical and
toxicological investigations, including a positive test for opioids in
urine. A second control group of 37 former chronic users of heroin, who had
allegedly been abstinent from the drug for several months, was also
investigated (group A2). Members of group A2 had applied for obtaining, or
reobtaining, a driving licence, and according to Italian law (D Les 285,
April 30, 1992) they had to be checked by physical examination and
toxicological screening of their urine, to verify abstinence. They had
undergone serial urine toxicological screening for opioids,
benzoylecgonine, amphetamines, benzodiazepines, methadone, barbiturates,
cannabinoids, and alcohol for about 40 days. Their hair was tested for
morphine and cocaine. The people in group A2 showed negative results in the
urine tests, but still had traces of morphine in their hair, suggesting
either persisting occasional use of heroin, short abstinence from the drug,
or exposure to opioids in the environment. As a "negative" control group,
we chose 20 employees of the Institute of Forensic Medicine, 15 men and
five women aged 27-44, with no evidence of exposure to opioids (group N).

A hair sample of about 150 mg (3-5 cm long) was collected from the vertex
posterior of each individual's head, by cutting with scissors as close to
the scalp as possible. Hair samples were collected with the informed
consent of the people in the control groups, and stored at room temperature
in paper envelopes until analysis. The sample collection and pretreatment
has been described in detail elsewhere.19 Hair specimens were washed with
50 mL ethyl ether and 50 mL 0·01 mol/L hydrochloric acid on a porous glass
filter, then dried, cut with scissors into small fragments, and weighed.
100 mg of hair was incubated overnight in a water bath containing 2 mL 0·25
mol/L hydrochloric acid at 45°C. The incubation mixture was then
neutralised with equimolar amounts of 1 mol/L sodium hydroxide, and twice
extracted into organic phase (18% dichloromethane, 18% dichloroethane, 64%
heptane, by volume) from a carbonate buffer pH 9, by use of a proprietary
liquid-liquid extraction system. After shaking for 15 min, and phase
separation by centrifugation at 500 g for 10 min, the organic phases from
the two extraction steps were collected, pooled, and evaporated under an
air stream.

The dried extract was reconstituted with 1 mL 0·05 mol/L phosphate buffer
pH 5. Qualitative analysis was done with RIA, and quantitative analysis was
done with high-performance liquid chromatography (HPLC). The threshold
positive level of morphine was 0·1 ng per mg of hair. The RIA was done with
a commercial kit (Coat-a-Count, Diagnostic Products Corporation, Los
Angeles, CA, USA). An iodine-125 labelled tracer was used, and apparatus
tubes were coated with antibody highly specific for free morphine, which
gave less than 2% cross-reactivity with morphine glucuronide and other
major opioids.19 All results were then confirmed by HPLC, based on
reverse-phase separation on a polymeric column (PLRP-S 5 µm, Polymer Labs,
Church Stretton, UK), and amperometric detection at a glassy carbon
electrode (+350 mV vs a silver/silver chloride reference), according to a
routine procedure used in our institute.20 Quantitative blood analysis for
morphine and cocaine was carried out by HPLC, with amperometric detection
of morphine and fluorimetric detection of cocaine.21,22 The above methods
have been used for years for routine testing of hair and blood samples.
Since 1990 the Institute of Forensic Medicine has participated in an
inter-laboratory study for the validation of hair analysis techniques,
promoted by the US National Institute of Standards and Technology, MD, USA.

For the statistical analysis of our results we used one-way ANOVA,
Student's t test for unpaired data, and the non-parametric Wilcoxon
signed-rank test. Statistical tests used StatView SE+ Graphics (version 1·03).

Results

All urine samples from group A1 were positive for opioids. The urine
samples from group A2 tested negative for all substances. All members of
group D tested positive for opioids in urine. Benzodiazepines, alcohol,
cannabinoids, amphetamines, and methadone were also found in the urine
samples from several members of group D, but there was no specific pattern.
Group N was not tested by urinalysis, because this information was not
relevant for our purposes, and because toxicological urinalyses of
employees are prohibited by law in Italy.

The mean blood concentration of free morphine in group D was 273·3 ng/mL
(SD 188·63 ng/mL, range 60-894 ng/mL). Blood morphine concentrations in
group D were in all cases above established levels of toxicity.13 All
members of group D were known to the police as heroin addicts. Based on
this information, on the necropsy data, and on the findings at the death
scene, all the deaths were attributed to acute overdose of heroin. In most
cases, recent injection marks were found on the body, suggesting heroin
intake by intravenous injection.

The measurement of morphine in the hair of the deceased (group D) gave a
mean value of 1·15 ng/mg (SD 2·35 ng/mg; range 0-12·25 ng/mg). There was no
morphine in the hair of group N members. The mean morphine content of the
hair of the active heroin addicts of group A1 was 6·07 ng/mg (SD 4·29;
range 1·15-17·00 ng/mg). The incompletely abstinent individuals of group A2
had a mean value of 0·74 ng/mg (SD 0·93; range 0·10-3·32; figure).

Mean hair morphine content (ng/mg)

Horizontal barsmeans. A1active heroin addicts. Ddeceased following
heroin overdose. A2incompletely abstinent individuals.

One-way ANOVA showed a highly significant difference between the mean hair
morphine contents found in groups A1, A2, and D (p(0·0001). The statistical
comparison of the group data found that hair morphine contents were
significantly lower in group D than in the active heroin consumers of group
A1 (Wilcoxon test p(0·0001). However, hair morphine content in group D did
not differ significantly from those of the incompletely abstinent subjects
of group A2 (Wilcoxon test p0·978).

As expected, the active addicts of group A1 showed higher mean morphine
content in hair than the incompletely abstinent subjects of group A2
(p(0·0001). By the least square method, no linear (R20·017), exponential
(R20·007), or logarithmic (R20·170) correlation was found between the
amounts of morphine in the hair and in the blood of the people who died
from heroin overdose.

Discussion

The link between drug use and drug accumulation in hair has been examined
in several earlier reports.23-26 To our knowledge, however, hair analysis
has not been used to investigate the recent addiction histories of people
who have died from heroin overdose. Our study was limited to the province
of Verona, and may have been subject to selection bias since suitable hair
samples were available for only 37 of the 91 addicts who died. However, we
have shown that most fatal heroin overdoses occurred in heroin users with a
much lower hair morphine content than that found in the hair of active,
chronic consumers of the narcotic.

On the assumption of a rough positive correlation between mean heroin
intake and morphine concentrations in hair, and a hair growth rate of 1
cm/month, our findings suggest that most individuals who died from heroin
overdose had virtually abstained from heroin during the 4 months preceding
death. Thus, the results of this hair analysis support a theory of high
susceptibility to opioid overdose after periods of intentional or
unintentional abstinence. This theory has been used to explain the high
number of deaths among addicts recently released from jail or on completion
of a detoxification programme.9,27 The reasons for increased susceptibility
to overdose remain unclear, but it is likely that a lower heroin tolerance
after a period of abstinence, or a low tolerance owing to light or
irregular heroin use, leads to a corresponding decrease in the size of a
fatal dose.

The difference in hair opioid content between groups A1 and A2 in our study
supports the continued use of hair testing in forensic analysis in cases of
heroin overdose. The results of our study should indicate to the medical
staff of detoxification programmes that there are risks inherent in relapse
to heroin intake following abstinence from the drug. In particular, we
point out the potential risk of "opioid free" detoxification programmes for
individuals at risk of relapse. Moreover, occasional or recreational heroin
use (eg, at weekends), an increasing heroin addiction pattern that is not
characterised by dependence and tolerance, could lead to more cases of
heroin overdose than is generally thought.

Contributors

Franco Tagliaro was mainly responsible for writing the article, and for
developing, and validating the analytical methods. Zeno De Battisti
coordinated medical and pathological investigations, sample collection, and
analysis. Frederick P Smith contributed to study design, and to statistical
data analysis. Mario Marigo discussed, revised, and approved all aspects of
the study, and the resulting paper. All researchers contributed to writing
the paper.

Acknowledgments

We thank Carla Neri for routine urinalyses and Giovanna Carli for help.

References

1 Data from the Annual Report of the "Direzione Centrale per i Servizi
Antidroga", Rome: Italian Ministry of Internal Affairs, 1996.

2 Baselt RC, Cravey RH. Disposition of toxic drugs and chemicals in man.
Foster City, CA, USA: Chemical Toxicology Institute, 1995.

3 Uges DRA. Therapeutic and toxic drug concentrations. TIAFT Bull 1996; 26:
5-34.

4 Spiehler V, Brown R. Unconjugated morphine in blood by radioimmunoassay
and gas chromatography/mass spectrometry. J Forensic Sci 1987; 32: 906-16.

5 Lora-Tamayo C, Tena T, Tena G. Concentrations of free and conjugated
morphine in blood in twenty cases of heroin-related deaths. J Chromatrogr
1987; 422: 267-73.

6 Staub C, Jeanmonod R, Frye O. Morphine in postmortem blood: its
importance for the diagnosis of deaths associated with opiate addiction.
Int J Legal Med 1990; 104: 39-42.

7 Addington WW, Cugell DW, Bazley ES, Westerhoff TR, Shapiro B, Smith RT.
The pulmonary edema of heroin toxicity--an example of the stiff lung
syndrome. Chest 1972; 62: 199-205.

8 Richards RG, Reed D, Cravey RH. Death from intravenously administered
narcotics: a study of 114 cases. J Forensic Sci 1876; 21: 467-82.

9 Harding-Pink D, Frye O. Risk of death after release from prison: a duty
to warn. BMJ 1988; 297: 596.

10 Questel F, Dugarin J, Dally S. Thallium-contaminated heroin. Ann Intern
Med 1996; 15: 616.

11 Levine B, Green D, Smialek JE. The role of ethanol in heroin deaths. J
Forensic Sci 1995; 40: 808-10.

12 Brashear RE, Kelly MT, White AC. Elevated plasma histamine after heroin
and morphine. J Lab Clin Med 1974; 83: 451-57.

13 Sachs H. Theoretical limits of the evaluation of drug concentrations in
hair due to irregular hair growth. Forensic Sci Int 1995; 70: 53-61.

14 Henderson GL. Mechanism of drug incorporation into hair. Forensic Sci
Int 1993; 63: 19-29.

15 Smith FP, Kidwell DA. Cocaine in hair, saliva, skin swabs, and urine of
cocaine users' children. Forensic Sci Int 1997; 83: 179-89.

16 Cone EJ, Yousefnejad D, Darwin WD, Maguire T. Testing human hair for
drugs of abuse--II: identification of unique cocaine metabolites in hair of
drug abusers and evaluation of decontamination procedures. J Anal Toxicol
1991; 15: 250-55.

17 Skopp G, Potsch L, Moeller MR. On cosmetically treated hair--aspects and
pitfalls of interpretation. Forensic Sci Int 1997; 84: 43-52.

18 US General Accounting Office. Drug use measurement: strengths,
limitations, and recommendations for improvement. Report to the Chairman,
Committee on Government Operations, US House of Representatives.
Washington, DC: US Government Printing Office, June, 1993.

19 Marigo M, Tagliaro F, Poiesi C, Lafisca S, Neri C. Determination of
morphine in the hair of heroin addicts by high performance liquid
chromatography with fluorimetric detection. J Anal Toxicol 1986; 10: 158-61.

20 Tagliaro D, De Battisti Z, Lubli G, Neri C, Manetto G, Marigo M.
Integrated use of hair analysis to investigate physical fitness to obtain a
driving licence: a casework study. Forensic Sci Int 1997; 84: 129-35.

21 Tagliaro F, Carli G, Cristofori F, Campagnari G, Marigo M. HPLC
determination of morphine with amperometric detection at low potentials
under basic pH conditions. Chromatographia 1988; 26: 163-67.

22 Tagliaro F, Antonioli C, De Battisti Z, Ghielmi S, Marigo M.
Reversed-phase determination of cocaine in plasma and human hair with
direct fluorimetric detection. J Chromatogr 1994; 674: 207-15.

23 Püschel K, Thomasch P, Arnold W. Opiate levels in hair. Forensic Sci Int
1983; 21: 181-86.

24 Miyazawara N, Uematsu T, Mizuno A, Nagashima S, Nakashima M. Ofloxacin
in human hair determined by higher performance liquid chromatography.
Forensic Sci Int 1991; 51: 65-77.

25 Uematsu T, Sato R, Suzuki K, Yamaguchi S, Nakashima M. Human scalp hair
as evidence of individual dosage history of haloperidol: method and
retrospective study. Eur J Clin Pharmacol 1989; 37: 239-44.

26 Kintz P, Mangin P. Hair analysis for detection of beta-blockers in
hypertensive patients. Eur J Clin Pharmacol 1992; 42: 351-52.

27 Püschel K, Teschke F, Castrup U. Aetiology of accidental/unexpected
overdose in drug-induced deaths. Forensic Sci Int 1993; 62: 129-34.

Checked-by: Richard Lake