TAVI Outcomes Linked to LV Recovery in Patients With Severe Dysfunction

Patients with severe left ventricular dysfunction undergoing TAVI for aortic stenosis who experience an improvement in their ventricle’s function achieve mid-term mortality similar to that seen among those without severe LV dysfunction prior to the procedure, according to international registry data.

Survival was, in fact, better at 3 years for those whose left ventricle completely normalized after TAVI when compared to TAVI-treated patients without significant LV dysfunction at baseline, report investigators. Importantly, the researchers also identified several key variables that help predict which patients with severely impaired LV function prior to TAVI would be expected to show an improvement in cardiac function.

“It's a dilemma that we face on a routine basis when we assess patients with our heart team,” lead investigator Guy Witberg, MD, MPH (Rabin Medical Center, Petah Tikva, Israel), told TCTMD. “Obviously, when you have a patient whose baseline LV function is worse, then you're a bit more wary for the procedure itself. There's always the dilemma of whether or not the patient’s symptoms are because of aortic valvular or myocardial disease.”

Witberg said it’s hoped the new findings can inform discussions about TAVI’s expected benefits in those with severe LV dysfunction. While operators can be confident that a patient with a healthy LVEF and classic symptoms of aortic stenosis will fare well with TAVI, that’s not the case in those with an LVEF ≤ 30%. In such instances, if operators aren’t certain the LV will improve following TAVI, they may end up performing a futile procedure, he said.

At present, there are few data to guide clinicians in patients with aortic stenosis and severely impaired LV function. Data from the PARTNER trials showed that roughly one-third of patients with impaired LVEF (< 50%) had an improvement in function and that this was associated with better survival at 5 years. Similar data were seen in the CoreValve trials. However, Witberg said the definitions of “impairment” used in those analyses were quite liberal, noting that LVEF ≤ 30% is considered the more acceptable definition of dysfunction. 

AMTRAC Registry

For this study, which was published recently in EuroIntervention, investigators turned to the Aortic+Mitral Transcatheter (AMTRAC) Registry, which includes 17 TAVI centers in Europe and Israel. In total, 10,872 patients were stratified according to baseline LV function (LVEF > or ≤ 30%) and ventricular response following TAVI, with 914 (8.4%) having impaired LV function at baseline. The response was characterized as no LV recovery, LV recovery (EF increased ≥ 10%), and LV normalization (EF ≥ 50% after TAVI).

Of those with LVEF ≤ 30% prior to TAVI, 59.5% had an improvement in function, with 26.7% experiencing a complete normalization of the left ventricle.

The 3-year mortality rate for those without LV dysfunction at baseline was 29.4%. Compared with this group, those with LV dysfunction at baseline who did not have any improvement in LVEF after TAVI had a higher relative risk of death at 3 years (adjusted HR 1.32; 95% CI 1.16-1.71). On the other hand, those who had LV recovery after TAVI had a similar risk of mortality compared with those without LV dysfunction at baseline, while those who experienced a complete normalization of LVEF had a lower risk of death compared to those with baseline LVEF > 30% (HR 0.80; 95% CI 0.56-0.98). 

Investigators also showed that those who had similar LV function after TAVI had similar rates of death at 3 years no matter their baseline LVEF.

In a multivariate regression model, no prior MI, estimated glomerular filtration rate > 60 mL/min, and mean aortic-valve gradient (AVG; expressed continuously or as a binary variable) were all associated with a greater likelihood of LV recovery post-TAVI. Those with impaired LV function at baseline who didn’t have a prior MI and those with mean AVG > 40 mm Hg had mortality outcomes equivalent to those without severe LV dysfunction. 

Assessing Likelihood of Good Outcomes

To TCTMD, Witberg said that TAVI-eligible patients with significant LV dysfunction can be placed on a spectrum, with those with ischemic cardiomyopathy on one end and those without any obstructive coronary heart disease on the other.

Many TAVI candidates, roughly 40% to 50%, don’t have any coronary disease or only nonsignificant disease, Witberg said. “For those patients, if they have significant LV dysfunction, then you know that the most likely explanation [for LV dysfunction] is the valvular disease.” In these patients, there is greater likelihood the left ventricle will recover following TAVI and outcomes will be favorable, he added.

There's always the dilemma of whether or not the patient’s symptoms are because of aortic valvular or myocardial disease. Guy Witberg

Sammy Elmariah, MD, MPH (University of California San Francisco), who has studied the impact of baseline LVEF in the PARTNER trials, said the new findings are important for risk-stratifying patients with aortic stenosis being considered for aortic valve replacement. However, they should not be interpreted to suggest that AVR should not be offered to patients with LV dysfunction who are predicted to have their LVEF remain low following TAVI or surgery.

“This concept is similar to prior studies showing that clinical outcomes are poor in patients without contractile reserve during dobutamine stress echo,” Elmariah told TCTMD. “But those poor outcomes post-AVR are still better than the comparative outcomes if AVR was withheld. So, AVR helps in these patients, just not as much.”

From a clinical perspective, Witberg said it’s important for physicians to understand the type of patient with aortic stenosis and LV dysfunction they are treating. If there is a strong chance the left ventricle can improve or normalize with TAVI, then there’s no reason not to treat. In patients for whom there is a higher likelihood that TAVI might be futile, such as those with a prior MI or poor renal function, it becomes ever more important to consider the risks of the procedure, as well as other variables that can also impact ventricular recovery. 

“That’s the point where your clinical judgement needs to come into account,” said Witberg. “Sometimes we deal with patients that we don’t have very good solutions for. In those cases, I think it is relevant and it is reasonable to offer a patient an intervention after you make sure that you know the risks and benefits and that they know the risks and benefits. [You also] need to have some discussion with the family, as well.”

Witberg noted that 40% of patients in their analysis had no improvement in LV function after TAVI. One gap in the data is how patients with severe LV dysfunction and aortic stenosis treated with medical therapy fare if compared with those who undergo TAVI, he noted. “We know from the coronary field that patients with severe ischemic cardiomyopathy don’t benefit from coronary revascularization—we know that from [REVIVED-BCIS2]—and whether or not it’s the same with TAVI is a very interesting question.”