PARTNER A: Even Mild Paravalvular Leaks Linked With Higher Mortality at Two Years

Shelley Wood

March 26, 2012

March 26, 2012 (Chicago, Illinois) — Paravalvular leaks--a phenomenon virtually unheard of in the surgical valve replacement era--have truly been propelled into the limelight today, with the presentation of two-year results from the PARTNER A transcatheter aortic-valve intervention(TAVI) trial [1]. At two years, rates of death from any cause were not statistically different between transcatheter aortic-valve replacement (TAVR) and surgical valve replacement, but paravalvular regurgitation was more common after TAVR than surgery. And strikingly, even mild paravalvular regurgitation was associated with increased late mortality.

Dr Susheel Kodali (Columbia University, New York, NY) presented the two-year results here at the American College of Cardiology (ACC) 2012 Scientific Sessions; they were published online simultaneously in the New England Journal of Medicine. One-year results from the trial debuted at last year's ACC meeting, as reported at the time by heartwire .

"TAVR should be considered an alternative to surgery with similar mortality and clinical benefits," Kodali said this morning. "Periprocedural stroke concerns after TAVR have diminished with longer follow-up, and TAVR valve hemodynamics have remained stable, although periprocedural [aortic regurgitation] (even mild) has emerged as a predictor of late mortality."

Co–principal investigator, Dr Craig Smith (Columbia University), commenting on the two-year results for heartwire , acknowledged: "The paravalvular-leak findings were a surprise. Of the concerns dogging TAVR, paravalvular leak was the one for which two answers were given: one, it will become less frequent with time, experience, and modifications in design; and, two, it doesn't appear to matter anyway. For stroke and vascular complications, [the second explanation] clearly never applied.  We're back to number one alone for paravalvular leak, which shouldn't surprise us."

Kodali called the lack of difference in mortality and stroke the "most important" news from the two-year results but acknowledged that the finding that mild regurgitation is linked with mortality is "clearly interesting" and new. "What we've identified is a potential target in terms of improving on this procedure," he told heartwire . The procedure will continue to evolve, he pointed out, and a wider range of valve sizes, better recognition of patient characteristics, better imaging techniques will all help reduce the incidence of aortic regurgitation, he said. "This is a new procedure, being performed by inexperienced operators, going up against a decades-old therapy in the hands of the best surgeons," Kodali added. He believes the paravalvular findings "are not going to dampen enthusiasm for TAVI. If anything, this will motivate us to improve on these procedures and these devices."

PARTNER A: Deaths, Strokes at Two Years

Between years one and two in the PARTNER A cohort, 32 more patients died in the TAVR group and 25 in the surgery group, with no statistical differences in deaths from any cause or from cardiovascular causes. For strokes, the end point that drove much of the conversation after one-year results were reported, an additional eight strokes occurred, four in each group. Two transient ischemic attacks (TIAs) occurred in year 2 among TAVI-treated patients as compared with just one in the surgery group, yielding an overall neurological event rate that was higher in the TAVR group than in the surgery group, reaching borderline statistical significance. For strokes alone, however, no significant differences were seen between groups at two years. By year 2, 19 patients treated surgically received a new pacemaker, compared with 23 in the TAVR group, a statistically insignificant difference.

Major Events at Two Years

Event Surgery (%) TAVR (%) p
All deaths 35 33.9 0.78
CV deaths 20.5 21.4 0.80
All stroke/TIA 6.5 11.2 0.05
Stroke 4.9 7.7 0.17
MI 1.5 0 0.05

Mortality was higher in the transapically treated patients than in the transfemorally treated patients, but Kodali stressed to a morning press conference that the trial was not powered to compare these groups, and the transapical group included the "learning-curve experience." And Dr Robert Guyton (Emory School of Medicine, Atlanta, GA), a surgeon and trial investigator, pointed out that in patients in the continued-access program (CAP) registry, which now includes over 800 patients treated via a transapical approach, mortality rates have come down. "As experienced developed, mortality actually ended up being numerically lower than in the transfemoral arm . . . and stroke rate was dramatically reduced, to less than 3%, which is one of the lowest stroke rates reported" in transcatheter trials.

Paravalvular Regurgitation in the Spotlight

The finding from PARTNER A year 2 that is raising eyebrows here today is the paravalvular-regurgitation data. Moderate to severe paravalvular regurgitation has already been identified as a predictor of mortality among TAVR-treated patients, but as Leon said in an interview with heartwire , this is the first time mild regurgitation has been linked with a higher death rate, and within a prospective, multicenter, randomized clinical trial, with independent core-lab echocardiographic review. "The effect of aortic regurgitation on mortality was proportional to the severity of the regurgitation, but even mild aortic regurgitation was associated with increased rate of late deaths," Kodali et al write.

Mortality According to Presence/Severity of Aortic Regurgitation

Aortic regurgitation None/trace Mild Moderate/severe P (log rank)
1-y mortality (%) 14.5 29.2 29.5 <0.001
2-y mortality (%) 24.8 39.2 41.1 <0.001

Dr Robert Bonow

Commenting on the paravalvular-regurgitation data, Dr Robert Bonow (Northwestern University, Chicago, IL) pointed out that moderate to severe paravalvular regurgitation has emerged as an important, possibly even the most important predictor of mortality in TAVI-treated patients. In patients who undergo surgical valvular replacement, the rate of moderate to severe aortic regurgitation is typically less than 1%. "Here you have 40% of TAVR-treated patients with mild aortic regurgitation, and another 10% with moderate to severe aortic regurgitation, so 50% of patients have mild regurgitation or greater--that's a potential Achilles heel."

Dr Martin Leon

Speaking with heartwire , senior author on the study, Dr Martin Leon (Columbia University), agreed but, like Kodali, sees a silver lining: "This is an important observation, but I would put it in perspective. If you look at overall mortality between surgery and TAVR, the rates are identical, and in no subgroup is overall mortality different," Leon told heartwire . "We're thinking paravalvular leak can be treated or prevented, and there's clear evidence that there was valve undersizing in PARTNER that we think can be corrected using postdilatation if done selectively and carefully. If you can reduce the incidence of paravalvular leak down to 'none or trace,' you could potentially reduce mortality by half."

And looking at mortality among patients with no paravalvular regurgitation vs patients with mild to severe paravalvular regurgitation, mortality rates were halved, he noted. "This suggests that this should become a very aggressive target, and new technology, improved procedure characteristics, and better sizing of the devices--those all may be important in reducing mortality as compared with surgery," Leon said.

Room to Expand

Dr Craig Smith

Bonow and Smith tended to agree with Leon and Kodali, with Smith, a surgeon, predicting that "the frequency of paravalvular leak will decrease with experience and will not be a major impediment. Design may also play a role but is a more complicated and theoretical discussion."

Bonow, likewise, said that interventionalists he's spoken with are now doing everything they can to make sure they "fix" a paravalvular leak when they see moderate to severe regurgitation, by reexpanding the device or even using a valve-in-valve procedure. But to date, he says, operators haven't been overly concerned about mild regurgitation--that should likely change.

Already, speculation is mounting as to whether the paravalvular data might be one reason the FDA has not yet announced a date for its advisory committee to review the PARTNER A data, where the sponsor is seeking to extend the indication for TAVR to patients who are also eligible for surgery.

"In a regulatory environment sensitized by lead fractures, failing steel hip prostheses, and other postapproval embarrassments, this is sure to be noticed and will probably slow approval in young, low-risk populations until long term success is better understood," Smith speculated.

Bonow had no disclosures; Guyton has previously disclosed receiving grant support from Edwards Lifesciencies; Kodali disclosed receiving grant funding from Edwards; Leon disclosed receiving travel support from Edwards; Smith disclosed receiving grant funding and travel support from Edwards. Disclosures for the coauthors are listed in the paper.