TAVR in the Real World: The Initial 18 months' Experience in the United States
Abstract & Commentary
Jeffrey Zimmet, MD, PhD
Associate Professor of Medicine, University of California, San Francisco, Director, Cardiac Catheterization Laboratory, San Francisco VA Medical Center
Dr. Zimmet reports no financial relationships relevant to this field of study.
Source: Mack MJ, et al. Outcomes following transcatheter aortic valve replacement in the United States. JAMA 2013;310:2069-2077.
Transcatheter aortic valve replacement (TAVR) was introduced commercially in the United States following the November 2011 FDA approval of the Edwards SAPIEN device. This approval was based on the PARTNER trial, a single, multicenter, randomized, controlled trial that compared TAVR with standard therapy in patients with inoperable and high-risk aortic stenosis. It is well-recognized that results obtained in such trials by selected high-volume centers with expert teams under controlled circumstances may not be matched once the device is approved and moved to a more "real-world" setting. The American College of Cardiology and the Society of Thoracic Surgeons have collaborated to create a process for rollout of this device, including not only training and credentialing of operators and centers, but also a national registry for postmarketing surveillance. Mack and colleagues report the results from this Transcatheter Valve Therapy (TVT) registry in this article.
Between November 2011 and May 2013, results were obtained from 7710 U.S. TAVR cases performed at 224 centers. Of the procedures reported, 20% were in patients who were considered "inoperable" in terms of open surgical AVR, and 80% were considered high-risk but operable. The median age of patients was 84 years, and 49% were female. The mean STS predicted risk of operative mortality (STS PROM) was 7% in both high-risk and inoperable groups. A large proportion of patients pre-procedure had advanced heart failure, with 81% reporting NYHA class III or IV heart failure symptoms.
In terms of the procedure itself, the transfemoral approach was the most common route of access (64%), followed by the transapical approach (29%). The procedure resulted in successful device implantation in 92% of cases. More than one transcatheter valve was used in 2.9% of procedures. Death during the procedure was uncommon (0.8%), as was conversion to open aortic valve surgery (1%). Despite the low intraprocedural death rate, in-hospital mortality was significantly higher at 5.5%. Major complications tracked in the registry included stroke (2.0%), major vascular injury (6.4%), acute renal insufficiency (5.5%), major bleeding (3.5%), and need for new pacemaker or ICD (6.6%). Median hospital stay was 6 days. At 30 days, the incidence of death rose to 7.6%, stroke occurred in 2.8%, and aortic valve reintervention was necessary in 0.5%. The authors concluded that in a U.S. postmarket registry, TAVR results are comparable to prior published, controlled trial data.
Commentary
Surgical aortic valve replacement (AVR) is the second most common cardiac surgical procedure performed in the United States. Multiple studies have demonstrated that well over 30% of patients with severe aortic stenosis are denied AVR due to a high risk of surgical complications. These numbers do not necessarily include patients who are never referred for surgery, either because their physicians believe they are too high-risk or because the patients themselves will not consider open heart surgery. In this vein, TAVR is a transformative procedure, opening the door to many patients who would not otherwise be treated.
Following FDA approval, the number of U.S. medical centers performing the TAVR procedure rapidly expanded from the 35 PARTNER hospitals to 224 clinical sites. How do the results reported in the TVT "real-world" registry compare with the results of the tightly controlled PARTNER trial? In terms of the patients themselves, the risk profile of patients reported in the registry appears lower than that in the trial. The STS-PROM in the registry was only 7% for both high-risk and inoperable patients, while PARTNER reported scores of more than 11%. While it is well recognized that the STS calculator fails to capture many elements of operative risk, this likely represents some element of so-called "risk creep," whereby the procedure is offered to somewhat lower-risk patients over time. Despite this, mortality rates were similar to those reported in PARTNER. Among those receiving the procedure via the transfemoral route, 30-day mortality rates in the inoperable and high-risk patients were 6.7% and 5.0%, which are comparable to the rates of 5% and 3.7% reported in similar groups in PARTNER. The 30-day mortality rate among nontransfemoral patients in the high-risk subgroup was 10.8%, which is slightly higher than the 8.7% rate reported in the as-treated analysis of PARTNER. In-hospital and 30-day stroke rates from the registry, reported at 2.3% and 2.5%, compared favorably with the 4.7%, 30-day stroke rate from the PARTNER inoperable group.
Notably, major vascular complications and major bleeding were significantly lower in the registry compared with the trial (6.4% and 3.5%, compared with 11% and 9.3%), despite use of the same first-generation TAVR devices with large sheath sizes. This most likely represents a combination of better patient screening and a shortening of the learning curve from both worldwide experience and the use of a mandated educational program run by the device manufacturer.
Overall, this report demonstrates the success of the U.S. strategy of controlled dissemination of TAVR technology. As experience grows and devices improve, continued reporting of data at this level of detail will assist in further refinement of the process and will influence future recommendations for TAVR use.
