Relationship between CK-MB and high-sensitivity cardiac troponin after PCI and the impact on the diagnosis of periprocedural myocardial infarction: insights from Multivessel TALENT trial

Background Periprocedural myocardial infarction (PPMI) is associated with increased mortality after percutaneous coronary intervention (PCI). Although diagnostic biomarkers for PPMI have evolved from creatine kinase-myocardial band (CK-MB) to high-sensitivity cardiac troponin (hs-cTn), the current troponin thresholds have not been updated, and the relationship between post-PCI CK-MB and hs-cTn not re-evaluated. Consequently, the diagnosis and prognostic stratification of PPMI needs reassessment. Purpose We evaluated the relationship between post-PCI hs-cTn and CK-MB, and compare the frequency and clinical impact of PPMI across different definitions in patients with de novo three-vessel disease in a randomized PCI trial (Multivessel TALENT: NCT04390672). Methods This post-hoc sub-analysis from the Multivessel TALENT trial included patients with paired post-procedural CK-MB and hs-cTn measurements. Peak post-PCI CK-MB and hs-cTn values were used and expressed as ratios of the assay-specific upper limit of normal (ULN). PPMI rates were compared across three definitions: Society for Cardiovascular Angiography and Interventions (SCAI), Academic Research Consortium-2 (ARC-2), and Fourth Universal Definition of Myocardial Infarction (4th UDMI). Analyses using the SCAI definition were performed separately with CK-MB alone or hs-cTn alone. 1-year all-cause mortality was compared according to the presence of PPMI for each definition. Results Among 969 procedures (657 patients), paired post-PCI CK-MB and hs-cTnI or hs-cTnT measurements were available in 32.9% (288 procedures in 216 patients) and 67.1% (681 procedures in 441 patients), respectively. Post-PCI hs-cTn showed significant correlations with CK-MB on a log10-transformed scale, with a steeper slope for hs-cTnI than hs-cTnT (hs-cTnI: β=1.404 vs hs-cTnT: β=1.007, P<0.001). CK-MB thresholds of 5× and 10× ULN corresponded to hs-cTnT thresholds of 29× and 59×ULN, and to hs-cTnI thresholds of 101× and 266×ULN, respectively. Using the SCAI definitions, the incidence of PPMI was higher using hs-cTnI compared with CK-MB (15.6% vs. 1.7%, p<0.001), but not when using hs-cTnT compared with CK-MB (1.9% vs 1.0%, p=0.114). Using either hs-cTn did not affect the incidences of ARC-2 PPMI (hs-cTnI: CK-MB-based SCAI 1.7% vs ARC-2 3.8%, P=0.149; hs-cTnT: CK-MB-based SCAI 1.0% vs ARC-2 1.8%, P=0.267). However, the incidence of 4th UDMI PPMI was higher in both hs-cTn groups (hs-cTnI: CK-MB-based SCAI 1.7% vs 4th UDMI 12.5%, P<0.001; hs-cTnT: CK-MB-based SCAI 1.0% vs 4th UDMI 9.3%, P<0.001). Irrespective of the definition, PPMI was not associated with a significantly increased risk of all-cause mortality at 1 year. Conclusions Using hs-cTnI instead of CK-MB may overdiagnose PPMI by the purely biomarker-based SCAI definition. Refining the PPMI definition to incorporate differences between hs-cTnI and hs-cTnT is warranted.