TAVR-in-TAVR?

Historically, young patients with aortic stenosis were encouraged to undergo surgical aortic valve replacement with a mechanical valve, in part because the prospect of redo surgery for a degenerated bio-prosthesis is unappealing. This paradigm has shifted, and surgeons willingly implant bioprostheses in younger patients because of the expectation that transcatheter aortic valve replacement (TAVR) should avoid the need for redo surgery in the future.

Given enough time, transcatheter heart valves (THV) will inevitably degenerate. Surgical explantations for failed THV years later can be challenging, requiring extensive aortic endarterectomy and repair (1). TAVR-in-TAVR is an attractive solution but, until now, has largely remained a theoretical strategy without supporting data.

For this reason, we congratulate Landes et al. (2) for compiling a large multicenter experience of TAVR-in-TAVR procedures. Among 63,876 procedures performed at 37 centers in Europe, North America, and the Middle East, 212 (0.3%) were TAVR-in-TAVR, of which 138 were performed >1 year after the index TAVR.

The rate of peri-procedural complications was low. Notably, the 0.9% rate of coronary obstruction was surprisingly low compared to the 2.3% rate for TAVR-in-surgical aortic valve replacement in the VIVID (Valve-in-Valve International Data) registry (3). An important omission here is the denominator. How many patients considered for TAVR-in-TAVR were actually excluded because of unsuitable anatomy? Specifically, how many were excluded because of concern for coronary obstruction? How many instead underwent surgery with explantation of the first THV? How many were managed conservatively? These missing data are critical to understanding whether TAVR-in-TAVR is truly the panacea for failing THV or whether it is only an option for highly selected patients.

We hypothesize that the reported low rate of coronary obstruction is the result of better screening with pre-procedural computed tomography, as well as legacy THV designs and a historically deeper implantation technique. Most failing bioprostheses in this series were first-generation self-expanding valves (37%) or second-generation balloon-expandable valves with short stent frame (24%), which were typically implanted more ventricular. Modern THV are typically implanted more aortic and are thus more likely to cause coronary obstruction during TAVR-in-TAVR by sequestering the entire sinus of Valsalva (4). Even if coronary perfusion is maintained around the 2 layers of THV, coronary access for angiography and intervention is likely to be very difficult in many, if not most, cases. Leaflet modification strategies such as BASILICA (Bioprosthetic or native Aortic Scallop Intentional Laceration to prevent Iatrogenic Coronary Artery obstruction) may only be effective in selected cases for TAVR-in-TAVR (5).

Until we have a better understanding of the screen fail rate, we strongly caution against assuming that TAVR-in-TAVR is a realistic lifetime management strategy for every young patient considering TAVR, particularly with contemporary THV.