1. Is TCAR best under LA or GA
Is TCAR best under LA or GA? Results from the SVS-VQI TCAR Surveillance Project
Dipankar Mukherjee, MD, FACS, Devon T. Collins, MPH, Chang Liu, PhD, Jeffery Jim, MD, MSPH, FACS
Inova Fairfax Medical Campus, Department of Surgery, Department of Vascular Surgery Falls Church, VA
Section of Vascular Surgery, Washington University, School of Medicine, St. Louis, MO
June 11th, 2019

2. Disclosures
I have nothing to disclose.

3. Study Objectives
The objective of this study was to examine if there was a difference in patient outcomes between Transcarotid Artery Revascularization (TCAR) done under local anesthesia (LA) compared to general anesthesia (GA) utilizing a large national database.
Since TCAR is less invasive than carotid endartarectomy (CEA) and can provide better ‘neuroprotection’, but there are conflicting claims to be assessed.
We hope that the findings could help improve patient care and help guide best practices in the field.

4. Methods Retrospective national database analysis Statistical Analyses:
Prospectively collected Vascular Quality Initiative (VQI) data
Patients receiving TCAR under GA or LA
2016 to 2018
Primary Outcome: Composite measure of postoperative stroke, TIA, MI, and/or death occurrence.
Secondary Outcomes: Flow reversal time, radiation dose, contrast volume, total procedure time, and length of stay (days).
Statistical Analyses:
Baseline demographics and clinical factors of the study sample were described using medians and interquartile ranges [IQR] or Chi-squared/Fisher’s exact test for measuring differences in proportions.
Univariate and multivariate analyses were employed using one unmatched and 2 matched datasets with propensity score matching to compare the possible effects of GA vs. LA in TCAR patients.

5. Results – Characterizing the Study Population
N = 2,609 TCAR patients included
2,146 (82.3%) = GA
463 (17.7%) = LA
No statistically significant differences between GA and LA were observed:
For median [IQR]:
Age = 74 years [67 to 80]
BMI = 27.5 [24.3 to 31.2]
For proportions N (%):
Male = 1,670 (64.0%)
Caucasian = 2,375 (91.0%)
ASA class III = 1,735 (66.7%)
No present diabetes = 1,649 (76.5%)
Elective TCAR = 2,382 (91.3%)
Asymptomatic stenosis indication = 1,598 (61.4%)
From what was on previous slides from Dr. Mukherjee:
There were no significant differences between the two groups with regards to demographics or medical co-morbidities. The LA group had a higher proportion of patients with CHF and those requiring semi-urgent interventions.
Approximately 60% of all interventions were done for asymptomatic disease and 40% were for symptomatic disease.
Primary composite endpoints were reported in 3.2 % of patients in the GA group versus 2.8% in the LA group (p=0.808).
Postoperative Stroke, TIA and death occurred in 2.5% of patients done under GA vs 2.4% of those done under LA (p=0.998).
Secondary composite endpoints were also not statistically different between the two groups as enumerated in table 1.
Multivariate linear regression models demonstrated significantly less need for contrast with procedures done under GA (p= 0.039).

6. Results – Multivariate Analysis
Statistically significant (p<0.05)
Table 1: Multivariate analysis of outcomes comparing general versus local anesthesia in TCAR patients utilizing unmatched crude numbers or matched numbers based on propensity score matching.
Univariate Analyses: (from poster in case you want to read this out loud at some point).
-Median [IQR] for contrast across all 3 propensity matching models were statistically significant, showing a difference between GA and LA (p<0.005).
-Median [IQR] for flow reversal time was only statistically significant in the 1:3 matching model.

7. Results – Multivariate Analysis
Trending toward statistical significance
Table 1: Multivariate analysis of outcomes comparing general versus local anesthesia in TCAR patients utilizing unmatched crude numbers or matched numbers based on propensity score matching.
Univariate Analyses: (from poster in case you want to read this out loud at some point).
-Median [IQR] for contrast across all 3 propensity matching models were statistically significant, showing a difference between GA and LA (p<0.005).
-Median [IQR] for flow reversal time was only statistically significant in the 1:3 matching model.

8. Strengths
A large national database allows for statistical analyses at an inferential level versus basic descriptive analyses.
Real-world, accurate, and representative TCAR evaluation and of the patient population.
Will help support future analyses and other prospective randomized trials/cohort studies to further assess and confirm current results and practices in our field.
Prospective Research

9. Limitations
82.3% of the current sample size reflected TCAR was in majority done under GA vs. LA
Possibly a function of the field’s current practices
Need more LA data as time goes forward
Data recording and coding errors:
Data coding and clinical practice differences between institutions
Missing values or incorrectly entered into system
Large databases can only collect so many variables
Cannot account for some necessary confounding variables not required by the database for a researcher’s specific question
Since TCAR is a minimally invasive procedure it would be ideal to perform the operation under LA over GA.

10. Discussion and Conclusions
Excellent outcomes from TCAR for Carotid stenosis were noted in the VQI database.
Vascular surgeons should perform the TCAR procedure with the anesthetic technique most comfortable with.
A learning curve does exist to performing TCAR under LA, therefore it is perfectly appropriate to do the procedure under GA until the operator feels comfortable doing the procedure under LA as the results are equivalent.

11. Thank You!

12. Extra Slides

13. Introduction – sort of incorperated into the objectives slide at start.
There are conflicting claims of superiority with CEA done under LA vs GA.
TCAR is a new Carotid stenting procedure with Neuro protection done by cut down over the Common Carotid artery.
TCAR is less invasive than CEA with outcomes similar to CEA and better than TFCAS by some recent reports.
TCAR is currently being done both under GA and LA.