Editorial 

August 22/29, 2023

Extracranial-Intracranial Bypass Surgery for Stroke Prevention

Seemant Chaturvedi, J. Marc Simard

JAMA. 2023;330(8):697-698. doi:10.1001/jama.2023.11166

The concept of performing an extracranial-intracranial (EC-IC) arterial bypass operation for a patient with an internal carotid artery (ICA) or middle cerebral artery (MCA) occlusion is intuitively appealing. In these patients, it is believed that areas of hypoperfusion exist and anastomosis of the superficial temporal artery to the MCA is designed to augment brain perfusion. In vascular territories outside of the brain, bypass procedures performed to augment arterial flow typically improve clinical outcomes.¹ For reasons unknown, however, the brain appears to respond differently to bypass surgery.

In a landmark trial, the EC/IC Bypass Study Group enrolled 1377 patients, with randomization to surgery or best medical therapy.² Patients were enrolled with ICA or MCA stenosis or occlusion, along with a symptomatic event in the carotid territory within 3 months of study entry. The 30-day outcomes in the surgery group were commendable, with a major stroke rate of 2.5% and mortality of 0.6%. Over a 56-month follow-up period, however, there was no reduction in the stroke rate between the 2 groups, with a single stroke occurring in 18% of the medically treated patients and 20% of the surgery group. Because this was one of the earliest randomized trials in neurosurgery, the unexpected outcome caused considerable controversy, especially among neurosurgeons.

In the aftermath of this trial, several criticisms were raised, including the fact that the hemodynamic impact of the arterial stenosis or occlusion had not been considered. In the ensuing decade, longitudinal studies of patients with ICA occlusion were conducted. In studies that evaluated oxygen extraction on positron emission tomography scans, it was identified that there was a high stroke rate in patients with stage 2 cerebral hemodynamic failure, also known as misery perfusion. In these patients, there is an elevated oxygen extraction rate ipsilateral to the occlusion compared with the contralateral hemisphere. The increased oxygen extraction rate signifies a compensatory response to relative oligemia and represents an attempt to maintain oxygen supply to a hemisphere with decreased perfusion. In a study of 81 patients followed up for an average of 31.5 months, the stroke rate was significantly higher in patients with stage 2 failure compared with patients without hemodynamic failure (31% vs 7%).³ This group seemed to be an obvious group that might benefit from EC-IC bypass surgery.

In the subsequent Carotid Occlusion Surgery Study, 195 patients with documented ICA occlusion and stage 2 hemodynamic failure on positron emission tomography were assigned to bypass surgery or medical treatment.⁴ The primary end point was a composite of stroke or death within 30 days of surgery (or randomization in the medical group) plus ipsilateral stroke up to 2 years. There was no demonstrable benefit of surgery observed, with the primary end point occurring in 21.0% of the surgery group and 22.7% of the medical group. Notably, there was a stroke or death rate of 14.4% within 30 days of surgery in the surgical group compared with 2.0% in the nonsurgical group. Even in patients with multiple transient ischemic attacks or strokes in the territory of the ICA occlusion, there was still no benefit of surgery.⁵

In this issue of JAMA, Ma et al⁶ report the results of the Carotid and Middle Cerebral Artery Occlusion Surgery Study (CMOSS) randomized clinical trial comparing EC-IC bypass surgery vs medical therapy alone in 324 patients (median age, 52.7 years; 79.3% male). The study was motivated by an attempt to reduce the surgical complication rate and refine patient selection, hoping that these improvements could identify a group with demonstrable benefit from surgery. Patients were required to be symptomatic during the previous 12 months, and an ICA or MCA occlusion was required for study entry. Hemodynamic factors were assessed with computed tomography perfusion, and a 4-second delay on computed tomography perfusion and a relative cerebral blood flow ratio of less than 0.95 (comparing the ipsilateral vs contralateral hemisphere) were required. To optimize surgical results, strict credentialing of surgeons was undertaken, and patients were required to be at least 3 weeks beyond the qualifying stroke. The primary end point was a composite of stroke or death within 30 days of surgery (or randomization) and ipsilateral stroke from 30 days to 2 years.

The primary end point occurred in 8.6% of the surgical group and 12.3% of the medically treated patients (incidence difference, −3.6% [95% CI, −10.1% to 2.9%]; hazard ratio, 0.71 [95% CI, 0.33-1.54]). The difference was not clinically or statistically significant. Post hoc analysis of subgroups showed that, within 30 days, more strokes occurred in the surgery group (6.2% vs 1.8%), whereas in the time period of 30 days to 2 years, more strokes occurred in the medical group (10.3% vs 2.0%). Three fatal strokes all occurred in the surgery group. Within various demographic, medical, and imaging subgroups, there was not a subgroup for whom surgery was clearly beneficial; those with a relative cerebral blood flow ratio of less than 0.8 had a trend toward possible benefit, although the confidence interval around the effect estimate crossed the null and the test for interaction was not statistically significant. At 2 years, the surgical anastomosis patency rate was 93.6%.

Are we witnessing the same old story? Yes and no. The top-line result is now similar in 3 randomized trials—there is no clear benefit of EC-IC bypass surgery. Despite careful attention to hemodynamic factors and high rates of graft patency, EC-IC bypass surgery has yet to be shown to confer long-term protection from stroke.

The changing aspect of the story is the improved results with medical therapy, consisting of antithrombotic therapy and treatment of risk factors such as dyslipidemia, hypertension, and diabetes. Clinicians need to keep in mind the systemic nature of ICA and MCA atherosclerosis. In a previous population-based study of patients with ICA occlusion, the 5-year results for stroke, myocardial infarction, and death were 14%, 24%, and 29%, respectively.⁷ Even if there is a subgroup of patients who may benefit from EC-IC bypass surgery, the systemic nature of atherosclerosis signifies that the long-term threat of cardiac events and vascular death remains. Intensive medical therapy could mitigate the long-term risk of major vascular events, including cardiac events.

Another important observation is that the Ma et al⁶ study used medium-intensity medical therapy. The target for low-density lipoprotein cholesterol was less than 100 mg/dL. At present, one could argue for a low-density lipoprotein cholesterol target of less than 55 mg/dL or even less than 40 mg/dL based on the high risk for atherosclerotic events in these patients.^8-10 Targeted blood pressure control, dietary modification, and regular physical activity would also be expected to have salutary effects. A multimodality approach could result in an even lower event rate in patients receiving intensive medical therapy alone,¹¹ raising the bar for EC-IC bypass surgery.

In short, the results with EC-IC bypass surgery in randomized trials remain unimpressive. Until a better understanding of the unique hemodynamic features of the brain is achieved, it will be difficult for neurosurgeons to continue offering this procedure to patients with ICA or MCA occlusion. Intensive, multifaceted medical therapy remains the first-line treatment for patients with ICA or MCA occlusion.¹²