Eastern Vascular Society
EVS Home    |    Contact

Annual Meeting Home
 
Final Program
 
Past & Future Meetings

Back to 2016 Annual Meeting


Preoperative Risk Score Model to Characterize Predictors of Spinal Cord Ischemia Following Thoracic Endovascular Aortic Repair (TEVAR)
ALBEIR MOUSA, MD.,FACS.,RPVI.,MPH.,MBA1, Mike Broce, BA2, Ramez N. Morkous1, Michael Yacoub, MD.,FACS.,RPVI1, Andrew Sticco, MD.,FACS.,RPVI1, Mina Baskharoun, MD.1, Jenna Kazil, MD.1, Aravindra Nanjundappa, MD., FACC.,RVT1, Glibert R. Upchurch, Jr, MD3, Ali AbuRahma, MD.,FACS.,RPVI1.
1West Virginia University, Charleston, WV, USA, 2Charleston Area Medical Center Health Education and Research Institute, Charleston, WV, USA, 3University of Virginia Health System, Charlottesville, VA, USA.

Objectives:Spinal cord ischemia (SCI) is well recognized, but potentially preventable complication after TEVAR. Our objective was to determine significant predictors of SCI after TEVAR, and to develop a pre-operative risk score model to be used to anticipate post-operative events.
Methods: A retrospective review of a prospectively collected database from the SVS/PSO Virginia’s group was performed for all patients undergoing TEVAR. Preoperative demographics, procedure-related variables, and clinical details related to SCI were examined. A risk score model was developed based on multivariable logistic regression results to predict SCI. C statistics, equivalent to the area under the receiver operating characteristic curves, were used to determine if the raw risk scores or a categorical variable of low, medium or high risk could adequately summarize the original multivariate model.
Results: Of the 405 patients who underwent TEVAR during the study period (2012-2015), 32 patients (7.9%) developed SCI (temporary), of which 13 (3.2% overall and 40.1% of original) remained at discharge. Mean age of this cohort with SCI was 65.7 + 14.8, range, 18-89 years, and the majority were male (n=254, 62.7%). Fifty-two TEVARs were performed either urgently or emergently due to contained rupture (12.8%). Univariately, there was a strong trend for current smokers to develop SCI (11.1 versus 5.8%, p=0.060). Multivariately, there was a marginal association for subclavian coverage odds ratio [(OR):2.4, confidence interval (CI): 0.8-7.1, p=0.118)], independent predictors of SCI were celiac coverage (OR:5.2, CI: 1.8-14.7, p=0.002), dual subclavian and celiac coverage (OR:4.6, CI: 1.2-18.1, p=0.030), trauma indication (OR:9.2, CI: 2.9-29.3, p<0.001) and pre-operative spinal drain (OR:2.3, CI: 1.1-5.0, p=0.044). Probabilities from the original regression, raw score and raw score categories of low, moderate and high risk resulted in C statistics of 0.74, 0.74 and 0.71, respectively, indicating raw scores or categories adequately summarized the data (see Figure).
Conclusions: SCI following TEVAR is not benign with 40% remaining at discharge. Our results suggest that celiac or dual coverage, trauma and preoperative spinal drain greatly increased the odds of post-operative SCI. A raw SCI risk score or categories of low, moderate and high risk categories can be used to predict SCI.


Back to 2016 Annual Meeting