The CMOSS is the first RCT sponsored by a developing country to evaluate the efficacy and safety of EC-IC bypass surgery. It is also the first RCT to evaluate the cerebrovascular haemodynamics by CTP in patients with ICA or MCA occlusion. It is well-known that intracranial arterial stenosis or occlusion in Asia is much more common than in Europe or North America. This could greatly encourage patient enrolment. The CMOSS also selects more experienced neurosurgeons to work in the trial to minimise perioperative complications.
For the sample size calculation, we consulted the figures for stroke risk from MCA occlusion, which is different from ICA occlusion. In 1996, Yamauchi et al. [
17] investigated patients with symptomatic ICA or MCA occlusive diseases. They found that the incidence of ipsilateral ischaemic strokes in patients with increased OEF and normal OEF was 57.1% and 6.0%, respectively. (They also suggested that patients with increased OEF had decreased CBF and CBF/CBV values.) Kuroda et al. [
19] reported that the annual risks of total and ipsilateral stroke in patients with ICA or MCA occlusion and decreased rCBF and relative cerebral vasoreactivity (rCVR) were 35.6% and 23.7%, respectively. Actually, the total annual risk of ipsilateral stroke in patients with decreased rCBF only was 26.1%. In 2002, Ogasawara et al. [
18] presented a study to prospectively evaluate the relationships between cerebral haemodynamics and outcome of patients with symptomatic major cerebral artery occlusion (including ICA and MCA) by quantitative measurement of CBF using
133Xe and single photon emission tomography (SPECT). During the follow-up, the recurrence rate of stroke was 34.7% in patients with reduced rCVR at entry and less than 6% in patients with normal rCVR. All these data indicate that the annual risk of stroke is relatively higher in MCA than in ICA occlusion. Recently, aggressive medical management has been more widely used, and according to the results of a follow-up by our centre over the last 3 years, the risk of stroke in ICA or MCA occlusion is about 28%, which conforms to the data mentioned. Compared with the medical treatment group (28%), a further 50% reduction in the surgical treatment group is considered to be clinically significant; that is 14%. Regarding perioperative morbidity, in the COSS [
20] the perioperative rate for ipsilateral ischemic stroke was 14.4%, while the rate for the primary end point was 21.0%. This indicates that the stroke rate between 30 days to 2 years was about 7%. In the CMOSS, we have much stricter licensing of neurosurgeons; attending neurosurgeons are certified by their roles as chief surgeon in at least 15 consecutive previous EC-IC bypass surgeries. Using this level of expertise means that in China we can achieve an anastomosis graft patency rate of greater than 95% and a perioperative stroke and death rate of less than 10%.
In addition, because of the continuity of the ‘Five-twelfth’ National Science and Technology Support Program, the follow-up for these patients could prolonged to more than 2 years. This could reveal more information after analysis of the ‘cross-over curves’ in the COSS.