When Chosen Wisely, 90-Year-Olds Do Well After TAVR

Chicago, IL—New analysis of data from the PARTNER 1 randomized and registry cohorts shows that patients aged 90 years or older who undergo TAVR do not have any increase in mortality or major cardiovascular events compared with their younger peers. In fact, TAVR-treated nonagenarians are no less likely to survive over the long-term than similarly aged people in the United States without aortic stenosis.  

Take Home. When Chosen Wisely, 90-Year-Olds Do Well After TAVRAge alone should not be a barrier to TAVR, said lead author Creighton W. Don, MD, PhD (UW Medicine, Seattle, WA), during a moderated poster presentation today at the American College of Cardiology 2016 Scientific sessions. 

Don reported that he has begun implementing this lesson in his day-to-day practice. Speaking with patients who are above 90 and don’t have a lot of comorbidities, he said, “I’ve started to counsel [them] that in our study we have actually found that your outcomes are possibly, once you get through the procedure, fairly similar to what a 90-year-old’s would be who didn’t have your stenosis.” 

The findings come on the heels of a paper published last month using data from 24,000 patients in the Society of Thoracic Surgeons/American College of Cardiology Transcatheter Valve Therapy (TVT) Registry, which showed a "clinically modest" increase in mortality for nonagenarian TAVR patients. That paper too, however, concluded that age should not be the deciding factor in whether to offer transcatheter valve procedures to these elderly patients.  

Not on Death’s Door

For the study, Don and colleagues analyzed data on 2,621 patients who received the Sapien valve (Edwards Lifesciences) while enrolled in PARTNER 1 randomize and registry cohorts between 2007 and 2012. Patients aged ≥ 90 years tended to be frailer compared with patients aged 80 to < 90 years or > 80 years. They also had higher mean logistic EuroSCORE and STS scores at baseline. But interestingly, they had better quality of life on the Kansas City Cardiomyopathy Questionnaire (KCCQ) and were less likely to have major comorbidities such as prior CABG or PCI, history of MI, and diabetes.  

“The reason for this is there is some selection bias clearly,” Don said. First of all, these are the patients healthiest enough to survive to their 90s, he explained. “Secondly, if I’m going to do TAVR on a 90-year-old, they probably don’t have a ton of other comorbidities and they’re not on death’s door.” 

In-hospital outcomes were similar among the 3 groups, though the nonagenarians were less likely to be discharged home than those aged 80 to < 90 or < 80 years (76%, 79%, and 91%, respectively; P < 0.01). Thirty-day mortality and stroke risks also were equivalent regardless of patient age; the only 30-day outcome to differ was major bleeding (11%, 9.8%, 6.9%; P = 0.04), though even that disparity did not remain after adjustment for baseline characteristics associated with age. 

Patients in each age group experienced a similar degree of KCCQ improvement compared with baseline. “The oldest patients had highest quality of life at the end of the study, and this was very interesting,” Don said. 

Echocardiographic outcomes were less favorable for nonagenarians, who had a higher prevalence of moderate/severe paravalvular leak at discharge (10.4%, 9.4%, 7.2%; P = 0.04) and 1 year (9.0%, 7.3%, 5.7%, P = 0.03). They also had lower mean gradient at discharge (10.3, 10.4, 11.8 mm Hg; P < 0.01). 

The best news was that older patients fared better than predicted, Don said. “STS is very sensitive to changes in age, such that the older patients are given higher STS-predicted mortality.” 

Operative mortality turned out to be approximately 30% lower than expected for patients in the two older groups. Kaplan-Meier estimated survival curves were similar among the three age groups through 4 years. Strikingly, unlike younger TAVR patients, those aged at least 90 years “had survival equal to what is expected by age matched US population estimates,” Don said. 

Different Priorities for the Eldest Patients

Asked by an attendee whether he now advises elderly patients who are less symptomatic or asymptomatic to undergo TAVR, Don was circumspect. Since the PARTNER patients were, by design, symptomatic, the current analysis cannot directly answer that clinical question, he said. 

“This is actually I think the subject of the study that Columbia is going to be spearheading, looking at more asymptomatic patients. There were some recent retrospective studies suggesting there might be benefit in these patients,” but these are not sufficient to justify that choice yet, he said. “There are a lot of 90-year-olds who don’t have symptoms and their quality of life is very good. It’s unclear what TAVR would be offering them, if [for example] they’re going to have a hip fracture in the next 2 years and have that comorbidity.” 

The discussion over what to do in the eldest patient uniquely does not hinge on the “5-year durability argument” but rather on acute procedural risk and quality of life, said comoderator Susheel K. Kodali, MD (Columbia University Medical Center, New York, NY). Based on clinical outcomes, the study “does support continuing to treat those patients,” he concluded. “Whether it supports it based on resource utilization and healthcare cost to society [level] is a separate question.” 

Comoderator Raoul Bonan, MD (Montreal Heart Institute, Canada), took issue with the durability argument, stressing that it no longer holds water in 2016. “We have enough data now that these two valves we’ve been using now for more than 10 years have some durability,” he commented. 

Durable yes, Kodali agreed, but for “how long?” TAVR only has 5-year data, he said. “Surgical valves have data for 15 years. So I think that’s really where the argument [lies]. I’m not saying I agree with it or not, but that’s something at least in the discussion with the patient occurs.” 

Bonan countered that, in the absence of negative findings on durability, the argument that was “valuable” in 2007 or 2009 should be put to rest. 


  • Don CW. TAVR age paradox: oldest patients have better than expected outcomes in the PARTNER Study. Presented at: American College Cardiology 2016 Scientific Sessions. April 2, 2016. Chicago, IL.


  • Don reports receiving consultant fees/honoraria from Edwards Lifesciences and Medtronic.
  • Kodali reports receiving fees/honoraria for consulting for Edwards Lifesciences, Caret Medical, Meril Lifesciences, and other fees from Thubrikar Aortic Valve Inc and VS Medtech
  • Bonan reports receiving consulting fees/honoraria from Medtronic. 

Related Stories: