Reduced Cerebrovascular Reserve at CO2BOLD MR Imaging Is Associated with Increased Risk of Periinterventional Ischemic Lesions during Carotid Endarterectomy or Stent Placement: Preliminary Results1

PURPOSE: To determine whether any initial reductions in cardiovascular reserve (CVR) normalize after carotid revascularization and-because reduced CVR represents a risk factor for ischemic events-whether patients who develop periinterventional infarction have more severely reduced pretreatment CVR than those who do not. MATERIALS AND METHODS: Ethics committee approval and informed consent were obtained. Twenty-four consecutive patients with symptomatic high-grade internal carotid artery stenosis (seven women; mean age, 73.1 years +/- 9.4 standard deviation) were recruited from a prospective, randomized trial that compared carotid artery stent placement with endarterectomy. Magnetic resonance (MR) imaging, including CO(2) blood oxygen level-dependent (BOLD) MR, was performed 1-3 days before, 1-3 days after, and 1 month after carotid revascularization (carotid artery stent placement, n = 13; carotid endarterectomy, n = 11). RESULTS: Mean CVR in the ipsilateral middle cerebral artery (MCA) territory was reduced prior to treatment (mean DeltaT2* in ipsilateral territory, 1.92% +/- 1.18; mean DeltaT2* in contralateral territory, 2.28% +/- 1.15 P .05; mean DeltaT2* 1 month after treatment in ipsilateral territory, 2.27% +/- 1.05; that in contralateral territory, 2.14% +/- 0.96 P > .05). Those patients who developed new periinterventional infarcts (n = 7 with punctate foci of restricted diffusion) had greater reduction of CVR in the ipsilateral MCA territory prior to treatment (relative reduction, 32.5% +/- 46.0; P < .05) than those who did not develop infarction (n = 17; relative reduction, 9.2% +/- 55.9). CONCLUSION: CO(2) BOLD MR imaging could be used successfully to monitor the hemodynamic effects of carotid revascularization; initial reductions in CVR normalized after carotid revascularization. Severely reduced pretreatment CVR was associated with increased occurrence of new periinterventional therapy infarction.