Article Text

Download PDFPDF
106 Virtual-PCI With iFR-Pullback can Plan Coronary Intervention in Complex Coronary Disease with a Prediction of Functional Gain
  1. Sukhjinder Nijjer,
  2. Sayan Sen,
  3. Justin Davies,
  4. Ricardo Petraco,
  5. Rasha Al-Lamee,
  6. Christopher Broyd,
  7. Ghada Mikhail,
  8. Iqbal Malik,
  9. Jamil Mayet,
  10. Darrel Francis
  1. Imperial College London


Background Coronary intervention is increasingly performed in complex and tandem coronary disease where physiological optimisation is less practical. Advances in computation allow physiological maps of stenosis-specific coronary pressure loss to be mapped upon angiographic images. Virtual-PCI conceptually allows selective removal of a stenosis and predict the functional gain of a given stenting strategy. We sought to assess the change in resting haemodynamics after intervention and the feasibility of Virtual-PCI to predict post-PCI iFR.

Method Pressure and flow velocity measurements were made before and after PCI to 75 stenoses. Data was used to model the change in pressure ratios after PCI. Intracoronary motorised pressure-wire iFR-pullbacks were performed at rest in a separate cohort of 32 coronary arteries with tandem and diffuse disease to iFR changes onto the angiogram. Virtual-PCI predicted a post-PCI iFR (iFRpredicted) assuming the perfect removal of a given stenosis (Figure 1). This was compared to the observed iFR post real-world PCI (iFRobserved).

Abstract 106 Figure 1

A. Pre-PCI iFR-Pullback in a vessel with tandem and diffuse disease. B. Virtual PCI algorithms model the effect of removing a stenosis

Results Resting flow velocity over the wave-free period changed by only 4.4 cm/s; a demonstrable change was noted when stenoses had pre-PCI FFR <0.60 but trivial change (1.3 cm/s) in moderate stenoses. In contrast, pressure gradients and ratios rose after PCI without a change in flow suggesting the interaction between stenoses at rest is minimal. In the tandem disease cohort, mean vessel iFR pre-PCI was 0.78 ± 0.03. Virtual PCI predicted an iFRexpected of 0.94 ± 0.01. The observed iFR after real-world PCI was iFRobserved 0.93 ± 0.01 with no significant difference between the predicted and observed increase (ΔiFRexpected:0.16 ± 0.03 vs ΔiFRobserved 0.13 ± 0.03 p = 0.48). No significant systematic bias between iFRexpected and iFRobserved was noted (Bland-Altman mean difference 0.016 ± 0.004).

Conclusions Virtual-PCI enables selective removal a stenosis from an iFR-pullback physiological map of the vessel and predict the haemodynamic effect of that chosen intervention. The post-PCI physiological result can be predicted without systematic bias permitting virtual assessment of the functional gain of different stenting strategies. Virtual-PCI using iFR-Pullback is feasible in the catheter laboratory and real-time algorithms may enable live co-registration with the angiogram to make an advance in how intervention is performed in the future.

  • iFR
  • Prediction
  • stenting

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.