KOLs boost Early Aortic Valve Intervention in Patients with Asymptomatic AS--Despite Dubious Evidence
JAMA Cardiology featured a strong editorial urging cardiologists to intervene early in patients who have aortic valve stenosis and no symptoms. I do not agree.
Sir Robert Hutchison (1871-1960) offered “the physician’s prayer'“
From inability to let well alone, from too much zeal for the new and contempt for what is old, from putting knowledge before wisdom, science before art and cleverness before common sense, from treating patients as cases and from making the cure of the disease more grievous than the endurance of the same, good Lord deliver us.
In November of last year, I offered critical appraisal of an extremely flawed EARLY TAVR trial. Proponents and the company that make the aortic valve designed the trial to be positive.
The question: should a patient with severe aortic stenosis but no symptoms be watched (clinical surveillance) or operated on? In the case of EARLY TAVR, the operation would be with a transcatheter aortic valve replacement (TAVR) not a surgical replacement. (TAVR vs SAVR is a related debate).
But the issue at hand today is early intervention or clinical surveillance of these patients who have no complaints.
This month, three key opinion leaders, from Vanderbilt, Harvard, and Mayo clinic, penned an editorial in JAMA-Cardiology proposing that the time has come for early intervention in asymptomatic aortic stenosis.
It would be a major change in practice. For years, we waited for symptoms before intervening on these patients. Shortness of breath with exertion, angina and fainting are the triad of symptoms from AS.
But the KOLs, two of whom, report financial conflicts of interest with a maker of aortic valves, cite the EARLY-TAVR trial and two other trials to conclude that the
…data supporting prompt intervention are substantially stronger than the data supporting clinical surveillance (CS). In our opinion, the routine and preferred strategy for asymptomatic patients with severe AS should shift from CS (current default strategy) to prompt performance of AVR with some exceptions…
Implications of this proposal
I do not agree with their conclusion. But before I tell you why, I want to set out the significance for this proposed practice change.
The likely result in lowering the threshold for valve intervention will be to increase the number of transcatheter procedures much more than surgical interventions. The picture comes from a recent paper on trends in aortic valve interventions. Blue is surgery; orange is TAVR.
There are many reasons for this trend towards more TAVR. The least nefarious is that patients prefer not having open heart surgery if they can avoid it. The tension comes in offering TAVR to younger patients because there are serious questions about valve durability and the need for re-operation sooner with TAVR vs SAVR. That’s a debate for another day.
Suffice to say that any movement to operate early on AS will lead to greater use of TAVR procedures and valves.
Critical Appraisal of the Evidence for Early AV intervention
I refer you now to my Nov 11 column on EARLY TAVR. In short, the trial compared clinical surveillance to early TAVR in 900 patients. The issue was in the choice of the primary endpoint, which was a composite of death, stroke, or “unplanned hospitalization for cardiac causes.”
The topline results reported a statistically significant 50% reduction in the primary endpoint with the early TAVR group. The sole driver of the primary endpoint was unplanned hospitalization, mostly for symptoms of AS. Death and stroke were not different.
The problem was that this was unblinded trial so patients in the surveillance arm knew that they had severe AS but were not treated whereas the other group got “fixed.” It’s a classic problem with a phenomenon called faith healing and subtraction anxiety.
The key opinion leaders also cited two other trials in support of their call for early intervention.
Trial 1: One is called the EVOLVED RCT of 224 patients with asymptomatic AS and evidence of scar tissue in the left ventricle. This UK/Australian trial compared clinical surveillance to early aortic valve intervention with either TAVR (25%) or SAVR (75%).
A primary outcome event of death or unplanned AS-related hospital admission occurred in 18% of the early group vs 23% of the clinical surveillance arm. That 21% relative risk reduction did not meet statistical significance. (HR 0.79, 95% confidence intervals 0.44-1.43). Death rates were similar.
How did our KOLs deal with a nonsignificant trial? Easy. They wrote that the trial was underpowered with wide confidence intervals, which was true, but does not mean it would have been positive with more data. The KOLs also say the long time from diagnosis to intervention (5 months) favored the control arm. They write that 9 events in the early intervention group occurred before the valve intervention—implying that surgery would have saved patients from these events, which is pure speculation.
Trial 2: The other trial cited in favor of early intervention was—at least—positive. The Serbian AVATAR trial randomized 157 patients with asymptomatic AS to clinical surveillance vs surgical AVR. The primary endpoint of death, MI, stroke or unplanned hospitalization for heart failure occurred 15% of patients in the surgery arm vs 34% in the surveillance arm. This 54% relative risk reduction was statistically significant. HR 0.46 95% CI 0.23-0.90). Death was 50% lower and HF hospitalization was 68% lower in the surgery arm but neither reached statistical significance alone. Curious though was that death due to cardiac causes were identical.
The authors rightly called this evidence “preliminary support” for early surgery. It is preliminary because of the very small numbers of patients and events, as well as, the lack of difference in cardiac deaths.
My Summary
The KOL’s proposal to change practice of patients who complain of nothing--from clinical surveillance to a major intervention rests on dubious evidence. One trial, EVOLVED, is clearly nonsignificant, albeit underpowered. Another trial, the EARLY TAVR, was so biased in design that it was useless for clinical translation. And the third trial, AVATAR, was rightly deemed preliminary by its authors.
Lowering the bar of operating on asymptomatic patients will help the makers of aortic valves, the doctors who implant them and the hospitals that provide care.
But I hope to have persuaded you that the evidence that patients are better off being operated on based solely on an ultrasound image rather than symptoms is insufficient.
The approach to patients with asymptomatic aortic valve stenosis should remain conservative. That is, a good doctor pays close attention to symptoms and intervenes when the patient describes limitations due to the tight valve.
This does require careful questioning and listening, as some patients change their life to accommodate to the low flow through the valve.
I'm speaking as a patient whose consulting physician at the end was Eric Topol, but also as decades-long financial journalist and many would say is known for having a skeptic view on things – so please don't dismiss me!
I have been intrigued with the early trial, but mostly because I was in the watchful/waiting camp with my aortic insufficiency for more than 40 years. And I viewed myself as asymptomatic. (Tho as most asymptomatics will likely tell you, post-surgery they realize they probably were! I exercised regularly and over the decades saw top-notch cardiologists as I moved for my job from coast-to-coast and in between.
But it was my original diagnosis by a member of one of the most prestigious pioneering heart-related families in St. Paul that woke me up to what I'm about to say...
His view when he heard the murmur was, "It's nothing." This was in the early days of HMOs and hearing I had a "murmur" got my attention. After all, it was MY heart and I was learning for the first time in my 20s that it wasn't perfect!
I had no idea what a murmur was, but demanded an ultrasound - and had to fight to get it. From that point on, I watched it annually, with physicians showing varying degrees of interest and concern as I moved from city to city - Chicago, New York, San Francisco, where I was always seen at the top institutions.
As medicine progressed, I received CTs, MRIs - never a TEE, interestingly enough. Every year for a decade one cardiologist did a stress echo. Why a stress echo annually? I don't know. I read and researched to the point that I'm sure my regular cardiologist, in the final few years until surgery, dreaded having to see me and having me ask SO many questions.
All imaging over the past two decades suggested I had a "probable bicuspid valve" with increasingly severe stenosis. This includes imaging in Cleveland pre-surgery.
It wasn't until about six month pre-surgery that I started to have symptoms (it took a suprisingly long time to recovery my breath after walking up a steep hill. On one hand I thought nothing of it, but the more it happened, the more I hoped it wasn't what it turned out to be). By that point, because of also having an aortic aneurism that was edging toward the 5.0 mark, I was on the every 3-month scan plan.
This is where things get interesting...
My cardiologist here in San Diego suggested coming back in yet another three months, but my body suggested I get a second opinion. Thanks to Eric's intro, I had been going back/forth over the years with Lars Svensson at Cleveland. Each year, after reviewing the images, he would suggest I continue to watch it. This time, after seeing the results, he suggested I fly in for further tests and consultations. Within I month I did and his conclusion after reviewing the results, when I suggest we wait maybe eight months until after a grandchild was born was: "I would do this soon." Not because of the aneurism, which was hovering at 4.9, but the valve combined with my symptoms.
A month later I had the full slice-and-dice. I was in good health and the surgery and recovery were surprisingly fantastic.
But here's the kicker: I was 67. When he opened me up, he found I had a unicuspid valve, but that my heart had expanded to an athlete's heart to compensate, which is why I was asymptomatic well beyond when most unicuspid valves would have been replaced. My point: Despite all the great imaging, perhaps because of the way I'm built, they couldn't see that until they got inside.
They also did a bypass of my marginal obtuse artery, but the overall consensus is that it was my valve that was causing my breathlessness after walking up hills, not the blockage.
Moral of my story: I'm among those who watched it like a hawk and educated myself, keeping my fingers crossed at every echo that the measurements of my left ventricle would be stable. I don't believe most people do, and my bigger concern is physicians who discover murmurs and tell the patient, "Don't worry about it," when in the very least they need to get a baseline. In fact, it wasn't until about seven years before surgery, after moving back to San Diego from Connecticut, where I saw a Yale cardioligst, that a doctor even suggested I get a solid baseline on the size of my aorta... which is when we realized IT needed to be watched. That should have been SOP on any bicuspid patient, but nobody ever mentioned it.
I was skeptical of the early trials for the reason you were. I'm a huge fan of Edwards and I and I genuinely think heart failure is a big issue that can be prevented. I'm just dubious that most internists/family med doctors will make the referrals to a cardiologist on discovering a "murmur." And if they do, that most patients will realize WHY they should monitor it going forward. After all, you don't feel aortic insufficiency if it's minor or even moderate – and if like me, you feel (or felt) fine. I simply don't think the risk of heart failure or other ramifications are communicated enough to valve patients.
Again, I speak as a non-med professional but as someone with a keen and very personal interest in the topic.
Thanks for reading, assuming you got this far!
Excellent counterpoint to the recent JAMA Viewpoint paper. The underlying problem is the proliferation of confusing combined primary endpoints that seem designed to increase chances of getting to a "positive" trial rather than revealing unbiased medical truth. EARLY TAVR was particularly egregious - defining TAVR as a therapeutic procedure in the intervention group and an adverse event in the control arm. Subsets of hospitalizations should not be accepted as an endpoint, as they are clearly biased toward intervention. What good is an intervention that reduces CV or heart failure hospitalization but increases non-CV/HF hospitalization? More in the CV community need to call out this problem, or it will continue.