Patient-specific computer modelling to predict anatomical risk factors preventing post transcatheter aortic valve implantation coronary re-access in bicuspid aortic valve; a modelling study

Abstract Background Post Transcatheter aortic valve implantation (TAVI) coronary re-access is predicted to become more common as TAVI expands into lower risk cohorts. The use of a self-expanding valve (SEV) has been recognised as an independent predictor for unsuccessful coronary cannulation. Other predictors include reduced transcatheter heart valve (THV) to coronary distance and small sino-tubular junction. Despite this however a patient cohort remains in which the use of a SEV is preferable due to anatomical complexity or extensive calcification. Purpose Using patient-specific computer modelling, we aimed to assess how the deployment of different sized THVs and at different depths affected the distance from valve to coronary and valve to sino-tubular junction in bicuspid aortic valve cases. Methods In this modelling study we have used pre-procedural, CT-derived, patient-specific finite element analysis computer models to predict the eventual valve result. We studied two anatomical measures for both right and left coronary sinuses. The distance from coronary ostium to nearest structure, either THV or displaced native leaflet – Sinus of Valsalva (SoV) free space. The second distance was from the STJ to THV – STJ free space. THVs were modelled at both high and medium implant positions. THV size was based upon annular dimensions. Where the annular dimensions fell near an anatomical grey-zone, a second size THV was modelled. For each patient and measure, a mean of all distances was calculated, in addition to both minimum and maximum distances. Results The free space SOV left main coronary artery (LMCA) mean was 6.39 mm (4.87-7.91 mm). The difference between maximum and minimum was 0.9461 mm (0.54–1.36 mm), p<0.01. The largest patient-specific difference between minimum and maximum was 2.41 mm, using the same THV size but at different depths. The free space SOV right coronary artery (RCA) mean was 5.73 mm (4.17-7.30 mm). The difference between maximum and minimum was 0.789 mm (0.49-1.09 mm), p<0.01. The STJ free space LMCA mean was 4.3 mm (2.35–6.27 mm), The difference between maximum and minimum was 0.99 mm (0.63-1.35 mm), p<0.01. The STJ free space RCA mean was 3.76 mm (2.2-5.32 mm).The difference between maximum and minimum was 1.18 mm (0.9-1.56 mm), p<0.01. The largest patient-specific difference between minimum and maximum was 2.02 mm, using the same valve size but at different depths. Data presented as mean (95% confidence intervals), n=16 patients/48 models. Conclusions This modelling study has indicated a significant difference in all four measured parameters when different THVs are modelled at different depths. Note some of the largest differences are seen with the same size THV but at a different depth. We would highlight the small mean effect size seen across the results, however in cases where the difference is >2mm, this could have an effect on coronary re-access and warrants further investigation.