Carfentanil Blood Concentrations in Impaired Driving Investigations in Ontario, Canada

Abstract

Carfentanil is a potent opioid with no medical use in humans; it presents a serious threat to public health and road safety due to its presence in the illicit drug supply, the potency of the drug and instances of use prior to the operation of a motor vehicle. The identification and quantitation of carfentanil using liquid chromatography with tandem mass spectrometry was performed in blood samples obtained through impaired driving-related investigations from 2017 to 2019. In a series of 66 cases submitted to the Centre of Forensic Sciences in Ontario, Canada, blood concentrations of carfentanil ranged from <0.04 to 2.1 ng/mL in the population studied. Driving behaviors frequently came to the attention of concerned citizens, decreased levels of consciousness were commonly reported and a variety of medical interventions were required, in some cases, to preserve life due to apparent opioid toxicity.

Introduction

In 2008, the Drug Evaluation and Classification Program was included in the Criminal Code of Canada to assist police officers with the investigation of suspected drug-impaired drivers. At the conclusion of the evaluation, if the drug recognition expert (DRE) determines that impairment is present and due to a drug, a bodily fluid sample is obtained from the individual. In the province of Ontario, this sample is submitted to the Centre of Forensic Sciences (CFS) for toxicological analysis. In the years following the implementation of this legislation, the number of drug-impaired driving cases submitted to the CFS has increased from ∼300 cases per year from 2009 to 2015 to >2,000 cases in 2020. The recent upward trend of this case type is likely due to several factors including an increased number of police officers trained in standardized field sobriety testing and as DREs, and the pervasiveness and increase in recreational use of opioids and other drugs. The rise in drug-impaired driving investigations may also have been impacted by changes to the Criminal Code of Canada impaired driving laws in 2018. These changes enabled police officers to legally demand blood samples, in some cases, without the need for a drug evaluation; additionally, the amendments included blood drug per se limits and zero tolerance for any detectable level of certain drugs (1). While the detection of carfentanil can form part of an impaired driving investigation, this drug was not included in the amendments and is, therefore, not currently subject to blood drug per se limits or zero-tolerance restrictions.

Carfentanil is an opioid, which, based on an animal model (i.e., Wister rats), is estimated to have an analgesic potency of ∼10 000 times that of morphine (2). Carfentanil is used in veterinary medicine to incapacitate wildlife, notably in large animals, for examination and procedures. Carfentanil is not approved for therapeutic use in humans; however, in the last few years, it has become an illicit, recreational drug. Carfentanil was first detected in CFS case submissions in 2017 in both post-mortem and ante-mortem case types. However, it is likely that carfentanil cases in Ontario pre-dated 2017 when improvements were made to our detection capabilities for this drug (3). Furthermore, in Health Canada analysis of seized drug exhibits, carfentanil was observed as early as 2016 (4).

Reports of the effects of carfentanil on humans are limited. The first report of human exposure involved a veterinarian accidentally exposed to carfentanil while attempting to sedate an elk (5). Liquid from a tranquilizer dart splashed in his face, eyes and mouth and, despite immediate decontamination within two min, the individual developed drowsiness. Naltrexone was administered parenterally, and his symptoms resolved. Another postulated exposure originated from aerosolized carfentanil and remifentanil used during an attempt to resolve a hostage-taking; 125 hostages died because of the exposure (6). Symptomology included rapid loss of consciousness, inhibited tendon, pupil and corneal reflexes, respiratory depression and cyanosis; in less impacted individuals, observations included disorientation, vomiting, pinpoint pupils, bradycardia and hypotension. Loss of pain sensation was also reported.

Carfentanil blood concentrations arising from recreational use can often be associated with toxicity and overlap with concentrations attributed to fatal intoxication. In a case report of recreational use of carfentanil where the patient was found unconscious and taken to the hospital, the serum concentration was 0.6 ng/mL (7); other symptoms included hypotension, tachycardia, hypopnea and cyanosis. Treatment with naloxone enabled the individual to regain consciousness. Carfentanil blood concentrations ranged from 0.0115 to 0.125 ng/mL in seven impaired driving investigations (8); naloxone was administered in three of these cases. Additionally, in two cases of drug-impaired driving, carfentanil blood concentrations of 0.41 and 0.63 ng/mL were detected in combination with other drugs (9); naloxone was administered in one of these cases. Blood carfentanil concentrations ranged from 0.11 to 0.47 ng/mL in a subset of 12 driving under the influence of drugs cases, with mean and median concentrations of 0.23 and 0.19 ng/mL, respectively (10). In comparison, postmortem blood concentrations ranged from 0.0233 to 0.529 ng/mL in seven deaths attributed to carfentanil (11). In another study, femoral blood concentrations of <0.1 and 0.84 ng/mL were detected in two deaths attributed to carfentanil intoxication, whereas the cause of death was smoke inhalation and drowning in two individuals with femoral blood carfentanil concentrations of 0.90 and 0.14 ng/mL, respectively (3). The overlap between carfentanil concentrations associated with recreational use and fatal toxicity is consistent with other opioid drugs (12–14).

In efforts to expand on the available literature, carfentanil blood concentrations, demographics, co-occurring drugs, symptomology and driving behavior are explored. This study is a retrospective compilation of drug-impaired driving cases where a blood sample was available for analysis and carfentanil was detected.

Methods

The selected cases include drug-impaired driving submissions to the CFS for the period of 2017 to 2019 for which a blood sample was available for a toxicological analysis. In conjunction with each case, a synopsis of the events, observations of the driver and, in some cases, an evaluation by a DRE were included. The analytical techniques employed by the CFS are capable of identifying and quantitating a large variety of over-the-counter, illicit, recreational and prescription drugs. The extent of testing performed on the blood samples submitted was determined by reviewing the synopsis provided and the requests of the submitter, in addition to considerations regarding the volume of the sample available for analysis. Where sample volume permitted, drug screening was performed by liquid chromatography quadrupole time-of-flight mass spectrometry (LC–QTOF-MS; N = 63); in all cases (N = 66), an analytical procedure capable of identification and quantitation of 54 commonly occurring drugs and metabolites was performed by liquid chromatography with tandem mass spectrometry (LC–MS-MS). In some cases, additional screening was performed using immunoassay, gas chromatography–mass spectrometry, quantitative analysis using a targeted LC–MS-MS procedure and/or headspace gas chromatography with flame ionization detection for ethanol. Ethanol analysis was not performed on all cases; ethanol use may not have been suspected during the police investigation, it may have been excluded by testing performed elsewhere or limited sample volume may have precluded analysis.

The identification and quantitation of carfentanil was performed using a multidrug targeted LC–MS-MS method previously described (3). In 2017, the laboratory switched from relying on the LC–QTOF-MS for the detection of carfentanil, which had a limit of detection (LOD) of ∼0.5 ng/mL to using a targeted method with an LOD of 0.05 ng/mL. In April 2019, the LOD and the lower limit of quantitation (LOQ) for carfentanil were reduced from 0.05 to 0.018 ng/mL and 0.10 to 0.0375 ng/mL, respectively. The detection of carfentanil above the LOD but less than the LOQ will be referred to as <0.1 ng/mL prior to the change in methodology and <0.04 ng/mL after the improvement in sensitivity. Where possible, quantitation was performed in duplicate, and the average of the results is reported.

Results

A total of 66 impaired driving cases were identified from 2017 to 2019 in which carfentanil was detected in a blood sample. The subjects ranged from 19 to 78 years old (mean 37, median 36) and were predominantly male (84.9%). Table I outlines the time of the incident and the time of year (both divided into four broad categories) and the reason for police involvement which included collisions, concern for the driving behavior and/or the well-being of the individual. The driving behavior was predominantly described as weaving, swerving or erratic; concern for the person included decreased levels of consciousness described as “in and out” of consciousness, passed out, sleeping, slumped over the wheel, a belief that the person could be overdosing and/or that a medical event was occurring. As the investigations unfolded, drug intoxication and drug use were frequently identified through symptomology, presence of powders and/or liquids and paraphernalia including a tourniquet, syringes or smoking apparatus. The powders and liquids present in the vehicles were a variety of colors such as purple, white, blue, brown and yellow; the analysis of drug exhibits is not within the scope of testing of the CFS (this service is provided by Health Canada). Subjects admitted to the administration of methamphetamine, cocaine, fentanyl and heroin. Some individuals required medical interventions such as cardiopulmonary resuscitation (CPR), oxygen, naloxone, intubation and additional medical care. There were some individuals who were apneic, expelling frothy fluid from their airways and/or had a blue tinge to their nails and required immediate assistance. Naloxone was administered multiple times in some cases to achieve a sufficient opioid antagonistic effect.

Table I.

Summary of Incident Details for Carfentanil-Involved Drug-Impaired Driving Investigations (N = 66) in Ontario from 2017 to 2019

Time of daya	Midnight–6 a.m.	6 a.m. to noon	Noon to 6 p.m.	6 p.m. to midnight
	6 (9.2%)	11 (16.9%)	29 (44.6%)	19 (29.2%)
Time of year	January–March (27.3%)	April–June (53.0%)	July–September (16.7%)	October–December (3.0%)
2017 (6.1%)	0	1	1	2
2018 (3.0%)	1	0	1	0
2019 (90.9%)	17	34	9	0
Reason for police involvement	Collision 35 (53.0%)	Enforcementb	Health concernc
	Single 21 (31.8%)	16 (24.2%)	15 (22.7%)
	Multi-vehicle 10 (15.2%)
	Not specified 4 (6.1%)
Further observations	Intoxication/drug used	Medical interventionse	Driver distressf
	56 (84.8%)	38 (57.6%)	51 (77.3%)

^aTime not provided in a single case.

^bCitizen complaint regarding a possible impaired driver, random sobriety check and traffic violations.

^cDecreased levels of consciousness, suspected drug overdose or medical intervention required.

^dObservations included the presence of pipes, syringes, tourniquet, powders and/or liquids.

^eNaloxone, CPR, oxygen, intubation and/or further treatment at the hospital.

^fDecreased levels of consciousness, respiratory impairment and/or suspected drug overdose.

Carfentanil-involved cases

A summary of the blood carfentanil concentrations detected is presented in Table II. The first detection of carfentanil in a blood sample from an impaired driving investigation occurred in 2017. Both 2017 (4 cases) and 2018 (2 cases) had very few cases, but in 2019 (60 cases) carfentanil involvement increased. There were 3 and 12 cases, respectively, where the carfentanil was detected below the 0.1 and 0.04 ng/mL LOQs; these cases were excluded from the mean and median, as provided in Table II. Selected cases are further explored in Table III due to findings of particular significance, specifically, the first blood carfentanil identification in a driver (Case 1), the highest carfentanil concentration (Case 2), carfentanil-only cases where other drugs including ethanol were not detected (Cases 3 and 4) and cases with minimal other drug involvement (Cases 5–9). All cases in Table III were analyzed using the LC–QTOF-MS screening procedure and the targeted LC–MS-MS method (3); ethanol and cannabinoid analyses were performed unless otherwise indicated.

Table II.

Blood Carfentanil Concentrations (ng/mL) in Drug-Impaired Driving Investigations in Ontario from 2017 to 2019

Year	Number	Minimum	Maximum	Mediana	Meana
2017	4	0.13	0.31	0.28	0.25
2018	2	<0.1b	0.13
2019	60	<0.04b	2.1	0.16	0.25
All years	66	<0.04	2.1	0.16	0.25
All yearsc	65	<0.04	0.62	0.16	0.21

^aData less than the lower LOQ excluded from the calculation.

^bLower LOQ change in 2019.

^cExclusion of an unusually high concentration (2.1 ng/mL).

Table III.

Selected Cases of Interest

Case (sex, age)	Toxicological findings	Concentration (ng/mL)	Summary of occurrence
1 (male, 24)a	Carfentanil Methadone Fentanyl Acetaminophen	0.13 105 <1.3 Qualitative	Δtb 2 h and 55 min Speeding, collision Syringe and tourniquet in vehicle Constricted pupils, droopy eyelids, elongated speech pattern, fresh and old puncture marks; no pain or discomfort despite injuries On methadone program and admitted to injecting fentanyl cDrug evaluation categories: central nervous system depressant and narcotic analgesic
2 (female, 22)a,d	Carfentanil Cocaine Benzoylecgonine Phenacetin Acetaminophen	2.1 49 1,215 Qualitative Qualitative	Δt 4 h and 45 min Erratic driving, weaving between lanes and into oncoming traffic Crack cocaine, pipe and pills located in vehicle Falling asleep, slurring, coughing, jittery movements, confused, “out of it” Complaining of breathing difficulties and that she may have a seizure and transported to hospital
3 (male, 33)	Carfentanil Ethanol	0.056 Not detected	Δt 3 h and 50 min Drove vehicle into oncoming lane, hit ditch on the opposite side of the road, crossed entire roadway into another ditch and a stream Breathing but verbally unresponsive, gazing ahead, constricted pupils, in and out of consciousness, vomited and transported to hospital Droopy eyelids, raspy speech, pupils 1.5 mm in all three light conditions Admitted to smoking “a couple of points” of fentanyl cDrug evaluation categories: central nervous system depressant and narcotic analgesic Could not provide a urine sample
4 (male, 28)	Carfentanil Ethanol	0.41 Not detected	Δt 50 min Mounted sidewalk, causing extensive damage to trees and his vehicle Transported to hospital exhibiting opioid toxicity including pinpoint pupils, decreased respiration and oxygen saturation and level of consciousness, treated with intravenous naloxone and recovered, second naloxone dose was required ∼80 min later
5 (male, 25)d	Carfentanil Fentanyl Ethanol	0.31 <1.3 Not detected	Δt not provided Vehicle running, driver’s foot on brake, on an entrance ramp Syringe and tourniquet in vehicle Slumped over wheel, unresponsive to all forms of stimuli, without vital signs, CPR started, regained consciousness, sweating profusely and transported to hospital Fresh puncture marks, continually scratching and vomiting, difficulty staying awake Admitted to injecting heroin
6 (male, 55)a	Carfentanil Fentanyl Benzoylecgonine	0.13 <1.3 31	Δt 47 min Slumped over wheel with vehicle running Syringe in vehicle Breathing but making gurgling noises, unconscious, brought to hospital
7 (male, 28)a	Carfentanil Benzoylecgonine	0.056 21	Δt 2 h and 34 min Unconscious in vehicle in an intersection Drug residue and paraphernalia in vehicle In and out of consciousness (“on the nod”), pinpoint pupils, eyes rolling backward, slurred speech, transported to hospital
8 (male, 29)a	Carfentanil Tetrahydrocannabinol Carboxytetrahydrocannabinol Acetaminophen	0.46 1.9 Qualitative Qualitative	Δt 1 h and 25 min Speeding, ran stop sign, left roadway, crashed into stop sign and a parked car in the driveway of a house Syringe in vehicle; liquid in syringe blue in color and identified by a Health Canada laboratory as carfentanil (plus caffeine and dimethyl sulfone) Unconscious with foot on the gas pedal and tires spinning, transported to hospital
9 (male, 38)	Carfentanil Nordiazepam Ethanol	0.18 <50 Not detectede	Δt 1 h and 5 min Vehicle left roadway after failing to navigate a ramp Sleepy, confused, shaky In the police cruiser, eyes rolled back, fell asleep, snoring loudly, unresponsive to verbal commands and transported to hospital

^aEthanol analysis not performed.

^bΔt = time elapsed between incident and blood collection.

^cDrug evaluation performed.

^dCannabinoid analysis not performed.

^eAnalysis performed by the hospital laboratory.

Many of the cases (97.0%) involving carfentanil also had multiple other drugs/metabolites detected (additional compounds detected ranged up to 10 drugs/metabolites with a mean and median of 5). Common co-occurring drugs/metabolites were benzoylecgonine (47.0%), fentanyl (42.4%), methamphetamine (40.9%), amphetamine (33.3%), methadone (27.3%) and cocaine (24.2%). Supplementary Table S1 contains concentration ranges for the drugs/metabolites encountered in the carfentanil-involved cases; compounds without quantitative results are also listed in the descending order of the number of cases and percent of cases where the drug/metabolite was identified. Naloxone detection in the blood samples is likely an underestimation. There were cases where the history indicated naloxone was administered and it was not identified in the blood sample; this is likely due to the LC–QTOF-MS sensitivity for naloxone or time of administration in relation to the time of sample collection. Additionally, the reported naloxone use is likely an underestimation since not all case histories were comprehensive and medical intervention, generally, was a common requirement.

Discussion

Carfentanil was first detected in illicit drug seizures in the European Union in 2012 (15) and in Canada in 2016 (4). In Ontario, near the end of 2016, it seemed that the CFS may be missing detection of a potent opioid being implicated in casework. The screening technique (LC–QTOF-MS) used at the CFS could detect carfentanil; however, due to the potency of the drug and resulting low blood concentrations, the method was unlikely to be sufficiently sensitive. A 2021 publication by Solbeck et al. outlines the response of the CFS to this issue by utilizing an existing method to improve detection capabilities (3).

In the first report of human exposure (5), George et al. mused about how the future misuse or abuse of carfentanil remained to be seen, following up with predictions that “Intentional oral or parenteral exposures to this highly potent veterinary narcotic could be expected to cause a severe intoxication or fatality without prompt treatment.” The potency of carfentanil, its intermittent presence in the illicit drug supply and the prevalence of recreational opioid use highlight the health hazards of this opioid. In the current study, it was often reported (77.3% of cases) that individuals had decreased levels of consciousness, respiratory impairment or were suspected of overdosing while operating a motor vehicle, after the vehicle had come to a stop or while being in the driver’s seat. Many of these cases could have become fatalities but for the actions of concerned citizens, involvement of first responders and medical intervention prior to or at the hospital.

In 2019, Ontario saw a significant increase in the detection of carfentanil in opioid-related deaths (16), in the illicit drug supply (4) and in drug-impaired driving cases. The 2019 data from this study suggest that the presence of carfentanil in drug-impaired drivers was predominantly in the first half of the year and decreased by the end of the year (October–December). Similarly, this declining trend is also reflected in opioid-related deaths in Ontario where, in 2020, carfentanil was detected in 11 cases compared to a peak of 488 in 2019 (16) and in the illicit drug supply in Ontario when it was detected in 23 submissions in 2020 compared to 1,019 submissions in 2019 (4). A similar trend was also noted in the USA; however, it appears that their peak occurred in 2017 (17). Predominantly daytime (6 a.m. to 6 p.m.; 61.5%) compared to nighttime (6 p.m. to 6 a.m.; 38.5%) occurrences in this series of cases were similar to another carfentanil impaired driving publication where daytime and nighttime incidents were 56.1% and 43.9%, respectively (8). The common co-occurring drugs such as fentanyl (42.4%), cocaine (24.2%) and benzoylecgonine (47.0%) were also detected in that series of case reports with 49%, 22% and 32% positivity respectively (8); however, in comparison, methadone (27.3%), methamphetamine (40.9%) and amphetamine (33.3%) were more prevalent in this study.

The blood carfentanil concentrations detected in this study are comparable with other literature involving drug-impaired drivers (8–10, 18). The highest blood concentration of 2.1 ng/mL reported in Case 2 can be compared to concentrations in two drivers of 1.3 and 2.7 ng/mL, which were considered outliers and excluded from statistical analysis (10), as well as the detection of 1.2, 1.5 and 8.2 ng/mL in three impaired drivers (18). For context, a fatal carfentanil concentration has been reported as low as 0.0233 ng/mL (11). Elevated concentrations of opioids in some individuals are not novel and, typically, are attributed to tolerance. In addition to tolerance, Sofalvi et al. considered the possibility that receptor polymorphism, variation in the number of receptors and other drug interactions could potentially mitigate the severity of carfentanil toxicity (10). Mata and Coleman highlighted a case where a co-occurring central nervous system stimulant (namely, methamphetamine) could have assisted with the ability to function in an individual with a fentanyl concentration of 303 ng/mL (14). In Case 2, both cocaine and benzoylecgonine were present and, given a delay of 4 h and 45 min from incident to sample collection, the short half-life and in vitro degradation of cocaine, it is likely that the cocaine concentration was significantly higher at the time of the incident compared to the time of the analysis; this individual was taken to the hospital, but the nature of the medical treatment was not part of the case history submitted.

Correlating a specific drug with the observed effects is often problematic due to the co-occurrence of psychoactive compounds. Polypharmacy was a typical finding in this study with the detection of at least one other drug or metabolite in 97.0% of cases; the most common combinations involved opioids (fentanyl and methadone) and stimulants (cocaine, benzoylecgonine, methamphetamine and amphetamine). However, there were some examples where carfentanil was the only drug detected (Cases 3 and 4) or was the most toxicologically relevant finding, in that, contributions from any other drugs/metabolites identified were unlikely to have been significant (Cases 5–9). In Cases 3 and 4, opioid effects were obvious including pupillary constriction, decreased consciousness and, in the latter case, decreased respiration reversed by naloxone; other opioid-related observations in Case 3 were vomiting, raspy voice, droopy eyelids and urinary retention. In a report of 61 impaired driving cases involving carfentanil, only 4 involved carfentanil alone (8). In the single carfentanil-only case that included a quantitative analysis, the concentration was 0.0501 ng/mL; the individual had rear ended a car and regained consciousness after naloxone administration (8). A decreased ability to handle the complexities of operating a motor vehicle was clearly demonstrated in the cases where the sole or predominant factor was drug impairment by carfentanil.

The analytical approach to drug-impaired driving cases in this study is a potential limitation in some of the cases presented. The general drug screen by LC–QTOF-MS was not performed in three cases; in 10 cases, there was no immunoassay or targeted LC–MS-MS analysis to detect cannabis use and in 32 cases, analysis for ethanol (or other volatiles) was not performed. This may underestimate the prevalence assessment of co-occurring drugs and the potential contribution to observations of the individual’s physical state and driving capabilities (for more information on which drugs/metabolites were impacted, see Supplementary Table S1). It is not unusual to receive small sample volumes (<2 mL) in relation to these cases, particularly when blood samples are seized from a hospital under a police warrant, as opposed to collected by a medical practitioner specifically for the purpose of forensic toxicology testing. Therefore, sample volume may have precluded additional analysis in many of these cases. In others, there may have been other evidence (e.g., results from breath alcohol testing or blood/urine hospital analysis) that influenced the decision regarding additional testing.

The current study adds to the scientific literature by describing symptoms and driving behaviors in carfentanil-involved drug-impaired driving investigations. Samples of blood were analyzed, and carfentanil concentrations ranging from <0.04 to 2.1 ng/mL were detected. While the carfentanil blood concentrations in this study could be considered as examples of tolerance to opioids, signs of opioid toxicity were present in many of the cases, with some requiring immediate medical intervention. The observations included pinpoint pupils, lack of response to verbal or physical stimuli, decreased levels of consciousness, impairment of respiratory function and, in some cases, hypoxia and absence of vital signs. This series of cases provides evidence of the serious implications of recreational carfentanil use including the detrimental impact on the operation of a motor vehicle.

Supplementary data

Supplementary data are available at Journal of Analytical Toxicology online.

Funding

There was no funding for this research or publication.

Data Availability

The data underlying this article are available in the article and in its online supplementary material.