A proposed grading system for spinal cord arteriovenous shunts

Abstract Background The clinical outcomes of microsurgery for spinal cord arteriovenous shunts (SCAVSs) exhibit fluctuations due to varying patient selection criteria, underscoring the importance of a standardized surgical grading system that can effectively stratify the feasibility of SCAVSs resection. Methods A cohort of 308 consecutive patients with surgically treated SCAVSs was randomly divided into a modeling group and a validation group. The surgical grading system was developed based on the independent risk factors of incomplete resection identified in the modeling group and subsequently verified in the validation group. The system’s specificity and sensitivity were tested through Receiver Operating Characteristic (ROC) analyses. Results Multivariate analysis indicated that metameric AVSs (p=0.007), AVSs with maximum length ≥3 cm (p=0.017), embedded AVSs (p=0.032) and anterior sulcal artery supply (p=0.013) were independent risk factors of incomplete resection. Subsequently, each of the four parameters is assigned one point, and the SCAVSs grade is calculated by aggregating all parameter scores. The area under ROC curve (AUC) of modeling group and validation group was 0.856 (95% confidence interval [95% CI], 0.794-0.919) and 0.819 (95% CI, 0.747-0.892) respectively. Across the entire cohort, patients with scores ranging from 0 to 4 exhibited complete resection rates of 88.7%, 66.7%, 30.6%, 4.5% and 6.3%. The corresponding rates of severe treatment-related deterioration were found to be at levels of 6.0%, 12.0%, 12.9%, 31.8% and 25.0 %, respectively. Conclusion: The proposed grading system effectively stratifies the surgical feasibility of SCAVSs based on both the probability of achieving complete resection and the treatment risk. Its simplicity renders it a valuable tool for clinical decision-making, as well as a reference point for evaluating treatment outcomes across different centers and surgical techniques.