Recanalization of Vertebral Artery Dissection
Background and Purpose— We investigated the predictors and time course for recanalization after vertebral artery dissection.
Methods— We prospectively studied 61 consecutive patients with confirmed diagnoses of vertebral artery dissection without intracerebral hemorrhage. Neuroimaging and clinical follow-up were performed at presentation and at 3, 6, and 12 months.
Results— We included 61 patients with confirmed vertebral artery dissection; 19 were evaluated and followed up with conventional angiography, 24 with MR angiography, and 18 with CT angiography. Fifty-one patients had a stenotic dissection, 7 had an occlusive dissection, one had a double-lumen image, and 2 had a pseudoaneurysm. The estimated rate of complete recanalization after vertebral artery dissection was 45.9% at 3 months, 62.3% at 6 months, and 63.9% at 12 months. We found no association between outcome and complete or partial recanalization nor did we find any factors associated with recanalization.
Conclusions— These results suggest that recanalization of vertebral artery dissection occurs mainly within the first 6 months after the onset of symptoms regardless of the location or pattern of the dissection.
Vertebral artery dissection (VAD) is a well-recognized cause of ischemic stroke in the vertebrobasilar circulation territory of young and middle-aged adults.1,2 Little is known about the natural history because most of the studies are based on spontaneous carotid artery dissection (SCAD). Despite the dearth of information, recurrent dissection has only rarely been reported, and prognosis tends to be good.3,4
Selective digital subtraction angiography has long been the gold standard for diagnosis and follow-up. However, this technique is not without risk.5 A noninvasive method is therefore needed to detect, monitor, and follow-up this condition. In a recent ultrasound follow-up study in patients with SCAD, Nedeltchev et al6 found that recanalization occurs mainly within the first 6 months after the onset of symptoms. However, the evaluation of VAD by using duplex ultrasonography remains difficult because of the small size of the vertebral arteries, their deep location, frequent anatomic variations, and surrounding structures.7,8 MR angiography (MRA) and CT angiography (CTA) offer potential advantages for the noninvasive assessment of extracranial vascular disease and have previously been proven useful for the detection of VAD.9–11 Sequential neuroimaging examinations can indicate recanalization or normalization of the blood flow and may be helpful for the decision to discontinue antithrombotic therapy. However, the appropriate timing for follow-up examinations has not yet been defined. We undertook this study to investigate the predictors and the time course of recanalization of VADs using digital subtraction angiography, MRA, and CTA.
Subjects and Methods
Prospectively collected data for all patients presenting with first-time VAD at the National Institute of Neurology (Mexico City) from January 2003 through December 2008 were reviewed. All patients underwent an etiologic workup during the acute phase, which included physical and neurological examination, routine blood examinations, an electrocardiogram, and an assessment of the cerebral arteries of the neck. Inclusion criteria consisted of (1) compatible clinical signs and symptoms of VAD with definite angiographic, CTA, or MRA findings of dissection in the vertebral artery; (2) acute infarcts within the vertebrobasilar circulation territory on diffusion-weighted imaging; and (3) no evidence of luminal irregularities, stenosis, or occlusions in vertebral artery or the other cervical and intracranial arteries that would be suggestive of atherosclerosis. Patients presenting with subarachnoid hemorrhage were treated in a neurosurgical unit and were not included in the present study.
Risk factors for VAD and ischemic stroke were defined as follows: a history of antihypertensive, antidyslipidemic, or antidiabetic treatment or blood pressure >140/90 mm Hg; total cholesterol value >200 mg/dL; low-density lipoprotein >100 mg/dL; or triglycerides >150 mg/dL without treatment or fasting blood glucose >100 mg during the hospital stay. The other evaluated vascular risk factors evaluated were current smoking status, which was derived from patient interviews or medical records, and history of migraine, which was determined using the standard diagnostic criteria.
The subtypes of cerebral infarcts were classified as follows: (1) vertebral perforator infarcts: a small perforating arterial infarct in the medulla, including lateral medullary infarct and medial medullary infarct; (2) basilar perforator infarcts: a small perforating arterial infarct in the pons, including paramedian pontine infarcts; and (3) and (4) small scattered infarct and small infarct: multiple or unique small (largest diameter <1 cm), scattered dot-shaped infarcts involving the territory of the vertebral artery, basilar artery, posterior inferior cerebellar artery, anterior inferior cerebellar artery, superior cerebellar artery, and/or posterior cerebral artery.
The location of the dissection was defined by standard anatomic categories. The V1, V2, and V3 segments (V1 from the origin to C6, V2 from C6 to C2, and V3 from C2 to the dura) and the V4 segment from the dura to the basilar artery. Three groups were defined according to the proximal site of the dissection: group: (1) extracranial vertebral artery dissections originating from the extradural segment and group; (2) intracranial vertebral artery dissections originating from the intradural segment; and (3) simultaneous involvement of both the extracranial and intracranial segments. VADs were also categorized according to the history of trauma.
The initial neuroradiological studies were performed within the first week of diagnosis (range, 1 to 12 days after the onset of VAD symptoms). The choice of the initial and subsequent imaging procedures was at the discretion of the treating neurologist. Follow-up examinations to assess recanalization were performed at 3, 6 and 12 months. Complete recanalization was diagnosed when imaging showed normal findings, and partial recanalization was diagnosed when imaging revealed a reduction in the degree of stenosis or a change from occlusion to a stenosis. Once the follow-up imaging showed complete recanalization, no further examinations were performed.
Clinical follow-ups were also performed at 3, 6, and 12 months. No selection criteria with respect to aspirin or anticoagulants were applied, and all treatments were initiated once the diagnosis was confirmed. Stroke severity on admission was assessed with the National Institutes of Health Stroke Scale12 and functional outcomes were assessed using the modified Rankin Scale. Modified Rankin Scale scores of 0 to 2 were defined as “favorable” outcomes and scores of 3 to 6 were defined as “unfavorable” outcomes. The local ethics committee approved this study.
Digital subtraction angiography with selective catheterization of the vertebral arteries was performed in 19 patients (19 VADs) within 1 week of the onset of the neurological symptoms. Angiographic findings of a double lumen (the presence of a false lumen or an intimal flap), stenosis involving an irregular long or short segment (the pearl and string sign), occlusion involving either the entire vertebral artery or only one segment of the artery, or a pseudoaneurysm associated with a narrowed arterial lumen were used as reliable angiographic findings indicative of VAD.13 Of the other reported abnormal findings suggesting vertebrobasilar artery dissection, a string sign, tapered narrowing, total occlusion with proximal distension, and retention of contrast medium, were not considered pathognomonic signs. Although a finding of resolution of stenosis on follow-up angiography was considered a reliable angiographic sign of vertebrobasilar artery dissection, patients with initial arteriograms that did not reveal the double-lumen sign or the pearl and string sign were not included in this study.
CTA was performed in 18 patients (18 VADs). Axial source images were evaluated to confirm the findings of MIP images or to explore new findings missed on the MIP images. The window widths and levels were adjusted to allow a clear delineation of the enhanced lumen, intramural hematoma, and calcification. The criteria for diagnosing a VAD using CTA were as follows: (1) a narrowed centric or eccentric lumen surrounded by crescent-shaped, mural thickening and an associated increase in the external diameter; (2) an abrupt or tapered occlusive lumen and an associated increase in the external diameter; or (3) an aneurysmally dilated lumen or a dilated and narrowed lumen with or without crescent-shaped mural thickening or an intimal flap. Increased external diameter was determined by comparing it with the segment proximal to the dissection.10 Follow-up CTAs were available for 18 patients.
Magnetic Resonance Angiography
MRA examinations (24 patients; 24 VADs) were performed on a 1.5-T scanner and included both carotid arteries and the V1 to V4 segments of the vertebral arteries; the aortic arch was not within the excitation volume. The images were reconstructed using an MIP algorithm to render the projection MR angiograms from different viewing angles. If the vertebral artery was not adequately delineated on the projections or the source images, additional 2-dimensional time-of-flight-sensitive fast low-angle shot images were obtained. Occlusion referred to nonvisualization of flow in either sequence. The criterion for a mural hematoma was an eccentric residual lumen surrounded by a semilunar signal alteration. A dissecting aneurysm was diagnosed when an aneurysmal dilatation was accompanied by a mural hematoma. The location of the arterial changes was defined according to the criteria used for CTA that is described previously.9,11
Data analysis was performed with the statistical package SPSS 13. The χ2 test was performed for crosstabulation analysis. The following variables were analyzed: age, sex, history of trauma, baseline National Institutes of Health Stroke Scale score, smoking status, hypertension, diabetes mellitus, hypercholesterolemia, history of migraine, and location of VAD. Next, forward stepwise analysis was performed to determine the independent associations between recanalization and other clinical and radiological factors. All variables that had a probability value of <0.25 on the univariate analysis were included in the multivariate analyses. The Kaplan-Meier method was used to determine the likelihood of recanalization at 3, 6, and 12 months.
We evaluated 74 consecutive patients with VAD. All patients presented with symptoms of ischemia in the vertebrobasilar territory. Of the 74 patients, 13 patients were not included because the baseline and follow-up neuroimaging techniques used were different and 2 patients had vertebral artery occlusion of unknown cause. Therefore, a total of 61 consecutive patients (64% men) aged 20 to 48 years (mean 34.6±7 years) with confirmed VAD were included in the study. The most common vascular risk factors were smoking (28%), alcohol use (21%), and hypertension (16%). Nontraumatic dissections were more frequent than traumatic ones (64% versus 36%); 19 of the trauma cases (29%) were related to minor trauma and 4 (6%) to major trauma. A total of 8 patients were treated with aspirin and 53 were anticoagulated.
Of the 61 VADs, 19 were evaluated and followed up with digital subtraction angiography, 24 with MRA, and 18 with CTA. With respect to neuroimaging features, 51 (83%) were classified as a stenotic dissection (including 27 tapered) and 7 (11.5%) were classified as an abrupt occlusive dissection (4 in V1, one in V2, and 2 in the V4 segment). One case (1.6%) showed a double-lumen image and 2 (3.3%) had pseudoaneurysms. Twenty-one (34%) were in the V1 to V4 segments, 6 (9.8%) in the V2 to V4 segments, 19 (31%) in the V3 to V4 segments, 11 (18%) in the V1 to V2 segments, and 4 (6.6%) in the V4 segment.
Recanalization of VAD
All patients underwent follow-up imaging studies equivalent to the ones used for diagnosis during hospitalization. Imaging showed complete recanalization in 28 (46%) cases at the 3-month follow-up, in 38 (62%) at the 6-month follow-up, and in 39 (64%) at the 12-month follow-up. The presenting clinical characteristics of cases with and without recanalization during follow-up are shown in Table 1. Cardiovascular risk factors, a history of migraine, and minor or major trauma did not differ between the 2 groups. None of the initial neuroimaging findings were more frequent in the group of VADs with complete recanalization. Of the extracranial VADs, 7 (63.2%) had complete recanalization, whereas of the 4 intracranial VADs, 3 (75%) and 29 (63%) of the 46 extra-/intracranial VADs, respectively, showed complete recanalization (P>0.5) and at follow-up.
The estimated rate of complete recanalization according to the Kaplan-Meier method was 45.9% (95% CI, 39.5% to 52.3%) at 3 months, 62.3% (95% CI, 56.1% to 68.9%) at 6 months, and 63.9% (95% CI, 57.8% to 70%) at 12 months. The Figure shows in a histogram the time when recanalization was observed at Month 3, Month 6, and Month 12.
None of the patients had further ischemic neurological symptoms during follow-up. Favorable outcomes (modified Rankin Scale ≤2) were observed in 55 (90.2%) of the 61 patients and were as frequent for VADs with complete recanalization (35 of 55 patients) as for those without complete recanalization. Also, good outcomes were equally frequent for extracranial and intracranial VADs. After multiple regression analysis, sex, history of minor trauma, vascular risk factors, a history of migraine, location of VAD, vertebral artery occlusion, and the type of antithrombotic treatment were not independently associated with clinical outcome (Tables 1 and 2⇓).
Angiography is the current imaging method of choice for the diagnosis of VAD. However, findings with conventional angiography are helpful but not always diagnostic of VAD.5,13 Therefore, the diagnosis may be suspect unless the angiographic findings are interpreted in conjunction with clinical presentation or follow-up studies.12,13 With advances in CT and MRI techniques, CTA and MRA are gradually replacing angiography for diagnosis. In this prospective observational study, CTA and MRA seem to be reliable tools for diagnosis and follow-up of VAD, especially when combined with the patient’s history and clinical examination. CTA and MRA have the advantage of being less invasive and may facilitate decisions on the duration of therapy with anticoagulants or antiplatelet agents.
Similar to previous reports,1,14 the present study confirms that VADs occur predominantly in middle-aged adults. The mean age of our patients was 34.6 years. However, we emphasize that VADs, according to other studies,13 occurs not only in young adults, but also in older adults. In the present series, 64.5% of the patients were men, which is in accordance with previous reports.15,16 Although the role of sex in VAD is controversial, some authors report evidence of a male preponderance as high as 83%.14 There are several genetic and environmental factors that could potentially contribute to these sex differences that need to be further studied.
Although extracranial VADs have been reported to be more common than intracranial VADs,16 in the present study, we found a surprisingly high simultaneous involvement of both the extracranial and intracranial segments. In our series, the isolated involvement of the intracranial segment was only 6.6%. Intracranial extension probably occurs more frequently than previously thought. The location of the vascular abnormality (intracranial versus extracranial) does not seem to play a role in the likelihood of recanalization (P=0.27) or even in the functional prognosis of intracranial dissections presenting with ischemic symptoms. Previously, De Bray et al17 described a poorer prognosis for intracranial VADs but we found no difference.
With regard to the pattern of dissection, a double-lumen image and pseudoaneurysms were found in 4.9% of the patients, and stenotic dissection was observed in 84% of patients with 44% of stenotic dissection images being tapered, results that are very similar to the respective values of 83% and 35% that were reported in a previous population-based study.18 None of these patterns had an impact on the probability of recanalization (P>0.05 for string sign ending in total occlusion, pseudoaneurysm, and double lumen; P=0.28 for string sign).
In an earlier series of 130 consecutive cases,16 we found that recanalization occurs over a period of 6 months in only 36% of the patients, which stands in contrast to a rate of up to 88% in 3 to 6 months reported by Mas et al.5 However, in the present study, complete recanalization occurred in 64% of patients within the first 6 months. The reason for this discrepancy in 2 series from the same institution is not clear. The discrepancy could be based on the fact that our previous report included patients seen as early as the late 1970s using different and older imaging techniques.
The complete recanalization rate that we observed is similar to the results of previous studies.7,19 In the Wessels et al series,7 complete recanalization was found in 64% of patients, whereas De Bray et al17 reported complete recanalization in 9 (50%) of 18 patients with intracranial VAD and in 4 (36%) of 11 with extracranial VAD. Using digital subtraction angiography as the follow-up technique, Lee et al18 found complete recanalization in 63% of the 19 patients with VAD with a mean time to achieve complete recanalization of 8.6 months (range, 3 weeks to 56 months) by combining data from 37 VAD cases.
Complete recanalization of VADs was observed mainly within the first 6 months after the onset of symptoms. These findings are in accordance with the results of SCAD studies6 reporting that 60% of vessel recanalization occurs within 6 months. The ideal timing of initial and follow-up investigations for evaluation of VAD and vessel patency remains unknown. It is know that recanalization is a dynamic process that may begin days or weeks after the initial dissection.13 Due to the time periods at which the investigations were conducted, our study was unable to identify the exact timing of recanalization. However, the time course of VAD recanalization observed in this series supports the current clinical practice of maintaining antithrombotic therapy only for the first 6 months after the onset of symptoms or until complete recanalization is observed.
In the study by Nedeltchev et al,6 an occlusive SCAD was an independent negative predictor of complete recanalization, which occurred in only 45% of the occlusive dissections studied, but we observed no difference in the rate of complete recanalization when complete occlusion was present. Also, we did not observe any association between hypertension and other cardiovascular risk factors and the recanalization rates. However, our patients were younger than the patients included in the studies that reported such an association.19
The present study is not capable of addressing the role of oral anticoagulation or antiplatelet therapy in recanalization rates. In accordance with previous results in VAD and SCAD patients, our results show that complete recanalization did not influence functional outcomes in patients with stroke.
The present study has several limitations. The choice and duration of antithrombotic therapy were at the discretion of the treating physician, but it can be argued that current evidence has not compared different antithrombotic treatments. Another limitation is the use of different imaging techniques for follow-up, but again, the decision about the choice of imaging method was at the discretion of the attending doctors at the time of arrival and depended on the availability of equipment. To address this, we used the same imaging method at diagnosis and follow-up, and we excluded those patients who did not meet this requirement. Finally, and more importantly, the lack of asymptomatic patients and patients with transient ischemic attack is something to be taken into account. This limitation, however, was unavoidable due to the fact that our center is a tertiary referral hospital. It is very likely that most asymptomatic patients or patients with transient ischemic attack were evaluated and discharged at primary or secondary centers for symptoms/signs that did not raise suspicion of VAD and thus were never referred to our center.
In conclusion, similar to SCAD, complete recanalization of VADs usually occurs within 6 months after the onset of symptoms independent of the location or the pattern of the vascular defect. MRA and CTA appear to be safe and accurate methods for following up of patients with VAD. However, a patient series with standardized follow-up imaging methods is needed to determine the use of each technique and to detect possible variations between the different imaging methodologies.
- Received September 23, 2009.
- Revision received November 23, 2009.
- Accepted November 24, 2009.
Nedeltchev K, Der Maur TA, Georgiadis D, Arnold M, Caso V, Mattle HP, Schroth G, Remonda L, Sturzenegger M, Fischer U, Baumgartner RW. Ischaemic stroke in young adults: predictors of outcome and recurrence. J Neurol Neurosurg Psychiatry. 2005; 76: 191–195.
Touzé E, Gauvrit JY, Moulin T, Meder JF, Bracard S, Mas JL. For the multicenter survey on natural history of cervical artery dissection. Risk of stroke and recurrent dissection after a cervical artery dissection: a multicenter study. Neurology. 2003; 61: 1347–1351.
Mas JL, Bousser MG, Hasboun D, Laplane D. Extracranial vertebral artery dissections: a review of 13 cases. Stroke. 1987; 18: 1037.
Nedeltchev K, Bickel S, Arnold M, Sarikaya H, Georgiais D, Sturzenegger M, Mattle HP, Baumgartner RW. Recanalization of spontaneous carotid artery dissection. Stroke. 2009; 40: 449–504.
Auer A, Felber S, Schmidauer C, Waldenberger P, Aichner F. Magnetic resonance angiographic and clinical features of extracranial vertebral artery dissection. J Neurol Neurosurg Psychiatry. 1998; 64: 474–481.
Chen CJ, Tseng YC, Lee T-H, Hsu HL, See LC. Multisection CT. Angiography compared with catheter angiography in diagnosing vertebral artery dissection. AJNR Am J Neuroradiol. 2004; 25: 769–774.
Leclerc X, Lucas C, Godefroy O, Nicol L, Moretti A, Leys D, Pruvo JP. Preliminary experience using contrast-enhanced MR angiography to assess vertebral artery structure for follow-up of suspected dissection. AJNR Am J Neuroradiol. 1999; 20: 1482–1490.
Brott T, Adams HPJ, Olinger CP, Marler JR, Barsan WG, Biller J, Spilker J, Holleran R, Eberle R, Hertzberg V. Measurements of acute cerebral infarction: a clinical examination scale. Stroke. 1989; 20: 864–832.
Hosoya T, Adachi M, Yamaguchi K, Haku T, Kayama T, Kato T. Dissection clinical and neuroradiological features of intracranial vertebrobasilar artery. Stroke. 1999; 30: 1083–1090.
De Bray JM, Penisson-Besnier I, Dubas F, Emile J. Extracranial end intracranial vertebrobasilar dissections: diagnosis and prognosis. J Neurol Neurosurg Psychiatry. 1997; 63: 46–51.
Lee VH, Brown RD, Mandrekar JN, Mokri BL. Incidence and outcome of cervical artery dissection: a population-based study. Neurology. 2006; 67: 1809–1812.