Introduction

Transcatheter aortic valve replacement (TAVR) is an established therapy for patients with symptomatic severe aortic stenosis (AS) who are at extreme, high, or intermediate risk for surgical aortic valve replacement (SAVR). How TAVR compares with SAVR in patients who are at low risk for surgery is now the subject of multiple clinical trials. The LRT (Low Risk TAVR) trial was the first Food and Drug Administration–approved Investigational Device Exemption trial in the United States to evaluate feasibility of TAVR in low-risk patients. Early results of the LRT trial confirmed safety of TAVR in low-risk subjects with symptomatic severe AS (1). At 30 days, there was zero mortality and zero disabling stroke. The rate of permanent pacemaker (PPM) implantation was low (4.8%) and only 1 patient had greater than mild aortic regurgitation (0.5%). TAVR compared favorably to a historical control cohort of SAVR patients from the enrolling institutions using data from the Society of Thoracic Surgeons (STS) database (1). Leaflet thickening was detected in 14% of the patients at 30 days. Herein, we present 1-year clinical outcomes and transcatheter heart valve hemodynamics of the TAVR cohort from the LRT trial.

Methods

The LRT trial ( NCT02628899) was a prospective investigator-initiated multicenter feasibility trial to test the safety of transfemoral TAVR in low-risk patients with symptomatic severe AS who were deemed eligible for transfemoral TAVR. The trial design has been described previously (2). The research protocol was approved by relevant Institutional Review Boards (Online Table 1). All patients gave written informed consent and were evaluated before enrollment by an independent Clinical Review Committee to ensure low-risk status and clinical and anatomical eligibility for transfemoral TAVR, which was performed using commercially available transcatheter heart valves. Pre- and post-procedural echocardiograms and computed tomography (CT) studies were analyzed by an independent core laboratory. The primary endpoint of the study was all-cause mortality at 30 days. Important secondary endpoints included stroke, paravalvular regurgitation, PPM implantation, and need for reintervention at 1 year (±30 days). Trial data were independently 100% monitored from screening to 30 days, and a risk-based plan was used from 30 days to 12 months. All clinical endpoints were adjudicated by an independent Clinical Events Adjudication Committee comprising an interventional cardiologist, a cardiothoracic surgeon, and a neurologist using Valve Academic Research Consortium-2 definitions (3). Subjects in the LRT trial will be followed for 5 years after TAVR with yearly clinical visits and echocardiography.

Statistical analysis

Continuous variables are presented as mean ± SD, categorical variables as proportions. All analyses were performed with SAS version 9.4 (SAS Institute, Cary, North Carolina).

Results

Baseline characteristics are summarized in Table 1. Compared with earlier pivotal TAVR studies, LRT trial subjects were younger, had lower STS Predicted Risk of Mortality score, and had fewer comorbidities, in keeping with their low-surgical-risk status. Procedural details have been published elsewhere (1). In brief, moderate sedation was used in 75.5%, transfemoral access in 100%, and balloon-expandable valves in 88.2% of subjects. Nearly all (95.1%) subjects received a 23 mm transcatheter heart valve or larger. Post-procedure length of stay was 2.0 ± 1.1 days. At 30 days, there was zero mortality and zero disabling stroke. Table 2 summarizes procedural complications and clinical outcomes at 30 days and 1 year. One subject withdrew consent from the study and 2 were lost to follow-up at 1 year. After a median follow-up of 366 (interquartile range: 351 to 386) days, all-cause mortality at 1-year was 3.0% (n = 6 of 197), and cardiovascular mortality was 1.0% (n = 2 of 197). Both cardiovascular mortalities were sudden deaths at home. Figure 1 shows Kaplan-Meier survival for the TAVR cohort. Two subjects required reintervention, following diagnosis of aortic bioprosthetic valve endocarditis, 3 months and 7 months after TAVR, respectively. Both underwent surgical excision of the transcatheter heart valve and implantation of a new surgical bioprosthesis with uneventful recovery. Ten subjects required PPM implantation after TAVR and a further 4 subjects required PPM after the 30-day follow-up visit. No subject required percutaneous coronary intervention in the year following TAVR.

Transcatheter heart valve hemodynamics were excellent post-procedure and were maintained at 1 year (Central Illustration, Table 3). Mean left ventricular ejection fraction at 1 year was 63.1 ± 8.3%. Hypoattenuated leaflet thickening (HALT) was observed in 14.0% of subjects (n = 27 of 193) with an evaluable CT or transesophageal echocardiography at 30 days. HALT was observed in subjects who received balloon-expandable valves only, and not self-expanding valves. The Central Illustration, Table 3 summarize hemodynamics according to presence or absence of HALT. At 30 days, mean valve area and dimensionless index were lower in subjects with HALT, with a trend toward higher mean gradient. At 1 year, these differences appeared to resolve. At 1 year, there was a numerically higher rate of stroke in subjects with HALT (3.8% vs. 1.9%; p = 0.53), although the absolute number of events was small in both groups (1 of 27 with HALT vs. 4 of 166 with no HALT).

Discussion

LRT is the first prospective trial to report 1-year outcomes of TAVR in low-risk patients in the United States. It is to be expected that procedural outcomes in low-risk patients with symptomatic severe AS should be better than in intermediate- or high-risk patients. The LRT trial confirmed this with excellent in-hospital and 30-day TAVR results that were maintained at 1 year, with low mortality and stroke rates.

As the TAVR implantation procedure has become safer, attention has turned to other factors, including PPM implantation, paravalvular regurgitation, hemodynamics, long-term outcomes, and access to the coronaries. These factors are considered all the more important in younger low-risk patients who have longer life expectancy. In the NOTION (Nordic Aortic Valve Intervention) trial, all-cause death at 1 year in a mixture of intermediate- and low-risk subjects was 4.9% (4). The rate of new PPM implantation at 1 year was 38.0%. The rate of significant aortic regurgitation (defined as greater than mild in severity) was 15.7% using the first-generation self-expandable CoreValve (Medtronic, Minneapolis, Minnesota).

One-year mortality in the LRT trial was 3.0%. This compares favorably with the transfemoral access TAVR cohort of the PARTNER 2A (Placement of Aortic Transcatheter Valves) trial (10.0%) (5) and the TAVR cohort of the SURTAVI (Surgical or Transcatheter Aortic-Valve Replacement in Intermediate-Risk Patients) trial (7.0%) (6) in intermediate-risk patients, suggesting that TAVR is safe in low-risk patients. Transcatheter heart valve hemodynamics in the LRT trial were excellent immediately post-implantation, and these excellent results persisted at 1-year follow-up. PPM implantation rate was the lowest of any TAVR study to date using commercially available transcatheter heart valves. Clinically significant (greater than mild) paravalvular aortic regurgitation occurred in only 1 subject at 30 days.

Leaflet thrombosis has been observed with both surgical and transcatheter bioprostheses (7). A largely subclinical phenomenon at present, there is no clear signal that leaflet thrombosis is associated with excess cerebrovascular accidents or premature structural valve deterioration. In the LRT trial, leaflet thickening was a secondary endpoint. Subjects in the LRT trial underwent follow-up time-resolved contrast-enhanced cardiac CT at 30 days to evaluate for HALT and reduced leaflet motion, which were found in 14.0% and 11.2% of subjects, respectively. At 30 days, valve area and dimensionless index were lower in subjects with HALT (Table 3). This could be explained by the fact that HALT was only observed in subjects who received a balloon-expandable valve. It is well documented that self-expanding valves with supra-annular leaflets have lower gradients and larger valve areas than do balloon-expandable valves with intra-annular leaflets (8,9). Alternatively, these differences could be due to leaflet motion restriction related to possible thrombosis. However, the presence of HALT did not appear to have an impact on valve hemodynamics at 1 year in this study. The stroke rate was twice as high in patients with HALT, but the proportions belie the low event rate (1 stroke in 27 subjects with HALT vs. 4 strokes in 166 subjects without HALT). There is clearly a potential mechanistic link between aortic valve leaflet thrombosis and stroke, but our study was not powered to detect differences and the stroke rate was very low in this low-risk population. Whether HALT truly is associated with a higher rate of stroke could be derived from a meta-analysis including the 2 industry-sponsored randomized trials of TAVR ( NCT02675114, NCT02701283) in low-risk patients that are expected to report their findings in 2019. Both trials include nested substudies with 30-day follow-up CT scans to evaluate for leaflet thrombosis. The cumulative data from these substudies and the 2 LRT trials will be useful to address this yet unanswered question. In addition, longer-term data are needed to evaluate whether HALT or subclinical leaflet thrombosis leads to premature structural valve deterioration and to determine the optimal antiplatelet or anticoagulation regimen after TAVR. Further, dedicated studies are warranted to examine the optimal anticoagulation regimen to minimize leaflet thrombosis at 30 days. The currently enrolling investigator-initiated LRT 2.0 trial ( NCT03557242) seeks to answer these questions by randomizing low-risk patients to aspirin or warfarin for 30 days after TAVR.

With these excellent results, the argument against dissemination of TAVR to low-risk patients is primarily based on the lack of long-term durability data of transcatheter heart valves. All bioprostheses—surgical and transcatheter—have a finite lifespan before their leaflets inevitably degenerate, leading to stenosis or regurgitation. From the PARTNER 1 trial, 5-year echocardiography data in surviving subjects have revealed no signal for premature structural valve deterioration (10–12). The paucity of data regarding durability of transcatheter heart valves is often underscored as a weakness of TAVR, but the data for surgical bioprostheses are also sparse. Recently, 2 standardized definitions have been proposed in Europe and North America (13,14), but until these are incorporated into clinical trial endpoint definitions, the data will remain difficult to parse. Ultimately, it is not clear that durability is at the top of the list of priorities for patients with symptomatic severe AS. Another concern with TAVR in younger and healthier patients is catheter access to the coronary arteries, which is more challenging with transcatheter than with surgical aortic bioprostheses in situ. In the LRT trial, the mean age of the patients was 73.6 years, and none of them required coronary intervention within the first year. The onus is now on the industry to design transcatheter heart valves that will facilitate easy access to the coronary arteries should patients require future coronary intervention.

Study limitations

The LRT trial was not a randomized study. A historical SAVR cohort from the STS database served as control for the primary endpoint of 30-day mortality (1). Unfortunately, the STS database does not capture any data beyond 30 days, so it is not possible to perform a comparison of TAVR versus SAVR outcomes beyond 30 days. To enable future studies using this methodology, it would be useful for the STS database to capture 12-month clinical outcomes, akin to the STS/American College of Cardiology Transcatheter Valve Therapy registry.

Conclusions

TAVR appears to be safe in low-risk patients with symptomatic severe AS, with low rates of in-hospital, 30-day, and 1-year complications. HALT, observed in a minority of patients at 30 days, did not appear to have an impact on valve hemodynamics at 1 year. More data are required to determine whether HALT is associated with increased risk of stroke.

Footnotes