In the not-so-distant past, assessing the best candidates for transcatheter aortic valve replacement (TAVR) was pretty straightforward. They tended to be the sickest and oldest patients, those judged physically unable to withstand the rigors of invasive surgery for their severe aortic stenosis. But with TAVR finding a new comfort zone in younger, healthier patients, the calculus is changing and determining the most eligible recipients for the replacement valve is becoming a much more nuanced—and challenging—exercise.
“Just trying to determine who shouldn’t get TAVR is now the hardest question,” says Kyle Buchanan, MD, an interventional cardiologist with AMITA Health Medical Group Heart and Vascular Center in Hinsdale, Ill. “It’s becoming a matter of what the patient’s goals are: Do they want to feel better, live longer or both?”
A flurry of risk scoring systems and calculators—many of them app-based—has been introduced in recent years as the field has attempted to keep up with the TAVR groundswell. They include the American College of Cardiology (ACC)/Society of Thoracic Surgeons (STS) TAVR In-Hospital Mortality Risk App, the TAVR 30-day readmission rate calculator (TAVR-30, by makeadent.org) and various frailty scales and indices designed to help risk-stratify TAVR candidates. But as less invasive valvular surgery continues to evolve in terms of its technology, safety and patient set, are these tools measuring the right variables and targeting the right outcomes?
“If a patient with a 20-year life expectancy after valve replacement is at higher risk for mild paravalvular leak, does that portend higher clinical significance than it would for a higher risk patient whose life expectancy may be only three to seven years?” asks Chandan Devireddy, MD, an interventional cardiologist at Emory University Hospital Midtown in Atlanta and a member of the Society for Cardiovascular Angiography and Interventions Structural Heart Disease Council. “I think the next step in studying TAVR is going to be better targeting decision-making specific to low-risk patients.”
DEVELOPING THE NEW METRICS
In the new cosmos of structural heart disease (see our accompanying article: As Surgery Gives Way to Transcatheter Procedures, Is the Cardiology Cash Cow in Jeopardy?), ease of use and utility are becoming catchwords for predictive model development. To create their 30-day readmission calculator, the first tool of its type, a multi-institution research team seized on a metric associated not only with greater healthcare expenditures and resource utilization, but with poorer long-term outcomes. The measure was particularly appropriate in the case of TAVR, which, the team found was responsible for a nearly 18 percent 30-day all-cause readmission rate, using the Nationwide Readmissions Database (NRD) (EuroIntervention 2019;15:155-63).
“We realized that identification of patients at greatest risk of readmission within 30 days could result in the intensification of post-discharge care so that perhaps the patient would see a cardiologist within a week instead of two to four weeks, or the nurse would call the patient within three days to make sure they’re taking their medication,” says Sahil Khera, MD, MPH, assistant director of the structural heart program at Mount Sinai Medical Center in New York and lead author of the EuroIntervention study outlining the methodology behind the 30-day readmissions calculator. Available as an IOS app, the tool is a weighted point system for determining the probability of 30-day readmission based on nine clinical variables frequently matched to the metric, including anemia, chronic lung disease, chronic renal disease, chronic liver disease, atrial fibrillation, acute kidney injury and length of hospital stay.
The ACC/STS TAVR In-Hospital Mortality Risk App approaches risk assessment from more of an outcomes perspective. Available free from various app stores, the tool allows clinicians to evaluate through variants closely associated with dying the mortality risk of patients considering TAVR and compare it to the national average based on TVT Registry data. “Even as the likelihood of good outcomes improves from the lower and lower risk groups we’re now seeing, you still want to get a handle on mortality so you can do everything possible to reduce risk in these patients,” explains John Carroll, MD, director of interventional cardiology at University of Colorado Hospital and co-chair of the ACC/STS workgroup that introduced the first mortality risk calculator for TAVR in 2015 and continues to update the app through data collection.
Carroll’s group has found a timely new application for the registry-generated mortality data: measuring the performance of hospitals performing TAVR across the country, thus joining the ongoing debate over CMS site and operator volume requirements for starting and maintaining a TAVR program. By developing models that can weigh and compare site performance, Carroll’s group is attempting to move the discussion from just volume metrics to actual outcomes. “Our thrust is to help [TAVR] sites assess their performance and identify ways they can improve,” he observes.
WEIGHING QUALITY OF LIFE
As TAVR becomes the default for aortic valve replacement, the question of which patients are the best candidate turns increasingly on the issue of post-procedure quality of life. Simply put, is the procedure worth it for patients already struggling with advanced stages of frailty, dementia, cognition deficit, arrhythmia, diabetes and other comorbidities?
Buchanan has studied these issues and firmly believes that frailty should be a top-of-the-mind consideration. “Being frail shouldn’t necessarily exclude someone from having the procedure,” he allows, “but it does provide the patient and physician with more information on what the prognosis may be.” His research found that patients deemed “frail” based on both a full frailty assessment and an STS risk score were five times more likely to die within 30 days of TAVR and 2.75 times more likely to die within a year when compared to patients with an STS risk score of 8 or above. Frailty was judged on the presence of three of five criteria: body mass index below 20 kilograms per square meter; serum albumin below 3.5 grams per deciliter; Katz Index of independence in daily living activities of less than 4/6; low grip strength; and a slow 15-foot walk time (Am Heart J 2018;200:11-16).
Are TAVR candidates being adequately sized up today for frailty? “I certainly think physicians try to pay attention to frailty,” Buchanan says, “but is it objectively measured for every patient? Definitely not.”
Part of the problem, he concedes, is that frailty testing is a luxury not all hospitals can afford. “It takes 15 to 20 minutes to perform and document the test results, and while this might not seem like much effort, it becomes cumbersome and difficult when patients are only allotted 15 to 30 minutes per clinic visit,” Buchanan explains. And without the time or resources to ask all the questions and conduct all the tests needed to comprehensively measure the condition, he adds, clinicians typically rely on an admittedly subjective standby—the eyeball test.
“Frailty has always been the elephant in the room when assessing surgical risk,” acknowledges Devireddy, “and the method of assessment has varied from center to center. This may not play as big a role in stratifying low-risk patients, but for intermediate- and high-risk patients the assessment may as quick as doing albumin, hemoglobin and walking tests, and assessing cognition. Perhaps we need more specific tools in low-risk patients, however, to predict poorer outcomes or alternative treatment choices.”
CHARTING THE FUTURE OF RISK ASSESSMENT
If experts in the field agree on one thing, it’s that risk calculation is still in its infancy, primed to grow apace with TAVR itself. As Buchanan maintains, TAVR hasn’t been around long enough and generated enough data to definitively say which metrics should be universal. “We don’t even have enough long-term data,” he says, “to talk with certainty about the durability of TAVR valves,” a particularly important issue with younger patients who may need 20 or more years of usability from their device. “The technology is just too young.”
What is taking tangible shape, however, is the format for future programs and devices. Risk algorithms that reside on physicians’ smartphones and can be easily accessed to obtain with minimal input a valid risk score for TAVR patients are becoming the gold standard. Younger physicians who grew up in the digital age are particularly at home with this mode. Tools that are linked and constantly updated to patient electronic medical records also are certain to prevail.
While metrics and imaging are two important pillars of any effective TAVR evaluation, what shouldn’t be lost in the process, Khera says, is a decidedly low-tech criterion—a sitdown with the patient and family. “You can’t make a decision without really understanding what the patient’s goals are in life and how they might benefit, or not benefit, from this surgery,” he explains. “And that discussion must include family as well as patient expectations.”