What killed George Floyd: Part 9 – Fentanyl and methamphetamine

The autopsy report indicated that there was a significant load of fentanyl in Mr. Floyd as well as a relatively small amount of methamphetamine.

Fentanyl and sudden death:

The classic story of an opioid/opiate death due to central nervous system suppression is that of someone who takes the drug and becomes increasingly drowsy and obtunded. The victim then becomes unconscious and his or her breathing becomes increasingly labored, often with snoring.  The breathing becomes irregular, and then finally stops, and the decedent dies.  I’ve seen videos of people who have died this way, either under surveillance or by suicide.  It takes awhile.

Fentanyl is a little different.  Often the death is sudden, seconds to minutes after taking the drug.  Here’s an investigative narrative from one of my medicolegal death investigators regarding a case.  The details are changed a little so it’s not a “real” case — but it’s as close as I can make it to one of my cases without breaking the rules:

The decedent was a 40-year old male who was attending a party at the home of a friend. The friends observed him go to a bedroom alone. A short time later the decedent came out of the bathroom, turned blue in color and suffered a collapse. The friends called 911.

This is a typical story in my jurisdiction.  The decedent will be fine, will take some drugs, and then a few minutes later just collapse.  No long, drawn out respiratory collapse.  Don’t get me wrong, there’s plenty of that kind, too.  But it doesn’t have to happen that way, and frequently does not.  As one CDC Morbidity and Mortality Weekly Report noted in fentanyl overdoses(1):  “Although some persons sought out fentanyl and others attempted to avoid it, a majority of respondents reported that opioid-seeking behaviors were not altered in response to the emergence of fentanyl. A majority of respondents who witnessed a suspected fentanyl overdose (75%) described symptoms as occurring rapidly, within seconds to minutes. “

Because of this association of fentanyl with sudden death, it received a particular street name in Chicago in the early 2000s(2): The fentanyl was clandestinely produced in a lab in Toluca, Mexico and sold by the Mickey Cobra street gang. The term “Drop Dead” was coined for this “tainted heroin.”  

There are a number of possible mechanisms for this kind of rapid death in opiate/opoid death in general.  One is sudden respiratory failure due to pulmonary edema. Heroin, in particular is associated with rapid respiratory failure associated with fulminant pulmonary edema.  The mechanism is not completely clear.  Heroin increases pulmonary histamine release, which increases vascular permeability(3). While this has been observed primarily in heroin, it has also been reported in fentanyl intoxications(4). Heroin use has also been associated with rapid development of acute respiratory distress syndrome with diffuse alveolar damage.

While pulmonary edema is common in most drugs deaths, at autopsy I don’t tend to see the florid edema that I associate with fatal drug-induced pulmonary edema. Typically in these deaths associated with severe pulmonary edema, there is foam in the mouth and airways at the time of death, and the lungs are very heavy and full of fluid.  What I tend to see is a lesser degree of pulmonary edema, though there is almost always some. There’s probably more going on with fentanyl.

One finding in fentanyl-related deaths that is is not seen often in other opiate deaths is so-called “wooden chest syndrome.”  This has been described repeatedy in the context of anesthesia involving fentanyl, but has also been described in the context of abuse.  This is a syndrome where there is diffuse muscular rigidity that includes the chest wall, resulting in the immediate inability to breathe.  This rigidty can be so severe that it makes mechanical ventilation impossible(5). Most articles suggest that this is mediated through the central nervous system. This is an unlikely mechanism in the death of Mr. Floyd specifically because he did not display this diffuse rigidity

The search for the molecular mechanism of sudden death in fentanyl is ongoing.  Some research suggests that fentanyl binds to a receptor that regulates ionic currents.  One study found that fentanyl affects the potassium ion channels in the heart (the hERG channel). Alterations of this channel are associated with changes in the QTc interval and sudden death.  This study notes(6): Fentanyl blocks hERG and prolongs the cardiac AP. The post-mortem blood concentration in fentanyl-overdose related death ranges from 9 -1140 nM. While our data are not sufficient to conclude that hERG block is exclusively responsible for fentanyl-related death, our data do provide a mechanism for fentanyl-induced death, especially for individuals whose cardiac electrophysiological functions are already compromised.

Another study by Tschirhar and Zhang noted that this blockade is exacerbated by other stresses such as hypoxia, hypokalemia, alkalosis, and variant hERG alleles.

Thus, in addition to the traditional respiratory depression, fentanyl can result in direct cardiac death — and this is potentiated by methamphetamine.  Did Mr. Floyd die of a fentanyl overdose?  No.  Did fentanyl have an effect on the heart that made it more vulnerable?  Almost certainly. People who claim that fentanyl did not contribute to death because a decedent did not display typical lengthy respiratory distress signs are missing perhaps the most common mechanism of death.


Methamphetamine is another drug that can induce arrhythmia.  Methamphetamine is called a “sympathomimetic” drug because it mimics hormones (like adrenalin) that stimulate the sympathetic nervous system (the “Flight or Fight” system) associated with excitation, arousal, increased blood pressure, and such.  In traditional overdose cases it results in seizures, but it is also has direct cardiac effects.  Because of these cardiac effects there is no real safe-all-the-time dosage.  It can kill at any dose if the individual is sufficiently vulnerable — or it may take a lot if the person is not vulnerable.

Again, there are multiple mechanisms whereby methamphetamine can cause death.  Because it is a sympathomimetic drug, it can cause death by causing fatal serotonin effects — often mediated by hyperthermia (high body temperature).  Hyperthermia can result in cell death, particularly of muscle cells (“rhabdomyolysis”) which in turn is toxic to a number of organs, including the kidneys.  This in turn can lead to multiorgan failure, disorders of coagulation, and death.   One mechanism that has been proposed is that methamphetamine-induced hyperthermia results in temperature-dependent leakage of small vessels in the brain, resulting in fulminant cerebral edema (which I see a lot in these cases). (7)

Methamphetamine use can result in spikes in blood pressure, resulting in any number of hypertensive related catastrophies, from intracranial hemorrhage to hypertensive arrhythmia.  Methamphetamine can cause vasoconstriction, resulting in ischemia of organs including the bowel, brain, or heart, resulting in bowel necrosis, stroke, and sudden cardiac death, particularly in someone who has pre-existing coronary artery disease.

Of importance, methamphetamine is directly cardiotoxic.  It directly induces cell death of myocardial cells (8).  Methamphetamine cardioxicity is directly related to both acute cardiac death and progressive long term heart damage and failure (9,10). Methamphetamine can cause stress cardiomyopathy(11):

The cardiac complications of methamphetamines are hypothesized to arise from a variety of mechanisms. In animal studies, administration of methamphetamine to rats causes hypertension, tachycardia and myocardial toxicity with cellular death, fibrosis and contraction band necrosis. In human autopsy specimens, severe interstitial fibrosis and scar formation has been documented. Physiological pathways implicated include catecholamine surges acutely during methamphetamine intoxication with contingent hypertensive crises, longer-term upregulation of the sympathetic axis, and myocardial toxicity with impaired cellular metabolism. Depending on which process predominates, different patterns of pathology may develop, particularly in the case of methamphetamine-associated cardiomyopathy

As previously mentioned, a variety of echocardiographic abnormalities are reported in methamphetamine abusers (Table 2)[table not included in quote]. Dilated cardiomyopathy has most commonly been reported. Other patterns include both increased frequency of hypertensive cardiomyopathy, and a stress cardiomyopathy pattern, which incorporates both Takotsubo and reverse Takotsubo patterns. In addition to various patterns of non-ischaemic cardiomyopathy, the progress of ischaemic cardiomyopathy also appears to be accelerated. Valvular lesions may also be increased in frequency.

In fact, in some studies, stress cardiomyopathy was among the most common cardiac effects(12):

Methamphetamine (METH) users have higher incidences of cardiovascular disease. Recent epidemiology studies suggest that cardiovascular disease is the leading cause of natural death in METH users, and the second leading cause of death overall in METH users. Of these cases, left ventricular hypertrophy and Takotsubo cardiomyopathy were among the most common types of cardiovascular disease .

As a teaser, it should also be noted that death in these cases can often present as PEA.  Many folk believe that drug-related deaths and sudden cardiac death must present as ventricular fibrillation.  For readers who don’t know, PEA stands for “pulseless electrical activity” and means that the electrical system of the heart is telling the heart to beat, but it just isn’t responding.  In ventricular fibrillation, the signals that tell the heart to beat become disorganized.  People have claimed that PEA is a hallmark of asphyxial deaths.  It is not. In these cases, methamphetamine-induced cardiac arrest can present as PEA. In one study of 103 patients with stress cardiomyopathy with cardiac arrest, of 84 that presented with cardiac arrest (as opposed to developing arrest in the hospital), 36 presented with PEA, 37 presented with ventricular fibrillation, and 11 presented with ventricular tachycardia (13).

In addition, in experimental studies, methamphetamine may cause cardiovascular collapse as a result of toxic effects in the rostral ventrolateral medulla causing central cardiac dysregulation. Methamphetamine users who are not in extremis can present with a broad array of electrocardiographic abnormalities, most commonly QTc prolongation.

As one review notes, chronic use of methamphetamine has a “stacking” effect on cardiotoxity, so at the time of death the acute drug level may be low and the death is due to this continued toxicity rather than some short term acute intoxication (14):

Cardiovascular effects reflect a heightened catecholaminergic state (ie, elevated heart rate [HR], blood pressure [BP], and myocardial contractility), direct vasoconstriction or vasospasm, and possibly modulation of reactive oxygen species, inflammation, and reduced NO‐mediated vasodilation… Although the euphoric effect often begins to recede by 4 hours, increases in HR and BP may persist for >24 hours. Therefore, repetitive dosing can lead to “stacking” of the cardiovascular effects. “Binge” use, defined as use for several days after which a period of abstinence ensues, has been associated with sensitization to the HR and BP effects in animals, resulting in cardiac inflammation and necrosis.

The effects of methamphetamine on the cardiovascular system may be enhanced by the concomitant use of other drugs, mainly alcohol, cocaine, and opioids….Opioids, which reduce oxygenation via their respiratory depressant effects, may dysregulate the myocardial oxygen supply and demand balance when used in conjunction with methamphetamines.

This “stacking” of the is a cardiotoxic component of methamphetamine intoxication  is not as dose-dependent as traditional overdose.  Because it can occur at a relatively low concentration of methamphetamine in the blood at the time of death, it’s contribution to death may not be immediately appreciated.

The combination of opiates and methamphetamine act synergistically to result in sudden cardiac death.

Finally, since I am writing about methamphetamine, I am obligated to say a word about so-called “Excited Delirium.”  This is a syndrome of agitated delirium and hyperthermia resulting in a serotonin-syndrome-like death in long term users of sympathomimetics, especially methamphetamine.  The diagnosis is somewhat controversial, primarily for political rather than practical reasons, but it is not necessary to go into it further here because Mr. Floyd was simply not delirious, and thus any diagnosis of delirium would be incorrect.

It should also be pointed out that there are a number of presentations that are related, and distinguishing between them is sometimes more of a nosologic than “real” issue.  A number of years ago, Chan sort of covered them all in what he called “Sympathomimetic poisoning syndrome” in that they were all basically the result of an amphetamine/methamphetamine/adrenaline/ecstacy/etc storm that resulted in a sympathetic overload.  A lot of these different syndromes really reflect the details of what system failure is most dramatic.  If it’s cardiac failure, then it’s stress cardiomyopathy.  If it’s hyperthermia, then it’s serotonin syndrome.  If it’s associated with delirium, then it’s excited delirium (until that becomes to politically toxic to use — then folk will just call it “catecholamine-induced agitated delirium” or something like that so people won’t play word games about it, and move on).  One review, for instance, distinguishes five different “drug-induced hyperthermia (DIH) syndromes:(15)

DIH syndromes are a rare and often overlooked cause of body temperature elevation and can be fatal if not recognized promptly and managed appropriately. There are five major DIH syndromes: (1) neuroleptic malignant syndrome, (2) serotonin syndrome, (3) anticholinergic poisoning, (4) sympathomimetic poisoning, and (5) malignant hyperthermia. The differential diagnosis of DIH syndromes can be challenging because symptoms are generally nonspecific, ranging from blood pressure changes and excessive sweating to altered mental status, muscle rigidity, convulsions, and metabolic acidosis.

Similarly, Chan et al. break it into four categories(16):

This article discusses four relatively common syndromes as examples of drug-induced hyperthermia: (1) familial malignant hyperthermia (MH), (2) neuroleptic malignant syndrome (NMS), (3) sympathomimetic poisoning, and (4) anticholinergic poisoning syndrome.

They continue regarding “sympathomimetic poisoning”:

Thirty million Americans have experimented with cocaine, methamphetamines, and amphetamines, with more than 7.5 million people abusing these drugs on a regular basis. Mild to severe hyperthermia may be associated with recreational or pharmaceutical sympathomimetics use. All centrally acting sympathomimetics produce their effects by raising, to various degrees, synaptic concentrations of norepinephrine, dopamine, and serotonin, although their noradrenergic effect is emphasized by the term, sympathomimetic.

… Hyperthermia complicates or is primarily responsible for many of the clinical findings present in patients suffering from amphetamine or cocaine toxicity. Mental status changes, such as agitation, confusion, panic, and hallucinations, are common in sympathomimetic poisoning and may progress to coma or status epilepticus, contributing to a poor neurologic outcome in many of these patients (Evans SE, Williams S, Clark RF, unpublished data). Exertional hyperthermia in the agitated and often restrained patient may be accompanied by rhabdomyolysis, DIC, hyperkalemia, hypoxia, myocardial dysfunction, and the rapid onset of cardiovascular collapse and death shortly after arrival at the hospital or while still in custody of police or paramedical personnel. For those who are admitted to the hospital, the expected complications of hyperthermia and rhabdomyolysis are commonly encountered, including renal failure, acidosis, DIC, and hepatic necrosis, In the case of cocaine, hepatic necrosis may also result from its metabolite, norcocaine.

Some of the people who say that one of these or the other “doesn’t exist” are basically playing word games — the general pattern of sympathomimetic poisoning is what it is.  While classical sympathomimetic poisoning is essentially a hyperthermic issue, other manifestations, particularly stress cardiomyopathy and arrhythmia due to cardiotoxicity does not necessarily require incredibly elevated body temperature.

Since I am obviously building a case for a multifactorial stress-related death, the key here in Mr. Floyd’s death is that it is yet another stressor to add on to the pile in a vulnerable person, and like all the others, it is not unethical, racist, or “character assassination” to note these various stressors.

Methamphetamine and fentanyl as self-medication:

As mentioned previously, some people look at drug taking behavior as self-medication rather than addictive abuse.  These are not mutually exclusive, and my personal belief is that they both occur alone and they can occur together.

A little known fact about methamphetamine is that it is a fairly good analgesic.  In some instances, it is a better pain-killer than opiates. Amphetamines decrease pain perception by modulating transmission at the spinal level. As one animal study notes(17):

Manganese-enhanced magnetic resonance imaging (MEMRI) has been extensively used in studying the
structural and functional features of the central nervous system (CNS). Divalent manganese ion (Mn2+ ) not only enhances MRI contrast, but also enters cells via voltage-gated calcium channels or ionotropic glutamate receptors, which represents an index of neural activities. In the current mouse model, following the repeated amphetamine (Amph) treatment, a reduction of reactivity to thermal pain stimulus was noticed. Since the spinal dorsal horn is the first relay station for pain transmission in CNS, we examined the changes of neural activity in the dorsal spinal cord, particularly the superficial dorsal horn, by analyzing manganese-enhanced T1-weighted MR images (T1WIs). Our data revealed a temporal correlation between reduced pain sensitivity and increased MEMR signals in the spinal dorsal horn subsequent to repeated Amph treatments

In clinical use, amphetamines experimentally decreases the pain associated with dental procedures, is used to treat abdominal pain and other forms of chronic pain. Methamphetamine specifically has been used to control pelvic pain following in vitro fertilization and has been used to treat diabetic neuropathic pain. Addition of sympathomimetics to opiate/opioid therapy in cancer patients potentiates analgesia, allowing the use of lower doses of opiates and counteracts opioid mental impairment.

The same is true for sickle cell disease, where one author notes(18):

Many drugs that are not routinely thought of as analgesics can be of great benefit in treating patients with sickle cell anemia in crisis or who are experiencing chronic pain. The most important psychostimulant drugs used in the acute pain setting are the amphetamines dextroamphetamine and methylphenidate.

Psychostimulant drugs have been shown in human and animal studies to-possess intrinsic analgesic properties and to have the ability to enhance the analgesic properties of opioids when both types of drugs are given in combination. In addition, amphetamines produce a decrease in somnolence and an increase in general cognitive abilities.

In fact there is a subset of methamphetamine addicts who use the drug primarily for self-medication to treat chronic pain rather than for its stimulant effect. As one article in the psychiatric literature notes, one subject preferred the analgesic effect of methamphetamine to that of opioids(19):

Six months after his physician visit for increased back pain, the patient presented to an outside hospital with altered mental status that improved with administration of naloxone; he was then transferred to our hospital for further management. Urine drug screen was positive for benzodiazepines, oxycodone, and amphetamines. When these results were discussed with him, he denied the intentional use of benzodiazepines or opioids, but noted that ingestion of these substances was possible, as he had taken a handful of unknown pills shortly before presentation to the outside hospital. He noted that, within the preceding 6-9 months, he had discontinued the use of prescription opioids and moved solely to methamphetamine use every 3-4 days to palliate his chronic back pain.

A more recent study of stimulant abusers found that 52% of stimulant users used the drug to self-medicate for pain. This was most common in cases of chronic neuropathic pain, such as in HIV infection.(20) In animal studies, administration of methamphetamine to monkeys resulted in over a doubling of pain tolerance(21). Similar results are seen in mice.

Of course, I do not know if Mr. Floyd on the day of his death knew he was taking either of these drugs.  In my jurisdiction, people die because their drug dealers lie to them about what they are being sold.  They think that they are taking heroin, and it turns out they are taking a combination of fentanyl and heroin or fentanyl and methamphetamine.

However, if Mr. Floyd *was* knowingly taking both of these drugs, it makes sense from a self-medication point of view.  For a person with chronic low-grade sympathomimetic poisoning from a functional paraganglioma, or simply as a person who natural had high endogenous catacholamines resulting in chronic hypertension, etc. It makes perfect sense to self-medicate with drugs that would decrease the manifestations of this imbalance.  Unfortunately, it would also make him more vulnerable when placed in a more acutely intense stressful situation.

Next, I’ll discuss the stress associated with being arrested and restrained independent of the issue of prone position.

  1. Somerville NJ, O’Donnell J, Gladden RM, et al.  Characteristics of fentanyl overdose — Massachusetts, 2014-2016  MMWR 2017 66(14): 382–386
  2. Denton JS, Donoghue ER, McReynolds J, Kalekar MB. An Epidemic of Illicit Fentanyl Deaths in Cook County, Illinois: September 2005 through April 2007 Journal of Forensic Sciences 53(2):452-454
  3. Radke JB, Owen KP, Sutter ME, Ford JB, Albertson TE. The effects of opioids on the lung. Clin Rev Allergy Immunol. 2014 Feb;46(1):54-64. doi: 10.1007/s12016-013-8373-z. 
  4. Ng B, Godwin J, Purssell R. Noncardiogenic pulmonary edema associated with ultrapotent opioid overdoses. BCMJ. 2019;61:256-9.
  5. Levine R, Veliz S, Singer D. Wooden chest syndrome: Beware of opioid antagonists, not just agonists.  Am j Emerg Med 2020 38(2):411.e5-411.e6
  6. Tschirhart JN, Li W, Guo J, Zhang S.  Molecular mechanisms of fentanyl-mediated sudden death.  Canadian J Cardiology 2018 34(10):S75-S76 DOI:https://doi.org/10.1016/j.cjca.2018.07.301
  7. Kiyatin EA, Sharma HS Chapter Six – Leakage of the blood-brain barrier followed by vasogenic edema as the ultimate cause of death induced by acute methamphetamine overdose  International Review of Neurobiology 2019 146:189-207.
  8. Sun X, Wang Y, Xia B, et al.  Methamphetamine produces cardiac damage and apoptosis by decreasing melusin Toxicol Applied Pharmacol 2019 378:114543 https://doi.org/10.1016/j.taap.2019.03.015
  9. Akhgari, M., Mobaraki, H. & Etemadi-Aleagha, A. Histopathological study of cardiac lesions in methamphetamine poisoning-related deaths. DARU J Pharm Sci 25, 5 (2017). https://doi.org/10.1186/s40199-017-0170-4,
  10. Schurer S, Klingel K, Sandri M.Clinical Characteristics, Histopathological Features, and Clinical Outcome of Methamphetamine-Associated CardiomyopathyJ Am Coll Cardiol HF. 2017 Jun, 5 (6) 435–445
  11. Paratz ED, et al. The Cardiac Complications of Methamphetamines. Heart, Lung and Circulation (2015), http://dx.doi.org/10.1016/j.hlc.2015.10.019
  12. Freeling JL, McFadden LM. The emergence of cardiac changes following the self-administration of methamphetamine. 2020 Drug and Alcohol Dependence 21,108029
  13. Gili S, Cammann VL, Schlossbauer SA et al. Cardiac arrest in takotsubo syndrome: results of the InterTAK registry. European Heart J. 2019 40:2142-2151.
  14. Reddy PK, Ng TMH, Oh EE, et al. Clinical Characteristics and Management of Methamphetamine‐Associated Cardiomyopathy: State‐of‐the‐Art Review  J Am Heart Assn 2020;9:e016704 doi: 10.1161/JAHA.120.016704
  15. Megan E. Musselman, Pharm.D., BCPS, Suprat Saely, Pharm.D., BCPS, Diagnosis and treatment of drug-induced hyperthermia, American Journal of Health-System Pharmacy, Volume 70, Issue 1, 1 January 2013, Pages 34–42, https://doi.org/10.2146/ajhp110543
  16. Chan, T. C., Evans, S. D., & Clark, R. F. (1997). DRUG-INDUCED HYPERTHERMIA. Critical Care Clinics, 13(4), 785–808. doi:10.1016/s0749-0704(05)70369
  17. Lei BH, Chen JH, Yin HS. Repeated amphetamine treatment alters spinal magnetic resonance signals and pain sensitivity in mice. Neurosicence Letters 2014 583:70-75
  18. Yaster M, Kost-Byerly S., Maxwell LG. The management of pain in sickle cell disease. Pediatric Clinics NA 2000 47(3)::699-710.
  19. Opperman PJ, Derr J, Karre VMM, Balasanova AA Stimulant Use for Analgesia: A Case Series. Prim Care Companion CNS Disord. 2021 Jan 14;23(1):20l02620. doi: 10.4088/PCC.20l02620.
  20. Cathleen M. Beliveau, Vanessa M. McMahan, Justine Arenander, Martin S.Angst, Margot Kushel, Andrea Torres, Glenn-Milo Santos & Phillip O. Coffin (2021): Stimulant use for self-management of pain among safety-net patients with chronic non-cancer pain, Substance Abuse, DOI: 10.1080/08897077.2021.1903654
  21. Weitzman ED, Ross GS. A behavioral method for the study of pain perception in the monkey: The effects of some pharmacological agents. Neurology 264-272.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.