Validation of the newly introduced CASTLE Score for predicting successful CTO recanalization

Background: The new EuroCTO CASTLE Score was validated against the widely adopted Japanese Multicenter CTO Registry (J-CTO) score in predicting technical success in percutaneous coronary intervention (PCI) for coronary chronic total occlusions (CTO). Methods: A total of 463 patients treated by CTO PCI were included in a retrospective analysis. Result: The mean CASTLE score was 2.23 ± 1.1 and J-CTO score was 2.84 ± 1.0. The overall technical success rate was 83.2%. At 30 days follow up, a primary composite safety endpoint showed a low proportion of stent thrombosis (0.2%) and re-hospitalization (0.4%). Moreover, an improvement of clinical symptoms was found in 83% of patients. Receiver operating characteristic analysis (ROC) demonstrated a comparable overall discriminatory performance in predicting technical outcome: CASTLE score, area under the ROC curve (AUC) 0.668, 95% CI: 0.606–0.730; J-CTO score AUC 0.692, 95% CI: 0.631–0.752; Comparison of AUCs: p = 0.324. Those findings were even consistent in more complex procedures CASTLE Score ≥ 4 and J-CTO score ≥ 3: CASTLE Score AUC 0.514, 95% CI: 0.409–0.619; J-CTO score, AUC 0.617, 95% CI: 0.493–0.741; Comparison of AUCs: p = 0.211. Furthermore, increasing score values are accompanied by a longer examination and fluoroscopy time, more contrast medium and a higher dose area product. Conclusion: Compared to the widely accepted J-CTO score, the new introduced EuroCTO CASTLE score demonstrated a comparable overall discriminatory performance in predicting technical outcomes in CTO PCI.