Predictors of fractional flow reserve/instantaneous wave-free ratio discordance: impact of tailored diagnostic cut-offs on clinical outcomes of deferred lesions

Patient-related and lesion-related factors may influence instantaneous wave-free ratio (iFR)/fractional flow reserve (FFR) concordance, potentially affecting the safety of revascularization deferral.Consecutive patients with at least an intermediate coronary stenosis evaluated by both iFR and FFR were retrospectively enrolled. The agreement between iFR and FFR at their diagnostic cut-offs (FFR 0.80, iFR 0.89) was assessed. Predictors of discordance were assessed using multivariate analyses. Tailored iFR cut-offs according to predictors of discordance best matching an FFR of 0.80 were identified. The impact of reclassification according to tailored iFR cut-offs on major cardiovascular events (MACE: cardiovascular death, myocardial infarction or target-lesion revascularization) among deferred lesions was investigated.Two hundred and ninety-nine intermediate coronary stenosis [FFR 0.84 (0.78-0.89), iFR 0.91 (0.87-0.95), 202 left main/left anterior descending (LM/LAD) vessels, 67.6%] of 260 patients were studied. Discordance rate was 23.4% (n = 70, 10.7% iFR-negative discordant, 12.7% iFR-positive discordant). Predictors of discordance were LM/LAD disease, multivessel disease, non-ST-elevation myocardial infarction, smoking, reduced eGFR and hypertension. Lesion reclassification with tailored iFR cut-offs based on patient-level predictors carried no prognostic value among deferred lesions. Reclassification according to lesion location, which was entirely driven by LM/LAD lesions (iFR cut-offs: 0.93 for LM/LAD, 0.89 for non-LM/LAD), identified increased MACE among lesions deferred based on a negative FFR, between patients with a positive as compared with a negative iFR (19.4 vs. 6.1%, P = 0.044), whereas the same association was not observed with the conventional 0.89 iFR cut-off (15 vs. 8.6%, P = 0.303).Tailored vessel-based iFR cut-offs carry prognostic value among FFR-negative lesions, suggesting that a one-size-fit-all iFR cut-off might be clinically unsatisfactory.