Other Relevant Articles
Original Research2019 Feb 25;12(4):335-342.
JOURNAL:JACC Cardiovasc Interv. Article Link
Szijgyarto Z, Rampat R, Werner GS et al. Keywords: chronic total occlusion; coronary artery disease; percutaneous coronary intervention; scoring system
OBJECTIVES - The aim was to establish a contemporary scoring system to predict the outcome of chronic total occlusion coronary angioplasty.
BACKGROUND - Interventional treatment of chronic total coronary occlusions (CTOs) is a developing subspecialty. Predictors of technical success or failure have been derived from datasets of modest size. A robust scoring tool could facilitate case selection and inform decision making.
METHODS - The study analyzed data from the EuroCTO registry. This prospective database was set up in 2008 and includes >20,000 cases submitted by CTO expert operators (>50 cases/year). Derivation (n = 14,882) and validation (n = 5,745) datasets were created to develop a risk score for predicting technical failure.
RESULTS - There were 14,882 patients in the derivation dataset (with 2,356 [15.5%] failures) and 5,745 in the validation dataset (with 703 [12.2%] failures). A total of 20.2% of cases were done retrogradely, and dissection re-entry was performed in 9.3% of cases. We identified 6 predictors of technical failure, collectively forming the CASTLE score (Coronary artery bypass graft history, Age (≥70 years), Stump anatomy [blunt or invisible], Tortuosity degree [severe or unseen], Length of occlusion [≥20 mm], and Extent of calcification [severe]). When each parameter was assigned a value of 1, technical failure was seen to increase from 8% with a CASTLE score of 0 to 1, to 35% with a score ≥4. The area under the curve (AUC) was similar in both the derivation (AUC: 0.66) and validation (AUC: 0.68) datasets.
CONCLUSIONS - The EuroCTO (CASTLE) score is derived from the largest database of CTO cases to date and offers a useful tool for predicting procedural outcome.
Copyright © 2019 American College of Cardiology Foundation. Published by Elsevier Inc. All rights reserved.