Intra-arterial thrombolysis is delivered through an endovascular catheter proximal to the site of arterial occlusion or from within the thrombus (Fig. 2.2). The main advantage of intra-arterial thrombolysis is that an elevated dose of thrombolytic agent can be delivered at the site of occlusion while reducing the systemic dose of drug compared to intravenous therapy. Intra-arterial therapy was effective when delivered up to 6 h following stroke onset, hence it may be utilized in patients who present outside of the 4.5-h window for intravenous rtPA therapy. However, this technique requires an interventional neuroradiology team to place the catheter in the occluded arterial branch prior to thrombolysis, potentially delaying therapy onset. No trial has directly compared intravenous and intra-arterial therapy to date.

The evidence supporting intra-arterial therapy includes two Phase III randomized controlled trials, the North American Prolyse in Acute Cerebral Thromboembolism study part II (PROACT-II) and the Japanese Middle Cerebral Artery Embolism Local Fibrinolytic Intervention Trial (MELT) [26, 27]. PROACT-II utilized 9 mg of recombinant pro-urokinase or heparin (control group) given as an intra-arterial infusion over 2 h in cases of middle cerebral artery (MCA) occlusion presenting less than 6 h following stroke onset [26]. In this study, 40 % of treatment patients reached a modified Rankin score (mRS) of <2 % vs. 25 % of control patients. Recanalization rates were 66 % in the treatment group vs. 18 % in controls. Symptomatic intracranial hemorrhage was higher in the treatment group at 10 % vs. 2 % for controls; however, mortality was equal in the two groups. The MELT trial utilized urokinase delivered in bolus doses every 5 min until recanalization was achieved in cases of MCA occlusion presenting within 6 h of stroke onset [27]. The control arm of the MELT trial received standard medical therapy that did not include intravenous rtPA as it was not approved in Japan during the trial. The MELT trial was halted early when intravenous rtPA was approved for use in acute ischemic stroke in Japan. The subsequent analysis of the truncated dataset was promising but did not reach significance for the primary endpoint, mRS 0–2 at 90 days, where 49 % of the treatment arm reached this endpoint vs. 39 % in the control group. However, secondary analyses were significantly improved including excellent outcome (mRS 0–1 at 90 days) achieved in 42 % of treatment patients vs. 23 % of controls and a National Institutes of Health Stroke Scale (NIHSS) score of 0–1 at 90 days was 5.3 % in the treatment arm and 3.5 % in the control arm (P = 0.017). Rates of intracerebral hemorrhage (ICH) were 9 % in treatment patients and 2 % in controls.

No thrombolytic therapy has received regulatory approval for intra-arterial use in the United States. Therefore, the use of intra-arterial thrombolytics is considered off label [28]. Furthermore, pro-urokinase is not clinically available in the United States and urokinase has not been approved for use in acute ischemic stroke, leaving rtPA as the thrombolytic of choice for intra-arterial thrombolysis in most centers. The American College of Chest Physicians has included intra-arterial rtPA as an option for acute ischemic stroke <6 h in an MCA distribution or in the basilar distribution in centers with expertise in endovascular therapy [28].

Combination therapy with intravenous rtPA followed by intra-arterial thrombolysis is a strategy that provides the benefit of early intravenous thrombolytic dosing followed by intra-arterial administration of thrombolytic drug directly to the thrombus [29]. This strategy has been studied in the Interventional Management of Stroke (IMS) II trial comparing reduced dose (0.6 mg/kg) rtPA intravenous followed by intra-arterial rtPA (maximum dose 22 mg) to historical controls from the NINDS study [30]. IMS II patients had a mortality of 16 % vs. 24 % in the NINDS placebo arm. Furthermore, all functional measures were improved in IMS II patients compared to NINDS placebo patients and the rate of intracerebral hematoma was not different between the IMS II cohort and the intravenous rtPA arm of NINDS. Combination strategies utilizing various dosing of intravenous and intra-arterial thrombolytics have been reported in noncontrolled studies with promising results [31–34]. A Phase III trial (IMS III) comparing standard intravenous rtPA treatment vs. reduced dose intravenous rtPA followed by intra-arterial rtPA for patients presenting within 3 h of stroke onset is ongoing (NCT00359424, Clinicaltrials.gov).

Mechanical thrombolysis and embolectomy are strategies that employ an intra-arterial device to either physically break down or secure and retrieve a clot from the occluded artery (Figs. 2.2 and 2.3) [35, 36]. These therapies may be combined with stenting to maintain recanalization after embolectomy [37]. Generally, these therapies are combined with intravenous and/or intra-arterial thrombolytics if the stroke presents within 4.5 h of onset or without pharmacologic thrombolysis in later presentations. The techniques employed in mechanical thrombolysis include clot disruption with an endovascular wire or sonothrombolysis where ultrasound pulses are delivered with thrombolytics at the tip of the catheter [38]. Embolectomy can be achieved with wire loops to capture emboli, retrievable stent strategies (Fig. 2.3e) or aspiration catheters that suction clot from the vessel (Fig. 2.2e) [38]. Devices to achieve vessel recanalization have largely been approved based on safety and recanalization results in large animal models and in small human studies, limiting functional outcome measures in support of these strategies. The Mechanical Embolectomy in Acute Ischemic Stroke (MERCI) trial is the main evidence supporting mechanical clot retrieval [39, 40]. The MERCI device is an intra-arterial coiled wire that is “corkscrewed” around or into the clot to trap it and then used to physically retrieve the clot by pulling it from the parent artery. In the MERCI trial patients presenting within 3 h of stroke onset in any vascular distribution were given intravenous rtPA at standard doses. If the occlusion did not resolve on follow-up angiography, then the MERCI device was utilized. The device resulted in a 57 % recanalization rate that increased to 70 % with additional intra-arterial rtPA. An mRS outcome of 0–2 was achieved in 36 % of patients, the mortality rate was 34 % and the risk of procedure-related morbidity was 5.5 %. In summary, mechanical means of recanalization show promise in early clinical trials and warrant further study.

In vascular distributions other than the MCA, there is less evidence to guide the use of intra-arterial thrombolytic therapy and/or mechanical thrombolysis. In multiple case series of intra-arterial therapy for basilar artery occlusion, a potentially positive effect has been demonstrated when recanalization is achieved [41–44]. Stent-assisted recanalization of basilar occlusion has been reviewed in case series and has been shown to achieve improved recanalization rates with an acceptable risk profile up to 8 h after stroke onset [45, 46]. Randomized controlled trials for basilar occlusion might clarify the role for these treatments in this indication; however, such a trial would be difficult to execute given the poor outcomes experienced by these patients, the relatively infrequent presentation of these cases, and the lack of a standard therapy to act as a control.

In cases where reperfusion fails to reverse the evolution of stroke or where stroke progresses despite recanalization, there is a risk for malignant edema and elevated intracranial pressure (ICP) that may be life threatening (Fig. 2.4). Mortality in this cohort of patients has been as high as 78 % [47]. In severe cases there is a role for decompressive craniectomy to relieve brain shift, herniation, elevated ICP, and death. For malignant MCA stroke, the procedure generally includes a large craniectomy, durotomy, and subtemporal decompression (Fig. 2.4e, f). Results for decompressive craniectomy in case series and randomized trials demonstrate a significant decrease in mortality after this procedure [48, 49]. Patient age is the only definitive predictive factor of outcome after decompressive hemicraniectomy for stroke, where younger patients have better functional outcomes [50]. Other factors including timing of surgery, side of stroke, and the presence of herniation have not proven to be significant predictors of outcome. A potential downside of decompressive craniectomy for stroke is that the procedure saves lives but leaves patients in severely disabled or persistent vegetative states [51, 52]. Both the benefits and potential harms of this procedure must be considered carefully in the context of a specific case before treatment is undertaken.

In the case of cerebellar infarction, there is a risk of direct brainstem compression and hydrocephalus if swelling occurs (Fig. 2.5) [53]. Suboccipital craniectomy, in addition to expansion duroplasty, can be used to accommodate cerebellar swelling. In severe cases of cerebellar edema, infarcted tissue can be excised to provide more room for swelling (Fig. 2.5e, f). Early craniectomy in these cases may be justified to prevent impending herniation or brainstem compression.

Hemorrhagic stroke is broadly organized into ICH that involves bleeding into the parenchyma of the brain (Fig. 2.6), SAH that involves bleeding in the subarachnoid space from a ruptured vessel and intraventricular hemorrhage (IVH) where bleeding occurs into the ventricular system. All three can be caused by trauma, but spontaneous cases are classified as strokes. In all three cases the acute management relates to the avoidance of secondary deterioration to preserve neurological function remaining after the primary injury. The common etiologies of spontaneous ICH include hypertensive hemorrhages, amyloid angiopathy, coagulopathy or anticoagulation therapy-related hemorrhages, bleeding into ischemic stroke, bleeding from ruptured vascular malformations, and bleeding from tumors. As discussed in the introduction there is no convincing evidence for an ischemic penumbra surrounding an ICH, although there may be a metabolic penumbra that promotes cell death in the subacute phase [54]. Generally, secondary deterioration in ICH relates to enlargement of the clot and compression of adjacent structures. Therefore the central tenet of treatment of ICH is prevention of clot expansion and treatment of the underlying source of bleeding and/or decompression of the hematoma in cases where ICP is elevated.

Spontaneous SAH is generally related to ruptured brain aneurysms (about 85 % of cases), a group of idiopathic cases (about two-thirds being benign perimesencephalic SAH) and occasionally from arteriovenous or other brain vascular malformations. In aneurysmal SAH, there is a risk of re-hemorrhage with clot expansion and a risk of delayed cerebral ischemia during which time there is an increased risk of stroke [55]. Hence, treatment of SAH involves rapid diagnosis and treatment of the source of bleeding and treatment of aneurysmal SAH involves prevention and treatment of angiographic vasospasm, which is the main cause of delayed cerebral ischemia.

IVH is generally related to a preexisting ICH or SAH with extension into the ventricles and is associated with worse outcome in the case of ICH [56]. In IVH, deterioration relates to obstructive or communicating hydrocephalus and elevated ICP necessitating cerebral spinal fluid diversion [57]. The rest of the chapter will focus on the treatment of ICH.