Mechanical Prosthetic Valve Sparing for Aortic Root Abscess Complicated by Infective Endocarditis

Patients and Methods

From October 2014 through November 2017, 41 patients at our center underwent surgery for aortic root abscess complicated by IE of a mechanical prosthetic aortic valve. All patients were evaluated pre- and intraoperatively to determine their eligibility to undergo the valve-sparing technique (Table I). Twenty-one patients did not meet those criteria and underwent abscess drainage, débridement, and valve replacement surgery. Twenty patients were eligible for the mechanical valve-sparing technique. The protocol for this retrospective study was approved by the local ethics committee.

TABLE I Criteria for Using the Mechanical Prosthetic Valve-Sparing Technique

View Fullscreen

Clinical diagnosis of IE was based on suspicious history and clinical examination according to modified Duke criteria. The presence of aortic root abscess was confirmed by means of transesophageal echocardiography. In-hospital laboratory tests included white blood cell counts, inflammatory marker assays, and blood cultures. Table II summarizes the demographic and clinical characteristics of the 20 patients, all of whom had a St. Jude bileaflet mechanical aortic valve.

TABLE II Demographic and Clinical Characteristics of the 20 Patients Who Underwent Valve-Sparing Surgery

View Fullscreen

Surgical Technique

Median sternotomy and cannulation of the ascending aorta were performed in standard fashion. Next, 19 patients underwent 2-staged right atrial venous cannulation. The remaining patient underwent bicaval cannulation because concomitant mitral valve replacement was needed. Cardiopulmonary bypass was started, and hypothermia was maintained at 32 °C. Myocardial protection was achieved with antegrade blood cardioplegia. After aortotomy, the mechanical valve was carefully examined to confirm that no vegetations were present and that the leaflets moved freely. Finally, we ensured that any paravalvular leakage involved no more than one third of the circumference of the valve cuff.

The valve-sparing technique began with draining the abscess, aggressively débriding infected and necrotic tissue, and repairing the resulting defect with a Gelweave patch (Vascutek), as described in a previous report.⁷ The patch suture, performed in continuous running fashion, was started at the healthy aortic wall and continued to the cuff of the mechanical prosthetic valve (Fig. 1), with care taken to avoid injuring the affected coronary ostium.

View Figure

• Download as Powerpoint
• Download Figure

Postoperatively, patients underwent 6 weeks of antibiotic treatment based on their blood culture findings and sensitivity.

Results

Annular aortic root abscesses were found in all 20 patients who underwent valve-sparing surgery. The mean interval between valve implantation surgery and reoperation was 7.35 ± 4.09 years (range, 1.5–20 yr).

The valve-sparing technique was successful in all the patients. Combined procedures included mitral valve replacement in one patient and coronary artery bypass grafting (CABG) in another. The valve-sparing procedure was elective in 11 patients (55%) and urgent in 9 (45%) because of increasing dyspnea, New York Heart Association class II to IV function, and the need for diuretics. Surgical details are provided in Table III.

TABLE III Surgical Details for the 20 Patients

View Fullscreen

Low cardiac output syndrome necessitating inotropic support developed in 3 patients (15%). New temporary atrioventricular heart block necessitated implantation of a temporary pacemaker in 2 patients (10%), both of whom recovered completely. One patient had postoperative renal impairment but did not need dialysis. One patient had delayed neurologic recovery due to brain edema. Two patients (10%) died early (≤30 d postoperatively) of low cardiac output syndrome with progressive heart failure, superadded septicemia, and multisystem organ failure. Postoperative results are summarized in Table IV.

TABLE IV Postoperative Results in the 20 Patients

View Fullscreen

At 1-year follow-up, 18 patients (90%) were symptom free and had a well-functioning prosthetic aortic valve without paravalvular leak. Two patients had experienced recurrent IE. One of them was treated with antibiotics alone. The other underwent aortic root replacement with a composite graft 8 months after the valve-sparing procedure.

In the 4 patients who died or had a recurrent infection during the postoperative and follow-up periods, the infectious organisms were Staphylococcus aureus (n=3) and enterococci (n=1). The latter infection occurred in one of the patients who died early.

Discussion

Aortic root abscess in patients with an infected mechanical prosthetic aortic valve has long been a surgical challenge. In an autopsy study from 1976, Arnett and colleagues⁸ reported frequent aortic root abscesses, with or without prosthetic valve dehiscence and extension of infection to adjacent structures. The natural history and expected clinical outcome of progressive heart failure and septicemia in such patients mandates urgent surgery.

Different surgical approaches have been used to manage aortic root abscess complicated by IE of a prosthetic aortic valve. Reitz and associates⁹ successfully treated 3 patients by translocating the prosthetic aortic valve, closing the right and left coronary ostia, and then performing CABG with saphenous vein grafts. The groups of Frantz¹⁰ and VanHooser¹¹ reported replacing the aortic root with a composite prosthetic valve–Dacron tube conduit. Glazier and colleagues¹² successfully treated 21 patients by means of homograft aortic root replacement. Symbas and associates¹³ replaced an infected aortic valve and closed the abscess cavity with a synthetic patch. Yet, despite the good results, these techniques were surgically challenging, necessitating lengthy aortic cross-clamp and operative times, which can affect outcomes in such critically ill patients.^9–13

Our valve-sparing technique was performed within relatively short cross-clamp (mean, 67 min) and total bypass (mean, 103 min) times (Table III). In comparison, Nottin and colleagues¹⁴ reported a mean aortic cross-clamp time of 138 min and a mean total bypass time of 236 min for their technique, which involved aortic valve translocation and distal CABG.

The 1-year mortality rate of 10% in our cohort of 20 patients was low, given their increased surgical risk, critical condition, and need for reoperation. In a study of 168 patients over a 13-year period, Elgalad and associates¹⁵ found that outcomes were not affected by surgical complexity or valve substitute. In contrast, Masur and Johnson¹⁶ reported that only one of 14 patients with prosthetic endocarditis survived after medical therapy and surgical prosthetic valve replacement after débridement. Ivert and colleagues¹⁷ reported an overall mortality rate of 36% in a study of 33 patients diagnosed with prosthetic valve endocarditis and treated surgically; of note, 44% of the survivors needed reoperation. In a prospective observational study of patients with IE, Hill and associates¹⁸ found that age, staphylococcal infection, and contraindication to surgery predicted 6-month mortality. In our study, the 1-year recurrence rate of endocarditis after surgical débridement and repair was 10% (2 of 20 patients). We considered it satisfactorily low for our relatively simple and less time-consuming valve-sparing technique.

Although our study included only a small number of patients who were followed for only one year, its results show that our mechanical valve-sparing technique for treating aortic root abscess complicated by IE is safe and less time-consuming than conventional débridement and valve replacement. Studies in larger populations are warranted.