Cannabis-related Stroke

Introduction

Cannabis, which is the most widely used recreational substance in the world, is considered by many consumers as safe with few negative side effects.¹ This opinion is somehow strengthened by the fact that cannabis was also shown to have therapeutic applications.² Cannabis is obtained from the plant Cannabis sativa and its varieties Cannabis indica and Cannabis americana.³ The 2 main preparations derived from cannabis are marijuana and hashish.² The principal psychoactive cannabinoid in cannabis is delta 9 tetrahydrocannabinol⁴, and the potency of different preparations of cannabis that relates to tetrahydrocannabinol content is extremely variable.³ The plasma half-life of tetrahydrocannabinol isμ56 hours in occasional users and 28 hours in chronic users.⁵ Psychopharmacological acute effects associated with cannabis use are euphoria, increased self-confidence, relaxation, and a general sense of well being.³ Except for nausea associated with cancer chemotherapy, most of the potential beneficial effects are not approved by many administrations around the world. Indeed, the more common effects described as beneficial are glaucoma, analgesia, appetite in AIDS patients, tremor, Parkinson disease, spasticity in multiple sclerosis, epilepsia, anxiolytic, or antidepressive actions.^1,3 However, several important negative side effects associated with cannabis are also observed. Indeed, in selected patients, acute psychiatric and behavioral abnormalities, such as anxiety, panic, and attentional abnormalities, have been reported.^3,6 Risk of psychotic disorders or symptoms is higher in regular users of cannabis.⁶ Furthermore, psychological and physical dependence are described as chronic effects of cannabis use.⁶ As for other drugs, cannabis withdrawal syndrome, including anxiety, depressed mood, and sleep difficulties, may occur in heavy users on cessation.^6,7 Also, somatic negative effects, such as cardiovascular complications (myocardial infarction, ventricular tachycardia, and sudden death), peripheral events (peripheral arteritis and kidney infarction), and neurological complications (eg, stroke), have been reported.^1,8–10 Interestingly, despite its widespread abuse, cannabis-associated cerebrovascular disease is only infrequently reported.³ The purpose of this review is to analyze the different aspects of neurovascular complications in cannabis users as described in the literature. We searched on PubMed for articles associating the terms stroke or ischemic stroke (IS) and cannabis or marijuana. We have included articles published in English, French, Spanish, and those from our own files. There are 59 cannabis-related stroke cases^11–17 in 30 published articles including 4 reviews^3,13,14,16 and only 1 report linking cannabis use and cardiovascular events.¹

Neurological Complications of Cannabis Abuse

Usual etiologies of stroke in young population include cervical dissection and cardioembolism, but there is no obvious cause in 15% to 45% of cases.¹⁸ Using a cross-sectional design, a population-based study of hospitalized patients reported an adjusted odds ratio of 1.76 for cannabis exposure associated with IS.¹⁹ Cannabis could thus be considered as a cardiovascular risk factor inducing IS. In the literature, 59 case reports of cannabis-related stroke (mean age, 33±12.4 years) were described, and among them there was a majority of men with a sex ratio of 4.9:1. Most of stroke were IS (n=49)^{11,12,14–17,20–37} or transient ischemic attacks (n=5).^14,38,39 There is only 1 patient who was described with an hemorrhagic stroke associated with cannabis use.⁴⁰ For 4 patients, a diagnosis for stroke was suspected but not confirmed because there was no neuroimaging.^41–43Table resumes data of these 59 case reports; the majority of articles reported an isolated clinical case or at the most 3 cases. There were only 2 studies (1 retrospective and 1 prospective) describing, respectively, 17 and 10 cannabis-associated stroke patients.^16,17

When a relationship between cannabis consumption and stroke is suggested, one important argument is the temporal link between cannabis use and the occurrence or recurrence of stroke.^3,28 Only Thanvi et al³ proposed that stroke occurrence in cannabis users could be a coincidental event. Indeed, various reports suggest that strokes associated with cannabis abuse occurred while the drug was actually being smoked.^16,21,28,38 Furthermore, in other cases, strokes have occurred up to half an hour after the last joint has been smoked.^{12,16,20,25,27} These results are in accordance with one large study showing that cannabis increases myocardial infarction 4.8-fold during the hour following its intake.⁴⁴ In a review of the literature, Peyrot et al⁹ reported 55 cases of peripheral cannabis arteritis resembling Buerger disease, with a worsening of the vascular disease closely associated with periods of heavy marijuana use and a suspended progression of the disease when patients gave up cannabis consumption. Supporting this hypothesis, several studies reported recurrence of stroke in patients when withdrawal of cannabis was not done.^16,17,28 For instance, Mateo²⁸ and Lawson³⁸ described, respectively, a patient with 3 repeated IS and another patient with 3 transient ischemic attacks before stroke in close temporal relationship with cannabis and alcohol intake. In some articles, episodes of transient neurological symptoms were reported before stroke, sometimes simultaneous to cannabis use.^22,29,35,38

Proposed Mechanisms of Stroke in Cannabis Abusers

Neurocardiovascular Effects of Cannabis

The following cardiovascular effects of cannabis may be involved in proposed mechanisms of stroke relating to cannabis use: orthostatic hypotension with secondary impairment of the autoregulation of cerebral blood flow, altered cerebral vasomotor function, supine hypertension and swings in blood pressure, cardioembolism with atrial fibrillation, or other arythmias,^{1,20,27,33,45–47} vasculopathy (toxic or with immune inflammatory), vasospasm,²¹ reversible cerebral vasoconstriction syndrome (RCVS), ^3,20 or multifocal intracranial stenosis (MIS).¹⁷ Zacchariah²⁰ proposed that in patients with IS, elevation of blood pressure could be a reaction to cerebral vasospasm preceded by hypotension. However, in those cases of cannabis-related stroke, only one of the imaging studies demonstrated a border zone infarction compatible with hypoperfusion.²⁹ There is no convincing description of cerebral vasculitis in stroke population. From our point of view, in IS associated with cannabis abuse, there could be several mechanisms acting together. There is only one prospective study that has evaluated the relationship between cannabis use and IS. In this series of 48 young patients with IS, MIS was associated with cannabis use in 10 patients.¹⁷ Deplanque et al⁴⁸ suggest that the low frequency of neurovascular complications in cannabis users may reflect a genetic predisposition in some individuals. In case reports of cannabis-related stroke, cannabis users are often chronical users^14,16,17 and less frequently occasional abusers.²⁸ Potential triggering factors of stroke in cannabis users are described as sexual activity,^25,30 concomitant alcohol consumption,^{1,16,17,21,22,28,38} or unusually high consumption of cannabis.^{17,20,22,27,28} Binge drinking can be proarythmogenic¹ and can induce dehydration after excessive intake.⁴⁹ When cannabis is used, pure tetrahydrocannabinol is the only substance that is being rarely ingested. Cannabis is often smoked with tobacco with further exposure to additional toxic chemicals⁵⁰ and a wide variety of products of combustion are also inhaled.

Neuroimaging of Stroke in Cannabis Users

In the literature, it is frequently difficult to evaluate the mechanisms of stroke in cannabis users because in most of the cases, no exhaustive neurovascular investigations in acute phase of stroke and no radiological follow-up were performed. This situation has lead some authors to conclude that structural imaging in drug abusers was indistinguishable from those obtained in patients with other causes of stroke.⁵¹ As reported in Table, 55 of 59 cannabis-related stroke patients underwent a parenchymal computed tomography scan and MRI in, respectively, 45% (n=25) and 84% (n=46). There were 50 strokes confirmed by cerebral imaging (49 IS, and 1 hemorrhagic stroke), and 5 transient ischemic attacks without cerebral abnormality. IS or territory transient ischemic attacks were posterior in 46% (n=25), anterior in 50% (n=27), and both in 2 cases. Cerebral magnetic resonance angiography (MRA) was performed in 37 cases, transcranial Doppler in 17, and computed tomography scan angiography in 4 cases. Conventional angiography (CA), which is the reference to evaluate intracranial stenosis, was performed in 19 cases, associated with a noninvasive vascular investigation in 16 cases.^{12,14,16,17,20,21,25,29,31,33,35,40–43} Vascular investigations showed abnormalities in only 54% (n=25) of patients with vascular imaging. Stenosis was located on posterior cerebral (n=15), middle cerebral (n=13), superior cerebellar (n=8), anterior cerebral (n=4), basilar (n=2), anterior choroidal (n=1), and posteroinferior cerebellar (n=1) arteries. Cerebral MRA may not be suitable to evaluate stenosis, however. For instance, in one case previously described, the radiologist suggested that abnormalities on MRA were artifacts but CA confirmed the presence of MIS.¹⁷ It is difficult to recognize intracranial stenosis on cerebral MRA when medical doctors are not aware or sensitized to this possibility. Moreover, there was no vascular imaging in 22% of cases (n=13) in the acute phase. Inμ27% (n=16) there was a cerebral radiological follow-up, and a cerebral arterial imaging in all cases but one (MRA in 12, computed tomography scan angiography in 1, CA in 1, and transcranial Doppler in 1) with a partial or total reversibility of stenosis in 11 patients of 15.^{11,16,17,28,30,31,33,40} Urinary screening was performed in 78% of all clinical cases (n=46), and only 2 were negative for cannabis. The outcome was usually favorable (5% mortality among the 59 patients), but it is difficult to conclude about this point because the follow-up was variable, with different delays and without standardized clinical assessment in most of the studies.

Singh et al¹⁶ recently reported a large retrospective series of 17 patients with cannabis-related stroke and reviewed neurovascular imaging in acute stage but without control imaging. In 5 patients of 17, the examination was not completed by vascular imaging. Only 2 patients of 17 were described as having vascular abnormalities (left vertebral stenosis and right posterior artery stenosis). Wolff et al¹⁷ previously published a cohort of 48 consecutive young patients with IS, which is the largest clinical and radiological prospective series evaluating the relationship between IS and cannabis use. In this study, cannabis use might be associated with IS in young adults with a specific vascular pattern; MIS on MRA confirmed on CA with 3-dimensional reconstructed images. As a matter of fact CA was more sensitive than MRA to detect stenosis on superior cerebellar and on posteroinferior cerebellar arteries. Vascular imaging showed that posterior cerebral (n=9) and superior cerebellar (n=7) arteries were the most frequently involved vessels (Figure 1). Characteristics of this angiopathy were the involvement of multiple arteries and a partial or total reversibility of vasoconstriction within 3 to 6 months (Figure 2). These results showed that the term vasospasm is inappropriate in such cases because usually a spasm is transient and not prolonged. Therefore, in this population, the term reversible intracranial stenosis should be preferred to vasospasm. Topography of IS is more frequent in vertebrobasilar territory in 7 of 10 patients in this series, suggesting a susceptibility of posterior circulation. One explanation could be that posterior circulation might be more vulnerable and has a poor autoregulatory mechanism compared with anterior circulation.^16,28,32 The discrepancy between our previous results and the other studies, concerning stroke topography (more frequently in posterior territory), stenosis location (specially on posterior cerebral and superior cerebellar arteries), or the reversibility of intracranial stenosis (in 6 out of 10 patients), could be explained by our exhaustive systematic acute neurovascular investigations with a systematic follow-up.

• Download figure
• Download PowerPoint

• Download figure
• Download PowerPoint

Cannabis, IS, and RCVS

The above-mentioned results raise the question of the relationship between the vasoconstriction pattern observed in our patients and the RCVS recently described.^30,52 RCVS is a clinicoradiological syndrome with recurrent strong headaches and sometimes a neurological focal deficit related with stroke, and reversible vasoconstrictions on intracranial vascular imaging within 3 months. Calabrese described in his narrative review of RCVS a case of a 46-year-old man, cannabis abuser, who developed recurrent episodes of thunderclap headaches and a bilateral IS in occipital lobes, cerebellar hemispheres, and right frontal lobe.³⁰ Vascular imaging showed multifocal segmental stenosis, and follow-up MRA showed resolution of the cerebral vasoconstriction. These data confirm those which suggested that there is a temporal relationship between drug use and acute neurological deficit. Reversibility of the vasoconstriction within 12 weeks is a key point of this syndrome, but if one condition associated with vasoconstriction is still present, like cannabis consumption, stenosis could not be reversible. In a series of 67 patients with RCVS revealed by multiple thunderclap or unusual headaches, 20 were cannabis abusers, suggesting that cannabis could be a precipitating factor associated with RCVS, but no one experienced a stroke.⁵² In the literature, there is only 1 case associating brain hemorrhage and cannabis use and this patient had reversible MIS related to RCVS.⁴⁰ The author suggests that a possible mechanism in that case is that when stenosis subsides and perfusion is restored, artery rupture occurs.⁴⁰ There is, however, no evidence for arteritis in stroke related with cannabis abuse, and the vascular abnormalities observed in cannabis users with MIS were probably because of pharmacodynamic effects of the drug. The long duration of stenosis argues in favor of a drug-induced immunoallergic vasculitis rather than vasospam. Our point of view is that this disorder may be considered as a variant of RCVS.

Discussion

In light of this review, cannabis has to be considered as harmful and the cerebrovascular risk when cannabis is consumed is probably underestimated. There are <60 clinical cases of stroke associated with cannabis, and in most of them exhaustive neurovascular investigations were not carried out. Reversible cerebral angiopathy involving several arteries, associated with cannabis consumption in association with tobacco and alcohol use, is the most convincing mechanism of IS in young adults consuming cannabis. We also suggest that reversible MIS induced by cannabis is probably a variant of RCVS. The main similarity between MIS¹⁷ and RCVS⁵² is the presence of reversible multiple intracranial stenosis. The differences between these syndromes are based on the sex ratio of patients (more male in MIS and more female in RCVS), the presence of thunderclap headache in RCVS,⁵² and the location of intracranial stenosis (more frequent on posterior circulation in MIS). Some authors suggest that some genetic predisposition in few individuals could explain the very low frequency of such complications, but this hypothesis has to be confirmed. The high proportion of MIS associated with cannabis use in the study of Wolff et al¹⁷ could be explained by the exhaustive screening including a strong questioning about drug use, a systematic urine analysis, and a precise repeated arterial imaging. Indeed, in the literature, the most frequently presented characteristics of cannabis users who experienced a stroke are young male, chronic tobacco and cannabis abusers who have had an unusual high consumption of cannabis and alcohol just before stroke.

Conclusion

In regard to the literature, cannabis-related stroke is not a myth, and a likely mechanism of stroke in most cannabis users is the presence of reversible MIS induced by this drug. The reality of the relationship between cannabis and stroke is, however, complex because other confounding factors have to be considered (ie, lifestyle and genetic factors). To confirm that cannabis may be a precipitating factor of RCVS with severe complications, an epidemiological study to determine the incidence of MIS, complicated or not by stroke, in the general population and in the cannabis users is necessary.

As a recommendation we suggest asking all young patients with cryptogenic stroke about their drug consumption and making urine analysis for cannabinoids systematic in this population. It is necessary to have systematic exhaustive investigations in these patients, including MRA in acute stage of stroke to search intracranial stenosis. Thereafter, control vascular imaging is important to evaluate the reversibility of vascular abnormalities. A careful examination of MRA is necessary, and it would certainly be sufficient in most cases to evaluate intracranial stenosis. However, when there is a doubt between stenosis and arterial artifacts, CA with 3-dimensional reconstructed images is needed to confirm diagnosis of stenosis.

Acknowledgments

The author would like to acknowledge Rodrigue Galani (PhD) for writing assistance.

Footnotes