European Journal of Cardio-Thoracic Surgery Advance Access published July 15, 2014
Surgical management of infective endocarditis: an analysis of early and late outcomes† Philip Y.K. Pang*, Yoong Kong Sin, Chong Hee Lim, Teing Ee Tan, See Lim Lim, Victor T.T. Chao and Yeow Leng Chua Department of Cardiothoracic Surgery, National Heart Centre Singapore, Singapore * Corresponding author. National Heart Centre Singapore, 5 Hospital Drive, Singapore 169609. Tel: +65-6704-8896; fax: +65-844-9063; e-mail:[email protected] (P.Y.K. Pang). Received 30 March 2014; received in revised form 19 May 2014; accepted 10 June 2014
Abstract OBJECTIVES: To review our experience of surgical management of infective endocarditis (IE) over a 13-year period and analyse the outcomes and associated prognostic factors. METHODS: A retrospective review was conducted for 191 consecutive patients who underwent surgery for native and prosthetic valve endocarditis (PVE) between January 2000 and December 2012. Surgical outcomes were reviewed to include survival and postoperative complications. Follow-up was complete for 172 of 179 patients (96.1%) surviving to hospital discharge, with a mean follow-up of 6.6 ± 3.7 years. RESULTS: Mean age was 47.4 ± 14.9 years with 113 (63.9%) males. Native valve endocarditis was present in 177 patients (92.7%). Sixtythree patients (33.0%) presented with embolic complications. The brain was the most common site of embolism, involving 25 patients (13.1%). Streptococcus viridans was the most common infective organism, isolated in 68 patients (35.7%), followed by Staphylococcus aureus in 30 patients (15.7%). Eighty-seven patients (45.5%) had active endocarditis at the time of surgery. The mitral valve was infected in 136 patients (71.2%), the aortic valve in 66 (34.6%), the tricuspid valve in 29 (15.2%) and multiple valves in 38 (19.9%). Nineteen patients (9.9%) were intravenous drug users (IVDU). Twelve IVDUs (63.2%) suffered from tricuspid valve IE, compared with 7 of 162 patients (4.3%) in the non-IVDU population (P < 0.001). The most common indication for early surgery was intractable cardiac failure. Twelve patients (6.3%) died during the hospital stay for surgical treatment of IE. Logistic multivariate analysis identified preoperative creatinine clearance and stroke as independent predictors of in-hospital mortality. Overall 10-year survival and freedom from valve-related reoperation were 74.8 and 90.3%, respectively. Age, PVE, S. aureus endocarditis and postoperative left ventricular ejection fraction (LVEF) ≤45% were factors influencing long-term survival. CONCLUSIONS: Surgical management of endocarditis continues to be challenging and is associated with significant morbidity and mortality. This report of 191 patients who underwent valve surgery for IE shows that in-hospital mortality is influenced by preoperative renal function and stroke at the time of presentation. The optimal timing for surgery in patients with stroke remains controversial. Long-term survival was negatively influenced by increasing age, moderate to severely impaired LVEF, prosthetic valve IE and S. aureus infection. Keywords: Endocarditis • Infective • Heart valve prosthesis implantation
INTRODUCTION Despite substantial improvements made in the diagnosis and management of infective endocarditis (IE), it remains a serious condition that is associated with significant morbidity and mortality. Compared with antibiotic treatment alone, surgery for IE has greatly increased survival [1]. Surgery for IE is required in 25–30% of cases during the acute phase and in 20–40% during the convalescent phase [2, 3]. The most common indications for surgery in IE include intractable heart failure, uncontrolled infection related to perivalvular extension and resistant organisms, recurrent embolic events and presence of prosthetic material [4, 5]. †
Poster presentation at Chest 2013, 30 October 2013, Chicago, IL, USA.
This study was conducted to evaluate the impact of perioperative clinical variables on early and late outcomes in patients undergoing surgery for IE at a tertiary referral centre. With an increasing number of high-risk patients undergoing surgery for IE, the identification of perioperative prognostic factors will be useful for risk assessment and surgical decision-making.
MATERIALS AND METHODS Following approval from the local institutional review board (reference: 2011/880/C), a retrospective case-note and database review was performed on consecutive patients who had undergone surgery for IE between January 2000 and December 2012 at our tertiary referral centre.
© The Author 2014. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ADULT CARDIAC
ORIGINAL ARTICLE
European Journal of Cardio-Thoracic Surgery (2014) 1–7 doi:10.1093/ejcts/ezu281
P.Y.K. Pang et al. / European Journal of Cardio-Thoracic Surgery
2
Definitions In-hospital mortality was defined as all-cause mortality during the hospital stay for the surgical treatment of IE. Operations were considered emergent if performed within 24 h of hospital admission. Active IE was defined as ongoing infection in a patient who was still receiving antibiotic therapy. Renal failure was defined as serum creatinine clearance (CrCl) levels lower than 60 ml/min as calculated with the Cockroft–Gault formula, or the need for renal replacement therapy. Respiratory failure was defined as the need for mechanical ventilation for more than 24 h or the need for reintubation. Moderate to severely impaired left ventricular ejection fraction (LVEF) was defined as LVEF of 45% or less.
Patients During the period January 2000 to December 2012, 191 consecutive patients underwent surgery for IE at our institution. The medical records of all 191 patients were accessible and reviewed. This study focused exclusively on patients with native or prosthetic valve IE treated with open heart surgery. Cases of IE related to nonvalvular cardiovascular devices, such as pacemakers and catheters, were excluded. Preoperative clinical variables of interest are shown in Table 1. All 191 patients had a preoperative transthoracic echocardiogram. Echocardiographic parameters of interest included the presence of vegetation, maximum length of vegetation, abscess, infected valve and LVEF. The New York Heart Association (NYHA) functional class was assessed in all patients before surgery. Nineteen patients (9.9%) were in NYHA Class I, 86 (45.0%) in Class II, 36 (18.8%) in Class III and 50 (26.2%) in Class IV. Treatment for IE consisted of an antibiotic regimen dictated by the primary physician, consulting infectious disease specialist, and surgeon, and varied according to species identified and sensitivities. Whenever possible, each patient completed a standard regimen of a 4- to 6-week course of IV antibiotics after diagnosis. Analysis of
long-term outcomes focused on all-cause mortality, valve-related reoperation and recurrent endocarditis. Patients with IE who presented with haemorrhagic stroke referred to our unit were not offered surgery as all of them had invariably suffered severe neurological damage, either comatose or showing signs of coning. All patients with preoperative stroke who underwent surgery had embolic stroke. Prior to surgery, these patients underwent at least one computed tomography (CT) scan of the brain to exclude haemorrhagic transformation. Neurologists and neurosurgeons were involved in a multidisciplinary team approach in caring for this high-risk group of patients. Surgery was reserved for patients without severe neurological injury and CT evidence of haemorrhagic transformation, who had at least one remaining indication for surgery such as cardiogenic shock, congestive cardiac failure, severe valvular regurgitation, paravalvular abscess, recurrent systemic emboli or fungal endocarditis.
Statistical analysis Statistical analyses were performed with the Statistical Package for Social Science, version 17 (SPSS, Chicago, IL, USA). Continuous variables were expressed as either means with standard deviation or median with interquartile range, as appropriate. These were compared using two-tailed t-test or Mann–Whitney U-test, respectively. Categorical variables, expressed as percentages, were analysed with χ 2 or Fisher’s exact test. For factors that trended towards significance (P < 0.10), a logistic regression analysis was conducted to determine the independent predictors of in-hospital mortality. Survival function and freedom from morbid events were presented using Kaplan–Meier survival curves and comparisons performed with log-rank test. Cox multivariate regression analysis was used to identify the independent predictors of long-term outcomes. Preoperative and operative variables with a univariate P < 0.10 or those judged to be clinically important were entered into the multivariate Cox model. All two-tailed P-values
Surgical management of infective endocarditis: an analysis of early and late outcomes† Philip Y.K. Pang*, Yoong Kong Sin, Chong Hee Lim, Teing Ee Tan, See Lim Lim, Victor T.T. Chao and Yeow Leng Chua Department of Cardiothoracic Surgery, National Heart Centre Singapore, Singapore * Corresponding author. National Heart Centre Singapore, 5 Hospital Drive, Singapore 169609. Tel: +65-6704-8896; fax: +65-844-9063; e-mail:[email protected] (P.Y.K. Pang). Received 30 March 2014; received in revised form 19 May 2014; accepted 10 June 2014
Abstract OBJECTIVES: To review our experience of surgical management of infective endocarditis (IE) over a 13-year period and analyse the outcomes and associated prognostic factors. METHODS: A retrospective review was conducted for 191 consecutive patients who underwent surgery for native and prosthetic valve endocarditis (PVE) between January 2000 and December 2012. Surgical outcomes were reviewed to include survival and postoperative complications. Follow-up was complete for 172 of 179 patients (96.1%) surviving to hospital discharge, with a mean follow-up of 6.6 ± 3.7 years. RESULTS: Mean age was 47.4 ± 14.9 years with 113 (63.9%) males. Native valve endocarditis was present in 177 patients (92.7%). Sixtythree patients (33.0%) presented with embolic complications. The brain was the most common site of embolism, involving 25 patients (13.1%). Streptococcus viridans was the most common infective organism, isolated in 68 patients (35.7%), followed by Staphylococcus aureus in 30 patients (15.7%). Eighty-seven patients (45.5%) had active endocarditis at the time of surgery. The mitral valve was infected in 136 patients (71.2%), the aortic valve in 66 (34.6%), the tricuspid valve in 29 (15.2%) and multiple valves in 38 (19.9%). Nineteen patients (9.9%) were intravenous drug users (IVDU). Twelve IVDUs (63.2%) suffered from tricuspid valve IE, compared with 7 of 162 patients (4.3%) in the non-IVDU population (P < 0.001). The most common indication for early surgery was intractable cardiac failure. Twelve patients (6.3%) died during the hospital stay for surgical treatment of IE. Logistic multivariate analysis identified preoperative creatinine clearance and stroke as independent predictors of in-hospital mortality. Overall 10-year survival and freedom from valve-related reoperation were 74.8 and 90.3%, respectively. Age, PVE, S. aureus endocarditis and postoperative left ventricular ejection fraction (LVEF) ≤45% were factors influencing long-term survival. CONCLUSIONS: Surgical management of endocarditis continues to be challenging and is associated with significant morbidity and mortality. This report of 191 patients who underwent valve surgery for IE shows that in-hospital mortality is influenced by preoperative renal function and stroke at the time of presentation. The optimal timing for surgery in patients with stroke remains controversial. Long-term survival was negatively influenced by increasing age, moderate to severely impaired LVEF, prosthetic valve IE and S. aureus infection. Keywords: Endocarditis • Infective • Heart valve prosthesis implantation
INTRODUCTION Despite substantial improvements made in the diagnosis and management of infective endocarditis (IE), it remains a serious condition that is associated with significant morbidity and mortality. Compared with antibiotic treatment alone, surgery for IE has greatly increased survival [1]. Surgery for IE is required in 25–30% of cases during the acute phase and in 20–40% during the convalescent phase [2, 3]. The most common indications for surgery in IE include intractable heart failure, uncontrolled infection related to perivalvular extension and resistant organisms, recurrent embolic events and presence of prosthetic material [4, 5]. †
Poster presentation at Chest 2013, 30 October 2013, Chicago, IL, USA.
This study was conducted to evaluate the impact of perioperative clinical variables on early and late outcomes in patients undergoing surgery for IE at a tertiary referral centre. With an increasing number of high-risk patients undergoing surgery for IE, the identification of perioperative prognostic factors will be useful for risk assessment and surgical decision-making.
MATERIALS AND METHODS Following approval from the local institutional review board (reference: 2011/880/C), a retrospective case-note and database review was performed on consecutive patients who had undergone surgery for IE between January 2000 and December 2012 at our tertiary referral centre.
© The Author 2014. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ADULT CARDIAC
ORIGINAL ARTICLE
European Journal of Cardio-Thoracic Surgery (2014) 1–7 doi:10.1093/ejcts/ezu281
P.Y.K. Pang et al. / European Journal of Cardio-Thoracic Surgery
2
Definitions In-hospital mortality was defined as all-cause mortality during the hospital stay for the surgical treatment of IE. Operations were considered emergent if performed within 24 h of hospital admission. Active IE was defined as ongoing infection in a patient who was still receiving antibiotic therapy. Renal failure was defined as serum creatinine clearance (CrCl) levels lower than 60 ml/min as calculated with the Cockroft–Gault formula, or the need for renal replacement therapy. Respiratory failure was defined as the need for mechanical ventilation for more than 24 h or the need for reintubation. Moderate to severely impaired left ventricular ejection fraction (LVEF) was defined as LVEF of 45% or less.
Patients During the period January 2000 to December 2012, 191 consecutive patients underwent surgery for IE at our institution. The medical records of all 191 patients were accessible and reviewed. This study focused exclusively on patients with native or prosthetic valve IE treated with open heart surgery. Cases of IE related to nonvalvular cardiovascular devices, such as pacemakers and catheters, were excluded. Preoperative clinical variables of interest are shown in Table 1. All 191 patients had a preoperative transthoracic echocardiogram. Echocardiographic parameters of interest included the presence of vegetation, maximum length of vegetation, abscess, infected valve and LVEF. The New York Heart Association (NYHA) functional class was assessed in all patients before surgery. Nineteen patients (9.9%) were in NYHA Class I, 86 (45.0%) in Class II, 36 (18.8%) in Class III and 50 (26.2%) in Class IV. Treatment for IE consisted of an antibiotic regimen dictated by the primary physician, consulting infectious disease specialist, and surgeon, and varied according to species identified and sensitivities. Whenever possible, each patient completed a standard regimen of a 4- to 6-week course of IV antibiotics after diagnosis. Analysis of
long-term outcomes focused on all-cause mortality, valve-related reoperation and recurrent endocarditis. Patients with IE who presented with haemorrhagic stroke referred to our unit were not offered surgery as all of them had invariably suffered severe neurological damage, either comatose or showing signs of coning. All patients with preoperative stroke who underwent surgery had embolic stroke. Prior to surgery, these patients underwent at least one computed tomography (CT) scan of the brain to exclude haemorrhagic transformation. Neurologists and neurosurgeons were involved in a multidisciplinary team approach in caring for this high-risk group of patients. Surgery was reserved for patients without severe neurological injury and CT evidence of haemorrhagic transformation, who had at least one remaining indication for surgery such as cardiogenic shock, congestive cardiac failure, severe valvular regurgitation, paravalvular abscess, recurrent systemic emboli or fungal endocarditis.
Statistical analysis Statistical analyses were performed with the Statistical Package for Social Science, version 17 (SPSS, Chicago, IL, USA). Continuous variables were expressed as either means with standard deviation or median with interquartile range, as appropriate. These were compared using two-tailed t-test or Mann–Whitney U-test, respectively. Categorical variables, expressed as percentages, were analysed with χ 2 or Fisher’s exact test. For factors that trended towards significance (P < 0.10), a logistic regression analysis was conducted to determine the independent predictors of in-hospital mortality. Survival function and freedom from morbid events were presented using Kaplan–Meier survival curves and comparisons performed with log-rank test. Cox multivariate regression analysis was used to identify the independent predictors of long-term outcomes. Preoperative and operative variables with a univariate P < 0.10 or those judged to be clinically important were entered into the multivariate Cox model. All two-tailed P-values
