PARTNER 3

Published

Mack MJ, et al. "Transcatheter Aortic-Valve Replacement with a Balloon-Expandable Valve in Low-Risk Patients". The New England Journal of Medicine. 2019. 380(18):1695-1705.
PubMed • Full text • PDF • ClinicalTrials.gov

Clinical Question

In patients with symptomatic, severe aortic stenosis who are low risk surgical candidates (expected periprocedural mortality <4%, mean 1.9%), is transcatheter aortic valve replacement (TAVR) non-inferior to surgical aortic valve replacement (SAVR) with respect to the primary composite outcome of death, stroke, or rehospitalization at 1 year?

Bottom Line

In patients with symptomatic, severe aortic stenosis who are low-risk surgical candidates, TAVR was associated with a 6.6% absolute reduction in death, stroke, or rehospitalization at 1 year when compared to SAVR. The superiority of TAVR was driven by symmetric reductions in each component of the primary endpoint, including a 1.5% absolute reduction in overall mortality.

Major Points

The PARTNER series of randomized controlled trials has firmly established the role of TAVR with the balloon-expandable Edwards Sapien valve in patients with severe symptomatic aortic stenosis (AS) at prohibitive risk of surgery (PARTNER A), high risk for surgery (PARTNER B), and intermediate risk for surgery (PARTNER 2). Analogous results have been demonstrated with the CoreValve and Evolut families of self-expandable TAVR valves. On the strength of these prior studies, the 2017 ACC Expert Consensus Statement recommends a multidisciplinary Heart Team assessment to determine the relative benefit of proceeding with TAVR versus SAVR in individuals with at least intermediate risk for perioperative mortality (Society of Thoracic Surgeons [STS] score ≥4%) with SAVR . Whether TAVR is an effective alternative to SAVR in patients at low surgical risk (STS score <4%) remained unknown.

The 2019 PARTNER 3 trial randomized 1000 patients with symptomatic severe AS to TAVR versus SAVR and assessed for a primary outcome of death, nonfatal myocardial infarction, or rehospitalization. At 1 year, TAVR was associated with a 6.6% absolute reduction in death, stroke, or rehospitalization when compared to SAVR. The superiority of TAVR was driven by symmetric reductions in each component of the primary endpoint, including a 1.5% absolute reduction in overall mortality. Median length of stay (3 days versus 7 days) and percentage of patients discharged to home (95.8% versus 73.1%) also favored TAVR. In terms of safety, TAVR was associated with a 15.7% absolute increase in new left bundle-branch block, while SAVR was associated with a 21.1% absolute increase in major bleeding. At 1 year, the percentage of patients with mild paravalvular leak was higher with TAVR (29.4%) than with SAVR (2.1%). Of note, all patients who underwent TAVR were transfemoral access candidates, which is known to be associated with improved outcomes when compared to other access sites. Furthermore, the primary analysis was per-protocol, and therefore findings may have been impacted by bias due to differential drop out rates after randomization (43 in surgery group, 7 in TAVR group). Reassuringly, however, in a secondary intention-to-treat imputation analysis, the main study findings persisted.

Thus, the results of PARTNER 3, along with the similar results of the Evolut Low Risk trial using the Evolut R self-expanding TAVR valve, strongly suggest that TAVR is not only a suitable alternative to SAVR in low-risk patients, but may be a superior first-line option associated with lower adverse cardiovascular event rates. Importantly, these results were obtained exclusively using TAVR with transfemoral access and therefore they cannot be generalized to patients who are not transfemoral access candidates. Taken together, TAVR has now been shown to be either non-inferior or superior to SAVR across the entire spectrum of surgical risk (excepting individuals generally excluded or underrepresented in TAVR trials such as bicuspid aortic valve and mixed aortic stenosis/insufficiency). Even so, long-term outcomes data regarding the durability of TAVR valves (particularly given increased rates of mild paravalvular leak at 1 year with TAVR) will likely be required before TAVR is universally accepted as the approach of choice for patients with symptomatic severe AS.

Guidelines

As of April 2019, no guidelines have been published that reflect the results of this trial.

Design

• Multicenter, randomized, controlled trial
• N= 1,000
  • TAVR (n=503)
  • Surgery (n=497)
• Setting: 71 centers in the United States, Canada, Australia, New Zealand, and Japan
• Enrollment: March 2016 to October 2017
• Mean follow-up: 1 year
• Analysis: As-treated
• Primary outcome: composite of death from any cause, stroke, or rehospitalization (related to the procedure, valve, or heart failure) at 1 year after procedure

Population

Inclusion Criteria

• Severe calcific aortic stenosis, defined as both of:
  • AVA ≤1.0cm2 or AVA index ≤0.6cm2/m2
  • Jet velocity ≥4.0m/s OR mean gradient ≥40mmHg
• Symptomatic or equivalent, defined as any of:
  • NYHA functional class ≥2
  • Exercise tolerance test demonstrating limited exercise capacity, abnormal BP response, or arrhythmia
  • Asymptomatic with LVEF <50%
• Low surgical risk (STS for Mortality <4% and agreement by the site heart team and the trial case review committee)
• Anatomy suitable for TAVR with transfemoral placement of the balloon-expandable SAPIEN 3 system

Exclusion Criteria

• BMI >50 kg/m2
• MI ≤1 month before randomization
• Unicuspid, bicuspid, or non-calcified aortic valve
• Severe aortic or mitral regurgitation
• Moderate or greater mitral stenosis
• Pre-existing mechanical or bioprosthetic valve in any position
• Complex coronary artery disease
• Symptomatic carotid or vertebral artery disease or successful treatment of carotid stenosis within 30 days
• History of bleeding diathesis, coagulopathy, hypercoagulable state, or severe cytopenia
• Hemodynamic or respiratory instability
• Hypertrophic cardiomyopathy with obstruction
• LVEF <30%
• Intracardiac mass, thrombus, or vegetation
• Inability to tolerate antiplatelet/anticoagulation following procedure
• Stroke or TIA within 90 days
• Renal insufficiency with eGFR < 30mL/min or requirement for renal replacement therapy
• Severe lung disease or pulmonary hypertension
• History of cirrhosis or any active liver disease
• Significant frailty as determined by the Heart Team
• Life expectancy <24 months

Baseline Characteristics

From the TAVR group

• Demographics: age 73.3, male 67.5%, nonwhite 7.7%
• Cormorbidities: BMI 30.7, STS 1.9, EuroSCORE II 1.5, NYHA III or IV 31.2%, CAD 27.7%, stroke 3.4%, MI 5.7%, carotid disease 12.7%, PAD 6.9%, COPD 5.1%, creatinine >2 0.2%, DM 31.2%, AF 15.7%, PPM 2.4%, pulmonary hypertension 4.6%
• Aortic valve: mean AVA 0.8, mean gradient 49.4, LVEF 65.7%, moderate or greater AI 3.9%, systolic annular perimeter 78.1 mm, systolic annular area 473.5 mm2

Interventions

• Randomized 1:1 to transfemoral TAVR with the SAPIEN 3 system vs surgical bioprosthetic aortic-valve replacement
  • Randomization stratified according to site
• All TAVR procedures used transfemoral route
• Balloon valvuloplasty before and after TAVR performed at operator's discretion
• Patients received aspirin (81mg) and clopidogrel (≥ 300mg) before TAVR and were advised to continue taking these medications for at least 1 month after the procedure
• All patients underwent neurological exams at baseline and 30 days
• Rehospitalization was defined as any hospitalization related to the procedure, valve, or heart failure
• All components of the primary endpoint and key secondary end points were adjudicated by a clinical events committee whose members were aware of the treatment assignments

Outcomes

Comparisons are transcatheter aortic-valve replacement vs. surgical aortic-valve replacement.

Primary Outcomes

Composite of death from any cause, stroke, or rehospitalization at 1 year

42 (8.5%) vs 68 (15.1%) [HR 0.54; 95% CI 0.37-0.79; P=0.001]

Secondary Outcomes

Stroke

0.6% vs 2.4% [HR 0.25; 95% CI 0.07-0.88; P=0.02]

Death or stroke

1.0% vs 3.3% [HR 0.30; 95% CI 0.11-0.83; P=0.01]

New-onset atrial fibrillation at 30 days

5.0% vs 39.5% [HR 0.10; 95% CI 0.06-0.16; P<0.001]

Length of the index hospitalization

3 days vs 7 days [95% CI -4.0 to -3.0; P<0.001]

Subgroup Analysis

• The benefit of TAVR was similar in subgroup analyses based on age, sex, STS-PROM score, LVEF, NYHA class, presence of AF, and baseline KCCQ summary score.

Adverse Events

Life-threatening or major bleeding

3.6% vs 24.5% [HR 0.12; 95% CI 0.07-0.21]

New permanent pacemaker

36 (7.3%) vs 24 (5.4%) [HR 1.39; 95% CI 0.83-2.33]

Procedural death

2 (0.4%) vs 4 (0.9%)

Access site infection

2 (0.4%) vs 6 (1.3%)

Criticisms

• Outcomes adjudication was not blinded, potentially allowing for bias in outcomes ascertainment
• The primary analysis was per-protocol, and therefore findings may have been impacted by bias due to differential drop out rates after randomization (43 in surgery group, 7 in TAVR group). Reassuringly, however, in a secondary intention-to-treat imputation analysis, the main study findings persisted.

Funding

• Study sponsored by Edwards Lifesciences, who funded all trial-related activities and participated in site selection, data collection and monitoring, and statistical analysis.