Current Transthoracic Supra-Aortic Trunk Surgical Reconstruction Has Similar 30-Day Cardiovascular Outcomes Compared to Extra-Anatomic Revascularization but With Higher Morbidity Burden

Abstract

Operative risk for supra-aortic trunk (SAT) surgical revascularization for occlusive disease, particularly transthoracic reconstruction (TR), remains ill-defined. This study sought to describe and compare 30-day outcomes of TR and extra-anatomic (ER) SAT surgical reconstruction for an occlusive indication across the United States over a contemporary 15-year period. Using the National Surgical Quality Improvement Program, TR and ER performed during 2005-2019 were identified. Procedures performed for nonocclusive indications and those concomitant with coronary or valve operations were excluded. Rates of stroke, death, myocardial infarction (MI) and these as composite outcome (S/D/M) were compared. Logistic regression with stabilized inverse probability weighting (IPW) was used to compare groups via average treatment effect (ATE) while adjusting for covariate imbalances. Over the 15-year period, 166TR and 1,900 ER patients were identified. The majority of ERs were carotid-subclavian bypass (n=1,344; 70.7%) followed by carotid-carotid bypass (n=261; 13.7%) and subclavian/carotid transpositions (n=123; 6.5%). TR consisted of aorto-SAT bypass (n=120; 72.3%) and endarterectomy (n=46; 27.7%). The median age was 64years for TR and 65years in ER (P=0.039). Those undergoing TR were more often women (69.0% vs. 56.9%; P=0.001) and less likely to have undergone previous cardiac surgery (9.2% vs. 20.8%; P=0.006). TR were also less frequently hypertensive (68.1% vs. 75.4%; P=0.038) and had statistically lower preoperative creatinine levels (0.86 vs 0.91; P=0.002). Unadjusted rates of MI (0.6% vs. 1.3%; P=0.72) and stroke (3.6% vs. 1.9%; P=0.15) were similar between groups with mortality (3.6% vs. 1.5%; P=0.05) and S/D/M (6.6% vs. 3.9%; P=0.10) trending higher with TR. IPWs could be calculated for 1,754 patients (148TR; 1,606 ER). The estimated probability of S/D/M was 3.8% in the ER group and 6.2% in TR; no difference was seen in ATE (2.4%; 95% confidence interval [CI]: -1.5 to 6.2; P=0.23). No differences were seen in individual component ATEs (stroke: 3.0% vs. 1.7%; ATE=1.3%; 95% CI: -3.9 to 1.3; P=0.32; mortality: 3.8% vs. 1.4%; ATE=2.4%; 95% CI: -5.6 to 0.7; P=0.13). Secondary outcomes showed TR patients were more likely to have non-home discharge (18.7% vs. 6.6%; ATE=12.1%; 95% CI: 5.0-19.2; P<0.001) and longer lengths of stay (6.1 vs. 4.0; ATE=2.2days; 95% CI: 0.9-3.4; P<0.001). Moreover, TR patients were more likely to require transfusion (22.7% vs. 5.0%; ATE=17.7%; 95% CI: 10.2-25.2; P<0.001) and develop sepsis (2.7% vs. 0.2%; ATE=2.5%; 95% CI: 0.1-5.0; P=0.04). Transthoracic and extra-anatomic surgical reconstruction of the SATs for occlusive disease have similar operative cardiovascular risk. However, morbidity tends to be higher with TR due to higher transfusion requirements, sepsis risk, and need for facility stay. These results suggest ER as a first-line approach in those with proper disease anatomy is reasonable with lower morbidity, while TR remains justified in appropriate patients.