Ten-year single center experience with semi-eversion carotid endarterectomy: analysis of a cohort of patients with tips and tricks of our technique

The two most common surgical techniques to prevent ischemic stroke in patients with asymptomatic significant extracranial atherosclerotic carotid disease, or in symptomatic patients, are conventional CEA with patch closure (C-CEA) and eversion CEA (E-CEA) (1,2). We describe and present our ten-year experience on the efficacy and safety of semi-eversion CEA (SE-CEA) technique. CEA was offered to 1273 patients for significant (>70%, according to NASCET criteria), or symptomatic, primitive carotid artery stenosis during 2008-2018 (3). All had a preoperative carotid duplex ultrasound scan (DUS). Few patients with DUS of difficult interpretation received a preoperative computed tomography or magnetic resonance angiography. Long-term surveillance was based on carotid DUS at 3, 6 months, and each year thereafter. Restenosis was defined as a reduction in the diameter of the target vessel of at least 70%, diagnosed by a peak systolic velocity of at least 3 m/sec (4). All patients received prior informed consent from the operator or assistant, in compliance to the Helsinki informed consent declaration. Surgery is performed under general anesthesia and near-infrared spectroscopy (NIRS) for cerebral monitoring of regional oxygen saturation (rSO2). Surgical exposure is achieved in the usual fashion. Unfractionated heparin (2500 IU) are administered to the patient. The common carotid artery (CCA) is carefully exposed and a silastic vessel loop is double passed around it. The dissection is continued distally to the carotid bifurcation, where a local injection of a few ml of lidocaine is often required to avoid bradycardia. The hypoglossal nerve is recognized and respected, and the dissection is continued to the ICA, external carotid artery (ECA), and superior thyroid artery (STA). Sufficient distal dissection of the ICA past the finger-palpable end of the plaque is required to allow clamping of the artery on the free-plaque portion. A silastic vessel loop is double passed around the ICA and ECA. Vascular clamps are applied to CCA, ECA and STA, and 1 minute neurological assessment is performed. If rSO2 decreased to < 20% of the baseline value, the procedure is converted to C-CEA with patch closure, and a Pruitt carotid shunt (LeMaitre Vascular, Inc) is inserted and open. If no rSO2 decrease occurs after clamping, a longitudinal arteriotomy of the postero-lateral wall of the carotid bifurcation, slightly extended to ICA, is performed (Figure 1). After checking its back-flow, a vascular clamp is applied to ICA. Then, with the aid of a Watson-Cheyne dissector, the CCA, carotid bifurcation, and ICA are dissected away from the plaque. The plaque is completely divided at the level of CCA (Figure 2). The plaque in the ECA is broken with the help of the vascular clamp, distal to its origin, and then pulled out of the artery with the help of a surgical forceps. With the aid of the dissector, the distal ICA is carefully dissected away from the distal part of the plaque, which is cranially followed up to its distal endpoint. The wall of the ICA is distally everted by the second operator with two forceps, and the plaque is removed in one piece by the operator with the surgical forceps (Figure 3); any residual intimal debris can be removed under direct vision before replacing the ICA to its normal configuration. If the operator is not satisfied with his visualization of the endpoint of the plaque, the arteriotomy is extended distally along the ICA, and a C-CEA with patch closure and a Pruitt carotid shunt (LeMaitre Vascular, Inc) is employed. It is important to create a smooth dissection between the arterial wall and the plaque to limit the development of intimal flaps and debris, allowing the one piece-plaque removal. After checking the back-flow from the ICA and moving the clamp cranially, the arteriotomy is primarily closed with a continuous 6-0 Premio (Peters Surgical International Co, Ltd) round trip suture (Figure 4). Right before ending the suture, flushes from all carotid vessels are performed. The clamps are then removed from the ECA and CCA, and after some seconds from the ICA. The operating time is concluded in the usual fashion. Demographic, clinical, surgical, outcome variables, and follow-up of the entire cohort are presented in the Table. The EVEREST multicenter randomized trial compared C-CEA and E-CEA (3), finding no significant difference in postoperative mortality rate or in the onset of new neurological deficit between these two techniques, and demonstrating that C-CEA requires shorter clamp times than E-CEA and, when a patch was used during C-CEA, both techniques had a similar low rate of restenosis. We demonstrated a low ICA restenosis rate at a mean follow-up of 38 mo (the mean time of restenosis detection was 12 mo), comparable to any CEA restenosis rate reported in a recent systematic review (5). With SE-CEA a low restenosis rate may be ensured by the surgeon’s skill in performing the arteriotomy. Indeed, when performing the arteriotomy on the lateral-posterior wall of the carotid bifurcation, slightly extended to ICA, it is possible to suture the carotid in a wide area of the arterial wall itself, avoiding to incur in suture defects. Furthermore, by removing the plaque from ICA under the direct view of the endpoint it is unlikely leaving out the ending part of the plaque, causing residual stenosis. In our experience, patients treated with SE-CEA present a perioperative stroke and death rates widely less than 3%. Despite the encouraging results, our technique presents some limitations. In case of very long plaque it may be difficult to clearly define an endpoint into the ICA, and SE-CEA proves to be impractical. Conversion to C-CEA with patch closure is recommended in these cases, and we prefer to shunt when using a patch. Furthermore, in cases of shunt required for rSO2 decrease to less than 20% of the baseline value, we prefer to extent the arteriotomy distally into the ICA, for more safety when inserting the Pruitt shunt, hence performing a conversion to patch closure. This may represent another limitation of our technique. However, this scenario occurred only in few cases, indeed the most critical patients (i.e., those with occlusion of the contralateral ICA, and those with significant decrease of cerebral rSO2 detected by NIRS during carotid cross-clamping).