Prosthetic valve sewing-ring sealing with antibiotic and fibrin glue in infective endocarditis. A prospective clinical study
http://www.100md.com
《血管的通路杂志》
a Department of Cardiovascular Surgery B, Claude Bernard University, INSERM E226, Louis Pradel Hospital, 28, avenue du doyen Lepine, 69677 Bron Cedex, France
b Department of Cardiology, Louis Pradel Hospital, Claude Bernard University, UMR MA103, Bron, France
Abstract
To analyze the benefits of prosthetic valve sewing ring sealing with antibiotic and fibrin glue in endocarditis, a prospective study was undertaken in 16 consecutive patients. Valvular lesions concerned aortic valve in seven patients, mitral valve in three and both in six. Preoperative antibiotic treatment was started 15±15 days before surgery. Peroperative findings showed six isolated valvular lesions and ten perivalvular extensions. Before the prostheses were sutured to the annulus, the DacronTM sewing rings were covered with antibiotic and fibrin glue. The antibiotic was chosen according to the micro-organism identified preoperatively. We didn't have any peroperative mortality or paravalvular leakage (PVL). One patient died suddenly on postoperative day (POD) nine (non-contributive autopsy). Another patient was reoperated on POD 31 for PVL. Peroperative local bacteriological examinations were sterile. Mean follow-up was of 9.3±6.8 months and was completed in all cases. Actuarial survival rate was 94±6% and freedom from reoperation was 92±7%. Antibiotic treatment was stopped after 59±64 days for 10 patients. None of them had recurrence of endocarditis or PVL. Prosthetic valve sewing ring sealing with adapted antibiotic and fibrin glue can constitute a good weapon in the armamentarium of surgical treatment of active endocarditis.
Key Words: Endocarditis; Antibiotic/fibrin glue sealant; Parvalvular abscess; Prosthesis infection
1. Introduction
Prosthetic valve infection is a rare (1–9%) but severe complication after valve replacement in active infective endocarditis [1,2]. Regardless to the preoperative condition, it can constitute, in the best cases, a risk factor for paravalvular leakage (PVL) without hemodynamic repercussion, and in the worst, a dramatic sepsis leading to iterative valve replacement and subsequent high mortality risk [1,3].
For this reason, perioperative adapted systemic antibiotherapy is mandatory to prevent the risk of prosthetic valve infection and subsequent PVL, but does not provide a total protection against this risk [4]. One possible reason for this complication is the low antibiotic level in the sewing ring of the prosthesis and the adjacent valvular annulus, despite an adapted antibiotic serum level [5]. Some authors have previously reported in experimental models the benefit of an antibiotic and fibrin glue sealant for the prosthetic sewing ring before valve replacement in active endocarditis [6]. Nevertheless, clinical results remain questionable.
Herein, we report the results of a prospective clinical study of patients treated surgically for active endocarditis and discuss the benefits and risks of the prosthetic ring sealing with an antibiotic and fibrin glue combination.
2. Patients and methods
Between November 2004 and May 2006, 16 consecutive patients (12 male/4 female) operated by one surgeon for active endocarditis were analyzed prospectively. Mean age was 60±9 years, Mean New York Heart Association (NYHA) score was 3.2±0.8 and mean EuroSCORE was 8.9±2.5. Past medical history was marked by renal insufficiency needing dialysis for one, colic neoplasia in two, myeloma in one, arterial hypertension in four, B hepatitis in one and liver transplantation for cirrhosis in one. Among the patients, five had previous cardiac surgeries (four valvular and one coronary).
In all cases, the patients had preoperative transesophageal echographies (TEEs) assessing the valvular lesions and preoperative CT-scans searching for peripheral embolisms. Valvular lesions concerned the aortic valve in seven patients, the mitral valve in three and both valves in six (Table 1). Micro-organisms responsible for the endocarditis were identified preoperatively in all but two patients. Preoperative antibiotic treatment was started for a mean period of 15±15 days (Table 2).
All the interventions were done under general anesthesia, full sternotomy and TEE control (six interventions were emergent, eight were urgent and two were delayed). Arterial cannulation was femoral in two cases and venous cannulation was central for all patients. After cardiopulmonary bypass (CPB) start and cross clamping, the valves were exposed and the lesions analyzed. The cusps (or the previously implanted prostheses) were removed and all the infected tissues were debrided. While six patients had isolated valvular lesions, ten had perivalvular extensions: three had abscesses of the interventricular septum below the right coronary artery, three had lesions of the left fibrous trigone (including one with a left ventricle-to-right atrium fistula), one had an extension to the aortic left anterior commissure and three had mitral commissural lesions (one anterior and two posterior). All the patients with aortic paravalvular involvement had large resection of the suspect lesions followed by exclusion of the culprit zone with an autologous pericardial patch. Before the running suture of the patch was tightened, the cavity underneath was filled with and antibiotic/fibrin glue solution (TissucolTM, Baxter Healthcare).
All the prostheses were attached with 2/0 TeflonTM felt sutures, aortic valves being positioned in supra-annular position. Before the valve was sutured to the annulus, the DacronTM sewing ring was imbibed by a solution of antibiotic (generally 1 g, diluted in 10 ml of saline solution), then it was covered with fibrin glue. The antibiotic was chosen according to the micro-organism identified preoperatively (vancomycin was used otherwise). The rest of the intervention was carried out in a standard manner.
3. Results
3.1. Peroperative data
We did not note any intraoperative mortality. Six patients had aortic valve replacements, three had mitral valve replacements, six had double valve replacements (including one aortic annulus enlargement according to the technique described by Manouguian) and one had mitral valve repair in addition to an aortic valve replacement. Of over 22 implanted prostheses, 18 were mechanical and 4 biological. Intraoperative TEE did not show any paravalvular leakage. Mean CPB and cross-clamping times were, respectively, 78±22 and 57±18 min. Eight patients were weaned from CPB with inotropic support (stopped after a mean period of 2.4±1.6 days).
3.2. Early postoperative course
We did not note any clinical embolic complications or anaphylactic reaction due to the fibrin glue. Mean Troponin I level at 24 h was 5.8±5.6 μg/l and mean blood loss at 24 h was 412±184 ml. Mean intensive care unit stay was 7.8±18.8 days. Bacteriological, as well as pathological examinations, done on valvular fragments or explanted prostheses, confirmed the preoperative blood cultures in all cases. For one patient without preoperative micro-organism identification, Propionibacterium acnes was identified on bacteriological analysis. Thus, the postoperative antibiotherapy type was carried on comparably to the preoperative treatment in most of the patients.
One patient died suddenly in the ward on postoperative day (POD) nine after dialysis (chronic renal failure). The autopsy did not show any anatomical lesion explaining the fatal evolution. Another patient was reoperated on POD 31 for aortic valve disinsertion and PVL. Peroperative findings confirmed a mechanical dehiscence between the left anterior commissure and the right coronary artery without evidence of endocarditic lesions. Reinsertion of the prosthesis, in association with a repair of the aortic root, was performed with an excellent anatomical result. Peroperative local bacteriological examinations were sterile. Six patients required per or postoperative transfusions (red blood cells: 4.3±3.4 units, fresh frozen plasma: 1.0± 1.7 units and platelets: 0.2±0.4 units). Two patients presented with a postoperative atrioventricular block (one needed a permanent pace maker implantation on POD 5). One patient presented with a transitory thrombopenia. Postoperative TEE examination done before discharge did not show any PVL. All the surviving patients were discharged to postoperative rehabilitation for a mean period of one month. Individual data of postoperative evolution are represented in Table 3.
3.3. Late postoperative course
All the patients were contacted via their referent cardiologist for enquiry. All had regular echographic control. Mean follow-up was 9.3±6.8 months and was completed in all cases (median of eight months). Actuarial survival rate after one year was 94±6% and freedom from reoperation was 92±7% (Fig. 1). Antibiotic treatment was stopped after a mean period of 59±64 days postoperatively for 10 patients (one died and five are still ongoing). Mean NYHA score at follow-up was 1.3±0.6. None of them had recurrence of endocarditis of PVL.
4. Discussion
Three indications are widely accepted for urgent or emergent surgical treatment of active endocarditis, overdriving the inherent risk of contamination of the prosthesis itself. These indications are septic condition unresponsive to adapted antibiotherapy, severe hemodynamic state due most of the time to valvular insufficiency or to prosthetic valve thrombosis, and repeated embolisms. Even if the outcome of the patients treated surgically for endocarditis limited to the valve is acceptable and has improved over the past 20 years [7,8], the involvement of the perivalvular tissue adds significant hazard in the early postoperative period.
The extension to the paravalvular tissue represents the natural evolution of the infection of the valvular cusps in the case of native valve, or the prosthetic ring in the case of a pre-existing prosthesis. The degree of the para-annular involvement is tightly linked to the virulence of the micro-organism and to the duration of the infection before antibiotic treatment. This condition can lead to left ventricular-aortic dehiscence, aortic-to-atrial fistula, left fibrous trigone abscess, or ventricular septal defect (VSD). For these reasons, some authors have strongly recommended a large excision of the infected tissues and their replacement by pericardial patches [10], while others have reported complete mitro-aortic monobloc replacements in case of mitro-aortic infection with left fibrous trigone extension [11]. In our opinion, the paradigms of an ideal surgical treatment of active endocarditis remain the same: to debride infected tissue to eradicate the infectious focus, to exclude the abscess cavity from the circulation and subsequently from the prosthesis, to restore ventricular-aortic continuity (in case or aortic endocarditis) without mechanical tension, to ensure a solid anchorage of the prosthesis, and to prevent local recurrence of the infection. This last problem remains the most challenging one. For this reason, the present prospective clinical study was undertaken.
Karck et al. have previously reported the use of an antibiotic and fibrin glue sealant in an experimental study [6]. In vitro, DacronTM circles treated with a combination of a gentamicin derived antibiotic (ATB) and fibrin glue (FG) released ATB for three weeks after treatment. DacronTM fragments implanted into the descending aortas of pigs, after being contaminated with a Staphylococcus aureus solution, showed a persistent concentration of ATB after one week in case of association with FG, contrarily to those treated with ATB alone. Moreover, only 50% of the ATB+FG DacronTM fragments were infected with Staphylococcus aureus after one week of implantation, contrarily to the fragments treated with ATB alone, FG alone, or saline solution (100% of infections). Later on, Watanabe et al., from the same group, used routinely an association of Nebacetin (bacithracin + neomycin) and Tissucol for abscesses and prosthetic sewing rings in patients with active endocarditis [12]. Nevertheless, their attitude was not to adapt the antibiotic to the micro-organism incriminated in the infectious process.
In our series, all the patients but two had identified micro-organisms preoperatively on blood cultures. For this reason, the antibiotic used during the operation for the prostheses sewing rings was adapted as much as possible. For two patients without preoperative documentation, vancomycin was used because of its wide range of action on the most commonly present micro-organisms in valvular endocarditis. The second point concerns the implantation of the aortic prostheses in the case of extension of the endocarditis to the left fibrous trigone. We performed for these patients a systematic exclusion of the infected lesions by using an autologous pericardial patch and by filling the abscess cavity with antibiotic and fibrin glue [13]. In addition, we did a trans-aortic fixation of the sewing ring with TeflonTM felt U-stitches passed through the aortic wall at the level of the non-coronary cusp after dissection of the aortic root from the left atrium. The major benefits of the trans-aortic wall stitches are to preserve the aortic ring fragilized by acute endocarditis and to avoid attaching the valve directly against the abscess. It also gives the possibility of easy reconstruction of the trigone without mechanical stress. The suturing technique using interrupted TeflonTM felt U-stitches could also reduce by itself the incidence of postoperative leakage as suggested by Englberger et al. [14]. The AVERT investigator group has reported that the use of pledgets was an independent factor for reducing PVL after aortic or mitral replacement. In this trial, the SilzoneTM mechanical valve (St. Jude Medical, Minneapolis, USA), a valve with a silver-coated sewing ring known for its antimicrobial protection, presented a higher incidence of postoperative PVL leading to its suspension and recall from market. Nevertheless, the AVERT trial did not show that SilzoneTM coating was an independent factor for PVL.
One pitfall of our approach could be the risk of embolism of fibrin glue fragments in the bloodstream. We did not encounter any clinical evidence of embolism in the immediate postoperative period, even if we did not perform a systematic CT-scan to eliminate such a complication. We did not note any recurrence of endocarditis or PVL during the follow up period. Four patients were still ongoing antibiotic treatment without problems of tolerance.
In conclusion, prosthetic valve sewing ring sealing with antibiotic and fibrin glue in active endocarditis could be an interesting weapon in the armamentarium of the valvular endocarditis surgery. In our series, survival is acceptable and the case of paravalvlar leakage did not seem to be related to a local recurrence of endocarditis. Furthermore, application of fibrin glue on the sewing rings does not seem to constitute a risk for systemic embolisms. Longer follow-up and larger series are mandatory to affirm the interest of this technique, as well as a prospective randomized study.
References
Leblebicioglu H, Yilmaz H, Tasova Y, Alp E, Saba R, Caylan R, Bakir M, Akbulut A, Ara B, Eses S. Characteristics and analysis of risk factors for mortality in infective endocarditis. Eur J Epidemiol 2006; 21:25–31.
Farinas MC, Perez-Vasquez A, Farinas-Alvarez C, Garcia-Palomo JD, Bernal JM, Revuelta JM, Gonzalez-Macias J. Risk factors of prosthetic valve endocarditis: a case-control study. Ann Thorac Surg 2006; 81:1284–1290.
Ivert TS, Dismukes WE, Cobbs CG, Blackstone EH, Kirklin JW, Bergdahi LA. Prosthetic valve endocarditis. Circulation 1984; 69:223–232.
Witchitz S, Reidiboym M, Bouvet E, Wolff M, Vachon F. [Outcome of prognostic factors of infectious endocarditis over a 16-year period. A propos of 471 cases]. Arch Mal Coeur Vaiss 1992; 85:959–965.
Lau JK, Robles A, Cherian A, Ross DB. Surgical treatment of prosthetic endocarditis. Aortic root replacement using a homograft. J Thorac Cardiovasc Surg 1984; 87:712–716.
Karck M, Siclari F, Wahling H, Sperling U, Schmid C, Haverich A. Pretreatment of prosthetic valve sewing-ring with antibiotic/fibrin glue sealant compound as a prophylactic tool against prosthetic valve endocarditis. Eur J Cardiothorac Surg 1990; 4:142–146.
D'Agostino RS, Miller DC, Stinson EB, Mitchell RS, Oyer PE, Jamieson SW, Baldwin JC, Shumway NE. Valve replacement in patients with native valve endocarditis: what really determines operative outcome. Ann Thorac Surg 1985; 40:429–438.
Remadi JP, Najdi G, Brahim A, Coviaux F, Majhoub Y, Tribouilloy C. Superiority of surgical versus medial treatment in patients with Staphylococcus aureus infective endocarditis. Int J Cardiol 2005; 99:195–199.
Ergin MA, Raissi S, Follis F, Lansman SL, Griepp RB. Annular destruction in acute bacterial endocarditis. Surgical techniques to meet the challenge. J Thorac Cardiovasc Surg 1989; 97:755–763.
David TE. The surgical treatment of patients with prosthetic valve endocarditis. Semin Thorac Cardiovasc Surg 1995; 7:47–53.
Obadia JF, Henaine R, Bergerot C, Ginon I, Nataf P, Chavanis N, Robin J, André-Fouet X, Ninet J, Raisky O. Monobloc aorto-mitral homograft or mechanical valve replacement: a new surgical option for extensive bivalvular endocarditis. J Thorac Cardiovasc Surg 2006; 131:243–245.
Watanabe G, Haverich A, Speier R, Dresler C, Borst HG. Surgical treatment of active infective endocarditis with paravalvular involvement. J Thorac Cardiovasc Surg 1994; 107:171–177.
Symbas PN, Vlasis SE, Zacharopoulos L, Lutz JF. Acute endocarditis: surgical treatment of aortic regurgitation and aortico-left ventricular discontinuity. J Thorac Cardiovasc Surg 1982; 84:291–296.
Englberger L, Schaff HV, Jamieson WR, Kennard ED, Im KA, Holubkov R, Carrel TP. the AVERT Investigators. Importance of implant technique on risk of major paravalvular leak (PVL) after St. Jude mechanical heart valve replacement: a report from the Artificial Valve Endocarditis Reduction Trial (AVERT). Eur J Cardiothorac Surg 2005; 28:838–843.(Fadi Farhata,, Marion Duranda, Franois D)
b Department of Cardiology, Louis Pradel Hospital, Claude Bernard University, UMR MA103, Bron, France
Abstract
To analyze the benefits of prosthetic valve sewing ring sealing with antibiotic and fibrin glue in endocarditis, a prospective study was undertaken in 16 consecutive patients. Valvular lesions concerned aortic valve in seven patients, mitral valve in three and both in six. Preoperative antibiotic treatment was started 15±15 days before surgery. Peroperative findings showed six isolated valvular lesions and ten perivalvular extensions. Before the prostheses were sutured to the annulus, the DacronTM sewing rings were covered with antibiotic and fibrin glue. The antibiotic was chosen according to the micro-organism identified preoperatively. We didn't have any peroperative mortality or paravalvular leakage (PVL). One patient died suddenly on postoperative day (POD) nine (non-contributive autopsy). Another patient was reoperated on POD 31 for PVL. Peroperative local bacteriological examinations were sterile. Mean follow-up was of 9.3±6.8 months and was completed in all cases. Actuarial survival rate was 94±6% and freedom from reoperation was 92±7%. Antibiotic treatment was stopped after 59±64 days for 10 patients. None of them had recurrence of endocarditis or PVL. Prosthetic valve sewing ring sealing with adapted antibiotic and fibrin glue can constitute a good weapon in the armamentarium of surgical treatment of active endocarditis.
Key Words: Endocarditis; Antibiotic/fibrin glue sealant; Parvalvular abscess; Prosthesis infection
1. Introduction
Prosthetic valve infection is a rare (1–9%) but severe complication after valve replacement in active infective endocarditis [1,2]. Regardless to the preoperative condition, it can constitute, in the best cases, a risk factor for paravalvular leakage (PVL) without hemodynamic repercussion, and in the worst, a dramatic sepsis leading to iterative valve replacement and subsequent high mortality risk [1,3].
For this reason, perioperative adapted systemic antibiotherapy is mandatory to prevent the risk of prosthetic valve infection and subsequent PVL, but does not provide a total protection against this risk [4]. One possible reason for this complication is the low antibiotic level in the sewing ring of the prosthesis and the adjacent valvular annulus, despite an adapted antibiotic serum level [5]. Some authors have previously reported in experimental models the benefit of an antibiotic and fibrin glue sealant for the prosthetic sewing ring before valve replacement in active endocarditis [6]. Nevertheless, clinical results remain questionable.
Herein, we report the results of a prospective clinical study of patients treated surgically for active endocarditis and discuss the benefits and risks of the prosthetic ring sealing with an antibiotic and fibrin glue combination.
2. Patients and methods
Between November 2004 and May 2006, 16 consecutive patients (12 male/4 female) operated by one surgeon for active endocarditis were analyzed prospectively. Mean age was 60±9 years, Mean New York Heart Association (NYHA) score was 3.2±0.8 and mean EuroSCORE was 8.9±2.5. Past medical history was marked by renal insufficiency needing dialysis for one, colic neoplasia in two, myeloma in one, arterial hypertension in four, B hepatitis in one and liver transplantation for cirrhosis in one. Among the patients, five had previous cardiac surgeries (four valvular and one coronary).
In all cases, the patients had preoperative transesophageal echographies (TEEs) assessing the valvular lesions and preoperative CT-scans searching for peripheral embolisms. Valvular lesions concerned the aortic valve in seven patients, the mitral valve in three and both valves in six (Table 1). Micro-organisms responsible for the endocarditis were identified preoperatively in all but two patients. Preoperative antibiotic treatment was started for a mean period of 15±15 days (Table 2).
All the interventions were done under general anesthesia, full sternotomy and TEE control (six interventions were emergent, eight were urgent and two were delayed). Arterial cannulation was femoral in two cases and venous cannulation was central for all patients. After cardiopulmonary bypass (CPB) start and cross clamping, the valves were exposed and the lesions analyzed. The cusps (or the previously implanted prostheses) were removed and all the infected tissues were debrided. While six patients had isolated valvular lesions, ten had perivalvular extensions: three had abscesses of the interventricular septum below the right coronary artery, three had lesions of the left fibrous trigone (including one with a left ventricle-to-right atrium fistula), one had an extension to the aortic left anterior commissure and three had mitral commissural lesions (one anterior and two posterior). All the patients with aortic paravalvular involvement had large resection of the suspect lesions followed by exclusion of the culprit zone with an autologous pericardial patch. Before the running suture of the patch was tightened, the cavity underneath was filled with and antibiotic/fibrin glue solution (TissucolTM, Baxter Healthcare).
All the prostheses were attached with 2/0 TeflonTM felt sutures, aortic valves being positioned in supra-annular position. Before the valve was sutured to the annulus, the DacronTM sewing ring was imbibed by a solution of antibiotic (generally 1 g, diluted in 10 ml of saline solution), then it was covered with fibrin glue. The antibiotic was chosen according to the micro-organism identified preoperatively (vancomycin was used otherwise). The rest of the intervention was carried out in a standard manner.
3. Results
3.1. Peroperative data
We did not note any intraoperative mortality. Six patients had aortic valve replacements, three had mitral valve replacements, six had double valve replacements (including one aortic annulus enlargement according to the technique described by Manouguian) and one had mitral valve repair in addition to an aortic valve replacement. Of over 22 implanted prostheses, 18 were mechanical and 4 biological. Intraoperative TEE did not show any paravalvular leakage. Mean CPB and cross-clamping times were, respectively, 78±22 and 57±18 min. Eight patients were weaned from CPB with inotropic support (stopped after a mean period of 2.4±1.6 days).
3.2. Early postoperative course
We did not note any clinical embolic complications or anaphylactic reaction due to the fibrin glue. Mean Troponin I level at 24 h was 5.8±5.6 μg/l and mean blood loss at 24 h was 412±184 ml. Mean intensive care unit stay was 7.8±18.8 days. Bacteriological, as well as pathological examinations, done on valvular fragments or explanted prostheses, confirmed the preoperative blood cultures in all cases. For one patient without preoperative micro-organism identification, Propionibacterium acnes was identified on bacteriological analysis. Thus, the postoperative antibiotherapy type was carried on comparably to the preoperative treatment in most of the patients.
One patient died suddenly in the ward on postoperative day (POD) nine after dialysis (chronic renal failure). The autopsy did not show any anatomical lesion explaining the fatal evolution. Another patient was reoperated on POD 31 for aortic valve disinsertion and PVL. Peroperative findings confirmed a mechanical dehiscence between the left anterior commissure and the right coronary artery without evidence of endocarditic lesions. Reinsertion of the prosthesis, in association with a repair of the aortic root, was performed with an excellent anatomical result. Peroperative local bacteriological examinations were sterile. Six patients required per or postoperative transfusions (red blood cells: 4.3±3.4 units, fresh frozen plasma: 1.0± 1.7 units and platelets: 0.2±0.4 units). Two patients presented with a postoperative atrioventricular block (one needed a permanent pace maker implantation on POD 5). One patient presented with a transitory thrombopenia. Postoperative TEE examination done before discharge did not show any PVL. All the surviving patients were discharged to postoperative rehabilitation for a mean period of one month. Individual data of postoperative evolution are represented in Table 3.
3.3. Late postoperative course
All the patients were contacted via their referent cardiologist for enquiry. All had regular echographic control. Mean follow-up was 9.3±6.8 months and was completed in all cases (median of eight months). Actuarial survival rate after one year was 94±6% and freedom from reoperation was 92±7% (Fig. 1). Antibiotic treatment was stopped after a mean period of 59±64 days postoperatively for 10 patients (one died and five are still ongoing). Mean NYHA score at follow-up was 1.3±0.6. None of them had recurrence of endocarditis of PVL.
4. Discussion
Three indications are widely accepted for urgent or emergent surgical treatment of active endocarditis, overdriving the inherent risk of contamination of the prosthesis itself. These indications are septic condition unresponsive to adapted antibiotherapy, severe hemodynamic state due most of the time to valvular insufficiency or to prosthetic valve thrombosis, and repeated embolisms. Even if the outcome of the patients treated surgically for endocarditis limited to the valve is acceptable and has improved over the past 20 years [7,8], the involvement of the perivalvular tissue adds significant hazard in the early postoperative period.
The extension to the paravalvular tissue represents the natural evolution of the infection of the valvular cusps in the case of native valve, or the prosthetic ring in the case of a pre-existing prosthesis. The degree of the para-annular involvement is tightly linked to the virulence of the micro-organism and to the duration of the infection before antibiotic treatment. This condition can lead to left ventricular-aortic dehiscence, aortic-to-atrial fistula, left fibrous trigone abscess, or ventricular septal defect (VSD). For these reasons, some authors have strongly recommended a large excision of the infected tissues and their replacement by pericardial patches [10], while others have reported complete mitro-aortic monobloc replacements in case of mitro-aortic infection with left fibrous trigone extension [11]. In our opinion, the paradigms of an ideal surgical treatment of active endocarditis remain the same: to debride infected tissue to eradicate the infectious focus, to exclude the abscess cavity from the circulation and subsequently from the prosthesis, to restore ventricular-aortic continuity (in case or aortic endocarditis) without mechanical tension, to ensure a solid anchorage of the prosthesis, and to prevent local recurrence of the infection. This last problem remains the most challenging one. For this reason, the present prospective clinical study was undertaken.
Karck et al. have previously reported the use of an antibiotic and fibrin glue sealant in an experimental study [6]. In vitro, DacronTM circles treated with a combination of a gentamicin derived antibiotic (ATB) and fibrin glue (FG) released ATB for three weeks after treatment. DacronTM fragments implanted into the descending aortas of pigs, after being contaminated with a Staphylococcus aureus solution, showed a persistent concentration of ATB after one week in case of association with FG, contrarily to those treated with ATB alone. Moreover, only 50% of the ATB+FG DacronTM fragments were infected with Staphylococcus aureus after one week of implantation, contrarily to the fragments treated with ATB alone, FG alone, or saline solution (100% of infections). Later on, Watanabe et al., from the same group, used routinely an association of Nebacetin (bacithracin + neomycin) and Tissucol for abscesses and prosthetic sewing rings in patients with active endocarditis [12]. Nevertheless, their attitude was not to adapt the antibiotic to the micro-organism incriminated in the infectious process.
In our series, all the patients but two had identified micro-organisms preoperatively on blood cultures. For this reason, the antibiotic used during the operation for the prostheses sewing rings was adapted as much as possible. For two patients without preoperative documentation, vancomycin was used because of its wide range of action on the most commonly present micro-organisms in valvular endocarditis. The second point concerns the implantation of the aortic prostheses in the case of extension of the endocarditis to the left fibrous trigone. We performed for these patients a systematic exclusion of the infected lesions by using an autologous pericardial patch and by filling the abscess cavity with antibiotic and fibrin glue [13]. In addition, we did a trans-aortic fixation of the sewing ring with TeflonTM felt U-stitches passed through the aortic wall at the level of the non-coronary cusp after dissection of the aortic root from the left atrium. The major benefits of the trans-aortic wall stitches are to preserve the aortic ring fragilized by acute endocarditis and to avoid attaching the valve directly against the abscess. It also gives the possibility of easy reconstruction of the trigone without mechanical stress. The suturing technique using interrupted TeflonTM felt U-stitches could also reduce by itself the incidence of postoperative leakage as suggested by Englberger et al. [14]. The AVERT investigator group has reported that the use of pledgets was an independent factor for reducing PVL after aortic or mitral replacement. In this trial, the SilzoneTM mechanical valve (St. Jude Medical, Minneapolis, USA), a valve with a silver-coated sewing ring known for its antimicrobial protection, presented a higher incidence of postoperative PVL leading to its suspension and recall from market. Nevertheless, the AVERT trial did not show that SilzoneTM coating was an independent factor for PVL.
One pitfall of our approach could be the risk of embolism of fibrin glue fragments in the bloodstream. We did not encounter any clinical evidence of embolism in the immediate postoperative period, even if we did not perform a systematic CT-scan to eliminate such a complication. We did not note any recurrence of endocarditis or PVL during the follow up period. Four patients were still ongoing antibiotic treatment without problems of tolerance.
In conclusion, prosthetic valve sewing ring sealing with antibiotic and fibrin glue in active endocarditis could be an interesting weapon in the armamentarium of the valvular endocarditis surgery. In our series, survival is acceptable and the case of paravalvlar leakage did not seem to be related to a local recurrence of endocarditis. Furthermore, application of fibrin glue on the sewing rings does not seem to constitute a risk for systemic embolisms. Longer follow-up and larger series are mandatory to affirm the interest of this technique, as well as a prospective randomized study.
References
Leblebicioglu H, Yilmaz H, Tasova Y, Alp E, Saba R, Caylan R, Bakir M, Akbulut A, Ara B, Eses S. Characteristics and analysis of risk factors for mortality in infective endocarditis. Eur J Epidemiol 2006; 21:25–31.
Farinas MC, Perez-Vasquez A, Farinas-Alvarez C, Garcia-Palomo JD, Bernal JM, Revuelta JM, Gonzalez-Macias J. Risk factors of prosthetic valve endocarditis: a case-control study. Ann Thorac Surg 2006; 81:1284–1290.
Ivert TS, Dismukes WE, Cobbs CG, Blackstone EH, Kirklin JW, Bergdahi LA. Prosthetic valve endocarditis. Circulation 1984; 69:223–232.
Witchitz S, Reidiboym M, Bouvet E, Wolff M, Vachon F. [Outcome of prognostic factors of infectious endocarditis over a 16-year period. A propos of 471 cases]. Arch Mal Coeur Vaiss 1992; 85:959–965.
Lau JK, Robles A, Cherian A, Ross DB. Surgical treatment of prosthetic endocarditis. Aortic root replacement using a homograft. J Thorac Cardiovasc Surg 1984; 87:712–716.
Karck M, Siclari F, Wahling H, Sperling U, Schmid C, Haverich A. Pretreatment of prosthetic valve sewing-ring with antibiotic/fibrin glue sealant compound as a prophylactic tool against prosthetic valve endocarditis. Eur J Cardiothorac Surg 1990; 4:142–146.
D'Agostino RS, Miller DC, Stinson EB, Mitchell RS, Oyer PE, Jamieson SW, Baldwin JC, Shumway NE. Valve replacement in patients with native valve endocarditis: what really determines operative outcome. Ann Thorac Surg 1985; 40:429–438.
Remadi JP, Najdi G, Brahim A, Coviaux F, Majhoub Y, Tribouilloy C. Superiority of surgical versus medial treatment in patients with Staphylococcus aureus infective endocarditis. Int J Cardiol 2005; 99:195–199.
Ergin MA, Raissi S, Follis F, Lansman SL, Griepp RB. Annular destruction in acute bacterial endocarditis. Surgical techniques to meet the challenge. J Thorac Cardiovasc Surg 1989; 97:755–763.
David TE. The surgical treatment of patients with prosthetic valve endocarditis. Semin Thorac Cardiovasc Surg 1995; 7:47–53.
Obadia JF, Henaine R, Bergerot C, Ginon I, Nataf P, Chavanis N, Robin J, André-Fouet X, Ninet J, Raisky O. Monobloc aorto-mitral homograft or mechanical valve replacement: a new surgical option for extensive bivalvular endocarditis. J Thorac Cardiovasc Surg 2006; 131:243–245.
Watanabe G, Haverich A, Speier R, Dresler C, Borst HG. Surgical treatment of active infective endocarditis with paravalvular involvement. J Thorac Cardiovasc Surg 1994; 107:171–177.
Symbas PN, Vlasis SE, Zacharopoulos L, Lutz JF. Acute endocarditis: surgical treatment of aortic regurgitation and aortico-left ventricular discontinuity. J Thorac Cardiovasc Surg 1982; 84:291–296.
Englberger L, Schaff HV, Jamieson WR, Kennard ED, Im KA, Holubkov R, Carrel TP. the AVERT Investigators. Importance of implant technique on risk of major paravalvular leak (PVL) after St. Jude mechanical heart valve replacement: a report from the Artificial Valve Endocarditis Reduction Trial (AVERT). Eur J Cardiothorac Surg 2005; 28:838–843.(Fadi Farhata,, Marion Duranda, Franois D)