Leak to Shock
Dr. Kent SO
An 81-year-old male with hypertension and chronic kidney disease presented with heart failure symptom. Echocardiogram showed severe aortic stenosis and mild aortic regurgitation with peak pressure gradient 92mmHg and mean pressure gradient 52mmHg. The calculated aortic valve area was 0.9cm2. His STS score for SAVR was 8.3% for Predicted Risk of Mortality.
CT showed occluded right external iliac artery, diseased left external iliac artery and 4.7cm infra-renal aortic aneurysm. The annulus perimeter and perimeter derived mean diameter are 78mm and 24.8mm respectively (22 x 28mm), with severely calcified aortic leaflets. The planned procedure was TAVI by direct aortic approach using 29mm Evolut R valve under trans-esophageal echo (TEE) guidance.
Direct Aortic TAVI was performed resulted in good implantation depth (5mm over non-coronary cusp and left coronary cusp-LCC), but severe paravalvular leak (PVL) (AR index 20.9) despite post-dilatation using a 24mm Nucleus balloon (Figure 1, Movie 1). Further post-dilatation using 25mm Nucleus balloon (recommended size limit) resulted in no change in severity of PVL. Yet in view of stable haemodynamic and contrast limit, we decided to keep observe at that juncture and plan stage percutaneous closure if needed.
However, on the night after the procedure, patient developed refractory cardiogenic shock while attempting to wean off ventilator. Intensivist used dopamine and noradrenaline, which possibly further aggravated the PVL. Repeated echocardiogram ruled out pericardial effusion but severe PVL. Patient was brought to cath lab for emergency reassessment of valve position and percutaneous PVL closure. TEE showed a crescentic shape leak of 18mm x 4.6mm arising from LCC likely due to unevenly distributed native valve calcium.
After obtaining the optimal projection that separate the prosthetic valve and the leak, we crossed the prosthetic valve at high stent struct and then the PVL using an angled Terumo wire under MP catheter. A Confida wire was then exchanged into LV as anchor wire. 8-French Shuttle sheath was then inserted through the Confida wire (Figure 2). Two 8mm AVP-II plugs were inserted but TEE showed residual 8mm x 4.6mm leak. We decided to leave one 8mm AVP-II plug in-situ and used another AVP-III plug to optimally seal the leak. However, while manipulating, the anchor wire was lost (Movie 2). Rewiring was challenging with one AVP-II device in-situ, and finally was successful with repeated attempts. However, this time 8-French and 7-French shuttle sheaths could not be inserted due to poor support. A 6-French shuttle sheath was finally inserted, which unfortunately did not allow the use of suitable AVP-III plugs. The largest 6-French compatible AVP-II plug (14mm) was then implanted and resulted in mild PVL only (Figure 3, Movie 3, Movie 4). The AR index improved to 40 with diastolic BP improved from 39mmHg to 60mmHg. Patient was then extubated and discharged 10 days after the index procedure.
The mechanisms of PVL after TAVI can be broadly divided into unopposed prosthesis due to calcification, high or low implantation or prosthesis under-sizing. Different strategies need to be adopted to tackle the specific mechanism. Our case illustrated that percutaneous PVL closure is feasible, yet challenging, in treating PVL due to unopposed prosthesis, despite adequate post-dilatation. Specific techniques used in our procedure included high stent struct crossing and anchor wire technique for multiple devices use.