Article Text

Download PDFPDF

Endovascular therapy for acute basilar artery occlusion: a review of the literature
  1. A M Mortimer,
  2. M Bradley,
  3. S A Renowden
  1. Department of Neuroradiology, Frenchay Hospital, North Bristol NHS Trust, Bristol, UK
  1. Correspondence to Dr A M Mortimer, Department of Neuroradiology, Frenchay Hospital, Frenchay Park Road, Bristol, BS16 1LE, UK; alex_mortimer{at}


Basilar artery occlusion is an infrequent form of acute stroke; clinical outcomes are heterogeneous, but the condition can be fatal. There is a lack of randomized controlled trial data in this field. Case series suggest that patients who are recanalized have much better outcomes than those who are not, and it is generally accepted that intra-arterial techniques achieve high rates of recanalization. Controversially, several studies, including a meta-analysis and registry-based investigation, that have compared intravenous thrombolysis (IVT) and intra-arterial treatment suggest similar outcomes. However, there are many potential sources of bias in each of these studies, precluding a firm conclusion. Indeed, there are many confounding factors that can influence the outcome including severity of presentation, site of occlusion, clot load, degree of collateral flow, timing of therapy, agent used for recanalization and dose of thrombolytic agent. Additionally, pretreatment infarct core imaging using diffusion-weighted imaging and the posterior circulation Acute Stroke Prognosis Early CT Score (pc-ASPECTS) scoring systems have been shown to predict outcome and therefore may be useful in selecting patients for aggressive therapy. Protocols combining intravenous agents such as glycoprotein IIb/IIIa receptor antagonists or thombolytics agents with intra-arterial techniques (‘bridging’ therapy) have shown encouraging improvements in neurological outcome and survival. Furthermore, initial case series describing the use of mechanical clot extraction devices or aspiration catheters suggest high rates of recanalization. What would be useful is a randomized trial comparing IVT, endovascular approaches and a combined IVT/endovascular approach. However, the small numbers of patients and multiple confounding factors are barriers to the development of such a trial.

  • Acute stroke
  • endovascular
  • basilar occlusion
  • clot extraction
  • thrombolysis
  • stroke
  • thrombectomy

Statistics from


Basilar artery occlusion is an uncommon cause of stroke, accounting for only 6–10% of large vessel strokes.1 The most common underlying mechanism is atherosclerosis, but the basilar artery is also affected by embolism from both cardiac and proximal arterial sources, dissections, migraine and inflammatory conditions.2 The condition presents with a variety of clinical signs and symptoms.3 Most patients suffer prodromal symptoms in the preceding 2 weeks including headache, vertigo or both. In the acute presentation, altered levels of consciousness, dysarthria, paresis, hemiplegia or tetraplegia, ataxia, cranial nerve deficits and supranuclear occulomotor disturbances are the commonest presenting signs. Although prognosis is variable, basilar occlusion has traditionally been thought to be associated with a poor outcome.1 4 5 More recently there has been an emphasis on acute therapeutic recanalization to re-establish perfusion, with several case series suggesting a reduction in mortality and improved functional outcomes following thrombolytic treatment. There is controversy as to the best mode of treatment4–8; both intravenous thrombolysis (IVT) and/or endovascular techniques including intra-arterial thrombolysis (IAT) and mechanical clot extraction are advocated, though each mode of treatment has its proponents.4–8 In this study we explore the literature comparing IVT and IAT and examine the evidence pertaining to the use of endovascular techniques such as mechanical thrombectomy and thromboaspiration used to assist recanalization.

Natural history

The majority of patients presenting with basilar artery occlusion are aged 50–80 years, although women tend to present at an older age than men.2 Significant risk factors for basilar occlusion include hypertension, diabetes mellitus, smoking, hyperlipidemia, coronary and peripheral vascular disease and previous stroke. The majority of patients have transient ischemic attacks preceding the onset of stroke.2 It is difficult to obtain a universal view from the literature regarding the prognosis of this population when managed conservatively. Comparisons across observational studies are difficult because of differences in eligibility criteria and in classifications of favorable outcome. Bearing this in mind, some studies have reported favorable outcomes in 44–77% of patients while others have described poor outcome rates in 54–95% of patients (table 1).2 9–16 The results highlight the fact that prognosis is diverse, and that this condition should be considered as a disease spectrum with various factors influencing the prognosis (see below). Indeed, as the natural history of basilar occlusion is so varied, there is considerable difficulty in assessing the positive and negative impact of treatment; this is probably the source of much of the controversy pertaining to treatment of this condition. What is certain is that the outcome can be very poor and that strategies to select those who will respond to treatment are warranted.

Table 1

Outcome in conservatively treated patients with basilar stroke

Voetsch et al2 assessed outcome relative to etiology of occlusion and presentation in the New England Medical Center Posterior Circulation Registry. This prospectively-collected series of 407 consecutive patients with posterior circulation ischemia identified 55 patients with moderate (50–70%) or severe (>70%) basilar stenosis and 32 patients with basilar occlusion as documented by MR angiography, conventional angiography or transcranial Doppler sonography. Factors that were significantly associated with more severe outcomes by univariate analysis were distal territory involvement, embolic etiology and basilar artery occlusion. One explanation for a poorer outcome in embolic occlusions is that emboli most often lodge in the rostral basilar artery and that these patients have less time to develop collateral circulation to maintain perfusion. Likewise, patients with basilar stenoses who have residual flow are at lower risk of severe outcomes than those with basilar occlusion.

As well as pathological predictors of poor outcome, clinical predictors include decreased level of consciousness, tetraparesis and abnormal pupils. In the New England Medical Center Posterior Circulation Registry2 these clinical signs increased the probability of major disability or death 3–4-fold. Devuyst et al9 found that decreased consciousness was the single most powerful prognostic factor in basilar disease, followed by pupillary disorders, bulbar signs and dysarthria.

Key points

  • Prognosis in conservatively managed patients is varied but most series suggest that it is poor in the majority of patients.

  • There is a spectrum of presentation and outcome.

  • Factors associated with poor outcome in conservatively managed patients include:

    • Occlusion rather than stenosis

    • Embolic etiology

    • Distal basilar artery occlusion

    • Presenting with coma or depressed level of consciousness, tetraparesis, bulbar signs and abnormal pupils.

Intra-arterial thrombolysis (IAT): what is the evidence?

The first description of the use of IAT was published in the early 1980s.17 However, randomized clinical trial data for intra-arterial thrombolytic therapy for basilar artery occlusion are lacking and are limited to a single small study of only 16 patients that was prematurely terminated because of slow recruitment and withdrawal of the sale of urokinase in Australia.18 Patients were randomized to receive treatment with intra-arterial urokinase or control. All received heparin for a minimum of 2 days and then oral warfarin. In the eight patients who received thrombolysis, four died and the modified Rankin scale (mRS) score was 1 in the survivors. In eight patients in the control group, four died and the mRS score was 3 in the survivors. This showed no significant difference in outcome between the two groups although it did suggest a benefit for thrombolytic therapy. It is, however, certainly too small to draw meaningful conclusions and the results are not sufficient to provide randomized clinical trial evidence for or against the general application of IAT.

There have, however, been multiple series describing the use of IAT in the treatment of basilar occlusion. Unfortunately, each series differs in the method and timing of follow-up and method of documentation of favorable or poor outcome. It is therefore problematic to attempt a reliable comparison of clinical outcomes across studies and, for this reason, mortality—a relatively easily comparable outcome measure—has been used as a comparison across studies.

A large meta-analysis1 that incorporated 10 studies including 316 patients reported an overall recanalization rate of 64% and overall mortality of 56%. The mortality was 87% in non-recanalized patients and 37% in recanalized patients, resulting in a 48% absolute risk reduction of death (p<0.001). The rate of symptomatic intracranial hemorrhage in this analysis was only 7%.

In the most recent and the largest series to date,19 the data of 180 adult patients with angiographically-confirmed basilar occlusion treated with IAT at five German stroke centers were retrospectively evaluated. Patients with partial or complete recanalization had a significantly better neurological outcome than non-recanalized patients (p<0.001). For 38 (86%) of 44 patients who had no recanalization, the post-treatment outcome measured using the mRS was 5 or 6; such a poor outcome was present in only 32 (33%) of 97 patients with full recanalization and in 43% of patients with partial recanalization. Conversely, in patients with a good post-treatment outcome (mRS score 0–2), only 2% of patients had no recanalization whereas 34% had partial and 29% had complete recanalization. These results emphasize the notion that complete or at least partial recanalization of the occlusion is essential for a favorable neurological outcome. Furthermore, even partial recanalization resulted in a relatively satisfactory post-treatment mRS score of 0–4 in 57% of patients compared with 14% in non-recanalized patients. In keeping with this finding, a meta-analysis of all case series involving >10 patients with basilar occlusion who underwent IAT with urokinase and/or recombinant tissue plasminogen activator (rtPA) during the 1987–97 period (n=164 patients) found that failure to recanalize was associated with higher mortality.20

Key points

  • Randomized trial data are lacking.

  • Multiple series have been published.

  • Meta-analysis1 suggests a recanalization rate of 65%, mortality of 56% and symptomatic intracranial hemorrhage rate of 7% with IAT.

  • Outcome is much worse in non-recanalized patients19: recanalization is the key to improved outcome.

  • Partial recanalization also results in improvement in clinical outcome.19 20

Factors impacting on outcome in patients treated with IAT

As well as failure to recanalize, an additional factor associated with higher mortality in Schulte-Altedorneburg's large series was the presence of coma on presentation (RR 1.95, 95% CI 1.26 to 2.99; n=145).19 Intriguingly, however, it has recently been shown that low levels of consciousness assessed using the Glasgow Coma Scale (GCS) did not correlate with poor neurological outcome in 40 patients with acute basilar artery thrombosis managed with intra-arterial therapy. Equal numbers of patients with a good neurological outcome (mRS score 1; n=12) had a GCS of ≤6 or >6. In those with GCS ≤6 for >3 h, 33.3% had a good neurological outcome and this was a similar rate to that of the overall cohort. The authors concluded that it is therefore not appropriate to exclude patients from intra-arterial therapy on the basis of a low presenting GCS.21

The presence of collateral flow via posterior communicating arteries has been shown to correlate with neurological outcome in patients treated with IAT.22 Clot location, mechanism of formation and volume have also been shown to impact on outcome: in patients treated with IAT, thrombus location in the distal one-third of the basilar artery is associated with a lower mortality than thrombus location in the proximal and/or middle portions of the basilar artery (RR 0.52, 95% CI 0.31 to 0.86; n=126).23 Likewise, where studies have classified etiology,23 24 higher recanalization rates are seen in embolic occlusion than in atherothrombotic occlusion (66% and 72% vs 19% and 60%, respectively). Better neurological outcome following IAT for embolic occlusions has also been demonstrated. This should be tempered by the findings of others who have failed to show a significant difference relating to etiology.19 Thrombus volume has also been shown to impact on recanalization following IAT; as thrombus volume exceeds 300 μl, the recanalization rate decreased significantly.19

There is no conclusive evidence regarding timing of therapy. Eckert et al23 found a significantly better clinical outcome in patients with acute basilar occlusion treated within 6 h of symptom onset than in patients treated after >6 h (p=0.005). Sixteen of 33 patients treated within 6 h survived, 12 (36%) with a favorable outcome (Barthel Index >90) and four (12%) with an unfavorable outcome (Barthel Index <90). Mortality was 52% (n=17). In late therapy onset (n=43), only three patients survived with a favorable outcome (7%), 10 survived with an unfavorable outcome (23%) and 30 patients died (mortality 70%). Sixteen patients were only treated after 10 h, 12 of whom died (mortality 75%), and all four survivors had an unfavorable outcome. They concluded that performing IAT is not reasonable if the patient has had a severe neurological deficit for >6 h. However, Arnold et al25 found no significant association between time to treatment and favorable clinical outcome (mRS score 0–2) in 40 patients treated with IAT within 12 h of symptom onset. Likewise, Schulte-Altenbourge et al19 found that neither the duration of brain stem stroke symptoms nor the coma duration was an independent variable for a favorable neurological outcome (mRS score 0–2). Noufal et al26 described several patients presenting >8 h after symptom onset with poor clinical predictors including depressed levels of consciousness who had good outcomes following mechanical thrombectomy.26 Importantly, these patients did not show evidence of extensive brain stem infarction on diffusion-weighted imaging (DWI). The point has been made previously that patients who survive for a longer period following symptom onset may have better collateral pathways.22 The current findings would therefore suggest that strict application of rigid temporal exclusion criteria is not warranted. Indeed, lower rates of complicating post-therapeutic hemorrhage in the posterior circulation and the frequent catastrophic natural history of this condition lead many clinicians to favor a less conservative approach in these patients. Perhaps application of imaging to elucidate the extent of pretreatment core infarct will help select patients for treatment.

Indeed, several studies have shown that radiological predictors, including the brain stem DWI score or the posterior circulation Acute Stroke Prognosis Early CT Score (pc-ASPECTS) based on CT angiography source images (CTASI), are independent predictors of clinical outcome in patients treated with IAT or other means of endovascular therapy.27–29

CTASI can improve the conspicuity and therefore the sensitivity of acutely ischemic areas over that depicted by non-enhanced CT imaging and, effectively, the source images provide information on brain perfusion. Puetz et al27 aimed to quantify early ischemic changes seen on CTASI to predict functional outcome in patients with basilar artery occlusion using the pc-ASPECTS score, which allocates the posterior circulation 10 points. One point each is subtracted for hypoattenuation on CTASI in the left or right thalamus, cerebellum or posterior cerebral artery territory, respectively, and two points each are subtracted for hypoattenuation on CTASI in any part of the midbrain or pons. A pc-ASPECTS score of 10 indicates absence of visible posterior circulation ischemia; a score of 0 indicates hypoattenuation in all pc-ASPECTS territories. Of 46 patients with basilar artery occlusion, 52% (12/23) with a CTASI pc-ASPECTS score ≥8 but only 4% (1/23) with a score ≤8 had a favorable functional outcome (mRS score 0–2). This difference was consistent in 21 patients with angiographic recanalization. In contrast, patients with a CTASI pc-ASPECTS score ≥8 were less likely to die. The authors conclude that the CTASI pc-ASPECTS score can identify patients who will have a poor clinical outcome despite recanalization. In a further study, the same group have shown that the extent of hypoattenuation on the initial CTASI predicts the final infarct extent (based on the pc-ASPECTS score) in patients with basilar artery occlusion.28

Similarly, Cho et al29 evaluated the extent of the brain stem DWI lesion load on pretreatment MRI according to the affected arterial territories of the medulla, pons and midbrain—the ‘brain stem DWI lesion score’. This comprises scores between 0 and 22, with a score of 0 indicating the absence of visible posterior circulation ischemia and a score of 22 indicating DWI restriction in the entire brain stem. The number of arterial territories with abnormal DWI was assessed at each brain stem level: 0–8 in the medulla, 0–6 in the pons and 0–8 in the midbrain. Of the 29 patients included, the authors found that the DWI score was able to predict the clinical outcome (favorable outcome defined as mRS score 0–2) despite recanalization being achieved in 76% of patients. The authors concluded that this scoring system may be able to identify patients most likely to benefit from treatment. Extensive brain stem damage may indicate a poor prognosis regardless of recanalization success or failure. Conversely, they suggest that patients with limited DWI lesions might benefit from late reperfusion therapy.

Several studies have compared recanalization rates using various thrombolytic agents, most commonly urokinase or tPA. Although numbers are small in some studies and statistical significance was not reached, tPA trends toward a more effective recanalization agent.23 30–32 Schulte-Altedorneburg et al32 also found a significantly higher bleeding rate for tPA than for urokinase. In particular, the administration of high-dose tPA (>80 mg) caused a significant increase in intracranial hemorrhage compared with urokinase (46% vs 21%, p<0.01). The neurological outcome was worse in patients with post-interventional bleeding (p<0.001). The authors recommend a cut-off point of 80 mg for the dosage of tPA.

The same group also recorded patterns of infarction following intra-arterial therapy with and without successful recanalization.32 The occipital lobe, thalamus and the superior part of the cerebellum were significantly more often affected by ischemic lesions in the case of successful recanalization. This result corresponds to the angiographic finding that occlusion of the post-communicating segment of one or both posterior cerebral arteries (PCAs) was frequently found with successful recanalization but not with absent recanalization (39% vs 0%). Their findings suggest that the successful recanalization by fibrinolytic agents comes with the price of distal migration of thrombotic material into the PCA, superior cerebellar arteries and perforators branching from the pre-communicating segment of the PCA. The authors suggest that mechanical thrombectomy may achieve more effective recanalization with less distal embolization.

Key points

  • In some studies patients presenting in coma have the worst prognosis overall19 but, presenting low GCS did not correlate with poor outcome in patients treated with intra-arterial therapy in a recent study.21

  • Recanalization rates are better in patients with lower clot volume, embolic etiology and distal basilar occlusion.23 24

  • The neurological outcome tends to be better in patients with collateral flow.22

  • Early treatment has the best outcomes but late presentation does not preclude treatment and does not universally result in a poor outcome.23–26

  • Imaging with DWI or CT angiographic source images may help to predict the outcome independent of recanalization.27–29

  • Treating patients with low DWI scores despite late presentation may result in a successful outcome.26 29

  • Recanalization is better with tPa than with urokinase, although doses of tPA >80 mg are associated with higher rates of symptomatic intracranial hemorrhage.32

  • Distal embolic events may complicate IAT.32

IAT versus IVT: some controversial findings

There is controversy in the stroke literature as to the optimal treatment for patients with basilar artery occlusion. Critics of IAT claim that procedural risks and expenses cannot currently be justified.33 Proponents suggest higher rates of recanalization with IAT and potentially lower rates of symptomatic intracranial hemorrhage.1 4 5 Indeed, transcranial Doppler monitoring suggests recanalization in only 30% of cases of basilar artery occlusion treated with IVT.34 Two principal studies have attempted to compare IVT and intra-arterial therapy—a meta-analysis undertaken by Lindesberg and Mattle in 20068 and a prospective multicenter registry-based trial, the Basilar Artery International Cooperation Study (BASICS), performed in 2009.35

Lindesberg and Mattle8 compared the outcomes of basilar artery occlusion in 76 patients treated with IVT and 344 patients treated with IAT, with the following findings:

  • The outcome for death or dependency was very similar (78% (59/76) and 76% (260/344), respectively).

  • Recanalization was achieved more frequently with IAT than with IVT (225/344 (65%) vs 40/76 (53%), p=0.05).

  • Survival rates after IVT and IAT were similar (38/76 (50%) and 154/344 (45%), respectively, p=0.48).

  • A total of 24% of patients treated with IAT and 22% treated with IVT obtained favorable outcomes (p=0.82). Favorable outcome was defined as mRS score 0–2, Barthel Index 95–100 or GCS 5.

  • The likelihood of a good outcome without recanalization was very low (2%).

The authors concluded that recanalization occurs in more than half of patients with basilar occlusion treated with IAT or IVT and 45–55% of survivors regain functional independence. The effect of IVT did not appear to be much different from that of IAT.

However, this analysis had certain limitations. The IAT and IVT group sizes were remarkably different and 50 of 76 patients treated by IVT were drawn from an earlier study by Lindsberg et al36 in which occlusion was diagnosed by the clinical syndrome combined with time-of-flight MR angiography. MR angiography is insensitive to slow and low residual flow and may show discontinuity (skip sign) attributable to technical reasons. It is therefore plausible that the basilar artery was either completely or only partially occluded, with a potentially life-saving residual flow maintaining the perfusion of the brain stem in the patients in the study by Lindsberg et al. Angiography confirming occlusion was available in only 5 of 50 patients, so this study—which provided most of the patients for the IVT group—could easily have introduced bias into the meta-analysis.

The BASICS trial36 was a prospective multicenter non-randomized observational registry of 592 consecutive patients who presented with an acute symptomatic and radiologically-confirmed basilar artery occlusion during a 5-year period. Patients were divided into three groups according to the treatment they received: (1) antithrombotic treatment only (antiplatelet drugs or systemic anticoagulation, n=183); (2) primary IVT including subsequent IAT (n=121); and (3) intra-arterial therapy (which comprised thrombolysis, mechanical thrombectomy, stenting or a combination of these approaches, n=288).

Stroke severity at the time of treatment was categorized as severe (coma, locked-in state or tetraplegia, n=347) or mild to moderate (any deficit that was less than severe, n=245). The outcome was assessed at 1 month. A poor outcome was defined as mRS score of 4 or 5 or death. For mild to moderate deficits, comparison with anitithrombotic treatment showed that the risk of a poor outcome after IVT (adjusted RR 0.94, 95% CI 0.60 to 1.45) was about the same as that after IAT (adjusted RR 1.29, 95% CI 0.97 to 1.72), but the outcome was worse after IAT than after IVT (adjusted RR 1.49, 95% CI 1.00 to 2.23). For severe deficits, comparison with anitithrombotic treatment showed that there was a lower risk of a poor outcome after IVT (adjusted RR 0.88, 95% CI 0.76 to 1.01) or IAT (adjusted RR 0.94, 95% CI 0.86 to 1.02). Outcomes regarding rates of recanalization and outcome of patients were similar after treatment with IAT and IVT (adjusted RR 1.06, 95% CI 0.91 to 1.22). The authors concluded that their results did not support unequivocal superiority of IAT over IVT and encouraged clinicians to treat patients who have acute symptomatic basilar occlusion and a mild to moderate deficit with IVT.

However, several criticisms can be made of this study. Importantly, patients in the IVT group had less severe strokes and were treated earlier, indicating a bias toward the preferred use of IVT in patients with shorter and less severe symptoms and a bias toward a better outcome in the IVT group. The mean initial National Institute of Health stroke scale score was 21 in the IVT group and 25 in the IAT group, and 55% of patients in the IVT group were treated within the first 3 h after symptom onset compared with only 23% of the IAT group.

Furthermore, confounding factors were not measured or controlled for. Thrombus volume, length of the occluded vessel and the existence and degree of residual or collateral flow are all known to have an impact on the ability to recanalize and on the clinical outcome.22–24 However, in the BASICS registry the IVT group did not include sufficient vascular imaging to control for this.

It is also uncertain as to whether true basilar artery occlusion or partial but collateralised stenosis or emboli causing only a temporary occlusion were treated in the IVT group. This factor may account for the relatively high recanalization rate of 67% in the IVT group in the BASICS registry relative to published rates of recanalization.34

Another important point is that 41 patients in the IVT group were treated with a bridging approach (IVT followed by IAT), but these patients were included in the IVT group alone with regard to assessment of recanalization and outcome. This could have had a significant impact on the results and could potentially have altered the statistical significance if these patients were included in an IAT/endovascular group.

If the potential biases described are borne in mind, it is difficult to draw firm conclusions from the available data comparing the delivery of thrombolytic agents. What is required is a trial comparing IVT, IVT plus IAT and IAT (including clot extraction devices).

Key points

  • A meta-analysis8 and a large prospective registry-based study (BASICS)36 have compared IAT and IVT.

  • While conclusions drawn in these studies suggest little difference between the two strategies, flaws in study design preclude making these conclusions.

  • The meta-analysis8 group sizes were very different and MR angiography (insensitive to slow flow) was used to assess the IVT group.

  • The BASICS study36 included patients with less severe strokes in the IVT arm and these were treated earlier. In addition, patients treated with a bridging approach were included in the IVT group.

  • A randomized trial would be useful to determine whether IAT adds to improved clinical outcomes over IVT alone or in combination with IVT.

Bridging therapy

One of the major disadvantages of the intra-arterial approach is the time loss to initiate the treatment. An attempt to overcome this disadvantage has been to combine the intravenous route with the intra-arterial access using the so-called ‘bridging’ concept. Furthermore, endovascular therapy is only offered in a few institutions—often tertiary referral centers—so, if treatment can be started early before transfer to these centers, the outcome may be improved.

Eckert et al37 and Nagel et al38 have both performed studies assessing outcomes in patients treated initially with an intravenous bolus, then infusion of the glycoprotein IIb/IIIa receptor antagonist abciximab (an initial bolus of 0.25 mg/kg followed by 0.125 μg/kg/min for 12 h) followed by intra-arterial therapy including tPA and, where appropriate, angioplasty with or without stenting for underlying atherothrombotic lesions. In the study by Eckert et al,37 half-dose intra-arterial tPA (median 20 mg) was used in the bridging group and full-dose intra-arterial tPA (median 40 mg) was used in the control group. In the study by Nagel et al,38 mechanical disruption of the thrombus was also performed in both groups. Both studies demonstrated improved rates of recanalization in the bridging groups as well as improved survival compared with IAT alone (58–62% vs 25–32%), and higher rates of favorable outcome (mRS score 0–3) were seen in the bridging groups (35% vs 12.5–17%). Although overall bleeding complications were greater in the study by Eckert et al, symptomatic intracranial hemorrhage did not differ significantly with a rate of 3–14%. These results are intriguing and introduce a potentially useful protocol for care; however, in both studies, patients in the bridging group were more frequently treated with percutaneous transluminal angioplasty/stent placement. This could have biased the results and prevents firm conclusions on the bridging concept including pre-intervention abciximab. Indeed, the results may reflect the fact that angioplasty/stenting is potentially a very useful technique in the treatment of atherthrombotic occlusions.

Pfefferkorn et al39 treated a small cohort of 16 patients presenting with basilar occlusion with a standardized protocol combining IVT with consecutive on-demand endovascular mechanical thrombectomy. Seven of the 16 patients required endovascular treatment because of persistent basilar artery occlusion following systemic thrombolysis. Approximately half of these patients had a good outcome using this regime: 44% with no or mild deficit (mRS score 0–2) and 13% with a moderate deficit (mRS score 3). Only one intracranial hemorrhage occurred in a patient who had received IVT alone.

The same group have instituted a ‘drip, ship and retrieve’ cooperative treatment protocol in which thrombolysis was initiated in a community hospital with simultaneous referral to a central stroke center for endovascular mechanical recanalization as required (n=26), and compared the outcomes with a group given IAT with or without tirofiban bridging between 2003 and 2005 (n=26).7 The rate of symptomatic intracranial hemorrhage was 12% in the historical cohort and 8% in those treated under the new protocol. Recanalization rates were similar (92% in the historical cohort and 85% in those treated with the new protocol); 38% of these had recanalization after IVT alone. Functional outcome was better among those treated with the new protocol, with more patients achieving a mRS score of 2 (38% vs 12%, p=0.03) and a mRS score of 3 (50% vs 23%, p=0.04). ‘Drip, ship and retrieve’ appeared to be feasible and safe in acute basilar artery occlusion. Patients appear to benefit from initiation of IVT before transfer for endovascular care. The speed of delivery, coupled with a ‘softening’ of the thrombus prior to intervention, may prove effective. The risk of intracerebral hemorrhage remains the largest potential drawback, but results from this initial work suggest that this risk may not be as significant as previously thought. Further studies are needed to confirm the effectiveness of the bridging approach.

Key points

  • A bridging approach of intravenous abciximab (glycoprotein IIb/IIIa inhibitor) prior to intra-arterial treatment of tPA has shown some clinical outcome benefit over IAT alone in non-randomized trials.37 38

  • Confounding factors limit firm conclusions as it is difficult to quantify the potential effect of angioplasty in the bridging arms.

  • A ‘drip, ship and retrieve’ regimen of initial IVT in a community hospital followed by IAT in an interventional neuroradiology center has also demonstrated additional benefit over IAT alone.39

Mechanical clot extraction

As the success of local thrombolysis in basilar thromboembolism depends on the volume and composition of the clot, and since thrombolysis can also be time-consuming and cause bleeding, various investigators have assessed the use of mechanical thrombectomy devices in the basilar artery. Most of the literature pertaining to mechanical thrombectomy/thromboaspiration features a small number of devices including MERCI40 and the Penumbra aspiration catheter.41 Simple mechanical clot disruption42 and several other devices including the Possis Angiojet catheter43 and the Solitaire stent retriever44 have also been used in the posterior circulation.

Lutsep et al40 extracted data on basilar artery occlusion from both MERCI and Multi MERCI trials. Patients received treatment up to 8 h after symptom onset. Recanalization was determined after retriever use and adjunctive therapy. Mortality and good outcomes (mRS score 0–3) were determined at 90 days in patients who were recanalized and not recanalized. Recanalization occurred in 21 of 27 (78%) patients. Mortality was 44% and good outcomes were seen in 41%. Eighteen patients with basilar occlusions were included in the largest series using the Penumbra aspiration catheter.41 44 Although these series suggest recanalization rates of approximately 85%, data specific to the vertebrobasilar system were not separately published.

Mayer et al43 assessed the use of the Possis Angiojet catheter in 12 patients presenting with vertebrobasilar occlusion. Combined treatment with the Angiojet and additional fibrinolysis or angioplasty resulted in a recanalization rate of 100%. Three symptomatic and two asymptomatic hemorrhages were detected by CT. Five of the 12 patients died; 50% of patients obtained a moderate to excellent outcome (mRS score 0–3). The authors concluded that this device has the potential to increase the recanalization rate, especially in cases of extensive thrombosis, and thus to improve patient outcomes. Encouraging results have also been obtained using the Solitaire stent retriever (figure 1).45 46 A 90% recanalization rate was achieved in one series that included nine basilar occlusions among 40 patients45 and, in another series,46 15 of 56 patients. In the latter series a favorable outcome rate (mRS score 0–2) of 33% was achieved, lower than that seen in anterior circulatory strokes in the same series. Proponents of the Solitaire stent retriever suggest that it has several advantages over other interventional techniques—not only is there a very high rate of successful recanalization, but implantation of the stent retriever typically produces immediate recanalization.

Figure 1

An example of the use of the Solitaire stent retriever. Left: Digital subtraction angiogram showing acutely occluded distal basilar artery. Center, left: Solitaire deployment in left posterior cerebral artery with partially recanalized right posterior cerebral artery. Center, right: Successful recanalization of basilar and posterior cerebral arteries. Right: Solitaire with extracted thrombus following this procedure.

Key points

  • High recanalization rates of 75–100% can be achieved with various thrombectomy/thromboaspiration devices.43–46

  • Further trials are necessary to elucidate whether this translates into improved clinical outcomes in appropriately selected patients.


Although there is a lack of randomized data, what is clear is that, although variable, basilar occlusion can be catastrophic and the outcome is better in patients who are treated with recanalization therapy. Patients with more distal occlusions of embolic etiology and with poor collateral flow do badly without treatment. Large clot loads are difficult to treat, even with IAT, and mechanical methods of extraction/aspiration may be needed to facilitate recanalization. Patients with a depressed level of consciousness (although a marker of poor prognosis) may still do well with treatment despite the presentation. Treating patients early is optimal, although delayed treatment may still be successful and exclusion from treatment based strictly on timing is not warranted. Perhaps assessment of pretreatment infarct core with diffusion-weighted MR imaging or CTASI will be a useful tool in selecting for aggressive therapy those patients who present late or who have clinically severe strokes. We suggest that the recanalization rates of IAT and mechanical devices are generally higher relative to IVT, and it is not unreasonable to assert that, for acutely unwell patients with poor prognostic indicators and heavy clot loads, endovascular therapy is currently the only way to achieve effective recanalization satisfactorily in the majority of patients.


Basilar artery occlusion is a heterogeneous condition with variable clinical severity; however, it can be fatal. Outcomes are improved in those who undergo recanalization. Many other factors influence the outcome including severity of presentation, site of occlusion, clot load, degree of collateral flow, core infarct distribution, timing of therapy, agent used for recanalization and dose of thrombolytic agent. This makes comparison of different therapeutic modalities difficult and probably renders firm conclusions as to the best mode of therapy (IAT vs IVT, IAT plus IVT, mechanical extraction vs thrombolytic, or mechanical extraction plus thrombolytic) impossible based on current data. It is probably true that all methods should be available to tailor treatment for individual patients as part of multidisciplinary care. With increasing use of mechanical clot extraction devices, much of the data we currently have pertaining to treatment of this condition no longer reflects current practice in many institutions. As the use of mechanical clot extraction devices increases, outcomes following the use of this technology should be assessed on a large scale. Initial results with these devices are encouraging. A randomized trial could help to clarify some points and, if therapies are compared, we could get closer to the claim that one treatment is superior to another. However, the relatively small numbers of patients, multiple confounding factors and lack of clinical equipoise are barriers to the development of such a trial and it is unlikely that a study of this quality will be available in the near future.



  • Competing interests None.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.