Composite Aortic Root Replacement for Complex Prosthetic Valve Endocarditis: Initial Clinical Results and Long-Term Follow-Up of High-Risk Patients

Transcription

1 Composite Aortic Root Replacement for Complex Prosthetic Valve Endocarditis: Initial Clinical Results and Long-Term Follow-Up of High-Risk Patients Manuel Wilbring, MD, Sems Malte Tugtekin, MD, Konstantin Alexiou, MD, Klaus Matschke, MD, and Utz Kappert, MD Department of Cardiac Surgery, University Heart Center Dresden, Dresden, Germany Background. Little information is available regarding the clinical and long-term results of patients with complex prosthetic valve endocarditis (PVE) involving the aortic root who undergo root replacement based on the Cabrol or Bentall procedures. Methods. Between January 2007 and December 2011, 148 patients underwent cardiac operations for PVE. The analysis included 31 patients with complex PVE and concomitant destruction of the aortic root. Of these, 13 patients were treated by the Cabrol procedure and 18 patients by the Bentall procedure. The mean EuroSCORE for mortality was 50.7% 3.8%. Mean follow-up was years (range, < 8.0 years), with a total of 97 patient-years. Results. The observed 30-day mortality was 12.9% and 5-year survival was 75.3%. The most common cause of death was septic multiple organ failure (42.9%). Independent predictors of mortality in multivariate analyses were terminal renal failure (odds ratio [OR], 4.8; p < 0.01), type 2 diabetes mellitus (OR, 4.6; p < 0.01), postoperative renal failure (OR, 4.0; p < 0.01), and staphylococcal infection (OR, 2.1; p 0.01). The prevalence of freedom from reinfection was 100.0% and that from valve-related events was 93.5%. Conclusions. Complex PVE is associated with quite high mortality and morbidity. Composite aortic root replacement provided good clinical and long-term outcomes as well as a low prevalence of reinfection and valve-related events. These results seem not to be inferior to those reported for noncomplex PVE. If the Bentall button technique was not feasible, the Cabrol procedure also provided excellent results. (Ann Thorac Surg 2012;94: ) 2012 by The Society of Thoracic Surgeons Infective endocarditis is possibly the most serious complication after valve replacement. About 1% to 6% of patients with valve prostheses can subsequently be diagnosed with prosthetic valve endocarditis (PVE) [1]. Despite improvements in medical and surgical therapy over the past decade, PVE is associated with a high prevalence of morbidity and mortality. Medical treatment alone is associated with poor outcome, so combination with surgical therapy is quite common. Nonetheless, surgical treatment of PVE has a mortality of 20% to 80% [2 6]. Patients who also have aortic root or coronary ostia are a particularly high-risk subgroup within patients with PVE. Composite aortic root replacement using mechanical or biological composite grafts or homografts is often needed but is associated with poor outcome [7]. Decision making regarding the indications for surgical treatment in patients with PVE is often based on common sense and experience [6]. Further risk analyses and characterization of perioperative features as well as their influence on the postoperative course are needed to Accepted for publication June 7, Address correspondence to Dr Wilbring, Department of Cardiac Surgery, University Heart Center Dresden, Fetscherstrasse 76, Dresden, Germany; manuel.wilbring@gmail.com. improve surgical results [6]. Studies on the long-term results of patients with PVE treated with aortic composite replacement are lacking. In particular, for the Cabrol procedure the prevalence of reinfection and the fate of the intercoronary graft remain unclear. We describe here a series of high-risk patients with PVE with aortic root or coronary ostia treated with modified versions of the Cabrol procedure and Bentall procedure. Characterization of preoperative and intraoperative data as well as the initial clinical and long-term results was a particular focus of the present study. Patients and Methods Ethical Approval of the Study Protocol The study protocol was approved by the Ethics Committee of University Heart Center Dresden (Dresden, Germany). All patients gave informed consent for their data to be used. The study was designed according to the Declaration of Helsinki. Study Design and Patients Between January 2007 and December 2010, 148 patients underwent cardiac operations for PVE. Of these, by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc

2 1968 WILBRING ET AL Ann Thorac Surg COMPLEX PROSTHETIC VALVE ENDOCARDITIS 2012;94: Abbreviations and Acronyms CABG coronary artery bypass grafting COPD chronic obstructive pulmonary disease CVVH Continuous venovenous hemofiltration EuroSCORE European System for Cardiac Operative Risk Evaluation IABP intraaortic balloon pump ICU intensive care unit LCOS low cardiac output syndrome OR odds ratio PVE prosthetic valve endocarditis ROC receiver operating characteristic patients (20.9%) presented with complex PVE with additional destruction of the aortic root or coronary ostia, or both. Aortic root replacement was undertaken using the Bentall procedure in 18 patients (58.1%) and a modified Cabrol procedure in 13 patients (41.9%). The study was designed as a retrospective data analysis. Perioperative data were collected from the database of our hospital and evaluated. Follow-up was by telephone or contact with the patient or his/her cardiologist. All survivors were successfully followed. The mean follow-up time was years (range, years), with a total of 97 patient-years. Follow-up included survival, cause of death, valve-related events, and recurrence of endocarditis. Preoperative Management and Indications for Surgical Treatment The indications for surgical treatment were based on clinical and hemodynamic presentation, as well as echocardiographic and microbiological findings. This method was in accordance with current guidelines set by the American College of Cardiology/American Heart Association and the European Society of Cardiology [8, 9]. Typical indications for operation were severe hemodynamic impairment resulting from septic conditions, valvular dysfunction due to vegetations, recurrent major embolism or large vegetations, and abscess or perivalvular involvement [6]. Blood for cultures was taken preoperatively and transesophageal echocardiography was performed in each patient. Early PVE was defined as PVE that occurred within the first 12 months after initial valve replacement. PVE that occurred later than 12 months after initial valve replacement was defined as late PVE. Surgical Management All procedures were carried out by median sternotomy and use of extracorporeal circulation. Aortic cannulation was typically carried out in the aortic arch or brachiocephalic trunk. Venous cannulation was performed using a 2-stage cannula inserted at the right atrial appendage. All procedures were done under normothermia. Antegrade crystalloid cardioplegia was used for myocardial protection. After excision of the infected valve, radical debridement of the infected tissue was completed. The conduits were implanted using interrupted stitches with pledgets in all cases. Excised tissue and valves were sent for microbiological examination. Composite aortic root replacement was made using biological or mechanical conduits. Because of a restricted availability at our institution, homografts were not used. Aortic root replacement was typically done as a button procedure (a modified version of the Bentall approach) [10], as described by Kouchoukos and colleagues [11]. Only in the case of coronary ostia or tight adhesions carrying a risk of strong traction at the level of the reimplanted ostia (which could cause uncontrollable hemorrhage) did we opt for the Cabrol method of root replacement [12]. The Cabrol method was undertaken primarily in the conventional manner; only slight modifications were used. The major concern was to avoid kinking of the intercoronary anastomosis and consequent coronary malperfusion. First the anastomosis of the left coronary artery was performed in a 90-degree shape. Afterward we retained some volume in the right ventricle to estimate the optimal length of the intercoronary graft. Subsequently, the anastomosis of the right coronary artery was done in a 45-degree shape instead of a 90-degree shape to prevent kinking. For the coronary reattachment we used a 6-mm Gore-Tex (Gore-Tex & Associates, Flagstaff, AZ) prosthesis instead of the commonly used 8-mm polyester prosthesis. We believe the Gore-Tex prosthesis to be less susceptible to kinking (Fig 1). Postoperative Management Antibiotic treatment was adapted according to microbiological findings. According to current guidelines of the European Society of Cardiology, a triple combination using vancomycin, gentamicin, and rifampicin was initiated. Intravenous antibiotic treatment was administered for 6 weeks or more after operation. Fig 1. Intraoperative image showing the Cabrol procedure. The ostial anastomosis in right coronary artery is modified by carrying out a 45-degree shaped anastomosis to avoid kinking (arrow).

3 Ann Thorac Surg WILBRING ET AL 2012;94: COMPLEX PROSTHETIC VALVE ENDOCARDITIS 1969 Table 1. Baseline Characteristics and Associated Mortality of the Patient Population Variable n % Overall Mortality p Value All patients % 22.6%... Logistic EuroSCORE (%) Age (y) a Male sex % 24.0% 0.46 Atrial fibrillation % 30.8% 0.25 Type 2 diabetes mellitus % 45.5% 0.04 a COPD None Extracardiac arteriopathy % 33.3% 0.31 Chronic heart failure (EF 30%) 2 6.5% 50.0% 0.19 Pulmonary hypertension (systolic PAP 60 mm Hg) % 50.0% 0.17 Coronary artery disease % 25.0% 0.68 Renal impairment (creatinine level 200 mol/dl) % 16.7% 0.89 Preoperative dialysis 2 6.5% 100% 0.04 a Preoperative neuroembolic event % 40.0% 0.48 Embolic myocardial infarction % 0% 0.35 a Significant p value. COPD chronic obstructive pulmonary disease; EF ejection fraction; EuroSCORE European System for Cardiac Operative Risk Evaluation; PAP pulmonary artery pressure. Statistical Analyses Univariate analyses for the risk factors of death were carried out. Relevant factors in univariate analyses were included in multivariate testing. Multivariate analyses were undertaken in a normal logistic regression model. Statistical analyses were carried out with SAS JMP 9.0 software (SAS Institute, Cary, NC). Categorical data are presented as absolute numbers and percentages. Numerical variables are expressed as means with the standard error of means. Continuous variables were compared using the Student s t test and dichotomous variables by Fisher s exact test. Time-to-event analyses were made using Kaplan-Meier analyses and the log-rank test. For relevant numerical variables, a receiver operating characteristic (ROC) analysis was also made. A p value less than 0.05 was considered significant for all tests. If the p value was between 0.05 and 0.10, a trend was assigned. Results Patient Characteristics and Preoperative Course Demographic data and baseline patient characteristics are summarized in Table 1. The mean age of patients was years. The logistic EuroSCORE for mortality was computed to be 50.7% 3.8%. Previous cardiac operations involved isolated aortic valve replacement in 17 patients (54.8%), the Bentall procedure in 12 patients (38.7%), and combined aortic valve and mitral valve replacement in 2 patients (6.4%). Concomitant coronary artery bypass grafting (CABG) had been performed in 2 patients. In 3 patients (9.7%), an indication for previous surgical treatment had been native valve endocarditis. The mean interval from previous cardiac operations was years (range, 0 19 years; median, 1.5 years). Preoperative echocardiography revealed abscess formation that affected the perivalvular annulus and destroyed the aortic root in all patients. Nine patients (29.0%) presented with additional infection of the mitral valve, and 1 patient had nonendocarditic (but relevant) tricuspid regurgitation. Early PVE was documented In 5 patients (16.1%). Major preoperative complications were cerebral embolic events in 5 patients (16.1%) and embolic myocardial infarction in 4 patients (12.9%). Microbiological Findings Blood for cultures and intraoperative specimens were taken from all patients. Positive blood culture results revealed the microbiological agent in 83.9% (n 26/31) of patients. Intraoperative specimens identified the relevant bacteria in only 19.4% of cases. Staphylococcal species were found in 45.2% of patients (n 14/31), predominantly Staphylococcus aureus (n 10/14). Enterococcus faecalis was identified as the causative microorganism in 4 patients (12.9%). A further 9 cases (29.0%) were caused by streptococcal species. In 12.9% (n 4/31) of patients, no pathogen could be identified. Staphylococcal infection was predictive for higher mortality (Table 2). Postoperative Antibiotic Treatment Each patient was initially treated with a triple combination of vancomycin, gentamicin, and rifampicin. Vancomycin was administered for at least 6 weeks postoperatively, with a target serum level between 25 and 30 mg/l. Gentamicin (dose of 3 mg/kg/d) was given for at least 2 weeks postoperatively and then replaced by ciprofloxacin (dose of 400 mg twice a day) for a further 4 weeks. Rifampicin (dose of 900 mg once a day) was administered for 6 weeks postoperatively.

4 1970 WILBRING ET AL Ann Thorac Surg COMPLEX PROSTHETIC VALVE ENDOCARDITIS 2012;94: Table 2. Postoperative Course and Associated In-Hospital Mortality Variable n % 30-Day Mortality (%) p Value Hospital deaths Septic multiple organ failure 3 75 LCOS 1 25 ICU stay 24 h h h Ventilation time 12 h h h Rethoracotomy Bleeding (total amount) (ml) Blood units Deaths: a Survivors: Fresh frozen plasma units Renal failure (needing CVVH) a Respiratory failure (needing noninvasive ventilation or reintubation) LCOS (needing IABP) Postoperative stroke Postoperative delirium Postoperative pacemaker implantation Wound healing complications a Significant p value. CVVH continuous venovenous hemofiltration; IABP intraaortic balloon pump; ICU intensive care unit; LCOS low cardiac output syndrome. Operative Data and Surgical Pathologic Features Nearly half of the procedures were emergencies (n 14/31 [45.2%]). A further 17 procedures were urgent. All patients presented with abscess formation involving the aortic annulus or aortic-mitral junction. Nine patients had mitral valve involvement, but no ventricular septal defects were observed. In 18 patients (58.1%), composite aortic root replacement was performed using the Bentall procedure. The remaining 13 patients (41.9%) were treated with the Cabrol procedure. Therefore the indications for the Cabrol procedure were severe destruction of the aortic root caused by the formation of a subcoronary abscess or involvement of the coronary ostia. A mechanical conduit was implanted in 58.1% of patients, whereas a biological conduit was implanted in 41.9% of patients. All substitute grafts were implanted with interrupted sutures. Concomitant repair or replacement of the mitral valve became necessary in 9 patients (29.0%), and 1 patient underwent additional repair of the tricuspid valve. Surgical treatment of concomitant coronary artery disease was completed in 2 patients. The mean time of the procedure was minutes and did not differ significantly between the Cabrol procedure and the Bentall procedure ( minutes and minutes, respectively; p 0.145). A prolonged procedure longer than 270 minutes was associated with higher mortality (p 0.02). The predictive trend in univariate analyses was doublevalve endocarditis (p 0.06) (Table 3). In-Hospital Morbidity and Mortality Four patients died (12.3%) within the first 30 postoperative days. The reasons for death were persistent sepsis with consecutive multiple organ failure in 3 patients and low cardiac output syndrome (LCOS) in 1 patient. No patient died during the surgical procedure. Most patients had a complex postoperative course with a stay in the intensive care unit (ICU) greater than 72 hours (n 1 8/31 [58.1%]). Only 6 patients (19.5%) were referred to a peripheral hospital ward within 24 hours. Accordingly, most patients had long ventilation times (Table 2). Severe preoperative septic conditions led to unregulated postoperative coagulation and a mean bleeding volume of ml. Consecutive units of blood ( units) and units of fresh frozen plasma ( units) were administered. In 6 patients (19.4%) rethoracotomy became necessary because of postoperative bleeding. The most common postoperative complication was respiratory failure with the need for noninvasive ventilation or even reintubation in 8 patients (26.7%). Postoperative renal failure necessitating hemofiltration was observed in 6 patients (20.0%). Postoperative stroke was diagnosed in

5 Ann Thorac Surg WILBRING ET AL 2012;94: COMPLEX PROSTHETIC VALVE ENDOCARDITIS 1971 Table 3. Operative Data and Intraoperative Disease Variable n % Overall Mortality (%) p Value Second cardiovascular operation Third cardiovascular operation Fourth cardiovascular operation Previous replacement of the aortic valve Previous concomitant aortic operation Previous concomitant mitral operation Previous concomitant CABG Previous operation due to endocarditis Site of infection Aortic valve Aortic and mitral valves Involvement and destruction of aortic root Bentall procedure Cabrol procedure Concomitant repair of the mitral valve Concomitant replacement of the mitral valve Concomitant repair of the tricuspid valve Concomitant CABG Biological substitute Mechanical substitute Procedure time a 270 min Deaths: min Survivors: min 0.02 b a Determined by receiver operating characteristic analyses; area under the curve b Significant p value. CABG coronary artery bypass grafting. 2 patients (6.7%). Postoperative delirium was noted in 7 patients (23.3%). Implantation of an intraaortic balloon pump became necessary in 3 patients (9.7%) with postoperative LCOS. One patient (3.2%) needed pacemaker implantation because of total atrioventricular arrest. Four patients (12.9%) experienced wound healing complications. Postoperative renal failure indicated poor inhospital outcome (p 0.03). Patients who died within 30 days after operation received more units of blood than did patients who survived operation 30 days after the procedure ( and units, respectively; p 0.04). Table 2 summarizes the postoperative course and variables associated with in-hospital mortality. Morbidity and Mortality During Follow-Up None of the survivors were lost to follow-up. The mean duration of follow-up was years (range, years), with a total of 97 patient-years. After discharge from hospital, a further 3 patients died, leading to an overall mortality of 22.6% (n 7/31) (Fig 2). The estimated 5-year survival was 75.3%. Causes for death after primary discharge from the hospital were severe metabolic disorder resulting from renal failure (n 1/3), respiratory failure due to severe chronic heart failure (ejection fraction 25%, n 1/3), and pneumonia resulting from immobilization resulting from preoperative stroke (n 1/3). The significant clinical variables found in univariate analyses for predicting death are summarized in Table 4. Multivariate analyses identified preexisting dependence on dialysis, type 2 diabetes mellitus, staphylococcal infection, and postoperative renal failure and respiratory failure as significant variables for predicting high overall mortality (Table 5). The overall prevalence of freedom from the need for reoperation was 93.5% (Fig 3). No patient treated with the Bentall procedure needed reoperation; the prevalence of reoperation was significantly higher in patients who underwent the Cabrol procedure (15.4% and 0.0%, respectively; p Fig 2. Kaplan-Meier survival curve for all patients.

6 1972 WILBRING ET AL Ann Thorac Surg COMPLEX PROSTHETIC VALVE ENDOCARDITIS 2012;94: Table 4. Significant Variables for Overall Mortality in Univariate Analyses Variable n Overall Mortality (%) p Value All patients Age 67 years a Preexisting type 2 diabetes mellitus Preexisting dialysis-dependent renal impairment Staphylococcal infection Double-valve endocarditis Procedure time b 270 min Deaths: min 0.02 Survivors: min Postoperative renal failure (needing CVVH) Postoperative respiratory failure (needing noninvasive ventilation or reintubation) Administration of 7 blood units c a Determined by receiver operating characteristic (ROC) analyses; area under the curve (AUC) c Determined by ROC analyses; AUC CVVH continuous venovenous hemofiltration. b Determined by ROC analyses; AUC 0.048). One patient experienced valve deterioration of the aortic bioprosthesis with a consecutive second graded regurgitation and needed additional repair of the mitral valve and CABG 3 years after operation. A second patient had a covered rupture of the proximal anastomosis 34 weeks after operation and underwent reoperation by implantation of a Medtronic Freestyle bioprosthesis (Medtronic, Minneapolis, MN). None of the patients had recurrent infection. Survival did not differ significantly between patients undergoing the 2 procedures. Freedom from angina was 100% in those who underwent the Bentall procedure and 92.3% in those who had the Cabrol procedure. The onset of angina in the patient needing reoperation (see earlier) was related to peripheral coronary stenosis; the Cabrol graft was intact. Another patient underwent coronary angiography because of nonspecific thoracic pain. Coronary angiography revealed an intact anastomosis and perfusion of the coronary arteries 2 years after completion of the Cabrol procedure (Fig 4). Comment PVE is associated with high mortality and morbidity, which is in accordance with our studies [2, 6, 14]. The treatment options for PVE are controversial, but several studies have clearly demonstrated that medical treatment alone is not sufficient in most cases [6, 13]. In most patients, combination with surgical treatment is not feasible. Despite developments over the past decades, the mortality reported for surgical treatment of PVE is 20% to 80% [2 6]. The present study focused on a high-risk subgroup within PVE: patients having destruction of the aortic root as well. Information concerning this particular subgroup in the literature is scant. Our observed 30-day mortality of 12.9% and overall mortality of 22.6% ranks at the lower end of reported values for mortality. Despite the complex pathologic features of PVE with aortic involvement, our observed mortality values were not higher than those reported for normal PVE [6]. The estimated 5-year survival of 75.3% was superior to the value of 58% 9% reported by Jassar and colleagues [7] for aortic composite graft replacement in cases of complex active endocarditis. Comparable data for PVE in the literature are not available. Table 5. Significant Variables for Overall Mortality in Multivariate Analyses Variable Odds Ratio p Value Preexisting type 2 diabetes mellitus Preexisting dialysis-dependent renal impairment Postoperative respiratory failure (needing noninvasive ventilation or reintubation) Postoperative renal failure (needing CVVH) Staphylococcal infection CVVH continuous venovenous hemofiltration. Fig 3. Prevalence of freedom from reoperation for all patients.

7 Ann Thorac Surg WILBRING ET AL 2012;94: COMPLEX PROSTHETIC VALVE ENDOCARDITIS 1973 Fig 4. Postoperative coronary angiography 2 years after the Cabrol procedure. (A) Regular opening of the mechanical prosthesis. (B) Regular aspect of the distal anastomosis (arrow). (C) Regular perfusion of the left coronary artery. (D) Regular perfusion of the right coronary artery. Several research teams have advocated the use of homografts in PVE [14, 15]. They state a theoretical advantage in resisting reinfection through a lack of artificial materials [7, 14]. Because of restricted availability at our institution, we did not use homografts. Nonetheless, the prevalence of the freedom from reinfection was 100% in the present series. This was an excellent result when one considers the observed complexity of PVE with destruction of the aortic root and also in comparison with a prevalence of recurrence of 6% to 20% reported in the literature [6, 16]. Jassar and colleagues could not identify a significant difference between nonhomograft biological and mechanical conduits and homografts with respect to reinfection rates [7]. Aortic root replacement allows radical debridement of infected tissue, which might have been for the cause of a low prevalence of reinfection. These results support our radical debridement first policy for active endocarditis. The observed prevalence of freedom from reoperation in the present study was 93.5%, which was comparable to those in other series. All reoperations were in patients who underwent the Cabrol procedure. One was necessary because of early valve deterioration of the aortic bioprosthesis with concomitant peripheral coronary artery disease and mitral regurgitation, and the other reoperation was needed because of a covered rupture of the proximal anastomosis of the aortic conduit. Whether those reasons for reoperation were due to the Cabrol procedure is not known. Nonetheless, none of the patients who underwent the Bentall procedure needed reoperation. This supports the results observed by Gelsomino and colleagues [17]. They described a nonnegligible incidence of early and long-term complications after the Cabrol procedure in the treatment of ascending aortic aneurysms and dissections. However the reasons for reoperation in our series were not associated with the Cabrol procedure itself. Additionally, a bias must be assumed because of the more complex pathoanatomic features leading to the indications for the Cabrol procedure. The Cabrol procedure provides good results in patients with a more complex pathoanatomic process if the Bentall procedure is not feasible. Likewise, Gelsomino and colleagues advocated that aortic root replacement using the Cabrol procedure should be used only if the button procedure is not feasible. This concept is in accordance with our policy. The postoperative course of the patients reflected the complex disease pattern. We documented long ventilation times, long stays in the ICU, and a high prevalence of postoperative complications (Table 2). Univariate analyses identified some variables that could be used to predict mortality (Table 4), but multivariate analyses ruled out several of these variables (Table 5). The remaining variables (preexisting diabetes mellitus, dependence on dialysis, postoperative renal failure, and respiratory failure) are associated with high surgical risks rather than being specific risk factors in cases of endocarditis. The single endocarditis-associated factor was staphylococcal infection. In accordance with the results of other research teams, we observed a high prevalence of staphylococcal infections, which were also predictive for death [1, 6, 18]. The higher prevalence of abscess formation in cases of staphylococcal infections might explain these observations [6]. Patients who died in hospital usually did so within the first few days after operation. This shows that postoperative septic multiple organ failure (which was the most common cause of death) is associated with a very poor prognosis. Encouraging observations were that a long

8 1974 WILBRING ET AL Ann Thorac Surg COMPLEX PROSTHETIC VALVE ENDOCARDITIS 2012;94: stay in the ICU and long ventilation times were not predictive for death. These findings indicate that intensive care medicine could successfully manage most complications of this septic disease pattern. Study Limitations The small number of patients and its retrospective design mainly limited the present study. Nonetheless, PVE necessitating aortic root replacement forms a particular subgroup within a rare disease pattern, so recruiting a large group of patients is inherently difficult. Prospective (or even randomized) studies of patients with PVE carry the risk of ethical problems and are therefore very unlikely to be carried out. Conclusions Composite aortic root replacement in cases of complex PVE is associated with quite high mortality and morbidity, and the postoperative course is also complex. Nevertheless, the clinical and long-term results seem not to be inferior to those reported for noncomplex PVE. Aortic root replacement for complex PVE was technically feasible in each patient and provided excellent freedom from reinfection or valve-related events. The Cabrol procedure undertaken if the Bentall button technique was not feasible produced excellent results for the treatment of destruction of the aortic root in cases of complex PVE. References 1. Alexiou C, Langley SM, Stafford H, Lowes JA, Livesey SA, Monro JL. Surgery for active culture-positive endocarditis: determinants of early and late outcome. Ann Thorac Surg 2000;69: Vongpatanasin W, Hillis LD, Lange RA. Prosthetic heart valves. N Engl J Med 1996;335: Piper C, Körfer R, Horstkotte D. Prosthetic valve endocarditis. Heart 2001;85: Stanbridge TN, Isalska BJ. Aspects of prosthetic valve endocarditis. J Infect 1997;35: Akowuah EF, Davies W, Oliver S, et al. Prosthetic valve endocarditis: early and late outcome following medical or surgical treatment. Heart 2003;89: Tugtekin S, Matschke K, Daubner D, et al. Prosthetic valve endocarditis: importance of surgical treatment. Thorac Cardiovasc Surg 2007;55: Jassar AS, Bavaria JE, Szeto WY, et al. Graft selection for aortic root replacement in complex active endocarditis: does it matter? Ann Thorac Surg 2012;93: Nishimura RA, Carabello BA, Faxon DP, et al. ACC/AHA 2008 guideline update on valvular heart disease: focused update on infective endocarditis: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines endorsed by the Society of Cardiovascular Anesthesiologists, Society for Cardiovascular Angiography and Interventions, and Society of Thoracic Surgeons. J Am Coll Cardiol 2008;52: Habib G, Hoen B, Tornos P, et al; ESC Committee for Practice Guidelines. Guidelines on the prevention, diagnosis, and treatment of infective endocarditis (new version 2009): the Task Force on the Prevention, Diagnosis, and Treatment of Infective Endocarditis of the European Society of Cardiology (ESC). Endorsed by the European Society of Clinical Microbiology and Infectious Diseases (ESCMID) and the International Society of Chemotherapy (ISC) for Infection and Cancer. Eur Heart J 2009;30: Bentall M, DeBono A. A technique for complete replacement of ascending aorta. Thorax 1968;23: Kouchoukos NT, Wareing TH, Murphy SF, Perrillo JB. Sixteen-year experience with aortic root replacement. Results of 172 operations. Ann Surg 1991;214: Cabrol C, Gandjbakhc I, Pavie A. Surgical treatment of ascending aortic pathology. J Card Surg 1988;3: Mahesh B, Angelini G, Caputo M, Jin XY, Bryan A. Prosthetic valve endocarditis. Ann Thorac Surg 2005;80: Sabik JF, Lytle BW, Blackstone EH, Marullo AG, Pettersson GB, Cosgrove DM. Aortic root replacement with cryopreserved allograft for prosthetic valve endocarditis. Ann Thorac Surg 2002;74: Musci M, Hübler M, Amiri A, et al. Surgical treatment for active infective prosthetic valve endocarditis: 22-year single-centre experience. Eur J Cardiothorac Surg 2010;38: Renzulli A, Carozza A, Romano G, et al. Recurrent infective endocarditis: a multivariate analysis of 21 years of experience. Ann Thorac Surg 2001;72: Gelsomino S, Frassani R, Da Col P, et al. A long-term experience with the Cabrol root replacement technique for the management of ascending aortic aneurysms and dissections. Ann Thorac Surg 2003;75: Grünenfelder J, Akins CW, Hilgenberg AD, et al. Long-term results and determinants of mortality after surgery for native and prosthetic valve endocarditis. J Heart Valve Dis 2001;10: