Is local coronary inflammation truly superior after PCI, or is the comparison driven by revascularization endpoints?

This commentary refers to ‘Local vs systemic inflammation and outcomes after percutaneous coronary intervention’, by Z. Ye et al., https://doi.org/10.1093/eurheartj/ehag731 and the discussion piece ‘Local and systemic inflammation after PCI: complementary risk signals beyond revascularization’, by Z. Ye and K. Dou, https://doi.org/10.1093/eurheartj/ehag932. We read with great interest the study by Ye et al. evaluating the relative prognostic value of local coronary inflammation, assessed by the pericoronary fat attenuation index (FAI), and systemic inflammation, assessed by high-sensitivity C-reactive protein (hs-CRP), in patients undergoing percutaneous coronary intervention (PCI).1 We congratulate the authors on this large, carefully adjusted analysis. We believe, however, that the conclusion that local inflammation shows a stronger association than systemic inflammation requires two clarifications. First, the patient-level MACE endpoint was dominated by unplanned repeat revascularization (130 of 164 events, 79%), and the vessel-level VOCE endpoint was dominated by target-vessel revascularization (97 of 118 events, 82%). Revascularization events are influenced by the completeness of revascularization, residual disease burden, operator judgment, and follow-up practice, rather than by inflammatory biology alone.2,3 Indeed, in the sensitivity analysis excluding revascularization from MACE, isolated high FAI-MAX with low hs-CRP no longer carried a significantly higher risk than low FAI-MAX with high hs-CRP (hazard ratio 1.70, 95% CI .69–4.15 vs 1.79, 95% CI .67–4.79). Although the dual-high phenotype remained the highest-risk subgroup, the claimed superiority of local over systemic inflammation was no longer apparent. We would welcome event-specific analyses for death and myocardial infarction and a formal evaluation of whether the incremental predictive value of FAI over hs-CRP persists after exclusion of revascularization, using reclassification metrics rather than area under the curve alone.4