Front Cardiovasc Med. 2026 May 26;13:1832830. doi: 10.3389/fcvm.2026.1832830. eCollection 2026.
ABSTRACT
METHODS: A total of 618 STEMI patients who underwent PCI were retrospectively enrolled. The primary endpoint was 1-year MACE. Feature selection was performed using the Boruta algorithm. Multivariable Cox regression analysis was conducted to evaluate the independent association between CTI and MACE. The incremental prognostic information provided by CTI was assessed using 365-day time-dependent receiver operating characteristic (ROC) curves and the DeLong test, while clinical utility was evaluated using decision curve analysis (DCA). Subgroup analysis was performed according to diabetes status. Sensitivity analyses using different adjustment strategies were conducted to assess the robustness of the results.
RESULTS: During follow-up, 139 patients (22.5%) experienced MACE. CTI was identified as an important predictor by the Boruta algorithm and remained independently associated with MACE in multivariable Cox regression analysis (HR = 1.443, 95% CI: 1.191-1.748, P < 0.001). Although the addition of CTI to the baseline model resulted in only modest improvement in discrimination and clinical net benefit, further component analysis showed that the association between CTI and MACE was mainly driven by acute heart failure events. Subgroup analysis demonstrated consistent associations between CTI and MACE regardless of diabetes status (P for interaction = 0.504). The association between CTI and MACE remained robust across multiple sensitivity analyses.
CONCLUSIONS: CTI was an independent predictor of MACE in STEMI patients undergoing PCI. Notably, the association appeared to be mainly driven by acute heart failure events, suggesting a close relationship between CTI and post-infarction ventricular dysfunction.
PMID:42273597 | PMC:PMC13246453 | DOI:10.3389/fcvm.2026.1832830