A predictive score for spinal cord ischemia after open thoracoabdominal aortic aneurysm repair

To develop a simple preoperative risk score and assess its performance to predict the occurrence of spinal cord ischemia (SCI), in patients undergoing open surgical repair of thoracoabdominal aortic aneurysm (TAAA).

A retrospective analysis was conducted on patients who underwent elective TAAA open repair between 2016 and 2020. A preoperative risk score (0-6 points) was constructed based on six binary variables: TAAA extent II, body mass index of ≥30, smoking history, preoperative diuretic use, age >70 years, and history of chronic kidney disease. Each variable contributed 1 point. The score was applied retrospectively to the study cohort. Discriminative performance was assessed by receiver operating characteristic analysis, and the area under the curve (AUC) was calculated with a 95% confidence interval (CI). Sensitivity and specificity were calculated using the optimal cut-off based on the Youden index.

Among the 246 patients, SCI occurred in 33 patients (13.4%). The incidence of SCI increased progressively with higher preoperative scores: 0/71 (95% CI, 0.0-5.1) for a score of 0, 1/93 (95% CI, 0.03-5.9) for a score of 1, 4/47 (95% CI, 2.4-20.4) for a score of 2, 8/15 (95% CI, 26.6-78.7) for a score of 3, and all patients with scores of 4 (11/11; 95% CI, 71.5-100) and 5 (9/9; 95% CI, 66.4-100). The optimism-corrected AUC was 0.902 (apparent AUC, 0.919) after bootstrapping internal validation. At the Youden-optimal cut-off of ≥2, sensitivity was 92.1%, specificity 78.9%, positive predictive value 42.7%, and negative predictive value 98.3%. Decision curve analysis confirmed a clinical benefit across a wide range of threshold probabilities.

This preoperative risk score, derived from a single-center cohort and internally assessed with bootstrap resampling, offers a simple tool to estimate SCI risk in patients undergoing open TAAA repair. These preliminary findings may support risk stratification and patient counseling, but external validation will be required before broader clinical use can be recommended.