July 2nd, 2012

Eroding Confidence in ASD Occlusion Devices


The FDA has identified more than 100 cases of cardiac erosion following atrial septal defect occlusion (ASO) device implantation. All were associated with implantation of the AMPLATZER (AGA Medical/St. Jude Medical) or related cribriform ASO devices; none have been reported thus far with the GORE HELEX (Gore Medical) ASO device.

Although the estimated occurrence of cardiac erosion is 0.1% to 0.3%, the precise incidence is unknown, because ASO manufacturers aren’t required to register or report device implantations, and the reporting of device-associated complications in the FDA Manufacturer and User Facility Device Experience (MAUDE) database is sporadic.

Cardiac erosion has been observed as early as within 24 hours of device implantation and as late as 8.6 years after implantation, with ≈50% occurring in the first week, 61% in the first month, and 87% in the first year after device placement. The erosions typically occur at the domes of the atria near the aortic root and cause immediate hemodynamic compromise due to pericardial effusion, tamponade, or both; patients typically present with pericarditis, hypotension, or even sudden death. Occasionally, the process is chronic; fibrosis develops as erosion progresses, and an atrial-to-aortic fistula develops. These individuals typically present with heart failure or a new-onset murmur. Regardless of the manner of presentation, prompt recognition and surgical intervention are indicated.

Except for ASO device type, the risk factors that predispose to device-related erosions are unknown. A deficient anterosuperior atrial rim (<5 mm) seems to be a risk factor, but the FDA Circulatory Device Advisory Panel determined that the data supporting this association were inconclusive and made the uncharacteristic recommendation that St. Jude remove deficient anterosuperior rim — which occurs in ≈40% of patients with a secundum ASD — as a contraindication to ASO device closure.

Limitations in our understanding of this life-threatening complication highlight the fact that a device registry is needed. In the meantime, patients considering ASO device implantation should be informed of this potential complication and its symptoms, so they can decide whether to proceed with percutaneous ASD closure or, alternatively, to undergo surgical repair.

Will these reports of cardiac erosions following ASD occlusion device implantation affect your recommendations regarding their use?

5 Responses to “Eroding Confidence in ASD Occlusion Devices”

  1. Isaac Vilayil Mammen, M.D., D.M. says:

    So if it’s my son I should do only a surgical closure?

  2. Thierry Legendre, MD says:

    This report is to be reminded when thinking about PFO closure in patients with stroke of unknown origin.

  3. Judith Andersen, AB, MD says:

    No –, just do not, as a default, use an Amplatzer device.There are several other device options that have obvious advantages. Although I am not comfortable with the data analysis of this trial, the conclusions are worrisome. Other options for PFO closure — for which fewer data are available — may be preferable –and we are all aware that no option is perfect. Weighing the risk of potential repetitive embolization vs closure suggests other medical options: decreasing the risk of peripheral venous thrombus formation, including peripheral venous insufficiency, thrombotic predispositions, etc., and now, the use of less demanding forms of antithrombotic therapy ( rivaroxaban, dabigatran, apixaban) — permitting a safer means of anticoagulation without dietary restrictions, monitoring demands, and worry about the impact of most concomitant medications — allows a means of diminishing risk of PFO-permissive transmission of stroke-inducing thrombi. PFOs, without closure, are forever, antithrombotic therapy is compliance-dependent, and definitive closure of a PFO still seems like a Holy Grail, even if we have not yet achieved the goal of acquiring it.

  4. Dan Hackam, MD PhD says:

    The trouble is that it is often very difficult to know whether PFO is related or unrelated to cryptogenic stroke in any given patient. Depending on the study you read, PFO is present in one quarter to one third of normals. Lifetime anticoagulation, on the other hand, can pose significant bleeding hazards even with the newer oral anticoagulants – all it takes is one road traffic accident, one major fall, one unanticipated event like an emergency surgery. We are talking about putting people on oral anticoagulation from their 30’s or 40’s for a half century or more. This is not a decision one should take lightly! If there is good corroborative evidence that the PFO was involved in the presenting event, then I believe closure would still be the best way to go, as it will prevent the need for lifetime oral anticoagulation and its attendant hassles. The patient can then forget about their issue and not live the rest of their life worried about the threat of another event or “medicalized” by taking chronic medication while still young.

  5. David Powell , MD, FACC says:

    Note that the “medical arm” of at least two of the trials included both antiplatelet therapy or VKAs, left to the discretion of the physicians. I do not think there is a head-to-head trial comparing anticoagulation options, though an old meta analysis favored VKAs. Nevertheless, some “expert” opinions support aspirin.
    The first step is to be as certain as possible that the “cryptogenic” CVA is consistent with an embolic etiology. This varies in my experience, with some neurologists referring patients for TEEs who have lacunar stroke. Then there is the PFO itself. Multiple characteristics have been variably associated with stroke, including size, the presence of an atrial septal aneurysm, a large Eustachian valve, and spontaneous right to left shunting (recently reported in JACC). I await information regarding these PFO characteristics among study participants. One also wonders whether the largest, most egregious PFOs were represented in the trials, as they may have been directly closed. For now, it is an uncertain art.