By Jeffrey Zimmet, MD, PhD

Associate Professor of Medicine, University of California, San Francisco Director, Cardiac Catheterization Laboratory, San Francisco VA Medical Center

Dr. Zimmet reports no financial relationships relevant to this field of study.

SYNOPSIS: Imaging-defined leaflet thrombosis was common and similar between transcatheter aortic valve replacement and surgical aortic valve replacement bioprosthetic valves. These findings showed no apparent relationship to valve hemodynamics or to clinical outcomes, including stroke.

SOURCE: Blanke P, Leipsic JA, Popma JJ, et al. Bioprosthetic aortic valve leaflet thickening in the Evolut Low Risk Sub-Study. J Am Coll Cardiol 2020;75:2430-2442.

The occurrence of subclinical leaflet thrombosis in bioprosthetic aortic valves has been reported for several years, usually characterized by CT scan as hypoattenuated leaflet thickening (HALT) and reduced leaflet motion (RLM). While few of these patients will develop significant valvular dysfunction from leaflet thrombosis, for the remainder the meaning of this phenomenon remains unknown.

The natural history of these abnormalities over time remains unknown, as does the significance of these findings regarding valve durability and risk of outcomes, including stroke. Additionally, although this phenomenon has been described in bioprosthetic valves of all types, most investigations to date have concerned transcatheter aortic valve replacement (TAVR), with no clear answer as to whether the phenomenon is relatively TAVR-specific.

As TAVR has expanded into younger and lower-risk patients, these questions have taken on more importance. As part of the assessment comparing TAVR and surgical aortic valve replacement (SAVR) in low surgical risk patients, the authors of the Evolut Low Risk trial undertook a prespecified substudy to address some of these uncertainties.

Of the 1,403 patients enrolled in the overall trial, 503 were randomized in the substudy, which sought to evaluate HALT and RLM by CT at one-month and one-year post-valve implant. Of these, 375 patients had evaluable CT studies performed within 30 days of the procedure (197 TAVR, 178 SAVR). A few patients (18 TAVR and 39 SAVR) were receiving anticoagulation at the time of the CT, and were excluded from the primary analysis. Among patients who did not receive oral anticoagulation, HALT at the 30-day CT scan was seen in 31 of 179 TAVR subjects and 23 of 139 SAVR patients (P = 0.856). At the one year, thrombotic findings were seen in a higher proportion of patients, with positive findings in 47 of 152 TAVR patients and 33 of 116 SAVR patients (P = 0.661).

Among TAVR patients, all received the self-expanding supra-annular CoreValve Evolut prosthesis. Smaller annular size (as measured by perimeter) was associated with greater incidence of HALT. However, no procedural characteristics, such as post-dilatation or resheathing/recapture of the valve, were related to HALT frequency. Among SAVR patients, sutureless valve types were associated with greater incidence of HALT.

During the one-year course of the study, valve function was preserved with low transvalvular gradients regardless of the extent of HALT. As in prior studies, surgical valves showed higher mean gradients than TAVR, but did not consistently go up with increasing extent of leaflet thrombosis. Clinical endpoints, including death, stroke, transient ischemic attack (TIA), and myocardial infarction, were extremely low and showed no relationship to the presence or extent of HALT or RLM.

The authors concluded that imaging-defined leaflet thrombosis after bioprosthetic AVR is “frequent but dynamic” in the first year after valve deployment, with no difference between TAVR and SAVR valves. The extent of HALT and RLM did not correlate with hemodynamic status and had no apparent association with clinical outcomes, including stroke and TIA.


Since its initial description in 2015, the issue of imaging-defined leaflet thrombosis after aortic bioprostheses has represented a problem of uncertain extent and significance. The wide variability in incidence reported in prior studies has, on its own, flummoxed attempts to characterize its effects. During Evolut Low Risk Substudy, the authors performed CT at two timepoints in the first year and used rigorous imaging criteria to define HALT and RLM in a relatively large cohort. The incidence described here (approximately 16% at 30 days and 30% at one year) likely is accurate for the valve types described in a low-risk population.

The equivalence of thrombotic findings between surgical and transcatheter valves is an interesting finding. Earlier reports, such as the SAVORY study and the RESOLVE registry, had suggested much higher rates of leaflet thrombosis among TAVR valves. This does not appear to be the case, at least among the low-risk patients in this cohort or for the CoreValve platform.

HALT was dynamic over the course of the study. Some patients with HALT at 30 days showed resolution at one year, even without oral anticoagulation, while some without HALT at 30 days showed those findings at one year. This may, in part, explain the large range of leaflet thrombosis reported in earlier studies with less-rigorous imaging timing and image analysis.

The antiplatelet techniques differed between SAVR and TAVR groups. While TAVR patients were treated with one month of dual antiplatelet therapy, followed by aspirin alone, SAVR patients were treated with aspirin only (unless there was another indication for oral anticoagulation). This confirms earlier reports suggesting no effect of dual vs. single antiplatelet therapy on HALT.

The apparent lack of any association between imaging-defined leaflet thrombosis and either early valve dysfunction or major clinical outcomes is unsurprising, considering the positive early clinical results with these valves to date. Whether there will turn out to be a link to longer-term structural valve deterioration, and whether oral anticoagulation will be part of a successful therapeutic strategy, is a question that remains unanswered.