Background and purpose: Overviews of randomized patch trials by the Cochrane Collaboration suggest that a policy of routine patching is preferable to routine primary closure. However, there is no systematic evidence that patch type, whether prosthetic or vein, influences outcome after carotid endarterectomy (CEA).
Methods: Two hundred seventy-three patients were randomized to vein or thin-walled Dacron patch (Hemashield Finesse) closure of the arteriotomy after 276 CEA procedures. Patients were reviewed clinically and with duplex ultrasound scanning at 1, 6, 12, 24, and 36 months or until death. No patients were lost to follow-up. Cumulative statistical analyses are presented for the 264 patients (269 CEAs) who actually received a randomized treatment allocation.
Results: Cumulative freedom from death or ipsilateral stroke at 3 years (including operative events) was 93.0% in the Dacron patch group and 95.5% in the vein group P =.42). Cumulative freedom from death or any stroke was 91.5% after Dacron patch closure and 93.9% after vein closure (P =.46). Cumulative freedom from recurrent stenosis greater than 70% or occlusion at 3 years was 92.9% for patients randomized to the Dacron patch group and 98.4% for patients randomized to the vein group (P =.03). At 3 years the incidence of stroke in the carotid territory not operated on was 1.0% in 93 patients with no contralateral internal carotid artery disease at randomization, and increased to 1.3% in 78 patients with 1% to 69% stenosis, and 2.0% in 51 patients with contralateral 70% to 99% stenosis. No late strokes occurred distal to 42 occluded contralateral internal carotid arteries.
Conclusions: Patch type has no influence on early operative risk, no association with enhanced patterns of thrombogenicity in the early postoperative period, and no influence on risk for ipsilateral or any stroke at 3 years. Dacron patches were, however, associated with a significantly higher incidence of recurrent stenosis at 3 years, with most occurring within 6 to 12 months of surgery. However, the higher incidence of recurrent stenosis was not associated with a parallel increase in late stroke, and in this study a program of serial ultrasound surveillance could not have prevented one ipsilateral stroke.