Oct 5, 2012 - KenicHi sato, M.d., PH.d.,1,4 MiKi FujiMura, M.d., PH.d.,3 taKasHi inoue, M.d., PH.d.,2. HiroaKi sHiMizu, M.d., PH.d.,2 aKira taKaHasHi, M.d., ...
J Neurosurg 118:131–139, 2013 ©AANS, 2013
Medullary infarction as a poor prognostic factor after internal coil trapping of a ruptured vertebral artery dissection Clinical article Hidenori Endo, M.D., Ph.D.,1,2 Yasushi Matsumoto, M.D.,1 Ryushi Kondo, M.D., Ph.D.,1 Kenichi Sato, M.D., Ph.D.,1,4 Miki Fujimura, M.D., Ph.D., 3 Takashi Inoue, M.D., Ph.D., 2 Hiroaki Shimizu, M.D., Ph.D., 2 Akira Takahashi, M.D., Ph.D., 4 and Teiji Tominaga, M.D., Ph.D. 5 Departments of 1Neuroendovascular Therapy and 2Neurosurgery, Kohnan Hospital; 3Department of Neurosurgery, National Hospital Organization Sendai Medical Center; and Departments of 4 Neuroendovascular Therapy and 5Neurosurgery, Tohoku University Graduate School of Medicine, Sendai, Japan Object. Internal coil trapping is a treatment method used to prevent rebleeding from a ruptured intracranial vertebral artery dissection (VAD). Postoperative medullary infarctions have been reported as a complication of this treatment strategy. The aim of this study was to determine the relationship between a postoperative medullary infarction and the clinical outcomes for patients with ruptured VADs treated with internal coil trapping during the acute stage of a subarachnoid hemorrhage (SAH). Methods. A retrospective study identified 38 patients who presented between 2006 and 2011 with ruptured VADs and underwent internal coil trapping during the acute stage of SAH. The SAH was identified on CT scanning, and the diagnosis for VAD was rendered by cerebral angiography. Under general anesthesia, the dissection was packed with coils, beginning at the distal end and proceeding proximally. When VAD involved the origin of the posterior inferior cerebellar artery (PICA) with a large cerebellar territory, an occipital artery (OA)–PICA anastomosis was created prior to internal coil trapping. The pre- and postoperative radiological findings, clinical course, and outcomes were analyzed. Results. The internal coil trapping was completed within 24 hours after admission. An OA-PICA anastomosis followed by internal coil trapping was performed in 5 patients. Postoperative rebleeding did not occur in any patient during a mean follow-up period of 16 months. The postoperative MRI studies showed medullary infarctions in 18 patients (47%). The mean length of the trapped VAD for the infarction group (15.7 ± 6.0 mm) was significantly longer than that of the noninfarction group (11.5 ± 4.3 mm) (p = 0.019). Three of the 5 patients treated with OA-PICA anastomosis had postoperative medullary infarction. The clinical outcomes at 6 months were favorable (modified Rankin Scale Scores 0–2) for 23 patients (60.5%) and unfavorable (modified Rankin Scale Scores 3–6) for 15 patients (39.5%). Of the 18 patients with postoperative medullary infarctions, the outcomes were favorable for 6 patients (33.3%) and unfavorable for 12 patients (66.7%). A logistic regression analysis predicted the following independent risk factors for unfavorable outcomes: postoperative medullary infarctions (OR 21.287 [95% CI 2.622–498.242], p = 0.003); preoperative rebleeding episodes (OR 7.450 [95% CI 1.140–71.138], p = 0.036); and a history of diabetes mellitus (OR 45.456 [95% CI 1.993–5287.595], p = 0.013). Conclusions. A postoperative medullary infarction was associated with unfavorable outcomes after internal coil trapping for ruptured VADs. Coil occlusion of the long segment of the VA led to medullary infarction, and an OA-PICA bypass did not prevent medullary infarction. A VA-sparing procedure, such as flow diversion by stenting, is an alternative treatment in the future, if this approach is demonstrated to effectively prevent rebleeding. (http://thejns.org/doi/abs/10.3171/2012.9.JNS12566)
A
Key Words • vertebral artery dissection • internal coil trapping • vascular disorders • endovascular treatment • medullary infarction • subarachnoid hemorrhage n intracranial VAD is a recognized cause of SAH or posterior circulation ischemia in young or middle-aged adults.3 In patients with ruptured VADs,
Abbreviations used in this paper: BTO = balloon test occlusion; DSA = digital subtraction angiography; mRS = modified Rankin Scale; OA = occipital artery; PCoA = posterior communicating artery; PICA = posterior inferior cerebellar artery; SAH = subarachnoid hemorrhage; VA = vertebral artery; VAD = VA dissection; WFNS = World Federation of Neurosurgical Societies.
J Neurosurg / Volume 118 / January 2013
previous studies have emphasized a high incidence of rebleeding and a high mortality rate at the time of rebleeding.16,29 Thus, it has been proposed that patients undergo early obliteration of the ruptured VAD by surgical or endovascular methods to prevent rebleeding.5,6 The advantage of endovascular treatment for a ruptured VAD is that the dissection may be occluded with coils (internal coil trapping) after diagnostic angiography using minimally invasive measures.5,6,12,20,21,24 Postoperative medullary in131
H. Endo et al. farctions have been reported as a complication of internal coil trapping.5,6,12 However, the exact frequency of these infarctions, their topographic pattern, and their relationship to clinical outcomes are still undetermined. The aim of this study was to address these issues relating to the use of internal coil trapping for ruptured VAD during the acute stage of SAH.
Methods Patient Population
We retrospectively reviewed the interventional neuroradiology and neurosurgery databases to identify patients with ruptured intracranial VADs who were treated at the Kohnan Hospital between 2006 and 2011. The inclusion criteria for this study were as follows: the patient had undergone internal coil trapping for a ruptured VAD, the patient’s SAH was identified by CT, and the patient’s diagnosis of VAD was rendered by cerebral angiography (findings included aneurysmal dilation of the intracranial VA with or without the pearl-and-string sign).8 Any cases involving the basilar artery or a traumatic VAD were excluded. Forty-two patients with ruptured VADs were admitted to our institution during the study interval. Thirty-eight of these patients met the inclusion criteria. Four patients who underwent other treatments were excluded from the study. We reviewed the clinical and radiological features of these 38 patients. The clinical and the radiographic data were gathered from both electronic and paper medical records.
Treatment Strategy
An early endovascular internal coil trapping procedure is our preferred treatment method for ruptured VADs. The patients in this study underwent diagnostic 4-vessel DSA prior to endovascular treatment. The vessels were assessed for the size, shape, and location of the VAD with respect to the major branches and collateral vessels. The presence or absence of the contralateral VA, PCoA, and PICA was also noted. The angiograms were examined for extensions of the dissection into the adjacent arterial segments, such as the PICA, basilar artery, anterior spinal artery, or perforating vessels. A BTO was performed using a nondetachable silicone balloon to occlude the parent VA proximal to the dissection, if both the contralateral VA and the PCoA were hypoplastic and the patient’s condition permitted.5,13,21 A neurological evaluation was conducted during the BTO, and collateral flow through the contralateral VA or PCoA was confirmed using angiography. Internal coil trapping was performed under general anesthesia after the diagnostic angiography. A guidewire was advanced to a position distal to the dissection, followed by a microcatheter to maintain the true lumen of the vessel. The microcatheter was pulled back to the distal portion of the dissection. The entire dissection was then occluded with coils, starting distally and proceeding proximally. The trapping was complete if the following criteria were satisfied: the ipsilateral vertebral angiogram showed total obstruction of antegrade flow, and the contralateral vertebral angiogram demonstrated 132
cessation of retrograde filling of the dissected segment. When the dissection involved the origin of the PICA with a large cerebellar territory, an OA-PICA anastomosis was created prior to internal coil trapping. The OA-PICA anastomosis and internal coil trapping were performed on the same day. The PICA was preserved during the internal coil trapping when the PICA arose from the margin of the dissection. The 2 cases involving the thin PICA, which supplied only a small part of the lower vermis, led to the PICA being sacrificed. The amount of heparin injected at the beginning of the procedure was 3000 U, and 1000 U was given every hour. Intraoperative anticoagulation was not used in patients with an episode of rebleeding. One patient with irregular stenosis without aneurysmal dilation of the affected VA was treated with surgical trapping and was excluded from the study. Three of the 8 patients examined by BTO were not considered to have ischemic tolerance. Two of these 3 patients were treated conservatively, and 1 patient was treated with stent-assisted coiling. These 3 patients were excluded from the study. Clinical and Radiological Outcome Measures
Conventional DSA and 3D DSA were performed using Innova 3131 technology (GE Healthcare). The 3D volume-rendering images were developed using an Advantage 4.4 workstation (GE Healthcare). The length of the trapped VAD and the distance from the distal end of dissection to the vertebrobasilar junction were measured using this workstation. All MRI studies were performed using a Signa 1.5- or 3-T system (GE Healthcare) with a standard head coil. The diffusion- and T2-weighted imaging studies were obtained the day after the procedure. The lesions with high signal intensity detected by diffusion- and/or T2-weighted imaging were regarded as areas of ischemia. The topographic patterns of the postoperative medullary infarctions were determined using the schemes outlined in previous reports.4,7,27 Additional MRI and DSA studies were obtained if symptomatic vasospasm was suspected during the early postoperative period. The follow-up 3-T MRI and MRA studies obtained after discharge from the hospital were scheduled at 6, 12, 24, and 48 months after treatment. Each patient’s functional status was evaluated using the mRS.26 To compare clinical outcomes, we divided the patients into 2 groups: those with favorable outcomes (mRS Scores 0–2) and those with unfavorable outcomes (mRS Scores 3–6). The functional status 6 months after the treatment was the time point for the assessment of the final outcome. Statistical Analysis
Statistical analysis was performed using JMP Pro 9 for Macintosh. A Mann-Whitney U-test was used to assess the continuous variables, and the Fisher exact test was used to assess the categorical variables. The following variables were analyzed: age, sex, side of the VAD, hypertension, diabetes mellitus, hyperlipidemia, preoperative WFNS score, any preoperative rebleeding, a postoperative medullary infarction, the presence of a ventriculoperitoneal shunt for hydrocephalus, symptomatic vasospasm, the location of the VAD and OA-PICA J Neurosurg / Volume 118 / January 2013
Internal coil trapping for ruptured vertebral artery dissection bypass if present, the length of the trapped VAD, and the distance between the distal end of the dissection to the vertebrobasilar junction. Univariate analysis was performed to determine the association between the clinical outcomes and other factors. Logistic regression analysis with a forward stepwise method was then performed to determine the independent association between the clinical outcomes and other factors. The univariate cutoff for inclusion in the logistic regression analysis was p < 0.10. Univariate analysis was also performed to determine the association between the medullary infarction and other factors. The limit for statistical significance was set at p < 0.05 for a 95% CI.
Results Baseline Characteristics and Preoperative Course
Thirty-eight patients (19 men and 19 women), ranging in age from 30 to 82 years (mean 53 years), were treated for ruptured VADs with internal coil trapping. A preoperative CT scan revealed SAH in all 38 patients. Preoperative rebleeding occurred in 15 patients (39.5%). Rebleeding in 6 of 15 patients was confirmed by CT scanning, and the rebleeding in the remaining 9 patients was diagnosed based on clinical symptoms (including sudden-onset coma with respiratory arrest). The preoperative WFNS grade distribution was as follows: 10 patients (26.3%) were Grade I, 9 (23.7%) were Grade II, 8 (21.1%) were Grade IV, and 11 (28.9%) were Grade V. Twentyeight patients (73.7%) were admitted to our institution on Day 0, 7 patients (18.4%) were admitted on Day 1, 1 patient (2.6%) was admitted on Day 2, and 2 patients (5.3%) were admitted on Day 15.
Angiographic Findings
Angiography showed that all 38 patients had unilateral VADs; in 27 patients (71.1%) the VADs occurred on the right side (Table 1). An aneurysmal dilation of the affected VA was confirmed in all 38 patients. The pearland-string sign was observed on angiograms for 23 patients (60.5%). The VADs were located between the origin of the PICA and the vertebrobasilar junction in 14 patients (36.8%), involved the origin of the PICA in 9 patients (23.7%), were located in the VA without definite PICAs in 8 patients (21.1%), and were located proximal to the origin of the PICA in 7 patients (18.4%).
Endovascular Treatment
Internal coil trapping was completed within 24 hours after admission. In all cases, complete occlusion of the dissection was confirmed by a final angiography study. An OA-PICA anastomosis followed by internal coil trapping was performed in 5 of 9 patients with PICA involvement during the acute SAH stage. The PICA was preserved during internal coil trapping in 2 patients with PICA involvement because the PICA arose from the margin of the dissection. In the other 2 patients, the thin PICA was sacrificed, because thin PICAs supplied only a small part of the lower vermis. Lumbar drainage or ventricular drainage of the CSF to improve the acute hydrocephalus after
J Neurosurg / Volume 118 / January 2013
SAH was performed in 15 and 4 patients, respectively. Postoperative rebleeding did not occur in any case during a mean follow-up period of 16 months (range 4 days–65 months). Three patients exhibited recanalization in the early postoperative period (postoperative Days 21, 27, and 31). One of these 3 patients, who had been initially treated with an OA-PICA anastomosis followed by internal coil trapping, underwent proximal clip occlusion of the affected VA without complications. The other 2 patients, who were initially treated using internal coil trapping with preservation of the PICA arising from the proximal end of the dissection, underwent OA-PICA anastomosis with additional coil embolization without complications. Two patients (5.3%) had symptomatic vasospasm, which was treated with intraarterial administration of fasudil hydrochloride. One of these patients developed a cerebral infarction in the left temporal lobe, resulting in a right hemiparesis with sensory aphasia. A ventriculoperitoneal shunt for hydrocephalus was placed in 10 patients (26.3%). Recanalization of the trapped VADs occurred in 2 of 24 patients who underwent 3-T MRI over a 6-month period. In these 2 patients, neither rebleeding nor progression had occurred by the 48-month follow-up point. Clinical Outcomes and Prognostic Factors
Postoperative MRI was performed in all patients, and medullary infarctions were observed in 18 patients (47.4%). The topographical pattern of medullary infarctions, which was determined using the schemes from previous reports,4,7,27 included 7 dorsolateral infarctions (38.9%), 5 large inferodorsolateral infarctions (27.8%), 3 hemimedullary infarctions (16.7%), 2 paramedian infarctions (11.1%), and 1 dorsal infarction (5.6%) (Fig. 1). Later, we investigated the risk factors for postoperative medullary infarctions (Table 1). Among the variables examined, which included age, sex, side, WFNS grade, preoperative rebleeding, length of the trapped VAD by platinum coils, distance from the distal end of the dissection to the vertebrobasilar junction, VAD location, and OA-PICA bypass, only age and the length of the trapped VAD were identified as significant factors for predicting the occurrence of a postoperative medullary infarction. The mean length of the trapped VAD in the infarction group was significantly longer than that in the noninfarction group. Next, we investigated the clinical outcomes and prognostic factors (Table 2). The clinical outcomes at 6 months were favorable (mRS Scores 0–2) in 23 patients (60.5%) and unfavorable (mRS Scores 3–6) in 15 patients (39.5%). The univariate analysis demonstrated that the following factors were significantly different in the favorable and unfavorable outcome groups: age, history of diabetes mellitus, preoperative rebleeding, and postoperative medullary infarction. Nine (47.4%) of 19 patients with high-grade SAH (WFNS Grade IV or V) had favorable outcomes, which were not significant risk factors for unfavorable outcomes. Furthermore, postoperative medullary infarction, preoperative rebleeding, and diabetes mellitus were independent risk factors for an unfavorable outcome in the logistic regression analysis. Of the 18 patients with postoperative medullary infarction, the outcomes were favorable for 6 patients (33.3%) and un133
H. Endo et al. TABLE 1: Risk factors for medullary infarction following endovascular trapping of the ruptured VAD Value* Variable mean age no. of men VAD on rt side WFNS Grade I or II preop rebleeding mean length of trapped VAD in mm mean length to vertebrobasilar junction in mm VAD location proximal to PICA distal to PICA absence of PICA PICA involvement OA-PICA bypass
Infarction Group (n = 18)
Noninfarction Group (n = 20)
56 ± 9 12 (63) 12 (44) 8 (42) 10 (67) 15.7 ± 6.0 10.9 ± 4.8
50 ± 14 7 (37) 15 (56) 11 (58) 5 (33) 11.5 ± 4.3 9.9 ± 4.3
3 (43) 6 (43) 5 (62) 4 (44) 3 (60)
4 (57) 8 (57) 3 (38) 5 (56) 2 (40)
Total (n = 38) 19 27 19 15
7 14 8 9 5
p Value (univariate) 0.036‡ 0.103† 0.724† 0.746† 0.096† 0.019‡ 0.45‡ 1.000† 0.745† 0.438† 1.000† 0.653†
* Mean values are presented as the mean ± SD. All other values are the number of patients (%). The percentages are based on the number of patients per category. † Fisher exact test. ‡ Mann-Whitney U-test.
favorable for 12 (66.7%) (Table 3). The large infarctions, especially hemimedullary and inferodorsolateral infarctions, usually caused unfavorable outcomes. Six patients (15.8%) died during the follow-up period; 3 deaths were due to pneumonia, which occurred later in the follow-up period; 1 death was caused by an acute respiratory problem, which occurred in the early postoperative period; 1 death was the result of initial brain damage caused by severe SAH; and 1 death resulted from a ruptured middle cerebral artery dissection that led to a second SAH. Four of the 6 patients who died of pneumonia or an acute respiratory problem were also diagnosed with postoperative hemimedullary or large inferodorsolateral medullary infarctions.
Case 14
Illustrative Case
This 51-year-old man presented with sudden onset of severe headache associated with transient left hemiparesis. Computed tomography imaging revealed an SAH predominantly localized in the posterior fossa. Cerebral angiography showed segmental aneurysmal dilation of the right intracranial VA with arterial narrowing proximal to the dilation (Fig. 2A), which confirmed the diagnosis of a ruptured VAD. A well-developed PICA arose from the distal end of the dissection. Subsequently, internal coil trapping of the dissection was performed under general anesthesia. A 6-Fr guiding catheter was introduced into the right VA, and a microcatheter was advanced coaxially into the VAD. The dissection and the affected VA were successfully occluded with bare platinum coils (Fig. 2B and C). The final angiographic study showed sufficient collateral flow to the posterior circulation, including the
134
right PICA from the left VA. A severe lateral medullary syndrome with right-sided hemiparesis became apparent after recovery from general anesthesia. Diffusion-weighted imaging was performed on postoperative Day 1 and revealed a large right inferodorsolateral medullary infarction (Fig. 2D). The postoperative course was uneventful. The patient did not develop symptomatic vasospasm or hydrocephalus. The patient required long-term rehabilitation due to right ataxic hemiparesis and dysphagia. The symptoms partially resolved, and the patient could walk with a cane at the 12-month follow-up (mRS Score 3).
Discussion
In this paper, we report the outcomes of ruptured intracranial VADs treated with internal coil trapping. Although the long-term efficacy of internal coil trapping to prevent recurrent bleeding after an SAH has been confirmed, the occurrence of medullary infarctions, which are treatment-related complications, correlated with unfavorable outcomes at the 6-month follow-up. Urgent surgical treatment has been justified because of the high incidence of recurrent bleeding in ruptured VADs without treatment.16,28,29 In a study by Yamada et al.28 of 24 conservatively treated patients with ruptured VADs, 14 patients (58%) experienced a total of 35 rebleeding episodes. Of these 14 patients, 13 died, and 11 of these deaths were directly attributable to rebleeding. The primary goal of surgical treatment is to prevent rebleeding, which can be accomplished surgically28 or endovascularly.1,5,6,12,13,19–21,24 Recently, the less invasive endovascular method is becoming the first line of therapy, especially during the acute phase after SAH. Proximal occlusion does not necessarily eliminate the risk of rebleeding,23 and internal coil trapping of J Neurosurg / Volume 118 / January 2013
Internal coil trapping for ruptured vertebral artery dissection
Fig. 1. Postoperative MRI studies of representative cases with medullary infarction as follows: hemimedullary infarction (Case 2) (A), large inferodorsolateral infarction (Case 6) (B), paramedian infarction (Case 18) (C), dorsolateral infarction (Case 4) (D), and dorsal infarction (arrowhead) (Case 16) (E). The case numbers correspond to those in Table 3. The values of the mean length of the trapped VA and the mean length from the distal end of dissection to the vertebrobasilar junction are shown and are presented in millimeters. The images in panels C and D have been reversed for convenience. TABLE 2: Predictors of clinical outcome for ruptured VAD treated by endovascular trapping Value* Variable mean age (yrs) male sex rt side history hypertension diabetes mellitus type 2 hyperlipidemia WFNS Grade IV or V preop rebleeding medullary infarction shunt for hydrocephalus symptomatic vasospasm PICA involvement OA-PICA bypass
p Value
Favorable Outcome (n = 23)
Unfavorable Outcome (n = 15)
Total (n = 38)
Univariate†
50 ± 12 10 (43) 16 (70)
57 ± 10 9 (60) 11 (73)
19 27
0.034 0.508 1.000
13 (57) 1 (4) 2 (9) 9 (39) 5 (22) 6 (26) 7 (30) 0 (0) 6 (26) 5 (22)
6 (40) 4 (27) 3 (20) 10 (67) 10 (67) 12 (80) 3 (20) 2 (13) 3 (20) 2 (13)
19 5 5 19 15 18 10 2 9 7
0.507 0.069 0.365 0.184 0.008 0.002 0.709 0.149 1.000 0.681
Regression (OR, 95% CI) 0.759 (1.015, 0.920–1.119) 0.795 (1.334, 0.136–12.261)
0.013 (45.456, 1.993–5287.595)
0.036 (7.450, 1.140–71.138) 0.003 (21.287, 2.622–498.242)
* Mean values are presented as the mean ± SD. All other values are the number of patients (%). Percentages are based on the number of patients per outcome. A favorable outcome is considered for patients with an mRS score of 0–2, and an unfavorable outcome is considered for patients with an mRS score of 3–6. † Difference between subgroups by Fisher exact text except for age (Mann-Whitney U-test).
J Neurosurg / Volume 118 / January 2013
135
136
63, M
48, M
49, F 63, F
59, M
52, M
52, M
82, M 68, F 56, M
45, F 52, M
51, M
58, M 60, M
46, F 55, M
2
3
4 5
6
7
8
9 10 11
12 13
14
15 16
17 18
rt lt
lt lt
rt
rt rt
rt lt rt
rt
lt
rt
rt lt
rt
rt
rt
IV V
IV I
I
I V
II V V
V
II
V
I IV
I
V
I
yes yes
no no
no
no yes
yes yes yes
yes
yes
yes
no no
no
yes
no
VAD Location
absence of PICA
absence of PICA distal to PICA
proximal to PICA trapping trapping trapping
trapping
trapping
trapping
trapping trapping
proximal to PICA trapping trapping
trapping
pearl & string PICA involvement bypass + trapping pearl distal to PICA trapping
pearl & string distal to PICA pearl & string distal to PICA
pearl
pearl & string PICA involvement bypass + trapping pearl & string PICA involvement trapping
pearl & string absence of PICA pearl distal to PICA pearl distal to PICA
pearl
trapping
trapping
Treatment
PICA involvement bypass + trapping
pearl & string absence of PICA
pearl
pearl pearl
pearl
pearl & string absence of PICA
pearl & string proximal to PICA
Angiography Finding
no no
no no
no
no yes
no no no
no
no
no
no no
no
no
no
PICA Sacrifice
no no
yes no
no
yes yes
yes no no
no
no
yes
yes no
no
no
no
VPS
* MCA = middle cerebral artery; M & M = Morbidity and Mortality; VPS = ventriculoperitoneal shunt; VS = vasospasm.
49, F
1
Case Age (yrs), WFNS No. Sex Side Score Rebleeding
TABLE 3: Characteristics of patients with postoperative medullary infarction*
no no
no no
no
no no
no no no
no
yes
no
no yes
no
no
no
Symptomatic VS
hemimedullary paramedian
dorsolateral hemimedullary large inferodor solateral dorsolateral large inferodor solateral large inferodor solateral dorsolateral dorsal
dorsolateral
large inferodor solateral dorsolateral
dorsolateral dorsolateral
paramedian
large inferodor solateral hemimedullary
Pattern of Medullary Infarction
6 3
0 1
3
1 5
5 6 2
5
3
6
0 6
1
5
6
Final mRS Score
none mild lt ataxia, rt sensory distur bance acute respiratory problem right hemiparesis, dysphagia
mild ataxia persistent disturbance of con sciousness, tetraparesis rt ataxic hemiparesis, dysphasia
rt hemiparesis & sensory apha sia caused by VS persistent disturbance of con sciousness, tetraparesis muscle weakness aspiration pneumonia rt moderate ataxia, lt numbness
persistent disturbance of con sciousness, tetraparesis mild lt hemiparesis, mild sen sory disturbance none brain death caused by 2nd SAH (ruptured MCA dissection) aspiration pneumonia
aspiration pneumonia
M & M at 6 Mos
H. Endo et al.
J Neurosurg / Volume 118 / January 2013
Internal coil trapping for ruptured vertebral artery dissection
Fig. 2. Case 14. A: Preoperative right VA angiogram showing right intracranial VAD. B: Postoperative left VA angiogram showing the complete obliteration of the dilated segment by platinum coils. C: Radiograph revealing the shape of the coil mass packed into the dissected site. D: Diffusion-weighted imaging study obtained on postoperative Day 1 showing a large right inferodorsolateral medullary infarction.
the dissection is considered an effective treatment option for ruptured VADs.5,6 In this study, the favorable outcomes for patients with ruptured VADs treated with internal coil trapping were observed in 60.5% of cases. Similar findings were described in recent studies regarding these clinical outcomes.6,12,21,24 We performed aggressive endovascular treatment during the acute stage of SAH, immediately after admission, in patients with high-grade SAH (WFNS Grade IV or V). High-grade SAH was reported to be a risk factor for unfavorable outcomes after endovascular treatment for ruptured VADs.12,24 However, high-grade SAH was not identified as a statistically significant predicting factor for unfavorable outcomes in this study. Nine (47.4%) of 19 patients with high-grade SAHs had favorable outcomes. Surgical indications for high-grade SAH should be determined by the response to CSF drainage and medical intervention.25 However, it would be reasonable to perform endovascular treatment as quickly as possible to prevent death in some of the patients with high-grade SAH, considering the high incidence of rebleeding prior to surgical treatment. Postoperative medullary infarctions have been reported as unresolved complications of internal coil trapping.5,6,12 The use of the coils to trap the dissection can lead to an occlusion of the perforating artery arising from the VA or PICA, thus resulting in the formation of a medullary infarction.5,6 Lateral medullary syndrome occurred in 5.6%–36.0% of cases of ruptured VAD after endovascular treatment.5,6,12 There may be unidentified cases of postoperative medullary infarctions because it is difficult J Neurosurg / Volume 118 / January 2013
to recognize symptoms of brainstem ischemia during the acute stage of an SAH. This study indicated that the MRI-based ischemic complication rate after internal coil trapping contributed to the high detection rate of postoperative medullary infarctions. The BTO might effectively assess the ischemic tolerance at the time of the trapping for the affected VA.5,6 However, it would be difficult to assess perforating artery ischemia using balloon occlusion of the proximal VA. Anatomical studies have shown that perforating vessels arising from the PICA or proximal VA are responsible for the vascularization of the posterior surface of the medulla oblongata. A lateral medullary infarction may occur if the artery is compromised.15 The anatomical variations of these perforators largely depend on the development of the PICA and the location of the PICA origin.15 In contrast, the critical perforators arise from the distal VA at a site approximately 14 mm proximal and 16 mm distal to the vertebrobasilar junction.14 These perforators enter the foramen cecum, and the occlusion of these vessels results in medial medullary infarctions.14 In this study, 16 patients were diagnosed with medullary infarctions involving the lateral side of the medulla oblongata, and 5 patients were diagnosed with hemimedullary or paramedian infarctions involving the medial side of the medulla oblongata. The development and location of the PICA and the distance from the dissection to the vertebrobasilar junction did not differ between the infarction group and the noninfarction group. The length of the trapped VAD by platinum coils was found to differ between the 2 groups. These results suggested that trapping of the longer segments of the intracranial VA was associated with the obliteration of more perforators that supply the medulla oblongata. To date, there have been no reports regarding the extent of VA occlusion in internal coil trapping. The histopathological analysis of the ruptured VAD revealed that most of the ruptured sites corresponded to the segments with the greatest dilations on the angiograms.17 Therefore, occlusion of these dilated segments of the VADs might be sufficient to prevent rebleeding. Ro et al.22 reported that the adventitia was distinctively observed at the ruptured lesion site in all 58 of their autopsy cases. However, multiple tears in the internal elastic lamina were recorded and extended over the ruptured focus of the adventitia.17,22 Thus, our strategy was to pack the entire segment with coils, including the dilated and stenotic segments. The locations of the PICA, anterior spinal artery, and perforators, which should be preserved during embolization, affect the length of the VA occlusion, although small perforators usually are invisible on angiograms.20 The dilated segment of the VAD should be adequately packed with coils to promote thrombosis of the rupture site and to serve as a point of origin for the proximal coils. The coil packing should be completed at the entrance to the dissection. The coils should not be overpacked at the proximal VA, because it may lead to an extension of the coil mass. Medullary infarction occurred in 3 patients treated with OA-PICA anastomosis followed by internal coil trapping, which indicates that OA-PICA anastomosis does not guarantee blood flow in the perforators. In cases 137
H. Endo et al. in which the perforators arise from the PICA, a medullary infarction could still occur after internal coil trapping with OA-PICA anastomosis. The OA was anastomosed to the caudal loop of the PICA, and the perforators arising from the proximal segment of PICA would be supplied by the retrograde flow from the bypass. However, these perforators could be occluded by the thromboembolism during the internal coil trapping procedure even after an uneventful bypass procedure. A medullary infarction may not be a treatment-related complication and may represent the natural course of the disease. This ischemic condition can occur concomitantly with the onset of the SAH, because VADs are major causes of medullary infarction.7,10,11 Kameda et al.7 reported that 29% of the lateral medullary infarctions and 21% of the medial medullary infarctions were caused by VADs. The sizes of the medullary infarctions in patients with VADs were larger than in patients with atherosclerosis. The greater size is due to the simultaneous occlusion of multiple perforators arising from the VA that occurs with the VADs.11 In this study, 9 of 18 patients with postoperative medullary infarctions had hemimedullary or large inferodorsolateral infarctions. The clinical outcomes of cases with unruptured VADs were relatively favorable.8 In a study of 102 patients with unruptured VADs who presented with ischemic symptoms, the clinical outcomes were favorable (mRS Scores 0 and 1) in 90.2% and unfavorable (mRS Scores 2–6) in 9.8% of the cases.8 However, only 61% of the cases had favorable outcomes (mRS Scores 0–2) in our study. The outcomes could have been affected by the medullary infarctions and the SAHs. The patients with large ischemic lesions tended to have severe dysphagia and aspiration pneumonia.9 In the current study, 4 patients with large medullary infarctions died of respiratory problems. In contrast, the patients with small ischemic lesions, such as the dorsolateral or dorsal type, resulted in favorable outcomes when the cases revealed low-grade SAH and had no other complications. In open surgery, the perforating artery could be sacrificed if the origin is involved in the dissection.16 Furthermore, lower cranial nerve palsy caused by surgical manipulation could occur postoperatively, particularly in lesions involving the PICA.2,16 Treatment with a combination of stents and coils or with stents alone preserves the parent artery, and it is a plausible alternative to an internal coil trapping procedure.1,19 The occlusion of the perforators did not occur in any of the 26 patients treated with a Neuroform stent or the balloon-expandable coronary stent.19 However, the follow-up angiography revealed complete obliteration of the VAD in only 50% of these cases. Narata et al.18 recently reported that a flow-diverting stent is an option when treating ruptured VADs. In that report, 3 patients were treated with Pipeline embolization devices, and no recanalization or rebleeding occurred during the 3-month follow-up. The authors proposed this approach as an option when occlusion is too risky or is not possible. There is a risk of acute occlusion after stenting during the acute stage of SAH because of insufficient antiplatelet therapy. Thus, the treatment selection of one of these techniques for individual patients might represent the future treatment approach for ruptured VADs. 138
Conclusions
A postoperative medullary infarction was associated with unfavorable outcomes after internal coil trapping for ruptured VADs. Coil occlusion of the long segment of the VA led to medullary infarction. An OA-PICA bypass did not prevent medullary infarction. A VA sparing procedure, such as flow diversion by stent, could be an alternative treatment in the future, if this approach is demonstrated to effectively prevent rebleeding. Disclosure The authors report no conflict of interest concerning the materials or methods used in this study or the findings specified in this paper. Author contributions to the study and manuscript preparation include the following. Conception and design: Endo. Acquisition of data: Endo, Matsumoto, Kondo, Sato. Analysis and interpretation of data: Endo, Matsumoto, Kondo, Sato. Drafting the article: Endo. Critically revising the article: all authors. Reviewed submitted version of manuscript: all authors. Approved the final version of the manuscript on behalf of all authors: Endo. Statistical analysis: Endo. Study supervision: Shimizu, Takahashi, Tominaga. References 1. Ahn JY, Han IB, Kim TG, Yoon PH, Lee YJ, Lee BH, et al: Endovascular treatment of intracranial vertebral artery dissections with stent placement or stent-assisted coiling. AJNR Am J Neuroradiol 27:1514–1520, 2006 2. Al-khayat H, Al-Khayat H, Beshay J, Manner D, White J: Vertebral artery-posteroinferior cerebellar artery aneurysms: clinical and lower cranial nerve outcomes in 52 patients. Neurosurgery 56:2–11, 2005 3. Arnold M, Bousser MG, Fahrni G, Fischer U, Georgiadis D, Gandjour J, et al: Vertebral artery dissection: presenting findings and predictors of outcome. Stroke 37:2499–2503, 2006 4. Bassetti C, Bogousslavsky J, Mattle H, Bernasconi A: Medial medullary stroke: report of seven patients and review of the literature. Neurology 48:882–890, 1997 5. Hamada J, Kai Y, Morioka M, Yano S, Todaka T, Ushio Y: Multimodal treatment of ruptured dissecting aneurysms of the vertebral artery during the acute stage. J Neurosurg 99: 960–966, 2003 6. Iihara K, Sakai N, Murao K, Sakai H, Higashi T, Kogure S, et al: Dissecting aneurysms of the vertebral artery: a management strategy. J Neurosurg 97:259–267, 2002 7. Kameda W, Kawanami T, Kurita K, Daimon M, Kayama T, Hosoya T, et al: Lateral and medial medullary infarction: a comparative analysis of 214 patients. Stroke 35:694–699, 2004 8. Kim BM, Kim SH, Kim DI, Shin YS, Suh SH, Kim DJ, et al: Outcomes and prognostic factors of intracranial unruptured vertebrobasilar artery dissection. Neurology 76:1735–1741, 2011 9. Kim JS: Pure lateral medullary infarction: clinical-radiological correlation of 130 acute, consecutive patients. Brain 126: 1864–1872, 2003 10. Kim JS, Han YS: Medial medullary infarction: clinical, imaging, and outcome study in 86 consecutive patients. Stroke 40: 3221–3225, 2009 11. Kim JS, Lee JH, Choi CG: Patterns of lateral medullary infarction: vascular lesion-magnetic resonance imaging correlation of 34 cases. Stroke 29:645–652, 1998 12. Lee JM, Kim TS, Joo SP, Yoon W, Choi HY: Endovascular treatment of ruptured dissecting vertebral artery aneurysms— long-term follow-up results, benefits of early embolization, and predictors of outcome. Acta Neurochir (Wien) 152:1455– 1465, 2010
J Neurosurg / Volume 118 / January 2013
Internal coil trapping for ruptured vertebral artery dissection 13. Leibowitz R, Do HM, Marcellus ML, Chang SD, Steinberg GK, Marks MP: Parent vessel occlusion for vertebrobasilar fusiform and dissecting aneurysms. AJNR Am J Neuroradiol 24:902–907, 2003 14. Mahmood A, Dujovny M, Torche M, Dragovic L, Ausman JI: Microvascular anatomy of foramen caecum medullae oblongatae. J Neurosurg 75:299–304, 1991 15. Mercier PH, Brassier G, Fournier HD, Picquet J, Papon X, Lasjaunias P: Vascular microanatomy of the pontomedullary junction, posterior inferior cerebellar arteries, and the lateral spinal arteries. Interv Neuroradiol 14:49–58, 2008 16. Mizutani T, Aruga T, Kirino T, Miki Y, Saito I, Tsuchida T: Recurrent subarachnoid hemorrhage from untreated ruptured vertebrobasilar dissecting aneurysms. Neurosurgery 36:905– 913, 1995 17. Mizutani T, Kojima H, Asamoto S, Miki Y: Pathological mechanism and three-dimensional structure of cerebral dissecting aneurysms. J Neurosurg 94:712–717, 2001 18. Narata AP, Yilmaz H, Schaller K, Lovblad KO, Pereira VM: Flow-diverting stent for ruptured intracranial dissecting aneurysm of vertebral artery. Neurosurgery 70:982–989, 2012 19. Park SI, Kim BM, Kim DI, Shin YS, Suh SH, Chung EC, et al: Clinical and angiographic follow-up of stent-only therapy for acute intracranial vertebrobasilar dissecting aneurysms. AJNR Am J Neuroradiol 30:1351–1356, 2009 20. Peluso JP, van Rooij WJ, Sluzewski M, Beute GN, Majoie CB: Endovascular treatment of symptomatic intradural vertebral dissecting aneurysms. AJNR Am J Neuroradiol 29:102–106, 2008 21. Rabinov JD, Hellinger FR, Morris PP, Ogilvy CS, Putman CM: Endovascular management of vertebrobasilar dissecting aneurysms. AJNR Am J Neuroradiol 24:1421–1428, 2003 22. Ro A, Kageyama N, Abe N, Takatsu A, Fukunaga T: Intracranial vertebral artery dissection resulting in fatal subarachnoid hemorrhage: clinical and histopathological investigations from a medicolegal perspective. Clinical article. J Neurosurg 110:948–954, 2009
J Neurosurg / Volume 118 / January 2013
23. Sugiu K, Tokunaga K, Ono S, Nishida A, Date I: Rebleeding from a vertebral artery dissecting aneurysm after endovascular internal trapping: adverse effect of intrathecal urokinase injection or incomplete occlusion? Case report. Neurol Med Chir (Tokyo) 49:597–600, 2009 24. Sugiu K, Tokunaga K, Watanabe K, Sasahara W, Ono S, Tamiya T, et al: Emergent endovascular treatment of ruptured vertebral artery dissecting aneurysms. Neuroradiology 47:158–164, 2005 25. Suzuki M, Otawara Y, Doi M, Ogasawara K, Ogawa A: Neurological grades of patients with poor-grade subarachnoid hemorrhage improve after short-term pretreatment. Neurosurgery 47:1098–1105, 2000 26. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J: Interobserver agreement for the assessment of handicap in stroke patients. Stroke 19:604–607, 1988 27. Vuilleumier P, Bogousslavsky J, Regli F: Infarction of the lower brainstem. Clinical, aetiological and MRI-topographical correlations. Brain 118:1013–1025, 1995 28. Yamada M, Kitahara T, Kurata A, Fujii K, Miyasaka Y: Intracranial vertebral artery dissection with subarachnoid hemorrhage: clinical characteristics and outcomes in conservatively treated patients. J Neurosurg 101:25–30, 2004 29. Yamaura A, Watanabe Y, Saeki N: Dissecting aneurysms of the intracranial vertebral artery. J Neurosurg 72:183–188, 1990
Manuscript submitted March 16, 2012. Accepted September 6, 2012. Please include this information when citing this paper: published online October 5, 2012; DOI: 10.3171/2012.9.JNS12566. Address correspondence to: Hidenori Endo, M.D., Department of Neuroendovascular Therapy, Kohnan Hospital, 4-20-1 Nagamachiminami, Taihaku-ku, Sendai 982-8523, Japan. email: hideendo@ gmail.com.
139
