Accuracy of the Euro CTO(CASTLE) score obtained on coronary computed tomography angiography for Predicting 30-minute wire crossing in chronic total occlusions

BACKGROUND: To investigate the feasibility and accuracy of the Euro CTO (CASTLE)(CTA) score obtained on coronary computed tomography angiography (CCTA) for predicting the success of percutaneous coronary intervention (PCI) and the 30-min wire crossing in chronic total occlusions (CTO). METHOD: One hundred and fifty patients (154 CTO cases; median age, 61 (interquartile range [IQR], 54–68) years; 75.3% male) received CCTA at the People's Hospital of Liaoning Provincce within 1 month before the procedure. The Euro CTO (CASTLE) score obtained on CCTA(CASTLE(CTA)) was calculated and compared with the Euro CTO (CASTLE) score obtained based on coronary angiography (CASTLE(CAG)) for the predictive value of 30-min wire crossing and CTO procedural success. RESULTS: In our study, the CTO-PCI success rate was 89.0%, with guidewires of 65 cases (42.2%) crossing within 30 min. There were no significant differences in the median CASTLE(CTA) and CASTLE(CAG) scores in the procedure success group (3 [IQR, 2–4] vs 3 (IQR, 2–3]; p = 0.126). However, the median CASTLE(CTA) score was significantly higher than the median CASTLE(CAG) score in the procedure failure group (4 [IQR, 3–5.5] vs 4 [IQR, 2.5–5.5]; p = 0.021). There was no significant difference between the median CASTLE(CTA) score and the median CASTLE(CAG) score in the 30-min wire crossing failure group (3 [IQR, 3–4] vs 3 [IQR, 2–4]; p = 0.254). However, the median CASTLE(CTA) score was significantly higher than the median CASTLE(CAG) score in the 30-min wire crossing group (3 [IQR, 2–3] vs 2 [IQR, 2–3]; p < 0.001). The CASTLE(CTA) score described higher levels of calcification than the CASTLE(CAG) score (48.1% vs 33.8%; p = 0.015). There was no significant difference between the CASTLE(CTA) score (area under the curve [AUC], 0.643; 95% confidence interval [CI], 0.561–0.718) and the CASTLE(CAG) score (AUC, 0.685; 95% CI, 0.606–0.758) for predicting procedural success (p = 0.488). The CASTLE(CTA) score (AUC, 0.744; 95% CI, 0.667–0.811) was significantly better than the CASTLE(CAG) score (AUC, 0.681; 95% CI, 0.601–0.754; p = 0.046) for predicting 30-min wire crossing with the best cut-off value being CASTLE(CTA) ≤ 3. The sensitivity, specificity, positive predictive value, and negative predictive value were 90.8%, 55.2%, 54.6%, and 87.0%, respectively. CONCLUSION: The CASTLE(CTA) scores obtained from noninvasive CCTA perform better for the prediction of the 30-min wire crossing than the CASTLE(CAG) score.