Midterm Surgical Outcomes of Noncomplicated Active Native Multivalve Endocarditis: Single-Center Experience

Transcription

1 ADULT CARDIAC Midterm Surgical Outcomes of Noncomplicated Active Native Multivalve : Single-Center Experience Takeyoshi Ota, MD, PhD, Thomas G. Gleason, MD, Stefano Salizzoni, MD, Lawrence M. Wei, MD, Yoshiya Toyoda, MD, PhD, and Christian Bermudez, MD Department of Cardiothoracic Surgery, Division of Cardiac Surgery, University of Pittsburgh, Pittsburgh, Pennsylvania Background. Surgical treatment for endocarditis is still challenging and only a handful of studies have analyzed surgical outcomes in patients with active multivalvular endocarditis. Methods. From June 1996 to October 2007, 152 patients underwent surgery for active native valve endocarditis; 117 patients with single-valve endocarditis and 35 patients with multivalvular endocarditis. Preoperative and postoperative data were retrospectively reviewed to determine risk factors for early and late mortality. Results. One-year and 5-year survival were 78.6% 3.8% and 54.8% 5.3% in the single-valve group, and 74.3% 7.4% and 64.8% 8.3% in the multivalve group, respectively (log-rank 0.64). The rates of valve replacement were significantly higher in the single-valve group than in the multivalve group in all sites except the pulmonary valve. Freedom from reoperation was 90.0% 3.6% in the single-valve group and 79.5% 8.5% in the multivalve group in 5 years (log-rank 0.30). No recurrence of endocarditis was noted in the multivalve group, while 4 patients (3.4%) had recurrence in the single-valve group during the follow-up period (p 0.57). There was no significant difference in postoperative complications between the two groups. Multivariate analysis of the multivalve group identified postoperative renal failure as a predictor of late mortality with no predictors identified for early mortality, reoperation, and recurrence. Statistical analysis of the overall cohorts showed multivalve endocarditis was not an independent predictor of early and late mortality. Conclusions. Surgical treatment for active, native, and noncomplicated multivalve endocarditis was associated with respectable early and late morbidity and mortality comparable with single-valve endocarditis, and was not an independent predictor of early and late mortality. (Ann Thorac Surg 2011;91:1414 9) 2011 by The Society of Thoracic Surgeons Treatment for active infective endocarditis is still challenging with high mortality and morbidity [1] and the incidence of infective endocarditis has not decreased in the last three decades despite the evolution of robust preventive medicine programs [1, 2]. Infectious endocarditis is associated with 10% to 20% in-hospital mortality [3 5], and 20% to 50% of patients eventually require surgical intervention [6, 7]. Although less frequent than single-valve endocarditis, multivalvular endocarditis is not an uncommon or clinically insignificant phenomenon. However, only a handful of studies have evaluated surgical outcomes in patients with multivalve endocarditis [8 11]. The purpose of this study was to summarize our experience with the surgical treatment of multivalve endocarditis and to determine the incidence, pathology, predictors of mortality, and short-term and long-term surgical outcomes within a comparison between singlevalve and multivalve endocarditis. Accepted for publication Dec 7, Address correspondence to Dr Bermudez, Department of Cardiothoracic Surgery, University of Pittsburgh, 200 Lothrop St, PUH C-900, Pittsburgh, PA 15213; bermudezc@upmc.edu. Patients and Methods This study was approved by the quality improvement review committee in the University of Pittsburgh and the need of patient consent was waived. From June 1996 to October 2007, 152 patients at the University of Pittsburgh Medical Center undergoing surgery for active native valve endocarditis were identified and data retrospectively collected from our cardiac surgical database and review of medical records. The mean follow-up period was years. Prosthetic valve endocarditis was excluded from this study due to the different nature of the disease and clinical characteristics. Follow-up was completed by phone interview and late outcome was cross-referenced using the Social Security Number Death Index. Multivalve endocarditis was present in 23% of patients (35 of 152) (multivalve group). Of these, 31 patients (20.4% of total cohort) had double valve endocarditis and 4 patients (2.6%) had triple valve endocarditis. was defined based on the Duke criteria [12]. was diagnosed as active when more than one of the following was confirmed: (1) septic shock; (2) positive blood culture; (3) operative findings including vegetation, abscess, and active tissue inflammation; (4) positive culture from excised tissues; and (5) positive Gram stain by The Society of Thoracic Surgeons /$36.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg OTA ET AL 2011;91: SURGICAL OUTCOMES OF MULTIVALVE ENDOCARDITIS Characteristics of both groups are detailed in Table 1. Overall, 70% of the patients (106 of 152 patients) were classified as New York Heart Association class III or IV. The primary surgical indications included the presence of mobile vegetation of more than 10 mm in 103 patients (67.8%), heart failure in 50 patients (32.9%), systemic emboli in 28 patients (18.4%), and persistent sepsis in 5 patients (3.3%). Thirty-three percent (51 of 152) of surgery was performed within 72 hours of diagnosis of active endocarditis. There were no statistically significant differences in preoperative characteristics between the multivalve group and the single-valve group. Responsible organisms are detailed in Table 2. Methicillin-sensitive Staphylococcus aureus was the most common organism in the single-valve group (35.9%). The methicillin-sensitive Staphylococcus aureus and Streptococcus viridans were the most common organisms seen in the multivalve group (22.9%). The rate of Streptococcus viridans infection was significantly higher in the multivalve group as compared with the single-valve group (p 0.027), while the other infecting organisms did not differ significantly. No organism was identified in 1 patient (2.9%) in the multivalve group and 5 patients (4.2%) in the single-valve group. Statistical Analysis Continuous variables were reported as the mean standard deviation and categoric variables were reported Table 2. Offending Organism Organism Multivalve (n 35) Single-Valve (n 117) MSSA 8 (22.9%) 42 (35.9%) 0.21 MRSA 7 (20.0%) 14 (12.0%) 0.26 Other Staphylococcus 4 (11.4%) 11 (9.4%) 0.74 species Streptococcus viridans 8 (22.9%) 9 (7.7%) a Streptococcus agalactiae 1 (2.9%) 5 (4.2%) 1.00 Other streptococcus species 2 (5.7%) 7 (6.0%) 1.00 Pseudomonas aeruginosa 0 (0%) 2 (1.7%) 1.00 Enterococcus faecalis 3 (8.6%) 10 (8.5%) 1.00 Haemophilus influenzae 0 (0%) 4 (3.4%) 0.57 Fungi 1 (2.9%) 3 (2.6%) 0.57 Others 0 (0%) 5 (4.2%) 0.58 Culture negative 1 (2.9%) 5 (4.2%) 1.00 a p MSSA methi- MRSA methicillin-resistant Staphylococcus aureus; cillin-sensitive Staphylococcus aureus as frequencies. The Fisher exact or 2 test was used to evaluate categoric variables for univariate analysis. The Student t test or the Mann-Whitney U test was used for continuous variables based on distribution of the data. p ADULT CARDIAC Table 1. Preoperative Patient Characteristics Characteristics Multivalve (n 35) Single-Valve (n 117) p Age (mean SD) year Male:Female 27:8 80: COPD 4 (11.4%) 22 (18.8%) 0.44 DM 12 (34.2%) 35 (29.9%) 0.68 Hyperlipidemia 4 (11.4%) 20 (17.1%) 0.59 Hypertension 15 (42.9%) 57 (48.7%) 0.56 Renal insufficiency (Cr 2.0 mg/dl) 16 (45.7%) 32 (27.4%) 0.06 Dialysis 11 (31.4%) 23 (19.7%) 0.17 CVA 10 (28.6%) 28 (23.9%) 0.66 PVD 4 (11.4%) 20 (17.2%) 0.60 IVDU 8 (22.8%) 26 (22.2%) 1.00 Previous CABG 1 (2.9%) 10 (8.5%) 0.46 NYHA II 7 (20.0%) 22 (18.8%) 1.00 III 10 (28.6%) 35 (29.9%) 1.00 IV 16 (45.7%) 45 (38.5%) 0.55 LVEF (5.7%) 2 (1.7%) 0.11 Operation within 72 hours after diagnosis 13 (37.1%) 38 (32.4%) 0.68 Surgical indications Congestive heart failure 12 (34.2%) 38 (32.5%) 0.84 Septic shock 2 (5.7%) 3 (2.6%) 0.33 Vegetation size 10 mm 20 (57.1%) 83 (70.9%) 0.15 Emboli 6 (17.1%) 22 (18.8%) 1.00 CABG coronary artery bypass grafting; COPD chronic obstructive pulmonary disease; Cr serum creatinine; CVA cerebrovascular accident; DM diabetes mellitus; IVDU intravenous drug user; LVEF left ventricular ejection fraction; NYHA New York Heart Association Classification; PVD peripheral vascular disease; SD standard deviation.

3 ADULT CARDIAC 1416 OTA ET AL Ann Thorac Surg SURGICAL OUTCOMES OF MULTIVALVE ENDOCARDITIS 2011;91: The Kaplan-Meier model with the log-rank test was used to calculate estimates for long-term survival, freedom from reoperation, and freedom from recurrent endocarditis. The Cox regression model or the logistic regression model was used to determine independent predictors of early (30-day and hospital mortality) and late mortality, reoperation, and recurrence. Only variables with a p value less than 0.05 from univariate analyses were retained for the multivariate regression tests. For all analyses, a p value less than 0.05 in a two-sided test were considered statistically significant. All statistical analyses were performed with SPSS (SPSS Inc, Chicago, IL). Results Statistical Analyses for the Overall Cohort Thirty-day mortality rate was 7.9% (12 patients). There were 13 in-hospital deaths (8.6%) defined as any death in the first admission where surgery was performed. Overall survival was 91.4% 2.4% at 1 year and 56.4% 5.3% at 5 years. Freedom from reoperation and recurrence at 5 years were 88.9% 3.3% and , respectively. In the multivariate analysis, postoperative dialysis was a significant independent predictor associated with the in-hospital mortality (p 0.006, odds ratio [OR] 19.25, confidence interval [CI] 2.31 to 160.0). The Cox regression analysis identified the presence of chronic obstructive pulmonary disease (p 0.001, OR 3.84, CI 1.98 to 7.46) and the history of previous coronary artery bypass grafting (p 0.023, OR 2.93, CI 1.16 to 7.36) as independent predictors of late mortality. No predictor was identified for reoperation and recurrence. Multivalve endocarditis was not an independent predictor of early and late mortality, and reoperation. Comparison of Multivalve and Single-Valve Surgical procedures are summarized in Table 3. Twentyfour out of 29 infected aortic valves (82.7%) in the multivalve group were replaced with prosthetic valves and all infected aortic valves (n 37, 100%) in the single-valve group were replaced. Sixty-four percent (20 of 31) of mitral valves and 79% (14 of 16) of tricuspid valves in the multivalve group were repaired, while only 33% (19 of 58) of mitral valves and 33% (7 of 21) of tricuspid valves in the single-valve group were repaired. At all infected sites except the pulmonary valve, the rates of valve replacement were significantly higher in the single-valve group than in the multivalve group (aortic valve: 100% singlevalve group versus 82.7% multivalve group, p ; mitral valve: 67.2% single-valve group versus 35.5% multivalve group, p ; tricuspid valve: 66.7% singlevalve group versus 21.4% multivalve group, p 0.015). With regard to the details of mitral repair, prosthetic mitral rings were used in 7 cases (7 of 20, 35%) in the multivalve group and 7 cases (7 of 19, 36.8%) in the single-valve group (p 1.00), leaflet plasty alone using autologous or bovine pericardium was performed in 11 patients (55%) in the multivalve group and 9 patients Table 3. Infected Sites and Surgical Procedures Multivalve (n 35) Single Valve (n 117) Infected site AV MV 21 (60.0%) AV 37 (31.6%) MV TV 6 (17.1%) MV 58 (49.6%) AV TV 4 (11.4%) TV 21 (17.9%) AV MV TV 4 (11.4%) PV 1 (0.8%) Surgical procedure AVR MVR 7 (20.0%) AVR 37 (31.6%) AVR MVP 12 (34.3%) AVP 0 (0%) AVR TVR 2 (5.7%) MVR 39 (33.3%) AVR TVP 1 (2.9%) MVP 19 (16.2%) AVP MVR 2 (5.7%) TVR 14 (12.0%) AVP TVP 1 (2.9%) TVP 7 (6.0%) MVR TVR 1 (2.9%) PVR 1 (0.8%) MVP TVP 5 (14.3%) AVR MVR TVP 1 (2.9%) AVR MVP TVP 2 (5.7%) AVP MVP TVP 1 (2.9%) AV aortic valve; AVP aortic valve plasty; AVR aortic valve replacement; MV mitral valve; MVP mitral valve plasty; MVR mitral valve replacement; PV pulmonary valve; PVR pulmonary valve replacement; TV tricuspid valve; TVP tricuspid valve plasty; TVR tricuspid valve replacement. (47.4%) in the single-valve group (p 0.75), and debridement alone was performed in 2 patients (10%) in the multivalve group and in 3 patients (15.8%) in the singlevalve group (p 0.66). Ten cases (28.6%) in the multivalve group and 18 cases (15.4%) in the single-valve group required debridement and closure of abscess cavities using autologous or bovine pericardium (p 0.09). Prosthetic mitral rings were used as per surgeons preference if felt that the complexity of the repair and some degree of annular dilation that could compromise the repair were present. Therefore, no mitral ring was used exclusively for treatment of secondary mitral regurgitation. Operative data and postoperative complications are detailed in Table 4. Concomitant coronary artery bypass grafting was performed in 8 patients (22.9%) in the multivalve group and 28 patients (23.9%) in the singlevalve group. As expected, cardiopulmonary bypass time and aortic cross-clamp time were significantly longer in the multivalve group than in the single-valve group. The mean length of postoperative hospital stay was days in the multivalve group and days in the single-valve group with no significant difference. There was no significant difference in postoperative complications, which included surgical bleeding, persistent bacteremia, deep sternal infection, pneumonia, prolonged intubation, transient ischemic attack, stroke, pulmonary embolism, renal insufficiency, heart block, gastrointestinal bleeding, 30-day mortality, in-hospital mortality, and recurrence, between the two groups. There were a total of 13 deaths in the multivalve group and 51 deaths in the

4 Ann Thorac Surg OTA ET AL 2011;91: SURGICAL OUTCOMES OF MULTIVALVE ENDOCARDITIS Table 4. Operative Data, Postoperative Complications, and Early-Late Mortality Multivalve (n 35) Single-Valve (n 117) p Cardiopulmonary bypass time (mean a SD; minutes) Aortic cross-clamp time (mean SD; a minutes) Length of postop stay (days) Concomitant CABG 8 (22.9%) 28 (23.9%) 1.00 Abscess closure 10 (28.6%) 18 (15.4%) 0.09 Takeback for bleeding 4 (11.4%) 7 (6.0%) 1.00 Persistent bacteremia 4 (9.3%) 16 (13.7%) 1.00 Deep sternal infection 1 (2.9%) 2 (1.7%) 0.55 Pneumonia 5 (14.3%) 10 (8.5%) 0.34 Prolonged intubation ( 72 hours) 13 (37.1%) 38 (32.5%) 0.68 TIA 0 (0%) 1 (0.9%) 1.00 Stroke 0 (0%) 3 (2.6%) 1.00 Pulmonary embolism 0 (0%) 1 (0.9%) 1.00 Renal failure required dialysis 5 (14.3%) 9 (7.7%) 0.31 Heart block 3 (8.6%) 14 (12.0%) 0.76 Permanent pacemaker implant 2 (5.7%) 12 (10.3%) 0.52 GI bleeding 3 (8.6%) 5 (4.3%) day mortality 4 (11.4%) 8 (6.8%) 0.47 In-hospital mortality 5 (14.3%) 8 (6.8%) 0.18 Recurrence 0 (0%) 4 (3.4%) 0.57 All death 13 (37.1%) 51 (43.5%) 0.56 Primary cause of death Cardiac 2 (5.7%) 8 (6.8%) 1.00 Infection 2 (5.7%) 6 (5.1%) 1.00 Neurologic 0 (0%) 4 (3.4%) 0.57 Multiorgan failure 4 (11.4%) 9 (7.7%) 0.49 Cancer 0 (0%) 3 (2.6%) 1.00 Respiratory 1 (2.9%) 6 (5.1%) 1.00 Other/Unknown 4 (11.4%) 15 (12.8%) ADULT CARDIAC a p CABG coronary artery bypass grafting; GI gastrointestinal; SD standard deviation; TIA transient ischemic attack. single-valve group during the follow-up period. The primary causes of death are detailed in Table 4. No significant difference in the primary causes of death between the two groups was noted. One-year survival was % in the single-valve group and % in the multivalve group, and 5-year survival was % in the single-valve group and % in the multivalve group, with no significant differences between the groups (log-rank 0.64) (Fig 1A). Five years after surgery, freedom from reoperation was 90.0% 3.6% in the single-valve group and 79.5% 8.5% in the multivalve group (log-rank 0.30) (Fig 1B). Freedom from recurrence was 100% in the multivalve group during the follow-up period and 93.0% 3.9% in the single-valve group in 5 years (log-rank 0.21). Statistical Analyses for Multivalve To assess risk factors and predictors of mortality, a series of statistic analyses was performed using only the multivalve group. In univariate analysis, statistically significant factors associated with late mortality were the presence of preoperative renal insufficiency (serum creatinine 2 mg/dl) (p 0.043), peripheral vascular disease (p 0.014), postoperative persistent bacteremia (p 0.014), and postoperative renal failure requiring dialysis (p 0.016). Cox regression analysis identified that postoperative renal failure (p 0.033, OR 10.25, CI 1.2 to 87.4) was an independent predictor of late mortality. No predictor for in-hospital mortality was identified by multivariate analysis; however, persistent bacteremia (p 0.006) and postoperative renal failure (p 0.006) were associated with in-hospital mortality by univariate analysis. Univariate analysis demonstrated no significant predictor for 30-day mortality and reoperation. There was no statistically significant relationship between the responsible organisms and late mortality. Comment In this study we reviewed our experience of surgical treatment of noncomplicated native multivalve endocar-

5 ADULT CARDIAC 1418 OTA ET AL Ann Thorac Surg SURGICAL OUTCOMES OF MULTIVALVE ENDOCARDITIS 2011;91: Fig 1. Kaplan-Meier curves. (A) Survival curve (multivalve versus single valve) (log-rank 0.64). (B) Freedom from reoperation (logrank 0.30). ( multivalve; --- single valve.) ditis, comparing with those surgically treated for singlevalve endocarditis. We found no significant differences in postoperative complications, early and late outcomes, freedom from reoperation, and recurrence of infection between single-valve and multivalve endocarditis. The early and late mortality of surgically treated multivalvular endocarditis were respectable and comparable with those with single-valve endocarditis. While the presence of multifocal intracardiac infection may, by logic, predict an increased risk of regional recurrence on new prosthetic materials (ie, multiple valve replacement-repair), we had no recurrence in the patients with multivalve endocarditis during the follow-up period and demonstrated no statistically significant difference in recurrence of infection compared with the single-valve endocarditis. In patients with multivalvular endocarditis, postoperative renal failure was an independent predictor for late mortality. In the analyses with the overall cohort, multivalve endocarditis was not a predictor of 30-day and in-hospital mortality, reoperation, and late mortality. Previous reports are somewhat discrepant on the issue of whether multivalvular disease imparts a poorer prognosis [8 10, 13]. The incidence of multivalvular endocarditis varies in individual series from 12.4% to 26.3% [11, 14]. Kim and colleagues [15] reported multivalvular endocarditis in 18% of nonsurgical endocarditis cases. The incidence of multivalvular endocarditis in our series was 23.0%, which is similar to the incidence seen in other series. Triplevalve or quadruple-valve endocarditis occurs infrequently [16, 17]. There were only 4 cases of triple-valve endocarditis in this series. Overall, the responsible infecting organisms did not differ significantly between single-valve and multivalve patients, with Staphylococcus species being most common, except that Streptococcus viridans infections occurred more frequently in cases of multivalve endocarditis as compared with single-valve endocarditis. This finding is consistent with previous reports [9, 11]. In the present study, 30-day mortality was 6.8% in the single-valve group and 11.4% in the multivalve group, in-hospital mortality was 14.3% in the single-valve group and 6.8% in the multivalve group, and 5-year survival was 54.8% 5.3% in the single-valve group and 64.8% 8.3% in the multivalve group with no statistically significant difference. While 30-day mortality of multivalvular endocarditis in our series is similar to the other series (range is from 12.5% to 16.0% [8, 9, 11]), the results of long-term survival are different from some reports by other investigators. Yao and colleagues [11] reported that the overall survivals at 5 and 10 years were 74% and 62%, and Mihaljevic and colleagues [9] reported 87% and 64% at 5 and 10 years, respectively. In the latter study of only 63 patients, however, 60% had a healed endocarditis at the time of surgery with no active infection. The study by Yao and colleagues had only 48 patients over 20 years, and 33% of the patients included in the study had healed endocarditis. This may be one of the factors that explain the discrepancy between the late outcomes of this study and the other studies. In our current series, all patients had active endocarditis with surgical intervention. Alternatively, Sheikh and colleagues [8] reported overall survival at 5, 7, and 10 years as 68%, 59%, and 49%, respectively, for cases of only active infective endocarditis, which is comparable with our results. Interestingly, we found that the rates of valve replacement were significantly higher in the single-valve group than in the multivalve group in all valves except the pulmonary valve. This may be reflective of various degrees of valve involvement at multiple foci of multivalve endocarditis. It is reasonable to assume that multivalvular endocarditis, at its onset, begins as single-valve disease with subsequent seeding of the second valve by the bloodborne pathogen. It is likely that the secondary involved valve is less affected, resulting in high probability of repair. In our series the most common combination in multivalvular endocarditis involved the aortic and mitral valves (21 patients, 60.0% of multivalve endocarditis). Twelve of the 21 patients (57.1 %) underwent aortic valve replacement and mitral valve repair; only 2 of the 21 aortic valves (9.5%) were repairable while 12 of the

6 Ann Thorac Surg OTA ET AL 2011;91: SURGICAL OUTCOMES OF MULTIVALVE ENDOCARDITIS mitral valves (57.1%) were repairable. In the multivalve group, 25 out of 29 aortic valves (86%) were replaced with prosthetic valves while all affected aortic valves (37 of 37) in the single-valve group were replaced (p 0.03). The high repair rates of the affected mitral and tricuspid valves in the multivalve group (69% and 79%, respectively) suggest that secondary involvement of the valve structure was less severe than the primary lesion, which, in turn, could contribute to the absence of significant difference in mortality and morbidity between the studied groups. The limitations of this study include the retrospective nature and analysis including a relatively small number of patients. However, this is one of the largest series of active native multivalvular endocarditis. Although this is a single-center study multiple surgeons used different techniques, which could impact both the short-term and long-term outcomes. The results in this study may be controversial and require further analysis based on the fact that multiple foci of infection and complicated surgical procedures theoretically could lead to higher morbidity and mortality than single focus of infection and simpler surgery, as opposed to our results. Possible explanations of this paradox include the following. (1) The exclusion from this study of prosthetic valve endocarditis that had a high probability of requiring a complex operation (eg, aortic root replacement, use of homografts) with increased perioperative morbidity and mortality resulted in a high proportion of noncomplicated cases in our series. (2) In our series, Streptococcus viridans was significantly more frequent in the multivalve group than in the single-valve group. Streptococcus viridans is generally accepted as less virulent and susceptible to pharmacologic treatment compared with Staphylococcus species [18], and this could have contributed to narrowing the gap in mortality between the multivalve and single-valve endocarditis. (3) Generally, mitral valve repair is associated with a better outcome than mitral valve replacement. The rates of mitral and tricuspid valve repair in the multivalve group were significantly higher than in the single-valve group, which could also have been a contributory factor to narrow the gap. (4) Our institution is fairly aggressive on early surgical intervention once the patients meet the criteria. In fact, 33% of the surgical cases were performed within 72 hours after diagnosis. We believe that the early surgical intervention resulted in the high rate of valve repair in the secondary involved valve, as described above, in turn leading to better outcomes in the multivalve group. (5) Similarly, based on the high rate of valve repair and counterintuitive conclusions on the multivalve endocarditis in this study, it should be important for the readers to take into account that our multivalve endocarditis cases might not have as extensive disease as those reported elsewhere. In summary, we presented our experience of acute native endocarditis, focusing on multivalve endocarditis. Surgical treatment of active, native, and noncomplicated multivalve endocarditis was associated with respectable short-term and long-term mortality and morbidity in our series. Multivalve endocarditis was not an independent predictor of early and late mortality, and reoperation. When comparing multivalve to single-valve endocarditis, there was no significant difference in early and late survivals, freedom from reoperation, and postoperative morbidity. Postoperative renal failure was an independent predictor of late mortality for multivalve endocarditis. While we believe that the present study showed informative findings, it is also important to note that our results are still controversial and further studies are necessary. References Mylonakis E, Calderwood SB. Infective endocarditis in adults. N Engl J Med 2001;345: Tleyjeh IM, Steckelberg JM, Murad HS, et al. Temporal trends in infective endocarditis: a population-based study in Olmsted County, Minnesota. JAMA 2005;293: Thuny F, Di Salvo G, Belliard O, et al. Risk of embolism and death in infective endocarditis: prognostic value of echocardiography: a prospective multicenter study. Circulation 2005;112: Netzer RO, Altwegg SC, Zollinger E, Täuber M, Carrel T, Seiler C. Infective endocarditis: determinants of long term outcome. Heart 2002;88: Netzer RO, Zollinger E, Seiler C, Cerny A. Infective endocarditis: clinical spectrum, presentation and outcome. An analysis of 212 cases Heart 2000;84: Vlessis AA, Hovaguimian H, Jaggers J, Ahmad A, Starr A. Infective endocarditis: ten-year review of medical and surgical therapy. Ann Thorac Surg 1996;61: Zamorano J, de Isla LP, Malangatana G, et al. Infective endocarditis: mid-term prognosis in patients with good in-hospital outcome. J Heart Valve Dis 2005;14: Sheikh AM, Elhenawy AM, Maganti M, Armstrong S, David TE, Feindel CM. Outcomes of double valve surgery for active infective endocarditis. J Thorac Cardiovasc Surg 2009;138: Mihaljevic T, Byrne JG, Cohn LH, Aranki SF. Long-term results of multivalve surgery for infective multivalve endocarditis. Eur J Cardiothorac Surg 2001;20: Gillinov AM, Diaz R, Blackstone EH, et al. Double valve endocarditis. Ann Thorac Surg 2001;71: Yao F, Han L, Xu ZY, et al. Surgical treatment of multivalvular endocarditis: twenty-one-year single center experience. J Thorac Cardiovasc Surg 2009;137: Durack DT, Lukes AS, Bright DK. New criteria for diagnosis of infective endocarditis: utilization of specific echocardiographic findings. Duke Service. Am J Med 1994;96: Grover FL, Cohen DJ, Oprian C, Henderson WG, Sethi G, Hammermeister KE. Determinants of the occurrence of and survival from prosthetic valve endocarditis. Experience of the Veterans Affairs Cooperative Study on Valvular Heart Disease. J Thorac Cardiovasc Surg 1994;108: David TE, Gavra G, Feindel CM, Regesta T, Armstrong S, Maganti MD. Surgical treatment of active infective endocarditis: a continued challenge. J Thorac Cardiovasc Surg 2007; 133: Kim N, Lazar JM, Cunha BA, Liao W, Minnaganti V. Multivalvular endocarditis. Clin Microbiol Infect 2000;6: Kontogiorgi M, Koukis I, Argiriou M, et al. Triple valve endocarditis as an unusual complication of bacterial meningitis. Hellenic J Cardiol 2008;49: Seeburger J, Groesdonk H, Borger MA, et al. Quadruple valve replacement for acute endocarditis. J Thorac Cardiovasc Surg 2009;137: Murtagh B, Frazier OH, Letsou GV. Diagnosis and management of bacterial endocarditis in Curr Opin Cardiol 2003;18: ADULT CARDIAC