Why Do Marijuana and Synthetic Cannabimimetics Induce Acute Myocardial Infarction in Healthy Young People?

The use of cannabis preparations has steadily increased. Although cannabis was traditionally assumed to only have mild vegetative side effects, it has become evident in recent years that severe cardiovascular complications can occur. Cannabis use has recently even been added to the risk factors for myocardial infarction. This review is dedicated to pathogenetic factors contributing to cannabis-related myocardial infarction. Tachycardia is highly important in this respect, and we provide evidence that activation of CB1 receptors in brain regions important for cardiovascular regulation and of presynaptic CB1 receptors on sympathetic and/or parasympathetic nerve fibers are involved. The prototypical factors for myocardial infarction, i.e., thrombus formation and coronary constriction, have also been considered, but there is little evidence that they play a decisive role. On the other hand, an increase in the formation of carboxyhemoglobin, impaired mitochondrial respiration, cardiotoxic reactions and tachyarrhythmias associated with the increased sympathetic tone are factors possibly intensifying myocardial infarction. A particularly important factor is that cannabis use is frequently accompanied by tobacco smoking. In conclusion, additional research is warranted to decipher the mechanisms involved, since cannabis use is being legalized increasingly and Δ9-tetrahydrocannabinol and its synthetic analogue nabilone are indicated for the treatment of various disease states.


Introduction
Cannabis is derived from leaves, stems or flowers of the Cannabis sativa and Cannabis indica plants, which have been known since ancient times, and like other primeval plants, e.g., opium poppy and ephedra, have been used over centuries and are still prominent today [1][2][3]. Mostly, dried flowers or leaves (marijuana) are inhaled through smoking or vaping, but edible products such as cookies, gummies or chocolate are used as well. Cannabis plants contain more than 400 different compounds and about 100 cannabinoids [4]; the latter ones, mainly ∆ 9 -tetrahydrocannabinol (THC; previous name ∆ 1 -tetrahydrocannabinol) are responsible for the biological effects of cannabis including euphoria, relaxation and changes in perceptions but also dysphoria, anxiety or psychotic symptoms. THC and the nonintoxicating cannabidiol (CBD) are the best-studied cannabinoids [4,5].
THC, CBD, its mixture and the synthetic cannabinoid nabilone are also available in purified or pure form and are used for medical purposes. Cesamet ® (nabilone) and Marinol ® (THC; IUPAC name: (-)-∆ 9 -trans-tetrahydrocannabinol; INN: dronabinol) are approved for anorexia and weight loss in HIV infection and for nausea and vomiting in cancer chemotherapy; Epidiolex ® (cannabidiol) is indicated for Lennox-Gastaut and Dravet syndrome and Sativex ® (mixture of CBD + THC, nabiximols) for neuropathic pain Figure 1. Cannabinoids and their affinities to the classical cannabinoid CB1 and CB2 receptors and to other receptors sensitive to cannabinoids, as well as to inhibitors of enzymes involved in the synthesis and/or degradation of AEA and 2-AG. Note that the numbers in the superscript indicate the appropriate reference . The figure presents only phytocannabinoids (green font), synthetic cannabinoids and other compounds discussed in this article (black font), endogenous cannabinoids (pink font) and inhibitors of the endocannabinoid synthesis and degradation (blue font) that have been considered in this review. ECS, endocannabinoid system; the "plus sign" indicates agonism and the "minus sign" antagonism, inverse agonism or inhibition versus the respective receptors/enzymes. The intensity of blue color next to the compound is higher the lower the values of Ki, IC50 or EC50 are (expressed in nM). Based on Pertwee et al. [39] unless stated otherwise (superscript). WIN55212-3, inactive S(-)enantiomer of WIN55212-2 [40]; AM404, an Figure 1. Cannabinoids and their affinities to the classical cannabinoid CB 1 and CB 2 receptors and to other receptors sensitive to cannabinoids, as well as to inhibitors of enzymes involved in the synthesis and/or degradation of AEA and 2-AG. Note that the numbers in the superscript indicate the appropriate reference . The figure presents only phytocannabinoids (green font), synthetic cannabinoids and other compounds discussed in this article (black font), endogenous cannabinoids (pink font) and inhibitors of the endocannabinoid synthesis and degradation (blue font) that have been considered in this review. ECS, endocannabinoid system; the "plus sign" indicates agonism and the "minus sign" antagonism, inverse agonism or inhibition versus the respective receptors/enzymes. The intensity of blue color next to the compound is higher the lower the values of K i , IC 50 or EC 50 are (expressed in nM). Based on Pertwee et al. [39] unless stated otherwise (superscript). WIN55212-3, inactive S(-)enantiomer of WIN55212-2 [40]; AM404, an inhibitor of anandamide transport [41]. Abbreviations: ∆ 9 -THC, ∆ 9 -tetrahydrocannabinol; 2-AG, 2-arachidonoylglycerol; abn-CBD, abnormal cannabidiol; ACEA, arachidonoyl-2'-chlorethylamide; ACPA, arachidonylcyclopropylamide; AEA, anandamide; CB 1 , cannabinoid CB 1 receptor; CB 2 , cannabinoid CB 2 receptor; CBD, cannabidiol; DAGL, diacylglycerol lipase; ECS, endocannabinoid system; FAAH, fatty-acid amide hydrolase; GPR18, G protein-coupled receptor 18; GPR55, G proteincoupled receptor 55; LPI, L-alpha-lysophosphatidylinositol; MethAEA, methanandamide; n.d., not determined; OEA, oleoylethanolamide; PEA, palmitoylethanolamide; TRPV1 transient receptorpotential cation-channel subfamily V member 1; URB597, inhibitor of fatty-acid amide hydrolase. A significant increase in global intake of both plant and synthetic cannabinoids has been observed in recent decades. In 2016, 192.2 million people aged 15-64 years used cannabis for nonmedical purposes across the globe (INCB 2018). The underlying reasons of this effect are growing legalization (e.g., in U.S. the number of states legalizing marijuana for recreational purposes is still rising [42]); the impact of mass culture with the common occurrence of marijuana and its symbols in everyday products; and recently, the overall increase in the use of psychoactive substances during the lockdown of the COVID-19 pandemic [43][44][45]. Moreover, in 2016 the WHO reported a dramatic (about tenfold) increase in the THC content of marijuana [11].
In an early review about the cardiovascular safety of THC [46], the conclusion was reached that harmful effects occur in people with pre-existing heart disease only. In more recent studies, acute exposure even of young healthy people to cannabis was reported to lead to severe cardiovascular events including myocardial infarction (MI), sudden cardiac death, cardiomyopathy, transient ischemic attack and stroke (e.g., [47,48]. For example, Bachs and Mørland [49] reported six cases of cardiac death in young adults in which THC was present in postmortem blood samples. One of the first convincing studies suggesting that marijuana acts as a trigger for myocardial infarction (MI) was performed by Mittleman et al. [50] who showed that marijuana smokers had a 4.8-fold increased risk of developing MI in the first hour after cannabis exposure. Moreover, a French study reported that cardiovascular disorders, including MI and fatal stroke, were observed among 9.5% of 200 cannabis-related hospitalizations [51]. Therefore, it was recommended that individuals with pre-existing cardiovascular conditions should avoid cannabis [48].
In the past five years, cannabis use has been listed among the risk factors of MI (also associated with an increased risk of cardiovascular mortality) in younger patients [52,53] and at least twelve reviews have appeared drawing attention to adverse cardiac effects of cannabinoids (Table 1). They are based on numerous case reports, epidemiologic or retrospective cohort studies encompassing millions of cannabis users, patients or hospitalizations. Their authors conclude that (1) cannabis use is an independent predictor of MI, heart failure and cerebrovascular accidents; (2) young cannabis users are more at risk with respect to hospitalizations for acute MI, arrhythmia and stroke; (3) medical cannabis authorization was associated with an increased risk of visits at emergency departments or hospitalizations for cardiovascular events including stroke and acute coronary syndrome; (4) screening for marijuana use should be performed in young patients with cardiovascular disease; and (5) the increasing risk of MI and other acute cardiovascular events among young cannabis users strongly needs further studies (including clinical trials) to assess cannabis-related cardiovascular implications and to determine the detailed pathophysiology of cardiac adverse events of cannabis (Table 1). Table 1. Reviews from the past 5 years highlighting the potential impact of cannabis intake on the increase in risk of myocardial infarction and other severe cardiovascular disorders.

Final Conclusion References
Systematic review 115 articles (81 case reports, 29 observational studies, 3 clinical trials and 2 experimental studies). 116   The aims of this review are (1) to give a short introduction of the endocannabinoid system, including cannabinoid receptors and their distribution in areas relevant for cardiovascular effects; and (2) to discuss sites and mechanisms responsible for tachycardia associated with cannabis use in humans. Moreover, we are going to assess whether cannabis use affects (3) the two prototypical pathogenetic mechanisms involved in the development of MI, i.e., thrombus formation and coronary constriction; and (4) further mechanisms which may worsen MI, such as a decrease in energy supply, an increase in energy demand, as well as proarrhythmogenic and cardiotoxic effects.

Endocannabinoid System in Humans and in Experimental Animals-Components and Anatomical Distribution
The main components of the endocannabinoid system (ECS) are shown in Figure 1. It consists of (i) the endogenous cannabinoids (endocannabinoids) such as anandamide (AEA), 2-arachidonoylglycerol (2-AG), noladin ether, virodhamine, oleamide or endocannabinoidlike compounds such as palmitoylethanolamide (PEA) or oleoylethanolamide (OEA), (ii) their receptors and (iii) the enzymes involved in their synthesis and degradation [63,64]. Cannabinoids act via two main types of cannabinoid receptors, CB 1 -R and CB 2 -R, that belong to the G protein-coupled receptor (GPCR) superfamily. Other targets of endocannabinoids are the orphan G-protein-coupled receptors GPR18 and GPR55. GPR18, which shares low sequence homology with the cannabinoid receptors, is also activated by endogenous L-α-lysophosphatidylinositol (LPI; [65]). Another important cannabinoidsensitive receptor, which, does not, however, belong to the GPCR superfamily, is the transient receptor-potential vanilloid-type 1 (TRPV1). AEA is among others synthesized by the N-acyl-phosphatidylethanolamine (NAPE) phospholipase D. For 2-AG synthesis, two isoforms of diacylglycerol lipase (DAGL), α and β, are necessary. Fatty-acid amide hydrolases (isoforms FAAH-1 and FAAH-2) and monoacylglycerol lipase (MAGL) are responsible for degradation of AEA and 2-AG, respectively. For removal of AEA, an AEA transporter may also be relevant [64,66].
With respect to THC, two extreme differences to the ECS have to be considered. First, THC has an affinity for CB 1 , CB 2 , GPR18 and GPR55 receptors only (without binding to e.g., TRPV1 receptors; Figure 1). Second, degradation of THC does not involve enzymes such as FAAH or MAGL. For this reason, data with TRPV1 receptors will not be considered in this article at all and results with FAAH and MAGL inhibitors will be discussed only if To compare acute cardiac functional effects (and their potential mechanisms) in human and in experimental animals we have concentrated on the comparison of the distribution of CB-Rs in the heart, coronary artery, platelets and in brain regions involved in cardiovascular regulation. In the human heart, gene and/or protein expression of CB 1 -Rs and CB 2 -Rs occurs in the left ventricle [67][68][69], right atrium [70] and epicardial adipose tissue [71]. CB 1 -Rs and CB 2 -Rs were also identified in hearts of the guinea pig [72], rat [73][74][75], and mouse [76,77]. In detail, these receptors occur in the left ventricle [78,79] and left atrium of the rat [80], and in the left ventricle of the dog [81] and the mouse [67,82,83].
The occurrence of the GPR55 was described in human [84] and mouse [85] hearts, in rat neonatal cardiomyocytes [86] and left ventricles [79] and in mouse ventricles [87]. The GPR18 occurs in rat left ventricles [79] and in rat fetal cardiac tissue, but not the maternal heart [88].
Receptors sensitive to cannabinoids have so far not been identified in hearts of rhesus monkeys, cats and rabbits. Moreover, the components of the ECS have not been detected in the cardiac conduction system of humans or experimental animals. To the best of our knowledge, GPR18 has not been identified in human cardiac tissue.

Tachycardia-General
Zuurman et al. [101] clearly showed in their review analyzing 165 articles, which are dedicated to the acute effects of cannabis or THC on the central nervous system (CNS) and HR in healthy volunteers and are based on 318 tests (or test variants), that an increase in HR is the most reliable biomarker to study the effects of cannabis. Tachycardia leading to complex adverse cardiac consequences such as a decrease in cardiac-stroke volume or myocardial oxygen supply-demand imbalance is regarded as a potent predictor of cardiovascular morbidity and mortality [102,103]. In this context, it is of interest that a reduction in heart rate (HR) induced by various drugs has a beneficial effect in patients with MI [103]. Table 2 shows that a THC-induced increase in HR occurs after: • smoking of cigarettes (joints) [104][105][106][107][108][109][110][111][112][113][114]    Significant tachycardia also occurs after smoking of UR-144, JWH-122 and JWH-210 [114,128] and other synthetic cannabimimetics (reviewed by Zawilska et al. [8] and Chung et al. [9]). Nonetheless, quantitative or even qualitative differences exist, depending on the route of administration, the compound under consideration and the pharmaceutical formulation. Thus, comparable increases in HR took place 0.5 h after smoking or vaporization but only 1.5 h after oral administration of the same dose of THC [129]. An increase in HR was induced by nabilone (Cesamet ® ) but not by higher doses of THC (Marinol ® ), because of the better bioavailability of nabilone compared to THC [121]. Namisol ® , another oral formulation of THC, led to only slight and not clinically relevant changes in HR [130,131].
Cannabinoid-induced tachycardia in humans is mediated by cannabinoid CB 1 receptors (Table 2), since it was antagonized by three different CB 1 -R antagonists including rimonabant [109], AVE1625 [116] and surinabant [118]. Moreover, cannabidiol (CBD), which possesses very low affinity to CB receptors (see Figure 1), failed to modify HR [119] or increased HR only at a much higher dose than THC did [126]. CBD (believed to be an antagonistic principle against some of the effects of THC; [132]) did not modify the tachycardia induced by THC [127].
In order to determine the mechanism and site of action of the tachycardic effect of THC, it is of interest to include experiments on experimental animals since many studies in humans are not possible for ethical reasons. Since THC was less frequently used in experiments on animals than in humans, not only cardiac effects of THC but also of various other CB-R ligands have been considered in Tables 3-5.
How can we explain such a drastic difference in the cardiac effects of THC between humans and experimental animals? Of course, species differences may be responsible.
In this context, one should keep in mind that the low resting HR (~70 beats/min) in humans results from strong parasympathetic dominance [165]. By contrast, basal HR in the experimental animals listed in Table 3 were higher than in humans, pointing to differences in the balance between sympathetic and parasympathetic tone. It amounted to (in beats/min) 100-130 [133],~130 [136], 140-160 [134] and 190-230 [135] in conscious rhesus monkeys,~90 in mongrel dogs [137], 350-400 in rats [142] and~400 in mice [25]. Thus, we cannot exclude the possibility that THC (and/or other cannabinoids) increases HR only in the case of a low basal HR such as in humans.

Tachycardia-Mechanisms
The site/mechanism of action involved in the tachycardic effect of THC in humans may be i. the heart itself; ii. the autonomic nervous system; iii. the central nervous system; iv. the baroreceptor reflex.

Heart
The possibility that THC elicits tachycardia via activation of CB 1 -Rs in the sinoatrial node is not plausible, since these receptors are G i/o protein-coupled, i.e., inhibitory. The possibility that THC acts via the β 1 -adrenoceptors activated by noradrenaline (NA) can be excluded, since THC is devoid of a sufficient affinity for this type of receptors [149,163].
Another possibility might be that THC affects the β 1 -adrenoceptor-mediated effect of NA via allosteric modulation. Maggo and Ashton [167] found in isolated rat right atria that WIN55212-2 and MethAEA slightly increased the chronotropic effect of NA. The authors suggest the potential involvement of CB 1 -Rs (as opposed to CB 2 -Rs) in this effect, but they did not use any antagonists. Wistar rats perfused heart 2 AEA ↔HR, ↔CF, ↓dp/dt max, ↓LVSP antagonists not used [175] Wistar rats perfused heart 2 oleamide ↑CF CB 1 -R suggested but no proven [176]   antagonists not used [182] rabbits platelets AEA HU-210 ↑aggregation ↔aggregation platelet aggregation induced by AEA independent from CB 1 -Rs (not antagonized by RIM but reduced by ASA) [190] mice homogenised hearts THC 100 µM ↓oxygen consumption antagonists not used [191] mice cardiac mitochondria THC 0.1 and 0.2 µM ↓oxygen consumption ↓mitochondria coupled respiration ↓oxygen consumption; not dependent on CB 1 -Rs (similar changes in CB 1 −/− mice) [192] beef cardiacmitochondria THC 120 µM ↓respiration ↓ oxygen consumption ↓mitochondrial oxygen consumption; antagonists not used [193] rats cardiac mitochondria THC, HU-210, AEA THC and HU-210 ↓oxygen consumption and ↓mitochondrial membrane potential ↑mitochondrial hydrogen peroxide production ↓mitochondrial oxygen consumption; antagonists not used [194] Wistar rats cardiac mitochondria THC up to 500 µM ↔ROS production, no mitochondrial swelling ↔membrane potential, no oxidative stress, no lipid peroxidation THC is not directly toxic in isolated cardiac mitochondria, and may even be helpful in reducing mitochondrial toxicity [195] SD rats neonatal ventricular myocytes CB13 prevents ET1-induced ↓mitochondrial bioenergetics and mitochondrial membrane depolarization improvement in cardiac mitochondrial function (precise mechanism unclear) [196] sheep Purkinje fibers THC ↑APD 90 antagonists not used [197]  In a similar study on rat atria [179], CP55940 (or CBD) did not affect the positive chronotropic effect of isoprenaline, an unselective β-adrenoceptor agonist. By contrast, both AM251 (CB 1 -R antagonist) and AM630 (CB 2 -R antagonist) increased the effect of isoprenaline at 1 µM and decreased it at 3 µM. It must be recalled in this context that a negative and positive inotropic effect occurs following CB 1 -and CB 2 -R activation in the isolated left atrium of the rat, respectively [180]. It is surprising that both antagonists influenced the effect of isoprenaline in an identical concentration-dependent manner in the study of Weresa et al. [179] and it is unclear whether the results obtained for the antagonists are of interest for the effect of THC.
However, THC and/or synthetic cannabinoids failed to modify the tachycardia induced by isoprenaline in spinal dogs [149], pithed rats [40,163] and rabbits [139]. They also failed to affect the bradycardia induced by the muscarinic (M) receptor agonist methacholine in pithed rats [40].
In conclusion, there is not much evidence that the conduction system of the heart plays a role in the THC-induced tachycardia.

Autonomic System
Several studies shown in Table 2 suggest that the tachycardia induced by THC in humans might result from sympathetic stimulation and/or parasympathetic inhibition, since it was diminished by previous administration of the β-adrenoceptor antagonist propranolol and the M receptor antagonist atropine [107,[122][123][124]. However, these results are not consistent. Thus, the THC-induced tachycardia led to a decrease in high-frequency heart-rate variability (HF-HRV), a measure of parasympathetic cardiac control; but to no changes in the pre-ejection period (PEP), a measure of sympathetic cardiac functioning [120]. However, THC shortened PEP in the early study by Kanakis et al. [123]. Moreover, Beaconsfield et al. [108] described an inhibitory effect by propranolol but not by atropine.
The question is whether THC directly acts via the autonomic system and/or via a central mechanism. The results by Gash et al. [110], who found that the maximal increases in HR and plasma noradrenaline (NA) level took place 10 and 30 min after THC smoking, respectively, suggest but do not prove that tachycardia is related to a peripheral mechanism. This conclusion may also be reached from the study by Strougo et al. [117], in which the average population equilibration half-lifes for HR and CNS responses were < 10 min and 40-85 min, respectively.
Provided that the autonomic system is the site involved in THC-induced tachycardia, a direct activation of the sympathetic nervous system and/or an inhibition of the parasympathetic system should occur. To clarify the mechanisms, again in vivo experiments on animals and in vitro experiments on tissues from humans and experimental animals have to be considered.
The possibility that the effect of THC on HR directly involves the peripheral autonomic system was examined in dogs, cats, rabbits and rats. In anaesthetized dogs [149], the THCinduced bradycardia was partially inhibited by spinal section at C2-C4 or by bilateral vagotomy (to destroy the sympathetic and parasympathetic parts of the autonomic nervous system, respectively) and abolished by the combination of both procedures. Vollmer et al. [151] showed that the THC-induced bradycardia in anaesthetized cats was diminished by cervical cardiac denervation but not by vagotomy. In pithed rabbits [139] and in pithed and vagotomized rats [40,[160][161][162][163], in which the CNS is mechanically destroyed, neither THC nor one of the synthetic or endogenous cannabinoids under study (i.e., WIN55212-2, CP55940, AEA and MetAEA) produced a fall in HR.
The study by Cavero et al. [149] on dogs also excludes that THC acts via autonomic ganglia. Another two mechanisms have been considered in in vitro studies. Thus, THC might lead to an increased availability of NA due to the inhibition of the neuronal NA transporter or to facilitation of carrier-mediated NA release. The latter two mechanisms are involved in the peripheral effects of cocaine and methamphetamine, respectively, which, as with THC, can elicit a marked tachycardia, sometimes even associated with MI [204,205]. However, an inhibitory effect of THC on the NA transporter in rat hypothalamic synaptosomes occurred at very high concentrations only [206] and THC, CP55940, WIN55212-2, AEA and 2-AG did not facilitate carrier-mediated NA release in rat and mouse renal tissue at all [207]. The results of the latter studies carried out on extracardiac tissues can be transferred to the heart, since the properties of the NA transporter do not differ between tissues.
Although there are some pieces of evidence that THC does not directly act via the autonomic nervous system, more recent data show that there is a direct effect anyway, i.e., via presynaptic receptors. These types of receptors were almost unknown when Cavero et al. [149] and Vollmer et al. [151] carried out their experiments in dogs and cats, respectively. In pithed animal preparations, presynaptic receptors on sympathetic and/or parasympathetic nerve fibers will be overlooked since an impulse flow along the neurones does no longer occur. When, however, the sympathetic outflow of pithed rats [40,160,161] and rabbits [139] was stimulated electrically, a CB 1 receptor-mediated inhibition of the neurogenic tachycardic response could be demonstrated. In pithed rabbits, a CB 1 receptor-mediated inhibition of vagal neuroeffector transmission in the heart was also described [139]. In harmony with the latter study, methylatropine, an M-receptor antagonist that does not penetrate the blood-brain barrier, diminished the bradycardia induced by WIN55212-2 in anaesthetized rats [159]. Presynaptic CB 1 -Rs leading to inhibition of atrial NA release have also been identified in vitro in atrial tissue from humans [168], guinea pigs [177] and rats [178]. Presynaptic facilitatory CB 1 -Rs cannot be expected, since CB 1 -Rs are G i/o protein-coupled and only G s and G q protein-coupled receptors lead to facilitation of neurotransmitter release [208]. The other three receptor entities activated by THC, i.e., CB 2 , GPR18 [177] and GPR55, do not serve as presynaptic receptors.
One can expect that sympathetic hyperactivity and reduced parasympathetic transmission are accompanied by several cellular pathologies, typical of MI-induced cardiac injury. Examples are oxidative stress, infiltration of inflammatory cells to the myocardium and peripheral ganglia, elevation of proinflammatory cytokines and nerve growth factor, and activation of satellite glial cells [209].
In conclusion, presynaptic inhibitory CB 1 -Rs are present on the cardiac sympathetic neurones of humans and animals and on the parasympathetic neurones of rabbits; their activation would lead to a decrease and increase in HR. The bradycardia usually elicited by THC in animals might be related to a predominant action on the presynaptic CB 1 -Rs on the sympathetic neurones or by a combined activation of CB 1 -Rs in the brain and in the autonomic system. The fact that tachycardia occurs in humans instead may be related to a stimulatory input from the CNS which overrides the brake due to the inhibitory CB 1 receptors on the sympathetic nerve endings. The potential role of presynaptic inhibitory CB 1 -Rs on the cardiac human parasympathetic neurons has so far not been examined.

Central Nervous System
In order to obtain a deeper insight into the brain mechanisms involved in the effect of THC on HR, experiments are of interest in which THC or another cannabinoid was administered to the cerebrospinal fluid (CSF) or directly into brain sites involved in cardiovascular regulation. In studies on dogs [149], cats [151], rabbits [140,141] and rats [210][211][212] in which THC or another cannabinoid agonist was injected into the cerebral circulation or into the CSF, a bradycardia occurred consistently (Table 5). Only in one study on anaesthetized rats, an agonist with preference for CB 1 receptors did not affect HR at all [210]. Table 5. Cardiovascular effects of acute cannabinoid administration into the cerebral circulation, the cerebrospinal fluid or directly into selected brain areas.

↓HR, ↓BP ↑HR, ↑BP
the centrally induced ↑HR and ↑BP is mediated by CB 1 -Rs in the PVN (reduced by AM251 given into the PVN) and can be masked by peripheral CB 1 -Rs; the direction of the response (↑ or ↓ of HR and BP) probably depends on the sympathetic tone [220,221] Wistar rats urethane PVN CP + AM251 1.7 mg/kg i.v. 10 ↑HR, ↑BP pressor response of CP (after blockade of peripheral CB 1 -Rs by AM251) mediated via NMDA-, GABA A -, β 2 -, TP-, AT 1 -Rs and NO (antagonized by the respective inhibitors given i.v.) [220,221]   An entirely different picture emerged when THC or an agonist was injected into brain areas relevant for cardiovascular regulation; all studies of that kind were performed on rats (Table 5; experiments involving the nucleus tractus solitarii (NTS) and the ventral medial prefrontal cortex (vMPFC) will be discussed under Section 4.4). Microinjection of the cannabinoid into the rostral ventrolateral medulla (RVLM) caused tachycardia [94,213], bradycardia [214] or did not affect HR at all [159,216]. In one study, injection of a GPR18 agonist into the RVLM increased HR [215]. HR was increased following microinjection of AEA into the dorsal periaqueductal grey (dPAG; [217][218][219]) but decreased upon microinjection of an inhibitor of AEA degradation into the bed nucleus of the stria terminalis (BNST; [222][223][224]); both effects were mediated via CB 1 -Rs.
In our own studies on the paraventricular nucleus of the hypothalamus (PVN; [220,221]; Table 5), cannabinoids led to brady-or tachycardia, dependent on the experimental conditions. Microinjection of the cannabinoid CP55940 into this brain region led to bradycardia, which was antagonized by the CB 1 -R antagonist AM251 given into the PVN. When AM6545, a CB 1 -R antagonist which does not penetrate the blood-brain barrier, was injected i.v., CP55940 microinjected into the PVN elicited tachycardia instead. One interpretation of these findings might be that CP55940 increases the sympathetic outflow, which is, however, inhibited by cardiac presynaptic inhibitory CB 1 -Rs, eventually leading to bradycardia. If the "brake" is removed by blockade of these receptors, tachycardia will occur instead. The latter studies are somewhat reminiscent of the paper by Szabo et al. [139] on conscious rabbits, in which WIN55212-2 injected i.v. to conscious rabbits produced bradycardia at lower doses (0.005 and 0.05 mg/kg) but tachycardia at the highest dose (0.5 mg/kg). One might assume that the bradycardia is the result of the central activation of the parasympathetic outflow but is reversed into a tachycardia when the presynaptic CB 1 -Rs on the vagal nerves are strongly activated.
THC does not only act via the autonomic nervous system but also activates the hypothalamus/pituitary/adrenal cortex (HPA) axis, eventually leading to an increased cortisol secretion [229]. Cortisol in turn sensitizes the adrenoceptors which lead to tachycardia and BP increase. THC acts via different pathways in the brain, namely via noradrenergic neurones projecting from the locus coeruleus and serotoninergic neurones projecting from the raphe nuclei to the PVN. In addition, the PVN, which represents the origin of the HPA axis, is also activated by neurones originating in higher brain regions (reviewed in El Dahan et al. [230]).
In conclusion, there is good evidence that the effect of THC on HR is the result of a combined action between central CB 1 receptors in areas of the brain involved in cardiovascular regulation and peripheral presynaptic inhibitory CB 1 receptors on sympathetic and/or parasympathetic nerve fibers. It is interesting that even within the same animal species cannabinoids can elicit either brady-or tachycardia. In humans, the interplay between central and peripheral CB 1 receptors appears to be such that tachycardia will be the only effect. Tachycardia is further increased by a THC-driven rise in cortisol secretion.

Baroreceptor Reflex
Administration of conventional (rapid release) formulations of the Ca 2+ channel blocker nifedipine led to marked tachycardia sometimes associated with MI [231]; this side effect is related to unloading of the baroreceptor reflex due to an abrupt fall in blood pressure. Table 2 shows that cannabinoids, although they decreased blood pressure in two studies on conscious humans, usually increase blood pressure or leave it unaffected. Moreover, the dose-dependent increases in BP induced by THC are better correlated to changes in HR than to the dose [104].
There are few studies on anaesthetized animals in which the effect of topical administration of cannabinoids into the NTS on HR and on baroreceptor sensitivity was examined (Table 5). HR was not affected in rats [41,226,227]. Baroreceptor sensitivity was not affected in dogs [225] and in rats in the study by Durakoglugil and Orer [226], increased in the studies by Seagard et al. [227] and Brozoski et al. [41] and even decreased in the study by Lagatta et al. [228]. In the latter study, which was performed on conscious rats and in which the CB 1 -R antagonist AM251 was topically administered to the vMPFC, HR remained constant but baroreceptor sensitivity increased, suggesting that the CB 1 -Rs in this brain region decrease its sensitivity [228].
In conclusion, it is unlikely that the baroreceptor reflex plays a role in the tachycardia elicited by THC and other cannabinoids.

Thrombus Formation and Coronary Constriction?
Since THC can lead to MI associated with tachycardia, the question arises whether THC also influences the two major causes of MI, i.e., thrombus formation and coronary constriction.
Multiple case reports have linked marijuana to thrombus formation, leading to acute MI [13], and the authors quote the publication by Deusch et al. [91] on human platelets in vitro in which CB 1 -and CB 2 -Rs are expressed and THC enhances expression of the platelet fibrinogen receptor (glycoprotein IIb-IIIa) and P-selectin; experiments with CB-R antagonists were not performed in that study. In other studies, THC by itself failed to modify the aggregation of human platelets [186] and even decreased aggregation of human and rabbit platelets induced by various factors [182]. Agonists such as ACEA, HU-210 and JWH-015 also failed to modify the aggregation of human [187,188] or rabbit platelets [190]; the same held true for the CB 1 -and CB 2 -R antagonists AM251 and AM630 [186,189] and for LPI (endogenous ligand of the GPR55), which, in addition, decreased the ADPinduced platelet aggregation [26]. Only the endocannabinoids 2-AG, AEA and virodhamine clearly activated human platelets and stimulated their aggregation [185][186][187][188]; however, in a manner dependent on thromboxane A 2 , which cannot be formed from THC and synthetic cannabinoids (for details, see Table 4). In the only in vivo study, WIN55212-2 given i.p. failed to affect thrombus formation in the ear venules of hairless mice whereas AEA reduced it, again by an arachidonic acid-derived agent ( [164]; Table 3).
The possibility that THC leads to alterations of the vascular wall and eventually to atherogenesis and atherotic plaque rupture was considered in human cells. CB 1 -R activation increases the formation of reactive oxygen species (ROS) and accumulation of lipid droplets in macrophages, ROS production and injury of endothelial cells and ROS formation and migration of vascular smooth-muscle cells. By contrast, CB 2 -Rs usually influence the three cell types in an opposite direction, i.e., they counteract the formation of lipid droplets in macrophages, inhibit the adhesive and infiltrative properties of endothelial cells and inhibit migration of vascular smooth-muscle cells (reviewed in El Dahan et al. [230]).
In conclusion, taking into account the experimental studies discussed above, there is not much evidence to suggest that THC and synthetic cannabinoids (unlike endocannabinoids) can lead to platelet aggregation as a process eventually leading to thrombus formation, but they may promote atherogenesis and atherotic plaque rupture as a prerequisite of thrombus formation. Whether a detrimental effect on the wall of the coronary arteries will really occur in vivo is, however, unclear since CB 2 -R activation, unlike CB 1 -R, has effects in the opposite direction.
Coronary flow (CF) is impaired by tachycardia, since blood supply to the coronary arteries can occur only during the diastole of the heart action. Consequently, in the rat perfused spontaneously beating heart, THC caused tachycardia that was accompanied by a decrease in coronary flow [169]. In this section, we will consider the possibility that THC and related compounds have a direct effect on the coronary arteries.
THC reduced CF in the rat perfused heart in which vasopressin (VP) was given to induce coronary tone [175]. By contrast, other cannabinoids enhanced CF both under standard conditions [77,174,176] and under VP ( [175]; for details, see Table 4). The increase in CF was mainly connected to changes in ventricular performance measured as a decrease in left ventricular developed pressure (LVDP, which is obtained by subtracting the enddiastolic pressure from the left ventricular systolic pressure; LVSP); this phenomenon might explain the increase in CF [77,174]. However, the increase in CF in coronary arteries preconstricted with VP was associated with an enhanced LVSP [175].
Which receptors are involved in the above effects? Surprisingly, the decrease in CF by THC was not antagonized by CB 1 -and CB 2 -R antagonists. On the other hand, CB 1 -Rs are involved in the enhancement of coronary vasodilation, since CB 1 -R antagonists such as rimonabant, AM251 and O-2050 but not the CB 2 -R antagonist SR144528 or the GPR18 antagonist O-1918 showed an antagonistic effect [77,175]. For the ACEA-induced increase in CF, Ford et al. [174] suggest the involvement of a novel site since this effect was reduced by rimonabant, AM281 (CB 1 -R antagonists) and SR144528 (CB 2 -R antagonist), but not by AM251 (CB 1 -R antagonist) and AM630 (CB 2 -R antagonist).
The results of experiments on isolated rat coronary arteries [90] are consistent with those obtained on isolated hearts. Thus, WIN55212-2 elicited vasodilation and this effect was antagonized by two CB 1 -R antagonists [90].
In conclusion, a direct dilatory effect of cannabinoids on coronary arteries has been shown in one in vitro model of the rat only. Although most cannabinoids increase CF, THC itself showed an inhibitory effect.

Increased Energy Demand and Decreased Energy Supply?
Tachycardia, thrombus formation and/or coronary constriction have been discussed as factors involved in the development of acute MI accompanying the use of THC or related compounds. The question arises whether other factors may contribute. An increase in energy demand and/or a decrease in energy supply may play a role.
An increase in energy demand might be caused by a positive inotropic effect of cannabinoids. Indeed, in isolated rat left atria the CB 2 -R agonist JWH-015 and AEA, examined in the presence of a CB 1 -R antagonist, induced a positive inotropic effect which is mediated by CB 2 -Rs [180]. Moreover, Walsh et al. [87] concluded from their experiments on GPR55-deficient mice that GPR55 increases the adrenoceptor-mediated positive inotropic response.
By contrast, a negative inotropic effect of THC, ∆ 8 -THC and HU-210 was obtained in the perfused rat heart (Table 4; [157,170,173]); in those studies, however, the authors did not use CB-R antagonists to determine the type of receptors involved. In the study by Sterin-Borda et al. [180] in rat left atria, the CB 1 -R agonist ACEA alone, as well as AEA in the presence of a CB 2 -R antagonist, had a negative inotropic effect. Likewise, AEA had a negative inotropic effect in human right atrial muscle, which was mimicked by its stable analogue MethAEA and by HU-210 and antagonized by the CB 1 -R antagonist AM251 [70]. The latter increased contractility of human right atrial muscles by itself [70], suggesting that these receptors are activated by endocannabinoids or are constitutively active. The lack of a positive inotropic effect of AEA in isolated human, as opposed to rat, cardiac tissue [70] may be due to species differences. However, Bonz et al. [70] have not examined the effect of AEA in the presence of a CB 1 -R antagonist or of a selective CB 2 -R agonist. An opposite influence of CB 1 -R and CB 2 -R activation in the rat heart might be the reason for the lack of changes in contractile function of the isolated rat heart in response to AEA alone [180] and in rat ventricular myocytes in response to the CB 1 -/CB 2 -R dual agonist CB13 [18]. Another two studies are difficult to interpret. Thus, in the rat perfused heart, both a CB 1 -and a CB 2 -R antagonist showed a negative inotropic effect [172] and in rabbit left ventricular myocytes the CB 2 -R agonist A-955840 had a CB 1 -R-and CB 2 -R-independent negative inotropic effect [38].
In conclusion, there is no evidence that THC and other cannabinoids elicit a positive inotropic effect in the human heart.
A decrease in energy supply might be caused by an impairment of oxygen transport by cannabis use. Since tobacco smoking is associated with an increased carboxyhemoglobin level resulting in decreased cardiac oxygen supply (reviewed in Dorey et al. [232]), the frequent combination of tobacco and cannabis smoking can explain the impairment of oxygen supply in many instances. However, cannabis use per se can also lead to signifi-cantly increased expired carbon oxide concentrations, provided that THC was administered by smoking but not when it was vaporized or taken orally ( [129]; Table 2). An increase in serum carboxyhemoglobin level was also observed in an animal study, i.e., in mice "smoking" cannabis cigarettes via a special smoke-inhalation system ( [148]; Table 3).
Smoking cannabis can decrease the oxygen transport to the heart and, in this respect, changes in mitochondrial oxygen consumption in response to cannabinoids are of interest. However, the results obtained so far are contradictory. Thus, on the one hand, THC or AEA and HU-210 not only at high (100-120 or 1-20 µM) but even at low (0.1 or 0.2 µM) concentrations led to a decrease in oxygen consumption in bovine [193], rat [194] and mouse [191,192] cardiac tissue or mitochondria. The latter was connected with a lower mitochondrial membrane potential and an enhanced mitochondrial hydrogen peroxide production ( Table 4). The detailed mechanism(s) of the above changes have so far not been examined. Although CB 1 Rs were detected in cardiac mitochondria, Mendizabal-Zubiaga et al. [192] excluded CB 1 -Rs, since similar changes were observed in CB 1 -/and CB 1 +/+ mice. By contrast, a detailed analysis of the toxic effects of a wide range of THC concentrations (1-500 µM) on isolated rat-heart mitochondria failed to detect any changes regarding an enhanced production of reactive oxygen species or lipid-peroxidation products, mitochondrial swelling or changes in mitochondrial membrane potential [195]; the authors even concluded that THC may be helpful in reducing mitochondrial toxicity. Moreover, the dual CB 1 /CB 2 receptor agonist CB13 prevented cardiac mitochondrial dysfunction (such as membrane depolarization and decreased mitochondrial bioenergetics) induced by endothelin-1 in neonatal rat ventricular myocytes [196].
Even if a direct effect of marijuana or cannabimimetics on mitochondrial function is controversial, indirect effects should be considered as well. Thus, all potential mechanisms involved in cardiac injury such as tachycardia, constriction of coronary artery and platelet aggregation, changes in action potential (mainly disturbances in calcium homeostasis), i.e., pathological conditions characterized by the deprivation of oxygen supply to cardiomyocytes, may impact adversely on mitochondrial function [233]. There is increasing evidence for an important role of cardiac (e.g., [234][235][236]) and coronary microvascular [237,238] mitochondria in MI.
In conclusion, independent from tobacco smoking, cannabis smoking can lead to an increase in carboxyhemoglobin and a subsequent reduction in oxygen transport. Although controversial data exist as to whether THC affects mitochondrial respiration directly, an indirect detrimental effect (e.g., due to tachycardia) is very likely.

Other Factors
MI can lead to life-threatening arrhythmias, and for this reason possible effects of THC and other cannabinoids on the conduction system of the heart are of great relevance.
In the study by Miller et al. ([125], Table 2), i.v. THC enhanced sinus automaticity and facilitated sinoatrial and atrioventricular nodal conduction in humans in vivo, most probably representing the positive chronotropic and positive dromotropic effect elicited by noradrenaline as a result of sympathetic stimulation; intra-atrial and intraventricular conduction was not affected. It is very plausible that noradrenaline will stimulate the cardiac conductive system, thereby eventually leading to a proarrhythmogenic effect under pathological conditions.
On the other hand, some effects on cardiac ion channels in cardiomyocytes may be beneficial in tachyarrhythmias. As shown in Table 4, cannabinoids might exert an antiarrhythmic effect related to the inhibition of cardiac voltage-gated inward L-type Ca 2+ currents; this verapamil-like effect was shown in rat tissue and is related to the activation of CB 1 -Rs [199]. The suppression of cardiac Na + /Ca 2+ exchanger (NCX1)-mediated currents in rat ventricular cardiomyocytes may contribute to an antiarrhythmogenic effect of JWH-133 under ischemic conditions (mediated via CB 2 -Rs; [201]). Moreover, prolongation of the AP in response to THC was found in sheep Purkinje fibers (receptor not determined; [197]).
The synthetic agonist CB13 inhibited the tachypacing-induced shortening of the rat atrial effective refractory period protecting against atrial fibrillation (receptor not identified; [80]).
By contrast, a decrease in AP duration was observed in rabbit sinoatrial-node samples in response to AEA (CB 1 -Rs involved; [198]) and in rabbit Purkinje fibers in response to a high concentration of JWH-030 (mechanism unclear; [156]). Moreover, the endogenous agonist of GPR55 receptors, l-α-lysophosphatidylinositol (LPI), increased Ca 2+ entry via L-type Ca 2+ channels in rat cardiomyocytes [86]; the involvement of GPR55 receptors has been proven using an appropriate antagonist (Table 4).
In conclusion, although THC may lead to tachyarrhythmia as a result of a high NA level, evidence for direct anti-and proarrhythmogenic effects is restricted to in vitro studies on preparations from animals.
The mechanism(s) of cardiotoxicity are still not entirely clear. When the study by Nahas and Trouve [169] appeared, CB-Rs had not yet been deciphered. The use of selective CB 1 -and CB 2 -R antagonists revealed that the apoptotic effect of THC is neither CB 1nor CB 2 -R-related [203] whereas CB 2 -Rs are involved in the effect of JWH-030 on cell viability [156]. Although THC is devoid of an inhibitory effect on cell viability in H9c2 cells, it may have such an effect in vivo. One can expect that, unlike in a cell line, THC is metabolized to THC-OH and THC-COOH. The metabolites, however, do not possess any affinity for CB 1 -or CB 2 -Rs.
In conclusion, some interesting cardiotoxic effects of THC and synthetic cannabinoids have been shown in vitro but it is unclear to which extent they play a role in humans in vivo.

General Conclusions
Cannabis contains ∆ 9 -tetrahydrocannabinol as its major psychotropic principle and, with respect to vegetative effects, its use for recreational purposes has been considered safe over a long time period. In recent years, however, an increasing number of studies revealed serious cardiovascular effects, even including acute myocardial infarction (MI) in healthy young people; indeed, cannabis has been listed among the risk factors of MI. The potential mechanisms induced by exposure to THC and other cannabimimetics triggering MI are shown in Figure 2. MI related to cannabis use is associated with tachycardia. Tachycardia is the most reliable biomarker of cannabis use and occurs independent of the route of administration. The reason why cannabis elicits tachycardia in humans but almost exclusively bradycardia in animals is unclear but may have to do with the relatively low heart rate level in humans. One explanation for the difference between humans and animals might be that the cannabinoid CB 1 receptor-driven central stimulation of the sympathetic system is inhibited markedly by presynaptic inhibitory CB 1 receptors on the sympathetic nerve fibers in animals but only slightly in humans. Cannabis use is frequently associated with tobacco smoking, thereby increasing the risk to develop MI. However, it is questionable whether the two most typical pathogenetic factors for the development of MI, i.e., thrombus formation and coronary constriction, play a role in the case of cannabisrelated MI, at least not on the basis of the few studies available. Administration of cannabis by smoking but not by other routes impairs energy supply by increasing the formation of carboxyhemoglobin; impairment of mitochondrial respiration is an additional factor. Worsening of MI by an increased energy demand because of a positive inotropic effect is unlikely. Proarrhythmogenic effects of cannabis per se are unlikely but may appear as a consequence of increased noradrenaline levels associated with tachycardia. The increasing use of cannabis preparations for recreational but also for therapeutic purposes warrants each effort to further elucidate cardiovascular mechanisms, in order to avoid severe side effects.
available. Administration of cannabis by smoking but not by other routes impairs energy supply by increasing the formation of carboxyhemoglobin; impairment of mitochondrial respiration is an additional factor. Worsening of MI by an increased energy demand because of a positive inotropic effect is unlikely. Proarrhythmogenic effects of cannabis per se are unlikely but may appear as a consequence of increased noradrenaline levels associated with tachycardia. The increasing use of cannabis preparations for recreational but also for therapeutic purposes warrants each effort to further elucidate cardiovascular mechanisms, in order to avoid severe side effects.

Conflicts of Interest:
The authors declare no conflict of interest.