New Ischemic Brain Lesions on MRI after Stenting or Endarterectomy for Symptomatic Carotid Stenosis: A Sub-Study of the International Carotid Stenting Study (ICSS)

Abstract

Conclusion: In a substudy of patients randomized in the International Carotid Stenting Study (ICSS) comparing carotid artery stenting with carotid endarterectomy for symptomatic carotid stenosis, patients randomized to the stenting group had three times more ischemic lesions found by post-treatment magnetic resonance imaging (MRI) diffusion-weighted imaging (DWI) than patients randomized to the endarterectomy group. Cerebral protection devices did not seem to be effective in preventing ischemic DWI lesions after stenting. Summary: The ICSS randomized patients with symptomatic carotid artery stenosis to carotid stenting or carotid endarterectomy. Of the 50 centers that participated in the ICSS study, 7 took part in an MRI substudy, the results of which were reported in this article. In the MRI substudy, MRI was done 1 to 7 days before treatment (endarterectomy or stenting), at 1 to 3 days after treatment (post-treatment scan), and again at 27 to 33 days. The primary end point of this substudy was the presence of at least one new ischemic brain lesion on DWI on the post-treatment scan. Analysis was per protocol, and investigators who read the scans were blinded to whether the patient had received endarterectomy or carotid stenting. The substudy comprised 231 patients (107 in the endarterectomy group and 124 in the stenting group). The detection of one new DWI lesion on post-treatment scans done a median of 1 day after treatment occurred more frequently in the stenting group (50% of the 124 patients) than in the endarterectomy group (17% of the 107 patients; odds ratio [OR], 5.21; 95% confidence interval [CI], 2.78-9.79; P < .0001). There were changes at 1 month on fluid-attenuated inversion recovery sequences in 28 of 86 patients (33%) in the stenting group and in 6 of 75 patients (8%) in the endarterectomy group, (adjusted OR, 5.93; 95% CI, 2.25-15.62; P = .0003). In centers with a policy of using cerebral protection devices, new DWI lesions were present after treatment in 37 of 51 patients (73%) in the stenting group and in 8 of 46 (17%) in the endarterectomy group (adjusted OR, 12.20; 95% CI, 4.53-32.84). In centers with a policy of using unprotected carotid artery stenting, new DWI lesions were present in 25 of 73 patients (34%) in the stenting group and in 10 of 61 patients (16%) in the endarterectomy group (adjusted OR, 2.70; 95% CI, 1.16-6.24; interaction P = .019). Subanalyses of age <71 years, sex, whether the qualifying event was a TIA or stroke, and whether there were DWI lesions present on the pretreatment scan all indicated higher rates of new DWI lesions after stenting compared with endarterectomy. Comment: Previous nonrandomized studies have also suggested higher rates of ischemic lesions detected on DWI imaging after stenting compared with endarterectomy. A meta-analysis of these studies is presented in Fig 4 of this article and indicated the OR of new ischemic lesion after treatment was 6.71 (95% CI, 4.57-9.87) favoring endarterectomy. The authors' OR of 5.21 is very similar to that obtained from the meta-analysis. The clinical significance of DWI lesions in the long run is unknown but is postulated to eventually lead to cognitive decline and dementia (Pendlbury ST et al [Lancet Neurol 2009;8:1006-18]). Of particular interest is the apparent lack of protection against DWI lesions afforded by cerebral protection devices. One might postulate, given the relative ORs for DWI lesions with and without protection devices, that protection devices actually increase the risk of DWI lesions! There are a number of reasons why cerebral protection devices may have been ineffective in this study. For instance, all protection devices in this study were of the filter type, and conclusions about other types of devices cannot be made. Another possible conclusion is that cerebral protection during carotid artery stenting is another idea that seems like a good idea but does not hold up to proper scrutiny when one actually examines data collected by unbiased observers.