| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
a Département de Cardiologie, Hôpital Timone, Marseille, France
b Service de Chirurgie Cardiaque, Hôpital Timone, Marseille, France
c LERTIM, Faculté de Médecine de Marseille, Hôpital Timone, Marseille, France
d Fédération de Microbiologie, Hôpital Timone, Marseille, France
Accepted for publication June 15, 2007.
* Address correspondence to Dr Habib, La Timone Hospital, Cardiology, Blvd Jean Moulin, Marseille, 13005, France (Email: gilbert.habib{at}free.fr).
| This article has been selected for the open discussion forum on the CTSNet Web Site: http://www.ctsnet.org/sections/newsandviews/discussions/index.html
|
 |
Abstract |
|---|
 Top Abstract IntroductionPatients and MethodsResultsCommentReferences |
|---|
Methods: All consecutive patients with definite diagnosis of aortic infective endocarditis operated on during the active phase were included. Endpoints were in-hospital mortality and a combined endpoint including infective endocarditis recurrence, prostheses dysfunction, or long-term cardiovascular mortality.
Results: Among 127 consecutive patients, mean age 57 ± 15 years, 87% male, 30% with preexisting aortic prosthesis, and 63 (50%) with annulus abscess, 54 (43%) were treated with aortic homograft and 73 (57%) with conventional prosthesis. Median time between diagnosis and surgery was 10 days. In-hospital mortality was 9%, not different between homograft and conventional prostheses (11% versus 8%, p[ = 0.6). By multivariable analysis, prosthetic valve endocarditis (8.5 95% confidence interval: 2.2 to 33.6, ]p = 0.001) was the only variable independently associated with in-hospital mortality, which was not influenced by type valvular substitute (p = 0.6), even in the subset with annulus abscess (p = 0.2). Ten-year survival free from the combined endpoint was 44% ± 10%, not different between homograft and conventional prostheses (log rank p = 0.2). By multivariable analysis, comorbidity index (2.6 [1.05 to 6.3], p = 0.04) and prosthetic valve endocarditis (2.3 [1.2 to 4.6], p = 0.02) were independently predictive of the combined endpoint, which was not determined by type of valvular substitute (p = 0.6) even in the subset with annulus abscess (p = 0.5).
Conclusions: Implantation of conventional prostheses during the active phase of aortic endocarditis yields similar low operative mortality and long-term prognosis as compared with aortic homografts, even in patients with annulus abscess.
Recent changing patterns in infective endocarditis (IE) include increased rates of early surgery, which is currently performed in almost 50% of all patients [1, 2] and a trend for decreased in-hospital mortality rates, the latter potentially related to the former [1]. Among patients operated on during the active phase, however, despite better understanding of the surgical approach [3] and standardization of antiobiotics regimen [4, 5], reported in-hospital mortality rates remain around 15% [2], and even higher among subsets such as severe heart failure patients, annulus abscess, or prosthetic valve endocarditis (PVE) [6]. Type of valvular substitute to be implanted has been sharply debated, and conflicting reports on its impact on outcome [7–10] did not allow clear recommendations [4, 5]. In the aortic position, cryopreserved homografts raised hopes in view of their ability to restore the aortic root in case of annulus abscess, to offer better intrinsic resistance to infection and to allow higher antibiotic penetration [5, 7, 11]. However initial enthusiasm has been tempered by low availability, technical issues [12], concerns regarding durability [13], reports of IE recurrences [14], and more importantly, by efficient alternative procedures using conventional prostheses [3, 6, 9].
Taking advantage of a prospectively enrolled population of patients with aortic IE operated on during the active phase, the present study aimed to analyze, among other determinants of outcome, the impact of the type of valvular substitutes on in-hospital mortality and long-term prognosis. We hypothesized that homograft implantation as compared with conventional prostheses is not a determinant of outcome, including among patients with annulus abscess.
|
Patients and Methods |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentReferences |
|---|
Clinical Definitions
Baseline clinical data were prospectively collected at study entry. Occurrence of cerebral embolic events was screened by systematic cerebral computed tomography scan at study entry and noncerebral systemic embolic events by systematic abdominal echography or computed tomography scan depending upon creatinin level. Associated baseline medical conditions were assigned weighs and summated to calculate a comorbidity index [17]. Aortic surgery during the active phase was indicated for at least one of the following conditions: hemodynamic impairment referring to heart failure or acute severe aortic regurgitation with signs of elevated left ventricular pressure, persistent sepsis despite adequate antibiotic treatment, or high embolic risk due to persistence of large and mobile vegetation [2, 4, 5]. Synergic antibiotic treatment was conducted intravenously as recommended [4, 5]. Postoperative follow-up events were prospectively collected during routine follow-up visits systematically performed after dismissal.
Microbiological Data
Blood cultures were systematically collected at admission, and serological assessment was completed for Coxiella burnetii, Bartonella sp, Chlamydiae sp, Legionella, Brucella, Mycoplasma pneumoniae, Candida sp, and Aspergillus sp. Rare microorganisms such as Escherichia coli, Enterobacter cloacae, Gemella morbillorum, Corynebactéries, Bartonnella henselea, Mycoplasma hominis, Propionibacterium acnes, and Aspergillus sp were labeled "atypical microorganisms." In addition, direct analysis and polymerase chain reaction since 1995 were performed on each explanted valve.
Echocardiographic Methods
All patients underwent both transthoracic and transesophageal comprehensive echocardiography in our laboratory within 24 hours of admission by a senior cardiologist. Transthoracic and transesophageal echocardiograms were systematically repeated once a week and when clinically indicated. Data were electronically stored and used as noted at the time of original examination without alteration. Degree of aortic regurgitation when present was assessed semiquantitatively or by quantitative methods for the most recent patients and classified as mild, moderate, or severe [18]. Diagnosis of aortic annulus abscess was suspected in front of a thickened area or mass with heterogeneous echogenic or echolucent appearance at the annulus level [19].
Surgical Procedures
Indications for aortic surgery during the active phase were hemodynamic impairment in 19 patients (15%), high embolic risk in 17 (14%), persistent sepsis in 37 (29%), and more than one of the aforementioned reasons in 54 (42%). When the infection was limited to the cusps of native aortic valves or to the leaflets of aortic prostheses, surgery consisted in simple valve replacement by conventional prostheses or in implantation of aortic homograft, at the discretion of the surgeon. When infection extended to surrounding structures, surgery consisted of extensive resection of all infected tissues and their reconstruction by either autologous fresh or glutaraldehyde-fixed bovine pericardium, Dacron fabrics, or aortic homograft [3, 6]. In case of annulus abscess with annulus disruption, its reconstruction was achieved depending upon surgeon preference either by fresh autologous or glutaraldehyde-fixed bovine pericardium followed by implantation of a conventional prostheses or by direct implantation of an aortic homograft. Before 1995, the subcoronary technique was used, whereas after 1995, homografts were uniformly implanted with the aortic root replacement techniques [20, 21].
Statistical Analysis
Patients data were compared among groups with 
2 or Fisher’s exact test for nominal variables and with Mann-Whitney U test for ordinal variables. Endpoints were in-hospital mortality—defined as mortality occurring postoperatively before dismissal—and a combined endpoint grouping IE recurrence, noninfective prostheses dysfunction or long-term cardiovascular mortality. The effect of potential risk factors on operative mortality were examined in a logistic regression model. The effect of potential risk factors on the combined endpoint were analyzed computing Kaplan-Meier estimates, and curves were compared using the log-rank test. Cox proportional hazards models were used to estimate the adjusted effect of prognostic factors on the time dependent endpoint. All statistical tests were two-sided. For univariate analysis, p values less than 0.05 was considered significant. Multivariable analysis was performed with a forward stepwise approach with p = 0.10 as the threshold for entering or removing variables. All results obtained by multivariable analysis were systematically adjusted for type of valvular substitute (homograft versus conventional prostheses). Analyses were performed with SPSS software (version 13.0 2004; SPSS, Chicago, Illinois).
|
Results |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentReferences |
|---|
Surgical procedures
Median time between diagnosis and surgery was 10 days (range, 0 to 46). Type of valvular substitute included aortic homografts in 54 patients (43%)—8 with subcoronary techniques—bioprosthesis in 50 (39%), and mechanical prosthesis in 23 (18%), including 1 Bentall procedure. In 14 patients (11%), extension of the lesions to mitral anterior leaflet required associated mitral procedures (repair in 4 patients, valve replacement in the remaining). Comparison of preoperative characteristics between patients with homograft and those with conventional prostheses is presented in Tables 1 and 2.
|
|
|
Long-Term Recurrence, Prostheses Dysfunction, or Cardiovascular Mortality
The 115 patients who survived the operative period were followed up during 313.13 person-years (median, 2.06 years; maximum, 12). Follow-up was complete in 98% of patients.
Recurrences of IE were observed in 8 patients, 4 with homograft and 4 with conventional prostheses. Three of those patients (all with aortic homograft) required reoperation. Micro-organisms involved were oral streptococci in 3, Staphylococcus aureus in 1, Streptococcus bovis in 1, enterococci in 1, HACEK in 1, and atypical in 1. All micro-organisms during recurrence were different from the original micro-organism. Noninfective prostheses dysfunction occurred in 5 patients (3 with bioprostheses and 2 with mechanical prostheses), 2 of whom required reoperation. Cardiovascular death occurred in 24 patients, 11 with homograft and 13 with conventional prostheses.
The combined endpoint including recurrence of IE, noninfective prostheses dysfunction, or cardiovascular death was reached by 35 patients (31% [20% of patients with homograft versus 18% of patients with conventional prostheses, p = 0.7]). Five- and ten-year survival free from the endpoint were 68% ± 5% and 44% ± 10% respectively, and were not significantly different between homograft and conventional prostheses (log rank p = 0.2; Fig 2). By univariate analysis, age (p = 0.02), comorbidity index (p = 0.009), and PVE (p = 0.02) were the only variables predictive of the combined endpoint. By multivariable analysis (hazards ratio, 95% CI), comorbidity index (2.6 [1.05 to 6.3], p = 0.04) and PVE (2.3 [1.2 to 4.6], p = 0.02) remained independently predictive of the endpoint. The multivariable model was systematically adjusted to type of valvular substitute, which did not influence long-term outcome (0.6 [0.3 to 1.4], p = 0.6, for homograft versus conventional prostheses; Fig 3).
|
|
Patients With Aortic Annular Abscess
Among 63 patients with annulus abscess, 41 (65%) were treated by aortic homografts and 22 (35%) were treated by pericardial annular reconstruction and implantation of either bioprosthesis (15 patients) or mechanical prosthesis (7 patients). Associated mitral procedures were performed in 5 patients. Among 23 patients with PVE and annulus abscess, 14 (61%) were treated by homograft and 9 (39%) by conventional prostheses. Staphylococcus aureus was the most frequent microbiological finding (21%) followed by Streptococcus bovis (16%), oral streptococci (11%) and Coxiella burnetti (6%). Microorganism remained unknown in 15% of cases.
Operative mortality
Seven patients died postoperatively before dismissal leading to 11% operative mortality, not significantly higher than in patients without annulus abscess (8% operative mortality, p = 0.5). Operative mortality appeared higher among patients with homograft as compared with conventional prostheses but the difference was not significant (14.5% versus 4.5%, p = 0.4). By multivariable analysis (odds ratio [95% CI]), PVE (12.4 [1.4 to 114.2], p < 0.001) was the only variable independently associated with operative mortality, which was not influenced by type of valvular substitute (4.7 [0.5 to 45.1], p = 0.2, for homograft versus conventional prostheses).
The 56 patients who survived the operative period were followed up during 167.85 person-years (median, 2.85 years; maximum, 11). During follow-up, recurrence of IE was observed in 5 patients, 4 with homograft and 1 with conventional prostheses (p = 0.7). The combined endpoint was reached by 19 patients (34%), and 5-year survival free from the endpoint was 68% ± 7%, and was not significantly different between homograft and conventional prostheses (log rank p = 0.5; Fig 4). Multivariable analysis (hazard ratio [95% CI]) identified Staphylococcus aureus infection (4.3 [1.2 to 15.2], p = 0.02) and negative blood cultures (6.2 [2.1 to 18.7], p = 0.001) as the only variables independently predictive of the endpoint, which was not influenced by type of valvular substitute (0.7 [0.2 to 2.0], p = 0.5, for homograft versus conventional prostheses).
|
|
Comment |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentReferences |
|---|
The recently reported trend in decreased in-hospital mortality [1] in patients with IE may find an explanation in improvements in the medical management of the disease with standardization of length, nature and mode of administration of antibiotic regimens [4, 5], but also in a multidisciplinary approach [22] involving cardiologists, infectious disease consultants, and cardiac surgeons. The increased rate of early surgery consistently reported in the recent years is indeed a fundamental evolution in the management of IE, with nearly half of patients operated on during the active phase [1, 2] and could be another explanation for the observed improved outcome [1]. However, in-hospital mortality rates among operated patients remain high, particularly among those with severe heart failure preoperatively, annulus abscess, or PVE [2, 6]. In the quest for determinants of postoperative outcome improvement, cryopreserved homografts fulfill all theoretical requirements, namely, ability of restoring the aortic root anatomy and of providing best hemodynamic profile with no need for anticoagulation, intrinsic resistance to infection, and better penetration of antibiotics [7, 23]. Promising results were accordingly reported, including in high-risk subsets such as patients with annulus abscess or PVE [7, 24].
These results were tempered however with those similarly satisfactory, obtained with conventional prostheses [6], including in patients with annulus abscess [3] or PVE [9]. Only a few studies [7, 10] compared respective use of both types of valvular substitutes and yields discordant results. Moreover, besides technical difficulties of aortic homograft implantation and unresolved issues regarding the best techniques [12], their low availability [2], reports of endocarditis recurrences [23], uncertainties regarding durability [13], signs of rejection, and absence of viability on explanted specimen [14] emphasized that homograft may not be the once expected panacea. Our data, which account for the very few [7, 10] comparing a reasonable number of patients with active aortic IE and indifferently treated in the same institution with either aortic homograft or conventional prostheses, failed to demonstrate the superiority of the former over the latter.
The absence of impact on outcome of aortic homograft, both in terms of operative mortality and long-term recurrences, prostheses dysfunction, or cardiovascular mortality, including in the subsets of patients with annulus abscess, is further enhanced by less severe preoperative clinical conditions in patients with homografts as attested by younger age, lower comorbidity index, and lower rate of heart failure. The larger number of homografts implanted in patients with annulus abscess, which reflects in part the preference of one of our surgeons who operated on the largest number of patients, may nevertheless reflect actual different anatomic presentation and the same echocardiographic finding (ie, annulus abscess) can hide different surgical lesions. We do not refute the fact that some patients with extreme annulus disruption and aortic root destruction are more easily treated by homografts implantation by a surgeon with specific training and skill in this tecnics. However, most of these lesions can also be treated by pericardial reconstruction of the annulus and aortic root replacement with composite graft, and although not randomized, our data show that when implantation of conventional material is feasible, results are similar to those obtained with homografts implantation including in the subset with annulus abscess. These data therefore do not support systematic use of homografts, including in patients with annulus abscess.
Radical, extensive extirpation of all infected tissues as already advocated [3, 6] and systematically performed in our surgical department rather than the type of valvular substitute per se could thus be a plausible explanation of the observed low in-hospital mortality and long-term complication rates. These results, which compare favorably with the recently published European data [2], could find another potential explication in aggressive surgical indications as attested to by higher rates of patients operated on during the active phase (62%) compared with those commonly reported [1, 2] and by short median time between diagnosis and surgery. This aggressive surgical management has been for long the policy of our department and does not reflect referral to our center of the most severe patients, as none was operated on in cardiogenic shock and only 5 were contraindicated owing to excessive operative risk during the study period. Early surgery before intractable heart failure or excessive extension of anatomic lesions may explain the unexpected [6] absence of excess mortality observed among patients with heart failure or annulus abscess, which will deserve future prospective studies to be confirmed. These encouraging results, however, should not mask the persistent high risk carried by PVE, consistently reported [6] and similarly observed in our series despite this aggressive surgical approach. The increased in-hospital mortality and long-term complication rates in patients with PVE seem to be mainly due to complexity of surgical treatment [6] and should recall to us the critical need for promoting valve repair over replacement [25].
Study Limitations
Obviously, our patients were not randomly allocated to either homografts or conventional prostheses, which does not allow definite statement on their respective impact on outcome. However, the choice of the type of valvular substitutes was mostly dependent on the surgeon preference and inclination for the type of prostheses he was comfortable with. Moreover, the two groups compare well, and the only significant differences in baseline comparison between homografts and conventional prostheses did not favor the latter. Lastly, referral bias in tertiary care centers is always possible; although our center is the main referral center for the disease in our region, the clinical, echocardiographic, and microbiological picture of our population was closed to that reported in a population-based setting [1].
In conclusion, among patients operated on during the active phase of aortic infective endocarditis for hemodynamic impairment, persistent sepsis, or high embolic risk, implantation of conventional material when technically feasible was associated with similar outcome as aortic homograft implantation. Low in-hospital mortality and long-term complications rates were similarly observed among all subsets but PVE, including patients with heart failure or annulus abscess, irrespective of the type of valvular substitutes. Early aggressive surgical management of aortic endocarditis, including complete large resection of all infected tissues, might be one explanation for the good short-term and long-term results obtained in this particularly high-risk population.
|
References |
|---|
|
TopAbstractIntroductionPatients and MethodsResultsCommentReferences |
|---|
Related Article
This article has been cited by other articles:
|
|
 |
|
  Endorsed by the European Society of Clinical Micro, Authors/Task Force Members, G. Habib, B. Hoen, P. Tornos, F. Thuny, B. Prendergast, I. Vilacosta, P. Moreillon, M. de Jesus Antunes, et al. 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) Eur. Heart J., October 1, 2009; 30(19): 2369 - 2413. [Full Text] [PDF]
|
|
|
|
 |
|
 H. A. Cooper, E. C. Thompson, R. Laureno, A. Fuisz, A. S. Mark, M. Lin, and S. A. Goldstein Subclinical Brain Embolization in Left-Sided Infective Endocarditis: Results From the Evaluation by MRI of the Brains of Patients With Left-Sided Intracardiac Solid Masses (EMBOLISM) Pilot Study Circulation, August 18, 2009; 120(7): 585 - 591. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Rekik, I. Trabelsi, I. Maaloul, M. Hentati, A. Hammami, I. Frikha, M. Ben Jemaa, and S. Kammoun Short- and long-term outcomes of surgery for active infective endocarditis: a Tunisian experience Interactive CardioVascular and Thoracic Surgery, August 1, 2009; 9(2): 241 - 245. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. Rahimtoola The Year in Valvular Heart Disease J. Am. Coll. Cardiol., May 19, 2009; 53(20): 1894 - 1908. [Full Text] [PDF]
|
|
|
|
 |
|
 T. Mihaljevic Invited commentary Ann. Thorac. Surg., December 1, 2007; 84(6): 1942 - 1942. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
