Ischemic Stroke After LAA Closure May Be Less Severe Than on DOACs

Patients with atrial fibrillation (AF) who experience an ischemic stroke after left atrial appendage occlusion (LAAO) have significantly better outcomes than those who have an event while being treated with direct oral anticoagulants (DOAC), a new retrospective analysis suggests.

Even though patients who underwent LAAO were older and tended to be a little sicker, the ischemic strokes that occurred in these individuals, as compared with in DOAC-treated patients, were less likely to be disabling/fatal at both hospital discharge and by 3 months.

“One of the reasons we did this analysis is that there is increasing data showing that there’s less mortality after appendage closure compared with anticoagulation, whether that’s coumadin or a NOAC [non-vitamin K antagonist oral anticoagulant],” senior investigator Vivek Reddy, MD (Icahn School of Medicine at Mount Sinai, New York, NY), told TCTMD. “I never really understood why. After warfarin, I just assumed it was because of less hemorrhagic stroke, but even that was a little bit weird because the number of hemorrhagic strokes that occur, even with warfarin, doesn’t really account for the magnitude of difference in mortality.”

These data, suggesting less severe ischemic stroke after LAAO, add another helpful piece to the puzzle. “If it's true that with appendage closure you have less severe ischemic stroke, and we know that bleeding in general [with anticoagulation] also increases mortality, then this makes a lot more sense to explain the mechanism of mortality benefit,” said Reddy.

The researchers acknowledge the study’s limitations, noting that this is a nonrandomized, retrospective study, subject to selection bias. Still, Reddy said they performed a propensity-matched analysis and several sensitivity analyses, and the results held up.

Andrew Goldsweig, MD (Baystate Medical Center, Springfield, MA), who wasn’t involved in the study, acknowledged its drawbacks but said he believes the new data show that LAAO is now a “mature treatment” for nonvalvular AF. 

“This isn't a one-off experimental thing,” Goldsweig told TCTMD, noting that the US Food and Drug Administration approved LAAO for patients with an appropriate rationale to look for an alternative to long-term oral anticoagulation. “This study certainly validates that left atrial appendage occlusion is absolutely indicated as superior to anticoagulation in that patient population. . . . The other thing I would add is that the findings may support a broader role for left atrial appendage closure. If strokes are less severe than with DOACs, then maybe it should be adopted more broadly,” he suggested.

There are additional data, said Goldsweig, showing that ischemic stroke is less severe after LAAO when compared with events that occur in warfarin-treated patients. The ongoing CHAMPION-AF and CATALYST trials are currently testing LAAO with the Watchman FLX (Boston Scientific) and Amplatzer Amulet (Abbott) devices, respectively, in a broader population of AF patients without contraindications to anticoagulation, he said.

In March, the Society for Cardiovascular Angiography and Interventions/Heart Rhythm Society (SCAI/HRS) released a new consensus document that continues to endorse LAAO for patients with nonvalvular AF at high thromboembolic risk who are not suited for long-term oral anticoagulation.

More Morbidity in LAAO Patients

The registry-based, retrospective study, which was published in JACC: Clinical Electrophysiology with Mohit Turagam, MD (Icahn School of Medicine at Mount Sinai), as lead author, included 125 consecutive patients (mean age 74.3 years; 55% male) with nonvalvular AF who had an ischemic stroke after being treated with LAAO—either the first- or second-generation Watchman/FLX or Amplatzer Cardiac Plug/Amulet LAA occluder—at eight centers in the US and Europe. The comparator arm included 322 patients with nonvalvular AF receiving DOAC therapy within 7 days before hospital admission with ischemic stroke, treated at the same eight centers.

The majority (55%) of DOAC-treated patients were on apixaban (Eliquis; Bristol-Myers Squibb), 29% on rivaroxaban (Xarelto; Bayer/Janssen), and 16% on dabigatran (Pradaxa; Boehringer Ingelheim), with 13.4% of all DOAC-treated patients on reduced dosages for various clinical reasons. Overall, the LAAO patients were older, had higher HAS-BLED scores, and were more likely to have a history of prior bleeding and hemorrhagic stroke. The CHA₂DS₂-VaSC scores were similar between the DOAC- and LAAO-treated patients.

Despite the differences in baseline morbidity, ischemic stroke that occurred after LAAO was less severe at hospital admission, discharge, and 3 months than in the DOAC-treated patients who had ischemic stroke. The incidence of disabling/fatal stroke (mRS score ≥ 3) at discharge was 38.3% versus 70.3%, respectively, and at 3 months was 33.3% versus 56.2%, both statistically significant differences. There was no difference in mortality during hospitalization, but the death rate was significantly lower in the LAAO group at 3 months (14.7% vs 32.1%; P = 0.002).

In a propensity-matched analysis that included 182 patients, the results were similar, with the LAAO group having significantly less severe stroke at discharge and 3 months compared with the DOAC-treated patients. Mortality at discharge and 3 months also favored treatment with LAAO (4.4% vs 14.4%; P = 0.023 and 12.5% vs 31.8%; P = 0.016).

In multiple sensitivity analyses, which excluded LAAO patients also receiving DOAC therapy (14.4% received both treatments), those on a reduced DOAC dose, and those nonadherent to medication, LAAO remained associated with fewer disabling/fatal strokes and less mortality at discharge and 3 months.

What’s Being Measured?

Goldsweig said that while the data are positive for LAAO, the study is difficult to interpret and ischemic stroke can occur for reasons unrelated to the LAA.

“In elderly people, which is essentially what this cohort is, they have plaque in the carotid, they have plaque in the aorta, and they have strokes for other reasons,” he said. “We know that A-fib-related strokes are, in general, worse than carotid plaque-related strokes because the left atrial appendage is a large structure. It’s interesting to see there is a difference in the severity of infarct here, and it makes you wonder whether the source of the infarcts is different in the different patients. That is, are the people in the left atrial appendage group having non-A-fib-related stroke?”

He pointed out that DOAC efficacy will always be subject to adherence: up to 25% of patients treated with apixaban, rivaroxaban, and dabigatran in the pivotal trials had stopped the drugs at 2 years. Unlike with LAAO, it’s possible that ischemic stroke in DOAC-treated patients may be related to their AF because they missed a dose or are ineffectively anticoagulated.     

“The bottom line,” said Goldsweig, “is it's tough to know what we're measuring here.”

Still, Goldsweig said that because the LAAO patients were sicker than those treated with medication, it’s possible that the reported difference in stroke severity might be underestimated if the study had captured similar patients.

Mechanistic Reasons for Findings

While LAAO is intended to reduce the risk of ischemic stroke in patients with nonvalvular AF, these events do occur. Reddy said the exact number isn’t known, given challenges with existing datasets, but that the National Cardiovascular Data Registry suggests the ischemic stroke rate is around 1.2%. That figure, he said, is “almost impossibly low,” likely due to underreporting. Data from the pivotal trials leading to FDA approval of Watchman and Amulet suggest the incidence is around 1.7%.  

To TCTMD, Reddy said there are a couple of potential reasons why LAAO may result in less severe ischemic stroke if such events occur. One possibility is less thromboembolic burden after appendage closure.

“If you have an appendage closure device, the clot burden with any residual stroke would tend to be smaller versus having a stroke on a NOAC when the appendage is open,” he said. “If the appendage is open, and if a stroke occurs, it could be a much bigger clot.”

Completeness of closure would also be a factor, he added. “We actually did not get enough imaging data to really do that analysis, to get an idea of the level of occlusion after an ischemic stroke. That's an analysis we'd love to do in the future.”

The other possibility is that there is a higher risk of hemorrhagic transformation in patients on DOACs who have an ischemic stroke, he said.

“When an ischemic stroke has bleeding into it, that expands the stroke,” said Reddy. “In fact, when we looked in our group and asked the question, ‘How many of these ischemic strokes undergo hemorrhagic transformation,’ it is very clear that it’s much higher in the DOAC patients versus those with appendage closure. This is true whether you look at the full cohort or in the propensity-matched cohort.”