Postoperative infection of a prosthetic aortic valve by Mycobacterium chelonae

Abstract number: 1134_04_276

Uçkay I., Garbino J., Matulionyte R., Djordjevic M., Rohner P., Khachatourian G., Pittet D.

Postoperative endocarditis can be caused sometimes by atypical pathogens. We present a case of an infection of a bio-prosthetic aortic valve/of the aortic root by Mycobacterium chelonae in an immuno-competent host.

Case report:  

A 72-year-old woman with aortic stenosis of degenerative origin underwent in February 2003 an elective replacement of the aortic valve by a bio-prosthesis Shelhighâ and of the enlarged aortic root by a tube Shelhigh N°23. In November 2003 she was re-hospitalised because of cardiac decompensation, inflammatory parameters and nocturnal transpiration without weight loss. Echocardiography revealed a displacement of the aortic valve without vegetations. Blood cultures, even with prolonged incubation periods, for microorganisms remained sterile. Angiography before surgery showed ‘sacculations’ of the aortic root; suspecting an infection. In December 2003 surgery was performed. The surgeons remarked several cavities of cold abscesses in the aortic root. The annulus was destructed by these abscesses protruding also into the pulmonary artery. Infection could not be excised in total. An intervention according to Bentall by insertion of a new Shelhighâ bio-prosthesis was performed and the coronaries re-implanted. An atrio-ventricular block III° necessitated a permanent pacemaker. Acridine strain revealed the presence of Mycobacteria spp. which did not grow in culture. PCR sequencing technique identified M. chelonae. A large empiric intravenous antibiotic therapy including amicacin, rifampicin, ethambutol, co-trimoxazol, clarithromycin and ciprofloxacin was begun. The i.v. regimen was changed to a peroral regimen three weeks after by omitting amikacin. After confirmation of M. chelonae the antibiotic sixtatherapy was reduced to a peroral regimen of co-trimoxazol, ciprofloxacin and clarithromycin, scheduled for at least 6 months. Clinically the treatment was fully successful already after 5 weeks. The origin of this unusual pathogen remained unclear, the possibility of contamination during the operation or contamination during fabrication could be a possibility. In March 2004 the patient suddenly died of ventricular fibrillation. Autopsy identified a dehiscence of the implantation of the right coronary artery as the cause. There were no signs of activity of the former infection. The abscesses had disappeared.

Conclusion:  

Successful empiric antibiotic therapy with surgery even for large size prosthetic aortic valve and aortic root is possible.