ACC 2020: Updates from the PARTNER 3 Trial — Dr Michael J Mack

 

Dr Michael J Mack (Baylor Scott & White Heart Hospital, Plano, TX, US) discusses the two-year clinical and echocardiographic outcomes from the PARTNER 3 low-risk randomized trial.

The goal of PARTNER 3 was to evaluate transcatheter aortic valve replacement (TAVR) compared with surgical aortic valve replacement (SAVR) among low-risk patients with symptomatic severe aortic stenosis.

Filmed remotely from Texas.

Questions 

1. Can you remind us of the objective and study design of Partner 3?
2. What did the one-year data show?
3. You are presenting the 2-year data for both clinical and echocardiographic outcomes, what does this data show?
4. What conclusions should be made with this additional data?
5. Do you think TAVR should be the default strategy for patients with symptomatic aortic stenosis?

 

Transcript Below : 

Question 1 : Can you remind us of the objective and study design of Partner 3?

The PARTNER 3 trial randomised 1,000 patients who were at low surgical risk with aortic stenosis. Half of the patients were randomised to surgery and half were randomised to transcatheter aortic valve implantation. The analysis was an as treated population. So that was 950 patients. The primary endpoint of the trial was a composite endpoint of death, stroke and cardiovascular re-hospitalisation at one year. The mean age of patients enrolled in the trial was 73 and the mean STS predicted risk score was 1.9. 

Question 2 : What did the one-year data show?

The one year data showed that TAVI was superior to surgery for the primary endpoint of death, stroke and cardiovascular re-hospitalisation. Looking at the individual components of the primary endpoint, both cardiovascular re-hospitalisation and stroke were statistically significant. Death was not a significant difference although it trended toward a benefit with TAVR. 

Question 3 : You are presenting the 2-year data for both clinical and echocardiographic outcomes, what does this data show?

So, the two year data does show some changes compared to the one year data. The two year data was still favourable for TAVR, but not nearly as significant as at one year. The hazard ratio was a 37% benefit for TAVR at two years. TAVI was still non-inferior to surgery at two years, but the main difference was due to a significant difference in the rate of re-hospitalisation. The rate of stroke and death had both narrowed between the two arms at two years so that they were close to identical at the two year period of time. What we also saw were that difference happened because there were more deaths and there were more strokes in the TAVR arm than in the surgery arm between year one and year two. We also saw that there was an increased incidence of valve thrombosis in the first two years with TAVI compared to surgery, there were 13 patients with VARC-2 definition of valve thrombosis at two years with TAVR and three with surgery. We also saw that both treatments remain durable at two years. There was no change in main valve gradients, no change in aortic valve areas and no change in the incidence of either mild or moderate paravalvular leak between the two arms. We also analysed it by new VARC-3 criteria and found that the incidence of structural valve deterioration or bioprosthetic valve failure was the same between the two arms. 

Question 4 : What conclusions should be made with this additional data?

So the conclusions of this trial is that both options, both TAVR and surgery remain viable options for patients who are low surgical risk. However, it has to be noted that there was an increase in both death, stroke, as well as valve thrombosis between one and two years in the TAVR arm. It should also be noted that these are only two year results and we really need to know longer term results and hence subsequent analysis would be done at three years, five years and at 10 years to be better informed about shared decision making with the patient. Also, we have the caveat that these results apply only to the population studied. So that means that patients that were excluded from the trials, so asymptomatic patients, patients with bicuspid valve disease, patients that were not candidates for a transfemoral approach for TAVR, patients with significant concomitant coronary artery disease were not studied and these results should not be extrapolated to those patients. So, all in all, the results are still positive for TAVR, just not quite as different as they were at one year. The time-to-event Kaplan-Meier curves begin to converge some between year one and year two. 

Question 5 : Do you think TAVR should be the default strategy for patients with symptomatic aortic stenosis?

So, from what we know now, TAVR will be the preferred approach in low surgical risk patients as long as they mimic the patients that were included in the trial. So, these results don't apply to patients that were younger or have bicuspid aortic valves. The mean age was 73, very few patients were under 65. When you have two different modalities of therapy and one is less invasive than the other, then the less invasive approach is always going to be preferred by the patients and their referring physicians. Therefore, from what we know right now, TAVR will be the preferred approach in low surgical risk patients. However, we've already seen some changes between year one and year two and we have to be open-minded that we may continue to see further changes as subsequent analyses are done and these patients live longer. The most important aspect is going to be durability. The younger you are as a patient, the longer you need your valve to last because the longer you're going to live. So the real important information is going to be 10 years outcomes of structural valve deterioration.