The saying goes, “what is too good cannot be entirely good.” The advantages of TAVI have been widely celebrated, reaching unprecedented implantation rates worldwide. This unbridled success stems from multiple factors, which, to this day, remain unexamined independently and objectively. Perhaps the truth, after all, is what hurts the most—and interests the least.
In this tendency towards a darker turn, the work before us aims to shed light and reveal the truth. And who better than registries for this purpose? In a globalized world with highly advanced and computerized centers, one can register nearly everything. This is the case for the EXPLANT-TAVR registry, a multicenter international study with the participation of 44 renowned institutions in the Western world. However, although Europe (primarily Germany, as well as France and Italy) and the USA remain the international leaders in this area, the absence of data from Asian countries and emerging cardiac intervention and surgery leaders such as African and, notably, South American countries is noticeable.
The registry focuses exclusively on TAVI device explants occurring during post-implant follow-up. The reasons for these explants included various indications categorized as infectious (prosthetic endocarditis) and non-infectious (structural degeneration, prosthesis-patient mismatch, paravalvular leak, and thrombosis). These categories align precisely with VARC criteria, and, for non-infectious cases, a committee (a true Heart Team) determined whether patients were surgical candidates rather than percutaneous solutions.
The study conducted straightforward yet insightful comparisons, providing data that, at the very least, will capture our attention. In total, 372 patients were included in the study, with explants performed between 2011 and 2022. These comprised 188 non-infectious cases and 184 endocarditis cases. A propensity analysis was performed to enable comparison, using the endocarditis group to compare outcomes based on whether it occurred before or after 18 months post-implantation.
Among the non-infectious explant causes, the most common were structural dysfunction (55.9%), paravalvular leak (43%), and prosthesis-patient mismatch (20.2%). Two of these factors are often regarded in the TAVI world as at least comparable, if not superior, to surgical bioprostheses. Of course, without knowing the proportion that these 372 explants represent relative to implant volumes—averaging 34 cases per year (<1 per center)—they may not seem very frequent (especially given the implant volume of the participating centers).
The authors found that patients requiring explantation due to endocarditis were older than those with dysfunction-related causes. The postoperative outcomes for endocarditis patients were, as expected, worse, with extended stays in both the intensive care unit and the hospital. Endocarditis significantly affected balloon-expandable prostheses more (61.5% vs. 38.5%; p < 0.001), while non-infectious dysfunction more commonly affected self-expanding devices (58.3% vs. 41.7%; p < 0.001). Some complication rates were notably high for explant procedures, such as the need for pacemaker implantation (18-19%), stroke (higher for endocarditis patients at 8.6% versus non-infectious dysfunction at 2.9%; p = 0.032), and severe bleeding (11-12%). All of this resulted in mortality rates, after group adjustment, of 15-17% at 30 days and 32-34% at one year, with no significant differences between the two etiologies. An interesting finding was the time from implantation to the adverse event, which was markedly shorter for endocarditis (mean 10.4 months, with a concerning accumulation of early cases) compared to structural dysfunction (mean 19.9 months, with a more extended time of occurrence); log-rank p < 0.001. Regarding the analysis of survival by early or late onset of endocarditis (cut-off at 18 months), results were surprisingly better for early cases, with a one-year mortality of 19.3% compared to 26% for late explants, p = 0.038. This finding seems to obscure the aggressiveness of endocarditis as the periprocedural mortality prevails due to the ease of device explantation, the primary Achilles’ heel of this complication, which we will analyze next.
This significant work sets some of the reference points for expected outcomes in addressing such complications. The authors conclude that the EXPLANT-TAVR registry confirms that TAVI explantation is a high-risk procedure for mortality and postoperative complications, particularly when endocarditis is the etiology.
COMMENTARY:
This registry confirms nearly identical results regarding the high mortality of patients undergoing cardiac surgery after receiving prior TAVI, comparable to the most relevant studies available to date and analyzed here on this blog (the STS database registry from 2012 to 2023 and the Michigan and STS Transcatheter Valve Therapy registries from 2012 to 2019). The most significant novelty of the EXPLANT-TAVR registry is that it demonstrates this high mortality even in cases without endocarditis, confirming the poor prognosis for these patients in the medium term.
Some arguments and critiques of the study have been made before: limited sample size for subgroup analysis, lack of representation of activity in less “elite” settings, lack of knowledge of actual rates at which these complications occur (which would allow them to be contextualized for clinical decision-making), and the absence of surgical technique details.
On the last point, I would like to offer a reflection on the technique. Given the current heterogeneity and the likely need to be prepared for root and coronary ostia injury, experience from some groups seems to be shaping the technique to face this “disaster” scenario. First, it is essential to differentiate between the explant of balloon-expandable and self-expanding prostheses. The primary difference lies in the aortotomy approach, which is performed at a similar height to that of a standard aortic valve replacement in the former. In the latter case, accessing the aortic valve posteriorly would be impeded if the aortotomy were performed above the stent level. This requires a superior aortic incision near the “fat whisker” of the aorta, releasing adhesions of the upper aortotomy to access the root. For both TAVI types, the technique of dividing the device by cutting the stent mesh seems to prevail. Removing the TAVI leaflets, leaving only the stent, facilitates the use of cutting tools on the stent itself. Various instruments have been suggested: powerful scissors, such as Mayo scissors, wire-cutting shears for sternotomy cerclages, etc. However, a specific type of scissor used in dentistry for cutting wires in dental bridges and orthodontic brackets could be highly useful and should be included in our instrument kits. Designed for steel wire cutting, they allow stent cutting cell by cell. Another useful instrument can be an endarterectomy spatula to separate the stent from the aortic wall and then from the native valve leaflets. Experience with Perceval® explantation has shown that it tends to coil up once sectioned. However, this does not seem to occur with TAVIs, which demonstrate greater radial force, necessitating the use of ligatures or sutures on the stent to fold it as it is separated from the aortic wall, or resorting to the martensitic state properties of nitinol when cooled (tips and tricks recently analyzed on our blog).
A second reflection stems from a section of the study that highlights concerns regarding TAVI-related endocarditis. Although the study does not provide microbiological data, two causal agents are noted in two specific periods. The first, staphylococci, relates to periprocedural contamination. Given that TAVI involves crimping and percutaneous approaches, contamination may occur more frequently. The authors call for extreme antiseptic measures, noting the high standards of the centers included in the registry. Having visited some of these centers, I can attest to the scrupulousness of the surgeons and especially interventional cardiologists compared to what we are accustomed to in our units. Indeed, the authors report a lower rate of staphylococcal endocarditis in surgical prostheses than in TAVIs, although they provide no data. In our setting, underdiagnosis and underreporting of this complication likely occur, as these patients are often deemed non-surgical. The second agent is enterococci, particularly faecalis, associated with a higher incidence of urinary tract infections or intestinal translocation in an elderly population. This is complemented by a call for antibiotic prophylaxis precautions in TAVI recipients and aggressive infection treatment.
Ultimately, the saga continues with the analysis of a complication as devastating as TAVI explantation. While this study shows some expected aspects regarding outcomes, it is still a field that combines lack of experience and sporadic presentation, lending a heroic undertone. Unfortunately, in surgery, good outcomes stem from standardization, and heroics are best left to the cinema. We must minimize the occurrence of these complications through optimal patient selection, bearing in mind that, although infrequent, they are not impossible. Surgeons are well aware of the outcomes of a first reintervention on a bioprosthesis, which are far less severe due to lower risk of injury to the aortic root structures. Combating the radial force of the stent is key. And since these procedures will likely increase, until then, may the force be with us… to keep cutting stent cells.
REFERENCE:
Marin-Cuartas M, Tang GHL, Kiefer P, Fukuhara S, Lange R, Harrington KB, et al.; EXPLANT-TAVR Investigators. Transcatheter heart valve explant with infective endocarditis-associated prosthesis failure and outcomes: the EXPLANT-TAVR international registry. Eur Heart J. 2024 May 31:ehae292. doi: 10.1093/eurheartj/ehae292.
