We screened all consecutive patients who had undergone CAS procedures in Chungnam National University Hospital from August 2012 to January 2017, retrospectively. The CAS procedures were done in patients with symptomatic carotid stenosis and asymptomatic carotid stenosis ≥75%. We excluded all patients underwent CAS with DPDs because of following reasons; 1) total occlusion of contralateral carotid arteries, 2) involvement of common carotid artery (CCA), 3) severe proximal tortuosity or difficult aortic arch, 4) immediate and severe intolerance with a PPD which lead to use DFDs, and 5) physician's preference to use DFDs. In addition, patients with insufficient data about the procedure were excluded even in patients who were treated with PPDs. All clinical and procedural characteristics were confirmed by reviewing medical and procedural records. The patients were divided into 2 groups according to the occurrence of NI during CAS procedures; neurologic tolerance (NT) and NI groups.

All patients were treated with aspirin (100 mg/day; Boryung Pharmaceutical Company, Seoul, Korea) and clopidogrel (75 mg/day; Bristol-Myers Squibb/Sanofi Pharmaceuticals, Bridgewater, NJ, USA) for at least 3 days before the procedure, and combination therapy with clopidogrel was maintained at least 1 month after the procedure. All procedures were performed percutaneously through the femoral artery. Baseline angiography was performed on the carotid arteries and intracranial vessels of both sides using a 5-Fr Judkins right diagnostic catheter (Cook medical Inc., Bloomington, IN, USA). We checked the presence of anterior communicating artery (ACOM) communication. In cases with invisible ACOM connection, contralateral carotid angiogram while compressing ipsilateral CCA manually was performed to find ACOM connection (compression test). After the angiographic confirmation of the carotid target lesion stenosis, conventional heparin was intravenously injected at 100 IU/kg. We inserted a 0.035-inch soft guidewire into the distal part of the external carotid artery (ECA), and a 5-Fr Judkins catheter was advanced in the ECA. After the stabilizing the diagnostic catheter in the distal ECA, a 0.035-inch Amplatzer Extra stiff guidewire was changed instead of the soft guidewire. A 5-Fr sheath was replaced by a 9-Fr standard sheath and systemic arterial blood pressure was monitored during the procedures using this sheath. The MO.MA device (MO.MA Ultra^®; Medtronic, Minneapolis, MN, USA) was positioned via a 9-Fr standard sheath in the CCA with the single marker of the distal balloon located in the ECA, aiming at fully occluding all side branches including superior thyroid artery. All devices, balloons for pre- and post-dilatation and stent, for CAS should be prepared on the table before the ECA occlusion. The distal ECA balloon was inflated, and then the ECA occlusion was confirmed by small contrast injection. The proximal CCA balloon was inflated. We recorded pressures in both femoral artery and common carotid artery occlusion pressure (CCAOP) simultaneously. CCAOP was measured immediately after expansion of the proximal balloon of the MO.MA catheter after the insertion into the CCA (Figure 1A and B). Femoral arterial pressure was measured by the blood pressure measured from the distal aorta through a 25-cm-long sheath inserted into the femoral artery (Figure 1B). These 2 pressures were monitored throughout the procedure (Figure 1C and D). A 0.014-inch guide wire was advanced through the lesion and positioned in the distal internal carotid artery and pre-dilatation, stent deployment, and post-dilatation were performed as soon as possible. After the stent deployment, aspiration of 60 mL of blood was performed through the guiding catheter, and if no debris was found in the last basket, the occlusion balloons were deflated. An additional 20 mL of blood aspiration was performed if debris was found in the last basket. The final angiography included ipsilateral biplanar carotid and intracranial views were obtained to exclude cerebral embolisms.

A neurologist observed all the procedures and continuously evaluated the neurological state of patients before and after the procedures. Diffusion-weighted magnetic resonance imaging (DW-MRI) of brain was performed in all patients except 5 within 24 hours after the procedure.

Symptomatic carotid stenosis was defined as carotid stenosis with having ipsilateral neurological ischemic events within 6 months before the intervention. Procedural success was defined as residual stenosis ≤30% after the CAS without immediate procedural complications. NI was defined as transient neurologic symptoms during balloon occlusion. Neurological symptoms occurring within 24 hours after the procedure were regarded as periprocedural events. The definition of procedural events was following:

We defined neurologic events as major or minor stroke during the procedure.

The data were analyzed using commercial software (SPSS version 22.0; IBM Co., Chicago, IL, USA). Continuous data were presented as the mean ± standard deviation. Categorical variables were described with numbers and percentages, and unless otherwise specified. A Student's t-test was performed for statistical comparison between the nominal measures, while a χ² test was performed for the categorical data. Threshold for categorization of systolic CCAOP was specified using best cut-off value from receiver operator characteristic (ROC) curve. Binary logistic regression was used in the prediction of NI, and multivariable analysis was performed on parameters that were significant on the univariate analysis. We performed multivariate analysis with 4 variables to prevent overfitting of the model. Statistical significance was assumed at p<0.05.

We screened a total of 183 patients underwent CAS from August 2012 to January 2017 (Figure 2). Sixty patients were excluded: 45 patients treated with DFDs, 14 with insufficient pressure data and 1 with dissecting aneurysm. So, we analyzed total 123 patients in this study. Their clinical characteristics were summarized in the Table 1. The mean age was 72±8 years old, 89% were male, and 78% had symptomatic carotid stenosis. Of total 123 patients, 49 patients (40%) had NI during CAS procedures (NI group). Univariate analysis was performed to find the factor that associated with the NI, and the results were expressed together in the Table 1. NI was significantly associated with symptomatic carotid stenosis (odds ratio [OR], 6.915; 95% confidential interval [CI], 1.947–24.560; p=0.003). There were no statistical differences in the presence of underlying diseases such as hypertension, peripheral arterial disease and ischemic heart disease between the 2 groups. The incidence of diabetes mellitus was seemed to be higher in the NI group than in the NT group, but it was statistically insignificant (49% vs. 32%; p=0.067).

Lesion and procedural characteristics were compared in the Table 1. The target lesions were similar in the 2 groups. The mean degree of diameter stenosis was 91±6%, which means relatively severe obstructive lesions. There was no difference in the presence of severe tortuosity, intraluminal thrombus, ulcerative lesion and moderate to severe calcified lesion in the 2 groups. Three patients, who had a type III aortic arch that was difficult to approach, developed NI.

The total procedure time was 43±12 minutes, and the mean occlusion time was 4.8±1.2 minutes (range, 3.0–9.7 minutes). The 2 groups did not show a difference in the mean protection time. NI was observed frequently in patients without the visualization of ACOM on pre-procedural transfemoral cerebral angiography (TFCA) (OR, 3.456; 95% CI, 1.266–9.434; p=0.015).

The mean femoral artery pressure was 176/80 mmHg and the mean CCAOP was 50/38 mmHg. The systolic and diastolic femoral artery pressures did not show significant differences between the 2 groups (p=0.759 and 0.910, respectively). In contrast, the NI group had significantly lower systolic and diastolic CCAOP (39±18 vs. 57±20 mmHg; p<0.001 and 31±14 vs. 42±14 mmHg; p<0.001, respectively) than the NT group. NI were significantly associated with low systolic CCAOP (OR, 0.950; 95% CI, 0.928–0.973; p<0.001). The best cut-off value of systolic CCAOP after receiver operating curve analysis was less than 42 mmHg in the prediction of NI with sensitivity of 74% and specificity of 73% (area under the curve [AUC], 0.753; p<0.001). The OR of systolic CCAOP ≤42 mmHg was 7.477 (95% CI, 3.307–16.904; p<0.001). After multivariate analysis, systolic CCAOP ≤42 mmHg (OR, 6.461; 95% CI, 2.716–15.370; p<0.001) and symptomatic carotid artery stenosis (OR, 5.549; 95% CI, 1.415–21.767; p=0.014) were significant predictors of the developing NI during CAS with PPDs (Table 2).

Periprocedural and clinical outcomes were listed in the Table 3. Comparing the amount of debris captured from blood that was suctioned after the procedure, debris was observed in 45% of the NT patients, while it was observed in 71% of the NI patients (p=0.005). More than 10 debris were captured in 43% of the NI group and 18% of the NT group (p=0.009).

The procedures were completed successfully for all patients. Pre- and post-procedural DW-MRI data were available in 121 patients. No differences in the rate of new high signal intensity (HSI) in post-procedural DW-MRI were found between the 2 groups (33% vs. 22%; p=0.215).

NI was persistent more than 30 minutes in 6 patients of the NI group, but it was resolved completely within 2 hours in 2 patients and defined as TIA. However, 4 patients experienced persistent neurologic event; 1 patient (1%) had a major stroke and 3 (3%) had minor strokes. All these 4 neurologic events occurred in the NI group, and the neurologic symptoms, which developed during the procedures, persisted more than 24 hours afterward. Therefore, stroke incidence significantly higher in NI group (8.2% vs. 0%; p=0.023). No patient had myocardial infarction or death during the procedure and during their hospital admission. No new event occurred between 24 hours and 1 month after the procedure.

All 4 patients, who experienced periprocedural neurologic events, had symptomatic carotid artery stenosis, and ACOM was not visualized in 3 patients on the pre-procedural magnetic resonance angiography (MRA) or TFCA. In addition, 3 patients had low systolic CCAOP of less than 42 mmHg. Four patients with strokes exhibited new HSI on DW-MRI. Of them, 2 patients had watershed infarction on the ipsilateral border zone.