Propensity-Score Matched Analysis of Three Years Outcomes of TransCarotid Artery Revascularization versus Carotid Endarterectomy in The Medicare Database
Sina Zarrintan, Nadin Elsayed, Bryan Clary, Mahmoud B. Malas*
Department of Surgery, University of California San Diego, San Diego, CA
Objective: Carotid endarterectomy (CEA) remains the gold standard procedure for carotid revascularization. Transfemoral carotid artery stenting (TFCAS) was introduced as a minimally invasive alternative procedure in patients who are at high risk for surgery. However, TFCAS was associated with increased risk of stroke and death compared to CEA. Transcarotid artery revascularization (TCAR) has outperformed TFCAS in several prior studies and has shown similar perioperative and one-year outcomes compared to CEA. There are no prior studies to compare the mid-term outcomes of TCAR and CEA. We aimed to compare the one-year and three-years stroke and death of TCAR vs. CEA in the Vascular Quality Initiative (VQI)-Medicare-Linked (Vascular Implant Surveillance & Interventional Outcomes Network [VISION]) Database.
Methods: The VISION Database was queried for all patients undergoing CEA and TCAR between September 2016 to December 2019. The outcomes included one-year and three-years death, stroke and stroke/death. One-to-one propensity-score matching (PSM) without replacement was used to produce well-matched cohorts based on patients’ demographics, comorbidities, preoperative medications, insurance, anesthesia type, symptomatic status, urgency, stenosis ≥ 80% and physician volume.
Results: A total of 43,714 patients underwent CEA and 8,089 patients underwent TCAR during the study period. Patients in the TCAR cohort were older and were more likely to have severe comorbidities. PSM produced two well-matched cohorts of 7,351 pairs of TCAR and CEA. In the matched cohorts, there were no differences in one-year death (HR=1.13, 95%CI=0.99-1.30; P=.065), stroke (HR=0.97, 95%CI=0.85-1.10; P=.601) and stroke/death (HR=1.04, 95%CI=0.94-1.14; P=.475) (Tab. 1). At three-years, TCAR was associated with increased risk of death (HR=1.16, 95%CI=1.04-1.30; P=.008). However, there was no difference in three-year stroke (HR=0.97, 95% CI=0.86-1.09; P=.619) or stroke/death (HR=1.06, 95%CI=0.97-1.15; P=.194) (Fig. 1; Fig. 2; Tab. 1). When stratifying by initial symptomatic presentation, the increased three-years death associated with TCAR persisted only in symptomatic patients (HR=1.33, 95%CI=1.08-1.63; P=.008). The three-year stroke and stroke/death outcomes persisted after stratifying by symptom status (Tab. 2).
Conclusions: In this large multi-institutional PSM analysis with robust Medicare-linked follow-up, the rate of stroke/death at three years was similar in TCAR and CEA regardless of symptomatic status. The slight increase in the risk of three-years death despite similar stroke rate in symptomatic patients undergoing TCAR is likely confounded by more severe comorbidities despite matching. Therefore, TCAR can be considered a safe procedure in patients at high risk for CEA. A randomized controlled trial comparing TCAR to CEA is necessary to further determine the role of TCAR in standard risk patients requiring carotid revascularization.
Table 1: One-year and three-years outcomes of TCAR vs. CEA
| Before Match | After Match* | |||||||
| Outcomes | TCAR | CEA | HR (95% CI) Ref=CEA | P-Value | TCAR | CEA | HR (95% CI) Ref=CEA | P-Value |
| One-Year Death | 8.1% | 5.7% | 1.45 (1.31-1.60) | <0.001 | 7.8% | 7.0% | 1.13 (0.99-1.30) | 0.065 |
| One-Year Stroke | 7.3% | 6.4% | 1.15 (1.04-1.26) | 0.005 | 7.2% | 7.3% | 0.97 (0.85-1.10) | 0.601 |
| One-Year Stroke or Death | 13.7% | 10.8% | 1.26 (1.17-1.36) | <0.001 | 13.2% | 12.7% | 1.04 (0.94-1.14) | 0.475 |
| Three-Year Death | 22.0% | 16.8% | 1.43 (1.32-1.56) | <0.001 | 21.3% | 18.8% | 1.16 (1.04-1.30) | 0.008 |
| Three-Year Stroke | 13.0% | 10.8% | 1.17 (1.07-1.28) | <0.001 | 13.0% | 13.1% | 0.97 (0.86-1.09) | 0.619 |
| Three-Year Stroke or Death | 29.2% | 23.7% | 1.29 (1.21-1.38) | <0.001 | 28.8% | 27.3% | 1.06 (0.97-1.15) | 0.194 |
*Matched based on age, sex, race, insurance, smoking, comorbidities (hypertension, diabetes, coronary artery disease, prior coronary artery bypass graft or percutaneous coronary intervention, congestive heart failure, chronic obstructive pulmonary disease, chronic kidney disease and dialysis), preoperative medications (aspirin, P2Y12 inhibitor, statin and beta-blocker), urgency, anesthesia type, ipsilateral stenosis ≥ 80%, symptomatic status and physician volume. CEA, carotid endarterectomy; CI, confidence interval; HR, hazard ratio; TCAR, transcarotid artery revascularizationTable 2: One-year and three-years outcomes of TCAR vs. CEA in matched cohorts stratified by symptomatic status
| Asymptomatic | Symptomatic | |||||||
| Outcomes | TCAR | CEA | HR (95% CI) Ref=CEA | P-Value | TCAR | CEA | HR (95% CI) Ref=CEA | P-Value |
| One-Year Death | 7.1% | 6.6% | 1.08 (0.92-1.27) | 0.343 | 10.1% | 8.1% | 1.25 (0.98-1.60) | 0.069 |
| One-Year Stroke | 5.8% | 6.3% | 0.93 (0.79-1.09) | 0.350 | 11.3% | 10.7% | 1.02 (0.82-1.25) | 0.880 |
| One-Year Stroke or Death | 11.3% | 11.4% | 1.00 (0.88-1.12) | 0.937 | 18.7% | 16.5% | 1.10 (0.93-1.30) | 0.258 |
| Three-Year Death | 20.3% | 18.9% | 1.10 (0.96-1.25) | 0.166 | 24.2% | 18.5% | 1.33 (1.08-1.63) | 0.008 |
| Three-Year Stroke | 11.6% | 12.1% | 0.94 (0.81-1.09) | 0.428 | 17.5% | 16.2% | 1.00 (0.82-1.22) | 0.971 |
| Three-Year Stroke or Death | 26.7% | 26.6% | 1.03 (0.92-1.14) | 0.630 | 34.8% | 29.6% | 1.12 (0.96-1.30) | 0.161 |
CEA, carotid endarterectomy; CI, confidence interval; HR, hazard ratio; TCAR, transcarotid artery revascularization
Figure 1: Freedom from stroke and death (Stroke-Free Survival) before propensity score matching in transcarotid artery revascularization (TCAR) vs. carotid endarterectomy (CEA) cohorts
Figure 2: Freedom from stroke and death (Stroke-Free Survival) after propensity score matching in transcarotid artery revascularization (TCAR) vs. carotid endarterectomy (CEA) cohorts
Back to 2023 Abstracts