PARTNER B - by theheart.org

San Francisco, CA - Two-year outcomes in the PARTNER B trial, testing transcatheter aortic-valve replacement (TAVR) using the Sapien device (Edwards Lifesciences) against best medical care, show that survival curves are continuing to separate and the number needed to treat to save one life dropped from five at one year to four patients at two.

"Two-year data continue to support the role of TAVR as the standard of care for symptomatic patients with aortic stenosis who are not surgical candidates," said Dr Raj R Makkar (Cedars-Sinai Medical Center, Los Angeles, CA), who presented the results here at TCT 2011.

The FDA approved the Sapien valve for the US market last week, based on PARTNER B results.

By two years, 67.6% of patients in the medical group had died, compared with 43.3% in the TAVR group, a difference of 24.3%, Makkar noted. In a landmark analysis looking only at deaths between the one- and two-year mark, 35% of patients randomized to medical management who survived one year were dead within two, whereas half that amount, just 18% of the TAVR group, died in this period.

A similar separation of curves was seen for cardiovascular mortality and repeat hospitalizations. In fact, said Makkar, "there was half the number of repeat hospitalizations in the TAVR group as in the standard-therapy group, despite the fact that there were many more patients alive in the TAVR group at one year." The difference in median days alive out of hospital between the two groups added up to a full year.

Two-year outcomes: PARTNER B

End point	Standard therapy (%)	TAVR (%)	Absolute difference	Log rank p
All-cause mortalitya	67.6	43.3	24.3	< 0.0001
Cardiovascular mortalityb	62.4	31.0	31.4	< 0.0001
Repeat hospitalizations	72.5	35	37.5	< 0.0001

a. Intention-to-treat analysis; crossover patients followed

b. Intention-to-treat analysis; crossover patients censured

A low number of cerebrovascular events continued to accrue in the TAVR arm, but not the standard-therapy arm, after the one-year mark. By two years, the "all-strokes" rate was twice as high in the TAVR group, at 11.2%, compared with 5.5% in the standard group.

Showing slides that used 30 days as the cut point—to separate periprocedural stroke from other stroke causes—Makkar showed that four hemorrhagic strokes and five ischemic strokes occurred between 30 days and two years in the TAVR group, compared with one hemorrhagic stroke and four ischemic strokes in the standard group.

"Beyond 30 days, the reason for strokes is multifactorial," he said, including things like medication use and falls.

Of note, an analysis looking at patients with mild or no paravalvular leaks and moderate or severe leaks found no mortality difference at two years.

Commenting on the study during a morning press conference, Dr Michael Mack observed, "This is absolutely a dramatic result that bolsters the one-year results. An extremely positive trial. You don't have to do anything more than look at the lines of survival between TAVR and control: there was a dramatic difference, with a 20% absolute improvement at one year, and the question is how long is the improvement going to be sustained? And the answer from this is two years . . . and hopefully longer than that."