Indications and Outcomes of Prophylactic and Therapeutic Extracranial-to-intracranial Arterial Bypass for Cerebral Revascularization.

BACKGROUND: Extracranial-to-intracranial (EC-IC) arterial bypass is a technically demanding procedure used to treat complex cerebral artery diseases. The indications, proper surgical techniques, and outcomes of this procedure have been under debate over the recent decades.
METHODS: Between January 2004 and December 2012, 28 patients, including patients with cerebral artery occlusion, intracranial aneurysm, cranial base tumor, and Moyamoya disease, underwent EC-IC bypass. Patients' records were retrospectively reviewed for demography, indications, complications, high-flow versus low-flow bypass, patency rate of bypass, and neurological outcome. The patients were sorted into prophylactic (n = 16) and therapeutic (n = 12) groups based on the preoperative presentation of their neurological symptoms. Follow-up evaluation was performed at a mean of 32.7 ± 24.3 months.
RESULTS: The overall patency rate of bypass was 100%, the postoperative stroke rate was zero, and the surgical complication rate was 14.3%. There was no significant difference in the bypass patency rate between the 2 groups or between the high-flow and low-flow bypass patients. Patients who underwent prophylactic bypass had minimal surgical and total complications (P = 0.03 and P < 0.01, respectively) and a better neurological outcome. Surgical complications were more common in patients who underwent therapeutic bypass (25%).
CONCLUSIONS: The collaboration of neurosurgeons and plastic surgeons in performing EC-IC bypass can result in excellent outcomes with a high bypass patency rate and few complications, particularly for prophylactic EC-IC bypass.

In 1970, Yasargil et al[1] reported the first extracranial-to-intracranial (EC-IC) arterial bypass, in which the superficial temporal artery (STA) was anastomosed to the middle cerebral artery (MCA) as a treatment for an occluded internal carotid artery (ICA). The procedure gained some popularity, and the indications have recently expanded to include not only cerebral ischemia but also Moyamoya disease (MMD, which is marked by the gradual stenosis of bilateral internal carotid arteries, typically affects teenagers and young adults, and is largely resistant to medical treatment),[2-5] intracranial aneurysms, and intracranial tumors.[6]

In 1985, a randomized trial failed to demonstrate the superiority of EC-IC bypass to the best medical treatments available for reducing the risk of ischemic stroke.[7] As a consequence, the frequency of these operations declined sharply.[8,9] However, this study was later criticized for several design flaws, including inappropriate cohort assignments and the involvement of mostly low-volume centers and surgeons.[10] A post hoc analysis confirmed that EC-IC bypass is indeed an effective treatment for MMD and for atherosclerotic disease, particularly for those with poor cerebrovascular reserve, inadequate collateral flow, and posterior circulation disease,[11] and Schaller[12] later demonstrated that neurologic function and stroke risk can be significantly improved by EC-IC bypass.[10] Currently, there are many conditions for which EC-IC bypass is becoming the standard therapy, including MMD, certain tumors encasing the branches of the ICA in the skull base,[13,14] and giant aneurysms that require ligation or the clipping of the ICA or its branches.[14,15]

Its resurgence in popularity notwithstanding, EC-IC bypass remains a technically demanding procedure that requires considerable microvascular skill and experience. In this light, we have previously described a beneficial collaboration between neurosurgeons and reconstructive microsurgeons performing EC-IC bypass, reporting 3 successful cases.[16] This study was performed to investigate the indications, technical considerations, complications, and clinical outcomes for both prophylactic and therapeutic EC-IC bypass.

PATIENTS AND METHODS

A retrospective review of 28 consecutive patients who underwent EC-IC bypass at the Chang Gung Memorial Hospital between January 2004 and December 2012 was conducted after Institutional Review Board approval (# IRB-103-4535B) (Table 1). To be clear, EC-IC bypass has been performed by neurosurgeons around the world and at our institution for many years, and the clinical decision to perform this procedure was made according to the judgment of the attending neurosurgeon(s). This series of patients differs only in that the procedure was performed in conjunction with a plastic surgeon with microvascular expertise. This retrospective study therefore served primarily to highlight the outcomes of this collaboration between neurosurgeons and microsurgeons. Secondarily, we explored whether outcomes were different in patients undergoing the procedure because they had had episode(s) of cerebral ischemia/infarction (ie, therapeutic intervention) or because they had been found to be at high risk of ischemia/infarction (ie, prophylactic intervention). As such, patients were retrospectively categorized into prophylactic (group A; Table 1) and therapeutic (group B; Table 2) groups.

Table 1.

Demography, Type of Procedure, and Functional Outcome of 28 Consecutive Patients Who Underwent Extracranial-to-intracranial Arterial Bypass

Table 2.

Preoperative Neurological Symptoms in 12 Patients Who Underwent Therapeutic Extracranial-to-intracranial Arterial Bypass

Patients’ records were reviewed for demography, indications, acute medical and surgical complications (especially postoperative stroke or neurological deterioration), high-flow versus low-flow bypass, patency rate of bypass, pre- and postintervention Glasgow Coma Scores, and the Medical Research Council grading of muscle power (Table 1).

Ligation of aneurysms, application of the adjustable clamp to the aneurysm, and excision of the skull base tumor were performed by the neurosurgeon, whereas the plastic surgeon focused on microvascular anastomoses and the harvest of the bypass graft when used. Postoperatively, all patients were observed in the intensive care unit under the supervision of the neurosurgeons. Graft flow was monitored using a handheld Doppler probe every hour for 24 hours and then every 2 hours for the following 3 days. No systemic anticoagulation was administered routinely, out of concern for intracranial hemorrhage. Graft patency was assessed upon discharge using angiography or magnetic resonance angiography if indicated. The length of the hospital stay was typically 1 week. Comprehensive follow-up evaluation was performed at a mean of 32.7 ± 24.3 months (range, 1–67 months).

Surgical Technique

Preoperative planning was guided by arteriography or magnetic resonance angiography. In the majority of cases, the targeted recipient artery was the insular segment of the MCA, and the donor vessel was usually the distal STA. However, the vascular imaging results and/or the intracranial pathology occasionally dictated the use of other arteries, including the anterior cerebral artery (ACA), posterior cerebral artery, and posterior inferior cerebellar artery as recipient arteries and the external carotid artery (ECA) and occipital artery as donor arteries. The different locations of the MCAs, ranging from M2 to M5, were also documented (Table 3).

Table 3.

Detailed Surgical Procedure Performed on Donor and Recipient Vessels with or without Vein Grafts and Location of Middle Cerebral Artery Used according to Availability

The procedure began with craniotomy/intracranial exposure (and tumor extirpation or aneurysmal clipping). The dura was then incised, and the sylvian fissure was exposed for the identification of the intracranial recipient artery by the neurosurgeons.[16] The target recipient site of the MCA (or alternate recipient artery) was dissected. The STA (or alternate donor artery) was prepared and dissected; a length of approximately 2–8 cm was mobilized. If the need for a vein graft was anticipated, a segment of the greater saphenous, lesser saphenous, or cephalic vein of 7–23 cm in length was harvested, marking the proximal (outflow) end of the graft. The donor artery was transected distally, and vigorous pulsatile bleeding was confirmed before flushing was performed with heparinized saline and a microvascular clamp was applied.

In the high-flow setting, the initial anastomosis was performed distally between the intracranial recipient artery and the outflow end of the vein graft in an end-to-side fashion. Two single vascular microclamps (S&T, Neuhausen, Switzerland) were applied to either end because there was rarely sufficient space to apply a double microclamp. Elliptical arteriotomy was performed to match the diameter of the vein graft, and the lumen was flushed with heparinized solution. End-to-side anastomosis began on the posterior wall followed by the anterior wall of the recipient vessel, using interrupted 11-0 nylon sutures. Blood flow through the recipient artery after anastomosis served to untwist the vein graft and allow for the better assessment of the vein graft length required to reach the donor artery. Standard end-to-end anastomosis was performed using a double clamp and interrupted 9-0 or 10-0 nylon sutures between the vein graft and the donor artery. The adventitia was routinely stripped from the vein graft, allowing for adequate pulsatile flow and facilitating flow from the STA through the graft.

In the case of low-flow bypass, the plastic surgeon performed a direct bypass between the donor artery and recipient intracranial vessel using end-to-side anastomosis as described above. The dura was repaired and tented, the calvaria was fixed with miniplates, and the burr holes were enlarged to accommodate the bypass and to prevent any compression. The scalp was closed over a suction drain without causing excessive tension.

Statistical Analysis

Statistical analysis was performed using the SPSS version 18.0 statistical software (SPSS, Chicago, Ill.). The significance was tested using Pearson’s chi-square test. A P value of < 0.05 was considered statistically significant.

RESULTS

The patients included 12 men and 16 women, with an average age of 48.3 ± 23.7 years (range, 5–93 years). Twelve patients underwent EC-IC bypass for vascular occlusion/stenosis, 8 for MMD, 6 for an aneurysm, and 2 for a tumor (Table 1). There were 21 STA-MCA and 4 ECA-MCA bypasses and 1 STA-ACA, 1 ECA-posterior cerebral artery, and 1 occipital artery-posterior inferior cerebellar artery bypass. Vein graft was used in 12 cases, with a mean graft length of 12.8 ± 5.9 cm (range, 7–23 cm) (Table 1). The ECA was significantly preferred as the donor artery in group A rather than in group B (P = 0.03), and MCA M4 was more frequently used as the recipient vessel in group A (P = 0.05) (Table 3). Postoperative cerebral angiography demonstrated a complete bypass patency rate of 100% (28 of 28). There were no occurrences of postoperative stroke or surgical-related mortality during follow-up.

Sixteen patients underwent prophylactic EC-IC bypass (group A), including 2 patients with aneurysms, 6 patients with MMD, 3 patients with ICA occlusion, 3 patients with MCA occlusion, 1 patient with bilateral ICA stenosis, and 1 patient with ICA occlusion and MCA stenosis. Seven patients in group A used a vein graft as a bypass conduit (Table 3). There were no bypass failures or neurologic deterioration observed in group A (Table 1). The Glasgow Coma Scale and muscle power scores also showed improvement or remained constant (Table 1). At a mean follow-up of 30.1 ± 22.4 months, there was one complication involving a persisting hydrocephalus that occurred in 16 patients (8.3%) (Table 4).

Table 4.

Outcome Analysis of 28 Consecutive Patients Who Underwent Either Prophylactic or Therapeutic Extracranial-to-intracranial Arterial Bypass

Group B included 2 patients with cranial base tumors, 2 patients with ICA aneurysms and subarachnoid hemorrhage (SAH), 2 patients with ICA occlusion and cerebral infarction, 1 patient with a posterior communicating artery aneurysm and SAH, 1 patient with a giant basilar aneurysm and SAH, 1 patient with a giant basilar trunk dissection, 1 patient with ECA occlusion, and 2 patients with MMD (Table 1). Seven of the 12 patients (58.3%) had associated SAH or cerebral infarction. Five of the 12 patients used a vein graft as a high-flow conduit (Table 1). The average graft length in group B was 11.6 ± 3.8 cm. Seven of the 12 patients (58.3%) showed improved clinical symptoms and a stable neurological status after therapeutic EC-IC bypass (Table 1). One patient (case no. 17) with a symptomatic giant basilar trunk dissection aneurysm had neurological deterioration after a failed first EC-IC bypass and underwent a second revisional bypass on post-op day 4, which remained patent (Table 1). The clinical course was further impaired by respiratory failure, pneumonia, and meningitis. However, with appropriate medical care, the patient’s symptoms remained stable.

Twelve of the 28 patients (42.9%) were treated with a high-flow bypass (9 great saphenous veins, two lesser saphenous and one cephalic vein). There were no significant differences in bypass patency, surgical complications, and total complications between the groups with and without vein grafts (Table 5).

Table 5.

Comparisons of Outcomes and Complications for Extracranial-to-intracranial Arterial Bypass with and without the Use of Vein Grafts

In total, 11 complications developed in 28 patients (39.3%) (Table 4). One patient with cranial base meningioma died at 6 months after surgery from tumor progression after being discharged to a nursing facility. One patient had a residual tumor after resection of the cranial base paraganglioma at 14-month follow-up. Outcome analysis between the prophylactic (group A) and therapeutic (group B) groups is summarized in Table 4. The surgical and total complications in group B were 25% and 62.5%, respectively, which were significantly greater than those in group A (P = 0.03 and P < 0.01, respectively) (Table 4). Although the overall complication rate was fairly high, it must be remembered that these were very morbid patients with a life-threatening condition in whom the risk of doing nothing was extremely high. Furthermore, as many complications were medical and not surgical, and hence attributable largely to the patients’ comorbidities rather than to the procedure, the risk-to-benefit ratio, in our opinion, was therefore favorable.

Case Presentations

A 48-year-old woman presented with symptoms of acute cerebral infarction with acute onset of limb weakness and moderate dysarthria and dysphagia. Emergent preoperative computed tomographic angiography demonstrated right temporal infarction with right MCA occlusion (Fig. 1). Right STA-MCA bypass (M3 segment) was performed with an 8-cm great saphenous vein graft (Fig. 2). Perioperative indocyanine green video angiography provided imaging of the patent anastomosis and the restoration of brain perfusion [See Video 1, Supplemental Digital Content 1, which displays how an indocyanine green (25 mg in 5 mL) was intravenously injected after the completion of anastomosis during the end-to-side EC-IC arterial bypass, http://links.lww.com/PRSGO/A95]. Postoperatively, the patient was neurologically stable, and the remaining hospital stay was uneventful. At the 8-month follow-up evaluation, the patient had resumed her normal lifestyle, and angiography showed collateral perfusion to the right MCA via the right thyrocervical trunk, deep cervical artery, right ECA, right STA, and EC-IC bypass and then to the right MCA, intracranial ICA, and ACA (Fig. 3).

Fig. 1.

A, Preoperative CT angiography of a 48-year-old woman with infarction in right middle cerebral arterial territory. Note tiny right anterior cerebral artery A1 segment (arrow) and significant focal stenosis of tiny anterior communicating artery (arrowhead). B, Preoperative CT angiography of a 46-year-old woman with infarction in right middle cerebral arterial territory. Note significant focal stenosis of tiny right posterior communicating artery (arrow). CT indicates computed tomography.

Fig. 2.

A, Vein graft from right lesser saphenous vein (yellow arrow), which was 8 cm in length and 3 mm in diameter, was anastomosed to the MCA M3 (green arrow), which was 1 mm in diameter, with 10-0 nylon in an end-to-side fashion. B, Right superficial temporal artery (green arrow), which was 2.5 mm in diameter, was anastomosed to the vein graft (yellow arrow) in an end-to-end fashion.

Fig. 3.

Status of 48-year-old woman with prior left common carotid arterial stenting and infarction in right middle cerebral arterial territory post EC-IC bypass. Contrast-enhanced MRA showed total occlusion of right vertebral artery V4 segment, right common carotid artery and right cervical internal carotid artery, high grade stenosis of right middle cerebral artery distal M1 segment, and hypoplastic anterior communicating artery and right anterior cerebral artery A1 segment. Collateral perfusion to right middle cerebral artery is shown via right thyrocervical trunk, deep cervical artery, right external carotid artery, right superficial temporal artery, and EC-IC bypass (arrow) and then to right middle cerebral artery, intracranial internal carotid artery, and anterior cerebral artery. MRA indicates magnetic resonance angiography.

See video, Supplemental Digital Content 1, which displays how an indocyanine green (25 mg in 5 mL) was intravenously injected after the completion of anastomosis during the end-to-side EC-IC arterial bypass. The video angiography was recorded by the infrared lens of the Leica Microscope FL720 (Wetzlar, Germany) with a 820-nm filter. Fluorescence began at the superficial temporal artery at 10 seconds post indocyanine green injection and then traveled through the EC-IC arterial bypass to perfuse the ischemic middle cerebral artery territory, http://links.lww.com/PRSGO/A95.

A 54-year-old man with right neck internal carotid dissection experiencing nearly total MCA occlusion for 1 month underwent EC-IC bypass (Fig. 4). A right STA of 2 mm in diameter was prepared with good sprouting and was directly anastomosed to the MCA M4, which was 1.2 mm in diameter (Fig. 5). At the 6-month follow-up evaluation, magnetic resonance angiography revealed collateral perfusion to the right MCA via the right ECA, right STA, and EC-IC bypass and then to the right MCA (Fig. 6).

Fig. 4.

A, A 54-year-old man with infarction in right middle cerebral arterial territory. CT angiography showed right cervical internal carotid artery dissection with nearly total occlusion (arrow). B, CT angiography showed total occlusion of right internal carotid artery. Vascular angulation with significant stenosis at right MCA origin is shown (arrow). CT indicates computed tomography.

Fig. 5.

A, The MCA M4 (green arrow), which was 1.2 mm in diameter, was used as the recipient vessel, and the superficial temporal artery with good sprouting (yellow arrow), which was 2 mm in diameter, was prepared as the donor vessel. B, The area immediately after microsurgical anastomosis was performed in an end-to-side fashion using double microclamp. C, Good perfusion of MCA M4 after anastomosis and the release of microclamp.

Fig. 6.

Status of 54-year-old man with infarction in right middle cerebral arterial territory post EC-IC bypass. Contrast-enhanced MR angiography showed total occlusion of right internal carotid artery and hypoplastic right anterior cerebral artery A1 segment. Collateral perfusion to right middle cerebral artery is shown via right external carotid artery, right superficial temporal artery, and EC-IC bypass (arrow) and then to right middle cerebral artery. MR indicates magnetic resonance.

DISCUSSION

One currently accepted indication for EC-IC bypass is flow augmentation in the setting of chronic occlusive/stenotic cerebrovascular disease.[17] Patients with untreated athero-occlusive disease have a 10–12% annual risk of stroke.[18] Schaller,[12] in a systematic review, reported that the risk of stroke or death in selected patients with symptomatic inaccessible stenotic lesions of the ICA or MCA treated with EC-IC bypass is significantly less than that of medically treated patients (1.3% vs 3.6% per year, respectively). Neurological function was improved in 84% of the patients in the review by Schaller.[12] In our study, 12 patients were treated for ICA and/or MCA occlusion/stenosis. Eight of these patients in group A were asymptomatic and were treated prophylactically; all fared well, with no stroke or neurologic deterioration and few complications observed. Of the 4 symptomatic patients (nos. 19, 24, 25, and 26) in group B, 3 improved neurologically, 1 remained stable, and none suffered a post-bypass stroke.

MMD is an additional indication for EC-IC bypass. In these patients, the frequency of transient ischemic attacks can be reduced, and strokes can be avoided, by creating an alternative circulatory pathway in the brain.[19] The Japan Adult Moyamoya Trial is currently underway, prospectively examining the prognoses of patients treated with EC-IC bypass. Pending the publication of these results, most cerebrovascular surgeons recommend bypass surgery.[20] In our study, there were 8 patients with MMD; all fared well, with no significant complications and no strokes observed.

An additional indication for EC-IC bypass is flow replacement in the context of therapy for complex aneurysms and skull base tumors.[16,17] Clip application or coil placement for an intracranial aneurysm is not always feasible. In such cases, the parent artery can be obliterated to induce thrombosis of the aneurysm, but if this involves one of the major cerebral arteries, EC-IC bypass is indicated for ipsilateral revascularization. A recent meta-analysis of 20 studies has demonstrated that neurologic function and subsequent stroke risk in patients with otherwise untreatable intracranial aneurysm are significantly improved by EC-IC bypass.[21] Our study included 6 patients with an aneurysm, including 4 with an associated SAH in group B. These 4 patients, along with 1 of the tumor patients, comprised the 5 total patients (nos. 17, 18, 20–22) in the study who demonstrated a postoperative diminution of muscle strength. These unfavorable outcomes likely reflect the grave nature of their neurologic condition rather than an untoward effect of the bypass procedure.

Prophylactic EC-IC bypass demonstrated a high patency rate and minimal complications in this study. However, therapeutic EC-IC bypass was associated with significantly greater surgical and total complications. It might be considered that there is a trade-off between the 100% patency rate and 25% surgical complication rate associated with therapeutic EC-IC bypass in this study and the gravity of the symptomatic cerebral artery disease present. The equally high successful patency rate of EC-IC bypass with or without vein graft indicated that a vein graft can be used safely depending on the availability of donor and recipient vessels.

The choice of revascularization conduit is dependent on the required amount of brain perfusion, the availability of the donor vessels, the availability of graft flow, and the size of the recipient vessel.[6] The ICA can be used, provided that there is sufficient collateral flow to allow for temporary occlusion. The ECA or one of its branches is used if the ICA is not available or if collateral flow is insufficient (Table 3).[6] Because of its good accessibility and close proximity, the STA is an ideal graft in patients with cerebro-occlusive disease, aneurysm, tumor, and MMD, ensuring that the low-flow volume is adequate for the recipient territory.[22] The size of the STA should be at least 1 mm in caliber, and the recipient vessel can be the M2 or M3 segment of the MCA or the superior cerebellar artery.[6] Patency rates of the STA were as high as 95%,[22,23] and in our study, they reached 100%. However, in cases of large parent vessel sacrifice, the STA is inadequate. For high-flow conduits, the reverse great saphenous vein graft is preferred because of its robust flow (70–140 mL/min). We prefer end-to-side anastomosis because it has been demonstrated that for venous repairs with a size discrepancy, end-to-side repairs are significantly better.[24] The lesser saphenous vein, which is similar to the greater saphenous vein but is smaller in diameter, may be another good option for decreasing the size mismatch at the end-to-side anastomosis of the MCA insular segment. In this study, the cephalic vein was used in one case because it matched the recipient vessel in size.

In a study of 137 revascularizations, Sekhar and Kalavakonda[25] reported an overall graft patency rate of 95.6% and an incidence of postoperative cerebral infarction of 16.8%. In addition to graft thrombosis, cerebral infarction can be caused by a prolonged ischemia time of the recipient segment of the MCA. Chazono et al[26] reported a case of intraoperative infarction with a clamping time of 65 minutes and no infarction in 3 identical cases with clamping times of 30–40 minutes. In partnership with neurosurgical colleagues, the plastic surgeon with well-honed microvascular skills is uniquely poised to perform these anastomoses accurately and efficiently to maximize the outcomes in these patients.

CONCLUSIONS

EC-IC bypass can be an effective procedure in the treatment of complex neurovascular conditions, including occlusive, stenotic, aneurysmal, and oncologic diseases. This study demonstrates that the collaboration of neurosurgeons and plastic surgeons in performing EC-IC bypass can result in excellent outcomes with few complications. Therapeutic EC-IC bypass was associated with greater total and surgical complications compared with prophylactic bypass.

ACKNOWLEDGMENT

We thank Chia-Yu Lin, MSc, for her assistance in the preparation of this article, video, and figures.