Do all patients with prosthetic valve endocarditis need surgery?

Abstract

A best evidence topic in cardiothoracic surgery was written according to a structured protocol. The question addressed was ‘do all patients with prosthetic valve endocarditis need surgery?’ Seventeen papers were found using the reported search that represented the best evidence to answer the clinical question. The authors, journal, date and country of publication, patient group studied, study type, relevant outcomes and results of these papers are tabulated. These studies compared the outcome and survival between surgically and non-surgically treated patients with prosthetic valve endocarditis. Of these studies, two were prospective observational studies and the rest were retrospective studies. The results of most of these papers were in accordance with the guidelines of the American College of Cardiology and American Heart association. These studies showed that unless a patient is not a surgical candidate, an operation is the treatment of choice in prosthetic valve endocarditis. Surgery should be performed as soon as possible, particularly in haemodynamically unstable patients and those who develop complications such as heart failure, valvular dysfunction, regurgitation/obstruction, dehiscence and annular abscess. In addition to the above indications and cardiac/valvularrelated complications of prosthetic valve endocarditis, infection with Staphylococcus aureus plays an important role in the outcome, and the presence of this micro-organism should be considered an urgent surgical indication in the treatment of prosthetic valve endocarditis. Surgery should be performed before the development of any cerebral or other complications. In contrast, in stable patients with other micro-organisms, particularly those with organisms sensitive to antibiotic treatment who have no structural valvular damage or cardiac complications, surgery can be postponed. The option of surgical intervention can also be revisited if there is a change in response to the treatment. This option is reserved for selected patients only and we conclude that as soon as the diagnosis of prosthetic valve endocarditis is made, cardiac surgeons should be involved.

INTRODUCTION

A best evidence topic was constructed according to a structured protocol. This is fully described in the ICVTS [1].

THREE-PART QUESTION

In [patients with prosthetic valve endocarditis (PVE) (mechanical/biological)] does [surgical intervention/surgery] result in [better outcome/survival rate].

CLINICAL SCENARIO

A surgical review has been requested for a 73-year old man with diabetes, renal failure and a previous stroke who had a mechanical aortic valve replacement 15 years ago. The patient was admitted with septicaemia and investigations confirmed endocarditis of the mechanical prosthesis associated with moderate aortic regurgitation. He has been started on broad-spectrum antibiotics and surgical intervention needs to be considered. A redo aortic valve replacement is considered high risk and you decide to check the literature yourself before reaching a decision.

SEARCH STRATEGY

Medline 1950 to June 2012 using OVID interface [exp prosthetic valve endocarditis/OR surgical intervention/surgery.mp] AND [exp prosthetic valve/OR treatment.mp OR exp prosthetic valve endocarditis/OR treatment.mp].

SEARCH OUTCOME

Seventeen papers that provided data to answer the question were found using the reported search. Non-English papers and those that did not compare surgery plus medical therapy with medical therapy alone were excluded. No randomized controlled trials were identified. There were only two prospective observational studies and the rest were retrospective reviews. These studies are presented in Table 1.

Table 1:

Studies on prosthetic valve endocarditis

Author, date, journal, country Study type (level of evidence)	Patient group	Outcomes	Key results	Comments and limitations
Calderwood et al. [3], 1986, J Thorac Cardiovasc Surg Retrospective (level III)	PVE 116; 45 non-surgical, 71 surgical	In complicated PVE better outcome observed with surgical treatment	Outcome difference (P <0.05)	Surgery was used mainly in complicated PVE and S. aureus infection Survival was influenced by congestive cardiac failure
Yu et al. [4], 1994, Ann Thorac Surg Prospective, observational, multi-centric (II)	PVE 74	Lower mortality with surgery compared with non-surgical treatment	Mortality (P <0.05);  • Non-surgical = 56%  • Surgical = 23%
Otaki [5], 1994, ASAIO J Retrospective (level III)	PVE 32; 7 non-surgical, 25 surgical	Non-surgical group all died Higher mortality with early onset PVE compared with late onset	In-hospital mortality (P < 0.05);  • Nonsurgical = 100%  • Surgical = 16%	Patients in non-surgical group were not surgical candidates due to cerebral complications or cardiogenic shock and all died subsequently
Wolf et al. [6], 1995, Chest Retrospective (level III)	PVE 122; 57 non-surgical, 65 surgical	Overall mortality with antibiotics only was higher compared with surgery, however, this significance was due to S. aureus PVE	Mortality at 4 months (P < 0.05);  • Non-surgical = 48%  • Surgical = 25% Mortality in S. aureus group (P < 0.05)  • Non-surgical = 100%  • Surgical = 55% Mortality in non-S. aureus (P = NS)  • Non-surgical = 81%  • Surgical = 89%	S. aureus was the main predictor of death Similar survival was observed between patients that were operated on within 8 days after admission compared with those operated on later
Ho et al. [7], 2010, Asian Cardiovasc Thorac Ann Retrospective (level III)	80 PVE; 46 non-surgical, 34 surgical	Surgery more beneficial in S. aureus and valve dysfunction	In-hospital mortality (P < 0.05)  • Non-surgical = 37%  • Surgical = 15%	Surgery required more in early onset PVE Heart failure was the only independent predictor of hospital mortality
John et al. [8], 1997, Clin Infect Dis Retrospective (level III)	33 PVE; (S. aureus) 9 non-surgical, 14 surgical	Surgical group with or without complications did better compared with the non-surgical group	In-hospital mortality (P < 0.05);  • Non-surgical = 63%  • Surgical = 14%	All patients with S. aureus PVE benefit from surgery even without evidence of cardiac complications Higher mortality with central nervous system involvement is seen and surgery was offered in one patient only
Sohail et al. [9], 2006, Am J Med Retrospective (level III)	55 PVE (S. aureus); non-surgical 23, surgical 32	Mortality was significantly higher in non-surgical group	In-hospital mortality (P < 0.05)  • Non-surgical = 48%  • Surgical = 28%	Surgical group were younger, had lower American Society of Anaesthesia grade Study population was collected for a 20-year period but no difference in the study outcome at the beginning and end of the study was found
Gordon et al. [10], 2000, Ann Thorac Surg Retrospective (level III)	77 PVE (early onset, within 12 months); 23 non-surgical, 54 surgical	Significantly better outcome with surgery in short- and long-term	PVE = 1% 30-day mortality (P < 0.05)  • Non-surgical = 35%  • Surgical = 8% 1-year mortality (P < 0.05)  • Non-surgical = 60%  • Surgical = 26% 2-year mortality (P < 0.05)  • Non-surgical = 63%  • Surgical = 35% 3-year mortality (P < 0.05)  • Non-surgical = 67%  • Surgical = 39%	Almost half of the patients in the non-surgical group were too sick to be offered surgery which influences the survival data Decreased trend of PVE over the recent years
Akowuah et al. [11], 2003, Heart Retrospective (level III)	66 PVE; 28 non-surgical, 38 surgical	Better outcome with surgery but some selected patients can be treated with Abx alone	PVE Incidence = 4.5% In-hospital mortality (P < 0.05)  • Non-surgical = 46%  • Surgical = 24% 10-year survival (P < 0.05)  • Non-surgical = 28%  • Surgical = 58% Freedom from endocarditis (P = NS)  • Non-surgical = 65%  • Surgical = 60% 29% in Abx group required surgery later	Selection bias: some were too sick for surgery and died
Alonso-Valle et al. [12], 2010, J Thorac Cardiovasc Surg Retrospective (level III)	122 PVE; 26 non-surgical, 107 surgical	Better outcome with surgery after 1 year. Patients with perivalvular abscess have worse prognosis Surgery is treatment of choice	PVE Incidence = 2.2% In-hospital mortality (P = NS)  • Non-surgical = 42%  • Surgical = 26% 1-year survival (P < 0.05)  • Non-surgical = 42%  • Surgical = 71%	Small percentage of patients in non-surgical group were included Early onset more common with mechanical valves Selection bias: non-surgical patients had no valvular dysfunction
Habib et al. [13], 2005, Heart Retrospective (level III)	104 PVE; 53 non-surgical, 51 surgical	Less mortality and better survival in staph group and complicated group with surgery compared with non-surgical	PVE incidence = 2% All patients in-hospital mortality (P = NS)  • Non-surgical = 17%  • Surgical = 25% Significantly lower in complicated group	Early onset more common with mechanical valves Selection bias exists Conservative treatment is preserved for patients with uncomplicated and non-S. aureus PVE
Kuyvenhoven et al. [14], 1994, Eur J Cardiothorac Surg Retrospective (level III)	70 PVE; 27 non-surgical, 43 surgical	Surgical group had better survival. Patients too ill for surgery were put in non-surgical group and all died	In-hospital mortality (P = NS)  • Non-surgical = 30%  • Surgical = 26%	Micro-organism and heart failure rather than other complications Mortality in surgical group correlated with abscess formation and dehiscence
López et al. [15], 2007, Eur Heart J Observational (level II)	66 PVE; 29 non-surgical, 37 surgical	No difference between surgical and non-surgical groups	In-hospital mortality (P = NS)  • Non-surgical = 45%  • Surgical = 32%	No long-term outcome Comparing surgical and non-surgical groups was not primary endpoint
Chirouze et al. [16], 2004, Clin Infect Dis Retrospective (multi-centric) (level III)	61 PVE; (S. aureus) 40 non-surgical, 21 early surgery	More heart failure in surgical group. No difference in mortality	PVE = 2.8% Heart failure (P = NS)  • Non-surgical = 30%,  • Early surgery = 42% Difference in mortality (P = NS) Overall mortality = 47%	Stroke was the only prognostic factor Lowest mortality with early surgery in group with cardiac complications Early surgery may not be beneficial in all
Tornos et al. [17], 1997, Clin Infect Dis Retrospective (level III)	Late onset PVE 59; 42 non-surgical, 17 surgical	PVE can be managed medically unless severe valve dysfunction is present	Late PVE = 11% In-hospital mortality (P = NS)  • Non-surgical = 24%  • Surgical = 29% Overall survival at 5-year (80%) and 10-years (63%) not different between groups (P = NS)	S. aureus PVE should undergo surgery Early surgery before cerebral complications recommended in these cases
Truninger et al. [18], 1999, Heart Retrospective (level III)	PVE 49; 10 non-surgical, 39 surgical	Haemodynamically stable patients with non-staphylococcal PVE can be treated with antibiotics	Mortality at mid-term (P = NS)  • Non-surgical = 18%  • Surgical = 14% Mortality at long-term (P = NS)  • Non-surgical = 16%  • Surgical = 14% Need for reoperation (P = NS)  • Non-surgical = 19%  • Surgical = 14%	More enterococcal in non-surgical group Valve dehiscence/abscess more in surgical group
Rekik et al. [19], 2009, Neth Heart J Retrospective (level III)	PVE 48; 28 non-surgical, 20 surgical	Higher mortality in surgical group but not statistically significant	In-hospital mortality (P = NS)  • Non-surgical = 14%  • Surgical = 35% 14% in non-surgical group required surgery at a later date	Longer delay to diagnosis, more heart failure were seen in surgical group

PVE: prosthetic valve endocarditis; ASA: American Society of Anesthesia; S. aureus: Staphylococcus aureus; CNS: central nervous system.

RESULTS

PVE is a serious condition and several factors influence the outcome. Based on the guidelines of the American College of Cardiology/American Heart Association, surgery is indicated for PVE resulting in haemodynamic instability, heart failure or valvular complications such as valve dysfunction/dehiscence, valvular obstruction/regurgitation or abscess formation (Class I). In persistent bacteraemia, relapsing infection and recurrent emboli, surgery is advisable (Class IIa). In cases of uncomplicated PVE with a sensitive organism, however, surgery is not indicated [2]. Our search showed that most studies published to date favour an aggressive approach to surgery for PVE.

In 1986, a study by Calderwood et al. [3] confirmed that patients with complicated PVE do significantly better with surgery compared with those treated with antibiotics alone. In that study, surgery was mainly used to treat complicated PVE associated with Staphylococcus aureus infection. Yu et al. [4] and Otaki [5] also showed better survival with surgery, however, the groups were not comparable and high-risk patients who were not surgical candidates were included in the antibiotic therapy group. Wolf et al. reported their results focusing on infection with S. aureus. Their analysis showed no difference with medical or surgical treatment of patients with PVE with non-staphylococcal organisms, whereas patients with S. aureus PVE had 100% mortality without surgery. They concluded that S. aureus PVE is the main predictor of outcome in PVE [6]. Similarly, Ho et al. [7] confirmed that surgery was more beneficial than medical therapy in S. aureus PVE. John et al. [8] and Sohail et al. [9] also reported the results of S. aureus PVE and showed significantly better outcome with surgery even in patients without valve-related complications.

In another study, 30-day, 1- and 3-year mortalities were shown to be significantly higher in a non-surgical group compared with those who had surgery, however, more than half of the patients in the non-surgical group were too ill to be operated on [10]. Ten-year survival was found to be significantly higher in the surgical group, but there was selection bias arising from the inclusion of non-surgical candidates in their analysis [11]. Other authors have shown better outcomes in these patients after 1 year when treated surgically, however, they failed to show any difference in the incidence of in-hospital mortality [12, 13]. In both of these studies, patients with valvular dysfunction were not included in the non-surgical category, and also conservative treatment was selected for patients with non-S. aureus PVE. Early onset of PVE was found to be more common with mechanical prostheses [12, 13]. Kuyvenhoven et al. [14] and López et al. [15] also showed no difference in the incidence of in-hospital mortality between treatment groups, but the type of micro-organism and the presence of heart failure were more important factors compared with other complications of PVE.

In contrast to the previous studies, Chirouze et al. showed that the overall mortality in S. aureus PVE was not different between the groups whether or not they had surgery. They included patients with heart failure in the surgical group and suggested that early surgery was only beneficial in patients with cardiac complications [16]. Tornos et al. studied late onset PVE and showed no difference between the groups with respect to in-hospital mortality, and 5- and 10-year survivals. They made a recommendation favouring early surgery for S. aureus PVE [17]. Truninger et al. also showed no difference in the mortality rate with or without surgery at mid- and long-term. His surgical group, however, included more patients with valvular complications whereas in his medical group, more patients with enterococcal PVE were included. He concluded that stable patients with non-staphylococcal PVE can be treated medically [18].

Only 1 of the 17 studies demonstrated a worse outcome with surgery, however, the difference did not reach statistical significance. A longer delay in establishing the diagnosis and more heart failure in the surgical group were noted in this study as factors that negatively affect the outcome of surgery [19].

CLINICAL BOTTOM LINE

Unless a patient is not a surgical candidate, in PVE, surgery is the treatment of choice. This is particularly the case in patients who develop complications as a result of PVE. These complications include heart failure, valvular dysfunction, valvular regurgitation or obstruction, valve dehiscence and annular abscess. In these patients, early surgery also improves the outcome and increases the survival.

Infection with S. aureus should be considered an indication for surgery in PVE even without cardiac or valvular complications. These patients should undergo surgery as soon as possible before cerebral complications develop. In patients with other micro-organisms, especially where the organism is sensitive to the particular antibiotic treatment regime and no structural damage or cardiac complications have occurred, surgery can be postponed or reconsidered if there is response to the treatment. This option should be considered for selected patients only and cardiac surgeons should be involved from the early stages of the diagnosis.

Conflict of interest: none declared.