Early Surgery Beats a Conservative Approach in Asymptomatic AS: RECOVERY

PHILADELPHIA, PA—Postponing surgical intervention in favor of conservative medical management in patients with asymptomatic but very severe aortic stenosis leads to increased rates of operative or cardiovascular death as well as death from any cause, according to a new randomized trial.

Duk-Hyun Kang, MD, PhD (Asan Medical Center, Seoul, Korea), who presented the RECOVERY findings here today in a late-breaking session at the American Heart Association (AHA) Scientific Sessions 2019, told TCTMD that these findings “provide the evidence to support preemptive AVR for asymptomatic severe aortic stenosis.”

Commenting to TCTMD, S. Chris Malaisrie, MD (Northwestern University Feinberg School of Medicine, Chicago, IL), who earlier this year published a retrospective study on the topic, said this study “further shows evidence that patients with asymptomatic aortic stenosis have a bad prognosis, and there’s an opportunity for us to treat the correct set of patients with either AVR or TAVR.”

Intervention Wins Over Waiting

For the trial, which was simultaneously published in the New England Journal of Medicine, Kang and colleagues randomized 145 asymptomatic patients with very severe aortic stenosis (defined as AVA ≤ 0.75 cm² with peak aortic jet velocity ≥ 4.5 m/s or mean transaortic gradient ≥ 50 mm Hg) from four Korean institutions to early surgery—within 2 months—or conventional treatment between July 2010 and April 2015.

Surgery was successful in all 72 patients in whom it was attempted, with half each receiving either a mechanical or biological valve. Median time from randomization to surgery was 23 days, and there was no operative mortality in this group.

Of the 72 patients randomized to conservative management, 74% ended up undergoing surgical AVR or TAVR over follow-up, mostly due to the development of symptoms. The median time from randomization to intervention in this cohort was 700 days, and of those who received an intervention, 17% were admitted from the emergency department. Again, there was no operative mortality.

Over a median follow-up of 6.2 years in the early-surgery group and 6.1 years in the conservative-care arm, fewer patients died from cardiovascular causes in the former cohort compared with the latter (1% vs 15%; HR 0.09; 95% CI 0.01-0.67). The number needed to treat to prevent one cardiovascular death over 4 years was 20. This led to an advantage with regard to the primary endpoint (combination of operative mortality and cardiovascular mortality on follow-up) for early surgery compared with conservative management (P = 0.003).

All-cause death was lower in the early-surgery arm as well (7% vs 21%; HR 0.33; 95% CI 0.12-0.90).

Patients randomized to early surgery versus conservative management were hospitalized for heart failure less often (0 vs 11%), and the cumulative incidence of the composite of any secondary outcome or AVR in those assigned to conservative care was 62% and 92% at 4 and 8 years, respectively.

Homing in on the ‘Why’

The findings were likely positive for a couple of reasons, according to Malaisrie. First, patients included had “not just severe aortic stenosis but very severe aortic stenosis.” And even though they were asymptomatic, a portion of patients probably had “some LV dysfunction and may have been harboring some symptoms they just didn’t know about. So these patients may already have indications to proceed with either AVR or TAVR,” explained.

Secondly, “these groups were pretty low-risk patients and the Koreans were able to do an AVR . . . with a very low operative risk, near 0%,” Malaisrie said. “But the long-term data is undeniable. You can see the prognosis of the patients who were randomized to the medical group had worse survival and worse primary endpoints compared to the AVR group, which did very, very well. So, I think for this group of patients who are asymptomatic with very severe AS, [who] had previously a class II indication for AVR, that may be worth bumping up to a higher recommendation for intervention.”

Discussing the study in the main session, Robert Bonow, MD (Northwestern University Feinberg School of Medicine, Chicago, IL), questioned whether “these excellent results obtained in this single trial can be transportable to the United States in large centers where sometimes the operative mortality is not zero and we do indeed have the risk of strokes as we follow our patients.”

He also suggested that there might be implications for the TAVR space. “These are younger patients below our usual threshold for consideration for TAVR for the most part, and indeed many of these patients had bicuspid valve, but this clearly has implications for the EARLY TAVR trial, which is ongoing in the United States.”

Before making any changes to guidelines or practice, Bonow recommends waiting to see more data. “In the meantime, for those of us who see patients, one could argue again that the majority of these patients do come to a surgical endpoint within a very short period of time, so from a clinical management point of view, I think we already have data suggesting we can move the ball forward. And now we have these excellent outcomes data from Korea as well.”

In an accompanying editorial, Patrizio Lancellotti, MD, PhD (University of Liege Hospital, Belgium), and Mani Vannan, MBBS (Piedmont Heart Institute, Atlanta, GA), note that it is “intriguing” that 22% of the patients in the conservative approach group never underwent surgery, writing that these people “may have been those who had less cardiac damage or in whom the diseased progressed slowly.”

Because of this, the question arises regarding how to “best assess risk among patients with aortic stenosis, formulate a follow-up plan, decide on the timing of intervention, and devise a management strategy,” according to the editorialists. “Given that there appears to be a continuous increase in risk starting at a mean aortic-valve gradient of approximately 20 mm Hg, staging aortic stenosis instead of classifying the valvular lesion only according to data from Doppler imaging appears to be the best approach,” they write, adding that this means assessing structural abnormalities, considering other hemodynamic cardiac abnormalities, and measuring biomarkers.

“In a time of rapidly evolving transcatheter valve therapies, this framework of risk assessment in patients with aortic stenosis is perhaps best achieved in clinics that are dedicated to the care of patients with more than mild valvular disease to maximize the benefit of timely treatment,” Lancellotti and Vannan conclude.

Confirm the Absence of Symptoms

For operators today struggling with how to manage patients with asymptomatic aortic stenosis, Malaisrie recommended first confirming the absence of symptoms. “There’s an underutilization of stress testing, and I think some data suggests that as low as 50% of patients will actually get a stress test when it is a class I indication,” he said, adding that about one-third of patients thought to be asymptomatic end up showing symptoms on a stress test. “Another thing we look for during a stress test is an abnormal blood pressure response, which is also an indication to proceed,” Malaisrie noted.

Bonow acknowledged that it’s possible that not all patients enrolled in RECOVERY were truly asymptomatic. “Dr. Kang indicated that there was selective use of stress testing, but this is what we deal with clinically,” he said. “I think we're taking careful histories of our patients and we can believe these patients were presumably not limited by symptoms.”

It’s also important to follow this cohort “very, very closely,” Malaisrie stressed. The 19% of patients in the conservative approach arm who ended up in the emergency department represent a “lost opportunity to pick up patients somewhere before death and somewhere before ending up in the emergency room, either with another echocardiogram that shows progression or when patients truly develop symptoms. That's the time when we should quickly intervene,” he advised.