NATIONAL HARBOR, Md. — Innovation in left ventricular assist devices (LVADs) has reached the point that routine left atrial appendage (LAA) occlusion has little to offer anymore in the way of reducing strokes, a MOMENTUM 3 analysis showed.
In this trial, HeartMate 3 device recipients getting LAA occlusion during LVAD implant surgery had the same likelihood of survival free of any stroke over 2 years as peers getting LVAD surgery alone (73.5% vs 76.6%, P=0.37), even in the subgroup of patients with pre-existing atrial fibrillation (Afib).
Similarly, the incidence of any hemocompatibility-related adverse event (i.e., nonsurgical bleeding, stroke) over follow-up was comparable between LAA occlusion and non-LAA occlusion groups (40.4% vs 41.9%, P=0.77), Afib or no Afib, reported Akinobu Itoh, MD, PhD, of Brigham and Women’s Hospital and Harvard Medical School in Boston.
Overall survival rates exceeded 80% in both groups over 2 years.
“We believe that the enhanced hemocompatibility of the HeartMate 3 pump, with ongoing background anticoagulation, likely reduces any major observable benefit of routine surgical LAA occlusion,” Itoh told the audience during a late-breaking trial session at the Heart Failure Society of America (HFSA) annual meeting.
Although the procedure is simple enough, LAA occlusion does come with extra costs that may now be saved during HeartMate 3 implant, he suggested.
HeartMate 3 is the only durable LVAD left on the market following the departure of HeartWare last year.
In the pivotal MOMENTUM 3 trial, the centrifugal-flow pump had been found to be superior to its predecessor, the axial-flow HeartMate II, for 2-year outcomes in people with advanced heart failure. These results held up at 5 years, when survival exceeded 58% in HeartMate 3 recipients.
LAA closure has historically been performed concomitantly during cardiac surgery to prevent ischemic strokes in people with Afib. During the era of older LVADs, it had been suggested that surgical LAA occlusion during implantation can help reduce hemocompatibility-related adverse events down the road.
Although the present analysis found this not to be true in contemporary practice, Itoh cautioned that it remains uncertain whether LAA occlusion makes anticoagulation management easier after LVAD surgery, particularly for people needing periprocedural discontinuation due to another operation, those predisposed to bleeding issues, and those nonadherent to anticoagulation.
For the current study, Itoh and colleagues analyzed the 2,200 HeartMate 3 recipients from the MOMENTUM 3 trial and its continued access protocol phase. They compared the 1,380 with no concurrent procedures against the 150 who underwent LAA occlusion (of these, the numbers of patients with Afib were 581 and 75, respectively).
Session co-moderator Mark Drazner, MD, MSc, of the University of Texas Southwestern Medical Center in Dallas, asked if the preference for LAA occlusion was site-specific and Itoh responded that it appeared site-specific and surgeon-specific. LAA occlusion had been left to the discretion of the surgeons, who in 2014 when the trial started, had been unaware of the benefits of LAA closure, he said.
His group found that LVAD recipients who underwent LAA closure tended to have more heart failure of an ischemic etiology and a history of Afib, among other differences at baseline. The investigators adjusted for these factors across their analyses.
On top of the null effect on survival and stroke, LAA occlusion had no significant effect on procedure times and length of stay.
A multivariable analysis suggested that the predictors of hemocompatibility-related adverse events were older age, female sex, and ischemic etiology of heart failure — not LAA occlusion or history of Afib, Itoh reported.
However, he acknowledged that the low frequency of concomitant surgical LAA occlusion in MOMENTUM 3 meant the analysis was underpowered to show a benefit from the start. LAA occlusion could also have benefits that emerge after 2 years, he added, though he noted similar findings from the 5-year data of the pivotal trial.
Session co-moderator Gerasimos Filippatos, MD, of Athens University Hospital in Greece, suggested further analysis of trial patients who underwent HeartMate 3 placement and LAA occlusion, as well as a third heart procedure such as coronary artery bypass grafting or a valve procedure.
MOMENTUM 3 was supported by Abbott.
Itoh had no personal disclosures.