Abstract
Objective:
The objectives of this study were to investigate patient characteristics, valve pathology, bacteriology, and surgical techniques related to outcome of patients who underwent surgery for isolated native (NVE) or prosthetic (PVE) mitral valve endocarditis.
Methods:
From January 2002 to January 2020, 447 isolated mitral endocarditis operations were performed, 326 for NVE and 121 for PVE. Multivariable analysis of time-related outcomes used random forest machine learning.
Results:
Staphylococcus aureus was the most common causative organism. Of 326 patients with NVE, 88 (27%) underwent standard mitral valve repair, 43 (13%) extended repair, and 195 (60%) valve replacement. Compared with NVE with standard repair, patients who underwent all other operations were older, had more comorbidities, worse cardiac function, and more invasive disease. Hospital mortality was 3.8% (n = 17); 0 (0%) after standard valve repair, 3 (7.0%) after extended repair, 8 (4.1%) after NVE replacement, and 6 (5.0%) after PVE re-replacement. Survival at 1, 5, and 10 years was 91%, 75%, and 62% after any repair and 86%, 62%, and 44% after replacement, respectively. The most important risk factor for mortality was renal failure. Risk-adjusted outcomes, including survival, were similar in all groups. Unadjusted extended repair outcomes, particularly early, were similar or worse than replacement in terms of reinfection, reintervention, regurgitation, gradient, and survival.
Conclusions:
A patient- and pathology-tailored approach to surgery for isolated mitral valve endocarditis has low mortality and excellent results. Apparent superiority of standard valve repair is related to patient characteristics and pathology. Renal failure is the most powerful risk factor. In case of extensive destruction, extended repair shows no benefit over replacement.
Keywords: mitral valve, endocarditis, isolated
Graphical Abstract
Survival after surgery for mitral valve infective endocarditis is worse than after surgery for aortic valve endocarditis, even after accounting for confounding factors.1–3 Although recommended team-based multidisciplinary management and early aggressive surgery might yield improved surgical outcomes,4–8 multiple studies suggest outcomes are better if the valve is repaired rather than replaced, with fewer reinfections and reoperations.9–14 Experienced mitral valve surgeons report repair in up to two-thirds of patients with native valve endocarditis (NVE).15,16 However, these operations were not limited to active disease, but included remote or healed endocarditis, bound to have less severe pathology. No one repairs all infected native mitral valves: replacement is required for the most advanced pathologies.17
The challenge is extensive pathology, including mitral anular invasion and calcification and tissue and cusp destruction,18,19 for which standard repair techniques are insufficient and, if repaired, require patch-repair techniques.15 Effects of disease and surgical factors on choice of repair versus replacement and their effects on outcomes in the context of underlying patient conditions are not well characterized in recent literature.
Therefore, the objectives of this study were to investigate patient characteristics, disease-related valve pathology and bacteriology, and surgical techniques among patients who underwent surgery for isolated mitral valve endocarditis and their association with early outcomes and time-related reinfection, reintervention, mortality, and, for patients who underwent mitral valve repair, non–infection-related reintervention and valve hemodynamics.
METHODS
Patients
From January 2002 to January 2020, 2303 operations were performed in adults with active infective endocarditis, 447 for isolated mitral valve endocarditis in 429 unique patients, which constituted the study cohort (Figure E1).
Data
Patient characteristics, operative and pathologic findings, procedure details, microbiology results, echocardiography measures, and hospital outcomes were extracted from prospective quality registries. These data were approved for use in research, with patient consent waived (institutional review board numbers 07–043 and 4826; approved January 31, 2007).
Pathology
Patient records, operative reports, and operative transesophageal echocardiographic data were reviewed to identify disease pathology and presence of invasive disease and its stage.20 Noninvasive disease was infection confined to valve cusps, and invasive disease extended into the anulus, atrioventricular groove, or surrounding structures. Pathology was coded as previously described.20
Management
Infective endocarditis is managed with a team-based, multidisciplinary approach, the core of which includes a cardiologist, infectious disease specialist, and cardiac surgeon.20 All patients undergo preoperative brain imaging, and those with neurological symptoms and positive brain imaging are seen by a neurologist. All persons who inject drugs (PWIDs) are seen by an addiction psychiatry specialist and other members of the addiction team. Other specialists are brought in by indication, most often a nephrologist. Surgery is recommended as soon as possible after an indication for it is established, provided the patient is operable and has no contraindication or reason for delaying surgery, such as stroke or intracranial hemorrhage. Risk of hemorrhagic conversion or embolic stroke is weighed against risk of hemodynamic deterioration and additional emboli. It is desirable that antibiotic sensitivity of the infecting organism is known preoperatively, or that patients are covered with broad-spectrum antibiotics.
At surgery, infected and necrotic tissue and foreign material are radically debrided, followed by generous irrigation. Mitral valve repair has been the procedure of choice whenever technically feasible for NVE, with valve replacement performed when extensive destruction prevents adequate repair.
We distinguish NVE repair using standard maneuvers normally used for degenerative mitral valve disease, including use of small, <1-cm cusp patches to close perforations not involving the free margin, from extended repair of valves with more extensive tissue destruction (Table E1).
End Points
End points were postoperative complications, defined according to the Society of Thoracic Surgeons National Cardiac Database (http://www.sts.org/registries-research-center/sts-national-database/adult-cardiac-surgery-database/data-collection), reinfection treated medically or surgically, mitral valve reintervention, non–infection-related reintervention, longitudinal mitral regurgitation grade and gradient, and time-related mortality. Cross-sectional follow-up, with a closing date of May 25, 2021, was conducted, with a median of 4.4 years of follow-up (Figure E2).
Data Analysis
For statistical analyses, we used SAS version 9.4 (SAS Institute, Inc) and R version 4.0.3 (R Foundation for Statistical Computing). Continuous variables are summarized as mean ± standard deviation or equivalent 15th, 50th (median), and 85th percentiles for skewed distributions; for comparisons, we used the Wilcoxon rank-sum test. Categorical data are summarized as frequencies and percentages; for comparisons, we used the χ2 test.
Time-related and longitudinal analyses.
Time-to-event probabilities were obtained using the Kaplan–Meier estimator. Patients experiencing isolated mitral valve reinfection were re-entered as separate observations. Low occurrence of repeated events precluded repeated-events analysis.
For temporal trends in postoperative mitral valve regurgitation grade and gradient, we used all echocardiograms obtained during follow-up visits at our institution. A nonlinear longitudinal decomposition mixed-effects model was used to characterize temporal patterns, with patients entered as random effects.21
Risk factor identification.
For variables associated with time-related outcomes, we used random forest machine learning (RandomForestSRC version 2.12.0; R Foundation for Statistical Computing), for which 5000 trees were grown using a randomly chosen subset of variables for each split from the 70 listed in Appendix E1.
RESULTS
Patient Characteristics, Organisms, and Pathology
Of the 447 operations for isolated mitral valve endocarditis, 326 were for NVE and 121 for prosthetic valve endocarditis (PVE). Patients presented in 3 groups: NVE with untouched valves (n = 282; 63%), NVE after previous valve repair (n = 44; 9.8%), and PVE (n = 121; 27%; Table 1). Patients with NVE after previous repair presented with more cardiac morbidity than those with untouched valves (Figure 1, A). Compared with NVE patients, those with PVE were older, had more cardiac and noncardiac comorbidities, and underwent more nonelective operations (Figure 1, B).
TABLE 1.
Characteristics of patients undergoing surgery for mitral valve infective endocarditis, and their disease
| NVE (n = 326) | NVE untouched valve (n = 282) | NVE previous repair (n = 44) | PVE (n = 121) | |||||
|---|---|---|---|---|---|---|---|---|
|
|
|
|
|
|||||
| Characteristic | n* | Value | n* | Value | n* | Value | n* | Value |
| Demographics | ||||||||
| Age, y | 326 | 55 ± 14 | 282 | 55 ± 14 | 44 | 56 ± 14 | 121 | 57 ± 15 |
| Female | 326 | 132 (40) | 282 | 116 (41) | 44 | 16 (36) | 121 | 55 (45) |
| Body mass index, kg/m2 | 325 | 26 (20, 34) | 281 | 26 (20, 35) | 44 | 25 (21, 32) | 119 | 27 (22, 34) |
| Clinical status | ||||||||
| Preoperative intra-aortic balloon pump | 326 | 16 (4.9) | 282 | 16 (5.7) | 44 | 0 (0) | 121 | 5 (4.1) |
| Surgical acuity | 326 | 282 | 44 | 121 | ||||
| Elective | 59 (18) | 56 (20) | 3 (6.8) | 12 (9.9) | ||||
| Urgent | 250 (77) | 210 (74) | 40 (91) | 101 (83) | ||||
| Emergency | 17 (5.2) | 16 (5.7) | 1 (2.3) | 8 (6.6) | ||||
| Cardiac comorbidity |
||||||||
| Mitral valve | ||||||||
| Stenosis | 325 | 22 (6.8) | 282 | 16 (5.7) | 43 | 6 (14) | 120 | 14 (12) |
| Regurgitation | 326 | 291 (89) | 282 | 261 (93) | 44 | 30 (68) | 121 | 67 (55) |
| Grade | 315 | 272 | 43 | 113 | ||||
| None/trivial | 24 (7.6) | 11 (4.0) | 13 (30) | 46 (41) | ||||
| Mild | 30 (9.5) | 20 (7.4) | 10 (23) | 16 (14) | ||||
| Moderate | 67 (21) | 58 (21) | 9 (21) | 19 (17) | ||||
| Severe | 194 (62) | 183 (67) | 11 (26) | 32 (28) | ||||
| Tricuspid valve regurgitation | 320 | 137 (43) | 278 | 114 (41) | 42 | 23 (55) | 120 | 71 (59) |
| Heart failure | 324 | 131 (40) | 280 | 112 (40) | 44 | 19 (43) | 121 | 66 (55) |
| Left ventricular ejection fraction, % | 316 | 58 ± 8.6 | 276 | 59 ± 8.5 | 40 | 54 ± 8.4 | 119 | 55 ± 8.5 |
| Complete heart block | 300 | 4 (1.3) | 263 | 4 (1.5) | 37 | 0 (0) | 106 | 6 (5.7) |
| Atrial fibrillation or flutter | 313 | 49 (16) | 269 | 35 (13) | 44 | 14 (32) | 114 | 38 (33) |
| Noncardiac comorbidity | ||||||||
| Intravenous drug use | 154 | 22 (14) | 132 | 18 (14) | 22 | 4 (18) | 57 | 16 (28) |
| Renal dialysis | 326 | 51 (16) | 282 | 47 (17) | 44 | 4 (9.1) | 121 | 18 (15) |
| Creatinine, mg/dL | 326 | 1.1 (0.74, 2.8) | 282 | 1.1 (0.74, 2.8) | 44 | 1.1 (0.71, 2.3) | 121 | 1.2 (0.86, 2.6) |
| Previous stroke | 325 | 122 (38) | 282 | 104 (37) | 43 | 18 (42) | 121 | 66 (55) |
| Hypertension | 325 | 207 (64) | 281 | 182 (65) | 44 | 25 (57) | 121 | 88 (73) |
| History of smoking | 316 | 167 (53) | 272 | 147 (54) | 44 | 20 (45) | 115 | 70 (61) |
| Chronic obstructive pulmonary disease | 326 | 61 (19) | 282 | 48 (17) | 44 | 13 (30) | 121 | 33 (27) |
| Peripheral artery disease | 325 | 50 (15) | 281 | 42 (15) | 44 | 8 (18) | 121 | 14 (12) |
| Diabetes | 322 | 92 (29) | 278 | 85 (31) | 44 | 7 (16) | 120 | 31 (26) |
| Bilirubin, mg/dL | 306 | 0.50 (0.30, 0.90) | 265 | 0.50 (0.30, 0.90) | 41 | 0.50 (0.30, 0.97) | 117 | 0.60 (0.30, 1.6) |
| Microorganism | 305 | 267 | 38 | 98 | ||||
| Staphylococcus aureus | 100 (33) | 88 (33) | 12 (32) | 25 (26) | ||||
| Coagulase-negative staphylococcus | 23 (7.5) | 15 (5.6) | 8 (21) | 17 (17) | ||||
| Enterococcus | 29 (9.5) | 25 (9.4) | 4 (11) | 13 (13) | ||||
| Streptococcus viridans | 68 (22) | 66 (25) | 2 (5.3) | 11 (11) | ||||
| Gram-positive cocci | 51 (17) | 45 (17) | 6 (16) | 8 (8.2) | ||||
| Fungal | 3 (0.98) | 2 (0.75) | 1 (2.6) | 8 (8.2) | ||||
| Polymicrobial | 8 (2.6) | 7 (2.6) | 1 (2.6) | 6 (6.1) | ||||
| Other | 23 (7.5) | 19 (7.1) | 4 (11) | 10 (10) | ||||
| Pathology | ||||||||
| Vegetations present | 326 | 298 (91) | 282 | 257 (91) | 44 | 41 (93) | 121 | 102 (84) |
| Invasive disease | 326 | 73 (22) | 282 | 64 (23) | 44 | 9 (20) | 121 | 41 (34) |
| Circumferential extent | 323 | 279 | 44 | 120 | ||||
| Less than one-third | 36 (11) | 33 (12) | 3 (6.8) | 15 (13) | ||||
| One-third to one-half | 32 (9.9) | 28 (10) | 4 (9.1) | 19 (16) | ||||
| Full circumference | 2 (0.62) | 0 (0) | 2 (4.5) | 6 (5.0) | ||||
| Stage | ||||||||
| Cellulitis | 326 | 63 (19) | 282 | 54 (19) | 44 | 9 (20) | 121 | 31 (26) |
| Abscess | 326 | 47 (14) | 282 | 43 (15) | 44 | 4 (9.1) | 121 | 19 (16) |
| Abscess cavity | 324 | 16 (4.9) | 280 | 16 (5.7) | 44 | 0 (0) | 121 | 6 (5.0) |
| Pseudoaneurysm | 324 | 1 (0.31) | 280 | 0 (0) | 44 | 1 (2.3) | 121 | 1 (0.83) |
Data are presented as No. (%), mean ± SD, or median (15th, 85th percentiles). NVE, Native valve endocarditis; PVE, prosthetic valve endocarditis.
Patients with data available.
FIGURE 1.
Graphical representation of differences in patient characteristics among 3 principal groups. These differences have been reduced to a standardized, and thus comparable, common metric using standardized mean differences. Standardized mean differences more negative than −10% or greater than 10% are considered important. Thus, these graphs show substantial and important differences in patient characteristics among these groups. A, Nativevalve endocarditis (NVE) with a history of previous mitral valve repair versus NVE with an untouched mitral valve. Those with an untouched valve versus a previously repaired valve had more severe mitral valve regurgitation (MR), which is off scale at 110, were more likely to have Streptococcus viridans infection, and more likely to undergo elective operation than those who had previous mitral valve repair. B, NVE of untouched and previously repaired valves versus prosthetic valve endocarditis (PVE). Those with PVE versus NVE were less likely to have severe mitral valve regurgitation (standardized difference 93, off the scale), but more likely to have had a stroke, be in heart failure, have a greater circumferential extent of infection, and have invasive disease. LVEF, Left ventricular ejection fraction; BMI, body mass index; COPD, chronic obstructive pulmonary disease.
Staphylococcus aureus was the most common causative organism, with minor differences in NVE subgroups (Table 1). Coagulase-negative staphylococci were more commonly identified in NVE patients with previously repaired valves and PVE. Streptococcus viridans was most common with NVE on untouched valves. Enterococcus was present in 42 (10%). Fungal infection was uncommon: 11 infections (2.7%), 8 PVE.
Most patients had vegetations. Invasive disease occurred in 114 (26%) and was more commonly encountered in patients with PVE than NVE, 41 (34%) versus 73 (22%), with more invasive disease in all stages.
The Operations
Of the 447 operations, 131 were valve repairs (29%) and 316 replacements (71%), 121 for PVE. Repairs for NVE increased during the study period and replacements declined (Figure E3).
Of 326 NVE operations, 88 (27%) were standard repairs and 43 (13%) extended repairs (2 after previous repair; Tables 2 and E1). Extended repairs increased over the study period (Figure E3), most commonly for extensive posterior cusp destruction, with the posterior P2 scallop most frequently patched; 56% of these patches (n = 23) involved the cusp free margin, 3 (7.0%) were supported by artificial chords, 5 (12%) required multiple patches, and 8 (19%) were performed in the presence of mitral anular calcification. Seven (17%) had invasive anulus disease requiring reconstruction. Extended repair was performed more often than standard repair in patients with more comorbidities, such as anemia, renal failure, diabetes, chronic lung disease, and heart failure (Figure 2, A).
TABLE 2.
Characteristics of patients undergoing surgery for mitral valve infective endocarditis, and their disease, according to major operative category
| NVE standard repair (n = 88) | NVE extended repair (n = 43) | NVE replacement (n = 195) | PVE replacement (n = 121) | |||||
|---|---|---|---|---|---|---|---|---|
|
|
|
|
|
|||||
| Characteristic | n* | Value | n* | Value | n* | Value | n* | Value |
| Demographics | ||||||||
| Age, y | 88 | 53 ± 14 | 43 | 50 ± 15 | 195 | 57 ± 14 | 121 | 57 ± 15 |
| Female | 88 | 30 (34) | 43 | 24 (56) | 195 | 78 (40) | 121 | 55 (45) |
| Body mass index, kg/m2 | 88 | 25 (20, 31) | 43 | 28 (20, 38) | 194 | 26 (20, 35) | 119 | 27 (22, 34) |
| Clinical status | ||||||||
| Preoperative intra-aortic balloon pump | 88 | 2 (2.3) | 43 | 1 (2.3) | 195 | 13 (6.7) | 121 | 5 (4.1) |
| Surgical acuity | 88 | 43 | 195 | 121 | ||||
| Elective | 28 (32) | 11 (26) | 20 (10) | 12 (9.9) | ||||
| Urgent | 57 (65) | 30 (70) | 163 (84) | 101 (83) | ||||
| Emergency | 3 (3.4) | 2 (4.7) | 12 (6.2) | 8 (6.6) | ||||
| Cardiac comorbidity | ||||||||
| Previous mitral valve surgery | 88 | 43 | 195 | 121 | ||||
| Untouched | 75 (85) | 41 (95) | 166 (85) | 0 (0) | ||||
| Previous repair | 13 (15) | 2 (4.7) | 29 (15) | 0 (0) | ||||
| Previous replacement | 0 (0) | 0 (0) | 0 (0) | 121 (100) | ||||
| Mitral valve | ||||||||
| Stenosis | 88 | 0 (0) | 43 | 0 (0) | 194 | 22 (11) | 120 | 14 (12) |
| Regurgitation | 88 | 79 (90) | 43 | 39 (91) | 195 | 173 (89) | 121 | 67 (55) |
| Grade | 84 | 41 | 190 | 113 | ||||
| None/Trivial | 5 (6.0) | 2 (4.9) | 17 (8.9) | 46 (41) | ||||
| Mild | 10 (12) | 2 (4.9) | 18 (9.5) | 16 (14) | ||||
| Moderate | 17 (20) | 10 (24) | 40 (21) | 19 (17) | ||||
| Severe | 52 (62) | 27 (66) | 115 (61) | 32 (28) | ||||
| Tricuspid valve regurgitation | 87 | 30 (34) | 42 | 12 (29) | 191 | 95 (50) | 120 | 71 (59) |
| Heart failure | 88 | 22 (25) | 43 | 16 (37) | 193 | 93 (48) | 121 | 66 (55) |
| Left ventricular ejection fraction, % | 85 | 58 ± 9.1 | 43 | 61 ± 7.4 | 188 | 58 ± 8.6 | 119 | 55 ± 8.5 |
| Complete heart block | 85 | 0 (0) | 40 | 0 (0) | 175 | 4 (2.3) | 106 | 6 (5.7) |
| Atrial fibrillation or flutter | 86 | 10 (12) | 42 | 7 (17) | 185 | 32 (17) | 114 | 38 (33) |
| Noncardiac comorbidity | ||||||||
| Intravenous drug use | 48 | 6 (13) | 20 | 7 (35) | 86 | 9 (10) | 57 | 16 (28) |
| Renal dialysis | 88 | 7 (8.0) | 43 | 9 (21) | 195 | 35 (18) | 121 | 18 (15) |
| Creatinine, mg/dL | 88 | 0.99 (0.78, 1.8) | 43 | 0.93 (0.68, 5.4) | 195 | 1.1 (0.72, 3.4) | 121 | 1.2 (0.86, 2.6) |
| Previous stroke | 88 | 32 (36) | 43 | 12 (28) | 194 | 78 (40) | 121 | 66 (55) |
| Hypertension | 88 | 48 (55) | 43 | 25 (58) | 194 | 134 (69) | 121 | 88 (73) |
| History of smoking | 87 | 39 (45) | 43 | 27 (63) | 186 | 101 (54) | 115 | 70 (61) |
| Chronic obstructive pulmonary disease | 88 | 7 (8.0) | 43 | 8 (19) | 195 | 46 (24) | 121 | 33 (27) |
| Peripheral artery disease | 88 | 8 (9.1) | 43 | 3 (7.0) | 194 | 39 (20) | 121 | 14 (12) |
| Diabetes | 88 | 19 (22) | 43 | 17 (40) | 191 | 56 (29) | 120 | 31 (26) |
| Bilirubin (mg/dL) | 86 | 0.40 (0.30, 0.70) | 39 | 0.50 (0.30, 0.80) | 181 | 0.50 (0.30, 1.0) | 117 | 0.60 (0.30, 1.60) |
| Microorganism | 83 | 40 | 182 | 98 | ||||
| Staphylococcus aureus | 25 (30) | 17 (43) | 58 (32) | 25 (26) | ||||
| Coagulase-negative staphylococcus | 3 (3.6) | 0 (0) | 20 (11) | 17 (17) | ||||
| Enterococcus | 7 (8.4) | 2 (5.0) | 20 (11) | 13 (13) | ||||
| Streptococcus viridans | 26 (31) | 5 (13) | 37 (20) | 11 (11) | ||||
| Gram-positive cocci | 9 (11) | 9 (23) | 33 (18) | 8 (8.2) | ||||
| Fungal | 2 (2.4) | 0 (0) | 1 (0.55) | 8 (8.2) | ||||
| Polymicrobial | 2 (2.4) | 1 (2.5) | 5 (2.7) | 6 (6.1) | ||||
| Other | 9 (11) | 6 (15) | 8 (4.4) | 10 (10) | ||||
| Pathology | ||||||||
| Vegetations present | 88 | 74 (84) | 43 | 39 (91) | 195 | 185 (95) | 121 | 102 (84) |
| Invasive disease | 88 | 7 (8.0) | 43 | 9 (21) | 195 | 57 (29) | 121 | 41 (34) |
| Circumferential extent | 87 | 43 | 193 | 120 | ||||
| Less than one-third | 3 (3.4) | 6 (14) | 27 (14) | 15 (13) | ||||
| One-third to one-half | 3 (3.4) | 3 (7.0) | 26 (13) | 19 (16) | ||||
| Full circumference | 0 (0) | 0 (0) | 2 (1.0) | 6 (5.0) | ||||
| Stage | ||||||||
| Cellulitis | 88 | 5 (5.7) | 43 | 8 (19) | 195 | 50 (26) | 121 | 31 (26) |
| Abscess | 88 | 4 (4.5) | 43 | 4 (9.3) | 195 | 39 (20) | 121 | 19 (16) |
| Abscess cavity | 88 | 1 (1.1) | 43 | 3 (7.0) | 193 | 12 (6.2) | 121 | 6 (5.0) |
| Pseudoaneurysm | 88 | 1 (1.1) | 43 | 0 (0) | 193 | 0 (0) | 121 | 1 (0.83) |
Data are presented as No. (%), mean ± SD, or median (15th, 85th percentiles). NVE, Native valve endocarditis; PVE, prosthetic valve endocarditis.
Patients with data available.
FIGURE 2.
Graphical representation of differences in patient characteristics among major operative groups, depicted as standardized mean differences as in Figure 1. These show substantial and important differences among groups. A, Among patients with native valve endocarditis (NVE), both untouched and with previous repair, standard repair versus extended repair. Those who underwent extended repair versus standard repair had more extensive endocarditis, more comorbidities, such as anemia, renal failure, diabetes, chronic lung disease, and heart failure, and were more likely to have more-invasiveorganisms. B, Among patients with NVE who underwent either standard or extended repair, repair versus valve replacement. Those undergoing valve replacement rather than repair had more circumferential extent and invasiveness of infection and more nonelective procedures. C, Comparison of patients with valve replacement for NVE versus those who underwent valvere-replacement for prostheticvalve endocarditis (PVE). Comparedwith patientswith PVE,thosewith NVE who underwent replacement had more concomitant procedures, less extensive circumferential disease, fewer strokes, fewer heart failure symptoms, and higher ejection fraction. MR, Mitral regurgitation; COPD, chronic obstructive pulmonary disease; BMI, body mass index; LVEF, left ventricular ejection fraction.
Among NVE patients, 195 (60%) underwent mitral valve replacement (Table 2). Compared with those who underwent repair, these patients were more likely to have invasive disease, 57 (29%) versus 16 (12%), have greater circumferential extent of disease, and were less likely to undergo elective surgery (Figure 2, B). They had more heart failure, chronic lung disease, and anemia. Of the 44 with NVE and previous mitral repair, 15 (34%) underwent re-repair and 29 (66%) valve replacement (not analyzed as a separate group).
All with PVE underwent valve re-replacement; 41 (34%) had invasive disease and 19 (16%) underwent anular patch reconstruction for this or anular calcification. PVE patients differed from those who underwent valve replacement for NVE, having greater extent of disease, more heart failure symptoms, more strokes, and other comorbidities, such as pulmonary and hepatic disease and heart failure; they were less likely to undergo elective operations (Table 2 and Figure 1, B).
Extended repair and valve replacement patients had more concomitant procedures and longer cardiopulmonary bypass and myocardial ischemic times than standard repair patients (Table 3). Patients who underwent valve replacement for either NVE or PVE were more likely than those who underwent repair to have concomitant tricuspid valve repair (69/316 [22%]) versus (9/131 [6.9%]).
TABLE 3.
Operative conduct, postoperative complications, and echocardiographic evaluation of surgery for mitral valve infective endocarditis
| Variable | NVE standard repair (n = 88) | NVE extended repair (n = 43) | NVE replacement (n = 195) | PVE re-replacement (n = 121) | ||||
|---|---|---|---|---|---|---|---|---|
|
|
|
|
|
|||||
| n* | Value | n* | Value | n* | Value | n* | Value | |
| Operative conduct | ||||||||
| Concomitant procedure | 88 | 18 (20) | 43 | 10 (23) | 195 | 87 (45) | 121 | 38 (31) |
| Myocardial ischemic time, min | 88 | 66 (47, 101) | 43 | 93 (60, 118) | 193 | 84 (56, 132) | 113 | 96 (68, 129) |
| CPB time, min | 87 | 85 (60, 123) | 43 | 109 (73, 148) | 195 | 109 (73, 155) | 120 | 123 (89, 182) |
| Postoperative intra-aortic balloon pump | 88 | 0 (0) | 43 | 0 (0) | 195 | 0 (0) | 121 | 2 (1.7) |
| Postoperative complications | ||||||||
| Hospital death | 88 | 0 (0) | 43 | 3 (7.0) | 195 | 8 (4.1) | 121 | 6 (5.0) |
| Sepsis | 87 | 1 (1.1) | 43 | 5 (12) | 195 | 9 (4.6) | 120 | 9 (7.5) |
| Stroke | 88 | 8 (9.1) | 43 | 1 (2.3) | 195 | 13 (6.7) | 120 | 5 (4.2) |
| Reoperation for valve dysfunction | 88 | 0 (0) | 43 | 3 (7.0) | 195 | 1 (0.51) | 120 | 1 (0.83) |
| Reoperation for bleeding/tamponade | 88 | 1 (1.1) | 43 | 0 (0) | 195 | 2 (1.0) | 120 | 2 (1.7) |
| Other noncardiac reoperation | 88 | 2 (2.3) | 43 | 5 (12) | 195 | 19 (9.7) | 120 | 10 (8.3) |
| Renal failure | 88 | 5 (5.7) | 43 | 3 (7.0) | 195 | 13 (6.7) | 120 | 13 (11) |
| New-onset renal failure requiring dialysis | 81 | 2 (2.5) | 34 | 1 (2.9) | 160 | 7 (4.4) | 102 | 5 (4.9) |
| Prolonged ventilation,>24 h | 87 | 6 (6.9) | 42 | 14 (33) | 193 | 66 (34) | 119 | 42 (35) |
| Atrial fibrillation | 76 | 17 (22) | 35 | 9 (26) | 153 | 47 (31) | 76 | 22 (29) |
| Lengths of stay | 88 | 43 | 195 | 121 | ||||
| Preoperative, d | 6 (0, 11) | 4 (0.6, 12) | 7 (2, 14) | 7 (3, 14) | ||||
| Intensive care unit, h | 45 (22, 150) | 65 (26, 476) | 75 (25, 273) | 73 (27, 252) | ||||
| Postoperative, d | 8 (5.0, 17) | 9 (6.0, 31) | 11 (7.0, 21) | 13 (8.0, 26) | ||||
Data are presented as No. (%) or median (15th, 85th percentiles). NVE, Nativevalve endocarditis; PVE, prosthetic valve endocarditis; CPB, cardiopulmonary bypass.
Number of patients with data available.
Early Outcomes
Hospital mortality was 3.8% (n = 17); 0 after standard valve repair, 3 (7.0%) after extended repair, 8 (4.1%) after native valve replacement, and 6 (5.0%) after prosthetic valve re-replacement (Table 3). Of the 17 hospital deaths, 12 were among patients who were receiving preoperative dialysis for acute (n = 1) or chronic (n = 11) preoperative renal failure; another 2 (who died) required postoperative dialysis. In all, 15 patients required new postoperative dialysis. Standard repair patients had less prolonged ventilation and shorter intensive care unit and postoperative stays. Extended repair patients underwent more reoperations for valve dysfunction. Patients who underwent valve replacement were more likely to have longer intensive care unit and postoperative stays.
Reinfections
During follow-up, 37 reinfections occurred; 4 patients had 2 or more (Figure E4, A). Two of the 37 occurred after standard repair, 5 (in 4 patients) after extended repair, 17 (in 14 patients) after native valve replacement, and 13 after prosthetic valve re-replacement. Actuarially (non–risk-adjusted), those who underwent standard repair for NVE had fewer reinfections than other operative groups, which were all similar (Figure 3, A).
FIGURE 3.
Non–risk-adjusted actuarial estimates of time-related events after surgery for isolated mitral valve infective endocarditis and risk-adjusted predicted estimates from random survival forest analyses. Estimates are stratified according to major operative group: standard repair for native valve endocarditis (NVE; blue), extended repair for NVE (green), valve replacement for NVE (yellow), and valve replacement for prosthetic valve endocarditis (PVE; red). In actuarial depictions, each symbol represents an event and vertical bars 68% confidence limits; dotted lines represent patients being followed, but without events as yet. The following table lists the number of patients at risk at periodic intervals. These numbers are for mortality; numbers for reinfection and reintervention are slightly less because these events happen before death. A, Actuarial depiction of first reinfection. B, Risk-adjusted partial dependency plot of predicted first reinfection. C, Actuarial depiction of first reintervention for a noninfective indication. D, Risk-adjusted partial dependency plot of predicted first reintervention for a noninfective indication. E, Actuarial survival. F, Risk-adjusted survival.
Bacteriology of 13/37 (35%) reinfections was concordant with the index hospitalization (Table E2). Among 14 reinfections in 13 PWIDs, 5 (36%) were concordant infections. Staphylococcus aureus was the most common bacterial isolate among reinfections (14/37 [38%]), followed by Streptococcus viridans (6/37 [16%]). Concordant infections were commonly treated with reintervention (9/13; 69%) and discordant ones with medical management (13/21; 62%).
The most important risk factor for first reinfection was more recent date of surgery, particularly since 2012 (Figures E5, A and E6, A). Other risk factors included smoking history (Figure E6, B), younger age (Figure E6, C), and elevated bilirubin level (Figure E6, D). Risk-adjusted reinfection for the 4 operative groups showed similar occurrence (Figure 3, B). Although the variable identifying PWID was not populated sufficiently to be included in the machine-learning analysis, 14 of 37 (38%) reinfections occurred in PWIDs.
Fifty-one percent of patients with reinfections (19/37) were managed medically and 49% (18/37) underwent reintervention (Figure E7). Extended repair was associated with more early reinterventions for reinfections than any other group.
Reinterventions for Noninfectious Indications
Twenty-seven mitral valve reinterventions were for noninfectious indications; 5 patients had 2 or more (Figure E4, B). Five occurred after standard and 5 after extended repair, all but 1 for failed repair (Table E3).
Like reinfection, the extended repair group exhibited early risk of reintervention (Figure 3, C). However, unlike reinfections, advanced endocarditis features were not important predictors (Figure E5, B). Risk-adjusted predicted reintervention for noninfectious indications according to operative group was similar (Figure 3, D).
Mitral Regurgitation and Gradient After Mitral Valve Repair
Mitral regurgitation after valve repair for NVE tended to occur early postoperatively, but regurgitation grade was higher after extended than standard repair (Figure E8). Similarly, mitral valve gradient was slightly higher after extended repair (Figure E9).
Survival
There were 188 deaths during follow-up. Ten-year survival was 67% after standard repair, 54% after extended repair, 47% after native valve replacement, and 41% after prosthetic valve re-replacement (Figure 3, E).
Reinfection, particularly when medically managed, and reintervention for noninfectious indications were associated with high early risk of mortality (Figure E10).
Landmark analysis showed excess mortality among extended-repair patients compared with standard-repair patients in the first 1.5 years (Figure E11). Of 9 deaths during the early interval, 6 occurred in patients who were receiving dialysis (1 early repair failure, 1 reinfection, 3 unrelated cause, and 1 unknown cause). Acute respiratory distress syndrome caused 1 early death, and 2 (1 in a PWID) were of unknown cause. Mortality after extended repair was similar to that after valve replacement for NVE or PVE (Figure 3, E).
Preoperative dialysis was the most important risk factor for mortality, irrespective of valve repair or replacement (Figures 4 and E5, C). Chronic obstructive pulmonary disease, diabetes mellitus, and older age were also risk factors. When stratified according to infectious etiology, survival was lowest among patients with fungal endocarditis (Figure E12). No survival difference was related to how the valve was managed after adjustment for patient and disease factors (Figure 3, F).
FIGURE 4.
Unadjusted survival after isolated mitral valve repair or replacement for isolated mitral valve endocarditis among patients not receiving renal dialysis (black) and those receiving renal dialysis (light red). Each symbol represents a death, and vertical bars are 68% confidence limits. Numbers below horizontal axis are number of patients remaining at risk. preop, Preoperative.
DISCUSSION
Principal Findings
Surgery for isolated mitral valve endocarditis, presenting in a heterogeneous population and performed by surgeons attempting to select the right operation for the right patient, can have low operative mortality. The apparent outcomes superiority of standard valve repair versus replacement is related to patient and pathology characteristics. In patients with more comorbidities and extensive valve destruction, however, extended repair has worse outcomes, similar to valve replacement. Renal failure is the most powerful factor determining outcome after surgery for NVE and PVE.
Repair or Replace the Mitral Valve for NVE?
No one questions superiority of repair when pathology allows use of standard repair techniques. It seems equally clear that when repair is more complex, its advantage over replacement diminishes. In the presence of comorbid disease and cardiac dysfunction, added tissue destruction makes valve reconstruction and restoration of cusp and anulus integrity technically demanding and intraoperative decision-making challenging. Encountered in an acute setting with potential for rapid clinical deterioration and end-organ dysfunction, these pathologic elements have historically translated into decreased valve reparability and worse patient outcomes compared with mitral disease of other etiologies.2
Our proportion of valve repair for NVE is lower than that recently reported.15,16 However, those studies included both remote and active endocarditis and endocarditis involving the aortic valve (typically easily patched smaller perforations [kissing lesions] in the anterior mitral valve cusp). All of our patients had active endocarditis, and all operations were performed by surgeons experienced in mitral valve repair, performing more than 25 repairs per year.22 Inspired by reports of higher repair proportions,14,16 our repairs increased in recent years. There was an association between worse pathology and worse patient characteristics explaining our worse outcomes with extended repairs, suggesting we might have tried too hard in some cases for no benefit.
Risk of Reinfection
Low risk of reinfection has been a strong argument in favor of valve repair over replacement.15 With a generally low occurrence of reinfection, we too observed fewer reinfections after valve repair for NVE than replacement, but this was only true for standard repair; extended repair was associated with an equally high risk of reinfection as valve replacement. PVE patients had the highest risk.
Although the variable related to addiction was populated insufficiently to include in multivariable analysis, one-third of reinfections occurred in PWIDs. Other important predictors—more recent date of surgery, particularly since 2012, smoking history, and younger age—support the effect of lifestyle. The significance of preoperative bilirubin is supported by recent studies demonstrating its usefulness as a predictor of adverse events, particularly mortality, in the endocarditis setting.23,24 In the setting of bacterial infections, hyperbilirubinemia might reflect more severe infections that might predispose patients to greater reinfection risk.
The organism causing reinfections was concordant with the original infection in just over a third of patients, including PWIDs. Concordant infections were more likely to be treated surgically than discordant ones.
Risk of Reintervention
Half of the reinterventions in our study were due to reinfection, with extended repair patients again experiencing the highest risk and, accordingly, number of cusps requiring patching and anterior cusp patching were predictive of reintervention.
When it came to reinterventions for noninfectious indications, extended repair patients had the highest risk. Other studies have emphasized that valve repair requiring pericardial patch reconstruction is a surrogate for worse valve pathology requiring complex repair.16,25 Questions surrounding the durability of patch repair compared with other valve repair techniques for endocarditis have also previously been raised.26–28
Recurrent Mitral Regurgitation and Gradients After Valve Repair for NVE
Our longitudinal echocardiography data showed an early progression of mitral regurgitation after valve repair, with extended repair again showing earlier and more regurgitation. In addition, extended repair was associated with marginally higher postoperative mean gradient, with 1 patient requiring reoperation for stenosis.
Prosthetic Mitral Valve Endocarditis
There is no alternative to valve re-replacement in patients with PVE. These patients are sicker, with more invasive disease, although mitral PVE is less invasive than aortic valve PVE.1,29 PVE patients differed in many respects from those undergoing valve replacement for NVE. They had more strokes, comorbidities, and circumferential extent of disease, whereas NVE patients had more cusp destruction and severe regurgitation.
Although reinfection was highest after PVE, reintervention after replacement for NVE and PVE was similar for infectious and noninfectious indications. Only 1 case of periprosthetic leakage with PVE occurred in the present study. We have shown in previous studies that periprosthetic leakage, even without recurrent infection, is a predictor of late mortality.30,31 However, questions persist concerning whether periprosthetic leakage is a sequela of the endocarditis healing process or a manifestation of reinfection. Shah and colleagues32 studied periprosthetic leaks and reported their relation to PVE, mitral position, and same-valve infection, and their association with worse outcomes.
Implications for Survival
Our standard repair patients had less advanced disease and were healthier than those who underwent replacement for NVE; correspondingly, there was no hospital mortality in that group, whereas deaths occurred in the other groups. Extended repair and valve replacement patients had more destructive and invasive disease and were sicker. Their higher mortality and worse survival were explained by patient and disease characteristics.
Preoperative dialysis and renal failure were the most important predictors of early and late mortality. Survival in patients receiving dialysis who underwent infective endocarditis surgery has been historically poor, reported at 60%, 38%, 25%, and 16% at 1, 3, 5, and 7 years in our 2017 publication.33 The benefit of surgery over nonoperative management remains unclear and is an incentive to identify factors associated with poor outcomes in this population.33
In patients with advanced invasive infective endocarditis, those with aortic valve endocarditis have technical advantages. After debridement, the root can be reconstructed with an infection-resistant allograft,34 exteriorizing the infected area and allowing drainage to the pericardium and better antimicrobial access to residual organisms. A mitral anular abscess can be cleaned out but not easily exteriorized and must be sealed off, allowing less adequate penetration of antimicrobials. Hemodynamic differences between aortic and mitral endocarditis offer an explanation of observed differences in invasiveness and clinical manifestations, because invasion is promoted by higher pressure.29 From a pathophysiological standpoint, severe acute mitral regurgitation directly floods the lungs, whereas with acute aortic regurgitation, a well preserved mitral valve protects the lungs until the left ventricle fails. Although this is not a direct comparison of aortic and mitral valve endocarditis, our outcomes suggest we have become better at managing patients with mitral valve endocarditis, and current outcomes may no longer be worse than for patients with aortic valve endocarditis, except perhaps for patients receiving dialysis.1–3
Clinical Implications of This Study
The equation that guides an individual surgeon’s choice of repair versus replacement for NVE must take into account all factors, including valve pathology, the surgeon’s ability and experience, and patient factors such as surgical risk, need for an expeditious and reliable operation, and life expectancy. That valve management was not a risk factor for outcomes is not evidence that it is irrelevant, but rather suggests that our surgeons’ judgment and choices were adequate and good, possibly with the exception of use of extended repair. For each patient presenting with mitral valve endocarditis, the surgeon must neutralize the risk factors as much as possible to decrease risks associated with each procedure. This is accomplished by choosing the most suitable surgical strategy specifically tailored to each unique patient and thus achieving the best short- and long-term results.
Strengths and Limitations
Although this is a single-center observational study of 447 operations for isolated mitral valve infective endocarditis, it might be the largest to report patient, microbiological, pathological, and surgical factors in the context of outcomes specific to the mitral valve. Coding of the pathologic entity was based on all available information and performed according to a uniform template by a surgeon, usually not the operating surgeon. The surgeons’ description of the pathology was less complete in patients who underwent replacement.
As an observational study, referral, selection, and institutional factors influence these findings, limiting generalizability. Although most of these patients resided within 100 miles of our hospital or were Ohio residents, being a high-volume referral center, a substantial number came from other states or countries. As a result, many had advanced disease and were referred for high-risk surgery and had been receiving antimicrobial therapy of variable duration before their arrival. For risk-adjusted comparison, we used machine learning with 70 variables to identify important predictors of outcomes. Data on addiction are incomplete, available in only half the patients, but in a previous study,35 only 6% of patients undergoing operation for isolated mitral valve endocarditis had a history of intravenous drug use versus 20% of all patients with intravenous drug use–related infective endocarditis.
CONCLUSIONS
Using a patient- and pathology-tailored approach—the right operation for the right patient at the right time—surgery for isolated mitral valve endocarditis can be performed with low mortality and excellent outcomes. Apparent inferiority of any approach except standard valve repair relates to patient characteristics and pathological progression of the disease. Dialysis and renal failure, along with other patient comorbidities and progression and disease sequelae, are powerful determinants of short- and long-term outcomes. In patients with more comorbidities and extensive valve destruction, extended repair may have no advantage over valve replacement.
Supplementary Material
CENTRAL MESSAGE.
A tailored approach to isolated mitral endocarditis surgery has excellent outcomes determined by patient characteristics, with dialysis predicting the worst outcomes.
PERSPECTIVE.
A patient- and pathology-tailored approach to surgery for isolated mitral valve endocarditis has low mortality and excellent results. Superiority of standard valve repair is related to patient characteristics and pathology. Dialysis and renal failure are the most powerful factors determining outcomes. In case of extensive destruction, extended valve repair shows no benefit over replacement.
Acknowledgments
See the Acknowledgments section for Endocarditis Study Group members.
This study was funded in part by the Peter and Elizabeth C. Tower and Family Endowed Chair in Cardiothoracic Research, James and Sharon Kennedy, the Slosburg Family Charitable Trust, the Stephen and Saundra Spencer Fund for Cardiothoracic Research, and the Drs Sidney and Becca Fleischer Heart and Vascular Education Chair. Dr Moore is a National Heart, Lung, and Blood Institute Clinical Research Scholar of the Cardiothoracic Surgical Trials Network whose MS degree is funded in part by the National Institutes of Health grant HL088955.
Abbreviations and Acronyms
- NVE
native valve endocarditis
- PVE
prosthetic valve endocarditis
- PWIDs
persons who inject drugs
Discussion
Presenter: Dr Ryan Moore
Dr James Davies (Birmingham, Ala). Thank you very much for asking me to review the paper, Dr Moore. Thank you for getting me the paper in a timely fashion. A couple of questions. One, after last discussion, I didn’t see anything in that paper about the use of IV drugs in this patient group. If you could give one comment on that.
Two, you discussed about the elevated level of renal dysfunction or dialysis. Do you have any specific numbers that you look at for the patients? Whether it’s a GFR or creatinine levels specifically that puts them at elevated risk, not just the dialysis patients.
And one final question is you describing invasive pathology, aggressive or invasive pathology, for the type of bacterial infection. You have more fungal infections in the replacement group versus the repair group. Can you comment about out that or should you eliminate the fungal group out of this? Because in our hands, the fungal endocarditis specifically is somewhat different and those patients tend to do worse.
Dr Ryan Moore (Cleveland, Ohio). Thank you, Dr Davies, for reviewing the paper and moderating this discussion.
The first question was about specific levels related to the findings of creatinine, etc. We found that a creatinine level that approached 1.2 was where it seemed to take off in terms of the decrease in survival. We didn’t go much further than that. We have some partial plots included in our study that depict this, but we didn’t quantify the levels themselves to see whether a specific level served as a threshold that importantly changed outcomes for patients.
I think the finding of dialysis being a poor prognostic indicator is supported by other studies we’ve seen. An overarching theme of our study is that the decision to repair or replace the valve did not turn out to be as important as we thought heading into the study. What turned out to be more important was that these patients’ baseline characteristics seem to be the primary driver of outcomes, which has made us reassess our approach to decision making in the operating room regarding whether to repair or replace the valve. And how that actually changes the patient’s postoperative trajectory.
Dr Davies. And regarding IV drug use?
Dr Moore. Yes, we do not have a lot of IV drug use in our population. I don’t have the numbers offhand, I’m sorry. But when we eliminated patients who had double valve endocarditis and got down to the isolated numbers, it was less than 1 or 2 percent of patients with IV drug use in our population.
Dr Davies. If your patient population has elevated creatinine level of 1.2—this significant level—you’ve got a very healthy group of patients that you’re operating on. Our level seems to be quite a bit higher than that. The last question was about fungal endocarditis, whether grouping it in, or whether it should be separated.
Dr Moore. We haven’t discussed how the fungal population affects our outcomes, but that’s something we should definitely discuss moving forward. Because as you see, that is a population that displays some different characteristics that can influence how we interpret our outcomes. So, I think that’s something we will discuss, but we have not discussed it extensively today.
Footnotes
Webcast
You can watch a Webcast of this AATS meeting presentation by going to: https://aats.blob.core.windows.net/media/Publications/AM21_A34%20-%20Endocarditis.mp4-PresPlusDisc.mp4.
Scanning this QR code will take you to the table of contents to access supplementary information. To view the AATS Annual Meeting Webcast, see the URL next to the webcast thumbnail.
Conflict of Interest Statement
Dr Wierup disclosed financial relationships with Edwards Lifesciences, Medtronic, and CryoLife. Dr Blackstone is a consultant for clinical trials at Edwards Lifesciences and Abiomed. All other authors reported no conflicts of interest.
The Journal policy requires editors and reviewers to disclose conflicts of interest and to decline handling or reviewing manuscripts for which they may have a conflict of interest. The editors and reviewers of this article have no conflicts of interest.
References
- 1.Hussain ST, Shrestha NK, Gordon SM, Houghtaling PL, Blackstone EH, Pettersson GB. Residual patient, anatomic, and surgical obstacles in treating active left-sided infective endocarditis. J Thorac Cardiovasc Surg. 2014;148:981–8. [DOI] [PubMed] [Google Scholar]
- 2.Said SM, Abdelsattar ZM, Schaff HV, Greason KL, Daly RC, Pochettino A, et al. Outcomes of surgery for infective endocarditis: a single-centre experience of 801 patients. Eur J Cardiothorac Surg. 2018;53:435–9. [DOI] [PubMed] [Google Scholar]
- 3.Perrotta S, Jeppsson A, Frojd V, Svensson G. Surgical treatment for infective endocarditis: a single-centre experience. Thorac Cardiovasc Surg. 2017;65:166–73. [DOI] [PubMed] [Google Scholar]
- 4.Pettersson GB, Coselli JS, Hussain ST, Griffin B, Blackstone EH, Gordon SM, et al. 2016 American Association for Thoracic Surgery (AATS) consensus guidelines: surgical treatment of infective endocarditis. Executive summary. J Thorac Cardiovasc Surg. 2017;153:1241–58. [DOI] [PubMed] [Google Scholar]
- 5.Baddour LM, Wilson WR, Bayer AS, Fowler VG Jr, Tleyjeh IM, Rybak MJ, et al. Infective endocarditis in adults: diagnosis, antimicrobial therapy, and management of complications: a scientific statement for healthcare professionals from the American Heart Association. Circulation. 2015;132:1435–86. [DOI] [PubMed] [Google Scholar]
- 6.Habib G, Lancellotti P, Antunes MJ, Bongiorni MG, Casalta JP, Del Zotti F. 2015 ESC guidelines for the management of infective endocarditis: the task force for the management of infective endocarditis of the European Society of Cardiology (ESC). Eur Heart J. 2015;36:3075–128. [DOI] [PubMed] [Google Scholar]
- 7.Nishimura RA, Otto CM, Bonow RO, Carabello BA, Erwin JP III, Guyton RA, et al. 2014 AHA/ACC guideline for the management of patients with valvular heart disease: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines. Circulation. 2014;129:2440–92. [DOI] [PubMed] [Google Scholar]
- 8.Kang DH, Kim YJ, Kim SH, Sun BJ, Kim DH, Yun SC, et al. Early surgery versus conventional treatment for infective endocarditis. N Engl J Med. 2012;366:2466–73. [DOI] [PubMed] [Google Scholar]
- 9.Muehrcke DD, Cosgrove DM III, Lytle BW, Taylor PC, Burgar AM, Durnwald CP, et al. Is there an advantage to repairing infected mitral valves? Ann Thorac Surg. 1997;63:1718–24. [DOI] [PubMed] [Google Scholar]
- 10.Pagani FD, Monaghan HL, Deeb GM, Bolling SF. Mitral valve reconstruction for active and healed endocarditis. Circulation. 1996;94:II133–8. [PubMed] [Google Scholar]
- 11.Hendren WG, Morris AS, Rosenkranz ER, Lytle BW, Taylor PC, Stewart WJ, et al. Mitral valve repair for bacterial endocarditis. J Thorac Cardiovasc Surg. 1992;103:124–9. [PubMed] [Google Scholar]
- 12.Toyoda N, Itagaki S, Egorova NN, Tannous H, Anyanwu AC, El-Eshmawi A, et al. Real-world outcomes of surgery for native mitral valve endocarditis. J Thorac Cardiovasc Surg. 2017;154:1906–12. [DOI] [PubMed] [Google Scholar]
- 13.Ruttmann E, Legit C, Poelzl G, Mueller S, Chevtchik O, Cottogni M, et al. Mitral valve repair provides improved outcome over replacement in active infective endocarditis. J Thorac Cardiovasc Surg. 2005;130:765–71. [DOI] [PubMed] [Google Scholar]
- 14.Perrotta S, Frojd V, Lepore V, Schersten H, Jeppsson A, Svensson G. Surgical treatment for isolated mitral valve endocarditis: a 16-year single-centre experience. Eur J Cardiothorac Surg. 2018;53:576–81. [DOI] [PubMed] [Google Scholar]
- 15.Solari S, De Kerchove L, Tamer S, Aphram G, Baert J, Borsellino S, et al. Active infective mitral valve endocarditis: is a repair-oriented surgery safe and durable? Eur J Cardiothorac Surg. 2019;55:256–62. [DOI] [PubMed] [Google Scholar]
- 16.de Kerchove L, Price J, Tamer S, Glineur D, Momeni M, Noirhomme P, et al. Extending the scope of mitral valve repair in active endocarditis. J Thorac Cardiovasc Surg. 2012;143:S91–5. [DOI] [PubMed] [Google Scholar]
- 17.Feringa HH, Shaw LJ, Poldermans D, Hoeks S, van der Wall EE, Dion RA, et al. Mitral valve repair and replacement in endocarditis: a systematic review of literature. Ann Thorac Surg. 2007;83:564–70. [DOI] [PubMed] [Google Scholar]
- 18.Elgharably H, Hussain ST, Shrestha NK, Blackstone EH, Pettersson GB. Current hypotheses in cardiac surgery: biofilm in infective endocarditis. Semin Thorac Cardiovasc Surg. 2016;28:56–9. [DOI] [PubMed] [Google Scholar]
- 19.Afonso L, Kottam A, Reddy V, Penumetcha A. Echocardiography in infective endocarditis: state of the art. Curr Cardiol Rep. 2017;19:127. [DOI] [PubMed] [Google Scholar]
- 20.Pettersson GB, Hussain ST, Shrestha NK, Gordon S, Fraser TG, Ibrahim KS, et al. Infective endocarditis: an atlas of disease progression for describing, staging, coding, and understanding the pathology. J Thorac Cardiovasc Surg. 2014; 147:1142–9.e2. [DOI] [PubMed] [Google Scholar]
- 21.Rajeswaran J, Blackstone EH. A multiphase non-linear mixed effects model: an application to spirometry after lung transplantation. Stat Methods Med Res. 2017; 26:21–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chikwe J, Toyoda N, Anyanwu AC, Itagaki S, Egorova NN, Boateng P, et al. Relation of mitral valve surgery volume to repair rate, durability, and survival. J Am Coll Cardiol. 2017;69:2397–406. 10.1016/j.jacc.2017.02.026 [DOI] [PubMed] [Google Scholar]
- 23.Diab M, Sponholz C, von Loeffelholz C, Scheffel P, Bauer M, Kortgen A, et al. Impact of perioperative liver dysfunction on in-hospital mortality and long-term survival in infective endocarditis patients. Infection. 2017;45:857–66. [DOI] [PubMed] [Google Scholar]
- 24.Wei XB, Wang Y, Liu YH, Huang JL, Yu DQ, Chen JY. Effect of conjugated bilirubin on clinical outcomes in infective endocarditis. Eur J Clin Microbiol Infect Dis. 2019;38:2259–66. [DOI] [PubMed] [Google Scholar]
- 25.Pettersson GB, Hussain ST, Ramankutty RM, Lytle BW, Blackstone EH. Reconstruction of fibrous skeleton: technique, pitfalls and results. Multimed Man Cardiothorac Surg. 2014;2014:mmu004. [DOI] [PubMed] [Google Scholar]
- 26.Evans CF, DeFilippi CR, Shang E, Griffith BP, Gammie JS. Fresh autologous pericardium for leaflet perforation repair in mitral valve infective endocarditis. J Heart Valve Dis. 2013;22:560–6. [PubMed] [Google Scholar]
- 27.Fukunaga N, Sakata R, Koyama T. Reoperative analysis after mitral valve repair with glutaraldehyde-treated autologous pericardium. InteractCardiovascThorac Surg. 2017;25:912–7. [DOI] [PubMed] [Google Scholar]
- 28.Schoen FJ, Levy RJ. Calcification of tissue heart valve substitutes: progress toward understanding and prevention. Ann Thorac Surg. 2005;79:1072–80. [DOI] [PubMed] [Google Scholar]
- 29.Hussain ST, Shrestha NK, Witten J, Gordon SM, Houghtaling PL, Tingleff J, et al. Rarity of invasiveness in right-sided infective endocarditis. J Thorac Cardiovasc Surg. 2018;155:54–61. [DOI] [PubMed] [Google Scholar]
- 30.Blackstone EH, Kirklin JW. Death and other time-related events after valve replacement. Circulation. 1985;72:753–67. [DOI] [PubMed] [Google Scholar]
- 31.Dhasmana JP, Blackstone EH, Kirklin JW, Kouchoukos NT. Factors associated with periprosthetic leakage following primary mitral valve replacement: with special consideration of the suture technique. Ann Thorac Surg. 1983;35:170–8. [DOI] [PubMed] [Google Scholar]
- 32.Shah S, Alashi A, Pettersson GB, Rodriguez LL, Gillinov AM, Grimm RA, et al. Characteristics and longer-term outcomes of paravalvular leak after aortic and mitral valve surgery. J Thorac Cardiovasc Surg. 2019;157:1785–92. [DOI] [PubMed] [Google Scholar]
- 33.Raza S, Hussain ST, Rajeswaran J, Ansari A, Trezzi M, Arafat A, et al. Value of surgery for infective endocarditis in dialysis patients. J Thorac Cardiovasc Surg. 2017;154:61–70. [DOI] [PubMed] [Google Scholar]
- 34.Witten JC, Houghtaling PL, Shrestha NK, Gordon SM, Jaber W, Blackstone EH,et al. Aortic allograft infection risk. J Thorac Cardiovasc Surg. Published online May 12, 2021 [Epub ahead of print]. 10.1016/j.jtcvs.2021.04.086 [DOI] [PubMed] [Google Scholar]
- 35.Shrestha NK, Jue J, Hussain ST, Jerry JM, Pettersson GB, Menon V, et al. Injection drug use and outcomes after surgical intervention for infective endocarditis. Ann Thorac Surg. 2015;100:875–82. [DOI] [PubMed] [Google Scholar]
- 36.Rajeswaran J, Blackstone EH. Competing risks: competing questions. J Thorac Cardiovasc Surg. 2017;153:1432–3. 10.1016/j.jtcvs.2017.03.017 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.