Long-term prognostic value of non-invasive lesion-specific hemodynamic index

Abstract

Abstract Background With advancement in computational fluid dynamics (CFD) technology, novel lesion-specific hemodynamic parameters can be estimated non-invasively. However, their long-term prognostic implications have not been fully defined. Purpose We sought to investigate the ten-year outcomes of lesion-specific hemodynamic indices derived-from coronary CT angiography (CCTA). Methods A total of 145 vessels (95 patients) with stable coronary artery disease who underwent fractional flow reserve (FFR) derived by coronary computed tomographic angiography (FFRCT) and invasive FFR measurement was included from the first-in-human study of FFRCT. Study participants were enrolled from October 2009 to January 2011 and were followed up until December 2020. A total of 340 lesions with % diameter stenosis ≥30% were identified, and wall shear stress (WSS) and change FFRCT across the lesion (ΔFFRCT) were obtained using CFD techniques by an independent core laboratory. The optimal cut-off for WSS and ΔFFRCT was applied for outcome analysis. The primary endpoint was target lesion failure (TLF) including cardiovascular death, target vessel myocardial infarction, and target lesion revascularization at 10 years. Results The median WSS and ΔFFRCT was 183.3 [112.8; 273.9] and 0.06 [0.03; 0.13]. WSS and ΔFFRCT was mildly correlated with FFRCT (r=−0.18, P=0.001 for WSS; r=−0.36, P<0.001 for ΔFFRCT). Of 179 lesions with deferral of revascularization at the index procedure of FFR measurement, TLF occurred in 16 (8.9%) lesions. In prediction of 10-year TLF, % diameter stenosis (per-lesion) ≥50%, FFRCT ≤0.80, WSS ≥256.1 dyn/cm2, and ΔFFRCT ≥0.06 were significant predictors. However, in multivariate analyses with those predictors, % diameter stenosis (per-lesion) ≥50%, FFRCT ≤0.80 became insignificant, and lesion-specific hemodynamic indices were only predictive of 10-year TLF (adjusted hazard ratio [aHR] 2.66, 95% confidence interval [CI] 0.98–7.22, P=0.055 for WSS ≥256.1 dyn/cm2; aHR 5.88, 95% CI 1.10–33.25, P=0.045 for ΔFFRCT ≥0.06). WSS ≥256.1 dyn/cm2 and ΔFFRCT ≥0.06 had higher information gain in predicting outcomes than % diameter stenosis (per-lesion) ≥50% and FFRCT ≤0.80, and both improved predictability for 10-year TLF risk of the model with % diameter stenosis (per-lesion) ≥50% and FFRCT ≤0.80 (P=0.068 for WSS ≥256.1 dyn/cm2; P=0.011 for ΔFFRCT ≥0.06) (Figure). Conclusions Non-invasive lesion-specific hemodynamic indices (i.e., high WSS and high ΔFFRCT) were the robust predictors of 10-year outcomes of a target lesion with incremental predictability over anatomical severity and low FFRCT. Clinical application of non-invasive hemodynamic indices will provide better long-term risk stratification over the current prognostication scheme before an invasive procedure. Funding Acknowledgement Type of funding sources: Private company. Main funding source(s): Abbott, Philips, HeartFlow