Emergency Cardiac Surgery Following TAVI

Implications for the Future

Craig R. Smith


Eur Heart J. 2018;39(8):685-686. 

In this issue, the European Registry on Emergent Cardiac Surgery during TAVI (EuRECS-TAVI) presents the world's first large reported experience with emergent cardiac surgery following transfemoral transcatheter aortic valve implantation (TF-TAVI).[1] During the 4-year period from 2013 through 2016, the frequency of TAVI more than doubled in the 79 centres represented, generating a cohort of 27 760 TF-TAVI patients, of whom 211 (0.76%) required emergency cardiac surgical procedures (ECS). Though distinctly infrequent, the need for ECS was highly consequential. Mortality was 34.6% within 72 h, 46% within the hospital stay, and 78% at 1 year. Is such high mortality a consequence of TAVI being the treatment of choice in the elderly and high risk, while surgery still reigns at the young and healthy end of the spectrum? Indeed, the average age of ECS patients was 82 and two-thirds were female; however, the average risk pre-TAVI was solidly in the intermediate range, and only 23% were considered high risk. Is dispersion of TAVI to new, low volume centres making complications more lethal? Perhaps not—low volume centres in EuRECS (annual median 28 TF-TAVI) appeared to perform as well as high volume centres (annual median 116 TF-TAVI), although the data do not reveal how many low volume centres are included.

The most common single event leading to ECS was guidewire perforation of the left ventricle (28%), closely followed by annular rupture (21%). There were no striking differences between valve types, except for more frequent annular rupture with balloon-expanded valves (P < 0.001) and more frequent dissection with mechanically expanded valves. A balloon-expandable valve was a significant risk factor in univariate analysis (odds ratio 2.1) but not in multivariable analysis. This assortment of technical features illustrates a clear opportunity for improvement with advances in experience and device design, which are certain to occur.

It should be noted that this registry is focused entirely on ECS, and thereby ignores peripheral extracorporeal membrane oxygenation (ECMO) and other temporary support that can be life saving but may also have consequences. It seems likely that some of those manoeuvres were used in the 60 patients in whom we are told that ECS was 'considered'. Also outside the focus of this report are peripheral vascular injuries requiring open surgical repair. Those procedures are more common than ECS, and have lower mortality and morbidity.[2,3] This is not to imply that including the contribution of every conceivable salvage manoeuvre will change the overall risk/benefit balance of TAVI significantly, since the consequences of various manoeuvres are included in overall mortality and morbidity. Even so, it illustrates the point that percutaneous interventions in structural heart disease, with TAVI as the example, carry a new level of procedural risks that is not encountered in performance of diagnostic catheterizations and percutaneous coronary interventions (PCIs).

How well does this registry represent the world's experience? The results reported from EuRECS are quite similar to the handful of previous reports[4–6] based on much smaller numbers, suggesting general consistency. Estimating TAVI volume outside the USA over the same 4-year period to be ~163 000, EuRECS represents ~17% of that experience. EuRECS is 'an investigator-initiated independent multicenter observational registry…' The authors are admirably clear in stating that the data are self-reported and not verified. However, it seems reasonable to assume that the investigators who participate in a registry are at least as conscientious and reliable as those who do not participate. Following the same logic, it seems likely that the ~80% of institutions outside EuRECS are unlikely to be substantially better, and more likely to be the same or worse than the minority reported here. If these assumptions are true, institutions in the uncounted majority should consider the EuRECS experience to be best-case, and should take the cautions raised by the authors very seriously.

There has already been dramatic progress in TF-TAVI away from general anesthesia with an intensive care unit (ICU) stay toward much simpler routines, accompanied by healthy debate[7–9] over each element of simplification. On that path, my institution has been insistent on careful risk stratification,[10] aimed at avoiding the horror of dealing with something like annular rupture in an awake patient who cannot be immediately placed on cardiopulmonary bypass. Avoiding situations that might make the procedural mortality reported by EuRECS even higher is very logical, yet it must be acknowledged that most of our risk stratification tools are better at identifying patients who will tolerate a structural heart complication more or less well than at predicting the complication.

Finally, the authors are silent on whether these results call for restricting TAVI to institutions that have on-site cardiac surgery. Although that is a grossly obvious inference to this cardiac surgeon, such restrictions are already the subject of debate that is certain to intensify. Opponents of restriction rely on two primary arguments—access and an analogy to PCI. The problem with 'access' is that it can have many different meanings. It becomes synonymous with 'convenience' when argued to be a patient's 'right' to a TAVI programme in every local hospital. The results from EuRECS suggest that patients under that scenario would be acquiring, along with convenience, access to unjustifiable risk. 'Access' should not be about providing every interventional cardiologist with access to new technology, and device manufacturers with access to that book of business. Cynicism aside, it is at the very least debatable whether ensuring access to expensive twilight-of-life procedures serves a purpose as noble as ensuring access to vaccinations, to treatment of infectious diseases, and to other elements of basic healthcare. Myriad ambiguities make 'access' the wrong flag to wave when trying to rationalize TAVI without on-site cardiac surgery.

In applying the PCI analogy, opponents can indeed argue that PCI has been shown to be acceptably safe in meticulous and highly selective centres without on-site cardiac surgery.[11,12] Hower, that is a poor comparison. The frequency of ECS following PCI (0.2–0.6%)[13–15] is not orders of magnitude lower than what is presented here for TAVI, but it is much less lethal. PCI in the setting of acute coronary syndromes has been associated with a higher frequency of ECS (2%),[13] but even in that setting the mortality (1–20%)[13,14] does not approach what EuRECS is reporting for ECS following TAVI. Much can be done to support global ischaemia complicating PCI while making preparations for definitive treatment. The most frequent complications of structural heart interventions requiring ECS are highly structural, salvage depends on rapid, skilled cardiac surgery, and the mortality is much higher. Even as TAVI marches through the Guidelines into younger and healthier populations, structural heart device complications will remain challenging. For these reasons, TAVI should aspire to, but will not achieve, the level of safety associated with PCI.