Assessment of a conduction-repolarisation metric to predict Arrhythmogenesis in right ventricular disorders

Figures

• 

  Fig. 1. The RVI metric in the case of block (A) and re-entry (B). Stylised action potentials are shown during steady state pacing (S1) followed by a short coupled premature beat (S2). As shown in the top panels, the wave-front travels from a proximal (P) to a distal (D) site. During the S1 beat the wave-front conducts from the proximal to the distal site along a direct short path. However, during the S2 beat, the wave-front is blocked in the vicinity of the proximal site; it travels around a line of block and back towards the proximal site. In the first case (A), when the wave-front arrives at the distal site, the proximal site is still refractory, i.e. ATd < RTp, and the re-entrant wave-front is blocked. In the second case (B), when the wave-front arrives at the distal site the proximal site has regained excitability, i.e. ATd > RTp, enabling the wave-front to conduct and initiate re-entrant ventricular tachycardia (VT). The RVI is the interval between proximal repolarization (RTp) and distal (adjacent) activation (ATd) (shadowed area). Small or negative RVI values will enable a higher probability of re-entry.

• 

  Fig. 2. Spatial distributions of RVI measured before VT (left) and AT maps during VT (right) from representative patients with ARVC (A), BrS (B) and focal VT (C), demonstrating co-localisation of regions of lowest RVI (red) with earliest endocardial activation during VT (red) apart from in C.

• 

  Fig. 3. (A) The distance between the region of lowest RVI and the region of earliest activation during VT, Dmin, was significantly lower in BrS and ARVC than in focal VT. (B) BrS/ARVC patients with clinical VT events had lower Global-RVI, RVIG, than those without VT or than patients with focal VT. (C) Patients with ARVC/BrS with inducible VT had significantly lower RVIG than those with no VT or than patients with focal VT. Mean ± standard deviation of data distribution are reported above each box, while the number of patients per group is reported on the bottom of the panel.

• 

  Supplementary Fig. 1. Calculation of RVI. Left: Neighbouring sites to an electrogram recording are identified within a pre-defined radius in a representative ventricular mesh from the clinical system. Proximal site i is shown in red, while distal neighbouring sites j are shown in black. Right: RVI at site i is the minimum difference between local RT at site i (black circle on red signal) and local AT at sites j (black crosses on black signals). RVI is computed in a premature beat following a short coupling S1S2 interval (vertical lines represent pacing artefacts).

• 

  Supplementary Fig. 2A. RV activation maps created with the Ensite array in a BrS patient (subject 6), showing initiation of VT. The RV is shown in posterior-anterior and right lateral views. The time scale of the isochrones is shown on the left of each map. Sampled areas are marked on the isochronal maps and corresponding virtual unipolar electrograms are shown below the map. An S2 beat during a VT stimulation study results in a line of functional block. This results in the initiation of a re-entrant circuit around this persistent line of functional block on the posterior RV endocardium.

• 

  Supplementary Fig. 2B. Pacing artefacts are shown in purple. The first pacing artefact shown is the S2 stimulus; the second would have been for an S3 beat but the stimulus failed to capture as VT had already commenced.

• 

  Supplementary Fig. 3. Receiver operating characteristic (ROC) analysis shows that Global-RVI was able to accurately discriminate (A) between BrS/ARVC patients with and without clinical VT events (Sensitivity = 0.90, Specificity = 0.79, Accuracy = 0.83 for RVIG < −40.8 ms) as well as (B) between BrS/ARVC patients with clinical VT events and patients with focal VT (Sensitivity = 0.80, Specificity = 0.89, Accuracy = 0.84 for RVIG < −43.9 ms).

• 

  Supplementary Fig. 4. RV activation maps during VT initiation, with the point of earliest endocardial activation shown. The unipolar electrogram at this point is displayed, along with the 12 lead ECG. Cases where the 12 lead ECG was available are displayed, with 4 BrS patients and 2 ARVC patients. Case numbers correspond to those detailed in Table S1. The first VT beat has an ECG morphology consistent with an RV origin and the unipolar EGM (with a Q wave) at the site of earliest endocardial activation precedes or is on time with the QRS onset.