TAVR: Raising the Bar to Achieve Good Outcomes

The FDA approved two types of devices for transcatheter aortic valve replacement (TAVR) after determining they were better than standard therapy for inoperable patients and an alternative to open surgery in high-risk patients. Nonetheless, this generally elderly patient population poses challenges and there are still potential missteps that could derail TAVR. Patient selection tools, 3D imaging and other approaches may minimize the likelihood of complications and poor outcomes occurring.

The frailty factor

Patient selection may not be the be-all, end-all for achieving good outcomes with TAVR, but it can go a long way toward maximizing the chances. Identifying those patients with severe aortic stenosis, whether inoperable or at high surgical risk, who truly will benefit from the procedure is critical to success. But given the patient population, physicians may wrestle with myriad clinical and patient factors when making their assessments.

“The centers and teams evaluating patients for transcatheter aortic valve replacement are recognizing more and more the complexity of these patients,” says Brian R. Lindman, MD, MSCI, of Washington University School of Medicine in St. Louis. “We have patients who we’re evaluating in their 80s and 90s who have multiple comorbidities. Not only that, in many cases they are frail.”

In PARTNER I (Placement of Aortic Transcatheter Valve), the pivotal trial for Edwards Lifesciences’ Sapien valve, patients in the inoperable cohort were a mean age of 83.1; 92.2 percent had New York Heart Association (NYHA) class III or IV symptoms; 21.2 percent had oxygen-dependent chronic pulmonary obstructive disorder; and 18.1 percent were considered frail (N Eng J Med 2010;363:1597-1607). The mean age for patients in Medtronic’s extreme risk trial for its CoreValve device was 83.2 years; 92 percent were NYHA class III or IV; 23.5 had severe lung disease; and many were frail (J Am Coll Card 2014;63[19]:1972-1981).

“When we talk to these patients, it is clear that they don’t feel well but it is not so clear how much of their quality of life is influenced by their valvular stenosis,” Lindman says. “We are trying to weigh, does this patient have severe aortic stenosis as the dominant cause of their poor health status or is the dominant cause for their poor health status their clinical and geriatric comorbidities?”  

Lindman and colleagues offer a framework for assessing TAVR candidates based on clinical risk stratification, geriatric risk stratification, expected clinical benefit and patient preferences and goals (J Am Coll Cardiol Intv 2014;7[7]:707-716). Clinical risk factors—a very high Society of Thoracic Surgeons score, severe concomitant valve disease, comorbidities and the like—have been emphasized in the past but physicians may not recognize the importance of geriatric risk factors. Those include frailty, impaired mobility, disability, low muscle mass, cognitive impairment and mood disorders.

Frailty may be the tipping point between good and bad outcomes in some patients. Candidates may appear low risk for TAVR based on clinical factors but their frailty may leave them with little reserve during recovery, Lindman says. “We are recognizing more and more that these issues that fall under the umbrella of frailty are important to try to more objectively characterize so that we can better estimate and anticipate risk and potential benefit of these procedures.”

Looking beyond survival

Survival alone may not be the ultimate goal for patients seeking TAVR, some physicians argue. Very elderly patients may value quality of life over longevity, wishing to get the best out of their remaining time with family and friends. TAVR that extends life but at the cost of a poorer quality might seem a bad tradeoff. Physicians have reliable tools to navigate the clinical terrain but they may end up bushwhacking when it comes to measuring a patient’s expectations.  

“Understanding the creatinine or their valve grading or whatever, that is so easy for us,” says Suzanne V. Arnold, MD, MHA, a cardiologist at Saint Luke’s Mid America Heart Institute in Kansas City, Mo. “We do that every day. But understanding what their quality of life is, and what it will be, and why that is so important, is more difficult.”

Arnold and colleagues developed and tested a model to predict poor outcomes in TAVR patients that included quality of life and clinical factors using data from the PARTNER cohorts A and B as well as the access PARTNER registry (Circulation online May 23, 2014). They determined that shorter distance on a walking test and higher mean aortic valve gradients were associated with poor outcomes. Clinical factors included oxygen-dependent chronic lung disease, renal dysfunction, cardiac arrhythmias and cognitive decline.

Their model identified 10 percent of the patient population with a more than 50 percent likelihood at six months after TAVR of a poor outcome—defined as death, a summary score of less than 45 on the Kansas City Cardiomyopathy Questionnaire or a decrease of more than 10 points from the baseline quality-of-life score. At one year, 8 percent had a more than 70 percent likelihood of a poor outcome. 

Arnold emphasizes that the model is not meant to be a mechanism for withholding treatment; while it identified those at high risk for bad outcomes it could not with certainty determine who would have a bad or good outcome. Instead, it may provide a starting point for physicians and patients to have an informed discussion about options and realistic expectations.

“If you tell them, ‘You have a really high risk of not doing well and not feeling any better than you did before the procedure,’ I am not sure they would want to go through it,” she says. 

Round peg, oval hole

Experience, a growing body of knowledge and technological advancements all are helping to make TAVR a safer procedure for patients. “All the complications are decreasing, so that TAVR is moving in the right direction,” says Samir R. Kapadia, MD, director of the Sones Catheterization Laboratories at the Cleveland Clinic and a PARTNER investigator.

One of those complications is paravalvular regurgitation, which post-procedure goes lock step with a higher risk of mortality. In an analysis of 45 TAVR studies, Kapadia and colleagues estimated the incidence of moderate or severe aortic regurgitation was 11.7 percent, was frequently paravalvular, and the risk was higher with CoreValve (J Am Coll Cardiol 2013;61:1585-1595). One common cause for paravalvular leak was undersizing the valve. 

Early protocols called for 2D echocardiography imaging to measure the annulus to determine appropriate valve sizing. The catch is, the annulus typically is elliptical, not circular, as shown by 3D transesophageal echo (TEE) and CT but not by 2D echo.

“3D echo allows you to get to the actual plane of the annulus,” explains Rebecca Hahn, MD, director of invasive and valvular echocardiography at Columbia University Medical Center/NewYork Presbyterian Hospital in New York City. “Otherwise, you are measuring a single dimension off a 2D plane.”

Compared with 2D TEE, 3D TEE and CT provide more accurate information to guide patient selection and sizing. For instance, incidence of paravalvular regurgitation in patients treated with the Sapien valve and imaged by 3D TEE was 1.4 percent vs. 10.3 percent for 2D TEE (J Am Coll Cardiol 2013;61[9]:908-916). Another CT-based analysis study found that almost half of patients imaged by 2D TEE and implanted with CoreValve devices received the wrong valve size (J Am Coll Cardiol Interv 2014;7:652-656).

“No one should use 2D,” Hahn advises. “All the annuli are typically oval.”

While concerns over paravalvular regurgitation may be abating, stroke remains “probably the most important complication that we don’t have a good handle on,” Kapadia says. In PARTNER, for instance, stroke rates for both inoperable and high-risk groups exceeded the controls. 

An analysis of 25 registries and 33 single-center studies by Kapadia and colleagues found a 30-day pooled stroke rate for Sapien valves of 3 percent in registries and 3.2 percent in single center studies—an improvement from PARTNER results (J Am Coll Cardiol 2014;63:2101-2110). Core-Valve’s rates were 2.4 percent and 3.8 percent, respectively.

Registry data suggest the rate in the U.S. may be as low as 2 percent, but there may be ways to chip the rate lower (JAMA 2013;310[19]:2069-2077). Filters placed in the carotid artery might capture or deflect embolic material before it reaches the brain, and several trials are underway to evaluate these devices.

Less is more

TAVR procedures using a minimalist approach may offer another opportunity for reducing risk, advocates propose. The minimalist approach calls for fully percutaneous transfemoral TAVR procedures that are done in a catheterization lab with local anesthesia, minimal sedation and transthoracic echo rather than in a hybrid operating room with general anesthesia and TEE.

“If you can do them [TAVR patients] transfemorally, it is the fastest recovery, the least morbidity on the patient, least complications and you can do them awake, which means a shorter length of stay and lower costs,” says Vasilis Babaliaros, MD, co-director of the Structural Heart and Valve Center at Emory University Hospitals in Atlanta.

Babaliaros and the TAVR team at Emory compared costs and outcomes for 142 patients who underwent elective transfemoral TAVRs at their center between November 2010 and September 2013. The center initiated a minimalist approach for transfemoral TAVRs for both inoperable and high-risk patients in May 2012 (J Am Coll Cardiol Intv online July 30).

Thirty-day mortality was not significantly different between the two groups, nor was survival at 435 days follow-up. The 30-day major stroke rate was 1.4 percent with the minimalist approach vs. 0 and the rate of moderate or severe paravalvular leak was 3 percent vs. 5.8 percent.

The tally showed procedure time, total intensive care unit (ICU) time and length of stay lower with a stripped down TAVR for a difference in costs of approximately $10,000. “It looks like it is just as safe but it saves more money,” Babaliaros observes, adding that only experienced centers should initiate such programs. Other benefits might include reduced risk of infections, no swallowing problems that result from intubation with general anesthesia and possibly even the ability to recognize and intervene during a stroke.

“You have no idea if they had a stroke when they were asleep until they wake up from anesthesia,” Babaliaros says. “If I see an on-table stroke, we can take that patient immediately to our stroke center and have them remove the embolism. Theoretically, you could take a devastating problem and take care of it on the table.” 

Introduction of smaller-profile devices such as Edwards’ Sapien XT likely will widen the pool of candidates for transfemoral TAVR who could be treated using a minimalist approach, he projects. But it is not for every transfemoral patient: People who are obese, can’t lie flat for the duration of the procedure or have breathing difficulties would not be suitable.

Hahn adds that under a minimalist approach with transthoracic echo, physicians may sacrifice the ability to anticipate and diagnose complications during a TAVR procedure. Her concerns are supported by research. An analysis of French registry data that compared the use of general and local anesthesia with TAVR found overall similar outcomes but attributed local anesthesia’s  higher risk for postprocedural aortic regurgitation to less frequent use of TEE during the procedure (Circ Cardiovasc Interv online July 8, 2014).

And Hahn points out that the stakes are even higher as TAVR proponents push to lower the bar of entry to include intermediate-risk patients with severe aortic stenosis. “The intraprocedural imaging by TEE is invaluable, particularly when you move toward the intermediate risk group in which a single death in 100 will reduce our survival and we will no longer be comparable to surgery,” she cautions. “One would hope our outcomes might be in this intermediate group as good as surgery.”