Internal carotid artery flow volume measurement and other intraoperative duplex scanning parameters as predictors of stroke after carotid endarterectomy

Abstract

Intraoperative duplex scanning (IDS) after carotid endarterectomy (CEA) has been shown to reliably identify major defects either by significant changes in peak systolic velocities or by B-mode imaging. To evaluate whether IDS could also predict postoperative strokes in technically flawless CEAs, we analyzed several hemodynamic parameters and correlated them with patient outcome. From March 2000 to February 2001, 226 consecutive primary CEAs were performed in 208 patients (120 men). Of these, 153 lesions were asymptomatic. General anesthesia and synthetic carotid artery patches were used routinely. Intraluminal shunts were used when internal carotid artery (ICA) back-pressures were <50 mm Hg (35% of cases). IDS consisted of B-mode and color-flow imaging and spectral analyses of the common, external, and internal carotid arteries. Volume flows were measured three times, and the mean flow rate was used for this study. The first set of data was analyzed when the twenty-ninth patient had the second immediate postoperative stroke. It was noted that the two patients who had postoperative strokes had mean ICA volume flows (MICAVF) of 48 mL/min and 85 mL/min. Only two additional patients had MICAVF <100 mL/min. The remaining 25 cases had MICAVF ranging from 102 to 299 mL/min, with a mean of 165 +/- 57 mL/min (+/-SD) (P <.02). Although there was a significant correlation between MICAVF and ICA peak systolic velocity (P <.01), the latter was not found to be a significant predictor of postoperative stroke. Moreover, end-diastolic velocities, resistive index, ICA diameter, and ICA back-pressure also did not correlate with neurologic events. These findings led us to change our protocol for patients with MICAVF <100 mL/min. This included a repeat set of volume flow measurements after 15 to 20 minutes, withholding the reversal of heparin, and the liberal use of completion arteriography. Of the following 197 CEAs, 26 (13%) were found to have MICAVF <100 mL/min (range 55 to 99 mL/min; mean 79 +/- 18 mL/min). Of these, five had arteriography that documented spasm of the intracranial portion of the ICA in four and a small-diameter ICA (<2 mm) in one. Except for the five cases, the remaining 21 cases had MICAVF >100 mL/min (range 105 to 158 mL/min, mean 127 +/- 20 mL/min [+/-SD]) on repeat study. Four patients with persistent ICA low flow (70 to 99 mL/min) were treated with postoperative anticoagulation. One of the last 197 patients had a stroke caused by hyperperfusion syndrome 2 weeks after operation. Overall, six of 226 cases (2.7%) required revision on the basis of abnormal B-mode imaging results or peak systolic velocities >150 cm/s. There were two common carotid artery flaps, two ICA stenoses, one ICA flap, and one localized thrombus. All six were successfully revised and had repeat normal IDS study results, and none of these patients had a postoperative stroke. IDS is helpful in identifying residual lesions or defects that may contribute to postoperative neurologic deficits. MICAVF <100 mL/min are suggestive of spasm that could lead to thrombus formation and stroke, particularly in the presence of synthetic patches. We suggest that heparin reversal should not be used unless ICA flow rates are >100 mL/min. ICA spasm is short lived in most patients undergoing CEA.