Treatment outcomes in cerebral artery dissection and literature review

Abstract

Methods

Patients with cerebral artery dissections were reviewed in a hospital setting from 2008 to 2015. Clinical presentations, lesion locations, treatment modalities, functional outcomes, and mortality were reviewed. Parent artery occlusion was the first choice for surgery or endovascular treatment of a hemorrhagic dissecting cerebral artery. Endovascular or surgical reconstructive treatment was indicated in patients whose parent artery could not be occluded. Favorable functional outcomes were determined using modified Rankin Scale (mRS) scores of 0–2.

Results

In total, 61 patients with cerebral artery dissections were admitted to the hospital. Seven (11.5%) had traumatic dissections. All traumatic dissections were located in the internal carotid arteries. Overall favorable outcome rate was about 57% (4/7). Spontaneous cerebral artery dissections were found in 54 patients. No difference in favorable outcomes was observed between parent vessel occlusion and selective occlusion with parent vessel preservation (or vessel reconstruction) (70% and 63%, respectively, p = 1.000). Patients who presented with spontaneous dissection without intracranial hemorrhage had more favorable outcomes than those with intracranial hemorrhage (79% and 52%, respectively, p = 0.045). The mortality rate of patients with spontaneous dissection was 7.4%.

Conclusions

Most of the traumatic dissections were located on the internal carotid arteries and spontaneous dissections were commonly located on vertebral arteries. Nonhemorrhagic spontaneous cerebral dissections had better functional outcomes after treatment. Endovascular and surgical management were effective treatments by parent vessel occlusion or reconstructions.

Background

Cerebral artery dissections can be categorized into two groups: traumatic dissection and spontaneous dissection. Both groups present with intracranial hemorrhagic or thromboembolic events. Locations and architecture of traumatic dissections vary depending on stress forces and traumatic mechanisms.^1–5 Patients presenting with hemorrhagic spontaneous dissecting arteries carry a risk of recurrent hemorrhage of about 70% while the mortality rate is about 8.3%.^6–9 The most commonly involved arteries are the vertebral arteries, and patients without intracranial hemorrhage reportedly have better functional outcomes. No standard recommendations have been established to treat cerebral artery dissection. Traditional therapies largely have consisted of parent vessel occlusion using surgical or endovascular procedures. Recently, endovascular reconstructive treatments such as stent-assisted coiling embolization or use of flow diverters have been reported to be effective.^5,10–15

Objective

This study aimed to identify lesion locations and evaluate clinical presentations, management, and functional outcomes in patients with cerebral artery dissection.

Materials and methods

Patients with cerebral artery dissection were reviewed in a hospital setting from 2008 to 2015. Clinical presentations, lesion locations, treatment modalities, functional outcomes, and mortality were reviewed. Parent artery occlusion was the first choice for surgery or endovascular treatment of a hemorrhagic dissecting cerebral artery. Endovascular or surgical reconstructive treatment was indicated in patients whose parent artery could not be occluded. In cerebral artery dissection without intracranial hemorrhage, treatment was based on individual clinical presentation, degree of stenosis, recurrent symptoms, or image findings. Favorable functional outcomes were determined using modified Rankin Scale (mRS) scores of 0–2.

Statistical analysis was carried out using IBM SPSS, version 22.0. Fisher’s exact test and the chi-square test were performed as appropriate. Statistical significance was set as p < 0.05.

Results

In total, 61 patients with cerebral artery dissections were admitted to the hospital. Seven (11.5%) had traumatic dissections. All traumatic dissections were located in the internal carotid arteries and the most common cause was a motor vehicle accident. Two patients with hemorrhagic presentations had low initial Glasgow Coma Scale (GCS) scores due to primary head injuries. In traumatic dissection without intracranial hemorrhage, all patients presented with proptosis and bruits from traumatic carotid-cavernous fistulas (CCF), which were incidental in findings on cerebral angiography. Four patients were treated with parent vessel occlusion and one patient was treated with coiling embolization. Overall favorable outcome rate was about 57% (4/7) (Table 1).

Table 1.

Summary of patients with traumatic cerebral artery dissection.

No.	Age	Sex	Location	Presentation	Associated injury	GCS	Management	Outcome (mRS)
1	15	M	Intradural and cervical ICA	Intracranial hemorrhage	Severe DAI, SAH, IVH, fracture at base of skull	4	Coiling occlusion, decompressive craniectomy	5
2	66	M	Cavernous-supraclinoid ICA	Intracranial hemorrhage	SAH, IVH	5	Surgical clipping	5
3	25	M	Cavernous ICA	Incidental finding	Traumatic CCF	15	Conservative	0
4	15	F	Petrous ICA	Incidental finding	Traumatic CCF	15	ICA coiling occlusion	0
5	23	M	Cervical ICA	Incidental finding	Traumatic CCF	15	ICA coiling occlusion	3
6	32	F	Cavernous-supraclinoid ICA	Incidental finding	Traumatic CCF	15	Aneurysm coiling embolization	0
7	12	F	Cervical ICA	Incidental finding	Traumatic CCF	15	Conservative	0

M: male; F: female; GCS: Glasgow Coma Scale; mRS: modified Rankin Scale; ICA: internal carotid artery; VA: vertebral artery; DAI: diffuse axonal injury; SAH: subarachnoid hemorrhage; IVH: intraventricular hemorrhage; CCF: carotid-cavernous fistula.

Spontaneous cerebral artery dissections were found in 54 patients. Twenty-five patients presented with intracranial hemorrhage (46.3%) and 29 exhibited spontaneous dissection without intracranial hemorrhage (53.7%). Locations of spontaneous dissections included vertebral arteries, 61.1% (33 patients), internal carotid arteries, 20.4% (11 patients), basilar arteries, 9.3% (five patients), middle cerebral arteries, 7.4% (four patients), and anterior cerebral artery, 1.9% (one patient) (Table 2, Figure 1). Forty-four patients had intracranial spontaneous dissections (81.5%). Ten patients had extracranial spontaneous dissections (18.5%) and all were located in the internal carotid arteries. Among 29 patients without intracranial hemorrhage, the predominant signs were ischemic symptoms, 48.3% (14 patients), incidental finding, 34.5% (10 patients), cranial nerve compression, 13.8% (four patients), and headache, 3.4% (one patient).

Table 2.

Baseline characteristics and treatment in patients with spontaneous dissection.

Baseline characteristics	Total (44 intracranial artery, 10 extracranial artery)	Intracranial hemorrhage	Nonintracranial hemorrhage
No. (%)	54 (100%)	25 (46.3%)	29 (53.7%)
Age: mean ± SD (range)	54.74 ± 15.39 (15–83)	55.20 ± 16.38 (15–83)	54.34 ± 14.74 (15–81)
Sex: M:F	26:28	11:14	15:14
GCS on admission: median (range)	15 (3–15)	15 (3–15)	15 (5–15)
Location (%)
VA (all intracranial VA)	33 (61.1%)	22	11
ICA (10 extracranial ICA, one intracranial ICA)	11 (20.4%)	0	11
BA	5 (9.3%)	0	5
MCA	4 (7.4%)	2	2
ACA	1 (1.9%)	1	0
Treatment
Conservative treatment	25	8	17
Endovascular treatment	22	12	10
Parent vessel occlusion	6	4	2
Parent vessel preservation	16	8	8
Surgical management	7	5	2
Surgical trapping with or without bypass	4	2	2
Clipping with parent vessel preservation	3	3	0

No.: number of patients; M: male; F: female; GCS: Glasgow Coma Scale; VA: vertebral artery; ICA: internal carotid artery; BA: basilar artery; MCA: middle cerebral artery; ACA: anterior cerebral artery.

Figure 1.

Summary of cerebral dissection management. n: number of patients; VA: vertebral artery; ICA: internal carotid artery; BA: basilar artery; MCA: middle cerebral artery; ACA: anterior cerebral artery.

Endovascular treatment was performed in 22 patients (six parent vessel occlusions, 16 reconstructions) and surgical treatment in seven patients (four trapping with or without bypass, three clipping reconstructions with parent vessel preservation) resulting in favorable outcomes in about 73% and 43% (16/22 and 3/7; p = 0.193), respectively. No difference in favorable outcomes was observed between parent vessel occlusion and selective occlusion with parent vessel preservation (or vessel reconstruction) (70% and 63%, respectively, p = 1.000). Patients who presented with spontaneous dissection without intracranial hemorrhage had more favorable outcomes than those with intracranial hemorrhage (79% and 52%, respectively, p = 0.045) (Tables 2 and 3). The mortality rate of patients with spontaneous dissection was 7.4%. Three patients died from recurrent hemorrhage and one died from ischemic stroke.

Table 3.

Treatment outcomes in spontaneous cerebral artery dissection.

Characteristics and treatment	Favorable (mRS) 0–2 (%)	Unfavorable (mRS) 3–5 (%)	p value
Age (n = 54)			0.569
<60 (n = 21)	13 (62%)	8 (38%)
≥60 (n = 33)	23 (70%)	10 (30%)
GCS on admission (n = 54)			0.273
≥13 (n = 44)	31 (70%)	13 (30%)
<13 (n = 10)	5 (50%)	5 (50%)
Presentation (n = 54)			0.045
Intracranial hemorrhage (n = 25)	13 (52%)	12 (48%)
Nonintracranial hemorrhage (n = 29)	23 (79%)	6 (21%)
Treatment (n = 29)			0.193
Surgical treatment (n = 7)	3 (43%)	4 (57%)
Endovascular treatment (n = 22)	16 (73%)	6 (27%)
Treatment (n = 29)			1.000
Parent vessel occlusion (n = 10)	7 (63%)	3 (37%)
Parent vessel preservation (n = 19)	12 (70%)	7 (30%)

n: number of patients; GCS: Glasgow Coma Scale; mRS: modified Rankin Scale.

Discussion

A dissecting artery is characterized by tearing of the internal elastic lamina (IEL) and/or media, with or without intramural hemorrhage, causing narrowing or dilation of the vessel, following chronic thickening of the intima caused by granulation tissue and repetitive intramural hemorrhage with ruptured fragile neovessels after 14 days.^4,5,11,16 In a study of ruptured intracranial vertebral artery dissection, IEL ruptures were longer than expected from angiographic results.¹⁷ It is assumed that the healing process with neointimal proliferation begins after one week and may not be complete even after one month, depending on the extent of the wall injury. Vascular wall repair by neointimal formation occurs from the disrupted ends of the media proper toward the ruptured portion.¹⁶ The majority of aneurysms have one entrance to the pseudolumen (entry-only type). This type is associated with an unstable clinical course. Some cerebral dissecting aneurysms have both an entrance and exit (entry-exit type). This type of aneurysm occasionally contains a constant flow of blood through the pseudolumen and is clinically more stable than entry-only aneurysms.¹⁸

Traumatic aneurysms were found in less than 1% of all intracranial aneurysms.¹⁹ Clinical presentation warning signs included a cervical bruit, stroke on the secondary computed tomography, LeFort II or III fracture, and basilar skull fracture; any one of these presentations indicated that the patient should receive angiography within 72 hours. Because of the high mortality rate associated with recurrent hemorrhage in about 50% of intracranial traumatic aneurysms, early diagnosis and treatment are recommended. Only 0.65% of traumatic aneurysms were found in blunt traumatic brain injury.^2,20,21 Motor vehicle accidents represented the most common cause of traumatic dissection. In patients without symptoms, antiplatelet therapy was recommended.²²

In this study, we found two patients with hemorrhagic traumatic dissecting aneurysms. Both had ruptured internal carotid artery dissections with poor clinical presentation from a primary brain injury (Table 1, Figure 2). All trauma patients without intracranial hemorrhage had incidental findings during traumatic carotid-cavernous fistula management. In addition, dissections were located on the extradural carotid artery, with good treatment outcomes (Table 1).

Figure 2.

A 15-year-old male presented with altered consciousness after a motor vehicle accident. Computed tomography of the brain showed traumatic subarachnoid hemorrhage and fracture at the base of the skull (a). Left common carotid artery angiography demonstrated an internal carotid dissection with irregular narrowing ((b) and (c)). Computed tomography angiography of the brain after coiling occlusion of the left internal carotid artery (d). Computed tomography of the brain after decompressive craniectomy (e).

In nonhemorrhagic spontaneous cerebral dissection, most cases healed spontaneously in three to six months with medical treatment, and healing after six months was significantly decreased.²³ Cerebral artery dissections accounted for 3.0% of all ischemic strokes. Mechanisms of ischemic symptoms involved thromboembolism, hemodynamic failure, mass effect, and compression of a perforated artery.^4,24,25 In spontaneous dissections, pain is most frequently the initial symptom and is most often characterized as a unilateral headache affecting the frontotemporal area. The typical presentation of internal carotid artery dissection includes the triad of ipsilateral facial pain, partial Horner syndrome (oculosympathetic palsy), and subsequent ischemia. Patients with vertebral artery dissections typically present with pain in the posterior half of the neck, followed by ischemic manifestations in the posterior circulation.^11,23 In atherosclerotic lesions, patients with dissection are usually younger, their lesions are typically isolated, and the stenosis is smooth. Duration of treatment is open to debate. Typically, because dissections heal within six months of onset and rarely recur, treatment should be continued for three to six months.²³

In cerebral dissection without intracranial hemorrhage, recurrent ischemic symptoms, significant stenosis, or dilated of dissecting vessels after follow-up with best medication, treatment with surgical or endovascular procedures is recommended. Some studies have reported ischemic recurrence in about 2%–14%.^5,26 Patients presenting with intracranial hemorrhage carried a risk of recurrent bleeding at about 70%, mostly within 24–72 hours; the mortality rate was about 8.3% but increased to 47% among patients with recurrent hemorrhage.^{6, 8} In this setting, early management should be conducted to prevent recurrent hemorrhage. Parent vessel occlusion was the first choice, either with surgery or endovascular methods. Endovascular trapping using coil embolization is a simple procedure that completely obliterates the dissection. This procedure can be performed in the internal carotid artery with an adequate circle of Willis or a vertebral artery with adequate contralateral flow. Recently, endovascular reconstructive treatment such as stent-assisted coiling embolization or flow diverter use has been reported as an effective treatment modality.^{5,13,15,27,28} Stent-assisted coiling embolization is recommended in dissection with the saccular portion; the coils can obliterate the weak point of the sac and the stent can promote endothelial growth and prevent narrowing of the dissection. However, this technique needs dual-antiplatelet drugs and should be used with caution among patients with hemorrhage (Figure 3).⁵ Parent vessel reconstruction using a flow diverter device might be safe and effective for cerebral dissection that cannot achieve dome protection by any other means.^29,30 The flow diverter device had an overall complete occlusion rate in dissecting aneurysms of about 67%–75%.³¹ Greater protection of proximal rather than distal lesions might be effective to treat dissections.¹⁷

Figure 3.

An 82-year-old woman presented with subarachnoid and intraventricular hemorrhage. Oblique view of right vertebral artery angiography showed multiple aneurysms at the right vertebral artery and basilar artery (a). Stent-assisted coiling embolization was performed ((b) and (c)). Post-embolization angiogram showed nearly total occlusion of the aneurysms (d).

Currently, surgical procedures are mostly reserved for complex lesions that cannot be treated in a minimally invasive fashion. Various surgical procedures have been performed, including direct clipping, trapping with bypass, proximal occlusion, resection with reanastomosis, transposition, aneurysmorrhaphy with thrombectomy, and wrapping. Trapping using a bypass may constitute ideal management, but involves major complications and the dissecting aneurysm may not be completely trapped because of insufficient working space.^13,32 In this study, seven patients were treated using surgical procedures, including three direct clippings, three trappings with bypass, and one trapping. Only three had favorable outcomes. In this study, we preferred endovascular treatment as first-line management for cerebral dissection (Figure 4). When comparing outcomes, endovascular procedures tended to have better functional outcomes than surgical procedures (73% and 43%, respectively, p = 0.193).

Figure 4.

Management strategy for cerebral dissection.

A total of five patients had basilar dissection and all had ischemic symptoms. Only two patients had favorable functional outcomes (Figure 5). Basilar artery dissections are rare lesions associated with significant morbidity and death. The natural course and treatment options for basilar artery dissections differ considerably from those for vertebral artery dissections. Disease management is controversial and difficult, and requires great caution.^23,33 Endovascular treatment is indicated in the acute setting when dissection results in acute arterial occlusion. Among these patients, attempts can be consistently made to restore flow.²³

Figure 5.

A 51-year-old man presented with quadriparesis, dysarthria, and facial palsy. Computed tomography showed dilation and hyperdensity at the basilar artery without contrast opacification (a). Towne and lateral views of right vertebral artery angiography showed no flow to the basilar artery (b). Recanalization of the basilar artery with smooth narrowing of the lumen was observed after mechanical thrombectomy (c). Magnetic resonance imaging showed an intimal flap with intramural hemorrhage (d).

Patients with fibromuscular dysplasia have an increased risk of cervical artery dissection and intracranial aneurysms; whether these patients also have an increased prevalence of intracranial artery dissection is unknown.⁵ Two patients presented with fibromuscular dysplasia in this study and both had cervical carotid dissections. One patient had severe narrowing of the vessel and was successfully treated by stenting (Figure 6).

Figure 6.

A 51-year-old woman presented with subarachnoid hemorrhage from an aneurysm of the superior hypophyseal artery. Lateral and anteroposterior views of left common carotid artery angiography showed a string-of-beads sign and intimal flap of dissection at the proximal internal carotid artery ((a) and (b)). Carotid stenting of the left internal carotid artery was performed (c). Lateral view of left common carotid artery angiography showed remodeling of the vessel wall after stenting (d).

Overall, more than 79% of patients with intracranial artery dissection and without hemorrhage had favorable functional outcomes (mRS ≤ 1 or ≤ 2, or equivalent).^{5,12,25,26,28} In this study, 79% of patients with spontaneous dissection without intracranial hemorrhage had favorable functional outcomes (mRS ≤ 2) after treatment. Only 52% of hemorrhagic patients had favorable outcomes. This confirmed that better functional outcomes occurred in nonhemorrhagic dissection. Because most cases with ruptured spontaneous dissection were located in vertebral arteries (88%), this might not reflect the outcome at other locations. More studies on dissections at each location should be reviewed and investigated.

In this study, the mortality rates of spontaneous dissection were 12% in the hemorrhagic group (three patients) and 3% in the nonhemorrhagic group (one patient). The overall mortality rate was 7.4%. Recent studies reported a mortality rate in intracranial dissection of about 13%–16%,^3,26 and the mortality rate in the nonhemorrhagic group was 0%–3%. The mortality rate for patients with hemorrhage ranged from 19% to 50%.^5,12,25

Conclusion

Most of the traumatic dissections were located on the internal carotid arteries and spontaneous dissections were commonly located on vertebral arteries. Nonhemorrhagic spontaneous cerebral dissections had better functional outcomes after treatment. Endovascular and surgical management were effective treatments by parent vessel occlusion or reconstructions.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Orcid ID

Karanarak Urasyanandana http://orcid.org/0000-0003-2679-3419.