Abstract
Cervical artery dissection (CAD) is characterized by an intramural haematoma due to a subintimal tear and accounts for up to 25% of ischaemic strokes in young and middle-aged adults. Data regarding intravenous thrombolysis and endovascular thrombectomy in CAD are scarce and observational – both are reasonably safe and probably recommended. Based on observational evidence, antithrombotic therapy is used to prevent first or recurrent cerebral ischaaemic events in acute or subacute CAD, and event rates are low with either antiplatelet or anticoagulant therapy. The long-term rate of recurrent cerebral ischaemic events or bleeding complications in CAD patients is small while under antithrombotic treatment. Cerebral vasculitis treatment is based on observational series. When primary angiitis of the central nervous system is confirmed by biopsy, a combination of glucocorticoids and cyclophosphamide should be started. Rituximab may be used in patients who are intolerant of cyclophosphamide. In atypical, non-biopsy-proven cases, treatment should be adapted to the severity of neurological involvement. For giant cell arteritis, initial high-dose prednisolone is recommended, beginning a slow taper after 2–4 weeks and continuing at a low dose for 1–2 years. Treatment of p-ANCA-positive and -negative systemic vasculitis with cerebral involvement includes induction corticosteroid therapy followed by addition of cyclophosphamide or other glucocorticoid-sparing drugs.
Extracranial cervical artery dissection and cerebral vasculitis are uncommon causes of ischaemic stroke. Both may occur at any age. Intracranial artery dissection is even less common and less well defined (Debette et al., 2015).
Cervical Artery Dissection
Cervical artery dissection (CAD) of the internal carotid or the vertebral artery is characterized by an intramural haematoma which is thought to be caused by a subintimal tear into the arterial wall (Engelter et al., 2017). CAD accounts for up to 2.5% of all ischaemic strokes (Debette and Leys, 2009). However, it is a major cause (up to 25%) of ischaemic stroke in young and middle-aged adults (Leys et al., 2002; Nedeltchev et al., 2005). The mean age at occurrence of CAD is about 45 years (Debette et al., 2011a; Bejot et al., 2014). There is a slight male predominance among CAD patients (A. J. Metso et al., 2012), and men are on average 5 years older than women when experiencing CAD (Schievink et al., 1993; Bejot et al., 2014). CAD may occur spontaneously or subsequent to mechanical trigger events (e.g. minor [sports associated] trauma, cervical manipulation or severe [poly-] trauma) (Engelter et al., 2013). Putative risk factors for spontaneous CAD have been identified in large cohort studies. Recent infection (Grau et al., 1999; Kloss et al., 2012), hypertension (Debette et al., 2011b), and migraine (T. M. Metso et al., 2012) have been associated with spontaneous CAD. In rare cases, CAD may occur in patients with hereditary connective tissue disorders, with vascular Ehlers–Danlos syndrome being the most common among these patients (Grond-Ginsbach and Debette, 2009).
CAD may present with ischaemic stroke, transient ischaemic attack, or local symptoms (e.g. Horner’s syndrome, cranial nerve palsy, tinnitus, or cervical root impairment) (Schievink, 2001; Debette et al., 2011a). Most commonly, CAD patients suffer from cervical pain or headache (Schievink, 2001). The mural blood accumulation in CAD may be located subadventitially, thereby causing local compression syndromes (such as Horner’s syndrome) or – in very rare cases of arterial rupture – it may result in subarachnoid haemorrhage. Stenosis or occlusion of the dissected artery may occur due to the intramural haematoma and subsequent arterial narrowing. This may lead to cerebral ischaemic events, which are more often embolic rather than due to haemodynamic compromise (Engelter et al., 2007).
If suspected, the diagnosis of CAD can be confirmed by the presence of at least one of the following, widely used and established neurovascular criteria: Visualization of a mural haematoma, aneurysmal dilatation, long tapering stenosis, intimal flap, double lumen or occlusion >2 cm above the carotid bifurcation revealing an aneurysmal dilatation or a long tapering stenosis after recanalization in the internal carotid or vertebral artery (Debette and Leys, 2009). Both internal carotid and vertebral artery dissection may be visualized by magnetic resonance imaging (MRI), computed tomography (CT), or neurosonography. Specific fat-suppressed T1 sequences in MRI can most accurately depict the mural haematoma in CAD (Figure 21.1). Although neurosonography in general has a lower sensitivity in the diagnosis of CAD, it can detect a mural haematoma at very early stages of the vascular changes (Figure 21.2), when MRI can be falsely negative (Nebelsieck et al., 2009).
Figure 21.2 A: Colour-coded duplex ultrasound of a right internal carotid artery. Arrows indicate the hypoechogenicity of the arterial wall representing the mural haematoma of an acute cervical artery dissection. B: Transverse ultrasound imaging (power-mode) of an acutely dissected left internal carotid artery. Arrows indicate the hypechogenic mural haematoma.
Intravenous Thrombolysis in Cervical Artery Dissection
Acute therapies such as intravenous thrombolysis (IVT) or endovascular recanalization therapy (EVT) have to be considered in CAD patients presenting with ischaemic stroke. Based on the pathophysiology of CAD, there might be the risk of an increasing mural haematoma of the dissected vessel if treated with IVT in the acute setting. This might lead to a haemodynamic worsening and to an infarct growth. However, regarding the existing evidence on IVT in CAD, this seems to be a theoretical concern and there is currently no convincing reason to withhold IVT or EVT in CAD patients. IVT or EVT increase the odds to induce recanalization of an occluded (dissected) artery or of a distal (intracranial) thrombosis in CAD patients, too.
Evidence: Comments
Although established as safe and efficacious in patients with ischaemic stroke from different aetiologies (Emberson et al., 2014; Wardlaw et al., 2014), the evidence for the use of IVT in CAD patients is scarce and based on observational, non-randomized data only. Current guidelines of acute stroke treatment do not recommend against IVT in CAD patients; it is considered reasonably safe within 4.5 hours and is probably recommended (Class IIa; Level of Evidence C) (Jauch et al., 2013; Demaerschalk et al., 2016).
IVT in non-CAD ischaemic stroke patients and in CAD patients was compared in observational, registry-based studies (Engelter et al., 2009; Zinkstok et al., 2011). In one of these studies, CAD patients showed a slightly (but statistically significant after adjustment for age, gender, and stroke severity) lower recovery rate than patients with a stroke attributable to another cause. In this study, only 36% of CAD patients versus 44% of non-CAD patients (odds ratio [OR]adjusted 0.50 [95% confidence interval, CI: 0.27–0.95], p = 0.03) reached an excellent outcome at 3 months (i.e. excellent outcome defined as a modified Rankin Scale [mRS] score of 0 or 1) (Engelter et al., 2009). There was a high rate (67.7%) of CAD patients with a large artery occlusion in this study. Known as a negative prognostic factor in IVT treated stroke patients, this higher rate of large artery occlusion might – at least in part – explain the lower recovery rate of CAD patients. Yet another study compared meta-analysed data from observational studies and case reports of IVT-treated CAD patients with data from age- and stroke-severity matched patient data from the Safe Implementation of Thrombolysis in Stroke-International Stroke Thrombolysis Register (SITS-ISTR) (Zinkstok et al., 2011). In this study, 3-month mortality, the rate of symptomatic intracranial haemorrhage (ICH), and the number of patients reaching excellent 3-month outcome did not differ between IVT-treated CAD and non-CAD patients.
Data on comparisons of CAD patients receiving IVT versus those who did not are scarce. Analyses on the data from the Cervical Artery Dissection and Ischemic Stroke Patients (CADISP) consortium showed identical rates of favourable recovery after CAD related ischaemic stroke in both IVT treated and non-IVT treated patients (ORadjusted 0.95 [95% CI: 0.45–2.00]). A meta-analysis across observational studies (n = 10) identified 174 CAD patients receiving IVT (or some other form of thrombolytic treatment, n = 26) who were compared to 672 CAD patients who did not receive thrombolysis. Most importantly, the odds for achieving a favourable 3-month outcome were similar in thrombolyzed and non-thrombolysed CAD patients (OR 0.782 [95% CI: 0.49–1.33], p = 0.441). Although there was a higher rate of intracranial haemorrhage in thrombolysed patients (OR 2.65 [95% CI: 0.49–1.33], p = 0.042), a symptomatic haemorrhage occurred in one non-thrombolysed patient only (Lin et al., 2016).
Endovascular Therapy in Cervical Artery Dissection
In anterior circulation ischaemic stroke (from any cause) with large vessel occlusion, EVT including mechanical thrombectomy with or without IVT has been shown superior to IVT alone in randomized controlled trials (Berkhemer et al., 2015; Campbell et al., 2015; Goyal et al., 2015; Jovin et al., 2015; Saver et al., 2015). Yet again, data on endovascular therapy specifically in CAD patients are scarce and derived from observational studies only. The endovascular approach seems feasible in CAD although there might be the risk that the false lumen of the dissected artery is chosen for recanalization therapy.
Evidence: Comments
The current evidence on EVT in CAD stroke patients is based on case series and non-randomized, observational studies and should therefore be interpreted very cautiously. In a recent study comparing 38 CAD patients receiving EVT (with or without IVT) to CAD patients receiving IVT, adjusted (age, sex) National Institutes of Health Stroke Scale (NIHSS) excellent outcome (mRS 0–1) was equally frequent in both groups (OR 2.23, 95% CI: 0.52–9.59; p = 0.278) (Traenka et al., 2018). However, partial or complete recanalization of the occluded intracranial artery was (numerically) more frequent in EVT-treated patients (84.2% vs 66.7%) (Traenka et al., 2018). Another study also compared EVT-treated CAD-patients (n = 24) to EVT-treated non-CAD-patients (n = 421) showing no difference in the odds of a favourable 3-month-outcome (OR 0.58 [0.19–1.78], p = 0.34) (Jensen et al., 2017).
In a recent meta-analysis across eight observational studies comparing EVT-treated to IVT-treated CAD patients, the likelihood for a favourable outcome (mRS 0–2) was similar in both groups (OR 0.97 [0.39–2.44], p = 0.96) (Traenka et al., 2018).
Endovascular treatment might be particularly important in patients presenting with tandem occlusion (i.e. occlusion of the dissected artery and a distally located intracranial artery). In a retrospective study of EVT-treated patients, 20 patients with tandem occlusion due to internal carotid artery dissection (ICAD) were compared to non-CAD patients with isolated intracranial artery occlusion. Recanalization rates were similar in both groups (p = 0.23). Likewise, favourable outcome was achieved equally frequent in both groups (CAD-patients 70% vs non-CAD patients 50%, p = 0.093) (Marnat et al., 2016). However, comparisons in this study were not adjusted for confounding variables or differences in baseline characteristics (e.g. stroke severity).
Recurrent Ischaemic Events and Prophylactic Antithrombotic Treatment in CAD
The rate of (recurrent) cerebral ischaemic events or bleeding complications in CAD patients is low while under antithrombotic treatment. In a randomized controlled trial comparing antiplatelet therapy (mostly aspirin) to anticoagulation (mostly warfarin) in CAD patients, the overall rate of ipsilateral (to the dissected artery) ischaemic stroke, death or major bleeding was 2% (4 of 196) in the per-protocol population (CADISS Trial Investigators et al., 2015). There is consensus on the need for any antithrombotic treatment as primary or secondary prophylaxis of (recurrent) cerebral ischaemic events in acute or subacute CAD. Unfortunately, at the current stage, there is still equipoise on the choice of the antithrombotic therapy (anticoagulation or antiplatelets).
Evidence: Comments
There are at least 5 meta-analyses, based on observational data, comparing antiplatelets to anticoagulants in CAD patients (Menon et al., 2008; Lyrer and Engelter, 2010; Kennedy et al., 2012; Sarikaya et al., 2013; Chowdhury et al., 2015). These meta-analyses used different statistical approaches and showed conflicting results. No difference with regard to occurrence of stroke or death was reported by Menon et al. in 2008. A non-significant trend in favour of anticoagulants was reported in a later Cochrane review with regard to the endpoint of death or disability (OR 1.77 [95% CI: 0.98–3.22], p = 0.06) (Lyrer and Engelter, 2010). However, in this analysis major bleeds (symptomatic intracranial haemorrhage [5/627; 0.8%] and major extracranial haemorrhage [7/425; 1.6%]) occurred solely in the anticoagulation group. In turn, Sarikaya et al. showed a beneficial effect of antiplatelets with regard to a composite outcome of ischaemic stroke, intracranial haemorrhage, or death (relative risk [RR] 0.32 [95% CI: 0.12–0.64]). In 2015, the first randomized controlled study comparing antiplatelet treatment to anticoagulants in CAD patients was published. The Cervical Artery Dissection in Stroke Study (CADISS) was designed as a prospective feasibility study randomly assigning CAD patients to either antiplatelet therapy (aspirin, dipyridamole, or clopidogrel alone or in combination) or to anticoagulation therapy (heparin followed by warfarin with a target international normalized ratio [INR] of 2–3) (CADISS TRIAL Investigators et al., 2015). 250 CAD patients, mainly presenting with stroke or transient ischaemic attack (n = 224), were included. With regard to the primary outcome (ipsilateral stroke or death) there was no statistically significant difference between the groups (intention-to-treat population: OR 0.335 [95% CI: 0.006–4.233], p = 0.63). There was one major bleed which occurred in the anticoagulation group. Central reading of the patient baseline imaging confirmed CAD diagnosis in 197 of the 250 study participants. However, the main results of the study did not differ in the per-protocol population. Based on the very low event rates of the purely clinical primary outcome in this study, the authors calculated that 4876 patients per group would be needed to show significant differences between groups. Hence, the use of a surrogate outcome might help to overcome the feasibility issue in a therapy trial in CAD patients. Indeed, there is another prospective, randomized multicentre trial investigating aspirin versus anticoagulation (phenprocoumon) in acute CAD. The “Biomarkers and Antithrombotic Treatment in Cervical Artery Dissection (TREAT-CAD, NCT0204640, www.clinicaltrials.gov) trial uses a composite primary outcome including both clinical and – more importantly – also imaging surrogate outcome measures. New ischaemic lesions on diffusion weighted imaging (DWI) in CAD patients were observed in up to 25% of patients undergoing repeated brain MRI (Gensicke et al., 2015). The TREAT-CAD study started recruitment in 2013. Study recruitment was completed in December 2018 and results are expected soon.
Cerebral Vasculitis
Cerebral or central nervous system (CNS) vasculitis is defined as the presence of inflammation in the brain-supplying arteries or the veins draining the brain, the spinal cord, or the meninges. It is characterized by the presence of leucocytes in the vessel walls – these leucocytes induce damage to mural structures (Jennette et al., 1994). There is a broad spectrum of clinical manifestations of the disease, going along with a variety of pathological entities. Primary systemic vasculitis is caused by an autoimmune response, while secondary systemic and cerebral vasculitis may be induced by several infectious agents, e.g. measles, human immunodeficiency virus, herpes zoster and varicella zoster viruses, Epstein–Barr virus, or bacterial infections such as listeriosis, syphilis, and many others causing vasculitis with involvement of the brain and the meninges (Ferro, 1998). CNS vasculitis may be referred to small, medium, or large vessels of the CNS. Small and large vessels are in most cases affected by autoimmune disorders, while medium-sized vessels are mainly affected by infectious diseases.
Primary nervous system vasculitis is restricted to the nervous system, and may involve the central as well as the peripheral nervous system (single organ vasculitis, idiopathic). Secondary nervous system vasculitis is caused by a systemic disorder or infection known to cause inflammatory vasculopathy with the presence of systemic vasculitis (multi-organ vasculitis) or may be present in a systemic disorder that is restricted to the nervous system without evidence of further systemic vasculitis (single organ vasculitis with known aetiology) (Siva, 2001).
Primary Angiitis of the Central Nervous System (PACNS)
Primary angiitis of the central nervous system (PACNS, Figure 21.3) is a rare disease. Its incidence rate is estimated to be about 0.24/100,000 per year (Salvarani et al., 2007). The lesions are limited to the small or large vessels of the brain and of the spinal cord. Its neurological presentation can be very heterogeneous.
Some patients will experience acute clinical syndromes such as ischaemic stroke, while others may have a presentation such as a diffuse encephalopathy, or even tumour-like symptoms with clinical progression. Brain biopsy may be crucial to make a histopathological diagnosis confirming segmental inflammation of small arteries and arterioles with intimal proliferation and fibrosis. However, the diagnostic yield of brain biopsy for suspected PACNS is modest and accounts for about 11% of biopsied cases (Torres et al., 2016). A wide range of differential diagnoses consisting of other conditions such as hypertensive subcortical arteriolopathy, cerebral amyloid angiopathy, sarcoidosis, primary brain lymphoma, metastatic brain disease, infectious encephalitis, multiple sclerosis, progressive multifocal leucoencephalopathy, or Creutzfeldt–Jakob disease have to be considered (Alrawi et al., 1999; Torres et al., 2016; Salvarani et al., 2017). To make the diagnosis, history of exposure to vasoactive substances, a postpartum state, history of migraine headaches, thunder-clap headaches, or manifestations typical of reversible cerebral vasoconstriction syndrome (RCVS) have to be excluded (Salvarani et al., 2015). The disease may be progressive over years. Five-year fatality is estimated to be 25%, and for those not dying, progressive disability may occur.
Evidence
To date, there are no data about treatment available based on randomized controlled trials (Salvarani et al., 2017). A recently published large observational series suggested basing therapeutic decisions on the fact of small/distal or large/proximal vessel involvement (Salvarani et al., 2015).
Steroid-induced changes of the patients’ mental status may confound the clinical picture of the disease itself, hence some experts suggest avoiding intravenous pulse glucocorticoids. They suggest beginning the patient on oral intake with the equivalent of predisone 1 mg/kg per day (to a maximum of 80 mg/day, or its equivalent) until the diagnostic evaluation is complete. If PACNS can be proven by biopsy, thereby showing a picture of granulomatous inflammation, treatment with a combination of glucocorticoids and cyclophosphamide should be started. Differential diagnoses should be challenged if there is treatment failure with these drugs. Further, rituximab may be used in patients who are intolerant of cyclophosphamide (de Boysson et al., 2014; Salvarani et al., 2015).
The majority of cases in clinical practice will be atypical, non-biopsy-proven PACSN cases. Treatment in these cases should be adapted to the severity and the extent of neurological involvement. These patients should be treated with an initial high dose of glucocorticoids (Salvarani et al., 2015) with a following slow reduction in daily dose. Whether to add cyclophosphamide to the treatment should be decided in each individual case, taking the extent and the severity of the neurological deficits into account. Alternatively, treatment initiation may be done with intravenous methylprednisone, 15 mg/kg each day for 3 days (Guillevin and Pagnoux, 2003). Peroral prednisone shall then be started on day 4. Well-known, possible treatment-related side effects of glucocorticoids (e.g. bone loss or opportunistic infections) have to be considered and might be encountered by prophylactic treatments. As part of a remission induction strategy, cyclophosphamide can be administered as daily oral or intermittent, monthly intravenous therapy (starting dose for oral therapy: 1.5–2 mg/kg/day; 600 to 750 mg/m2). As cyclophosphamide might induce leucopenia, the white blood cell count should be monitored closely. Intravenous cyclophosphamide is infused once a month. A dose reduction of cyclophosphamide is mandatory for patients with an estimated glomerular filtration rate less than 20 mL/min. Data on the use of rituximab in PACNS are scarce (De Boysson et al., 2013a; Salvarani et al., 2014). In three cases, improvement of the condition of the patients (radiological and clinical) was reported. Rituximab was used in a dose of either two infusions (each 1 g) separated by 14 days or as 375 mg/m2 weekly for 4 weeks (de Boysson et al., 2013a; Salvarani et al., 2014).
Comment
The overall incidence of this condition is too low to evaluate a specific therapy against control. Therapeutic recommendations come from case series and are highly empirical. However, there seems to be a clear response rate to immunosuppression with use of steroids. Nevertheless, the proposed doses as well as timed regimens should be evaluated in randomized controlled trials.
Large Artery Vasculitis, Giant Cell Arteritis, and Others
Giant cell arteritis (GCA, Figure 21.4) is the most common vasculitis affecting medium and large cerebral vessels arising from the aortic trunk. GCA occurs at an incidence of 7–18 cases per 100,000 individuals; women are affected twice as often as men. Usually, patients are over 50 years of age at first onset of the disease. GCA was initially described as temporal arteritis (Horton disease), but about 15–27% of patients have extended extracranial involvement, since the entire aorta and all its branches can be affected including the carotid, the subclavian, and the iliac arteries. The most concerning clinical features of the disease are anterior ischaemic optic neuropathy with visual loss and cerebral strokes in the anterior as well as in the posterior circulation (Ruegg et al., 2003) which are the result of vascular inflammation involving cranial arterial branches (Salvarani et al., 2002). Polymyalgia rheumatica (PMR) is an inflammatory disorder that can occur before, and simultaneously with, or develop after, clinical manifestations of GCA. It is two or three times more common than GCA and clinically characterized by girdle pain and stiffness. Population-based studies have shown that PMR occurs in about 50% of patients with GCA, and approximately 15–30% of PMR patients develop GCA (Salvarani et al., 2002; Puppo et al., 2014). In contrast, Takayasu arteritis primarily affects the aorta and its major branches (Figure 21.5). The inflammation and damage are often localized to a portion of the affected vessels, but extensive involvement such as nearly pan-aortitis can be seen. The onset of disease usually occurs before the age of 30 years.
Figure 21.4 Colour coded duplex sonography of a 77-year-old male patient with biopsy-proven GCA. Imaging of the left superficial temporal artery showed hypoechogenic vessel wall on axial (A) and longitudinal (B) artery imaging.
Evidence
Steroids have not been studied against placebo in randomized controlled trials, but the effectiveness is well established by observational studies in which steroids were reported to resolve symptoms. Initial empirical therapy for GCA is recommended as follows: oral prednisolone 40–60 mg daily as a single or divided dose should be started. In cases with recent or impending focal neurological dysfunction, such as visual loss or stroke, pulsed intravenous (i.v.) methylprednisolone 1000 mg every day for the first 3 days may be considered. The oral treatment should be maintained for 2–4 weeks and then gradually reduced every 1–2 weeks by 10% of the total daily dose (about 2.5–5 mg/day) until dose is 10 mg/day. It will be important to frequently monitor clinical symptoms to guide the management. This also comprises the erythrocyte sedimentation rate (ESR) and the C-reactive protein, although they are not always reliable markers of disease activity.
Further maintenance therapy is empirical: prednisolone 5–7.5 mg/day for about 1–2 years should be prescribed. If the patient is asymptomatic and ESR is normal, the dose can be reduced gradually by 1 mg/day every 2–3 months (Myles, 1992; Hayreh et al., 2002; Mazlumzadeh et al., 2006). Adverse effects of steroids are common and related to increasing age at diagnosis, female sex, an initial dose of prednisolone of more than 40 mg/day, a total cumulative dose of at least 2 g of prednisolone, and maintenance doses above 5 mg of prednisolone a day. Calcium and vitamin D supplementation should be given with corticosteroid therapy in all patients. In patients with reduced bone material density, bisphosphonates are indicated (Salvarani et al., 2002). Only sparse data are available on the treatment with further immune-modulating drugs such as methotrexate, azathioprine, cyclophosphamide, or TNF alpha inhibitors. However, limited evidence is also available for the use of biological agents such as tocilizumab, ustekinumab, and abatacept in GCA (Salvarani and Hatemi, 2019). Tocilizumab in combination with prednisolone has recently been tested in a prospective randomized double-blind placebo-controlled trial in 251 GCA patients. Tocilizumab combined with a 26-week prednisolone taper was proven superior to a 26-week or 52-week prednisolone taper and placebo with regard to the primary outcome (sustained remission at week 52) (Stone et al., 2017; Salvarani and Hatemi, 2019). Further evidence from randomized studies with these agents is needed to identify which GCA patients should be treated and for how long they should be treated (Salvarani and Hatemi, 2019). For now, the aforementioned agents may be an opportunity in cases of non-responsiveness to steroids or may be used for their glucocorticoid-sparing effect (Hoffman et al., 2002; Unizony et al., 2012; de Boysson et al., 2013b; Salvarani & Hatemi, 2019). Antiplatelet treatment with acetylic acid 100 mg once daily should be initiated as soon as diagnosis is established. It may prevent visual loss as well as cerebral ischaemic events (de Boysson et al., 2014; Salvarani et al., 2015).
Related posts:
Chapter 5 – Supportive Care during Acute Cerebral Ischaemia
Chapter 12 – Neuroprotection for Acute Brain Ischaemia
Chapter 9 – Acute Antiplatelet Therapy for the Treatment of Ischaemic Stroke and Transient Ischaemic Attack
Chapter 18 – Drugs, Devices, and Procedural Therapies to Prevent Recurrent Cardiogenic Embolic Stroke
Chapter 25 – Using Pharmacotherapy to Enhance Stroke Recovery
Chapter 26 – Electrical and Magnetic Brain Stimulation to Enhance Post-stroke Recovery
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
