Pathophysiology and Risk Factors of Cervical Artery Dissection: What Have We Learnt from Large Hospital-Based Cohorts?

REVIEW

CURRENT OPINION Pathophysiology and risk factors of cervical artery dissection: what have we learnt from large hospital-based cohorts?

Ste´phanie Debettea,b,c,d

Purpose of review Cervical artery dissection (CeAD) is a major cause of ischemic stroke in young and middle-aged adults, although relatively uncommon in the community. Recent large collaborative projects have provided new insights into mechanisms and risk factors of CeAD. Recent findings Pathologic changes observed at the media–adventitia border in temporal arteries of CeAD patients suggest a predisposing arterial wall weakness. In large multicenter series of CeAD patients, compared to age- matched healthy controls and patients with an ischemic stroke of another cause, hypertension and migraine, especially without aura, were confirmed as risk factors for CeAD, in addition to cervical trauma and recent infection. Hypercholesterolemia and being overweight were shown to be inversely associated with CeAD. Differences in risk factor profile and structural features between carotid and vertebral dissection suggest that their pathophysiology may partly differ. An association of CeAD with fibromuscular dysplasia and reversible cerebral vasoconstriction syndrome was described. Genetic risk factors of CeAD are still poorly understood. Summary Large cohorts of CeAD patients have refined our understanding of the pathophysiology and risk factors of CeAD, but the molecular mechanisms are still poorly understood. Ongoing high-throughput genetic projects will hopefully provide novel insight into the biological substrate of CeAD. Keywords cervical artery dissection, pathophysiology, risk factors, stroke

INTRODUCTION SUMMARY OF CLINICAL AND Cervical artery dissection (CeAD) corresponds to a RADIOLOGICAL FEATURES, TREATMENT hematoma in the wall of an internal carotid or AND OUTCOME OF CERVICAL ARTERY vertebral artery, and is a major cause of ischemic DISSECTION stroke in young and middle-aged adults [1–3]. The The mean age of occurrence of CeAD is 44 years; incidence of the disease is relatively low in the general although the disease can occur in children as well, it population, estimated around 2.6–3/100 000 indi- is very rare beyond age 65. Carotid dissections are viduals per year [4,5]. The pathophysiology of CeAD more common than vertebral dissections, with a is incompletely understood. Recently, data from large hospital-based cohorts of CeAD patients [6,7] have

provided new insight into the risk factors and mech- a b anisms of CeAD. In this review, we will primarily Department of Neurology, Lariboisie`re Hospital, Inserm Unit U740, cParis 7 University, DHU Neurovasc Sorbonne Paris-Cite´, Paris, France discuss arguments for an underlying arteriopathy, and dDepartment of Neurology, Framingham Heart Study, Boston environmental risk factors and triggers, genetic pre- University School of Medicine, Boston, Massachusetts, USA disposition, and the intriguing overlap with other Correspondence to Ste´phanie Debette, MD, PhD, Service de Neuro- vascular diseases. Prior to this, we propose a short logie, Hoˆ pital Lariboisie`re, 2 rue Ambroise Pare´, 75475 Paris Cedex 10, overview of the clinical and radiological features, France. Tel: +33 01 49 95 25 97; fax: +33 01 49 95 25 96 treatment and outcome of CeAD. Intracranial dissec- Curr Opin Neurol 2014, 27:20–28 tions are not discussed in this article. DOI:10.1097/WCO.0000000000000056

www.co-neurology.com Volume 27 Number 1 February 2014 Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Pathophysiology and risk factors of cervical artery dissection Debette

present as an arterial occlusion or a dissecting aneur- KEY POINTS ysm. In patients with arterial occlusion, if a mural CeAD patients often seem to have a predisposing hematoma cannot be seen, the diagnosis of CeAD arterial wall weakness, as suggested by various can only be confirmed retrospectively if the occlu- concomitant structural and functional arterial sion recanalizes and leads to a typical stenosis or a abnormalities described in association with CeAD, dissecting aneurysm. In approximately 15% of the including pathologic changes in temporal arteries cases, CeAD affects multiple arteries. The dissected predominating at the media–adventitia border. artery usually ‘heals’ within the first 6 months and CeAD is in most instances a multifactorial disease for radiological changes are seldom observed beyond which multiple environmental risk factors (cervical 1 year. Although dissected arteries can recanalize trauma, recent infection, hypertension, and migraine) completely, residual stenoses, occlusions, or aneur- and protective factors (hypercholesterolemia, ysms are common [4]. overweight) have been described. Treatment at the acute phase of CeAD consists of Rarely, CeAD can occur as a complication of a rare antithrombotic therapy to prevent cerebral or inherited connective tissue disorder, mostly vascular retinal ischemia, either by anticoagulants or by Ehlers–Danlos syndrome; genetic variants modulating antiplatelet agents. The choice between the two is the risk of CeAD as part of a multifactorial based on empirical arguments, such as presence, predisposition are currently being explored in genome- size, and distribution of brain infarcts [11,12]. A wide association studies. Cochrane systematic review and meta-analysis of An intriguing overlap of CeAD with other uncommon observational studies does not suggest any signifi- vascular disorders has been described, such as FMD cant difference between anticoagulants and anti- and RCVS; exploring the molecular underpinnings of platelet agents on death and disability, or on risk this overlap may provide useful insight into the of cerebral infarction [13]. A pilot study testing the pathophysiology of CeAD. feasibility of a randomized trial comparing anticoagulants to antiplatelet agents in acute CeAD is close to completion [14&]. Intravenous thrombolysis is not contra-indicated at the acute phase of CeAD ratio of approximately 1.7 to 1 in populations of [15,16,17&]. Whether it is as efficient as in other European origin [4]. A predominance of vertebral causes of ischemic stroke is unclear [17&]. Long-term dissection has been described in Asian cohorts [8,9]. therapy is usually initiated after 3–6 months, and A slight male predominance was reported in Euro- consists of long-term antiplatelet agents in the case pean series (53–57%) [1]. Men more often suffer of residual arterial abnormalities [1]. from carotid dissection and are on average 5 years The rate of recurrent or de-novo cerebral ische- older when CeAD occurs [1,6,10&]. mia after treatment initiation is low in CeAD About 67–77% of CeAD patients suffer from patients (3% in the largest observational studies) ischemic stroke or transient ischemic attack, [6,13,14&,18], with most events occurring in the first although this proportion may be overestimated as weeks following the diagnosis. Similarly, sympto- large CeAD cohorts were recruited through depart- matic CeAD recurrences are also rare and seem to be ments of neurology specializing in stroke care [1]. most frequent within the first months after the Other ischemic manifestations include retinal ische- initial event (2.1% at 3 months in a series of 900 mia in the case of carotid dissection and rarely spinal CeAD patients [6]); some dissections that were diag- cord ischemia in patients with vertebral artery dis- nosed as multiple might have occurred sequentially section. Cervical pain and headache are the most within a short time frame [1]. Some investigators common symptoms, and are very suggestive of consider early CeAD recurrences and multiple CeAD CeAD when they precede the onset of cerebral as the same entity. Late recurrences are possibly ischemia. Horner syndrome can occur in patients underestimated as studies with long-term follow with carotid dissection, because of the compression up are scarce (a 7% rate was reported after 7.4 years of sympathetic fibers by the enlarged artery. Other of follow up in 200 patients [19]) [1]. Mortality rates complications include tinnitus, cranial nerve palsy, after CeAD are less than 4% in recently published and subarachnoid hemorrhage. The latter is rarely cohorts. Functional outcome is usually good, seen when the dissection extends intracranially or approximately 75% of CeAD patients who suffered in the case of concomitant reversible cerebral vaso- a stroke being independent at 3 months [1]. How- constriction syndrome (RCVS). ever, the psychological and social impact of CeAD Radiologically, CeAD most often presents as a can be substantial in these young, active indivi- long, irregular stenosis, starting more than 2 cm duals, and patients often complain of profound above the carotid bifurcation (Fig. 1). CeAD can also fatigue during the months following the event.

1350-7540 ß 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins www.co-neurology.com 21 Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Cerebrovascular disease

(a) (c) (e)

(b)

(d)

FIGURE 1. MRI scans and magnetic resonance angiography (MRA) of CeAD patients. (a) Axial fat suppression MRI scan showing a mural hematoma with narrowed lumen and enlarged external diameter in the left internal carotid artery (patient 1), (b) MRA showing a stenosis of the left internal carotid artery, on a grade I kink of the left internal carotid artery (patient 1). (c) Axial diffusion weighted MRI scan showing an acute brain infarct in the left middle cerebral artery territory, caused by the dissection of the left internal carotid artery (patient 1). (d) Axial fat suppression MRI scan showing a mural hematoma without lumen narrowing and with slightly enlarged external diameter in the left vertebral artery, V2 segment (patient 2). (e) MRA showing two dissecting aneurysms of the left and right cervical internal carotid arteries – the left vertebral dissection is not seen as the arterial lumen is unchanged (patient 2).

ARGUMENTS FOR UNDERLYING ARTERIAL predominating at the media–adventitia border, WALL WEAKNESS including vacuolar degeneration, capillary neo- There is converging evidence suggesting that CeAD angiogenesis, and erythrocyte extravasation, were patients may have a predisposing arterial wall weak- detected in all CeAD patients and only one control. ness (Fig. 2). Indeed, various concomitant structural These findings could suggest that CeAD patients and functional arterial abnormalities were described have an underlying arteriopathy and that outer in association with CeAD, although mostly in small layers of the arterial wall may be primarily involved samples. These include larger aortic root diameter in the causation of the intramural hematoma. [20], increased stiffness of carotid wall material and Arterial dissection is a common feature of circumferential wall stress [21], endothelial dysfunc- certain rare inherited connective tissue disorders, tion [22], and arterial redundancies (kinks, coils, or such as vascular Ehlers–Danlos syndrome, Marfan loops) [23]. syndrome (MFS), or Loeys–Dietz syndrome [25–27]. Pathological specimens of cervical arteries are Hence, it has been hypothesized that CeAD patients difficult to obtain in CeAD patients, given low may have an arterial wall weakness as part of mortality rates and the fact that therapeutic some underlying connective tissue fragility. Skin approaches are largely based on medical treatment. biopsies performed in consecutive CeAD patients As a surrogate pathological marker, superficial have shown that more than half of CeAD patients temporal arteries from 14 patients with spontaneous have ultrastructural skin connective tissue CeAD have been compared to those of nine accident abnormalities on electron microscopy, the most victims without CeAD [24]. Pathologic changes common pattern being ‘Ehlers–Danlos type III like’

22 www.co-neurology.com Volume 27 Number 1 February 2014 Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Pathophysiology and risk factors of cervical artery dissection Debette

– Recent cervical trauma – Recent infection – Hypertension – Rarely monogenic (especially – Migraine vascular Ehlers–Danlos syndrome) – Low cholesterol – Mostly multifactorial – Low body mass index

Genetic risk factors Environmental risk factors

Underlying arteriopathy?

Concomitant structural and functional arterial Arterial wall weakness abnormalities: – Aortic root diameter – Arterial wall stiffness – Endothelial dysfunction – Arterial redundancies – Temporal artery changes at media–adventitia border Association with other vascular diseases: – Fibromuscular dysplasia – Reversible cerebral vasoconstriction syndrome Cervical artery dissection

FIGURE 2. Pathophysiology and risk factors of CeAD. This diagram represents putative mechanisms and risk factors of CeAD. composite collagen fibrils and fragmentation of minor cervical trauma and CeAD is often difficult elastic fibers [28]. Skin biopsies performed in to establish. healthy relatives of three index CeAD patients have An association of CeAD with recent infection suggested that these connective tissue changes may has been reported, potentially predisposing to CeAD be inherited according to an autosomal dominant via endothelial damage or prothrombotic mechan- pattern [29]. isms [33–35]. Acute CeAD is associated with high white blood cell counts compared with patients admitted for non-CeAD IS or with healthy controls ENVIRONMENTAL RISK FACTORS [36&], possibly reflecting a pre-existing inflamma- Cervical trauma is an important risk factor for CeAD tory state. On cervical high-resolution MRI, patients [30,31]. Rarely, CeAD can occur after a major pen- with spontaneous CeAD more often have periarte- etrating or blunt trauma. Often, CeAD patients rial edema, often in conjunction with elevated report a minor cervical trauma in the days or weeks C-reactive protein or erythrocyte sedimentation preceding the dissection. In a multicenter study on rate, than patients with traumatic CeAD, supporting 966 CeAD patients from the Cervical Artery Dissec- the hypothesis of a predisposing role of inflam- tions and Ischemic Stroke Patients (CADISP) con- mation in CeAD [37]. sortium, cervical trauma in the previous month was Until recently, data on the relationship of CeAD reported in 40.5% of CeAD patients [32&]. The with vascular risk factors were controversial [1]. trauma was minor, that is, not leading to a medical Associations had usually been examined in small visit or hospitalization, in 88% of the cases. Prior samples, compared mostly with young patients with cervical trauma was significantly more common in non-CeAD IS [21,23,33,34,38] and seldom healthy CeAD patients than in 651 patients who suffered an controls [39,40]. Within the CADISP consortium, ischemic stroke of a cause other than CeAD (non- the prevalence of vascular risk factors was compared CeAD IS) or in 280 healthy controls. Among the between three groups of age-matched, sex-matched, numerous types of trauma that were reported, cer- and country-matched participants, of which there vical manipulation therapy and cervical trauma were 690 with CeAD, 556 with non-CeAD IS, and during sports were more common in CeAD patients 1170 referents from the general population [41]. than in controls [32&]. A causal relationship between Compared with referents, CeAD patients were more

1350-7540 ß 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins www.co-neurology.com 23 Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Cerebrovascular disease

frequently hypertensive, and had a lower prevalence and more often men. In carotid dissection, infec- of hypercholesterolemia, obesity and overweight. tions in the week preceding the dissection appear An inverse association of CeAD with BMI was also to be more common, whereas neck trauma in the shown in an independent large cohort [40]. All previous month tends to be less often reported than vascular risk factors were less frequent in CeAD in vertebral dissection. Aneurysmal dilatation is patients compared to young patients with a non- more common and bilateral dissection less frequent CeAD IS. Elevated blood pressure could contribute in patients with carotid dissection. The mechanisms to CeAD risk via increasing carotid stiffness [21]. The underlying these differential features are speculative inverse association of CeAD with hypercholestero- [6], but could include: anatomic features, such as lemia and BMI, the young age of occurrence of anchoring of the vertebral artery to the cervical CeAD, and the heterogeneous echostructure of spine rendering it more vulnerable to mechanical carotid arteries in CeAD patients [21], all suggest solicitations, or closer proximity of carotid arteries that atherosclerosis is probably not a predisposing to upper respiratory tract infections; different condition to CeAD, in contrast with aortic dissec- embryonic origins, pericytes and smooth muscle tion [42]. With aging and arteriosclerosis, increased cells in carotid arteries being derived from the neural synthesis and reduced degradation of extracellular crest, whereas vertebral arteries emerge exclusively matrix components, as well as increasing collagen from the mesoderm [53]; partly distinct genetic and elastin cross-links, could be hypothesized to susceptibility factors, supported by the high intra- make the arterial wall of cervical arteries more resist- familial correlation for the affected vessels in ant to tears [43,44]. One could also speculate that familial CeAD [54]. lean persons have less adipose tissue protecting the arteries from minor cervical traumas. The association of CeAD with low cholesterol and BMI could GENETIC RISK FACTORS also reflect shared genetic risk factors. CeAD is believed to be a multifactorial, complex A seasonal pattern in the incidence of CeAD has disorder in the vast majority of cases. In rare instan- been described, with a higher incidence in autumn ces, CeAD can occur as part of a monogenic disorder. and winter. Increased prevalence of infection and higher blood pressure levels in cold seasons have been proposed as putative underlying mechanisms Monogenic forms of cervical artery [45&]. dissection Migraine is a risk factor for CeAD [46,47]. In a Seldom, CeAD can occur as a complication of systematic review and meta-analysis, migraine known, rare inherited connective tissue disorders, doubled the risk of CeAD [48]. Migraine is also a mostly vascular Ehlers–Danlos syndrome, probably risk factor for ischemic stroke of other causes [49], in less than 2% of the cases according to large but it is even more common in CeAD patients published series [25]. Vascular Ehlers–Danlos syn- compared with age-matched and sex-matched drome (vEDS) is a rare autosomal dominant disease, patients with non-CeAD IS (36 vs. 27%) [50&&]. due to a mutation in the COL3A1 gene (prevalence: Whereas migraine with aura is more strongly associ- 0.2–1/100 000) [26]. The diagnosis is suggested clin- ated with ischemic stroke in general, migraine ically by the presence of arterial rupture, intestinal without aura is more commonly associated with or uterine rupture, and family history of vEDS [55]. CeAD. Frequency of stroke, dissection site (carotid Other features include easy bruising, thin skin with or vertebral), clinical features, and outcome do not visible veins, and characteristic facial features. In appear to differ between CeAD patients with and addition, the diagnosis must be confirmed by the without migraine [50&&]. The mechanisms under- demonstration of either a mutation in the COL3A1 lying the association between CeAD and migraine gene or an abnormal type III procollagen synthesis. are unclear. Vascular mechanisms are thought to A positive diagnosis of vEDS has important implica- play a key role in the pathophysiology of migraine tions, as vEDS patients are at increased risk of vessel [51]. A common predisposition to migraine and rupture secondary to endovascular investigations or CeAD, for instance, via endothelial dysfunction procedures, at high bleeding risk under anticoagu- [38,52] and shared genetic determinants, could be lants, and celiprolol is recommended for the pre- hypothesized. vention of vascular complications [56]. Interestingly, although they largely overlap, the The occurrence of CeAD in patients with MFS or risk factor profile and structural features of CeAD are other known inherited connective tissue disorders, not identical for carotid and vertebral dissection, such as Loeys–Dietz syndrome or osteogenesis suggesting that their pathophysiology may partly imperfecta, seems even more rare than in patients differ [6]. Patients with carotid dissection are older with vEDS [25,27,57]. In MFS, CeAD may

24 www.co-neurology.com Volume 27 Number 1 February 2014 Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Pathophysiology and risk factors of cervical artery dissection Debette occasionally occur as an extension of a proximal recently been applied to a number of complex dis- aortic dissection into the brachiocephalic arteries. eases, including stroke, with notable successes Although there are a few individual reports of CeAD [71&,72,73]. in MFS patients occurring independently of aortic lesions (<1% of CeAD patients in large series), infor- mation of the criteria used for the diagnosis of MFS is OVERLAP WITH OTHER VASCULAR usually not available [25]. DISEASES Familial cases of CeAD are rare (<2.5% in pub- Despite its low incidence, CeAD shows an intriguing lished series) and usually do not appear to occur in correlation with other vascular or neurological dis- the context of a known inherited connective tissue eases [74&&]. disorder [58]. Linkage analyses have not identified CeAD appears to be associated more often than any significant linkage peak to date, but their power predicted by chance with RCVS, a rare, underdiag- was limited [57]. nosed cause of severe headache [75&&]. This syndrome Although CeAD is sporadic in most cases, one is characterized by segmental constriction of cerebral cannot formally exclude that some patients suffer arteries resolving within 3 months, and may be from a Mendelian disorder with incomplete pene- revealed by recurrent thunderclap headaches, seiz- trance, as could be suggested by the inherited abnor- ures, strokes, and nonaneurysmal subarachnoid malities of skin connective tissue seen in otherwise hemorrhage [75&&]. Histological examination of brain healthy relatives of CeAD patients [29]. Systematic arteries has been described as normal [75&&]. Putative search for mutations in COL3A1 in a total of 53 triggers of RCVS include exposure to vasoactive drugs CeAD patients [54,59–61], and in TGFBR1 and and the postpartum period [75&&]. In a recent series of TGFBR2 in 56 consecutive CeAD patients [62], ident- 173 RCVS cases and 285 CeAD patients, 20 patients ified potentially deleterious mutations in three (12% of RCVS and 7% of CeAD patients) had both patients without any clinical evidence of vEDS or disorders [76&&]. CeAD patients with RCVS were more Loeys–Dietz syndrome. often women (90%) and more often had vertebral artery dissection than CeAD patients without RCVS (83 vs. approximately 30–40%) [76&&]. A history of Complex forms of cervical artery dissection migraine, mostly without aura, was highly prevalent In most CeAD cases, there is no evidence for an in these patients (60 vs. 36% in CeAD patients with underlying monogenic disease and CeAD seems and without RCVS). Underlying mechanisms of to occur as part of a multifactorial predisposition. RCVS are unknown and reasons for the phenotypic Heritability estimates of apparently sporadic CeAD overlap with CeAD are elusive. Transient failure to are not available. In total, 18 genetic association regulate cerebral arterial tone, with sympathetic studies testing the association of CeAD with candi- overactivity, has been suggested [77]. RCVS could date genetic variants have been published, on be hypothesized to facilitate dissection by increasing relatively small samples [57,63,64]. Of these, five pressure upstream of multiple intracranial stenoses, have reported significant associations with three or by altering vessel wall vasa vasorum. Alternatively, different candidate genes: ICAM-1 (rs5498) [65]. CeAD could perhaps lead to the release of vasoactive COL3A1 (30UTR 2-bp deletion) [61], and MTHFR substances triggering RCVS [75&&]. (MTHFR-C677T) [66–68]. The first two were not Fibromuscular dysplasia (FMD) is a nonathero- replicated and a meta-analysis supports a modest sclerotic, noninflammatory vascular disease that association between the MTHFR-677TT genotype affects primarily the renal and extracranial carotid and CeAD [57]. However, these studies have been and vertebral arteries, although other arterial beds markedly underpowered, mainly due to the low can be involved [78,79&]. Several histological sub- prevalence of CeAD that made it difficult to reach types exist, the most common being medial fibro- sufficient sample sizes. Moreover, candidate gene plasia (80–90%), defined by alternating areas of association studies are unable to identify novel thinned media and thickened collagen-containing genetic variants involved in unsuspected pathways, medial ridges [80]. The vast majority of patients are as they are based on what is already known or women (90%) with a mean age at diagnosis of suspected about the pathophysiology of the disease approximately 52 years, and a high prevalence of [69]. A large, multicenter genome-wide association hypertension [81]. The disease is often considered to study (GWAS) of CeAD is currently underway, as be rare, but the prevalence of renal FMD was part of the CADISP consortium [70]. GWAS consist reported to be around 4% in kidney donors [82]. of genotyping large numbers of genetic variants In the largest published registry, on 447 patients distributed across the chromosomes without requir- with FMD, 18.3% of the patients suffered from ing any a-priori hypothesis. This approach has arterial dissection, the most common dissection site

1350-7540 ß 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins www.co-neurology.com 25 Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Cerebrovascular disease

being carotid arteries (13.0% of FMD patients) fol- Conflicts of interest lowed by renal arteries (3.6%), and vertebral arteries There are no conflicts of interest. (3.4%). Multiple dissections were slightly more common than in large CeAD series (18.5 vs. 15%). The frequency of FMD in CeAD registries varies widely, REFERENCES AND RECOMMENDED ranging between 5.6% in a large recent multicenter READING & Papers of particular interest, published within the annual period of review, have series of 983 CeAD patients [83 ], and 16.5% in an been highlighted as: older single center series of 102 CeAD patients [84], & of special interest && or even 21% in one of the first CeAD series on 62 of outstanding interest patients [85]. Of note, in the latter two series most 1. Debette S, Leys D. Cervical-artery dissections: predisposing factors, diagnosis, and outcome. Lancet Neurol 2009; 8:668–678. patients were investigated with digital subtraction 2. Leys D, Bandu L, Henon H, et al. Clinical outcome in 287 consecutive young angiography (DSA) [84,85] vs. less than 4% of the adults (15 to 45 years) with ischemic stroke. Neurology 2002; 59:26–33. & 3. Putaala J, Metso AJ, Metso TM, et al. Analysis of 1008 consecutive patients patients in the former [83 ], as DSA has now been aged 15 to 49 with first-ever ischemic stroke: the Helsinki young stroke replaced by less invasive methods for the diagnosis registry. Stroke 2009; 40:1195–1203. 4. Lee VH, Brown RD Jr, Mandrekar JN, Mokri B. Incidence and outcome of of CeAD, mainly magnetic resonance angiography cervical artery dissection: a population-based study. Neurology 2006; (MRA) or computed tomography angiography (CTA). 67:1809–1812. 5. Be´jot Y, Daubail B, Debette S, et al. Incidence and outcome of cerebrovas- However, these methods, especially MRA, may be less cular events related to cervical artery dissection: the Dijon Stroke Registry. Int sensitive for the diagnosis of FMD. One series also J Stroke 2013; doi: 10.1111/ijs.12154. [Epub ahead of print] 6. Debette S, Grond-Ginsbach C, Bodenant M, et al. Differential features of reported a higher rate of CeAD recurrence in patients carotid and vertebral artery dissections: the CADISP study. Neurology 2011; with FMD, requiring confirmation in independent 77:1174–1181. 7. Arnold M, Kappeler L, Georgiadis D, et al. Gender differences in spontaneous samples [84]. The cause of FMD remains unknown. cervical artery dissection. Neurology 2006; 67:1050–1052. Cigarette smoking and a history of hypertension are 8. Huang YC, Chen YF, Wang YH, et al. Cervicocranial arterial dissection: experience of 73 patients in a single center. Surg Neurol 2009; 72 (Suppl 2): associated with the disease and a genetic predisposi- S20–S27. tion is suspected [86,87]. The mechanisms under- 9. Nakajima S, Tsukahara T, Minematsu K. A study of vertebrobasilar artery dissection with subarachnoid hemorrhage. Acta Neurochir Suppl 2010; lying the association between CeAD and FMD are 107:45–49. unclear. FMD could predispose to CeAD via the arte- 10. Metso AJ, Metso TM, Debette S, et al. Gender and cervical artery dissection. & Eur J Neurol 2012; 19:594–602. rial wall weakening that it entails, but common risk This study describes gender-specific differences in the characteristics of CeAD. factors, including genetic, could also contribute to 11. Arnold M, Fischer U, Bousser MG. Treatment issues in spontaneous cervi- cocephalic artery dissections. Int J Stroke 2011; 6:213–218. the phenotypic overlap. 12. Engelter ST, Brandt T, Debette S, et al. Antiplatelets versus anticoagulation in cervical artery dissection. Stroke 2007; 38:2605–2611. 13. Lyrer P, Engelter S. Antithrombotic drugs for carotid artery dissection. Cochrane Database Syst Rev 2010; 10:CD000255. CONCLUSION 14. Kennedy F, Lanfranconi S, Hicks C, et al. Antiplatelets vs anticoagulation for & dissection: CADISS nonrandomized arm and meta-analysis. Neurology 2012; Collaborative projects enabling the analysis of large 79:686–689. cohorts of patients have been helpful in recent years This is a progress report from the CADISS (Cervical Artery Dissection in Stroke Study), testing the feasibility of a large-scale randomized trial at the acute phase in improving our understanding of the pathophysi- of CeAD, comparing antiplatelets to anticoagulation. Results from the non- ology and risk factors of CeAD, a major cause of randomized arm are reported. 15. Zinkstok SM, Vergouwen MD, Engelter ST, et al. Safety and functional ischemic stroke in young adults. However, the mol- outcome of thrombolysis in dissection-related ischemic stroke: a meta-ana- ecular mechanisms underlying this disease are still lysis of individual patient data. Stroke 2011; 42:2515–2520. 16. Qureshi AI, Chaudhry SA, Hassan AE, et al. Thrombolytic treatment of poorly understood. Ongoing collaborative and hy- patients with acute ischemic stroke related to underlying arterial dissection pothesis-free genetic projects will hopefully provide in the United States. Arch Neurol 2011; 68:1536–1542. 17. Engelter ST, Dallongeville J, Kloss M, et al. Thrombolysis in cervical artery novel insight into the biological substrate of CeAD. & dissection: data from the Cervical Artery Dissection and Ischaemic Stroke Refining the analysis of the phenotypic overlap Patients (CADISP) database. Eur J Neurol 2012; 19:1199–1206. This observational study reports on the outcome of patients with CeAD treated or between CeAD and other vascular and neurological not with thrombolysis, using a propensity score approach. disorders, and assessing the genetic overlap between 18. Arauz A, Ruiz A, Pacheco G, et al. Aspirin versus anticoagulation in intra- and extracranial vertebral artery dissection. Eur J Neurol 2013; 20:167–172. these may provide additional useful insight into the 19. Schievink WI, Mokri B, O’Fallon WM. Recurrent spontaneous cervical-artery pathophysiology of CeAD. Dissecting the mechan- dissection. N Engl J Med 1994; 330:393–397. 20. Tzourio C, Cohen A, Lamisse N, et al. Aortic root dilatation in patients with isms of CeAD is an essential step in order to optimize spontaneous cervical artery dissection. Circulation 1997; 95:2351–2353. treatment and prevention strategies. 21. Calvet D, Boutouyrie P, Touze E, et al. Increased stiffness of the carotid wall material in patients with spontaneous cervical artery dissection. Stroke 2004; 35:2078–2082. Acknowledgements 22. Baracchini C, Tonello S, Vitaliani R, et al. Vasomotion in multiple spontaneous cervical artery dissections. Stroke 2008; 39:1148–1151. The author wishes to acknowledge her colleagues from 23. Barbour PJ, Castaldo JE, Rae-Grant AD, et al. Internal carotid artery redun- the Cervical Artery Dissection in Ischemic Stroke Patients dancy is significantly associated with dissection. Stroke 1994; 25:1201– 1206. (CADISP) group and the CADISP-plus consortium for the 24. Volker W, Dittrich R, Grewe S, et al. The outer arterial wall layers are primarily very enriching collaborations. S.D. receives support from affected in spontaneous cervical artery dissection. Neurology 2011; 76:1463–1471. a Chaire d’Excellence grant from the French National 25. Grond-Ginsbach C, Debette S. The association of connective tissue dis- Research Agency (ANR). orders with cervical artery dissections. Curr Mol Med 2009; 9:210–214.

26 www.co-neurology.com Volume 27 Number 1 February 2014 Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Pathophysiology and risk factors of cervical artery dissection Debette

26. Germain DP. Ehlers-Danlos syndrome type IV. Orphanet J Rare Dis 2007; 55. Beighton P, De Paepe A, Steinmann B, et al. Ehlers-Danlos syndromes: 2:32. revised nosology, Villefranche, 1997. Ehlers-Danlos National Foundation 27. Rodrigues VJ, Elsayed S, Loeys BL, et al. Neuroradiologic manifestations of (USA) and Ehlers-Danlos Support Group (UK). Am J Med Genet 1998; Loeys-Dietz syndrome type 1. AJNR Am J Neuroradiol 2009; 30:1614– 77:31–37. 1619. 56. Ong KT, Perdu J, De Backer J, et al. Effect of celiprolol on prevention of 28. Brandt T, Orberk E, Weber R, et al. Pathogenesis of cervical artery dissec- cardiovascular events in vascular Ehlers-Danlos syndrome: a prospective tions: association with connective tissue abnormalities. Neurology 2001; randomised, open, blinded-endpoints trial. Lancet 2010; 376:1476–1484. 57:24–30. 57. Debette S, Markus HS. The genetics of cervical artery dissection: a systema- 29. Wiest T, Hyrenbach S, Bambul P, et al. Genetic analysis of familial connective tic review. Stroke 2009; 40:e459–e466. tissue alterations associated with cervical artery dissections suggests locus 58. Schievink WI, Mokri B, Piepgras DG, Kuiper JD. Recurrent spontaneous heterogeneity. Stroke 2006; 37:1697–1702. arterial dissections: risk in familial versus nonfamilial disease. Stroke 1996; 30. Rubinstein SM, Peerdeman SM, van Tulder MW, et al. A systematic review 27:622–624. of the risk factors for cervical artery dissection. Stroke 2005; 36:1575– 59. Kuivaniemi H, Prockop DJ, Wu Y, et al. Exclusion of mutations in the gene for 1580. type III collagen (COL3A1) as a common cause of intracranial aneurysms or 31. Dittrich R, Rohsbach D, Heidbreder A, et al. Mild mechanical traumas are cervical artery dissections: results from sequence analysis of the coding possible risk factors for cervical artery dissection. Cerebrovasc Dis 2007; sequences of type III collagen from 55 unrelated patients. Neurology 1993; 23:275–281. 43:2652–2658. 32. Engelter ST, Grond-Ginsbach C, Metso TM, et al. Cervical artery dissection: 60. van den Berg JS, Limburg M, Kappelle LJ, et al. The role of type III collagen & trauma and other potential mechanical trigger events. Neurology 2013; in spontaneous cervical arterial dissections. Ann Neurol 1998; 43:494– 80:1950–1957. 498. This study compares the frequency of recent cervical trauma between patients with 61. von Pein F, Valkkila M, Schwarz R, et al. Analysis of the COL3A1 gene in CeAD and two sets of controls and examines whether the characteristics of the patients with spontaneous cervical artery dissections. J Neurol 2002; dissection differ according to the presence of a recent trauma. 249:862–866. 33. Grau AJ, Brandt T, Buggle F, et al. Association of cervical artery dissection 62. Pezzini A, Drera B, Del Zotto E, et al. Mutations in TGFBR2 gene cause with recent infection. Arch Neurol 1999; 56:851–856. spontaneous cervical artery dissection. J Neurol Neurosurg Psychiatry 2011; 34. Guillon B, Berthet K, Benslamia L, et al. Infection and the risk of spontaneous 82:1372–1374. cervical artery dissection: a case-control study. Stroke 2003; 34:e79–e81. 63. Buss A, Pech K, Roelver S, et al. Functional polymorphisms in matrix 35. Genius J, Dong-Si T, Grau AP, Lichy C. Postacute C-reactive protein levels metalloproteinases-1, -3, -9 and -12 in relation to cervical artery dissection. are elevated in cervical artery dissection. Stroke 2005; 36:e42–e44. BMC Neurol 2009; 9:40. 36. Grond-Ginsbach C, Giossi A, Aksay SS, et al. Elevated peripheral leukocyte 64. Jara-Prado A, Alonso ME, Martinez Ruano L, et al. MTHFR C677T, FII & counts in acute cervical artery dissection. Eur J Neurol 2013; 20:1405– G20210A, FV Leiden G1691A, NOS3 intron 4 VNTR, and APOE epsilon4 1410. gene polymorphisms are not associated with spontaneous cervical artery In this article, the authors compare leukocyte counts in patients admitted at the dissection. Int J Stroke 2010; 5:80–85. acute phase of a CeAD or a stroke of another cause. 65. Longoni M, Grond-Ginsbach C, Grau AJ, et al. The ICAM-1 E469K gene 37. Naggara O, Touze E, Marsico R, et al. High-resolution MR imaging of polymorphism is a risk factor for spontaneous cervical artery dissection. periarterial edema associated with biological inflammation in spontaneous Neurology 2006; 66:1273–1275. carotid dissection. Eur Radiol 2009; 19:2255–2260. 66. Pezzini A, Grassi M, Del Zotto E, et al. Migraine mediates the influence of 38. Lucas C, Lecroart JL, Gautier C, et al. Impairment of endothelial function in C677T MTHFR genotypes on ischemic stroke risk with a stroke-subtype patients with spontaneous cervical artery dissection: evidence for a general effect. Stroke 2007; 38:3145–3151. arterial wall disease. Cerebrovasc Dis 2004; 17:170–174. 67. Pezzini A, Del Zotto E, Archetti S, et al. Plasma homocysteine concentration, 39. Pezzini A, Caso V, Zanferrari C, et al. Arterial hypertension as risk factor C677T MTHFR genotype, and 844ins68 bp CBS genotype in young adults for spontaneous cervical artery dissection. A case-control study. J Neurol with spontaneous cervical artery dissection and atherothrombotic stroke. Neurosurg Psychiatry 2006; 77:95–97. Stroke 2002; 33:664–669. 40. Arnold M, Pannier B, Chabriat H, et al. Vascular risk factors and morphometric 68. Arauz A, Hoyos L, Cantu C, et al. Mild hyperhomocysteinemia and low folate data in cervical artery dissection: a case-control study. J Neurol Neurosurg concentrations as risk factors for cervical arterial dissection. Cerebrovasc Dis Psychiatry 2009; 80:232–234. 2007; 24:210–214. 41. Debette S, Metso T, Pezzini A, et al. Association of vascular risk factors with 69. Zondervan KT, Cardon LR. Designing candidate gene and genome-wide cervical artery dissection and ischemic stroke in young adults. Circulation case-control association studies. Nat Protoc 2007; 2:2492–2501. 2011; 123:1537–1544. 70. Dichgans M, Markus HS. Genetic association studies in stroke: metho- 42. Golledge J, Eagle KA. Acute aortic dissection. Lancet 2008; 372:55–66. dological issues and proposed standard criteria. Stroke 2005; 36:2027– 43. Norman PE, Davis TM, Le MT, Golledge J. Matrix biology of abdominal aortic 2031. aneurysms in diabetes: mechanisms underlying the negative association. 71. Traylor M, Farrall M, Holliday EG, et al. Genetic risk factors for ischaemic Connect Tissue Res 2007; 48:125–131. & stroke and its subtypes (the METASTROKE Collaboration): a meta-analysis of 44. Shantikumar S, Ajjan R, Porter KE, Scott DJ. Diabetes and the abdominal genome-wide association studies. Lancet Neurol 2012; 11:951–962. aortic aneurysm. Eur J Vasc Endovasc Surg 2010; 39:200–207. This is the largest GWAS of ischemic stroke to date on over 12 000 patients, 45. Kloss M, Metso A, Pezzini A, et al. Towards understanding seasonal variability showing that most genetic risk variants for ischemic stroke are probably associated & in cervical artery dissection (CeAD). J Neurol 2012; 259:1662–1667. with individual subtypes, providing strong arguments for performing a GWAS of Here, the authors explore the potential mechanisms underlying the seasonal CeAD, a very well defined, important subtype of ischemic stroke in young and variability of CeAD. middle-aged adults. 46. Tzourio C, Benslamia L, Guillon B, et al. Migraine and the risk of cervical artery 72. Ikram MA, Seshadri S, Bis JC, et al. Genome wide association studies of dissection: a case-control study. Neurology 2002; 59:435–437. stroke. N Engl J Med 2009; 360:1718–1728. 47. Pezzini A, Granella F, Grassi M, et al. History of migraine and the risk of 73. Holliday EG, Maguire JM, Evans TJ, et al. Common variants at 6p21.1 are spontaneous cervical artery dissection. Cephalalgia 2005; 25:575–580. associated with large artery atherosclerotic stroke. Nat Genet 2012; 48. Rist PM, Diener HC, Kurth T, Schurks M. Migraine, migraine aura, and cervical 44:1147–1151. artery dissection: a systematic review and meta-analysis. Cephalalgia 2011; 74. Southerland AM, Meschia JF, Worrall BB. Shared associations of nonathero- 31:886–896. && sclerotic, large-vessel, cerebrovascular arteriopathies: considering intracra- 49. Schurks M, Rist PM, Bigal ME, et al. Migraine and cardiovascular disease: nial aneurysms, cervical artery dissection, moyamoya disease and systematic review and meta-analysis. BMJ 2009; 339:b3914. fibromuscular dysplasia. Curr Opin Neurol 2013; 26:13–28. 50. Metso TM, Tatlisumak T, Debette S, et al. Migraine in cervical artery dissection In this elegant review, the authors discuss the phenotypic overlap between various && and ischemic stroke patients. Neurology 2012; 78:1221–1228. nonatherosclerotic, large vessel, cerebrovascular arteriopathies, including CeAD. In this article, the authors compare the frequency of migraine, overall and by 75. Ducros A. Reversible cerebral vasoconstriction syndrome. Lancet Neurol subtype, between patients with CeAD and controls with an ischemic stroke of && 2012; 11:906–917. another cause. They also examine whether the characteristics and outcome of This is a landmark review on the pathophysiology, risk factors, diagnosis and CeAD differ between patients with and without migraine. management of RCVS. 51. Kurth T, Mohamed S, Maillard P, et al. Headache, migraine, and structural 76. Mawet J, Boukobza M, Franc J, et al. Reversible cerebral vasoconstriction brain lesions and function: population based Epidemiology of Vascular && syndrome and cervical artery dissection in 20 patients. Neurology 2013; Ageing-MRI study. BMJ 2011; 342:c7357. 81:821–824. 52. Lee ST, Chu K, Jung KH, et al. Decreased number and function of endothelial In this article, the authors describe the largest series of patients with both CeAD progenitor cells in patients with migraine. Neurology 2008; 70:1510–1517. and RCVS. 53. Etchevers HC, Vincent C, Le Douarin NM, Couly GF. The cephalic neural 77. Calabrese LH, Dodick DW, Schwedt TJ, Singhal AB. Narrative review: crest provides pericytes and smooth muscle cells to all blood vessels of the reversible cerebral vasoconstriction syndromes. Ann Intern Med 2007; face and forebrain. Development 2001; 128:1059–1068. 146:34–44. 54. Martin JJ, Hausser I, Lyrer P, et al. Familial cervical artery dissections: clinical, 78. Touze E, Oppenheim C, Trystram D, et al. Fibromuscular dysplasia of cervical morphologic, and genetic studies. Stroke 2006; 37:2924–2929. and intracranial arteries. Int J Stroke 2010; 5:296–305.

1350-7540 ß 2014 Wolters Kluwer Health | Lippincott Williams & Wilkins www.co-neurology.com 27 Copyright © Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited. Cerebrovascular disease

79. Persu A, Touze E, Mousseaux E, et al. Diagnosis and management of 83. Be´jot Y, Aboa-Eboule´ C, Debette S, et al. Characteristics and outcomes of & fibromuscular dysplasia: an expert consensus. Eur J Clin Invest 2012; & patients with multiple cervical artery dissection. Stroke 2013; in press. 42:338–347. In this article, the authors compared the features of patients with multiple vs. single This article is providing an expert consensus on the diagnosis and management of artery CeAD, and describe an increased frequency of FMD in the former. FMD, a nonatherosclerotic arteriopathy predisposing to cervical artery dissection. 84. de Bray JM, Marc G, Pautot V, et al. Fibromuscular dysplasia may herald 80. Olin JW, Sealove BA. Diagnosis, management, and future developments of symptomatic recurrence of cervical artery dissection. Cerebrovasc Dis 2007; fibromuscular dysplasia. J Vasc Surg 2011; 53:826.e1–836.e1. 23:448–452. 81. Olin JW, Froehlich J, Gu X, et al. The United States Registry for Fibromuscular 85. d’Anglejan Chatillon J, Ribeiro V, Mas JL, et al. Dissection of the extracranial Dysplasia: results in the first 447 patients. Circulation 2012; 125:3182– internal carotid artery. 62 cases. Presse Med 1990; 19:661–667. 3190. 86. Perdu J, Boutouyrie P, Bourgain C, et al. Inheritance of arterial lesions in renal 82. Blondin D, Lanzman R, Schellhammer F, et al. Fibromuscular dysplasia in living fibromuscular dysplasia. J Hum Hypertens 2007; 21:393–400. renal donors: still a challenge to computed tomographic angiography. Eur J 87. Slovut DP, Olin JW. Fibromuscular dysplasia. N Engl J Med 2004; Radiol 2010; 75:67–71. 350:1862–1871.