Leaflet Thickening and Impaired Motion No Worse With TAVR: PARTNER 3 CT

SAN FRANCISCO, CA—Hypoattenuated leaflet thickening (HALT) and abnormal motion were more evident at 30 days among patients with severe aortic stenosis treated with TAVR in the low-risk PARTNER 3 trial, but by 1 year, the rates did not differ between surgical and transcatheter treatment arms, according to a new CT substudy presented today at TCT 2019.

Both HALT and reduced leaflet motion resulted in slightly higher mean aortic valve gradients at 30 days and 1 year, but lead investigator Raj Makkar, MD (Cedars Sinai Medical Center, Los Angeles, CA), said this increase is “clinically insignificant,” noting there were no deaths or MIs in patients with subclinical leaflet thickening.

“These are low risk patients and there are not many clinical events so it’s hard to make a definitive conclusion, but the fact that clinical events are low is reassuring,” Makkar told TCTMD. “I think we’ll have to follow these patients long-term and do larger analyses, perhaps by pooling studies that have been embedded within different clinical trials, to make a definitive statement about whether or not HALT is related to clinical events.” 

Nonetheless, based on these short-term results, Makkar said he is comforted because “there is no association with any of the life-threatening clinical events, such as death, stroke, or MI.” Given the lack of a clear association between HALT and reduced leaflet motion with serious outcomes, the routine use of oral anticoagulation following TAVR is not justified.

“Whether we give anticoagulation or not to these patients, that question should be answered by clinical trials,” said Makkar. “Trials don’t just look at imaging, but also the cost, or the risk, of giving anticoagulation to those patients. Bleeding is not uncommon in patients on anticoagulants and bleeding is associated with higher mortality. .”

Clinically Insignificant HALT

In 2015, Makkar published data showing reduced leaflet motion in patients included in the Portico Resheathable Transcatheter Aortic Valve System US investigational device exemption (PORTICO-IDE) study, as well as in a pooled cohort of patients from the RESOLVE and SAVORY single-center registries. Those data raised concerns about subclinical leaflet thrombosis, a concern that was heightened with pooled data from RESOLVE/SAVORY showing a higher incidence of stroke or transient ischemic attack (TIA) in those with reduced leaflet motion.

The clinical significance of HALT and reduced leaflet motion following TAVR with regards to progression to clinical thrombosis, stroke/TIA, and valve durability, continues to be a hot topic, particularly given that younger low-risk patients are now eligible for TAVR. Thrombus on the bioprosthetic valve can present with HALT with relatively normal motion, HALT with reduced leaflet motion but normal aortic valve gradients, and clinical valve thrombosis with elevated valve gradients, said Makkar.  

The PARTNER 3 substudy included 408 patients, including 346 individuals with evaluable CTs at 30 days and 312 patients with CTs evaluable at 1 year. At 30 days, the incidence of HALT was significantly higher among patients treated with TAVR than among those treated with surgery (13.3% vs 5.0%; P = 0.03), but the incidence was similar at 1 year (27.5% with TAVR vs 20.2% with surgery; P = 0.19). For those with HALT, reduced leaflet motion was observed in all patients, with the vast majority of patients having partially restricted leaflets. 

Interestingly, of the 25 patients with HALT at 30 days, there was resolution of leaflet thickening at 1 year in 14 subjects, none of whom received oral anticoagulation. For the 217 patients without HALT at 30 days, 46 developed thickened leaflets at 1 year. There was no significant difference in the mean aortic gradient between those with HALT and without HALT at 30 days and 1 year. On average, mean gradients were roughly 1-2 mmHg higher among patients with HALT and/or impaired leaflet motion at 1 year. For those with greater severity of HALT, mean gradients were higher when compared with those without leaflet thickening. 

Clinical events were very low, but there was no significant increase in the risk of death or MI between patients with HALT and those without leaflet thickening. There were three patients with HALT at 30 days, and one patient without leaflet thickening, who developed clinical valve thrombosis. While the pooled rates of death, stroke, TIA, and thromboembolic events were numerically higher among patients with HALT, Makkar stressed these findings are simply hypothesis-generating and require longer-term follow-up to clarify the impact on clinical outcomes.

More on Anticoagulation

Given the evidence of HALT, Ajay Kirtane, MD (NewYork-Presbyterian/Columbia University Irving Medical Center, New York, NY), who chaired the morning press conference, said the question of oral anticoagulation is bound to be debated, but he also noted that interpreting the findings will be tricky given that more than half of the 25 patients with HALT at 30 days had resolution at 1 year without receiving any treatment. Kirtane raised the possibility that anticoagulation might be an option in the early period after TAVR.

Speaking with the media, Robert Bonow, MD (Northwestern University Feinberg School of Medicine, Chicago, IL), said there is a class IIa indication for oral anticoagulation for 3 to 6 months in patients who undergo surgical aortic valve replacement and also wondered whether oral anticoagulation might be an option in TAVR. He noted that HALT does alter the performance of the valve, albeit in a mild way, and it will be important to study patients long term to determine if these changes in aortic valve gradients translate into early valve deterioration.

“I think the jury is still out,” he said. “I mean, 20% of the surgical valves also had some leaflet thrombosis as well. I think we still need to wait and see if we need to consider oral anticoagulation in the surgical and TAVR realm.”

For his part, Makkar disagreed. “We’re looking at an imaging finding,” he said. “We could give a drug that could eliminate this but at what expense? The risk-benefit ratio of a therapy is best suited for a clinical trial.” He pointed out that HALT was evident at 1 year in 22% of patients who received a bioprosthetic valve, an amount that isn’t trivial. Rather than wait for 10 years to see that HALT adversely impacts valve durability, it would be best to design a clinical trial to assess the impact of oral anticoagulation, said Makkar.

“We need a trial in younger patients that are going to live a long time,” he said. “Not only do we need it to address the issue of clinical events, but also for valve durability. Whether or not [HALT] affects valve durability needs to be studied. We can either wait for many years to find out or we can design the studies now.”

The GALILEO trial, which was testing a rivaroxaban-based anticoagulation strategy against antiplatelet therapy in the TAVR setting, was stopped last year following a review by the data and safety monitoring board that showed a higher risk of death, thromboembolic events, and bleeding in the rivaroxaban arm. Makkar noted that GALILEO included older patients with significant comorbidities and advised that future trials investigating oral anticoagulation need to include younger patients at lower risk of bleeding.

Like Makkar, Michael Mack, MD (The Heart Hospital Baylor, Plano, TX), one of the PARTNER 3 investigators, said he doesn’t believe oral anticoagulation is justified after TAVR. In fact, he questioned the indication for oral anticoagulation following surgical aortic valve replacement. “Now that we have routine surveillance of surgical valves for the first time, I think it calls into question the IIa recommendation for 3 months of anticoagulation,” said Mack. “We’ve always done it, but I think we’ll stop doing it on the basis of [PARTNER 3 CT substudy].”

To TCTMD, Makkar said systematic CT assessment of TAVR patients is also unwarranted because the information doesn’t change clinical practice. CT scans, he said, should only be prompted by increased aortic valve gradients or thromboembolic events.