Can you share some of the background of your study?

Karim Al-Azizi, MD: For transcatheter aortic valve replacement (TAVR), we are well beyond getting patients safe and alive off the table. Obviously, that remains the priority, but it used to be a considerable factor in the past. We are now thinking about the second TAVR intervention. The big topic that everyone is discussing is the lifetime management of aortic valve stenosis. We shouldn’t treat an 85-year-old like a 65-year-old. If a 65-year-old patient needs a TAVR, and his or her valve area size is for a very small valve, it may be fine now. But the question is, in 10 to 15 years, when this patient needs a second valve, do I have enough room, technically, to put in a second valve safely, without endangering any of the neighboring structures, including coronary arteries?

Echocardiography is the gold standard in assessing valves. We know that the Evolut valve (Medtronic) or a supra-annular valve more generally has a better hemodynamic performance in smaller aortic valves as compared to balloon-expandable prosthesis in a smaller aortic valve. But does better hemodynamic performance mean that all the patients with smaller balloon-expandable valves or intra-annular valves (especially now that we have the Portico valve [Abbott Vascular]), have worse outcomes?

A 23 mm Evolut R valve may appear to have a larger effective orifice area (EOA), or a larger valve area, than the Sapien 3 (Edwards Lifesciences), but it is actually a similar size from a valve standpoint. For example, different engines may operate the same way, but have different sizes of cylinders. The difference for TAVR valves is the hemodynamic performance and how we should look at these hemodynamics is one question. Different valves have a different hemodynamic “fingerprint”. In general, however, when deciding on what device to choose for our patients, while we should include hemodynamics in the equation, it is not the primary reason to choose a particular platform. Hemodynamics in TAVR is very dependent on several things. Loading conditions and sedation may affect flow and gradients.

Can you tell us about your study?

Karim Al-Azizi, MD: We evaluated the long-term outcomes, mortality at 5 years, comparing patients with smaller indexed EOA and hence considered to have severe prosthesis-patient mismatch, compared to patients that don’t have severe prosthesis-patient mismatch, in a single healthcare system with over 3000 patients. We are currently running our sub-analyses comparing the balloon versus the self-expanding platform to evaluate the rates of heart failure rehospitalization and quality of life metrics around transcatheter valve therapy. These results will be presented at a future conference.

We sought to link valve areas being larger or smaller to outcomes, so we looked at the hemodynamic performance of the transcatheter heart valve (THV) and followed these patients out to 5 years to see what the association might be with having a smaller valve in relationship to the patient’s body surface area (BSA), which is called prosthesis-patient mismatch. Prosthesis-patient mismatch is defined by having an effective orifice area (EOA) that is too small in relation to patient’s BSA, therefore often resulting in high residual transvalvular pressure gradients.

The study is retrospective and from a single healthcare system, Baylor Scott and White in Texas. The primary outcome was all-cause mortality at 5 years. We pulled all the commercial patients that have been treated, beyond the pivotal studies, for both native and valve-in-valve TAVR. Between 2012 and 2021, we identified over 3100 patients, and these were all-comers. About 43% were females, about 36% had a body mass index >30, and nearly 74% had been treated with a balloon-expandable platform with all its generations, the Sapien, Sapien XT, and Sapien 3. The Medtronic self-expanding valve was used as well, with its three iterations of CoreValve, Evolut Pro, and Evolut R. We followed these patients longitudinally out to 5 years and identified their mortality. We were able to complete that data collection on all patients.

We identified which patients, by definition, had prosthesis-patient mismatch, based on certain criteria and echocardiographic criteria that are well known in the literature.

The concern around prosthesis-patient mismatch comes from the surgical literature. The surgical aortic valve replacement signal of prosthesis-patient mismatch has been linked to worse outcomes and to mortality. Investigators  have attempted to apply this idea to the transcatheter aortic therapy side, but not everything in the human body translates equally, so the theory was up for question. Several studies and registries thus far have resulted in mixed signals. Some did not make a difference and some had a signal leaning towards an influence for prosthesis-patient mismatch but only went out to one year. Our study is unique in that it goes out to at least 5 years and includes 3100 patients that are all comers. More importantly, the other studies are more of a sub-analysis of a standard study that had a core lab, whereas our study includes all comers, is across 6 centers in Texas, and includes the echoreaders from several institutions.

The PARTNER 2 substudy performed by Ternacle and Pibarot et al, evaluated the same idea in patients undergoing surgical aortic valve replacement (SAVR) and TAVR, and the signal of worse outcomes was more obvious in SAVR patients than TAVR patients. Not to mention that its incidence in TAVR patients was much lower. Our study confirmed similar outcomes in a real-world population.

We found that measuring prosthesis-patient mismatch based on the gradients identifies more patients that fall under that definition, as compared to when predicting a prosthesis-patient mismatch. A measured prosthesis-patient mismatch can be compared to a predicted prosthesis-patient mismatch. “Predicted” is achieved by dividing the reference value of the EOA for the model and size of that prosthetic valve by the BSA of the patient. Our study identified the incidence of predictive prosthesis-patient mismatch to be much smaller, at 1.32%, as compared to measured prosthesis-patient mismatch, with an incidence of 7.19%, in the entire cohort of 3100 patients.

How does that translate to outcomes? Following our study patients out to 5 years, we found no significant difference in outcomes. There was no signal that patients who had been identified with severe prosthesis-patient mismatch — whether measured or predicted — died sooner, as compared to patients that did not have severe prosthesis-patient mismatch. At 5 years in this cohort of patients, we did not see that definition that people are worried about translate into hard outcomes and patients dying sooner.

What should we take away from this study result?

Karim Al-Azizi, MD: We should carefully evaluate patients for TAVR and decide on the THV based on many factors and not solely the fear of prosthesis-patient mismatch. We need to incorporate it into our thinking about the lifetime management of aortic valve stenosis. Will I have enough room to put in another valve in 10 to 15 years? Would this current valve give me prosthesis-patient mismatch that I am not seeing? Because now, we are stacking valves inside valves, like Russian dolls. If the answer is, the patient does not have enough room for a second valve in 10 years, and he or she is 60 years old, I will be having a conversation with that patient with my heart team around considering heart surgery and recommending root enlargement. If a patient has significant coronary artery disease, coronary access remains a priority, as these patients may need future interventions.

What are the needs for future study?

Karim Al-Azizi, MD: The patients that were in this study are mostly high- and intermediate-risk patients. Only a third were low-risk patients. Treating low-risk patients has come about just in the last few years. We will eventually be interested in looking at prosthesis-patient mismatch in low-risk patients, because these patients continue to be followed longitudinally, and much further out, we may see a different signal. There is a lot more to the question that we asked in our study, but at least out to 5 years, those patients that fell under the definition of severe prosthesis-patient mismatch did not have a significantly different outcome from those patients who did not meet that definition, whether measured by echocardiography or predicted, in native or valve-in-valve TAVR.

Find more: 
TVT Newsroom (With a Special Focus on TAVR)

Find Dr. Al-Azizi’s slides here.