Discussion:
In our study, we found that in patients undergoing SAVR with preoperative MDCT imaging; (1) There was a strong correlation between annular measurements on MDCT and actual SAVR size implanted; 2) Using pre-procedural virtual valve implantation, all patients with AI were predicted to be low risk for ViV TAVR coronary obstruction; (3) In patients with AS, pre-procedural MDCT with virtual valve implantation found that 38% of patients with tricuspid AS would be at high risk, while no patients with bicuspid AS were at high risk.
To our knowledge, this is the first paper addressing the issue of whether a patient undergoing SAVR will be a candidate for ViV TAVR in the future based on pre-procedural MDCT. As our surgical and transcatheter techniques evolve, focus has shifted to lifetime management of valvular disease. Patients with bioprosthetic valve failure are frequently at high surgical risk for reoperation, which is why ViV TAVR has become a valuable option in such patients. (1) Coronary artery obstruction more commonly occurs with ViV TAVR, and is a potentially fatal complication that portends a mortality rate greater than 50%. (6) Patients at high risk for coronary obstruction during ViV TAVR still have options, however these options come at a higher risk. Chimney stenting is associated with higher risk of procedural and delayed clinical events. (8) Which prompted the advent of BASILICA (Bioprosthetic or native Aortic Scallop Intentional Laceration to prevent Coronary Artery obstruction). However, there are a significant percentage of patients that are not candidates for BASILICA, and the procedure may result in higher rates of procedural complications including stroke. (3) Therefore, if we can classify patients who will be high risk for future transcatheter procedures, it may factor into the shared decision making between the patient and heart team. Options may then include; 1) TAVR as primary strategy followed by SAVR for failure, 2) SAVR with surgical manipulation (e.g.- coronary re-implant, root enlargement) to decrease subsequent risk for coronary obstruction, 3) Standard SAVR (with the awareness that advanced adjunctive techniques will likely be necessary or transcatheter options limited in the future).
Though the majority of patients (74%) were correctly sized with MDCT, the rest of the patients (26%) were oversized compared to actual SAVR implanted. One would expect that virtual implantation of a larger valve would increase the chance of falsely predicting high risk anatomy for ViV TAVR. Interestingly, despite implanting a larger virtual valve in almost ¼ of patients, MDCT did not predict any patients with AI to be at high risk for coronary obstruction. These findings suggest that outside of MDCT for root sizing/coronary evaluation, preoperative MDCT is less helpful for patients with AI.
MDCT predicted the correct valve size implanted in all but one patient with AS (94%). There was a strong correlation between MDCT annular measurements and actual SAVR sizing. In patients with bicuspid AS, all patients that underwent virtual valve implantation were predicted to have low risk anatomy for ViV TAVR. Perhaps this is because root sizes are generally larger in bicuspid AS patients. With virtual valve implantation, 38% of patients with tricuspid AS were high risk for ViV TAVR. This warrants a bigger question; would preoperative MDCT change management in patients with tricuspid AS with plan for SAVR? Future studies are needed to 1) compare preoperative virtual valve implantation of the SAVR to postoperative MDCT and 2) determine if preoperative MDCT can safely change the lifetime management in tricuspid AS patients.
As the heart team plans for the treatment of each individual’s aortic valve pathology, it is important to determine what size surgical bioprosthetic valve can be implanted. Surgical prosthesis with small internal orifice diameters may predispose patients to transcatheter heart valve under-expansion at the time of ViV TAVR. Smaller sized surgical valves are associated with higher post-procedural gradients and increased mortality in patients undergoing ViV TAVR. (4, 5) The ability to predict SAVR sizing can therefore be useful to identify patients that require smaller valves. For example, if preoperative workup reveals that the annulus will only accommodate a 19mm valve, the heart team can then discuss the most suitable option for the lifetime management of each case. Options may include; standard SAVR, SAVR with root enlargement, implantation of a prosthetic valve with an expandable frame (i.e.- Inspiris, Edwards Lifescience, Irvine), mechanical valve, or the use of TAVR instead of SAVR. This knowledge is especially useful in patients with AS, who tend to have smaller annuli than patients with AI (annulus area; 554.77mm2 ± 123.2 vs 474.18mm2± 135.41, P=0.049).
Among our studies limitations was the lack of patients with preoperative MDCT. This is reflective of current standard practice, which does not necessitate preoperative MDCT. Larger numbers are needed to evaluate what percentage of AS patients would be high risk based on preoperative MDCT. We also concede that virtual valve implantation has never proven to predict feasibility for future ViV TAVR procedures. Though outer stent diameters may be the same between valve manufacturers, valve sizes may have different inner stent diameters, and therefore a patient may end up with a different sized valve depending on the manufacturer. If a smaller or larger prosthesis is implanted, it will skew measurements and virtual valve implantation may be less accurate. Additionally, valves may be canted one way or implanted at different depths, therefore studies will need to address a pre-SAVR virtual valve implantation with MDCT to post-SAVR measurements on MDCT. Unfortunately, it is not standard protocol to obtain post-MDCT, so we were not able to investigate this. However, the majority of valve implantations on pre-CT were correctly sized for actual intraoperative measurements (or within 1 size at least), so this should not factor in largely to the limitations.
We believe that our study addresses a pressing need to investigate the utility of MDCT to help guide management decisions for patients undergoing isolated SAVR. Though the surge of TAVR has led to a reduction in isolated SAVR volume, SAVR is still an option for low-risk patients. (1, 9) It is the low risk/younger patients that are more likely to need a valve re-intervention during their lifetime. Hence, shared decision making between the patient and heart team is essential in patients with high risk anatomy to discuss all options for lifetime management.