Less Leaflet Thrombosis in Low-Risk TAVR Patients Treated With Oral Anticoagulation
Preliminary findings hint that low-risk patients—where valve durability is key—might fare better with warfarin versus aspirin.
NATIONAL HARBOR, MD—Oral anticoagulation with warfarin after TAVR in patients with severe, symptomatic aortic stenosis at low risk for surgery appears to provide early protection against subclinical thrombosis and does so without increasing the risk of bleeding or vascular complications, according to the results of a small study presented this week at CRT 2020.
Compared with an aspirin-based strategy, there was less hypoattenuated leaflet thickening (HALT), reduced leaflet motion, and hypoattenuation affecting motion (HAM) in patients treated with warfarin for 1 month after TAVR. In a pooled analysis that combined patients in the randomized trial with those in a clinical registry, HALT was evident in 16.4% of 55 patients who received aspirin alone and in 3.1% of 65 patients treated with warfarin (P = 0.01). Reduced leaflet motion was seen in 9.4% of the aspirin-treated patients and in 1.5% of those in the anticoagulation arm (P = 0.05).
“We all worry that leaflet thrombosis may negatively impact transcatheter heart valve durability, although we haven’t been able to prove this yet,” said lead investigator Toby Rogers, MD, PhD (MedStar Heart & Vascular Institute, Washington, DC), during a press conference. “It makes sense—a leaflet that doesn’t open properly, that can’t be good for your bioprosthetic valve.”
Several trials are currently exploring use of oral anticoagulation after TAVR in patients, including POPular-TAVI, ATLANTIS, ENVISAGE-TAVI AF, and AVATAR. Recently, the GALILEO trial was stopped early for harm following an analysis showing that treatment with rivaroxaban (Xarelto; Bayer/Janssen) increased the risk of death or thromboembolic events—the primary composite efficacy outcome—when compared with antiplatelet therapy. The risk of all-cause mortality was higher with rivaroxaban, as was the risk of bleeding.
Rogers stressed they are not advocating oral anticoagulation for all patients after TAVR. Older patients, like those in GALILEO (mean age 80.6 years), are at too high a risk for bleeding. But in a lower-risk patient with a longer life expectancy—and in whom long-term valve durability is important—oral anticoagulation might be possible and worth considering, he suggested.
“One thing I wouldn’t want to do is to cause bleeding,” Rogers told TCTMD. “But absolutely, on the basis of [this study], I would lean toward the surgical guidelines which recommend anticoagulation, the duration of which we can discuss until the cows come home, but I do certainly think anticoagulation is reasonable.”
Lower Rates of HALT and HAM
Rogers and colleagues had previously conducted a study of 200 low-risk TAVR patients and found that 13.5% of those on antiplatelet therapy had evidence of subclinical leaflet thrombosis. In contrast, just 4.8% of patients who received oral anticoagulation had CT-documented HALT. That finding spurred the present study, known as the Low-Risk Trial 2.0, in which investigators randomized patients with an STS score 3% or lower to 1 month of anticoagulation after TAVR to determine if treatment could reduce the incidence of subclinical thrombosis when compared with antiplatelet therapy.
It makes sense—a leaflet that doesn’t open properly, that can’t be good for your bioprosthetic valve. Toby Rogers
Of 94 patients, 44 were randomized to oral anticoagulation (mean age 74.2 years) and 50 to aspirin (mean age 72.8 years). Clopidogrel was provided at the discretion of the treating physician but not mandated. If low-risk patients had a preexisting condition that required anticoagulation, or if they were excluded from randomization because of bleeding risks, they were placed in the registry arm. In total, 124 patients were included in the pooled data set from the randomized study and registry arm.
Rogers said they had planned a larger 200-patient trial, but the US Food and Drug Administration approval of TAVR for low-risk patients in 2019 slowed their enrollment so the researchers decided to lock the database and analyze the existing data.
In the randomized cohort, there was no risk of in-hospital life-threatening or major bleeding, and no risk of major vascular complications among patients treated with oral anticoagulation. In the pooled cohort, which included the randomized study plus registry patients, there was a slight uptick in the risk of bleeding and vascular complications in the aspirin arm, but “this is a quirk of the study design,” said Rogers. If a patient had a bleeding complication during TAVR, for example, they were included in the registry and not placed on oral anticoagulation. At 30 days, there was no increased risk of major bleeding or complications in the warfarin-treated patients in the randomized and pooled cohorts. There were no deaths in the study, but two nondisabling strokes occurred in the aspirin arm, one in-hospital and one by 30 days.
In the randomized cohort, the rates of HALT, reduced leaflet motion, and HAM were 16.3%, 10.4%, and 10.4% in the aspirin arm, respectively. For those who received warfarin, the corresponding rates were 4.7%, 2.3%, and 2.3%, respectively. The differences between the aspirin- and warfarin-treated patients were not statistically significant in the randomized cohort. However, when investigators analyzed the pooled cohort, the rate of HALT, reduced leaflet motion, and HAM favored oral anticoagulation and the difference compared with aspirin was statistically significant.
Looking only at the 94 patients in the randomized study, Rogers said the difference in HALT between the warfarin- and aspirin-treated patients “is clinically meaningful.” With a handful more patients, the between-group difference would have been statistically significant, “but everything points in the same direction and the trend goes very much in favor of anticoagulation for preventing or reducing the incidence of leaflet thrombosis in this patient population,” he explained.
The optimal antithrombotic treatment strategy after TAVR is one of the big questions in the field. Current US guidelines recommend 3 to 6 months of dual antiplatelet therapy after TAVR, a regimen that is largely based on data from the PCI realm; surgical guidelines recommend 3 months of warfarin for SAVR, but adherence to those recommendations in the era of bioprosthetic valves is mixed. The present study, said Rogers, raises questions about whether physicians are providing the best treatment course after TAVR in low-risk patients.
“Maybe what we’ve been doing in the past, which is aspirin plus or minus clopidogrel, isn’t the best thing for these patients,” he said. “Anecdotally, I’ve certainly heard from various surgeons who, based on the data from the TAVR studies, have gone back to starting their patients on warfarin after putting a bioprosthetic valve in because of these concerns about leaflet thrombosis.”
Subclinical Endpoints and Valve Durability
William Weintraub, MD (MedStar Heart & Vascular Institute), who wasn’t involved in the study, pointed out that HALT and the other measures are surrogate endpoints but what physicians and patients care most about are clinical events. With that in mind, he asked about the strength of association between the CT-derived measurements and clinical outcomes.
“We have absolutely no idea,” said Rogers in response. “That’s the truthful answer. It’s a subclinical endpoint. No one has been able to convincingly demonstrate that HALT is associated with stroke, which is the first thing we worry about. There are some signals from meta-analyses of various data sets that suggest it might be linked to an increased TIA rate, but that’s not good, high-quality data.”
However, rather than clinical events, it’s the issue of valve durability that most plagues interventional cardiologists with respect to HALT and HAM, and whether it will lead to structural valve deterioration down the line, he said. “The only way is to follow patients long-term,” said Rogers. “Again, it’s very hard, mechanistically, to believe that having a leaflet that doesn’t move is good for your valve long term, but that’s a presumption.”
Patients in Low-Risk TAVR 2.0 will be followed for several years, but there are no plans for serial CT assessments. Rogers noted that patients in the study received just a 30-day course of warfarin so whether HALT and reduced leaflet motion returns after stopping warfarin will not be known. “Clearly, this study isn’t going to give us an answer as to whether 30 days is the right duration—do we need 3 months, or do we need more—but in terms of generating a hypothesis, it certainly sends a strong signal.”
Following the late-breaking presentation of the data, Andreas Baumbach, MD (Queen Mary University of London and Barts Heart Center, London, England), agreed the study findings would be helped with a second CT evaluation after the completion of warfarin, perhaps at 60 days or beyond, because the “trigger” for thrombosis might still exist beyond the early stage.
Rogers concurred, noting that CT data have demonstrated there are patients clear of HALT, HAM, and reduced leaflet motion at 30 days who show evidence of impairment at 1 year.
Hendrik Treede, MD, PhD (University Hospital Bonn, Germany), praised the study investigators but also noted that the hypothesis-generating study only raises more questions about how long patients need to remain on oral anticoagulation to offset the risk of leaflet thrombosis. “Would you want to place all of these patients on lifelong Coumadin?” he wondered. “Then they grow older and develop bleeding complications.”
Rogers T, on behalf of the LRT 2.0 investigators. Anticoagulation versus antiplatelet therapy after transcatheter aortic valve replacement in low-risk patients. Presented at: CRT 2020. February 25, 2020. National Harbor, MD.
- Rogers reports consulting/proctoring for Edwards Lifesciences and Medtronic.