乌鸦传媒

A new scoring system that identifies the best candidates for transcatheter aortic valve replacement (TAVR) -- reflecting a more modern rethinking of various risk factors -- still had a few flaws, experts argued.

The point-based model incorporated frailty and disability as predictors of 30-day and 1-year mortality. The latter had a C-statistic of 0.79 (95% CI 0.59-0.95), , of St. Vincent's Heart Center of Indiana in Indianapolis, and colleagues originally reported in July 2016.

International groups seemed to welcome the new tool in several letters published in the January 31, 2017 issue of the Journal of the American College of Cardiology. Nevertheless, they noted several points of contention.

"The authors have appropriately emphasized the need for models developed entirely from a TAVR population. Their recognition of the importance of frailty and 'disabilities' is a particularly important aspect of risk modeling in the TAVR population, and we are gratified to see their attempts to include these elements in their models," wrote , of University of Florida College of Medicine, Jacksonville, on behalf of the Transcatheter Valve Therapy (TVT) Registry Steering Committee.

Real World Experiences?

But Edwards identified a weakness of Hermiller's model: it was derived from the high-risk cohort of the CoreValve U.S. pivotal trial that used Medtronic's CoreValve. This meant having a small study population, which was, in turn, perhaps linked to the wide C-statistic confidence interval.

"It should be mentioned that the model does not appear to predict mortality but rather the probability that patients will be stratified to a low-, medium-, or high-risk group," he added.

A response by Hermiller and colleagues in the same journal had the group maintaining that they checked the performance and validity of their model via several methods, though they agreed that they should test it in larger cohorts and with non-CoreValve devices.

In contrast, an alternative TAVR risk scale from Edwards and his TVT Registry colleagues included the TAVR experience of all commercially available devices in the U.S. "The TVT Registry model was designed to predict in-hospital mortality. It was developed from a 'real-world' population of 13,718 consecutive patients and validated by comparison with 6,868 patient records not used in model development," according to Edwards.

"Discrimination was better than previously reported models and calibration was excellent," he emphasized, noting that a 30-day mortality model -- including frailty metrics -- is on its way.

What About Aortic Regurgitation?

Separately, a Chinese group took issue with the exclusion of aortic regurgitation in Hermiller's scoring system.

"As we know, many [researchers] have confirmed that aortic regurgitation after TAVR is a predictor for poor outcomes including increased rates of rehospitalization, death from cardiac causes, and death from any cause at 1 year. This means the patients with aortic regurgitation after TAVR have a much higher mortality rate," wrote Xinqun Hu, MD, PhD, of The Second Xiangya Hospital of Central South University in China, and colleagues.

"However, in this study, the authors do not put this important factor into their research design, which will increase the bias and make the result not so convincing. It is better to clarify how many patients [had] aortic regurgitation after TAVR in this study and evaluate the impact of [these] patients on mortality during follow-up."

Post-TAVR aortic regurgitation may be an important factor in survival, but is that information available for TAVR candidates in the first place?

"The purpose of our risk score was to provide clinicians guidance during the patient screening and selection process, in order to guide patient discussions and provide insight into mortality risk that may not be completely defined by the standard risk scores such as the Society of Thoracic Surgeons' Predictor of Mortality score," Hermiller's group said.

"Therefore, we focused our model on only those clinical characteristics that would be available during the pre-procedural assessment."

Why Just Mortality?

It was a big step to include frailty when calculating the odds of death among elderly TAVR candidates, but mortality shouldn't be the end-all in the decision-making process, a Spanish group argued.

"The proposed score is focused on mortality, yet the concept of futility goes beyond this concept of mortality. In fact, it has been reported that a significant percentage of elderly patients do not see improvements in their overall health status after [TAVR]," according to , of Spain's Hospital Regional Universitario, and colleagues.

"As such, it is important to include some functional, cognitive, or quality of life parameters in the score in order to adequately gauge the impact of [TAVR] on the elderly, a concept that has been proposed by other authors," they wrote, adding that it is also "essential to consider the preferences of patients and/or their caregivers in a shared decision-making framework."

"Our risk score was designed to predict risk before the procedure with the assumption that the patient has been adequately prepared for the procedure," Hermiller and colleagues argued.

Even so, the creators of the proposed score agreed that quality of life is important for the elderly considering TAVR. Clinicians should strive for a heart team approach that incorporates the patient's desires, they emphasized.