Unique topological classification of complex reentrant atrial tachycardias enables optimal ablation strategy

Funding Acknowledgements: Type of funding sources: Public grant(s) – EU funding. Main funding source(s): ERC starting grant - SMARTHEART (Nele Vandersickel) B Reentry is a common underlying mechanism of atrial tachycardia (AT). Reentry loops form around anatomical or functional 'holes' in the electrophysiological substrate and dominate the atrial cycle. Terminating a 'full' loop may cause previously 'suppressed' loops around another hole to become dominant. Interpretation of electroanatomical maps can thus be tedious: a full loop may be easily identified, but predicting suppressed loops is complex. We propose that topological classification may aid identification of suppressed loops in complex AT and improve ablation procedures. O Given a map of reentrant clinical AT, we aim to unambiguously classify the AT and suggest an operator-independent optimal ablation strategy. M The left atrium can be described topologically as a closed surface with three ‘holes’: mitral valve, left and right pulmonary veins (similar for the right atrium). Non-conductive tissue may constitute additional topological holes in the substrate. Anatomical reentry can occur around any of the holes described above. Therefore, we uniquely categorize an AT by identifying the number of holes in combination with the presence of full and suppressed loops. Fig. 1-2 shows the finite number of possible combinations for 3 and 4 holes. An optimal ablation strategy should terminate both full and suppressed loops, by connecting their corresponding holes. Directed graph mapping (DGM) is a method and software implementation to analyze arrhythmia by constructing a directed network from local activation time maps. We extended DGM to perform topological classification of AT. We hypothesize that the full reentry loops will correspond to the cycles detected by DGM in the network, while suppressed loops are characterized by activity propagating around a hole for at least 60% of its circumference. To test our algorithms, we generated > 1000 simulations corresponding to 5 different scenarios of Fig. 1-2 whereby the circumference of the suppressed loop was varied between 50% and 90% by adding slow conductive tissue. We also classified 110 clinical cases of mapped AT retrospectively and compared our suggested ablation strategy with the clinical ablation lines which ended the AT. R Terminating solely the full loops by an ablation line led to the suppressed loops creating a slower AT in 95% of simulations. Further ablation to terminate the suppressed loop ended the AT. Following our suggested ablation strategy directly terminated the AT. The ablation strategy suggested by DGM was topologically equivalent to clinically successful ablations in 90% of clinical cases. In 5 cases, DGM correctly predicted that the chosen ablation lead to a suppressed loop becoming dominant, resulting in slower AT. C We presented a unique topological classification of AT. We showed that suppressed loops play a pivotal role as not ablating them leads to slower AT. Benefits include preventing redo maps, shorter and operator independent procedures, and possibly reducing AT recurrence. [Figure: see text] [Figure: see text]