The CatLet score, clinical variables and improved outcome predictions in acute myocardial infarction: a follow-up study of median 4.3-years

Abstract

Abstract Background The Coronary Artery Tree description and Lesion Evaluation (CatLet) score accommodating the variability in the coronary anatomy is a recently develpoed and comprehensive angiographic scoring system aiming to assist in risk-stratification of coronary artery disease patients. Our preliminary study demonstrated that the CatLet score predicted the clinical outcomes in patients with acute myocardial infaration (AMI) undergoing pirmary percutaneous intervention (PCI). Purpose The current study aimed to identify whether the clinical variables (CVs: age, creatinine and left ventricular ejection fraction) additionally improved the performance of the CatLet score with respect to the outcome predictions in AMI undergoing primary PCI. Methods The CatLet score was calculated retrospectively in consecutively enrolled 308 patients with AMI. The primary end point was major adverse cardiac or cerebrovascular events (MACCEs) at median 4.3-years. Secondary endpoints were all-cause death and cardiac death. Cox regression survival analysis was uesd to identify the associations between the CatLet score, clinical variables and the clinical outcomes. Discrimination was assessed by Harrell's C index, net reclassification improvement (NRI) and calibration was assessed by Hosmer-Lemeshow test and validation plots. Results The CatLet score remained a significant predictor for outcome predictions at a median 4.3-years follow-up after adjusting for the 3 CVs. The model incorporationg the CatLet score and these 3 CVs (CVs-adjusted model) performed better than the stand-alone CatLet model in terms of outcome predictions. Compared with the stand-alone CatLet model, Harrell's C index with the CVs-adjusted model significantly increased by 0.10 (P=0.002) in MACCE, by 0.14 (P<0.001) in all-cause death events and by 0.12 (P=0.001) in cardiac death events. When reclassifying patients with 4.3 year MACCE from the stand-alone CatLet model into the CVs-adjusted model, out of the 244 subjects who remained free of MACCE events, 101 were correctly reclassified to a lower risk category and 42 were reclassified to a higher risk category (categories: ≤10%, 10% to ≤20%, 20% to ≤40%, >40%); out of the 64 subjects who developed MACCE events, 26 were correctly reclassified to a higher risk category and 10 were reclassified to a lower risk category. The category-based overall NRI was 0.49 (P<0.001) and the continuous overall NRI was 0.74 (P<0.001). Hosmer-Lemeshow tested showed a better calibration in MACCE events in the CVs-adjusted model (X2=4.14, P=0.8440) than in the stand-alone CatLet model (X2=5.06, P=0.7515). Similar findings were found both in all-cause death and cardiac death. Conclusion The 3 CVs (age, creatinine and left ventricular ejection fraction) additionally improved the performance of the CatLet score with respect to the outcome predictions in AMI undergoing primary PCI. (ChiCTR-POC-17013536) Reclassification plot Funding Acknowledgement Type of funding source: Private grant(s) and/or Sponsorship. Main funding source(s): Sci-Tech Development Program, Grant/Award