Inhospital mortality leaps after TAVR commercialization

Outcomes for patients undergoing traditional surgical and less invasive treatments for aortic stenosis have improved over time, but a recent uptick in the inhospital mortality rate with transcatheter aortic valve replacement (TAVR) may raise concerns.

“Post-commercialization we are seeing that transcatheter aortic valve replacement results are still reasonable compared to surgery but they are substantially worse than they were in the trials,” J. Matthew Brennan, MD, MPH, of Duke University Medical Center in Durham, N.C., told Cardiovascular Business. He presented the results Jan. 27 at the annual meeting of the Society of Thoracic Surgeons in Orlando, Fla.

Brennan and colleagues analyzed data from the STS Adult Cardiac Surgery Database from January 2008 to June 2013 as well as the STS/American College of Cardiology Transcatheter Therapy Registry to evaluate case volume and outcomes at hospitals. They looked at trends over time for TAVR and surgical AVR procedures for patients classified as low risk (149,307 patients), intermediate risk (50,571 patients) and high risk (27,280 patients).

The study period included the FDA approval of the Sapien TAVR valve (Edwards Lifesciences) in November 2011 for inoperable patients with severe aortic stenosis and October 2012 for high-risk patients.

Over the study period, inhospital mortality declined in all risk groups, with the greatest drops seen in the high-risk surgical AVR patient group and at TAVR centers. Risk-adjusted analyses showed the observed-to-expected ratio of inhospital mortality for surgical AVR high-risk patients dipped close to TAVR results by the second quarter of 2013. That was in part due to a jump in TAVR mortality after the second quarter of 2011.

“We see an increase in adjusted risk of inhospital mortality comparing before vs. after commercialization,” Brennan said. “The observed-to-expected ratio goes from 0.36, indicating a much better observed inhospital mortality than predicted, to 0.61. It almost doubles, so you have a drastic change from pre-FDA approval to post-FDA approval.”

At a press conference, Brennan emphasized that patients are experiencing better outcomes with the availability of TAVR technology. He offered several possible reasons for the uptick in inhospital mortality. With many centers in addition to those in clinical trials coming on board, newer participants may be experiencing a learning curve; they may not have the volume to be as proficient as the trial centers; they may not be as well equipped to handle emergencies; and they may be selecting a different type of patient.

The score used to evaluate patient risk has been validated for surgical AVR and not TAVR, which also may a contributing factor.

“In the trial we developed very strict inclusion-exclusion criteria,” Brennan told Cardiovascular Business. For instance, patients on dialysis were excluded in the trials but may be considered for TAVR in current practice. “It is opening up to other types of patients. Once we risk adjust for most of the characteristics that you would expect would make a difference, though, we are still seeing mortality rates increasing, suggesting it may be more than just patient selection.”

At the press conference, he suggested that the mortality trend may correct itself with less experienced or lower volume centers gaining experience and with strategies such as providing guidance to help them select more straightforward cases. “There has been an unprecedented training effort behind this device and this technology,” he said. “Recognition of this increase should redouble those efforts.”