Late Bacterial Endocarditis in an Intravenous Drug User With an Amplatzer Septal Occluder

Case Report

In February 2018, a 26-year-old man presented at our emergency department with a 2-day history of left peripatellar pain and swelling. At age 6 years, he had undergone percutaneous closure of a secundum ASD with use of the buttoned device.⁷ Seven years later, a transesophageal echocardiogram (TEE) revealed that the fabric of the buttoned device had torn, leaving a 6-mm central gap that created a left-to-right cardiac shunt. The buttoned device was percutaneously replaced with an ASO. The new defect was sized with a balloon, and an ASO (waist size, 8 mm) was introduced through the center of the old device. The 2 occluder discs engulfed and enclosed the existing device, leaving no residual shunt.

When the patient was admitted to our institution almost 14 years later, he was tachycardic but had no fever or leukocytosis. He had no splinter hemorrhages, Janeway lesions, or audible murmurs. He had been taking drugs intravenously for one year, and his urine toxicology results were positive for benzodiazepine and cocaine. Polymerase chain reaction tests detected methicillin-sensitive Staphylococcus aureus in joint aspirate and blood cultures.

A TEE showed a freely oscillating, windsock-like vegetative mass (area, 2.25 cm²) on the left atrial septal wall, attached by a stalk to the peripheral wiring of the patient's ASO (Fig. 1A–B). No vegetation on the valve was noted. We prescribed 2 weeks of antibiotic therapy with rifampin, gentamicin, and nafcillin. Three days after therapy began, blood cultures showed no microorganism growth, and the regimen was changed to nafcillin alone for 6 weeks. A TEE then showed that the mass had shrunk to 1 cm² but had not resolved (Fig. 1C), so surgical removal was scheduled.

View Figure

• Download as Powerpoint
• Download Figure

At surgery, the ASO was grossly intact without areas of incomplete endothelialization. The mass, attached to the peripheral portion of the occluder, was excised at the lowest possible point, and the device was preserved. On gross examination, the mass was pink with a thin stalk and a smooth external surface. Histologic analysis revealed a dense thrombotic formation of mixed chronic and acute inflammatory cell infiltrate surrounded by endothelium (Fig. 2). Nafcillin was continued for 6 weeks, and the patient recovered uneventfully before his discharge from the hospital. A transthoracic echocardiogram at his one-month follow-up appointment showed no residual mass or new growth (Fig. 3). Blood cultures were free of organisms, as were an acid-fast bacilli smear and fungal and bacterial cultures of the excised vegetation. The patient was lost to follow-up thereafter.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

Discussion

To our knowledge, 14 years is the longest interval between ASO implantation and diagnosis of IE. Ours is also the first report of using surgical excision and débridement to treat an ASO-associated vegetation caused by bacterial endocarditis.

The ASO has been used to repair various hole defects, such as a ruptured sinus of Valsalva or Fontan fenestrations.⁸ Given the location and size of our patient's residual shunt, the use of an ASO to replace the failed device was highly effective. However, whether the replacement ASO increased his risk of IE is not clear.

Our literature review revealed 24 reports of ASD-related bacterial endocarditis, 18 of which involved an ASO.^9,10 Late bacterial endocarditis is defined as infection occurring after implantation and at least 6 months after prophylactic antibiotic therapy has been completed.^10–12 In preliminary studies in pigs, 100% endothelial coverage was observed within this time frame.¹³ Presumably, complete device endothelialization prevents future endocarditis; however, incomplete or no endothelialization may persist.^3–5 Therefore, the extent of endothelial coverage can vary among patients. No studies or established screening methods are available to determine what percentage of complete endothelialization effectively protects against infection.

Full endothelialization was reported in 3 of the 18 previous cases of ASO-associated IE.^14,15 Krantz and Lawton¹⁵ described IE affecting a patient's endothelialized ASO after a prolonged urinary tract infection from disseminated methicillin-resistant S. aureus. Gross and histologic examination of our patient's ASO revealed endothelial tissue extending over the mass and device and continuing along the septum. Given our patient's history and clinical course, his recent systemic methicillin-sensitive S. aureus infection probably caused the endocarditis. This suggests that possible bacterial seeding of a completely endothelialized device should be considered in cases of ASO-associated IE.

The preferred treatment for ASO-associated IE is to remove the device and replace it with a homologous patch.^{4,5,9,10,12,14,15} Other treatments have also been suggested. Aruni and colleagues¹¹ showed that, absent device failure, antibiotic therapy alone was sufficient to resolve a vegetative mass. In our patient, prolonged antibiotic therapy was not completely effective. His IVDU led us to débride the ASO surgically. He recovered uneventfully, and no vegetation was evident on follow-up.

The American Heart Association recommends a 6-month course of prophylactic antibiotics to prevent endocarditis in patients who have ASD closure devices.¹⁶ These guidelines presume complete endothelialization of the device after 6 months and give no recommendations for long-term follow-up care. Furthermore, guidance for managing incomplete endothelization is limited. On the basis of our case and recent reports, we recommend placing greater emphasis on preventive care for patients who have ASD closure devices. Performing TEE is warranted in the presence of bacteremia, and elective TEE should be considered in asymptomatic patients who exhibit high-risk behavior after device insertion.