Predictive Value of Preoperative Multimodal CT for Outcomes Following Thrombectomy in Anterior Circulation Stroke

Objective Although mechanical thrombectomy is widely used for acute large intracranial vessel occlusion, not all patients benefit from this intervention. We aim to utilize preoperative multimodal CT to predict outcomes and guide individualized management after mechanical thrombectomy in acute anterior circulation occlusive stroke. Methods A total of 68 patients with acute anterior circulation occlusive stroke who underwent mechanical thrombectomy and successful recanalization in Sanming First Hospital Affiliated to Fujian Medical University from January 1, 2022 to June 30, 2024 were continuously collected. According to the 3 month modified Rankin score since onset, the patients were divided into the effective recanalization group (0–3 points) and the ineffective recanalization group (4–6 points). The clinical and imaging datas of the patients were analyzed, and the best cut-off value was calculated for the statistically significant ordinal or continuous variables. Multivariate logistic regression analysis was used to establish the combined model named CNCTA based on clinical and multimodal CT imaging. A nomogram was constructed based on the final model, and the model's performance was evaluated using receiver operating characteristic (ROC) curve analysis, calibration curve analysis, decision curve analysis, and K-fold cross-validation. Results A total of 68 (46 males) patients were enrolled in our study. There were 47 patients (69.12%) in the effective recanalization group and 21 patients (30.88%) in the ineffective recanalization group. Compared with the effective recanalization group, the patients in the ineffective recanalization group were older ( P  = 0.006), the occlusion site was more likely to involve the internal carotid artery ( P  = 0.010), a longer thrombus ( P  = 0.008), Alberta stroke program early CT score (ASPECTS) and collateral score (CS) were lower ( P  = 0.003 and 0.001, respectively). The best cut-off values of age, ASPECTS, length of thrombus and CS were 61.5 years, 7.5 points, 19.9mm and 3.5 points, respectively. ASPECT-1 was dichotomized from ASPECTS using the optimal cutoff value. The CNCTA model variables were composed of age ( P  = 0.005), site ( P  = 0.020), ASPECT-1 ( P  = 0.006) and CS ( P  = 0.004). The AUC was 0.917 (95%CI: 0.852–0.983). The Hosmer-Lemeshow goodness of fit test of CNCTA model was χ ²  = 3.167, P  = 0.923. The calibration curve exhibited an intercept of 0.0427 and a slope of 0.9371. The mean AUC from 5-fold cross-validation was 0.90 ± 0.07. Conclusion The CNCTA model based on clinical-multimodal CT imaging showed excellent ability in predicting the prognosis of acute anterior circulation occlusive patients with successful recanalization. The visual nomograph of CNCTA model can quantitatively evaluate the prognosis of patients, with wide applicability, it is useful for screening the patients required surgery.