麻豆传媒

Years after surgical aortic valve replacement, valve-in-valve transcatheter aortic valve replacement (TAVR) may be the answer for failing surgical bioprostheses, according to a registry study.

The valve-in-valve procedure -- performed with a balloon-expandable Sapien XT device -- was associated with a 1-year mortality rate of 12.4% for a high-risk cohort, reported John G. Webb, MD, of St. Paul's Hospital in Vancouver, and colleagues in the Journal of the American College of Cardiology.

Death was more likely for patients with a stubborn post-TAVR gradient remaining at 20 mm Hg or higher (16.7% versus 7.7% for lower gradients, P=0.01).

In the shorter term, 30-day event rates were also "relatively low:" all-cause mortality (2.7%), stroke (2.7%), major vascular complications (4.1%), conversion to surgery (0.6%), coronary occlusion (0.8%), new pacemaker insertion (1.9%).

The investigators called their findings "impressive" in light of the average 9.1% Society of Thoracic Surgeons predicted 30-day risk of surgical mortality.

"The availability of valve-in-valve TAVR will likely influence surgical practice in the management of aortic valve disease, possibly lessening concerns about the durability of surgical bioprostheses. It is essential to realize, however, that specific characteristics of the chosen surgical bioprosthesis, (including its internal diameter, design characteristics, and proximity to the coronary ostia) will have an impact on whether a valve-in-valve procedure will or will not be a realistic future option," according to the investigators.

Webb's group took their data from the PARTNER II Valve-in-Valve Registry and the subsequent continued access registry. Patients in the latter had lower 30-day (0.7% versus 8.3%, P<0.0001) and 1-year death rates (9.8% versus 19.8%, P=0.006), suggestive of a learning curve for valve-in-valve TAVR.

By 1 year, greater-than-mild paravalvular regurgitation was at 1.9%. Several echocardiographic measures appeared better:

• Mean gradient was 17.6 mm Hg, down from 35.0 mm Hg at baseline (P<0.0001)
• Effective orifice area reached 1.16 cm² from 0.93 cm² (P<0.0001)
• Left ventricular ejection fraction went up, 50.6% to 54.2% (P<0.0001)
• Mass index dropped, 135.7 to 117.6 g/m² (P<0.0001)
• Reduced moderate-or-severe mitral regurgitation, 34.9% to 12.7% (P<0.0001)
• Lower rates of moderate-or-severe tricuspid regurgitation, 31.8% to 21.2% (P<0.0001)

Quality of life (average 43.1 to 77.0 points on the Kansas City Cardiomyopathy Questionnaire, P<0.0001) and 6-minute walk test distances also improved (163.6 to 252.3 meters, P<0.0001).

More worrying, however, was that the first post-TAVR echocardiogram showed severe patient-prosthesis mismatch (PPM) at a high rate of 58.4%.

"To put this in perspective, in patients undergoing SAVR for native aortic stenosis [as] part of the PARTNER 2A intermediate-risk trial, 33% had severe PPM," Webb and colleagues wrote. "The higher gradients and higher rates of PPM after valve-in-valve procedures likely represent a combination of pre-existing PPM due to an undersized surgical valve and transcatheter heart valve underexpansion due to the constraints of pre-existing surgical valves."

"Unfortunately, current real-world data show that more than one-third of all patients undergoing SAVR receive a bioprosthesis smaller than 23 mm. Over the longer term, constrained transcatheter heart valve expansion and malcoaptation within a bioprosthesis might lead to accelerated leaflet degeneration and reduced durability," they suggested.

An editorial by Blase A. Carabello, MD, of East Carolina Heart Institute in Greenville, N.C., called the findings "gratifying," commenting that valve-in-valve TAVR "was used in a safe appropriate fashion to rescue patients at high surgical risk."

That said, Carabello emphasized that "this approach is a far cry from a prospective strategy in which a relatively young patient is advised to undergo aortic valve replacement with a surgically implanted bioprosthesis with the hope of performing valve-in-valve TAVR if the first valve fails. Although PPM did not affect outcome in this series, post-procedure gradient did."

"To use a valve-in-valve strategy prospectively, a low gradient has to be assured and that seems less likely if a small valve is implanted," Carabello suggested, emphasizing the importance of good imaging (to optimally size the surgical valve) and competency in surgical enlargement of the aortic root (to avoid a high post-procedural gradient and its associated mortality).

The registry study included 365 patients who were enrolled from 2012 to 2014 (mean age 78.9; 64.1% men). Two caveats of his retrospective analysis are the limited follow-up and the selection of participating centers.

Two-thirds of bioprostheses requiring intervention were over 10-years-old. Failure was due to stenosis or mixed disease 76.5% of the time and regurgitation for the remainder.

On top of the authors' findings, Carabello drew further insight from the echocardiographic data. "Taken together, the data suggest that a very modest reduction in afterload can lead to important clinical improvement, despite that the aortic valve area index was still in the 'severe' range at 0.6 cm²/m²."

"It has to be true that it is the obstruction-caused pressure overload of aortic stenosis that is the disease's major pathological instrument. The data herein provide insight into how much pressure overload reduction is necessary to provide adequate clinical improvement. A reduction in mean gradient of as little as 25 mm Hg may be adequate in elderly, relatively sedentary patients, consistent with previous reports of temporary improvement after balloon aortic valvotomy," he wrote.

"Conversely, a residual gradient of 20 mm Hg may constitute a significant residual pressure overload, impairing prognosis and promoting morbidity," he stated.