Transmurality of post-myocardial infarction scar and the cycle length of ventricular tachycardia

Abstract Background Despite advances in the treatment of ST-elevation myocardial infarction (MI) with primary percutaneous coronary intervention, the risk of remote ventricular arrhythmias (VA) and sudden death remains significant. Reperfused MI may support faster ventricular tachycardia (VT)1–3, potentially due to smaller isthmuses. While cardiac MRI has an expanding role in risk assessment and planning VT ablation, the link between scar characteristics and VT cycle length is poorly defined. Purpose This study compared VT inducibility and cycle length in vivo in rabbits with MI or ischemia-reperfusion (IR) to controls, with structural and functional characterisation using high-resolution ex vivo imaging. Methods Adult male New Zealand White rabbits underwent either permanent coronary ligation (MI, n=16), temporary ligation for 60 minutes (IR, n=8), or no intervention (controls, n=7). In vivo PES was perfomed on a subset (n=20). Langendorff-perfused hearts were imaged using optical mapping (OM) with RH237 dye to identify late activating sites in sinus rhythm, and conduction discontinuities quantified by maximal phase difference (PDmax) in activation time4. Sustained VT was induced ex vivo through burst pacing, with flecainide (1µM) if required. Fixed MI hearts were scanned with 7T MRI for high-resolution T1-weighted imaging (125µm3) and histology with Masson’s Trichrome. OM and histology data were mapped onto 3D MRI-rendered structures. Scar volume, surface area and transmurality were calculated. Results Sustained VT (>30 seconds) in vivo was induced in 29% (n=2/7) of MI hearts, 50% (n=3/6) of I-R hearts and 0/7 control hearts (p=0.11). MI scars were more transmural (84±4% vs 44±13%, p<0.0001) and extensive (38±17% vs 24±11%, p=0.043) than IR, which had a preserved epicardial rim. Scar volume, surface area or transmurality did not impact VA inducibility. VT cycle length was positively correlated with scar transmurality (R2=0.97, p=0.003), while no correlation was found with scar volume or surface area. IR hearts often had late LV apical activation in sinus rhythm (6/8 vs. 2/16 MI hearts, p=0.0047 and 0/7 controls, p=0.0005). Apical conduction times were longer in MI/IR hearts with evidence of an epicardial rim (n=9/24: 16ms±2 vs. 6ms±1, p <0.0001). Late activation was not associated with VA inducibility, whereas conduction discontinuity (PDmax) was increased in hearts with sustained VA (2.6ms ± 0.9 vs 1.4ms ± 0.7; p=0.028). Sustained VT was inducible ex vivo in 16/16 MI, 7/8 IR hearts and 1/5 of control hearts (MI/IR vs. controls, p<0.0001). During VT, lines of conduction block aligned with regions of late activation in sinus rhythm or discontinuous conduction. Conclusions Scar transmurality strongly correlates with VT cycle length, with less transmural scars linked to faster VT. Critical VT substrate in hearts with subendocardial scars may be associated with an epicardial rim, thus standard endocardial ablation may not prevent VT recurrence.