Michael J. Link, M.D. and Grace A. Herzog,
P.A.-C.
Michael J. Link, M.D. is an Instructor
of Neurosurgery, Mayo Medical School.
Grace A. Herzog, P.A.-C. is with the Department of Neurologic
Surgery at Mayo Clinic Jacksonville.
The purpose of extracranial to intracranial (EC-IC) bypass is to augment cerebral blood flow. This procedure entails connection of the superficial temporal artery (STA), or a venous conduit, to a branch of the middle cerebral artery (MCA). EC-IC bypass has been available as a potential treatment for stroke for the past 30 years. During this time, the procedure and its indications have undergone intense scrutiny and reevaluation as a legitimate therapy, engendering controversy among neurosurgeons and neurologists. The procedure has been touted as a great technical achievement, and at other times a therapeutic failure.
The first EC-IC bypass was performed by Yasargil, in Zurich, Switzerland, in 1967.1 Following in his footsteps, other talented cerebrovascular surgeons in the United States adopted the procedure during the 1970's, and its use expanded to most major neurosurgical centers around the world. In 1977 an international, randomized, prospective study was instituted to assess the safety and efficacy of the procedure.2 The negative results of that study, published in 1985,3 had a dramatic effect on the number of bypass procedures performed, and to some degree redirected the thinking of all physicians involved in the treatment of stroke.
The surgical options available for the treatment of stroke are limited. Carotid endarterectomy has been proven in large, randomized, prospective trials to be an effective treatment in selected patients with asymptomatic4 or symptomatic5,6 extracranial carotid artery stenosis of at least 60% or 70%, respectively. Endarterectomy of the intracranial arteries has not met with much success, mainly due to difficulty in establishing a satisfactory distal intimal break point.7,8 Intracranial arterial transpositions, most commonly anastomosing the superior cerebellar and posterior cerebral arteries in a side-to-side fashion, is rarely indicated, rarely performed and technically quite challenging.8 MCA embolectomy held brief popularity but has been essentially replaced by endovascular techniques.9
EC-IC bypass however, has continued to evolve. As our understanding of the etiology of stroke has increased, so has our ability to select patients for EC-IC bypass. Additionally, advances in surgical technique and neuroanesthesia have increased the bypass options and decreased the morbidity of the procedure. There remains a subset of patients with ischemic cerebrovascular disease who are not candidates for conventional carotid endarterectomy (CEA) but who may benefit from cerebral revascularization with EC-IC bypass. This article will review the results of the EC-IC bypass trial, the ensuing controversy, and the current indications for EC-IC bypass.
The EC-IC bypass study was initiated to determine whether anastomosis of the superficial temporal artery to the MCA could reduce, stroke and stroke-related death among patients with symptomatic, surgically inaccessible (to CEA) atherosclerotic stenosis or occlusion of the internal carotid or MCA. Patients were eligible for the trial if they had had either a transient ischemic attack (TIA) or a minor stroke in the appropriate arterial distribution within three months of entry. Angiographically, they had to have stenosis or occlusion of the trunk of the MCA, or internal carotid artery (ICA) stenosis above the C-2 vertebral body, or ICA occlusion. A number of exclusion criteria were established to insure medical fitness of the patients randomized to surgery.2 A total of 1,495 patients were entered between 1977 and 1982, with 118 subsequently excluded. Of the 1,377 eligible patients, 663 underwent STA-MCA bypass in addition to best medical therapy and the remaining 714 were assigned to best medical therapy alone. The two groups were similar with regards to entry characteristics. Patients were examined every 6 months by a participating neurologist, and repeat angiograms were performed in 92% of surgical patients at a median time of 32 days post-operatively. The angiograms were reviewed to determine the size and patency of the anastomosis, the degree of retrograde flow in the MCA, and the number of branches of MCA filled. No patient was lost to follow-up, and the average duration of follow-up among surviving patients was 55.8 months.
After careful analysis of the primary events of fatal and nonfatal strokes, it was clearly demonstrated that STA-MCA bypass did not reduce the rate of subsequent events compared to best medical therapy in the patients studied. Fatal and nonfatal strokes occurred both more frequently and earlier in patients randomly assigned to surgery (Figure 1).3 Secondary analyses comparing the rates of major and fatal strokes, all strokes and all deaths, ipsilateral ischemic strokes, and major ipsilateral ischemic strokes between the two groups also showed no statistical benefit with surgery.3
![]() |
Figure 1. Results of the primary analysis (all strokes, both fatal and notfatal), showing the failure of bypass between the superficial temporal artery and the middle cerebral artery to reduce stroke in the surgical (663 patients) as compared with the medical cohort (714 patients) after an average follow-up of 55.8 months. The analysis uses Kaplan-Meier cumulative-failure curves. Reproduced with permission from The EC/IC bypass study group, N Engl J Med. 1985; 313:1191-1200. |
Two subgroups of patients fared worse after surgery: 1) Patients who continued to have ischemic events after ICA occlusion but before randomization and 2) patients with severe MCA stenosis. Additionally, 200 patients in the study who had the worst post-operative angiograms were compared with 225 patients with the best angiograms. Those with good perfusion fared no better than those with a small or occluded anastomosis. Technical results of surgery were excellent. On final review, 96% of patients who had post-operative angiograms had patent STA-MCA anastomoses.3 Perioperative major stroke occurred in 30 patients (4.5%) and was fatal in 7 (1.1%). Ten of the 30 major strokes occurred after randomization to surgery but before the operation was performed and were included in an intention-to-treat analysis.3 The negative results of the study led many to conclude, "microvascular anastomosis is a marvellous technique, but it has failed to emerge as a miraculous treatment."10
The EC-IC bypass study was well planned, methodologically sound, and well executed.11 Still, it generated great controversy and was widely criticized.12-16 In fact, an ad hoc committee of the American Association of Neurological Surgeons was convened to review the validity of the study.15 Of all the various criticisms raised the main concerns included: Was the trial diluted with too many low-risk patients? Can the STA carry enough flow, or should a larger bypass vessel (saphenous vein) have been used? Was therapeutic benefit in some patients masked by perioperative morbidity in other patients? Was there "observational" bias and "randomization to treatment" bias?17,18 Does secondary analysis of multiple subgroups have any merit when there is no significant difference in the sample at large? How can a single therapeutic modality be expected to be effective against widely different pathophysiologic mechanisms? Finally, were too many patients operated outside the trial (eligible, but not randomized) to make the surgical cohort representative of the population?19 Suffice it to say, the issues were debated extensively in and out of the literature, and the majority of the criticisms were answered.10,20 However, the latter two concerns, and primarily the last one, have proven to be the "Achilles heel" of the study.
The initial report described only 115 patients refusing entry and only 52 patients operated outside the trial.3 A telephone survey, however, revealed that 2,572 patients received STA-MCA bypass outside of the trial while the trial was being conducted.16 In the U.S. and Japan, 1,014 of 1,831 nonrand-omized patients receiving surgery were believed to have been eligible for the trial.16 While this does not affect the internal validity of the study, (i.e. the conclusions reached for the patients studied were valid), it raised many concerns regarding whether the results could be generalized to the population at large. Despite these criticisms, the results of the EC-IC bypass study had a profound effect on the management of cerebrovascular occlusive disease. The number of procedures performed markedly decreased, and government and third-party payers began to refuse coverage for any revascularization surgery, regardless of the indication. The procedure is rarely performed now for ischemia, even at major cerebrovascular surgical centers. Many neurosurgeons in training may only see one or two procedures during their entire residency.
Perhaps the most important issue surrounding consideration of EC-IC bypass in patients with stroke is the mechanism of the stroke. The importance of being able to distinguish between embolic and hemodynamic causes of stroke, and the difficulty in doing so, cannot be overemphasized.21-24 The EC-IC bypass trial may have failed to prove a benefit of surgery because many of the patients in the trial probably suffered embolic strokes. Also, the patients with ICA occlusion and persistent symptoms who fared worse after being randomized to surgery were likely having recurrent emboli from the occluded ICA stump. It certainly has not been proven, and would not be expected, that an arterial anastomosis, that is providing retrograde flow through the MCA could prevent embolic complications from a more proximal (ICA or MCA stem) occlusive source.25
The hemodynamic properties of performing an STA-MCA anastomosis are also important in considering why the study produced a negative result. The MCA stenosis cohort fared particularly poorly. It has since been demonstrated that there is a significant risk of thrombosis at the stenotic lesion when retrograde flow from the bypass graft is made to compete from the usual antegrade collaterals.26 At the same time, there are several documented cases of the stenosis resolving after bypass.26 Predicting how any particular stenosis will react to revascularization is not yet possible. Creation of a bypass to reproduce more physiologic flow, such as anastomosing to the supraclinoid ICA instead of the distal MCA may reduce this problem. Techniques are available to make this anastomosis with little or no occlusion of the recipient ICA, but these techniques have not met with wide-spread acceptance.27
Clearly, the biggest hurdles to successfully revascularizing the ischemic brain are forgoing patients likely suffering embolic stroke from ICA occlusion or MCA stenosis. These patients may be better served by anticoagulation, and having a better understanding of the hemodynamic effects of creating an anastomosis in the distal cerebrovascular circulation in relation to the lesion being treated.
Advances in surgical technique have greatly expanded the options available for performing an EC-IC bypass. The EC-IC bypass study involved only the relatively low flow STA-MCA bypass (Figure 2). Currently, there are many bypass options using a reversed saphenous vein to bypass various segments of the anterior or posterior circulation.28 Most commonly, the saphenous vein is anastomosed in an end-to-end fashion to the external carotid artery (ECA), passed through a preauricular subcutaneous tunnel and then anastomosed in an end-to-side fashion to a branch of the MCA.29-32 If the ECA is providing important collateral circulation to the eye or brain, then the vein can be sewn end-to-side to the common carotid artery so this physiologic collateral circulation is not disturbed33 (Figure 3). Typically, the proximal anastomosis is performed with 6-0 or 7-0 monofilament suture. The distal anastomosis to the MCA is done with 9-0 or 10-0 suture. The vein conduit has the advantage of being a much higher flow bypass compared to the STA, but requires two anastomoses, and harvesting of the vein. Additionally, the petrous or supraclinoid segments of the ICA can be used as a donor or recipient site, depending on the anatomical location of the stenosis.34-37 While all these techniques take considerable technical skill, the greatest challenge comes not from constructing a bypass, but in deciding which patients will benefit.
![]() |
![]() |
| Figure 2. Artist's illustration of STA-MCA bypass operation. The superficial temporal artery is anastomosed to the middle cerebral artery using 9-0 or 10-0 monofilament suture. Reproduced with permission from Sundt's Occlusive Cerebrovascular disease, second edition, W.B. Saunders, Philadelphia, 1994. | Figure 3. Intraoperative left common carotid angiogram shows the reversed saphenous vein has been anastomosed in an end-to- side fashion (arrow marks anastomosis). This patient had a left internal carotid artery occlusion and the external branches were providing significant collateral circulation to the left eye. |
V.P. is a 49 year-old woman with a history of hypertension and tobacco abuse who presented with 8-weeks of right hand "incoordination" and word-finding difficulty. Worsening expressive aphasia was the most prominent symptom 1 week prior to presentation. Magnetic resonance imaging (MRI) of the brain revealed stroke in the left basal ganglion and a parietal "watershed" stroke (Figure 4). She was admitted to the hospital and started on heparin anticoagulation. Work-up for hypercoagulable state and cardiac disease was negative. Cerebral angiography revealed high-grade left MCA stenosis (Figure 5). Xenon - CT scanning confirmed hypoperfusion (< 30 cc/100 gms brain / minute) in the left MCA distribution. She continued to have a fluctuating neurologic exam on anticoagulation and with induced hypertension. After extensive discussion, it was elected to proceed with a left EC-IC bypass using a reversed saphenous vein from the ECA to the MCA (Figure 6). Post-operatively she has done well with no further ischemic symptoms during 3-months of follow-up. This case demonstrates a patient with persistent symptoms of cerebral hypoperfusion, despite maximal medical therapy, who benefited from EC-IC bypass.
![]() |
![]() |
| Figure 4a
and 4b. Figure 4a (left): MRI of the brain reveals a
stroke in the left basal ganglion (arrow). Figure 4b
(right): A stroke is seen in the left parietal -
occipital watershed zone (arrow). |
|
![]() |
![]() |
![]() |
| Figure 5a (left): AP right carotid angiogram reveals normal filling of the intracranial anterior cerebral and middle cerebral vessels. There is a non-hemodynamic stenosis of the origin of the anterior cerebral artery (arrow). | Figure 5b (center): AP left carotid angiogram reveals severe stenosis of the proximal left middle cerebral artery (arrow) and slow filling of the distal middle cerebral branches. | Figure 6. (Right): Intraoperative angiogram shows a patent anastomosis between the reversed saphenous vein and the middle cerebral artery (arrow). There is excellent filling of the distal middle cerebral branches (arrowheads). |
Despite the negative results of the EC-IC bypass study3, most cerebrovascular neurologists and neurosurgeons believe there is a subgroup of patients that will benefit from revascularization surgery in the form of a bypass. There are several studies available which suggest hemodynamic failure can be predicted in patients with occlusive cerebrovascular disease, or those prior to therapeutic carotid occlusion, using SPECT scanning38, Xenon _ CT scanning39 or positron emission tomography (PET).40 More importantly, it has been shown that EC-IC bypass has the potential to reverse the hemodynamic failure and normalize cerebral blood flow (CBF).41-46 Of course, the demonstration of hemodynamic failure at baseline does not necessarily prove that all subsequent strokes are hemodynamically mediated. Low-flow states may predispose to the formation of thromboemboli or, alternatively, emboli may cause infarction more readily in areas with poor collateral circulation.47 The best evidence to date indicates that patients with impaired angiographic collaterals, watershed stroke on pre-operative imaging, radionuclide imaging evidence for impaired CBF, and those who have failed optimal medical therapy are the best candidates for EC-IC bypass. Each case must be individually and carefully reviewed, preferably by a multidisciplinary team which includes neurologists, neurosurgeons and neuroradiologists.
As the ability to predict hemodynamic failure has improved, there has been renewed interest in considering another EC-IC bypass trial. Grubb et al., have recently demonstrated that patients with an occluded ICA and ipsilateral increased oxygen extraction fraction (the fraction of oxygen in the blood that the brain extracts to maintain metabolism) by PET scanning have a significantly increased risk of future stroke.47 This patient population for instance, may be better served by EC-IC bypass, rather than the broad mix of patients in the original trial.2 Unfortunately, PET is not widely available. Also to be considered is the inherent difficulty in recruiting enough patients fitting such narrow criteria to prove statistical significance. Thus, this proposition is being considered cautiously.48
We feel EC-IC bypass has a definite role in the treatment of ischemic stroke in carefully selected patients fulfilling very strict criteria as detailed above. With improved ability to screen patients (eliminate patients with embolic etiology of stroke), improved measures of regional CBF to direct the anastomosis, improved methods to predict the hemodynamic effects of the bypass, and improved surgical techniques to provide physiologic long-term blood flow, the role of EC-IC bypass is expected to become clearer in the near future.
REFERENCES
What's New
·
Northeast Florida Medicine Journal ·
Know Your Physician
· Legal
& Legislative
·
DCMS Alliance ·
Academy of Medicine ·
Member Websites ·
Community Health
About the DCMS ·
Meetings Calendar ·
Member Benefits
·
Employment Connection ·
Home
Duval County Medical Society
·
555 Bishopgate Lane
·
Jacksonville, FL 32204
Phone: (904) 355-6561
·
FAX: (904) 353-5848
General Email: dcms@dcmsonline.org
·
Webmaster's Email: mdoran@dcmsonline.org
Privacy Policy
and Disclaimers