gms | German Medical Science

Objective: During carotid endarterectomies (CEA) intraoperative carotid artery (ICA) cross clamping is a critical event. Yet placement of a shunt itself is associated with higher rates of micro embolism and increased rates of postoperative cognitive impairment. Various warning criteria have been discussed to decide when temporary shunt placement is necessary to avoid insufficient collateral perfusion.

Methods: A retrospective review of all in 2014 performed CEA cases in our department of neurosurgery was done. Transcranial doppler sonography (TCD) flow velocities and median nerve somatosensory evoked potentials (SEP) amplitudes at several critical intraoperative events (baseline value, start of EEG burst suppression, changes in arterial blood pressure, cross clamping and vessel suture) have been analyzed. Furthermore a correlation to postoperative clinical outcome was done.

Results: In 2014 67 CEAs were performed on 44 male (66%) and 23 female patients (34%) at mean age of 72 years (SD +/- 9.8 years). 49 (73%) of all CEA procedures were monitored with both modalities (TCD and SEP). 17 (25%) surgeries were performed under SEP monitoring only and one patient with TCD monitoring only. At the critical time of ICA clamping TCD values on the surgical side were available in 44 (66%) of all patients whereas SEP values were available in 64 (96%). The criterion for arterio-arterial shunt was progressive reduction of the SEP amplitude or TCD flow velocity. In 4 patients (6%) an arterio-arterial shunt was placed. New permanent postoperative neurological motor or sensory deficits appeared in 3 patients (4%). Comparing both modalities during ICA clamping, SEP demonstrated more stable signals but TCD velocity a larger fluctuation. There was no significant correlation between both modalities. Combining patients with new postoperative neurological deficit and shunted patients the following observations could be made: The highest sensitivity (60%) was found applying a 30% relative change in TCD flow velocity as sole warning criteria. This low sensitivity was even compromised by a very low positive predictive value (27%) and a moderate specificity (79%). Contrary a 50% amplitude change in SEP as sole alarm criteria demonstrated an excellent specificity (100%), positive predictive value (100%) and negative value (92%) with the limitation of a low sensitivity.

Conclusion: The study demonstrates an excellent positive and negative predictive value, which might guide the decision whether to shunt or to proceed without shunting. Yet the number of intraoperative events was low which might have had an impact on the low sensitivity of our warning criteria. However we could demonstrate a very low rate of shunting compared to other series, with a low incidence of postoperative deficits.