Two-year PARTNER 3 outcomes show convergence between TAVI and surgery on stroke and death

Martin Leon

Two-year outcomes of the PARTNER 3 clinical trial, in which low surgical risk aortic stenosis patients were randomised to receive either transcatheter aortic valve implantation (TAVI) or surgery, found that initial differences in death and stroke, which favour TAVI, were diminished after two years. Despite this, the primary endpoint of death, stroke or rehospitalisation remains significantly lower for TAVI than surgery.

These were among the conclusions of a study published in the Journal of the American College of Cardiology (JACC) this week, authored by Martin B Leon (Columbia Irving Medical Center/New York Presbyterian, New York, USA) et al, which also found that patients who underwent TAVI had increased valve thrombosis at two years compared to those who underwent surgery.

PARTNER 3 enrolled nearly 1,000 patients undergoing TAVI using the Sapien 3 (Edwards Lifesciences) balloon-expandable valve or surgery for severe aortic stenosis, and one-year results demonstrated the superiority of TAVI for the primary endpoints of death, stroke or rehospitalisation.

The study’s primary endpoint was a composite of death from any cause, all stroke (disabling or non-disabling), or cardiovascular rehospitalisation at one year.

Key secondary endpoints for the two-year follow-up report were acute myocardial infarction (MI), new-onset atrial fibrillation, need for a new pacemaker, new left bundle branch block, coronary obstruction, aortic valve reintervention, aortic valve endocarditis, and valve thrombosis.

Of the 950 patients in the as-treated population, 496 underwent TAVI and 454 surgery, and the intended valve was implanted in 948. Patients had a mean age of 73 years, included more men (69.3%), had lower STS-PROM scores (mean: 1.9%), fewer severe symptoms (New York Heart Association functional classes III or IV: 27.6%), and fewer co-existing conditions than patients enrolled in previous TAVI trials, the study team documents in JACC.

Outlining primary endpoint events at two years, the study team notes that the composite of death from any cause, all stroke, or cardiovascular rehospitalisation occurred in 57 patients (11.5%) after TAVI and 78 patients (17.4%) after surgery (HR: 0.63; 95% CI: 0.45 to 0.88; p=0.007). Between one and two years, TAVI was associated with more deaths than surgery (7 vs. 3), more strokes (6 vs. 1), and a similar number of rehospitalisations (10 vs. 8). The combined endpoint of death or disabling stroke at 2 years for TAVI was 3% compared with 3.8% for surgery.

In terms of secondary endpoints, Leon and colleagues note that there were small changes between one and two years for both TAVI and surgery in most secondary endpoints, including aortic valve re-intervention and endocarditis. However, the rates of valve thrombosis, which were numerically higher at one year after TAVI (1%) compared with those of surgery (0.2%; p=0.13), continued to diverge through two years (TAVI: 2.6%; surgery: 0.7%; p=0.02).

Among the patients with valve thrombosis at two years, seven of 13 (54%) patients who underwent TAVI and 0 of three patients who underwent surgery had an echocardiographic aortic valve mean gradient >20mmHg, with an increase from post-treatment of >10mmHg.

Leon et al suggest that the two-year PARTNER 3 follow-up study throws up a number of key findings, chiefly the reduction in the primary endpoint by 37% after TAVI compared with surgery.

Further key points highlighted by the study team include that death from all causes and strokes were more frequent with TAVI between one and two years, such that cumulative event rates through two years were similar to surgery; and that valve thrombosis was more frequent after TAVI versus surgery through two years—associated with an increase in aortic valve gradients in 54% of TAVI cases.

In an editorial comment accompanying the study in JACC, Bernard Predergast, Simon Redwood and Tiffany Patterson (St Thomas’ Hospital, London UK) consider the implications of the findings for heart teams. They write: “TAVI is the clear treatment of choice in most patients age >80 years where concerns regarding durability are usually not applicable. Conversely, SAVR [surgical aortic valve replacement] remains the default in those age <70 years (and certainly <65 years) given the paucity of long-term TAVI durability data in younger patients. Both modes of intervention are effective in patients between these age thresholds, and the choice should be determined following a shared decision-making process incorporating the pros and cons of SAVR and TAVI guided by patient-specific anatomic and clinical considerations, the potential need for repeat valve interventions over a lifetime journey (including their mode, sequence, and associated risks), and the patient’s values and preferences.”

Asked to comment on the implication of the findings, Leon told Cardiovascular News: “I don’t believe that the catch-up in death/stroke events between one to two years suggests diminishing TAVI benefit or an expected trend and is more likely random chance, given the very low event rates for both therapies.

“Remember, the absolute death/stroke rates still favour TAVI after two years and the primary composite endpoint still indicated TAVI benefit cw surgery.  Both therapies (surgery and TAVI) are excellent choices in low-risk patients and the basis for decision-making still rests on an individual patient assessment accounting for all factors, including age, specific comorbidities and other clinical factors, and anatomical considerations favouring one or the other treatment.”


Please enter your comment!
Please enter your name here