Multicentre diagnostic performance of on-site workstation CT derived fractional flow reserve.

Background: On-site workstation based computed tomography derived fractional flow reserve (CT-FFR) is an emerging method to assess vessel specific ischaemia in coronary artery disease (CAD). Global data on its diagnostic performance when compared with CT coronary angiography (CTA) is limited. Purpose(s): To evaluate the on-site multicentre diagnostic performance of reduced-order CT-FFR at detecting vessel specific ischaemia. Method(s): This is a retrospective pooled analysis of 141 patients (204 vessels) with suspected CAD enrolled from 3 global centres who underwent CTA, onsite CT-FFR and invasive FFR. On-site CT-FFR was performed using a reduced order model on a standard desktop computer with dedicated software. The per vessel diagnostic performance of CT-FFR (<=0.8) for vessel specific ischemia (FFR<=0.8) was compared with CTA (>=50% stenosis). Result(s): Mean age was 65.8+/-9.9, 70.7% were male. FFR significant stenosis was present in 34.3% (70/204) of vessels. Pearson correlation of CT-FFR for invasive FFR was 0.52, P<0.001. Bland Altman analysis demonstrated a mean difference of 0.06+/-0.15 (95% limits of agreement -0.22 to 0.35). Per vessel diagnostic accuracy, sensitivity and specificity of CT-FFR and CTA were 79.9% vs 53.5%; 78.6% vs 85.7%; 80.6% vs 35.9% respectively. Diagnostic performance as assessed by area under the receiver operator curve (AUC) for CT FFR was superior to CTA (0.82 [95% CI 0.76-0.88] vs 0.61 [0.55-0.67]; P<0.001). Conclusion(s): On-site workstation CT-FFR demonstrated high per vessel diagnostic performance and was superior when compared with CTA in assessment of vessel specific ischaemia as assessed by invasive FFR in a multicentre setting.