- Email: [email protected]
Mitral Valve Repair Versus Replacement in Elderly With Degenerative Disease: Analysis of the STS Adult Cardiac Surgery Database
Mitral Valve Repair Versus Replacement in Elderly With Degenerative Disease: Analysis of the STS Adult Cardiac Surgery Database Ryan J. Hendrix, MD, Ricardo A. Bello, MD, PhD, Julie M. Flahive, MS, Nikolaos Kakouros, MD, PhD, Gerard P. Aurigemma, MD, John F. Keaney, MD, William Hoffman, MD, and Christina M. Vassileva, MD Division of Cardiac Surgery, Department of Surgery, University of Massachusetts Medical School, Worcester; Department of Quantitative Health Sciences, University of Massachusetts Medical School, Worcester; Division of Cardiovascular Medicine, Department of Medicine, University of Massachusetts Medical School, Worcester; and Department of Anesthesiology, University of Massachusetts Medical School, Worcester, Massachusetts
Background. It has been postulated that mitral valve repair in the elderly does not confer short-term benefits over mitral valve replacement with complete preservation of the chordal apparatus. Our purpose was to test this hypothesis using data from The Society of Thoracic Surgeons Adult Cardiac Surgery Database (STS ACSD). Methods. Patients aged 70 years or more undergoing primary isolated elective mitral valve repair or mitral valve replacement for degenerative disease were obtained from the STS ACSD versions 2.73 and 2.81. Patients with a concomitant tricuspid procedure, atrial fibrillation surgery, or atrial septal defect/patent foramen ovale repair were included. The two treatment groups were further stratified by age in years (70 to 74, 75 to 79, and 80 or more). Adjusted 30-day mortality rates were analyzed by mitral procedure and chordal preservation strategy. Results. The study included 12,043 patients, of whom 71% underwent mitral valve repair. Observed 30-day mortality after repair was lower than after replacement (2.2% versus 4.8%, respectively; p < 0.0001). Using repair
as reference, adjusted operative mortality was higher for replacement in the overall cohort (odds ratio 1.83, 95% confidence interval: 1.45 to 2.31). There was no significant difference in mortality between complete versus partial chordal preservation in repair (odds ratio 1.24, 95% confidence interval: 0.80 to 1.93). Mitral valve replacement with chordal preservation remained inferior to repair (odds ratio 1.66, 95% confidence interval: 1.28 to 2.14). The failed repair rate was 7.9%, with a 30-day mortality of 6%. Conclusions. In patients aged 70 years or more, degenerative mitral repair was associated with lower operative mortality compared with replacement, irrespective of chordal preservation strategy. Failed repairs reduced this short-term benefit compared with chordalsparing replacement as evidenced by the similar operative mortality on an intention to treat analysis.
C
primary mitral regurgitation (MR), the rate of MVP increased from 42% to 61% [5]. Despite improvements, these national trends fall far short of the experiences reported by high-volume centers, and implementation of MVP in the care of the elderly population has been less uniformly accepted [6, 7]. The benefits of preserving the chordal apparatus in mitral valve replacement (MVR) have been shown in a number of studies [8, 9], including a randomized controlled trial in 2002 [10]. Despite robust retrospective studies comparing mitral repair to replacement, many of
urrent American Heart Association/American College of Cardiology guidelines recommend mitral valve repair (MVP) over replacement for degenerative mitral valve disease. The advantages of MVP compared with replacement in this population have been well established [1–4]. Accordingly, the utilization of MVP has steadily increased. In an 8-year review (2000 to 2007) of The Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database (ACSD), in patients with isolated Accepted for publication Sept 6, 2018. Presented at the Fifty-fourth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 27–31, 2018. Address correspondence to Dr Vassileva, Division of Cardiac Surgery, Department of Surgery, University of Massachusetts Medical School, 55 Lake Ave N, Worcester, MA 01655; email: mitralrepair@gmail. com.
Ó 2019 by The Society of Thoracic Surgeons Published by Elsevier Inc.
(Ann Thorac Surg 2019;-:-–-) Ó 2019 by The Society of Thoracic Surgeons
The Supplemental Tables can be viewed in the online version of this article [https://doi.org/10.1016/ j.athoracsur.2018.09.018] on http://www.annalsthoracic surgery.org.
0003-4975/$36.00 https://doi.org/10.1016/j.athoracsur.2018.09.018
2
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
these did not include information on the chordal preservation strategy, if any [2, 11–16]. Furthermore, no prospective randomized clinical trials comparing outcomes of mitral repair versus replacement in patients with degenerative MR have been performed. As a result, it remains unclear whether MVP offers short-term benefits over complete chordal-sparing MVR, particularly in the elderly population. The purpose of this study was twofold. Our first aim was to evaluate the operative morbidity and short-term mortality associated with MVP and MVR for degenerative MR in the elderly population to assess whether the benefit of repair over replacement diminishes with age. Our second aim was to assess 30-day outcomes after mitral repair in the elderly compared with mitral replacement according to the chordal preservation strategy utilized.
Ann Thorac Surg 2019;-:-–-
was documented in 10,050 patients (83%) and therefore not incorporated into the definition of a failed repair; 668 patients (14%) did not undergo a postprocedural TEE, and this data field was not specified for the remaining 325 patients (3%). Mortality was defined as operative mortality or any death within 30 days after surgery in accordance with the standard STS definition. All unknown values for variables were set to missing. Missing variables were generally less than 0.3%.
Statistical Analysis
The STS ACSD versions 2.73 and 2.81 were queried from July 1, 2011, to June 30, 2016. Demographics, socioeconomic factors, chronic conditions, operative characteristics, and 30-day morbidity and mortality were included in the provider user files. The study was approved by the Institutional Review Board of the University of Massachusetts Medical School. Informed consent was not required given the retrospective nature of the review and the deidentified nature of the records.
Percentages were reported for categorical variables. For normally distributed continuous variables, mean and standard deviation were reported with p values from Student’s t tests. For skewed continuous variables, median and interquartile range were reported along with p values from the Wilcoxon rank sum test. The p values were considered statistically significant at an alpha level of 0.05. Logistic regression was used to model the odds ratio (OR) of mortality. The models were adjusted for the following variables: age, sex, body mass index, diabetes mellitus, last hematocrit, last creatinine, dyslipidemia, hypertension, chronic lung disease, liver disease, immunocompromise, peripheral vascular disease, cerebrovascular disease, heart failure, ejection fraction, atrial fibrillation, tricuspid insufficiency, preoperative intraaortic balloon pump, and previous myocardial infarction. The covariates included in the final adjusted models were clinically relevant variables and have been well validated in prior reports [5]. All analyses were performed using SAS software (version 9.4; SAS Institute, Cary, NC).
Patient Cohort
Results
All patients aged 70 years or more who were undergoing an elective primary isolated mitral valve operation for degenerative disease were considered for inclusion in the study. Under this definition, patients with concomitant ablation for atrial fibrillation, tricuspid repair or replacement, and closure of a patent foramen ovale or an atrial septal defect were included. Exclusion criteria were defined as a preoperative unresponsive neurologic state, myocardial infarction within 21 days of surgery, cardiogenic shock, endocarditis, mitral disease etiology other than degenerative, mitral stenosis, unplanned mitral valve procedure, urgent, emergent, or salvage procedures, and transcatheter mitral procedures.
Baseline Characteristics
Patients and Methods Data Sources
Data Variables All variables were adopted from the STS forms versions 2.73 and 2.81. Specific variable definitions are described as follows. Chordal preservation was subdivided into complete (preservation of both anterior and posterior leaflets), partial (preservation of posterior or anterior leaflet), and none. For 337 (9.6%) MVR procedures, the chordal approach was not defined, and these patients were excluded from regression analyses to estimate operative mortality. A failed mitral repair was defined as an unsuccessful repair requiring mitral replacement during the same operation. Performance of an intraoperative postprocedural transesophageal echocardiogram (TEE)
In all, 12,043 patients were included in the study, of whom 8,523 (71%) underwent MVP and 3,520 (29%) underwent MVR. Mean age for the entire cohort was 77 years. There was a significant disease burden in patients undergoing both mitral repair and replacement (Table 1, Supplemental Tables 1–4).
Intraoperative Variables Detailed operative characteristics are presented in Table 2. Median duration of cardiopulmonary bypass and median aortic cross-clamp time were significantly lower for patients undergoing repair compared with replacement (115 minutes versus 122 minutes, p < 0.0001, and 83 minutes versus 90 minutes, p < 0.0001, respectively). When these data were stratified by age, both cardiopulmonary bypass time and aortic cross-clamp time remained the same or decreased with age, and both were consistently lower for repair compared with replacement in all age groups. Overall, 83% of the patients underwent intraoperative postprocedural TEE. The majority of patients had an excellent result; however, 4.0% of patients had documented moderate/severe mitral regurgitation. Procedural details are provided in Table 3. The mitral repair failure rate was 7.9%. Among patients with failed mitral repair, median cardiopulmonary bypass time was 160 minutes (interquartile range: 121 to 210), and median
Ann Thorac Surg 2019;-:-–-
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
3
Table 1. Baseline Characteristics of Study Cohort Total Cohort (N ¼ 12,043) Characteristics Age, years, mean (SD) Male White Body mass index, mean (SD) Diabetes mellitus Last hematocrit, mean (SD) Last creatinine, mean (SD) Dyslipidemia Hypertension Chronic lung disease Liver disease Immunocompromised Peripheral vascular disease Cerebrovascular disease History of heart failure EF, median (IQR) Atrial fibrillation Tricuspid insufficiency Preoperative IABP Previous myocardial infarction
Overall (N ¼ 12,043)
MVP (n ¼ 8,523)
MVR (n ¼ 3,520)
p Value
77 (4.8) 5,587 (46) 11,054 (92) 27 (12) 1,737 (14) 40 (4.6) 1.0 (0.50) 7,627 (63) 9,134 (76) 903 (7.6) 249 (2.1) 289 (2.4) 661 (5.5) 1,349 (11) 6,907 (58) 60 (55, 64) 5,674 (47) 4,106 (35) 24 (0.20) 706 (5.9)
76 (4.7) 4,163 (49) 7,861 (93) 27 (13) 1,089 (13) 40 (4.4) 1.0 (0.45) 5,311 (62) 6,317 (74) 545 (6.5) 166 (2.0) 187 (2.2) 440 (5.2) 869 (10) 4,722 (56) 60 (55, 65) 3,859 (45) 2,811 (33) 14 (0.16) 459 (5.4)
78 (4.9) 1,424 (40) 3,193 (91) 27 (12) 648 (18) 39 (4.8) 1.1 (0.55) 2,316 (66) 2,817 (80) 358 (10) 83 (2.4) 102 (2.9) 221 (6.3) 480 (14) 2,185 (62) 60 (53, 63) 1,815 (52) 1,295 (37) 10 (0.28) 247 (7.0)
<0.0001 <0.0001 0.002 0.06 <0.0001 <0.0001 <0.0001 0.0005 <0.0001 <0.0001 0.18 0.02 0.02 <0.0001 <0.0001 0.001 <0.0001 <0.0001 0.18 0.001
Values are n (%) unless otherwise indicated. EF ¼ ejection fraction; replacement.
IABP ¼ intraaortic balloon pump;
IQR ¼ interquartile range;
MVP ¼ mitral valve repair;
MVR ¼ mitral valve
Table 2. Operative Characteristics Total Cohort (N ¼ 12,043) Characteristics Median sternotomy Blood product transfusion CPB, minutes, median (IQR) Aortic CCT, minutes, median (IQR) Lowest CPB temp, C, median (IQR) Lowest CPB HCT, mean (SD) Tricuspid procedure Replacement Repair Intraop IABP Atrial fibrillation surgery Intraop postprocedural TEE Postprocedural MR None Trace/trivial Mild Moderate Severe Unknown
Overall (N ¼ 12,043)
MVP (n ¼ 8,523)
MVR (n ¼ 3,520)
p Value
8,728 (73) 4,460 (37) 117 (90–151) 85 (65–111) 33 (31–34) 25 (4.5)
5,798 (68) 971 (28) 115 (88–148) 83 (63–107) 33 (31–34) 25 (4.5)
2,930 (83) 779 (46) 122 (94–160) 90 (69–122) 33 (31–34) 24 (4.3)
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
41 (1.6) 2,528 (98) 141 (1.2) 881 (7.3) 10,050 (83)
17 (0.92) 1,834 (99) 82 (0.96) 634 (7.4) 7,241 (85)
24 (3.3) 694 (97) 59 (1.7) 247 (7) 2,809 (80)
<0.0001
5,802 (58) 2,715 (27) 766 (7.6) 158 (1.6) 234 (2.3) 375 (3.7)
3,927 (54) 2,192 (30) 667 (9.2) 117 (1.6) 138 (1.9) 200 (2.8)
1,875 (67) 523 (19) 99 (3.5) 41 (1.5) 96 (3.4) 175 (6.2)
0.001 0.42 <0.0001 <0.0001
Values are n (%) unless otherwise indicated. CCT ¼ cross-clamp time; CPB ¼ cardiopulmonary bypass; HCT ¼ hematocrit; IABP ¼ intraaortic balloon pump; Intraop ¼ intraoperative; IQR ¼ interquartile range; MR ¼ mitral regurgitation; MVP ¼ mitral valve repair; MVR ¼ mitral valve replacement; temp ¼ temperature; TEE ¼ transesophageal echocardiography.
4
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
Ann Thorac Surg 2019;-:-–-
Table 3. Mitral Procedure Details By STS Version Procedure Details Mitral repair Repair type Annuloplasty Leaflet resection Leaflet type Triangular Quadrangular Other Leaflet plication Leaflet debridement Chordal/leaflet transfer Neochords (PTFE) Mitral commissuroplasty Folding plasty Leaflet extension/replacement/patch Sliding plasty Mitral cleft repair (scallop closure) Edge-to-edge repair Mitral commissurotomy Annular decalcification Failed repair ratea Mitral replacement Chordal preservation None Anterior Posterior Both Unknown Mitral implant Bioprosthetic valve Mechanical valve
Overall
2.73
2.81
8,523 (71)
5,262 (72)
3,261 (70)
8,142 (96) 3,765 (45)
4,982 (95) 2,231 (43)
3,160 (97) 1,534 (47)
2,233 (26) 1,215 (14) 251 (3.0) 276 (8.6) 30 (0.93) 230 (2.7) 1,926 (23) 163 (5.1) 159 (4.9) 105 (1.3) 715 (8.5) 468 (14) 482 (5.7) 87 (1.0) 128 (1.5) 732 (7.9) 3,520 (29)
1,247 (24) 759 (14) 169 (3) ... ... 152 (2.9) 1,005 (19) ... ... 64 (1.2) 449 (8.6) ... 318 (6.1) 65 (1.3) 81 (1.6) 441 (7.7) 2,095 (28)
986 (30) 456 (14) 82 (3.0) 276 (8.6) 30 (0.93) 78 (2.4) 921 (28) 163 (5.1) 159 (4.9) 41 (1.3) 266 (8.2) 468 (14) 164 (5.1) 22 (0.68) 47 (1.5) 291 (8.2) 1425 (30)
572 (16) 126 (3.6) 831 (24) 1,654 (47) 337 (10)
375 (18) 70 (3) 504 (24) 962 (46) 184 (8.8)
197 (14) 56 (3.9) 327 (23) 692 (49) 153 (11)
1,383 (39) 37 (1.1)
... ...
1,383 (39) 37 (1.1)
a
Calculated as failed repairs divided by total repairs plus failed repairs.
Values are n (%) unless otherwise specified. PTFE ¼ polytetrafluoroethylene;
STS ¼ The Society of Thoracic Surgeons.
aortic cross-clamp time was 123 minutes (interquartile range: 91 to 166).
Postoperative Outcomes The postoperative outcomes of the study cohort are presented in Table 4. Patients who underwent repair had fewer postoperative complications, including lower rates of prolonged ventilation (7.2% versus 13%, p < 0.0001), reintubation (3.6% versus 5.8%, p < 0.0001), need for reoperation (3.8% versus 6.4%, p < 0.0001), and need for blood transfusion (32% versus 50%, p < 0.0001). Many of the postoperative complications increased with age for both repair and replacement, but remained consistently lower for patients who underwent repair.
Operative Mortality The operative mortality rate was 3.0% overall, and lower for repair compared with replacement (2.2% versus 4.8%,
p < 0.0001). This trend was evident in all age cohorts, including patients more than 80 years of age. Operative mortality for patients with MVP failure was 6.0% and increased across age group cohorts.
Adjusted Operative Mortality Data COMPLETE VERSUS PARTIAL CHORDAL PRESERVATION. To evaluate potential advantages of chordal sparing in replacement, adjusted mortality rates were calculated. No significant difference in adjusted operative mortality for complete versus partial chordal sparing replacement was found (OR 1.24, 95% confidence interval [CI]: 0.80 to 1.93). Similar results were obtained when all replacements performed secondary to a failed mitral repair were excluded (OR 1.10, 95% CI: 0.67 to 1.80). Consequently, in subsequent adjusted mortality analyses, partial and complete chordal preservation were combined.
Ann Thorac Surg 2019;-:-–-
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
5
Table 4. Postoperative Outcomes Total Cohort (N ¼ 12,043) Outcomes Hospital LOS, days, median (IQR) ICU LOS, hours, median (IQR) Postoperative creatinine, mean (SD) Blood product transfusion Reintubation ICU readmission Reoperation Neurologic complications Prolonged ventilation Renal failure Dialysis (acute onset) Cardiac arrest Atrial fibrillation Postoperative IABP Operative mortality
Overall (N ¼ 12,043)
MVP (n ¼ 8,523)
MVR (n ¼ 3,520)
p Value
7 (5, 9) 49 (26, 93) 1.2 (0.75) 4,452 (37) 510 (4.2) 458 (3.8) 545 (4.5) 366 (3.0) 1,065 (8.9) 242 (2.0) 147 (61) 183 (1.5) 3,815 (32) 59 (0.49) 357 (3.0)
7 (5, 9) 47 (25, 78) 1.2 (0.67) 2,700 (32) 307 (3.6) 302 (3.6) 322 (3.8) 233 (2.7) 607 (7.2) 134 (1.6) 85 (64) 113 (1.3) 2,702 (32) 37 (0.43) 188 (2.2)
8 (6, 10) 67 (37, 115) 1.3 (0.92) 1,752 (50) 203 (5.8) 156 (4.5) 223 (6.4) 133 (3.8) 458 (13) 108 (3.1) 62 (57) 70 (2.0) 1,113 (32) 22 (0.63) 169 (4.8)
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.02 <0.0001 0.002 <0.0001 <0.0001 0.27 0.01 0.93 0.17 <0.0001
Values are n (%) unless otherwise indicated. IABP ¼ intraaortic balloon pump; ICU ¼ intensive care unit; IQR ¼ interquartile range; repair; MVR ¼ mitral valve replacement; TEE ¼ transesophageal echocardiography.
REPAIR VERSUS REPLACEMENT. After adjustment for baseline characteristics, mortality according to age was calculated (Table 5). We found that adjusted odds of mortality for mitral replacement was higher compared with repair in all age groups, but the relationship was not linear: for patients aged 70 to 74 years, OR 1.86 (95% CI: 1.23 to 2.82); for 75 to 79 years, OR 2.25 (95% CI: 1.48 to 3.42); and for 80þ years, OR 1.53 (95% CI: 1.05 to 2.23). Adjusted mortality according to mitral procedure type (repair or replacement) and chordal preservation strategy (none or partial/complete) is presented in Table 6. There was no significant interaction between age group and chordal preservation strategy (p ¼ 0.43). When analyzed according to the actual treatment received (as-treated cohort), mitral replacement nearly doubled the risk of mortality compared with mitral repair (OR 1.83, 95% CI: 1.45 to 2.31). Stratifying MVR on the basis of chordal preservation, any preservation of the subvalvular apparatus (partial or complete) was associated with a lower adjusted risk of mortality after replacement; however, this remained inferior to MVP with OR 1.66 (95% CI: 1.28 to 2.14). There was a higher adjusted mortality rate with no chordal preservation as compared with repair (OR 2.53, 95% CI: 1.67 to 3.84).
Table 5. Adjusted 30-Day Mortality Stratified by Age Age
OR (95% CI)
Overall 70–74 years 75–79 years 80 years CI ¼ confidence interval;
1.83 1.86 2.25 1.53 OR ¼ odds ratio.
(1.45–2.31) (1.23–2.82) (1.48–3.42) (1.05–2.23)
LOS ¼ length of stay;
MVP ¼ mitral valve
On intention-to-treat analysis, patients who underwent mitral replacement secondary to mitral repair failure were included in the mitral repair group. The benefits of repair on mortality was reduced but significant (OR 1.46, 95% CI: 1.14 to 1.86). There was an increased adjusted mortality rate with no chordal preservation (OR 2.07, 95% CI: 1.31 to 3.28). Preservation of the chordal apparatus was associated with improved survival compared with no chordal preservation, and similar survival compared with repair (OR 1.30, 95% CI: 0.99 to 1.72). In an ideal cohort of patients excluding all cases of failed MVP, mitral replacement was associated with significantly worse mortality compared with repair (OR 1.64, 95% CI 1.27 to 2.11). Any preservation of the chordal apparatus was associated with improved survival compared with no chordal preservation, but overall, replacement remained inferior to repair (OR 1.48, 95% CI: 1.11 to 1.96). There was an increased mortality rate with no chordal preservation (OR 2.32, 95% CI: 1.46 to 3.70). Adjusted mortality of patients who underwent MVR after failed repair was significantly increased compared with patients who underwent MVR without a repair attempt (OR 1.52, 95% CI: 1.03 to 2.26).
Comment It has been postulated that mitral repair may not confer short-term benefits over mitral replacement with complete chordal preservation, especially in the elderly population, whose reduced life expectancy may limit the long-term survival benefits of repair. In this population, short-term perioperative morbidity, mortality, and quality of life considerations may take precedence over longterm survival. The purpose of our investigation was to
6
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
Ann Thorac Surg 2019;-:-–-
Table 6. Adjusted 30-Day Mortality by Operative Technique Operative Technique Overall, replacement versus repair By chordal preservation versus repair Partial/complete chordal preservation No chordal preservation CI ¼ confidence interval;
As Treated OR (95% CI)
Intention to Treat OR (95% CI)
Ideal OR (95% CI)
1.83 (1.45–2.31)
1.46 (1.14–1.86)
1.64 (1.27–2.11)
1.66 (1.28–2.14) 2.53 (1.67–3.84)
1.30 (0.99–1.72) 2.07 (1.31–3.28)
1.48 (1.11–1.96) 2.32 (1.46–3.70)
OR ¼ odds ratio.
compare short-term outcomes and adjusted mortality data. Using the STS ACSD database, our cohort included 12,043 patients aged more than 70 years (mean 77) of whom 71% underwent repair. This incidence of repair is higher than previously reported and is likely a result of the stringent inclusion and exclusion criteria used. Patients who underwent mitral repair were slightly younger and had fewer comorbidities; however, it is unclear whether comorbidity profiles drove the operative decisions to repair or replace the degenerative valve. In this selected group of patients, we demonstrated that mitral repair for degenerative mitral valve disease is associated with lower 30-day morbidity and adjusted mortality compared with replacement. These findings remained consistent across all age group cohorts, including patients aged 80 years or more, and are consistent with prior reports [12–14]. However, our study is the first to compare mitral repair to replacement according to the chordal preservation strategy utilized in replacement. Although preservation of the subvalvular apparatus was associated with improved 30-day mortality after mitral replacement, outcomes remained inferior to repair. Interestingly, complete sparing of the subvalvular apparatus offered no additional benefit over partial chordal preservation. In comparison of all cases, no chordal preservation was associated with the worst 30day mortality. Our results are congruent with a prior meta-analysis of MVR [17]. Mitral repair was associated with lower adjusted operative mortality in the as-treated analysis, intentionto-treat analysis, and ideal analysis. The failed MVP rate in the study was 7.9%, and was associated with an increased risk of 30-day mortality (6%) compared with MVR (4.8%). In characterizing failed mitral repairs, we did not incorporate postprocedural TEE results or early repair failures later requiring reoperation. Therefore, we may be underestimating the mitral repair failure rate that is reported. Although we demonstrated that MVP offers a short-term survival advantage over MVR, this is predicated on execution of a successful MVP. The short-term detriment of a failed MVP is clear, but that should not discourage surgeons from offering repair to patients aged 70 years or more. Rather, this highlights the importance of patient selection and considering volume-outcome relationships when devising future guidelines. Our study provides important information about the influence of chordal preservation on short-term outcomes and 30-day mortality. A major advantage of mitral repair
is preservation of the subvalvular apparatus, which can also be accomplished with complete chordal sparing in mitral replacement. Despite this, adjusted operative mortality for replacement was higher compared with repair. Therefore, it appears that while chordal preservation remains important to reduce operative mortality for MVR, mitral repair is the preferred operation in the elderly population.
Study Strengths A major strength of this study includes stratification on the basis of degenerative MR and chordal preservation strategy, for which we had specific information in the data source. Our study included all STS participating hospitals for the time period, not just high-volume centers with experienced mitral valve operators. It is encouraging to note that operative mortality was low (3% overall) in our cohort, and lower still (2.2%) for mitral valve repair. Furthermore, the majority of patients had an excellent result with none/trace/trivial residual MR on postprocedural TEE.
Study Limitations Our study has limitations inherent to any retrospective analysis. Although the STS ACSD is the most rigorous clinical registry in cardiac surgery encompassing a comprehensive array of patients and centers, selection bias remains a major limitation. Obviously, patients with repairable mitral valves were more likely to undergo repair. However, we only included patients with degenerative disease and noncalcified valves. Theoretically, these valves should all be repairable. Identification of this specific subgroup is dependent on accurate data coding, which is subject to operator variability. Our adjusted model has been well validated in prior analyses of mitral valve operations using the STS ACSD [5]; however, we cannot definitively exclude the possibility for uncontrolled confounding variables. We focused our investigation on the elderly, a patient population typically underrepresented in other studies. As such, our results might not be generalizable to younger patients. Within the limitations of the STS database, our conclusions are confined to short-term outcomes and 30-day mortality, which are not appropriate endpoints to measure longterm survival. However, considering the main point of the investigation was to evaluate how chordal preservation strategy in replacement affects short-term mortality, we believe our study makes a significant contribution to
Ann Thorac Surg 2019;-:-–-
the existing body of literature on the topic. Failed repairs in the elderly could have a significant impact on late mortality and quality of life.
Conclusion Although there are many factors that deserve attention in any comparison between mitral repair and replacement, including surgeon and center volume associations, factors associated with repair failure, and long-term outcomes, these were beyond the scope of this investigation given the data source. We have, however, provided robust evidence that mitral repair is a good operation in elderly patients, and that expertise in mitral repair to minimize failures is paramount to improving patient outcomes. In conclusion, mitral repair is associated with lower 30-day postoperative morbidity and mortality, including adjusted mortality, compared with replacement in the elderly. That was true in all age cohorts, including patients aged more than 80 years, irrespective of analysis strategy—as treated (failed repairs included in MVR group), intention to treat (failed repairs included in the MVP group), or ideal (failed repairs excluded). Preservation of the subvalvular apparatus was associated with improved short-term mortality after mitral replacement, but complete sparing of the subvalvular apparatus offered no additional benefit over partial chordal preservation. In comparison with mitral repair, no chordal preservation was associated with the worst short-term mortality, but MVR with preservation of the subvalvular apparatus still remained inferior to repair. The MVP failure rate is substantial and limits the benefit of repair. These findings could provide supporting evidence to the concept of mitral valve centers of excellence. The data for this research were provided by The Society of Thoracic Surgeons National Database Participant User File Research Program. Data analysis was performed at the investigators’ institutions. Funding for this study was provided by the Division of Cardiac Surgery at the University of Massachusetts Medical School.
References 1. Braunberger E, Deloche A, Berrebi A, et al. Very long-term results of valve repair with Carpentier’s techniques in nonrheumatic mitral valve insufficiency. Circulation 2001;104: I8–11.
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
7
2. Daneshmand MA, Milano CA, Rankin JS, et al. Mitral valve repair for degenerative disease: a 20-year experience. Ann Thorac Surg 2009;88:1828–37. 3. Vassileva CM, Shabosky J, Boley T, et al. Cost analysis of isolated mitral valve surgery in the United States. Ann Thorac Surg 2012;94:1429–36. 4. Jokinen JJ, Hippel€ainen MJ, Pitk€ anen OA, et al. Mitral valve replacement versus repair: propensity-adjusted survival and quality-of-life analysis. Ann Thorac Surg 2007;84: 451–8. 5. Gammie JS, Sheng S, Griffith BP, et al. Trends in mitral valve surgery in the United States: results from The Society of Thoracic Surgeons Adult Cardiac Surgery Database. Ann Thorac Surg 2009;87:1431–9. 6. Vassileva CM, Mishkel G, McNeely C, et al. Long-term survival of patients undergoing mitral valve repair and replacement: a longitudinal analysis of Medicare fee-forservice beneficiaries. Circulation 2013;127:1870–6. 7. Vassileva CM, McNeely C, Spertus J, et al. Hospital volume, mitral repair rates, and mortality in mitral valve surgery in the elderly: an analysis of US hospitals treating Medicare feefor-service patients. J Thorac Cardiovasc Surg 2015;149: 762–8. 8. Hannein HA, Swain JA, McIntosh CL, et al. Comparative assessment of chordal preservation versus chordal resection during mitral valve replacement. J Thorac Cardiovasc Surg 1990;99:828–37. 9. Natsuaki M, Itoh T, Tomita S, et al. Importance of preserving the mitral subvalvular apparatus in mitral valve replacement. Ann Thorac Surg 1996;61:585–90. 10. Yun KL, Sintek CF, Miller DC, et al. Randomized trial comparing partial vs complete chordal-sparing mitral valve replacement: effects on left ventricular volume and function. J Thorac Cardiovasc Surg 2002;123:707–14. 11. Badhwar V, Peterson ED, Jacobs JP, et al. Longitudinal outcome of isolated mitral repair in older patients: results from 14,604 procedures performed from 1991 to 2007. Ann Thorac Surg 2012;94:1870–7. 12. Gaur P, Kaneko T, McGurk S, et al. Mitral valve repair versus replacement in the elderly: short-term and long-term outcomes. J Thorac Cardiovasc Surg 2014;148:1400–6. 13. Nloga J, Henaine R, Vergnat M, et al. Mitral valve surgery in octogenarians: should we fight for repair? A survival and quality-of-life assessment. Eur J Cardiothorac Surg 2011;39: 875–80. 14. Chikwe J, Goldstone AB, Passage J, et al. A propensity scoreadjusted retrospective comparison of early and mid- term results of mitral valve repair versus replacement in octogenarians. Eur Heart J 2011;32:618–26. 15. Gillinov AM, Blackstone EH, Nowicki ER, et al. Valve repair versus valve replacement for degenerative mitral valve disease. J Thorac Cardiovasc Surg 2008;135:885–93. 16. Moss RR, Humphries KH, Gao M, et al. Outcome of mitral valve repair or replacement: a comparison by propensity score analysis. Circulation 2003;108:90–7. 17. Sa M, Escobar R, Ferraz P, et al. Complete versus partial preservation of mitral valve apparatus during mitral valve replacement: meta-analysis and meta-regression of 1535 patients. Eur J Cardiothorac Surg 2013;44:905–12.
Background. It has been postulated that mitral valve repair in the elderly does not confer short-term benefits over mitral valve replacement with complete preservation of the chordal apparatus. Our purpose was to test this hypothesis using data from The Society of Thoracic Surgeons Adult Cardiac Surgery Database (STS ACSD). Methods. Patients aged 70 years or more undergoing primary isolated elective mitral valve repair or mitral valve replacement for degenerative disease were obtained from the STS ACSD versions 2.73 and 2.81. Patients with a concomitant tricuspid procedure, atrial fibrillation surgery, or atrial septal defect/patent foramen ovale repair were included. The two treatment groups were further stratified by age in years (70 to 74, 75 to 79, and 80 or more). Adjusted 30-day mortality rates were analyzed by mitral procedure and chordal preservation strategy. Results. The study included 12,043 patients, of whom 71% underwent mitral valve repair. Observed 30-day mortality after repair was lower than after replacement (2.2% versus 4.8%, respectively; p < 0.0001). Using repair
as reference, adjusted operative mortality was higher for replacement in the overall cohort (odds ratio 1.83, 95% confidence interval: 1.45 to 2.31). There was no significant difference in mortality between complete versus partial chordal preservation in repair (odds ratio 1.24, 95% confidence interval: 0.80 to 1.93). Mitral valve replacement with chordal preservation remained inferior to repair (odds ratio 1.66, 95% confidence interval: 1.28 to 2.14). The failed repair rate was 7.9%, with a 30-day mortality of 6%. Conclusions. In patients aged 70 years or more, degenerative mitral repair was associated with lower operative mortality compared with replacement, irrespective of chordal preservation strategy. Failed repairs reduced this short-term benefit compared with chordalsparing replacement as evidenced by the similar operative mortality on an intention to treat analysis.
C
primary mitral regurgitation (MR), the rate of MVP increased from 42% to 61% [5]. Despite improvements, these national trends fall far short of the experiences reported by high-volume centers, and implementation of MVP in the care of the elderly population has been less uniformly accepted [6, 7]. The benefits of preserving the chordal apparatus in mitral valve replacement (MVR) have been shown in a number of studies [8, 9], including a randomized controlled trial in 2002 [10]. Despite robust retrospective studies comparing mitral repair to replacement, many of
urrent American Heart Association/American College of Cardiology guidelines recommend mitral valve repair (MVP) over replacement for degenerative mitral valve disease. The advantages of MVP compared with replacement in this population have been well established [1–4]. Accordingly, the utilization of MVP has steadily increased. In an 8-year review (2000 to 2007) of The Society of Thoracic Surgeons (STS) Adult Cardiac Surgery Database (ACSD), in patients with isolated Accepted for publication Sept 6, 2018. Presented at the Fifty-fourth Annual Meeting of The Society of Thoracic Surgeons, Fort Lauderdale, FL, Jan 27–31, 2018. Address correspondence to Dr Vassileva, Division of Cardiac Surgery, Department of Surgery, University of Massachusetts Medical School, 55 Lake Ave N, Worcester, MA 01655; email: mitralrepair@gmail. com.
Ó 2019 by The Society of Thoracic Surgeons Published by Elsevier Inc.
(Ann Thorac Surg 2019;-:-–-) Ó 2019 by The Society of Thoracic Surgeons
The Supplemental Tables can be viewed in the online version of this article [https://doi.org/10.1016/ j.athoracsur.2018.09.018] on http://www.annalsthoracic surgery.org.
0003-4975/$36.00 https://doi.org/10.1016/j.athoracsur.2018.09.018
2
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
these did not include information on the chordal preservation strategy, if any [2, 11–16]. Furthermore, no prospective randomized clinical trials comparing outcomes of mitral repair versus replacement in patients with degenerative MR have been performed. As a result, it remains unclear whether MVP offers short-term benefits over complete chordal-sparing MVR, particularly in the elderly population. The purpose of this study was twofold. Our first aim was to evaluate the operative morbidity and short-term mortality associated with MVP and MVR for degenerative MR in the elderly population to assess whether the benefit of repair over replacement diminishes with age. Our second aim was to assess 30-day outcomes after mitral repair in the elderly compared with mitral replacement according to the chordal preservation strategy utilized.
Ann Thorac Surg 2019;-:-–-
was documented in 10,050 patients (83%) and therefore not incorporated into the definition of a failed repair; 668 patients (14%) did not undergo a postprocedural TEE, and this data field was not specified for the remaining 325 patients (3%). Mortality was defined as operative mortality or any death within 30 days after surgery in accordance with the standard STS definition. All unknown values for variables were set to missing. Missing variables were generally less than 0.3%.
Statistical Analysis
The STS ACSD versions 2.73 and 2.81 were queried from July 1, 2011, to June 30, 2016. Demographics, socioeconomic factors, chronic conditions, operative characteristics, and 30-day morbidity and mortality were included in the provider user files. The study was approved by the Institutional Review Board of the University of Massachusetts Medical School. Informed consent was not required given the retrospective nature of the review and the deidentified nature of the records.
Percentages were reported for categorical variables. For normally distributed continuous variables, mean and standard deviation were reported with p values from Student’s t tests. For skewed continuous variables, median and interquartile range were reported along with p values from the Wilcoxon rank sum test. The p values were considered statistically significant at an alpha level of 0.05. Logistic regression was used to model the odds ratio (OR) of mortality. The models were adjusted for the following variables: age, sex, body mass index, diabetes mellitus, last hematocrit, last creatinine, dyslipidemia, hypertension, chronic lung disease, liver disease, immunocompromise, peripheral vascular disease, cerebrovascular disease, heart failure, ejection fraction, atrial fibrillation, tricuspid insufficiency, preoperative intraaortic balloon pump, and previous myocardial infarction. The covariates included in the final adjusted models were clinically relevant variables and have been well validated in prior reports [5]. All analyses were performed using SAS software (version 9.4; SAS Institute, Cary, NC).
Patient Cohort
Results
All patients aged 70 years or more who were undergoing an elective primary isolated mitral valve operation for degenerative disease were considered for inclusion in the study. Under this definition, patients with concomitant ablation for atrial fibrillation, tricuspid repair or replacement, and closure of a patent foramen ovale or an atrial septal defect were included. Exclusion criteria were defined as a preoperative unresponsive neurologic state, myocardial infarction within 21 days of surgery, cardiogenic shock, endocarditis, mitral disease etiology other than degenerative, mitral stenosis, unplanned mitral valve procedure, urgent, emergent, or salvage procedures, and transcatheter mitral procedures.
Baseline Characteristics
Patients and Methods Data Sources
Data Variables All variables were adopted from the STS forms versions 2.73 and 2.81. Specific variable definitions are described as follows. Chordal preservation was subdivided into complete (preservation of both anterior and posterior leaflets), partial (preservation of posterior or anterior leaflet), and none. For 337 (9.6%) MVR procedures, the chordal approach was not defined, and these patients were excluded from regression analyses to estimate operative mortality. A failed mitral repair was defined as an unsuccessful repair requiring mitral replacement during the same operation. Performance of an intraoperative postprocedural transesophageal echocardiogram (TEE)
In all, 12,043 patients were included in the study, of whom 8,523 (71%) underwent MVP and 3,520 (29%) underwent MVR. Mean age for the entire cohort was 77 years. There was a significant disease burden in patients undergoing both mitral repair and replacement (Table 1, Supplemental Tables 1–4).
Intraoperative Variables Detailed operative characteristics are presented in Table 2. Median duration of cardiopulmonary bypass and median aortic cross-clamp time were significantly lower for patients undergoing repair compared with replacement (115 minutes versus 122 minutes, p < 0.0001, and 83 minutes versus 90 minutes, p < 0.0001, respectively). When these data were stratified by age, both cardiopulmonary bypass time and aortic cross-clamp time remained the same or decreased with age, and both were consistently lower for repair compared with replacement in all age groups. Overall, 83% of the patients underwent intraoperative postprocedural TEE. The majority of patients had an excellent result; however, 4.0% of patients had documented moderate/severe mitral regurgitation. Procedural details are provided in Table 3. The mitral repair failure rate was 7.9%. Among patients with failed mitral repair, median cardiopulmonary bypass time was 160 minutes (interquartile range: 121 to 210), and median
Ann Thorac Surg 2019;-:-–-
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
3
Table 1. Baseline Characteristics of Study Cohort Total Cohort (N ¼ 12,043) Characteristics Age, years, mean (SD) Male White Body mass index, mean (SD) Diabetes mellitus Last hematocrit, mean (SD) Last creatinine, mean (SD) Dyslipidemia Hypertension Chronic lung disease Liver disease Immunocompromised Peripheral vascular disease Cerebrovascular disease History of heart failure EF, median (IQR) Atrial fibrillation Tricuspid insufficiency Preoperative IABP Previous myocardial infarction
Overall (N ¼ 12,043)
MVP (n ¼ 8,523)
MVR (n ¼ 3,520)
p Value
77 (4.8) 5,587 (46) 11,054 (92) 27 (12) 1,737 (14) 40 (4.6) 1.0 (0.50) 7,627 (63) 9,134 (76) 903 (7.6) 249 (2.1) 289 (2.4) 661 (5.5) 1,349 (11) 6,907 (58) 60 (55, 64) 5,674 (47) 4,106 (35) 24 (0.20) 706 (5.9)
76 (4.7) 4,163 (49) 7,861 (93) 27 (13) 1,089 (13) 40 (4.4) 1.0 (0.45) 5,311 (62) 6,317 (74) 545 (6.5) 166 (2.0) 187 (2.2) 440 (5.2) 869 (10) 4,722 (56) 60 (55, 65) 3,859 (45) 2,811 (33) 14 (0.16) 459 (5.4)
78 (4.9) 1,424 (40) 3,193 (91) 27 (12) 648 (18) 39 (4.8) 1.1 (0.55) 2,316 (66) 2,817 (80) 358 (10) 83 (2.4) 102 (2.9) 221 (6.3) 480 (14) 2,185 (62) 60 (53, 63) 1,815 (52) 1,295 (37) 10 (0.28) 247 (7.0)
<0.0001 <0.0001 0.002 0.06 <0.0001 <0.0001 <0.0001 0.0005 <0.0001 <0.0001 0.18 0.02 0.02 <0.0001 <0.0001 0.001 <0.0001 <0.0001 0.18 0.001
Values are n (%) unless otherwise indicated. EF ¼ ejection fraction; replacement.
IABP ¼ intraaortic balloon pump;
IQR ¼ interquartile range;
MVP ¼ mitral valve repair;
MVR ¼ mitral valve
Table 2. Operative Characteristics Total Cohort (N ¼ 12,043) Characteristics Median sternotomy Blood product transfusion CPB, minutes, median (IQR) Aortic CCT, minutes, median (IQR) Lowest CPB temp, C, median (IQR) Lowest CPB HCT, mean (SD) Tricuspid procedure Replacement Repair Intraop IABP Atrial fibrillation surgery Intraop postprocedural TEE Postprocedural MR None Trace/trivial Mild Moderate Severe Unknown
Overall (N ¼ 12,043)
MVP (n ¼ 8,523)
MVR (n ¼ 3,520)
p Value
8,728 (73) 4,460 (37) 117 (90–151) 85 (65–111) 33 (31–34) 25 (4.5)
5,798 (68) 971 (28) 115 (88–148) 83 (63–107) 33 (31–34) 25 (4.5)
2,930 (83) 779 (46) 122 (94–160) 90 (69–122) 33 (31–34) 24 (4.3)
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001 <0.0001
41 (1.6) 2,528 (98) 141 (1.2) 881 (7.3) 10,050 (83)
17 (0.92) 1,834 (99) 82 (0.96) 634 (7.4) 7,241 (85)
24 (3.3) 694 (97) 59 (1.7) 247 (7) 2,809 (80)
<0.0001
5,802 (58) 2,715 (27) 766 (7.6) 158 (1.6) 234 (2.3) 375 (3.7)
3,927 (54) 2,192 (30) 667 (9.2) 117 (1.6) 138 (1.9) 200 (2.8)
1,875 (67) 523 (19) 99 (3.5) 41 (1.5) 96 (3.4) 175 (6.2)
0.001 0.42 <0.0001 <0.0001
Values are n (%) unless otherwise indicated. CCT ¼ cross-clamp time; CPB ¼ cardiopulmonary bypass; HCT ¼ hematocrit; IABP ¼ intraaortic balloon pump; Intraop ¼ intraoperative; IQR ¼ interquartile range; MR ¼ mitral regurgitation; MVP ¼ mitral valve repair; MVR ¼ mitral valve replacement; temp ¼ temperature; TEE ¼ transesophageal echocardiography.
4
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
Ann Thorac Surg 2019;-:-–-
Table 3. Mitral Procedure Details By STS Version Procedure Details Mitral repair Repair type Annuloplasty Leaflet resection Leaflet type Triangular Quadrangular Other Leaflet plication Leaflet debridement Chordal/leaflet transfer Neochords (PTFE) Mitral commissuroplasty Folding plasty Leaflet extension/replacement/patch Sliding plasty Mitral cleft repair (scallop closure) Edge-to-edge repair Mitral commissurotomy Annular decalcification Failed repair ratea Mitral replacement Chordal preservation None Anterior Posterior Both Unknown Mitral implant Bioprosthetic valve Mechanical valve
Overall
2.73
2.81
8,523 (71)
5,262 (72)
3,261 (70)
8,142 (96) 3,765 (45)
4,982 (95) 2,231 (43)
3,160 (97) 1,534 (47)
2,233 (26) 1,215 (14) 251 (3.0) 276 (8.6) 30 (0.93) 230 (2.7) 1,926 (23) 163 (5.1) 159 (4.9) 105 (1.3) 715 (8.5) 468 (14) 482 (5.7) 87 (1.0) 128 (1.5) 732 (7.9) 3,520 (29)
1,247 (24) 759 (14) 169 (3) ... ... 152 (2.9) 1,005 (19) ... ... 64 (1.2) 449 (8.6) ... 318 (6.1) 65 (1.3) 81 (1.6) 441 (7.7) 2,095 (28)
986 (30) 456 (14) 82 (3.0) 276 (8.6) 30 (0.93) 78 (2.4) 921 (28) 163 (5.1) 159 (4.9) 41 (1.3) 266 (8.2) 468 (14) 164 (5.1) 22 (0.68) 47 (1.5) 291 (8.2) 1425 (30)
572 (16) 126 (3.6) 831 (24) 1,654 (47) 337 (10)
375 (18) 70 (3) 504 (24) 962 (46) 184 (8.8)
197 (14) 56 (3.9) 327 (23) 692 (49) 153 (11)
1,383 (39) 37 (1.1)
... ...
1,383 (39) 37 (1.1)
a
Calculated as failed repairs divided by total repairs plus failed repairs.
Values are n (%) unless otherwise specified. PTFE ¼ polytetrafluoroethylene;
STS ¼ The Society of Thoracic Surgeons.
aortic cross-clamp time was 123 minutes (interquartile range: 91 to 166).
Postoperative Outcomes The postoperative outcomes of the study cohort are presented in Table 4. Patients who underwent repair had fewer postoperative complications, including lower rates of prolonged ventilation (7.2% versus 13%, p < 0.0001), reintubation (3.6% versus 5.8%, p < 0.0001), need for reoperation (3.8% versus 6.4%, p < 0.0001), and need for blood transfusion (32% versus 50%, p < 0.0001). Many of the postoperative complications increased with age for both repair and replacement, but remained consistently lower for patients who underwent repair.
Operative Mortality The operative mortality rate was 3.0% overall, and lower for repair compared with replacement (2.2% versus 4.8%,
p < 0.0001). This trend was evident in all age cohorts, including patients more than 80 years of age. Operative mortality for patients with MVP failure was 6.0% and increased across age group cohorts.
Adjusted Operative Mortality Data COMPLETE VERSUS PARTIAL CHORDAL PRESERVATION. To evaluate potential advantages of chordal sparing in replacement, adjusted mortality rates were calculated. No significant difference in adjusted operative mortality for complete versus partial chordal sparing replacement was found (OR 1.24, 95% confidence interval [CI]: 0.80 to 1.93). Similar results were obtained when all replacements performed secondary to a failed mitral repair were excluded (OR 1.10, 95% CI: 0.67 to 1.80). Consequently, in subsequent adjusted mortality analyses, partial and complete chordal preservation were combined.
Ann Thorac Surg 2019;-:-–-
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
5
Table 4. Postoperative Outcomes Total Cohort (N ¼ 12,043) Outcomes Hospital LOS, days, median (IQR) ICU LOS, hours, median (IQR) Postoperative creatinine, mean (SD) Blood product transfusion Reintubation ICU readmission Reoperation Neurologic complications Prolonged ventilation Renal failure Dialysis (acute onset) Cardiac arrest Atrial fibrillation Postoperative IABP Operative mortality
Overall (N ¼ 12,043)
MVP (n ¼ 8,523)
MVR (n ¼ 3,520)
p Value
7 (5, 9) 49 (26, 93) 1.2 (0.75) 4,452 (37) 510 (4.2) 458 (3.8) 545 (4.5) 366 (3.0) 1,065 (8.9) 242 (2.0) 147 (61) 183 (1.5) 3,815 (32) 59 (0.49) 357 (3.0)
7 (5, 9) 47 (25, 78) 1.2 (0.67) 2,700 (32) 307 (3.6) 302 (3.6) 322 (3.8) 233 (2.7) 607 (7.2) 134 (1.6) 85 (64) 113 (1.3) 2,702 (32) 37 (0.43) 188 (2.2)
8 (6, 10) 67 (37, 115) 1.3 (0.92) 1,752 (50) 203 (5.8) 156 (4.5) 223 (6.4) 133 (3.8) 458 (13) 108 (3.1) 62 (57) 70 (2.0) 1,113 (32) 22 (0.63) 169 (4.8)
<0.0001 <0.0001 <0.0001 <0.0001 <0.0001 0.02 <0.0001 0.002 <0.0001 <0.0001 0.27 0.01 0.93 0.17 <0.0001
Values are n (%) unless otherwise indicated. IABP ¼ intraaortic balloon pump; ICU ¼ intensive care unit; IQR ¼ interquartile range; repair; MVR ¼ mitral valve replacement; TEE ¼ transesophageal echocardiography.
REPAIR VERSUS REPLACEMENT. After adjustment for baseline characteristics, mortality according to age was calculated (Table 5). We found that adjusted odds of mortality for mitral replacement was higher compared with repair in all age groups, but the relationship was not linear: for patients aged 70 to 74 years, OR 1.86 (95% CI: 1.23 to 2.82); for 75 to 79 years, OR 2.25 (95% CI: 1.48 to 3.42); and for 80þ years, OR 1.53 (95% CI: 1.05 to 2.23). Adjusted mortality according to mitral procedure type (repair or replacement) and chordal preservation strategy (none or partial/complete) is presented in Table 6. There was no significant interaction between age group and chordal preservation strategy (p ¼ 0.43). When analyzed according to the actual treatment received (as-treated cohort), mitral replacement nearly doubled the risk of mortality compared with mitral repair (OR 1.83, 95% CI: 1.45 to 2.31). Stratifying MVR on the basis of chordal preservation, any preservation of the subvalvular apparatus (partial or complete) was associated with a lower adjusted risk of mortality after replacement; however, this remained inferior to MVP with OR 1.66 (95% CI: 1.28 to 2.14). There was a higher adjusted mortality rate with no chordal preservation as compared with repair (OR 2.53, 95% CI: 1.67 to 3.84).
Table 5. Adjusted 30-Day Mortality Stratified by Age Age
OR (95% CI)
Overall 70–74 years 75–79 years 80 years CI ¼ confidence interval;
1.83 1.86 2.25 1.53 OR ¼ odds ratio.
(1.45–2.31) (1.23–2.82) (1.48–3.42) (1.05–2.23)
LOS ¼ length of stay;
MVP ¼ mitral valve
On intention-to-treat analysis, patients who underwent mitral replacement secondary to mitral repair failure were included in the mitral repair group. The benefits of repair on mortality was reduced but significant (OR 1.46, 95% CI: 1.14 to 1.86). There was an increased adjusted mortality rate with no chordal preservation (OR 2.07, 95% CI: 1.31 to 3.28). Preservation of the chordal apparatus was associated with improved survival compared with no chordal preservation, and similar survival compared with repair (OR 1.30, 95% CI: 0.99 to 1.72). In an ideal cohort of patients excluding all cases of failed MVP, mitral replacement was associated with significantly worse mortality compared with repair (OR 1.64, 95% CI 1.27 to 2.11). Any preservation of the chordal apparatus was associated with improved survival compared with no chordal preservation, but overall, replacement remained inferior to repair (OR 1.48, 95% CI: 1.11 to 1.96). There was an increased mortality rate with no chordal preservation (OR 2.32, 95% CI: 1.46 to 3.70). Adjusted mortality of patients who underwent MVR after failed repair was significantly increased compared with patients who underwent MVR without a repair attempt (OR 1.52, 95% CI: 1.03 to 2.26).
Comment It has been postulated that mitral repair may not confer short-term benefits over mitral replacement with complete chordal preservation, especially in the elderly population, whose reduced life expectancy may limit the long-term survival benefits of repair. In this population, short-term perioperative morbidity, mortality, and quality of life considerations may take precedence over longterm survival. The purpose of our investigation was to
6
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
Ann Thorac Surg 2019;-:-–-
Table 6. Adjusted 30-Day Mortality by Operative Technique Operative Technique Overall, replacement versus repair By chordal preservation versus repair Partial/complete chordal preservation No chordal preservation CI ¼ confidence interval;
As Treated OR (95% CI)
Intention to Treat OR (95% CI)
Ideal OR (95% CI)
1.83 (1.45–2.31)
1.46 (1.14–1.86)
1.64 (1.27–2.11)
1.66 (1.28–2.14) 2.53 (1.67–3.84)
1.30 (0.99–1.72) 2.07 (1.31–3.28)
1.48 (1.11–1.96) 2.32 (1.46–3.70)
OR ¼ odds ratio.
compare short-term outcomes and adjusted mortality data. Using the STS ACSD database, our cohort included 12,043 patients aged more than 70 years (mean 77) of whom 71% underwent repair. This incidence of repair is higher than previously reported and is likely a result of the stringent inclusion and exclusion criteria used. Patients who underwent mitral repair were slightly younger and had fewer comorbidities; however, it is unclear whether comorbidity profiles drove the operative decisions to repair or replace the degenerative valve. In this selected group of patients, we demonstrated that mitral repair for degenerative mitral valve disease is associated with lower 30-day morbidity and adjusted mortality compared with replacement. These findings remained consistent across all age group cohorts, including patients aged 80 years or more, and are consistent with prior reports [12–14]. However, our study is the first to compare mitral repair to replacement according to the chordal preservation strategy utilized in replacement. Although preservation of the subvalvular apparatus was associated with improved 30-day mortality after mitral replacement, outcomes remained inferior to repair. Interestingly, complete sparing of the subvalvular apparatus offered no additional benefit over partial chordal preservation. In comparison of all cases, no chordal preservation was associated with the worst 30day mortality. Our results are congruent with a prior meta-analysis of MVR [17]. Mitral repair was associated with lower adjusted operative mortality in the as-treated analysis, intentionto-treat analysis, and ideal analysis. The failed MVP rate in the study was 7.9%, and was associated with an increased risk of 30-day mortality (6%) compared with MVR (4.8%). In characterizing failed mitral repairs, we did not incorporate postprocedural TEE results or early repair failures later requiring reoperation. Therefore, we may be underestimating the mitral repair failure rate that is reported. Although we demonstrated that MVP offers a short-term survival advantage over MVR, this is predicated on execution of a successful MVP. The short-term detriment of a failed MVP is clear, but that should not discourage surgeons from offering repair to patients aged 70 years or more. Rather, this highlights the importance of patient selection and considering volume-outcome relationships when devising future guidelines. Our study provides important information about the influence of chordal preservation on short-term outcomes and 30-day mortality. A major advantage of mitral repair
is preservation of the subvalvular apparatus, which can also be accomplished with complete chordal sparing in mitral replacement. Despite this, adjusted operative mortality for replacement was higher compared with repair. Therefore, it appears that while chordal preservation remains important to reduce operative mortality for MVR, mitral repair is the preferred operation in the elderly population.
Study Strengths A major strength of this study includes stratification on the basis of degenerative MR and chordal preservation strategy, for which we had specific information in the data source. Our study included all STS participating hospitals for the time period, not just high-volume centers with experienced mitral valve operators. It is encouraging to note that operative mortality was low (3% overall) in our cohort, and lower still (2.2%) for mitral valve repair. Furthermore, the majority of patients had an excellent result with none/trace/trivial residual MR on postprocedural TEE.
Study Limitations Our study has limitations inherent to any retrospective analysis. Although the STS ACSD is the most rigorous clinical registry in cardiac surgery encompassing a comprehensive array of patients and centers, selection bias remains a major limitation. Obviously, patients with repairable mitral valves were more likely to undergo repair. However, we only included patients with degenerative disease and noncalcified valves. Theoretically, these valves should all be repairable. Identification of this specific subgroup is dependent on accurate data coding, which is subject to operator variability. Our adjusted model has been well validated in prior analyses of mitral valve operations using the STS ACSD [5]; however, we cannot definitively exclude the possibility for uncontrolled confounding variables. We focused our investigation on the elderly, a patient population typically underrepresented in other studies. As such, our results might not be generalizable to younger patients. Within the limitations of the STS database, our conclusions are confined to short-term outcomes and 30-day mortality, which are not appropriate endpoints to measure longterm survival. However, considering the main point of the investigation was to evaluate how chordal preservation strategy in replacement affects short-term mortality, we believe our study makes a significant contribution to
Ann Thorac Surg 2019;-:-–-
the existing body of literature on the topic. Failed repairs in the elderly could have a significant impact on late mortality and quality of life.
Conclusion Although there are many factors that deserve attention in any comparison between mitral repair and replacement, including surgeon and center volume associations, factors associated with repair failure, and long-term outcomes, these were beyond the scope of this investigation given the data source. We have, however, provided robust evidence that mitral repair is a good operation in elderly patients, and that expertise in mitral repair to minimize failures is paramount to improving patient outcomes. In conclusion, mitral repair is associated with lower 30-day postoperative morbidity and mortality, including adjusted mortality, compared with replacement in the elderly. That was true in all age cohorts, including patients aged more than 80 years, irrespective of analysis strategy—as treated (failed repairs included in MVR group), intention to treat (failed repairs included in the MVP group), or ideal (failed repairs excluded). Preservation of the subvalvular apparatus was associated with improved short-term mortality after mitral replacement, but complete sparing of the subvalvular apparatus offered no additional benefit over partial chordal preservation. In comparison with mitral repair, no chordal preservation was associated with the worst short-term mortality, but MVR with preservation of the subvalvular apparatus still remained inferior to repair. The MVP failure rate is substantial and limits the benefit of repair. These findings could provide supporting evidence to the concept of mitral valve centers of excellence. The data for this research were provided by The Society of Thoracic Surgeons National Database Participant User File Research Program. Data analysis was performed at the investigators’ institutions. Funding for this study was provided by the Division of Cardiac Surgery at the University of Massachusetts Medical School.
References 1. Braunberger E, Deloche A, Berrebi A, et al. Very long-term results of valve repair with Carpentier’s techniques in nonrheumatic mitral valve insufficiency. Circulation 2001;104: I8–11.
HENDRIX ET AL MITRAL REPAIR VERSUS REPLACEMENT IN ELDERLY
7
2. Daneshmand MA, Milano CA, Rankin JS, et al. Mitral valve repair for degenerative disease: a 20-year experience. Ann Thorac Surg 2009;88:1828–37. 3. Vassileva CM, Shabosky J, Boley T, et al. Cost analysis of isolated mitral valve surgery in the United States. Ann Thorac Surg 2012;94:1429–36. 4. Jokinen JJ, Hippel€ainen MJ, Pitk€ anen OA, et al. Mitral valve replacement versus repair: propensity-adjusted survival and quality-of-life analysis. Ann Thorac Surg 2007;84: 451–8. 5. Gammie JS, Sheng S, Griffith BP, et al. Trends in mitral valve surgery in the United States: results from The Society of Thoracic Surgeons Adult Cardiac Surgery Database. Ann Thorac Surg 2009;87:1431–9. 6. Vassileva CM, Mishkel G, McNeely C, et al. Long-term survival of patients undergoing mitral valve repair and replacement: a longitudinal analysis of Medicare fee-forservice beneficiaries. Circulation 2013;127:1870–6. 7. Vassileva CM, McNeely C, Spertus J, et al. Hospital volume, mitral repair rates, and mortality in mitral valve surgery in the elderly: an analysis of US hospitals treating Medicare feefor-service patients. J Thorac Cardiovasc Surg 2015;149: 762–8. 8. Hannein HA, Swain JA, McIntosh CL, et al. Comparative assessment of chordal preservation versus chordal resection during mitral valve replacement. J Thorac Cardiovasc Surg 1990;99:828–37. 9. Natsuaki M, Itoh T, Tomita S, et al. Importance of preserving the mitral subvalvular apparatus in mitral valve replacement. Ann Thorac Surg 1996;61:585–90. 10. Yun KL, Sintek CF, Miller DC, et al. Randomized trial comparing partial vs complete chordal-sparing mitral valve replacement: effects on left ventricular volume and function. J Thorac Cardiovasc Surg 2002;123:707–14. 11. Badhwar V, Peterson ED, Jacobs JP, et al. Longitudinal outcome of isolated mitral repair in older patients: results from 14,604 procedures performed from 1991 to 2007. Ann Thorac Surg 2012;94:1870–7. 12. Gaur P, Kaneko T, McGurk S, et al. Mitral valve repair versus replacement in the elderly: short-term and long-term outcomes. J Thorac Cardiovasc Surg 2014;148:1400–6. 13. Nloga J, Henaine R, Vergnat M, et al. Mitral valve surgery in octogenarians: should we fight for repair? A survival and quality-of-life assessment. Eur J Cardiothorac Surg 2011;39: 875–80. 14. Chikwe J, Goldstone AB, Passage J, et al. A propensity scoreadjusted retrospective comparison of early and mid- term results of mitral valve repair versus replacement in octogenarians. Eur Heart J 2011;32:618–26. 15. Gillinov AM, Blackstone EH, Nowicki ER, et al. Valve repair versus valve replacement for degenerative mitral valve disease. J Thorac Cardiovasc Surg 2008;135:885–93. 16. Moss RR, Humphries KH, Gao M, et al. Outcome of mitral valve repair or replacement: a comparison by propensity score analysis. Circulation 2003;108:90–7. 17. Sa M, Escobar R, Ferraz P, et al. Complete versus partial preservation of mitral valve apparatus during mitral valve replacement: meta-analysis and meta-regression of 1535 patients. Eur J Cardiothorac Surg 2013;44:905–12.