Can carotid stenting registries help in determining predictors of neurological outcome?

Carotid stenting (CS) has proved to be a safe alternative to carotid endarterectomy, and subsets of patients who may benefit from carotid stents rather than surgery still need to be defined. Randomized trials and individual series are looking at predicting factors of early outcome after carotid angioplasty and stenting (CAS) that can hardly be applied to all series. We analyzed early results of the Eurocast registry and discussed the potential role of multicentric registries in determining such outcome predictors.

The Eurocast registry is an online prospective registry on CAS that collected 897 procedures from February 2000 to December 2005 across 20 centers. Statistical analyses were performed with the univariate Fisher's exact test on CS procedures, excluding sole balloon angioplasty. Risk factors reported were hypertension, hyperlipidemia, past or current history of smoking, diabetes mellitus, obesity, cardiac disease, other vascular disease, pulmonary disease, hostile neck and renal failure. Half of patients (50.2%) had symptomatic carotid lesions. Local anesthesia (91.8%) and retrograde femoral access (95.6%) were the preferred method. Overall procedures were achieved in 98.2%, of which 84.5% were performed under cerebral protection.

Intraoperative neurological complications occurred in 26 (3.1%) of 824 procedures achieved. No immediate deaths were reported. The intraoperative ipsilateral stroke/death rate was 1.7%. The overall in-hospital stroke/death rate was 3.5%. Significant predictors of early neurological complications were infarctions on preoperative cerebral CT scanning, a deficient circle of Willis, a too short preoperative length of medication (7 days or less), any additional intervention (stenting and/or balloon dilatation), bradycardia and hypotension. Less reliable predictors were: age >70 years, renal failure, preoperative speech or motor deficit, postradiation stenosis, lesion on the left carotid axis and lesion on the common carotid artery.

This analysis confirmed that CS is a feasible and secure technique to treat carotid stenosis. However, the determination of neurological outcome predictors remains subject to further confirmation. Other patient subgroups identified in other studies (female sex, diabetics, coronary disease) should be investigated more precisely. Finally, the Eurocast analysis showed that the setting up of risk scores would be dependent on local practices, physicians' experience and would not be reproducible.