Advertisement

Are homografts superior to conventional prosthetic valves in the setting of infective endocarditis involving the aortic valve?

Open ArchivePublished:January 23, 2016DOI:https://doi.org/10.1016/j.jtcvs.2015.12.061

      Abstract

      Background

      Surgical dogma suggests that homografts should be used preferentially, compared with conventional xenograft or mechanical prostheses, in the setting of infective endocarditis (IE), because they have greater resistance to infection. However, comparative data that support this notion are limited.

      Methods

      From the prospective databases of 2 tertiary academic centers, we identified 304 consecutive adult patients (age ≥17 years) who underwent surgery for active IE involving the aortic valve (AV), in the period 2002 to 2014. Short- and long-term outcomes were evaluated using propensity scores and inverse-probability weighting to adjust for selection bias.

      Results

      Homografts, and xenograft and mechanical prostheses, were used in 86 (28.3%), 139 (45.7%), and 79 (26.0%) patients, respectively. Homografts were more often used in the setting of prosthetic valve endocarditis (58.1% vs 28.8%, P = .002) and methicillin-resistant Staphylococcus (25.6% vs 12.1%, P = .002), compared with conventional prostheses. Early mortality occurred in 17 (19.8%) in the homograft group, and 20 (9.2%) in the conventional group (P = .019). During follow-up (median: 29.4 months; interquartile-range: 4.7-72.6 months), 60 (19.7%) patients died, and 23 (7.7%) experienced reinfection, with no significant differences in survival (P = .23) or freedom from reinfection rates (P = .65) according to the types of prostheses implanted. After adjustments for baseline characteristics, using propensity-score analyses, use of a homograft did not significantly affect early death (odds ratio 1.61; 95% confidence interval [CI], 0.73-3.40, P = .23), overall death (hazard ratio 1.10; 95% CI, 0.62-1.94, P = .75), or reinfection (hazard ratio 1.04; 95% CI, 0.49-2.18, P = .93).

      Conclusions

      No significant benefit to use of homografts was demonstrable with regard to resistance to reinfection in the setting of IE. The choice among prosthetic options should be based on technical and patient-specific factors. Lack of availability of homografts should not impede appropriate surgical intervention.

      Key Words

      Abbreviations and Acronyms:

      AV (aortic valve), CI (confidence interval), HR (hazard ratio), IE (infective endocarditis), PVE (prosthetic valve endocarditis)
      Figure thumbnail fx1
      Kaplan-Meier plots for freedom from reinfection according to type of implanted valve.
      Use of homografts showed no significant benefit compared with conventional prosthetic valves, in the setting of infective endocarditis.
      Among 304 adult patients with active endocarditis undergoing AV replacement, the use of homografts did not demonstrate significant benefit, compared with conventional prosthetic valves, in survival or freedom from reinfection. Therefore, patient-specific factors, such as patient preferences and technical elements, should be the principal drivers of choices for valve prostheses.
      See Article page 1251.
      See Editorial Commentary page 1249.
      See Editorial page 1230.
      Recurrence of infection after valve replacement for infective endocarditis (IE) is a strong concern, and accordingly, the optimal prosthesis in this setting has been debated for decades.
      • Jassar A.S.
      • Bavaria J.E.
      • Szeto W.Y.
      • Moeller P.J.
      • Maniaci J.
      • Milewski R.K.
      • et al.
      Graft selection for aortic root replacement in complex active endocarditis: Does it matter?.
      • Perrotta S.
      • Lentini S.
      In patients with severe active aortic valve endocarditis, is a stentless valve as good as the homograft?.
      • Sweeney M.S.
      • Reul Jr., G.J.
      • Cooley D.A.
      • Ott D.A.
      • Duncan J.M.
      • Frazier O.H.
      • et al.
      Comparison of bioprosthetic and mechanical valve replacement for active endocarditis.
      • Klieverik L.M.
      • Yacoub M.H.
      • Edwards S.
      • Bekkers J.A.
      • Roos-Hesselink J.W.
      • Kappetein A.P.
      • et al.
      Surgical treatment of active native aortic valve endocarditis with allografts and mechanical prostheses.
      Surgical dogma suggests that autologous or allogeneic tissue is preferable to synthetic material in an infected field. Given reluctance to use foreign artificial materials, as with conventional mechanical or stented xenograft valve prostheses, homografts have been advocated by some, with favorable outcomes reported, particularly in cases involving prosthetic valve endocarditis (PVE) and other complex and aggressive lesions, such as root abscess formation.
      • Perrotta S.
      • Aljassim O.
      • Jeppsson A.
      • Bech-Hanssen O.
      • Svensson G.
      Survival and quality of life after aortic root replacement with homografts in acute endocarditis.
      • Preventza O.
      • Mohamed A.S.
      • Cooley D.A.
      • Rodriguez V.
      • Bakaeen F.G.
      • Cornwell L.D.
      • et al.
      Homograft use in reoperative aortic root and proximal aortic surgery for endocarditis: a 12-year experience in high-risk patients.
      • Lund O.
      • Chandrasekaran V.
      • Grocott-Mason R.
      • Elwidaa H.
      • Mazhar R.
      • Khaghani A.
      • et al.
      Primary aortic valve replacement with allografts over twenty-five years: valve-related and procedure-related determinants of outcome.
      • Yankah A.C.
      • Klose H.
      • Petzina R.
      • Musci M.
      • Siniawski H.
      • Hetzer R.
      Surgical management of acute aortic root endocarditis with viable homograft: 13-year experience.
      Most of these reports, however, are single-armed observational studies without conventional prostheses that can be used as comparators.
      • Perrotta S.
      • Aljassim O.
      • Jeppsson A.
      • Bech-Hanssen O.
      • Svensson G.
      Survival and quality of life after aortic root replacement with homografts in acute endocarditis.
      • Preventza O.
      • Mohamed A.S.
      • Cooley D.A.
      • Rodriguez V.
      • Bakaeen F.G.
      • Cornwell L.D.
      • et al.
      Homograft use in reoperative aortic root and proximal aortic surgery for endocarditis: a 12-year experience in high-risk patients.
      • Lund O.
      • Chandrasekaran V.
      • Grocott-Mason R.
      • Elwidaa H.
      • Mazhar R.
      • Khaghani A.
      • et al.
      Primary aortic valve replacement with allografts over twenty-five years: valve-related and procedure-related determinants of outcome.
      • Yankah A.C.
      • Klose H.
      • Petzina R.
      • Musci M.
      • Siniawski H.
      • Hetzer R.
      Surgical management of acute aortic root endocarditis with viable homograft: 13-year experience.
      Obstacles to the use of homografts remain, however: Implantation is technically more complex and is not universally taught. Use of homografts in the current era is as a complete aortic root replacement in almost all cases, and the aggressive nature of this procedure may be the major drawback in recommending it to patients already at high risk of operative mortality.
      • Jassar A.S.
      • Bavaria J.E.
      • Szeto W.Y.
      • Moeller P.J.
      • Maniaci J.
      • Milewski R.K.
      • et al.
      Graft selection for aortic root replacement in complex active endocarditis: Does it matter?.
      • Musci M.
      • Weng Y.
      • Hubler M.
      • Amiri A.
      • Pasic M.
      • Kosky S.
      • et al.
      Homograft aortic root replacement in native or prosthetic active infective endocarditis: twenty-year single-center experience.
      In addition, durability of homografts is limited, which may be a limiting factor, particularly among young patients.
      • Fukushima S.
      • Tesar P.J.
      • Pearse B.
      • Jalali H.
      • Sparks L.
      • Fraser J.F.
      • et al.
      Long-term clinical outcomes after aortic valve replacement using cryopreserved aortic allograft.
      • Shah D.K.
      • Li Z.
      • Park S.J.
      • Daly R.C.
      • Dearani J.A.
      • Schaff H.V.
      • et al.
      Replacement of the infected composite aortic root prosthesis.
      Finally, such grafts are relatively scarce and require cryopreservation. Therefore, the notion that homografts are required may in practice present an obstacle to appropriate surgical management of patients who have IE.
      Despite strong opinion, limited data are available, comparing operative outcomes for homograft versus conventional prosthetic valves in the setting of IE. Given that we have access to data from 2 institutions with similar referral patterns from the same patient pool, and diverse views on the preferred prosthesis, we felt that we had a unique dataset for analysis. We therefore reviewed our combined experience to compare short- and long-term outcomes for homograft versus conventional prosthetic replacements in patients who have active IE involving the AV requiring surgical therapy.

      Methods

       Study Subjects and Outcome Measures

      Using separate cardiac surgery databases from Massachusetts General Hospital and Brigham and Women's Hospital, we identified 304 adult patients (age ≥17 years) who underwent AV replacement for active IE, in the period January 2002 to August 2014 (n = 141 and n = 163, for the 2 institutions, respectively). The patients were categorized based on the type of implanted AV: homograft (homograft group); mechanical prostheses (mechanical group); or xenograft prostheses (xenograft group).
      Primary outcomes of interest were death and valve-related complications, including reinfection or reoperation of the heart valves, thromboembolic events, and anticoagulation-related bleeding. The definitions of valve-related complications, as well as the “active form” of IE, were based on the guidelines of the Society of Thoracic Surgeons.
      • Akins C.W.
      • Miller D.C.
      • Turina M.I.
      • Kouchoukos N.T.
      • Blackstone E.H.
      • Grunkemeier G.L.
      • et al.
      Guidelines for reporting mortality and morbidity after cardiac valve interventions.
      Early mortality was defined as death occurring either in-hospital or within 30 days of surgery.
      Follow-up clinical data were obtained from a centralized clinical data warehouse of Partners Healthcare that encompasses all patient encounters within the system.
      • Murphy S.N.
      • Chueh H.C.
      A security architecture for query tools used to access large biomedical databases.
      For further verification regarding mortality, the Social Security Death Index was searched if necessary. The study protocol was approved by the Institutional Review Board of Partners Healthcare, which encompasses Massachusetts General Hospital and Brigham and Women's Hospital. The requirement for informed consent from individual patients was waived because of the retrospective nature of the study design.

       Statistical Analysis

      Categorical variables, presented as frequencies and percentages, were compared using χ2 analysis or the Fisher exact test. Continuous variables, expressed as mean ± standard deviation, or median with range (or interquartile range), were compared using the Student unpaired t test (between 2 groups) or a 1-way analysis of variance (among 3 groups), as appropriate. Kaplan-Meier analyses were used to assess the probability of survival or freedom from valve-related complications, and log-rank tests were used to compare intergroup differences.
      To adjust for differences in baseline characteristics between the homograft and conventional prostheses groups, propensity-score analyses were used as a weighting variable to create 2 well-balanced cohorts, using the weighting methodology of inverse probability of treatment.
      • Cole S.R.
      • Hernan M.A.
      Adjusted survival curves with inverse probability weights.
      • Normand S.L.
      Evaluating the optimal timing of angiography: landmark or off the mark?.
      • Robins J.M.
      • Hernan M.A.
      • Brumback B.
      Marginal structural models and causal inference in epidemiology.
      Propensity scores were generated using multiple logistic regression analysis. Variables used included all preoperative variables listed in Tables 1 and 2, as well as aorta replacement and coronary artery bypass grafting (both categorical). Concomitant root replacement was not included as one of the covariates in the model, because homograft replacement is not independent of root replacement. Discrimination and calibration of propensity scores were assessed with C statistics and Hosmer-Lemeshow statistics, respectively. Scores for patients receiving homograft replacement were weighted using the term 1/propensity score; those for patients receiving conventional prostheses were weighted using the term 1/(1 – propensity score). Weighted logistic regression and weighted Cox proportional hazard analyses were performed. We assumed that the impact of types of implanted valves on the risks of reinfection would differ between the homograft and conventional prosthetic groups, across the early and late postoperative periods. To address this issue, landmark analyses were performed, split into 2 intervals: ≤360 days; and >360 days after surgery.
      • Therneau T.M.
      • Grambsch P.M.
      Modeling survival data: extending the Cox model.
      Table 1Baseline demographic and clinical variables
      VariableHomograft

      (n = 86)
      Mechanical

      (n = 79)
      Xenograft

      (n = 139)
      P value
      OverallHomograft vs mechanicalHomograft vs xenograft
      Demographic
       Age (y)55.6 ± 16.647.2 ± 14.559.8 ± 14.6.001<.001.049
       Male gender63 (73.3)60 (75.9)105 (75.5).91.69.70
       Race.62.98.24
      White75 (87.2)68 (86.1)129 (92.8)
      Black3 (3.5)3 (3.8)4 (2.9)
      Asian2 (2.3)2 (2.5)3 (2.2)
      Other3 (3.5)2 (2.5)0
      Unknown3 (3.5)4 (5.1)3 (2.2)
       IV drug user15 (17.4)16 (20.3)16 (11.5).19.64.21
      Patient's medical history
       Diabetes mellitus
      No insulin therapy9 (10.5)4 (5.1)17 (12.2).23.20.69
      Insulin therapy5 (5.8)2 (2.5)13 (9.4).14.30.34
       Smoking
      Past28 (32.6)24 (30.4)46 (33.1).92.76.93
      Current15 (17.4)10 (12.7)16 (11.5).43.39.21
      NYHA functional class III or IV47 (54.7)27 (34.2)74 (53.2).011.008.84
      Serum creatinine level (mg/dL)1.32 ± 0.711.56 ± 1.471.52 ± 1.03.15.18.13
      Values are n (%), or mean ± standard deviation, unless otherwise indicated. IV, Intravenous; NYHA, New York Heart Association.
      Table 2Profiles on infective endocarditis and surgical procedure
      Profile dataHomograft

      (n = 86)
      Mechanical

      (n = 79)
      Xenograft

      (n = 139)
      P value
      OverallHomograft vs mechanicalHomograft vs xenograft
      Embolic events26 (30.2)15 (19.0)47 (33.8).065.095.58
       Cerebral23 (26.7)10 (12.7)36 (25.9).046.024.89
      Bacteriology.002.005.044
      Viridans streptococcus12 (14.0)30 (38.0)35 (25.2)
       Other Streptococcus9 (10.5)3 (3.8)16 (11.5)
      Staphylococcus
       Methicillin-susceptible15 (17.4)7 (8.9)27 (19.4)
       Methicillin-resistant22 (25.6)10 (12.7)16 (11.5)
      Entercococcus11 (12.8)11 (13.9)26 (18.7)
       Other8 (9.3)6 (7.6)12 (8.6)
       Negative culture9 (10.5)12 (15.2)7 (5.0)
      Multiple valves affected14 (16.3)33 (41.8)29 (20.9)<.001<.001.40
      Mitral valve affected11 (12.8)30 (38.0)26 (18.7)<.001<.001.24
      Tricuspid valve affected4 (4.7)4 (5.1)3 (2.2).45.90.30
      Vegetation diameter >10 mm36 (41.9)43 (54.4)66 (47.5).27.11.41
      Abscess formation58 (67.4)32 (40.5)41 (29.5)<.001<.001<.001
      Prosthetic endocarditis50 (58.1)31 (39.2)30 (21.6)<.001.015<.001
       Biological31 (36.0)14 (17.7)23 (16.5)<.001.008<.001
       Mechanical19 (22.1)17 (21.5)7 (5.0)<.001.93<.001
      Severe dysfunction of affected valves49 (57.0)57 (72.2)107 (77.0)<.001.042.002
      Left ventricular ejection fraction (%)57.8 ± 11.960.1 ± 9.560.0 ± 11.6.17.16.17
      Emergency surgery86 (28.3)18 (22.8)139 (45.7).91.68.75
      Preoperative intra-aortic balloon pump5 (5.8)5 (6.3)5 (3.6).61.89.43
      Associated surgical procedures
       Aortic root replacement85 (98.8)15 (19.0)15 (10.8)<.001<.001<.001
       Aorta replacement16 (18.6)11 (13.9)27 (19.4).58.41.88
       Coronary bypass30 (34.9)16 (20.3)14 (10.1).022.036<.001
      Cardiopulmonary bypass time (min)318.2 ± 146.5235.0 ± 129.4181.4 ± 136.6<.001<.001<.001
      Cardiac ischemic time (min)236.1 ± 100.7177.1 ± 91.7136.8 ± 89.5<.001<.001<.001
      Values are n (%), or mean ± standard deviation, unless otherwise indicated.
      Multivariable Cox proportional hazards models, using a stepwise, backward-elimination technique that included variables with a P value ≤.20 in univariable analyses, were run to elicit independent risk factors for our outcomes of interest. Only variables with a P value <.10 were included in the final model. For the PVE subgroup analyses, new propensity scores for homograft versus conventional prostheses were calculated using the same methodology. Weighted logistic regression and weighted Cox proportional hazard analyses were performed. R statistical software, version 3.1.2 (https://www.r-project.org/), was used for statistical analyses. All reported P values were the result of 2-tailed tests.

      Results

       Baseline Characteristics

      Homografts, mechanical prostheses, and xenograft prostheses were used in 86 (28.3%), 79 (26.0%), and 139 patients (45.7%), respectively. The baseline demographic and clinical profiles of patients, according to the type of AV implanted, are summarized in Table 1. Overall, patients in the xenograft group were oldest, followed in descending order by those in the homograft and mechanical prostheses groups (P < .001). Patients with advanced heart failure symptoms were more common in the homograft group, compared with the mechanical group (P = .008); however, the prevalence was similar in the homograft and xenograft groups (P = .84).
      In terms of IE pathogens, methicillin-resistant staphylococci was the most common in the homograft group (25.6%), whereas the viridans group streptococci was the leading cause of IE in the mechanical (38.0%) and xenograft groups (25.2%) (Table 2). Patients who received homografts had PVE (P < .001) and abscess formation more often (P < .001), compared with those who received conventional prostheses; however, multivalve involvement (P < .001) and severe valve dysfunction (P = .006) were more common in the latter group (Table 2). Conversely, among patients with PVE, 45.0% had homografts placed.
      All patients, except for one, received aortic root replacement in the homograft group (98.8%); whereas 30 of 218 (13.8%) patients undergoing conventional prosthetic replacement received a root replacement procedure (P < .001; Table 2). Cardiopulmonary bypass and aortic crossclamping times were significantly longer in the homograft group than in the other 2 groups (Table 2).

       Unadjusted Outcomes

      Early mortality occurred in 17 of 86 (19.8%) patients in the homograft group, and 20 of 218 patients (9.2%) in the conventional prostheses groups (P = .011). Follow-up was challenging in this cohort; follow-up data for valve-related complications were complete for 230 (75.7%) patients, with a median duration of 29.4 months (interquartile range: 4.7-72.6 months). During study observation, 60 (20.4%) deaths occurred, along with 47 (15.5%) valve-related complications (Table 3). Kaplan-Meier analyses of survival, and freedom from valve-related complications showed no significant differences among the 3 groups. Specifically, freedom from reinfection at 5 years was 84.9% ± 5.8%, 83.2% ± 6.7%, and 81.1% ± 5.1% in the homograft, mechanical, and xenograft groups, respectively (P = .65) (Figures 1 and 2; Figure E1, Online Data Supplement).
      Table 3Summary of adverse outcomes
      VariablesHomograft

      (n = 86)
      Mechanical

      (n = 79)
      Xenograft

      (n = 139)
      P value
      χ2 analysis for early mortality and log-rank test for overall death, valve-related complications, and composite of death and valve-related complications.
      OverallHomograft vs mechanicalHomograft vs xenograft
      Early (30-d or in-hospital) death17 (19.8)10 (12.7)10 (7.2).019.22.005
      Overall death23 (26.7)15 (19.0)22 (15.8).13.24.047
      Valve-related events14 (16.3)11 (13.8)22 (15.8).90.67.93
       Reinfection6 (7.0)5 (6.3)12 (8.6).65.68.63
       Valve reoperation11 (12.8)4 (5.1)9 (6.5).22.090.72
       Thromboembolism1 (1.2)3 (3.8)7 (5.0).15.24.073
       Anticoagulation-related hemorrhage1 (1.2)4 (5.1)5 (3.6).30.14.12
      Composite of death and valve-related events35 (40.7)23 (29.1)34 (24.5).035.12.010
      Values are n (%), unless otherwise indicated.
      χ2 analysis for early mortality and log-rank test for overall death, valve-related complications, and composite of death and valve-related complications.
      Figure thumbnail gr1
      Figure 1Unadjusted Kaplan-Meier plots for overall mortality, according to type of valve implanted. Longitudinal bar plots indicate 95% confidence intervals (solid = allograft; dashed = bioprosthetic; dotted = mechanical).
      Figure thumbnail gr2
      Figure 2Unadjusted Kaplan-Meier plots for cumulative rates of reinfection, according to type of valve implanted. Longitudinal bar plots indicate 95% confidence intervals (solid = allograft; dashed = bioprosthetic; dotted = mechanical).

       Adjusted Outcomes

      The propensity-score model for the overall cohort showed C statistics of 0.821, and a Hosmer-Lemeshow goodness-of-fit P value of .647; these indicate reasonable discrimation and calibration profiles of the propensity-score model. After weighting, baseline variables were well balanced among the treatment groups (Table E1, Online Data Supplement). After risk adjustment, no evidence was found that homograft usage significantly affects early or long-term mortality, or development of valve-related complications (Table 4).
      Table 4Comparive outcomes of homograft replacement versus conventional prosthetic implantation
      OutcomeCrudeAdjusted
      HR
      For all outcomes except early mortality, which is given as odds ratio. HR, Hazard ratio; CI, confidence interval; NA, not applicable.
      95% CIP valueHR
      For all outcomes except early mortality, which is given as odds ratio. HR, Hazard ratio; CI, confidence interval; NA, not applicable.
      95% CIP value
      Overall cohort (n = 304)
       Early mortality2.431.20-4.93.0131.610.73-3.40.23
       Overall mortality1.510.90-2.53.121.100.62-1.94.75
       Valve-related events0.910.48-1.70.760.800.43-1.48.47
      Reinfection0.970.44-2.14.931.040.49-2.18.93
      Valve reoperation1.740.78-3.92.181.570.70-3.52.28
      Thromboembolism0.200.03-1.55.120.200.03-1.22.082
      Anticoagulation-related bleeding0.230.03-1.83.160.120.01-1.78.13
       Death + valve-related events1.430.95-2.16.0891.070.69-1.66.75
      Prosthetic endocarditis subgroup (n = 111)
       Early mortality1.150.45-2.90.761.880.69-5.24.22
       Overall mortality1.040.51-2.11.911.230.58-2.60.58
       Valve-related events0.700.21-2.41.580.700.22-2.26.55
      Reinfection0.360.03-3.50.380.760.11-5.53.79
      Valve reoperation0.820.14-4.95.830.600.10-3.71.59
      ThromboembolismNANANANANANA
      Anticoagulation-related bleeding0.510.04-5.60.580.410.04-4.30.45
       Death + valve-related events1.030.55-1.91.941.190.62-2.26.60
      For all outcomes except early mortality, which is given as odds ratio. HR, Hazard ratio; CI, confidence interval; NA, not applicable.
      Adjusted Kaplan-Meier curves for rate of valve reinfection for the homograft versus conventional prosthetic groups indicate that the use of a homograft did not significantly affect overall risk of reinfection (Figure 3). Landmark analyses revealed that the risk of reinfection within 1 year showed a trend toward being lower (not statistically significant) in the homograft group (hazard ratio [HR]: 0.33; 95% CI, 0.05-2.01; P = .23), compared with the conventional prosthetic group. After 1 year, the HR was 1.59 (95% CI, 0.64-3.93), with P = .31.
      Figure thumbnail gr3
      Figure 3Adjusted Kaplan-Meier plots for cumulative rates of reinfection of homograft versus conventional prosthetic (biological or mechanical) replacements. Blue and red shading indicates areas within 95% confidence intervals, in the homograft and conventional prosthetic groups, respectively (purple shading indicates overlap). HR, Hazard ratio; CI, confidence interval.
      When the outcome analyses were further adjusted for root replacement procedure in the weighted regression models, early mortality (odds ratio: 0.65; 95% CI, 0.24-1.80; P = .40), overall death (HR 0.67; 95% CI, 0.33-1.35; P = .26), and reinfection (HR 0.47; 95% CI, 0.14-1.59; P = .22), as well as other valve-related events (P value range: .18-.96), did not significantly differ in the homograft versus conventional prosthetic groups.
      Stepwise multivariable Cox hazard analyses revealed that the following independent risk factors affected overall mortality (composite of early and late mortality): insulin-dependent diabetes mellitus (HR 2.91; 95% CI, 1.28-6.58; P = .010); baseline serum-creatinine level (HR 1.20; 95% CI, 1.02-1.42; P = .033); methicillin-resistant Staphylococcus infection (HR 2.55; 95% CI, 1.42-4.25; P = .001); multivalve IE (HR 2.55; 95% CI, 1.45-4.50; P = .001); PVE (HR 1.72; 95% CI, 1.01-2.95; P = .047); preoperative intra-aortic balloon pump placement (HR 3.68; 95% CI, 1.64-8.27; P = .002); and root replacement procedure (HR 2.03; 95% CI, 1.17-3.52; P = .012). The use of a homograft was not a significant predictor of mortality in the multivariable model (HR 0.70; 95% CI, 0.32-1.51; P = .36). Independent risk factors for reinfection in the final multivariable model were as follows: current intravenous drug user (HR 7.68; 95% CI, 3.30-17.88; P < .001); multivalve IE (HR 2.28; 95% CI, 0.97-5.38; P = .059); and severe dysfunction of the affected valves (HR 7.70; 95% CI, 1.03-57.39; P = .046). However, use of a homograft was not a significant predictor of reinfection (HR 0.49; 95% CI, 0.18-1.35; P = .167).

       Prosthetic Valve Endocarditis: A Subgroup Analysis

      Of the 111 patients with PVE, 50 (45.0%) patients received a homograft replacement, whereas the other 61 (55.0%) received either a mechanical (n = 31) or xenograft (n = 30) prosthesis. The early mortality rate among those with PVE was similarly elevated, at 22.0% for homografts and 19.7% for conventional prostheses (P = .76). Unadjusted analyses showed no significant differences in the incidence of individual or composite adverse outcomes for homografts versus conventional prosthetic valve replacement in patients with PVE (Table 4). After risk adjustment (propensity-score model performance (C statistic = 0.840; Hosmer-Lemeshow P = .940; Table E2, Online Data Supplement), outcomes were not significantly affected by the type of prosthesis implanted (Table 4).

      Discussion

      The present study demonstrates no advantage to using one prosthetic option rather than another in the setting of IE; more specifically, the use of homograft prostheses had no apparent advantage in survival or freedom from valve-related complications, including reinfection, compared with conventional mechanical prosthetic or xenograft valves. A higher early mortality rate overall was seen in the homograft group; however, this finding is attributable to the poorer baseline conditions of this group, as risk-adjusted analyses yielded similar results among the groups.
      Our findings suggest that patient-specific factors, such as age and implant preference, as well as technical reconstructive considerations, should drive prosthetic choice, rather than surgical dogma. Although homograft reconstruction is preferred by many surgeons, including some in our group, their implantation is technically more complex, unfamiliar to many surgeons, and requires access to a homograft bank, and these are not universally available. The implication of our results is that surgical intervention for complex IE should not be withheld based solely on the notion that homografts are required in this setting.
      The homograft option is appealing from both a theoretical and practical standpoint. Fundamental surgical principles support use of a tissue rather than a prosthetic option in an infected area, in the interest of minimizing the risk of recurrent infection. The homograft limits prosthetic material to the sutures themselves. By these same principles, aortic allografts have been advocated in the past, as an alternative to extra-anatomic reconstruction for repair of infected descending thoracic or abdominal polyester grafts.
      • Kieffer E.
      • Bahnini A.
      • Koskas F.
      • Ruotolo C.
      • Le Blevec D.
      • Plissonnier D.
      In situ allograft replacement of infected infrarenal aortic prosthetic grafts: results in forty-three patients.
      In recent years, however, satisfactory outcomes have been reported for use of polyester grafts in situ in this setting, perhaps in part thanks to improvements in antibiotic therapy.
      • Fatima J.
      • Duncan A.A.
      • de Grandis E.
      • Oderich G.S.
      • Kalra M.
      • Gloviczki P.
      • et al.
      Treatment strategies and outcomes in patients with infected aortic endografts.
      Similarly, enthusiasm is building for antibiotic therapy alone for some cases of PVE,
      • Lalani T.
      • Chu V.H.
      • Park L.P.
      • Cecchi E.
      • Corey G.R.
      • Durante-Mangoni E.
      • et al.
      In-hospital and 1-year mortality in patients undergoing early surgery for prosthetic valve endocarditis.
      a condition previously considered to be a clear indication for surgical intervention.
      • Nishimura R.A.
      • Carabello B.A.
      • Faxon D.P.
      • Freed M.D.
      • Lytle B.W.
      • O'Gara P.T.
      • 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.
      Accordingly, the theoretical advantage of eliminating prosthetic material in an infected field may be less compelling in the current era.
      From a practical technical standpoint, for those experienced in handling homograft material, it affords flexibility in accommodating difficult root anatomy after aggressive debridement of infected tissue.
      • Lund O.
      • Chandrasekaran V.
      • Grocott-Mason R.
      • Elwidaa H.
      • Mazhar R.
      • Khaghani A.
      • et al.
      Primary aortic valve replacement with allografts over twenty-five years: valve-related and procedure-related determinants of outcome.
      • David T.E.
      • Regesta T.
      • Gavra G.
      • Armstrong S.
      • Maganti M.D.
      Surgical treatment of paravalvular abscess: long-term results.
      In the setting of active IE, a core principle is complete debridement of septic tissue, which may leave significant tissue defects, prior to reconstruction. With a conventional prosthesis, a patch can be used, created by either pericardial tissue or polyester-woven grafts.
      • De Oliveira N.C.
      • David T.E.
      • Armstrong S.
      • Ivanov J.
      Aortic and mitral valve replacement with reconstruction of the intervalvular fibrous body: an analysis of clinical outcomes.
      For a homograft, however, the conduit is pliable, and the anterior leaflet of the donor's mitral valve can be employed to accommodate abscess cavities or other tissue defects, either in an orthotopic position or by rotation of the homograft, as needed. Furthermore, the entire homograft inflow can be sutured directly to the muscle of the recipient left ventricular outflow tract in cases of partial or complete aortoventricular disruption, which sometimes occurs, particularly in prosthetic endocarditis. For those surgeons who are unfamiliar with homografts, however, the pliability of the conduit may in fact pose technical challenges and therefore present a disadvantage.
      Our findings are consistent with those of Jassar and colleagues,
      • Jassar A.S.
      • Bavaria J.E.
      • Szeto W.Y.
      • Moeller P.J.
      • Maniaci J.
      • Milewski R.K.
      • et al.
      Graft selection for aortic root replacement in complex active endocarditis: Does it matter?.
      who evaluated 134 patients undergoing root replacement surgery for IE involving AV, and found that 90 had PVE. In their study, mechanical, nonhomograft biological and homograft prostheses were used in 43, 55, and 36 patients, respectively. They reported no significant differences in the incidences of in-hospital mortality or major complications. During a mean follow-up period of 32.1 months, the 3 patient groups showed comparable valve-related outcomes, including reinfection rates. Several other recent comparative studies on the use of homografts in IE, although varied in lesion location, sample size, and use of control groups, have similarly reported that homograft replacement was not superior to mechanical or xenograft prostheses in the treatment of IE.
      • Klieverik L.M.
      • Yacoub M.H.
      • Edwards S.
      • Bekkers J.A.
      • Roos-Hesselink J.W.
      • Kappetein A.P.
      • et al.
      Surgical treatment of active native aortic valve endocarditis with allografts and mechanical prostheses.
      • David T.E.
      • Regesta T.
      • Gavra G.
      • Armstrong S.
      • Maganti M.D.
      Surgical treatment of paravalvular abscess: long-term results.
      • Avierinos J.F.
      • Thuny F.
      • Chalvignac V.
      • Giorgi R.
      • Tafanelli L.
      • Casalta J.P.
      • et al.
      Surgical treatment of active aortic endocarditis: Homografts are not the cornerstone of outcome.
      • Hagl C.
      • Galla J.D.
      • Lansman S.L.
      • Fink D.
      • Bodian C.A.
      • Spielvogel D.
      • et al.
      Replacing the ascending aorta and aortic valve for acute prosthetic valve endocarditis: Is using prosthetic material contraindicated?.
      • Moon M.R.
      • Miller D.C.
      • Moore K.A.
      • Oyer P.E.
      • Mitchell R.S.
      • Robbins R.C.
      • et al.
      Treatment of endocarditis with valve replacement: the question of tissue versus mechanical prosthesis.
      The current study adds to the literature via its rigorous risk-adjustment methodology to overcome selection bias in our treatment cohort, as well as the relatively heterogeneous attitudes and preferences among the large group of surgeons involved. In general, this group had a tendency to use homografts in more-aggressive forms of IE, such as PVE (P < .001), methicillin-resistant Staphylococcus infection (P = .002), and the presence of root abscess (P < .001). When viewed conversely, however, 26 of 48 (54.2%) patients with this type of infection, 73 of 131 (55.7%) with root abscess, and 61 of 111 (55.0%) with PVE did receive conventional prosthetic replacement. In addition, multivalvular involvement and severe AV dysfunction were significantly more common among the conventional prosthetic group than the homograft group. This mixed overlap of high-risk patients between the homograft and conventional prosthetic groups provided what we believe are comparable cohorts, allowing us to use statistical methods for risk adjustment to evaluate postoperative outcomes. Our risk-adjusted groups were adequately balanced for measured sources of bias (Tables E1 and E2). These analyses revealed similar short- and long-term outcomes, regardless of which conduit was selected for AV replacement (Table 4).
      Prosthetic IE is particularly challenging given the reported high surgical mortality of 10% to 20%.
      • Jassar A.S.
      • Bavaria J.E.
      • Szeto W.Y.
      • Moeller P.J.
      • Maniaci J.
      • Milewski R.K.
      • et al.
      Graft selection for aortic root replacement in complex active endocarditis: Does it matter?.
      • Perrotta S.
      • Lentini S.
      In patients with severe active aortic valve endocarditis, is a stentless valve as good as the homograft?.
      • Musci M.
      • Weng Y.
      • Hubler M.
      • Amiri A.
      • Pasic M.
      • Kosky S.
      • et al.
      Homograft aortic root replacement in native or prosthetic active infective endocarditis: twenty-year single-center experience.
      In addition to the risks inherent in reoperation, the technical aspects of the procedure are complicated by debridement of the infected AV prostheses, including the sewing ring and surrounding infected tissue. In this setting, isolated AV replacement may be challenging, despite motivation to simplify the operation by avoiding root replacement. In our series, however, only 65 of 111 patients (58.6%) with PVE underwent root replacement: 49 of 50 (98.0%) in the homograft group, and only 16 of 61 (26.2%) in the conventional prosthetic group (P < .001). Even with this conservative approach in the conventional prosthetic replacement group, long-term outcomes were comparable in both crude and adjusted analyses (Table 4).
      These findings may suggest that PVE in the AV does not always necessitate root replacement, but that more conservative AV replacement may be possible if complete removal of infected tissue leaves adequate root geometry. The less-extensive isolated AV replacement was correlated with shorter aortic clamping and cardiopulmonary bypass times (P < .001), which may be correlated with lower perioperative risks in this high-risk cohort. This finding additionally implies that patients who are cared for in settings without access to homografts can still undergo surgery for their disease and should not be denied surgery owing to the notion that homografts are necessary in all cases.

       Limitations

      This study has limitations inherent to the retrospective observational design. As in any surgical series, one can anticipate surgical bias in the selection of conduits for valve replacement. Although we used appropriate statistical adjustment techniques, residual confounding caused by unmeasured covariates still may have affected our results. Our database includes cases from 2 different institutions, with diverse staff and practice, which helps ameliorate some of the difficulties with single-center designs. Nevertheless, the 2 hospitals share a geographic location, which may affect the generalizability of our findings to other populations.
      Although this series is one of the largest that has sought to evaluate the impact of homograft uses on surgical outcomes in the setting of IE, the study might have been underpowered for testing the study hypotheses, owing to an insufficient sample size. The study results, therefore, should be interpreted with caution and need to be validated by further, larger-scale studies.
      Obtaining reliable follow-up data regarding valve-related complications was challenging given the significant attrition of patients (>20%). The institutional review board did not approve direct contact with patients as a means to obtain follow-up information, owing to the involvement of sensitive (intravenous drug use) information, by which longitudinal data were restricted to use within Partners Healthcare–affiliated hospitals in Massachusetts. This lack of complete follow-up data may have affected the overall validity of the study results.

      Conclusions

      No significant benefits of homograft, compared with standard, prosthetic valves were demonstrated with regard to reinfection or survival, in the setting of IE affecting AV. These findings suggest that patient-specific factors, such as patient preferences and technical considerations, should be the principal drivers of choices of valve prostheses. Furthermore, lack of access to homografts should not be considered an obstacle to surgical therapy for this serious condition.

       Conflict of Interest Statement

      Thoralf M. Sundt, MD, is a consultant for Thorasos Therapeutics. All other authors have nothing to disclose with regard to commercial support.

      Appendix

      Figure thumbnail fx2
      Figure E1Unadjusted Kaplan-Meier plots for cumulative rates of valve reoperation (left), and composite of death and valve-related complications (right), according to types of valve implanted. Longitudinal bar plots indicate 95% confidence intervals (solid = allograft; dashed = bioprosthetic; dotted = mechanical).
      Table E1Baseline balance table for overall cohort, after adjustment using inverse-probability weighting based on propensity scores
      CharacteristicsHomograft

      (n = 86)
      Conventional prosthesis

      (n = 218)
      P value
      Age (y)53.1 ± 17.055.0 ± 16.2.49
      Male gender76.576.7.98
      Race.48
       White89.991.6
       Black1.32.8
       Asian3.82.3
       Other2.54.7
       Unknown2.52.8
      IV drug user18.215.8.70
      Diabetes mellitus
       Without insulin therapy7.48.8.68
       With insulin therapy5.06.3.66
      Past smoking34.833.2.84
      Current smoking18.213.5.48
      NYHA functional class III or IV46.047.8.82
      Creatinine (mg/dL)1.39 ± 0.861.48 ± 1.09.56
      Embolic events24.326.8.71
       Cerebral21.220.7.94
      Methicillin-resistant Staphylococcus infection16.115.7.94
      Multiple valves affected19.924.3.52
      Mitral valve affected17.721.7.55
      Tricuspid valve affected2.73.0.85
      Vegetation diameter >10 mm36.533.7.81
      Abscess formation47.242.9.60
      Prosthetic valve endocarditis39.535.7.62
       Biological23.921.6.71
       Mechanical15.614.1.78
      Severe dysfunction of affected valves70.271.0.91
      Left ventricular ejection fraction (%)59.8 ± 10.959.5 ± 10.9.86
      Emergency surgery23.923.7.98
      Preoperative IABP3.94.6.80
      Aorta replacement15.617.1.77
      CABG20.719.3.81
      Values are %, or mean ± standard deviation, unless otherwise indicated. IV, Intravenous; NYHA, New York Heart Association; IABP, intra-aortic balloon pump; CABG, coronary artery bypass graft.
      Table E2Baseline balance table in patients with prosthetic valve endocarditis, after adjustment using inverse-probability weighting based on propensity scores
      CharacteristicHomograft

      (n = 50)
      Conventional

      prostheses

      (n = 61)
      P value
      Age (y)60.1 ± 13.160.5 ± 15.7.90
      Male gender86.785.2.81
      Race.41
       White93.488.0
       Black3.34.0
       Asian1.60
       Other04.0
       Unknown1.64.0
      IV drug user9.49.8.95
      Diabetes mellitus
       Without insulin therapy11.911.0.89
       With insulin therapy7.46.8.92
      Past smoking36.837.9.93
      Current smoking7.46.8.92
      NYHA functional class III or IV40.443.9.76
      Creatinine (mg/dL)1.36 ± 0.771.41 ± 0.62.68
      Embolic events18.616.4.76
       Cerebral15.111.5.58
      Methicillin-resistant Staphylococcus infection26.427.0.96
      Multiple valves affected14.415.7.87
      Mitral valve affected11.914.2.76
      Tricuspid valve affected2.51.5.67
      Vegetation diameter >10 mm36.533.7.81
      Abscess formation10.865.9.68
      Prosthetic valve endocarditis100100NA
       Biological54.757.6.81
       Mechanical45.342.4.81
      Severe dysfunction of affected valves44.944.6.98
      Left ventricular ejection fraction (%)56.8 ± 10.757.0 ± 11.8.95
      Emergency surgery20.721.7.91
      Preoperative IABP3.55.9.51
      Aorta replacement25.522.5.77
      CABG36.131.4.68
      Values are %, or mean ± standard deviation, unless otherwise indicated. IV, Intravenous; NYHA, New York Heart Association; NA, not applicable; IABP, intra-aortic balloon pump; CABG, coronary artery bypass graft.

      References

        • Jassar A.S.
        • Bavaria J.E.
        • Szeto W.Y.
        • Moeller P.J.
        • Maniaci J.
        • Milewski R.K.
        • et al.
        Graft selection for aortic root replacement in complex active endocarditis: Does it matter?.
        Ann Thorac Surg. 2012; 93: 480-487
        • Perrotta S.
        • Lentini S.
        In patients with severe active aortic valve endocarditis, is a stentless valve as good as the homograft?.
        Interact Cardiovasc Thorac Surg. 2010; 11: 309-313
        • Sweeney M.S.
        • Reul Jr., G.J.
        • Cooley D.A.
        • Ott D.A.
        • Duncan J.M.
        • Frazier O.H.
        • et al.
        Comparison of bioprosthetic and mechanical valve replacement for active endocarditis.
        J Thorac Cardiovasc Surg. 1985; 90: 676-680
        • Klieverik L.M.
        • Yacoub M.H.
        • Edwards S.
        • Bekkers J.A.
        • Roos-Hesselink J.W.
        • Kappetein A.P.
        • et al.
        Surgical treatment of active native aortic valve endocarditis with allografts and mechanical prostheses.
        Ann Thorac Surg. 2009; 88: 1814-1821
        • Perrotta S.
        • Aljassim O.
        • Jeppsson A.
        • Bech-Hanssen O.
        • Svensson G.
        Survival and quality of life after aortic root replacement with homografts in acute endocarditis.
        Ann Thorac Surg. 2010; 90: 1862-1867
        • Preventza O.
        • Mohamed A.S.
        • Cooley D.A.
        • Rodriguez V.
        • Bakaeen F.G.
        • Cornwell L.D.
        • et al.
        Homograft use in reoperative aortic root and proximal aortic surgery for endocarditis: a 12-year experience in high-risk patients.
        J Thorac Cardiovasc Surg. 2014; 148: 989-994
        • Lund O.
        • Chandrasekaran V.
        • Grocott-Mason R.
        • Elwidaa H.
        • Mazhar R.
        • Khaghani A.
        • et al.
        Primary aortic valve replacement with allografts over twenty-five years: valve-related and procedure-related determinants of outcome.
        J Thorac Cardiovasc Surg. 1999; 117 (discussion 91): 77-90
        • Yankah A.C.
        • Klose H.
        • Petzina R.
        • Musci M.
        • Siniawski H.
        • Hetzer R.
        Surgical management of acute aortic root endocarditis with viable homograft: 13-year experience.
        Eur J Cardiothorac Surg. 2002; 21: 260-267
        • Musci M.
        • Weng Y.
        • Hubler M.
        • Amiri A.
        • Pasic M.
        • Kosky S.
        • et al.
        Homograft aortic root replacement in native or prosthetic active infective endocarditis: twenty-year single-center experience.
        J Thorac Cardiovasc Surg. 2010; 139: 665-673
        • Fukushima S.
        • Tesar P.J.
        • Pearse B.
        • Jalali H.
        • Sparks L.
        • Fraser J.F.
        • et al.
        Long-term clinical outcomes after aortic valve replacement using cryopreserved aortic allograft.
        J Thorac Cardiovasc Surg. 2014; 148: 65-72.e2
        • Shah D.K.
        • Li Z.
        • Park S.J.
        • Daly R.C.
        • Dearani J.A.
        • Schaff H.V.
        • et al.
        Replacement of the infected composite aortic root prosthesis.
        Ann Thorac Surg. 2011; 92: 1651-1655
        • Akins C.W.
        • Miller D.C.
        • Turina M.I.
        • Kouchoukos N.T.
        • Blackstone E.H.
        • Grunkemeier G.L.
        • et al.
        Guidelines for reporting mortality and morbidity after cardiac valve interventions.
        J Thorac Cardiovasc Surg. 2008; 135: 732-738
        • Murphy S.N.
        • Chueh H.C.
        A security architecture for query tools used to access large biomedical databases.
        Proc AMIA Symp. 2002; : 552-556
        • Cole S.R.
        • Hernan M.A.
        Adjusted survival curves with inverse probability weights.
        Comput Meth Progr Biomed. 2004; 75: 45-49
        • Normand S.L.
        Evaluating the optimal timing of angiography: landmark or off the mark?.
        Circulation. 2007; 116: 2656-2657
        • Robins J.M.
        • Hernan M.A.
        • Brumback B.
        Marginal structural models and causal inference in epidemiology.
        Epidemiology. 2000; 11: 550-560
        • Therneau T.M.
        • Grambsch P.M.
        Modeling survival data: extending the Cox model.
        Springer, New York2000
        • Kieffer E.
        • Bahnini A.
        • Koskas F.
        • Ruotolo C.
        • Le Blevec D.
        • Plissonnier D.
        In situ allograft replacement of infected infrarenal aortic prosthetic grafts: results in forty-three patients.
        J Vasc Surg. 1993; 17 (discussion 55-6): 349-355
        • Fatima J.
        • Duncan A.A.
        • de Grandis E.
        • Oderich G.S.
        • Kalra M.
        • Gloviczki P.
        • et al.
        Treatment strategies and outcomes in patients with infected aortic endografts.
        J Vasc Surg. 2013; 58: 371-379
        • Lalani T.
        • Chu V.H.
        • Park L.P.
        • Cecchi E.
        • Corey G.R.
        • Durante-Mangoni E.
        • et al.
        In-hospital and 1-year mortality in patients undergoing early surgery for prosthetic valve endocarditis.
        JAMA Intern Med. 2013; 173: 1495-1504
        • Nishimura R.A.
        • Carabello B.A.
        • Faxon D.P.
        • Freed M.D.
        • Lytle B.W.
        • O'Gara P.T.
        • 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: 676-685
        • David T.E.
        • Regesta T.
        • Gavra G.
        • Armstrong S.
        • Maganti M.D.
        Surgical treatment of paravalvular abscess: long-term results.
        Eur J Cardiothorac Surg. 2007; 31: 43-48
        • De Oliveira N.C.
        • David T.E.
        • Armstrong S.
        • Ivanov J.
        Aortic and mitral valve replacement with reconstruction of the intervalvular fibrous body: an analysis of clinical outcomes.
        J Thorac Cardiovasc Surg. 2005; 129: 286-290
        • Avierinos J.F.
        • Thuny F.
        • Chalvignac V.
        • Giorgi R.
        • Tafanelli L.
        • Casalta J.P.
        • et al.
        Surgical treatment of active aortic endocarditis: Homografts are not the cornerstone of outcome.
        Ann Thorac Surg. 2007; 84: 1935-1942
        • Hagl C.
        • Galla J.D.
        • Lansman S.L.
        • Fink D.
        • Bodian C.A.
        • Spielvogel D.
        • et al.
        Replacing the ascending aorta and aortic valve for acute prosthetic valve endocarditis: Is using prosthetic material contraindicated?.
        Ann Thorac Surg. 2002; 74 (discussion S92-9): S1781-S1785
        • Moon M.R.
        • Miller D.C.
        • Moore K.A.
        • Oyer P.E.
        • Mitchell R.S.
        • Robbins R.C.
        • et al.
        Treatment of endocarditis with valve replacement: the question of tissue versus mechanical prosthesis.
        Ann Thorac Surg. 2001; 71: 1164-1171

      Linked Article

      • Resistance to infection of long-term cryopreserved human aortic valve allografts
        The Journal of Thoracic and Cardiovascular SurgeryVol. 151Issue 5
        • Preview
          To analyze the in vitro antimicrobial activity of 3 antibiotic regimens (group A, gentamicin-piperacillin-vancomycin-metronidazole-amphotericin B; group B, gentamicin-piperacillin-flucloxacillin-metronidazole-amphotericin B; and group C, meropenem-vancomycin-tobramycin-colistin-amphotericin B) used in the processing of cryopreserved human ascending aortic tissue and aortic valves against Staphylococcus epidermidis and Staphylococcus aureus. The results were additionally compared with the infection resistance of cryopreserved ascending aortic tissue against Escherichia coli and Pseudomonas aeruginosa.
        • Full-Text
        • PDF
        Open Archive
      • Challenging allograft use for aortic valve infective endocarditis: Is it the allograft or the surgeon?
        The Journal of Thoracic and Cardiovascular SurgeryVol. 153Issue 2
        • Preview
          We read with interest and some concern the recent paper by Kim and colleagues1 and its accompanying editorials, which concluded that there was no significant benefit to allograft use with regard to resistance to reinfection after surgery for aortic valve infective endocarditis (IE). With wider ramifications to current surgical practice, they concluded that such surgery can be safely and appropriately managed in centers with standard aortic valve surgery experience, without access to or experience with allografts.
        • Full-Text
        • PDF
        Open Archive