No Volume-Outcome Link With Watchman for LAA Closure? Single-Center Study

Physicians with limited experience closing the left atrial appendage (LAA) with the Watchman device (Boston Scientific) in patients with atrial fibrillation do not appear to have worse clinical outcomes when compared with physicians who have greater experience performing the procedure, according to the results of a new analysis.

After excluding each operator’s first 10 cases, the clinical outcomes were similar for operators at all levels of experience, a finding that suggests there is not a protracted learning curve. Researchers say their findings may mean the Watchman LAA occlusion procedure is not subject to the same volume-outcomes relationship as other structural heart interventions.

“If this truly was such a complex procedure, the volume-outcome relationship should hold true,” said lead investigator Ashish Pershad, MD (Banner University Medical Center, Phoenix, AZ). He added that some hospitals could implant Watchman as a standalone procedure, without also offering other structural procedures, provided they have a cardiothoracic surgeon, hybrid operating room, imaging specialists, and a physician capable of performing transseptal puncture. “They should have all the tools they need to do a Watchman procedure safely,” he stressed.

To TCTMD, Pershad said that some have balked at the idea of centers doing Watchman LAA closure without also performing the more challenging aortic or mitral valve procedures. With excellent proctoring and a dedicated expert in transseptal puncture, the Watchman LAA closure can be safely performed at centers that may not provide TAVR or MitraClip (Abbott Vascular) procedures, he said.

“Some people consider it a really difficult procedure, but my take on it is that other than the piece that involves expertise in transseptal puncture, it’s not that difficult,” said Pershad.

Andrew Goldsweig, MD (University of Nebraska Medical Center, Omaha), who was not involved in the study, said the Watchman LAA procedure is more challenging than TAVR, with risks not observed in other interventional cardiology procedures, but that the absence of a volume-outcome relationship after excluding some of the earliest cases is “really exciting.”

“I think it reflects a couple of interesting features we’ve seen growing in structural heart disease,” said Goldsweig. “One is institutional knowledge. At [Banner], they had seven operators, with two learning under the supervision of more-experienced operators. They won’t experience that much of a learning curve because they’re not operating alone. There’s somebody there to advise them. Even in a broader sense, there is a collective experience [with Watchman] throughout the cardiology community.”

Using TAVR as an example, Goldsweig pointed to a 2016 analysis of the original PARTNER trial published by Oluseun Alli, MD (Novant Heart and Vascular Institute, Charlotte, NC), that highlighted declining procedure times, contrast volumes, and fluoroscopy times at institutions that enrolled patients later in the trial. “As a community, as a global network of cardiologists, people just knew more about how to do this procedure,” he said. “If you’re just starting off, there were people out there to advise you. You weren’t starting in a vacuum.”

Even the expertise of well-trained device representatives from manufacturers can help novice physicians obtain good results, said Goldsweig.  

No Difference by Operator Volume

The new analysis, which was published August 9, 2019, in Cardiovascular Revascularization Medicine, included 425 consecutive patients (mean age 75 years) with atrial fibrillation undergoing LAA occlusion from 2015 to 2018 by seven operators at a single institution. Nearly 60% of the procedures were performed by interventional cardiologists, and operator experience was stratified into tertiles (< 40, 41-80, and > 80 procedures per year). The mean CHA₂DS₂-VASc score was 4.5, and the mean HAS-BLED score was 3.9.

The primary outcome, a composite of all-cause mortality, postprocedural stroke, major bleeding, and vascular complications at 45 days occurred in 4.9% of treated patients. Although the MACE rate was higher than that what had been observed in the clinical trials, the composite endpoint did not differ by physician volume after the exclusion of the first 10 cases. Overall, the primary endpoint event rate was 6.5%, 5.2%, and 4.5%, respectively, for low-, mid-, and high-volume operators (P = 0.83). Technical success was also similar across physician volumes.

“The complication rates are a little bit higher than those previously reported,” said Pershad, referring to the randomized trials and continued access registries. As to why there is no volume-outcome relationship, the researchers note that Watchman LAA closure was approved in 2015 and say their data reflects their contemporary experiences with the device. Additionally, the device has undergone some modifications since the initial PROTECT-AF study with the intent of improving procedural safety, they add.

Outcomes also did not differ by operator specialty, with the MACE rate occurring in 6.0% of patients treated by interventional cardiologists and in 3.4% of those treated by electrophysiologists (P = 0.24). Technical success rates also were similar between interventionalists and electrophysiologists. Pershad said electrophysiologists have an advantage over interventional cardiologists because they are more experienced with transseptal puncture, but this skill did not translate into better clinical outcomes.

“The question is whether transseptal puncture can be taught,” said Pershad. “Can you teach an interventional cardiologist who doesn’t do a lot of MitraClip procedures to do a safe transseptal puncture to be able to the Watchman [procedure]?” The issue is contentious, with some arguing that only electrophysiologists perform safe transseptal punctures because it’s part of their skill set for atrial fibrillation ablation, he added. “There are some that argue that this should largely be an EP-driven procedure.”

Pershad asserted that the use of procedural imaging, such as transesophageal echocardiography, can minimize the risks of transseptal catheterization and that, after an initial learning curve related to the puncture, both electrophysiologists and interventionalists can perform the procedure equally well.   

Goldsweig noted that transseptal puncture for the Watchman procedure requires a 14-Fr device, which is considerably larger than the transseptal puncture required for MitraClip. “It’s a pretty big hole compared to the average catheter size for a coronary procedure,” he said. Previous studies have hinted at a learning curve with Watchman LAA closures, he added, including a 2017 analysis of the Manufacturer and User Facility Device Experience (MAUDE) database. In that study, the researchers reported a spike in Watchman complications following FDA approval in 2015 which they attributed to the learning curve.

While he “wouldn’t definitively say there is no learning curve with Watchman,” Goldsweig predicted it is likely short. As for the higher MACE rate, Goldsweig attributed this finding to sicker patients treated in the real world as compared to in the randomized trials, noting that the CHA2DS2-VASc score was 2.2 in PROTECT-AF and 3.8 in PREVAIL.