Improving Patient Selection for TAVI: A Work in Progress

Pinpointing ideal patients for TAVI is next step to improving outcomes

October 27, 2011

October 27, 2011 (London, United Kingdom) — How to best identify the most appropriate patients for transcatheter aortic-valve implantation (TAVI) to improve outcomes was the subject of much discussion at the PCR London Valves 2011 meeting last week.

The prevailing view was that better risk scores need to be developed, including a TAVI risk score, to help select the right candidates, as the ones that are currently used — generally surgical scores such as the EuroSCORE and Society of Thoracic Surgeons (STS) — do not lend themselves well to predicting outcomes in a percutaneous procedure.

Speakers said it will be important to home in on three particular patient groups: first, those who are too sick for surgery but also too weak to receive TAVI need to be identified; second, for those who are very high risk for surgery and therefore possible candidates for TAVI, more cost-effectiveness data are needed; and finally, in the lower risk, or so-called intermediate, patients — who in some countries are already undergoing TAVI — other issues will take precedence, such as demonstrating long-term durability of the transcatheter aortic valves and ensuring that adverse events are at least equivalent to those from a surgical procedure. Two new trials in this intermediate-patient population, with the two TAVI devices currently on the market in Europe, are about to start any day.

And all agreed that as well as looking at 30-day mortality after TAVI, much more attention should be paid to one-year death rates, as it is becoming apparent that there is a huge attrition rate in terms of mortality in the 11 months following the completion of the procedure.

Identify "cohort C": Some patients are too high risk, even for TAVI

Currently, TAVI is only officially indicated in elderly patients who are deemed inoperable — ie, they would not be considered for surgical aortic-valve replacement (SAVR) — or in those who are considered very high risk for SAVR.

The way in which patients are assessed for TAVI differs from country to country but usually involves a surgical risk score, which takes into account comorbidities and age, although surgeons and cardiologists often feel they can use the "eyeball" or "end-of-the-bed" test to identify the most suitable intervention for a patient, Dr Todd M Dewey (Medical City Dallas Hospital, TX), a surgeon, told the London meeting. "We think we know it when we see it, but although this gives us something," in fact, it is not sufficient, he remarked.

However, surgical risk scores are poor at predicting mortality at 30 days in TAVI — the STS score "tends to underestimate" it and the EuroSCORE "wildly overestimates it," Dewey explained. Newer scores are needed, and there is hope on the horizon in terms of a new UK model for predicting risk in TAVI (a TAVI risk score) and EuroSCORE II, "which, because it's in press, we can't get access to the 'under-the-bonnet' specifics, but this is a hugely impressive project that should give us a much better handle on things," he noted.

What is becoming increasingly evident, Dewey and all of the other presenters agreed, is that TAVI was initially being performed on inoperable patients who were too sick to tolerate even a percutaneous intervention.

It's becoming more and more our job not to do the patients we did earlier.

Commenting on this, surgeon Dr Michael Mack (University of Texas, Dallas) observed that although TAVI improved all-cause survival by 20% over medical therapy in the PARTNER cohort B trial, in which inoperable patients were randomized to best medical therapy or TAVI with the Sapien valve (Edwards Lifesciences), death rates in the TAVI group at 12 months were far from inconsequential.

Dr Michael Mack

"It is somewhat concerning that [TAVI] had a mortality of 30.7% at one year. We can improve the one-year outcome by better patient selection," Mack observed.

Everyone concurred that, going forward in the moving field that is TAVI, it will be important in these inoperable patients to identify those who are "cohort C" — dying "with" aortic stenosis but not "from" aortic stenosis — as opposed to cohort B.

Dr Martyn Thomas (St Thomas' Hospital, London, UK) observed, "We need to identify the extreme risk group, those who should not get TAVI," and Dr John Webb (St Paul's Hospital, Vancouver, BC) agreed. "It's becoming more and more our job not to do the patients we did earlier."

How to assess frailty: Not as easy as it seems

In his talk, Dewey spoke specifically about frailty, which is included to some extent in the STS score but not the EuroSCORE; interestingly, the former seems to be slightly better at predicting long-term outcomes from TAVI than the EuroSCORE, other speakers observed.

Although frailty has been shown to be a predictor of reduced survival in TAVI, it is not always a clear-cut indicator, Dewey said, observing, "Some frail patients do well." And it is not always apparent whether being frail should steer patients away from SAVR to TAVI or between TAVI and no TAVI, he remarked.

He explained that frailty is a function of two processes, chronic medical illness and age, but emphasized there are two different ages for a person — their chronological age and their biological age — and one does not always reflect the other. Frailty is currently assessed using measures such as gait speed (5-m walk), grip strength, and activities of daily living, but other biological markers are also increasingly employed, including total bilirubin, albumin, and pulmonary-function tests. And it is important not to underestimate the role of psychological resilience, said Dewey, "which may be equally as or more essential than physical resilience and remains untested."

We want to know what's going to happen a year or two down the line, it's just not good enough for us to be able to say we know it when we see it; we've got to quantify it.

"Quantifying frailty continues to be a challenge because it is difficult to extract the signal from the noise in an elderly, sick population," he observed. "We don't have a stranglehold on how to use this information and how much weight to give it."

He and others went on to discuss how they choose patients for TAVI in their institutions, with a couple saying they relied heavily on nurses to help.

Webb explained: "We have a full-time nurse coordinator, and she does a much better job than I do." Mack said, in his institution, "We do objective tests such as STS score and 5-m walk, but our head nurse coordinator also gives us input."

Thomas said that in the SOURCE XT registry, "every patient is being given an 'eyeball' test score of 0 to 10, and it will be fascinating to see how this correlates with outcomes." Mack agreed, noting, "We want to know what's going to happen a year or two down the line, it's just not good enough for us to be able to say we know it when we see it; we've got to quantify it."

Ideal cohort A patient still not identified: TAVI risk score in progress

Moving on to discuss the so-called cohort A patients, the ones in whom surgical risk is very high and some deem would be better served by TAVI, Thomas observed, "To my mind, we have not yet identified in a surgical group the ideal patient to undergo TAVI."

Examining mortality data from the PARTNER cohort A results and registry data between 30 days and a year and even out to two to three years reveals that the majority of these patients die from noncardiac causes, Thomas said, such as pulmonary, renal, or gastrointestinal problems or cancer.

And interestingly, he noted, some predictors of lower mortality have emerged from the same data sets, and these include carotid artery stenosis, hyperlipidemia, and systemic hypertension, "all conditions that concentrate the mind on medical therapy, and I do wonder if we've missed a trick on the use of medical therapy; rather, what we do is a mechanical solution," he commented.

Germane to helping to identify the most appropriate patients will be the development of a TAVI risk score, Thomas added, but this has been fraught with difficulty for two reasons. First, "we just don't have a large enough number of patients" to accurately formulate a TAVI risk score yet, he said, and second, "Have we measured the right parameters? Clearly no; frailty needs to be put in there.

I do wonder if we've missed a trick on the use of medical therapy.

"We still strive to find somewhere in the middle where we will have a mortality advantage of TAVI at a certain measure of risk and that will allow us to identify a cost-effective group of patients who will benefit," he noted.

Mack said the perception has been that surgeons are opposed to TAVI in high-risk surgical patients because of the risk of stroke, but in fact the rate of stroke and death combined was more or less the same in both the surgery and TAVI arms in PARTNER cohort A, in which TAVI was deemed noninferior to surgery for this patient group.

"Patients and their physicians will choose the less invasive therapy every time," Mack stressed, adding, "When your average life expectancy is six years and you spend six months recovering from surgery, that is 10% of your remaining life." However, he stressed that more investigation is needed to drill down into the ideal risk/benefit equation for TAVI in this patient group.

"All of the risk scores become part of the decision," Thomas observed, "but I don't believe a single risk score should preclude treatment. These risk measures can only ever be a guide, and good clinical judgment from a multidisciplinary team should never be underestimated. Can we improve upon patient selection? I'm not sure we've answered this yet."

Thoughts turn to intermediate-risk patients

Thoughts are also turning to TAVI for lower-risk patients, the meeting heard. Two important trials in these patient groups are due to start soon: the PARTNER IIa study with the Sapien valve and SURTAVI with the CoreValve (Medtronic).

Dr Susheel K Kodali (Columbia University Medical Center, New York), who is to be an investigator for PARTNER IIa, told heartwire: "PARTNER IIa and SURTAVI are really going to change the dynamic. Obviously 30-day mortality is important and we have to compete on that, but now we've got to think about durability [of the device], because we are talking about patients who have average survivals of eight to nine years" and longer-term outcomes.

In addition, the issue of stroke will be key in this patient population, says Kodali. "We can't have a dramatically higher stroke rate with TAVI when we get down to patients with an STS of 4. Surgeons are very good in that patient population, so transcatheter has to be just as good. That's where this is moving; that's what everyone is excited about."

Dewey and Thomas are consultants for Edward Lifesciences. Mack reports that some of his travel expenses have been paid by Edwards but no other has no other conflicts of interest. Kodali has served as an advisor or consultant for Edwards Lifesciences and Medtronic.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.