Không đáp ứng Clopidogrel giả mạo Viêm nội tâm mạc sau tắc nghẽn PFO: Kết hợp chọc hút, lấy vật liệu và xét nghiệm phòng thí nghiệm để đánh giá chẩn đoán và điều trị

Abstract

Percutaneous patent foramen ovale (PFO) closure lowers recurrent stroke risk but can rarely cause device thrombosis or infection. A 57-year-old woman had elective PFO closure with a GORE CARDIOFORM Septal Occluder device after cerebellar infarction. Despite aspirin and clopidogrel, she developed fever. Transesophageal echocardiography revealed large thrombi on the device. Multidisciplinary management achieved percutaneous retrieval 1-week postimplantation. Cultures were negative. Platelet testing confirmed aspirin effect but showed clopidogrel nonresponsiveness, leading to anticoagulation. Follow-up imaging demonstrated thrombus regression. This case highlights early device thrombosis mimicking infection, successful percutaneous retrieval, and the value of platelet function testing in guiding therapy.

Case Report

A 57-year-old woman with a history of right cerebellar infarction in May 2025, attributed to paradoxical embolism through a patent foramen ovale (PFO), was electively admitted in June 2025 for percutaneous PFO closure. Under fluoroscopic guidance and following a 7.500-unit heparin bolus (activated clotting time, 329 s), a 25-mm GORE CARDIOFORM Septal Occluder device was delivered via a 16-F femoral sheath. Immediately postimplantation, she received a 600 mg loading dose of clopidogrel. On the first postprocedure day, the patient developed an isolated fever of 38.7 °C without focal symptoms. Chest radiography showed no infiltrates, and urinalysis was negative for urinary tract infection. Therefore, empirical intravenous ampicillin/sulbactam was administered after obtaining blood cultures. To further evaluate the fever, a repeat transesophageal echocardiogram on day 3 confirmed correct device position but revealed a large echogenic mass (1.7 cm × 0.9 cm) adherent to the right atrial surface of the occluder (extending toward the superior vena cava) and a thin layer of mass on the left disc (Figure 1). No valvular vegetations were seen. In addition, given the absence of respiratory symptoms or oxygen desaturation, computed tomography pulmonary angiography was not performed, and clinically relevant pulmonary embolism was considered unlikely. However, small peripheral emboli cannot be fully excluded as a contributor to the early fever.

Given the large biatrial thrombotic burden and the persistent concern for device infection/culture-negative endocarditis (despite negative cultures), the heart team discussed continued anticoagulation versus device removal. Ultimately, because a potentially infected intracardiac foreign body and the risk for progressive thromboembolism were considered to outweigh the procedural risk (particularly early after implantation and before complete endothelialization) we proceeded with percutaneous extraction on day 7, with surgical backup and thrombus aspiration to minimize embolization risk, in addition to escalation of anticoagulation and antibiotics. During the procedure, under TEE guidance through a 16-F femoral sheath, multiple snares (30 mm, 10 mm, 5 mm, 2 mm) failed to free the firmly attached septal disc. Accordingly, a free thrombus was aspirated for microbiology, after which a bioptome mobilized the occluder disc, and further aspiration cleared an additional clot. Finally, a 5-mm snare captured the right atrial disc connection, permitting uneventful removal of the conformable GORE CARDIOFORM Septal Occluder device (Figure 2; Video 1). Sedation was discontinued immediately afterward, and the patient awoke neurologically intact. Subsequently, thrombus specimens and serial blood cultures remained sterile.

Under continued antibiotics, the patient’s inflammatory markers normalized, and her fever resolved. Full anticoagulation therapy was maintained with low–molecular-weight heparin, and then her treatment was transitioned to apixaban, and aspirin was withdrawn. To clarify the mechanism of thrombosis, platelet function testing by Born aggregometry showed markedly reduced platelet activation in response to arachidonic acid, confirming effective aspirin-mediated cyclooxygenase inhibition. However, capillary closure time remained within normal limits despite P2Y₁₂-inhibitor therapy, indicating insufficient platelet inhibition by clopidogrel (1).

Notably, there was no prior history of venous thromboembolism or known thrombophilia. Routine coagulation variables and platelet count were unremarkable. Taken together, we therefore considered clopidogrel hyporesponsiveness a key contributor to the rapid thrombosis. Consequently, given the patient’s clopidogrel hyporesponsiveness, future procedures will employ an alternative P2Y₁₂ inhibitor (for example, ticagrelor). On follow-up, TEE demonstrated marked regression of residual thrombotic masses, leaving only a thin (3 mm to 4 mm) layer on the atrial face of the previous closure site (Figure 3).

Discussion

This case illustrates a rare but serious early complication of percutaneous PFO closure and device‐related thrombosis with presumed superinfection despite negative cultures in a patient with a prior paradoxical cerebellar infarction. To our knowledge, this represents the first reported thrombus aspiration and percutaneous retrieval of a GORE CARDIOFORM Septal Occluder device performed as early as 1-week postimplantation 2–4.

Summary Points

High index of suspicion: unexplained fever postdevice implantation warrants prompt imaging to exclude device-related masses.

Percutaneous retrieval feasibility: complex extraction techniques (snares, bioptome, aspiration) can successfully remove infected/thrombosed occluders without open surgery.

Tailored antithrombotic therapy: platelet function testing is essential not only in stent thrombosis after percutaneous coronary intervention but also in identifying hyporesponders to clopidogrel and guiding selection of alternative P2Y₁₂ inhibitors.