Survival Implication of Left Ventricular End-Systolic Diameter in Mitral Regurgitation Due to Flail Leaflets

Transcription

1 Journal of the American College of Cardiology Vol. 54, No. 21, by the American College of Cardiology Foundation ISSN /09/$36.00 Published by Elsevier Inc. doi: /j.jacc Valvular Heart Disease Survival Implication of Left Ventricular End-Systolic Diameter in Mitral Regurgitation Due to Flail Leaflets A Long-Term Follow-Up Multicenter Study CME Christophe Tribouilloy, MD, PHD,* Francesco Grigioni, MD, PHD, Jean François Avierinos, MD, Andrea Barbieri, MD, Dan Rusinaru, MD,* Catherine Szymanski, MD,* Marinella Ferlito, MD, Laurence Tafanelli, MD, Francesca Bursi, MD, Faouzi Trojette, MD,* Angelo Branzi, MD, Gilbert Habib, MD, Maria G. Modena, MD, Maurice Enriquez-Sarano, MD, on behalf of the MIDA Investigators Amiens and Marseille, France; Bologna and Modena, Italy; and Rochester, Minnesota Objectives Background Methods Results Conclusions This study analyzed the association of left ventricular end-systolic diameter (LVESD) with survival after diagnosis in organic mitral regurgitation (MR) due to flail leaflets. LVESD is a marker of left ventricular function in patients with organic MR but its association to survival after diagnosis is unknown. The MIDA (Mitral Regurgitation International Database) registry is a multicenter registry of echocardiographically diagnosed organic MR due to flail leaflets. We enrolled 739 patients with MR due to flail leaflets (age years; ejection fraction: 65 10%) in whom LVESD was measured (36 7 mm). Under conservative management, 10-year survival and survival free of cardiac death were higher with LVESD 40 mm versus 40 mm (64 5% vs %; p 0.001, and 73 5% vs %; p 0.001). LVESD 40 mm independently predicted overall mortality (hazard ratio [HR]: 1.95, 95% confidence interval [CI]: 1.01 to 3.83) and cardiac mortality (HR: 3.09, 95% CI: 1.35 to 7.09) under conservative management. Mortality risk increased linearly with LVESD 40 mm (HR: 1.15, 95% CI: 1.04 to 1.27 per 1-mm increment). During the entire follow-up (including post-surgical), LVESD 40 mm independently predicted overall mortality (HR: 1.86, 95% CI: 1.24 to 2.80) and cardiac mortality (HR: 2.14, 95% CI: 1.29 to 3.56), due to persistence of excess mortality in patients with LVESD 40 mm after surgery (HR: 1.86, 95% CI: 1.11 to 3.15 for overall death, and HR: 1.81, 95% CI: 1.05 to 3.54 for cardiac death). In MR due to flail leaflets, LVESD 40 mm is independently associated with increased mortality under medical management but also after mitral surgery. These findings support prompt surgical rescue in patients with LVESD 40 mm but also suggest that best preservation of survival is achieved in patients operated before LVESD reaches 40 mm. (J Am Coll Cardiol 2009;54:1961 8) 2009 by the American College of Cardiology Foundation In Western countries, degenerative valve disease is the most frequent cause of organic mitral regurgitation (MR) (1,2). Severe organic MR is a progressive disease associated with Continuing Medical Education (CME) is available for this article. From the *Department of Cardiology, INSERM, ERI 12, and University Hospital, Amiens, France; Cardiovascular Department, University Hospital of Bologna, Bologna, Italy; Department of Cardiology, University Hospital of Marseille, Marseille, France; Department of Cardiology, University Hospital of Modena, Modena, Italy; and the Division of Cardiovascular Diseases and Internal Medicine, Mayo Clinic, Rochester, Minnesota. Supported by a grant from the University of Bologna, Italy, donated by the Foundation Luisa Fanti Melloni and contributions from the Banca del Monte Foundation, Italy. Dr. Grigioni has received honoraria and travel grants from Edwards Lifesciences. Dr. Enriquez-Sarano is a consultant for and has received grants from Pfizer, AstraZeneca, and Edwards Lifesciences. Manuscript received December 22, 2008; revised manuscript received June 3, 2009, accepted June 9, high morbidity and excess mortality under medical management (3 6). Mitral valve surgery is required for patients with severe MR and overt symptoms because these patients incur high mortality. However, symptoms are not sensitive for patients at high risk, and patients who undergo surgery for symptoms continue to incur high mortality even after See page 1969 successful surgery. Thus, there has been an ongoing search for markers of risk under medical management that would be more sensitive and incremental to symptoms and would not imply excess risk after surgery. Whereas left ventricular (LV) dysfunction (3,7 9) is considered useful for that

2 1962 Tribouilloy et al. JACC Vol. 54, No. 21, 2009 LV Size in Mitral Regurgitation November 17, 2009: Abbreviations and Acronyms AF atrial fibrillation EF ejection fraction LV left ventricle/left ventricular MR mitral regurgitation LVESD left ventricular end-systolic diameter purpose, the best therapeutic approach in patients with severe MR is still a subject of debate (6,10,11). In that regard, reduced preoperative LV ejection fraction (EF) is a predictor of mortality and post-operative LV dysfunction in patients with chronic MR (12 15) and is now included as a class I indication for surgery in guidelines (7,8). Conversely, whereas left ventricular end-systolic diameter (LVESD) is also included as an indication for surgery in guidelines, data on the prognostic importance of LVESD are scarce (7,15 20). To our knowledge, there are no data on the impact of LVESD on long-term survival after diagnosis in patients with severe organic MR. This probably explains the discordance between European and American guidelines with regard to the LVESD cut-off used for the timing of mitral valve repair, especially in asymptomatic patients with severe MR (7,8). The MIDA (Mitral Regurgitation International Database) registry was set up specifically as a multicenter study of the medical and surgical outcome of MR in routine practice. We identified retrospectively consecutive patients with echocardiographically diagnosed flail leaflet (4), a diagnosis usually associated with severe MR (8,21). The aims of the present analysis were to study the association of LVESD with survival after diagnosis and to analyze whether this association applied to follow-up under conservative management and after surgical correction of MR. Methods Study design. The MIDA registry was assembled by systematically merging the consecutive experience with MR due to flail leaflets of 5 centers: 4 tertiary centers in Europe (University Hospitals in Amiens and Marseille, France, and Bologna and Modena, Italy), and 1 center in the U.S. (Mayo Clinic, Rochester, Minnesota). Preliminary data from the European registry have been previously published (4). The process of forming each center s dataset involved retrospective identification of consecutive patients diagnosed with MR due to flail leaflet since inception of the echocardiographic database. Echocardiographic variables were obtained by download of standardized measurements, prospectively entered in the databases (4). We obtained institutional review board authorizations before conducting the study. The study was conducted in accordance with institutional policies, national legal requirements, and the revised Helsinki declaration. Patients were screened for the study if they had degenerative MR with flail leaflet diagnosed by 2-dimensional echocardiography between 1980 and Specific eligibility criteria (4) were: 1) presence of echocardiographically diagnosed flail leaflet; 2) availability of a comprehensive clinical/instrumental evaluation at the time of baseline echocardiography; 3) exclusion of ischemic MR; and 4) absence of significant concomitant aortic valve disease, congenital diseases, mitral stenosis, and previous valve surgery. Patients were excluded if they denied authorization for research participation. A comorbidity index summating the patient s individual comorbidities was calculated (22) and atrial fibrillation (AF) at baseline was determined by electrocardiogram. A total of 861 patients were enrolled in the registry. For the present analysis, we considered only patients in whom a measurement of LVESD was available (n 739). Echocardiography. Transthoracic echocardiograms were performed within routine clinical practice, using standard methods (4). Left ventricular dimensions were assessed from parasternal long-axis views by 2-dimensional guided M-mode using the leading edge methodology at end diastole and end systole. Severity of MR was assessed semiquantitatively on a scale from 1 to 4 by Doppler echocardiography (4). Diagnosis of flail leaflet was based on the failure of leaflet cooptation, with rapid systolic movement of the involved leaflet tip in the left atrium (3,4). Echocardiograms were used as collected at the time of the index echocardiography, without subsequent modification. Follow-up. Follow-up collection was complete in each center for 95% of enrolled patients. The main end point was survival after diagnosis starting at baseline echocardiographic evaluation and reaching up to last follow-up under medical management (censored at surgery). Other end points were cardiac mortality, overall survival encompassing medical and surgical management and post-operative survival in patients who underwent surgery. During follow-up, patients were monitored by their personal physicians. Events were ascertained by clinical interviews and/or by telephone calls to physicians, patients, and (if necessary) next of kin. Autopsy records and death certificates were consulted for attribution of causes of death. Statistical analysis. Continuous variables were expressed as mean 1 SD and compared with Student t tests. Categorical variables were summarized as frequency percentages and analyzed by chi-square tests. For the analysis of outcome under conservative treatment, data were censored at the time of cardiac surgery, if performed. The entire follow-up was used to analyze outcomes under conservative and surgical treatment. Event rates 1 SE were estimated according to the Kaplan-Meier method and compared using a 2-sided log-rank test. Univariate and multivariable analyses of time to events were performed using Cox proportional hazards models with LVESD as an independent variable in continuous and categorical format: cut-off value of 40 mm based on U.S. recommendations (7). For multivariable analyses of mortality we used predefined Cox proportional hazards multivariable models that included covariates considered of potential prognostic impact (age, sex, comorbidity index, symptoms at baseline, AF at baseline, and EF). We also conducted analyses stratified by

3 JACC Vol. 54, No. 21, 2009 November 17, 2009: Tribouilloy et al. LV Size in Mitral Regurgitation 1963 presence or absence of baseline characteristics representing current class I indications for surgery (symptoms or EF 60%) with measurement of statistical interaction between those characteristics and LVESD in predicting mortality. The proportional hazards assumption was confirmed using statistics and graphs based on the Schoenfeld residuals. For continuous variables, the assumption of linearity was assessed by plotting residuals against independent variables. We used penalized smoothing splines (P-splines) to illustrate the association of LVESD as a continuous variable and the risk of overall mortality (23). The effect of surgery on the outcome was analyzed as a time-dependent covariate in a Cox multivariable model with the use of data from the entire follow-up. A significance level of 0.05 was assumed for all statistical tests. All p values are results of 2-tailed tests. Data were analyzed with SPSS version 13.0 (SPSS Inc., Chicago, Illinois) and S-Plus version 8.0 (Insightful Inc., Seattle, Washington). Results Baseline characteristics and management. The baseline characteristics of the 739 patients with organic MR overall and according to LVESD are presented in Table 1. Ninetyfive percent of patients had grade 3 to 4 MR by Doppler echocardiography. Flail leaflet was attributable to a degenerative process in 661 patients (89%) and to infective endocarditis in the remaining 78 (11%). Eight percent of patients (n 60) had a history of clinical coronary artery disease. Of the total patients, 293 (40%) were in New York Heart Association functional class I and EF was 60% in 591 patients (80%). The mean LVESD was mm (median 36 mm; interquartile range 31 to 40 mm). In 545 patients (73.7%), LVESD was 40 and 40 mm in the remaining 194 patients (26.3%). Patients with larger LVESD were mostly men, had more severe symptoms and a higher Baseline Table 1Characteristics Baseline Characteristics of the 739 Patients of the 739 With Patients OrganicWith Mitral Organic Regurgitation Mitral Regurgitation Due to Flail Leaflets Due to Flail According LeafletstoAccording LVESD to LVESD All Patients (n 739) LVESD <40 mm (n 545) LVESD >40 mm (n 194) p Value Characteristics Age, yrs Male sex 69.7 (515) 63.9 (348) 86.1 (167) NYHA functional class I 39.6 (293) 41.9 (228) 33.5 (65) 0.16 II 29.5 (218) 28.1 (153) 33.5 (65) III 22.7 (168) 21.7 (118) 25.8 (50) IV 8.0 (59) 8.3 (45) 7.2 (14) Cardiac history and risk factors Coronary artery disease 8.1 (60) 7.3 (40) 10.3 (20) 0.19 Hypertension 34.5 (255) 36.2 (197) 29.9 (58) 0.11 Diabetes mellitus 0.9 (7) 0.9 (5) 1.0 (2) 0.89 Smoking 34.5 (255) 32.3 (176) 40.9 (79) 0.03 Dyslipidemia 30.2 (223) 31.4 (169) 28.1 (54) 0.39 Infective endocarditis 10.6 (78) 10.5 (57) 10.8 (21) 0.89 Charlson comorbidity index Atrial fibrillation 20.7 (153) 17.4 (95) 29.9 (58) Echocardiographic data Flail Posterior leaflet 81.9 (605) 83.3 (454) 77.8 (151) 0.23 Anterior leaflet 6.8 (50) 6.2 (34) 8.2 (16) Both leaflets 10.9 (81) 10.1 (55) 13.4 (26) Nonspecified 0.4 (3) 0.3 (2) 0.5 (1) Left ventricular end-diastolic diameter, mm Left ventricular end-systolic diameter, mm Ejection fraction, % Left atrial diameter,* mm Systolic pulmonary artery pressure, mm Hg Medical therapy Angiotensin-converting enzyme inhibitors 48.0 (355) 48.9 (266) 46.8 (89) 0.63 Beta-blockers 17.5 (129) 17.8 (97) 16.8 (32) 0.77 Diuretics 43.4 (320) 40.6 (221) 52.1 (99) Warfarin 29.0 (214) 24.8 (135) 41.6 (79) Antiplatelet agents 24.6 (182) 24.3 (132) 26.3 (50) 0.57 Values are mean SD or % (n). *Data available for 690 patients (93.4%). Data available in 422 patients (57.1%). LVESD left ventricular end-systolic diameter; NYHA New York Heart Association.

4 1964 Tribouilloy et al. JACC Vol. 54, No. 21, 2009 LV Size in Mitral Regurgitation November 17, 2009: Figure 1 Overall Survival According to LVESD in Patients With Organic MR Overall survival according to left ventricular end-systolic diameter (LVESD) in patients with organic mitral regurgitation (MR): (A) conservative management (B) medical and surgical treatment. frequency of AF, but no difference in comorbidity index (Table 1). Management was solely conservative in 187 patients (25.3%) and was medical followed by surgery in 552 patients (74.7%). Mean duration of follow-up with conservative treatment was years. Mean overall duration of follow-up was years. Outcome in conservatively managed patients. In patients managed conservatively, 74 deaths were recorded. Overall survival rates of 1, 5, and 10 years were 97 1%, 80 3%, and 58 5%, respectively. The 10-year survival rate was higher for patients with LVESD 40 mm than for patients with LVESD 40 mm (Fig. 1A). Older age, higher New York Heart Association functional class, coronary artery disease, comorbidity index, AF at baseline, EF, and diuretic use were also univariate predictors of mortality (all p 0.03). End-diastolic LV diameter was not associated with increased mortality (p 0.16). After adjustment for age, sex, and comorbidity, LVESD was independently associated with mortality (adjusted hazard ratio [HR]: 2.15, 95% confidence interval [CI]: 1.26 to 3.66 for LVESD 40 mm vs. LVESD 40 mm, and HR: 1.08, 95% CI: 1.04 to 1.12 per 1-mm increment in LVESD) (Table 2). Further adjustment for symptoms, AF and EF did not influence these independent relationships (Table 2). To estimate the character of the relationship between LVESD and the risk of overall mortality under conservative management, we used spline functions for LVESD (Fig. 2). In multivariable analysis, there was no increase in mortality risk with increasing LVESD when it remained 40 mm (adjusted HR: 1.03, 95% CI: 0.95 to 1.10 per 1-mm LVESD increment, p 0.48). With LVESD 40 mm, there was a steep increase in mortality risk with increasing LVESD (adjusted HR: 1.15, 95% CI: 1.04 to 1.27 per 1-mm LVESD increment, p 0.007). Compared with patients with LVESD 40 mm, those with LVESD between 40 and 45 mm (adjusted HR: 1.89, 95% CI: 0.98 to 3.4, p 0.058), as well as those with LVESD 45 mm (adjusted HR: 3.7, 95% CI: 1.7 to 7.3, p 0.002) displayed excess mortality under medical management. In the subgroup with symptoms or EF 60% at baseline (n 320), the mean follow-up under medical management was years and the delay between diagnosis and surgery was explained by improvement in symptoms under medical management. The risk related to LVESD was identical in these patients with potential guideline surgical indication Relative With Organic RiskRelative Mitral of Overall Regurgitation Risk Death of Overall anddue Death to From Flail andleaflets: Cardiac Death From Causes Results Cardiac Associated of Cox Causes Univariate With Associated LVESD and Multivariable With in Patients LVESD Analyses in Patients Table 2 With Organic Mitral Regurgitation Due to Flail Leaflets: Results of Cox Univariate and Multivariable Analyses Medical Management* Medical and Surgical Management Post-Operative Outcome* HR (95% CI) p Value HR (95% CI) p Value HR (95% CI) p Value LVESD 40 mm Overall death Adjusted for age, sex, and comorbidity 2.15 ( ) ( ) ( ) Adjusted for age, sex, comorbidity, symptoms, EF, and AF 1.95 ( ) ( ) ( ) Death from cardiac causes Adjusted for age, sex, and comorbidity 2.87 ( ) ( ) ( ) Adjusted for age, sex, comorbidity, symptoms, EF, and AF 3.09 ( ) ( ) ( ) 0.04 Per 1-mm increment in LVESD Overall death Adjusted for age, sex, and comorbidity 1.08 ( ) ( ) ( ) Adjusted for age, sex, comorbidity, symptoms, EF, and AF 1.07 ( ) ( ) ( ) 0.04 Death from cardiac causes Adjusted for age, sex and comorbidity 1.11 ( ) ( ) ( ) Adjusted for age, sex, comorbidity, symptoms, EF, and AF 1.13 ( ) ( ) ( ) 0.01 *Analyses of outcome under medical management and post-operative outcome included age, sex, Charlson comorbidity index, presence of symptoms at baseline, AF at baseline, and EF as covariates. Analysis of outcome with medical and surgical treatment included the previously mentioned covariates and surgery as a time-dependent covariate. AF atrial fibrillation; CI confidence interval; EF ejection fraction; HR hazard ratio; other abbreviations as in Table 1.

5 JACC Vol. 54, No. 21, 2009 November 17, 2009: Tribouilloy et al. LV Size in Mitral Regurgitation 1965 Figure 2 Association Between LVESD and the Risk of Overall Mortality Under Conservative Management Hazard ratio (solid line) and 95% confidence intervals were estimated in a Cox multivariable model with left ventricular end-systolic diameter (LVESD) represented as a spline function. at baseline (adjusted HR: 1.07, 95% CI: 1.01 to 1.11 per 1-mm LVESD increment, p 0.02) and in those with neither symptoms nor lowered EF (adjusted HR: 1.07, 95% CI: to 1.14 per 1-mm LVESD increment, p 0.035). There was no interaction between presence of symptoms or EF 60% and the effect of LVESD on survival (p 0.69). In patients in whom body surface area was available (n 555), an LVESD 22 mm/m 2 was associated with excess mortality under medical management (adjusted HR: 2.03, 95% CI: 1.06 to 3.89, p 0.03). There was an increase in mortality risk with increasing LVESD when LVESD was 22 mm/m 2 (adjusted HR: 1.12, 95% CI: 1.01 to 1.23 per 1-mm LVESD increment, p 0.01), but not when it remained 22 mm/m 2 (p 0.62). The model with unadjusted LVESD was slightly but significantly superior to that with LVESD normalized (p 0.045) so that the unadjusted value was used as the main independent variable in our analysis. The relationship between LVESD 40 mm and mortality under medical management was unchanged after excluding patients with moderate MR (adjusted HR: 2.09, 95% CI: 1.03 to 4.26, p 0.04) or patients with history of infective endocarditis (adjusted HR: 1.88, 95% CI: 1.01 to 2.95, p 0.044). During conservative follow-up, 49 deaths of cardiac causes occurred. Causes of death were LV dysfunction (n 31, 64%), unexplained sudden death (n 14, 28%), thromboembolism (n 2, 4%), myocardial infarction (n 1, 2%), and infective endocarditis (n 1, 2%). Cardiac mortality rates for 1, 5, and 10 years were 2 0.7%, 13 2%, and 29 5%, respectively, for the entire cohort. The 10-year cardiac mortality rate was significantly lower for patients with LVESD 40 mm (Fig. 3A). In multivariable analysis, a greater LVESD was independently predictive of death from cardiac causes (Table 2). There was a significant increase in cardiac mortality with increasing LVESD when it was 40 mm (HR: 1.16, 95% CI: 1.05 to 1.28 per 1-mm LVESD increment, p 0.004), but not when it remained 40 mm (p 0.20). Outcome with medical and surgical treatment. Mitral valve surgery was eventually performed in 552 patients (75%) on the basis of the following indications: dyspnea/ congestive heart failure in 371 patients (67%), patient and/or physician preference in 111 (20%), LV dilation in 17 (3%), infective endocarditis in 18 (3%), AF in 1 (0.2%), and miscellaneous reasons in the remaining 34 patients (6%). The mitral valve was repaired in 78% of patients and replaced in 22%. In 86 patients (15.6%), a coronary artery bypass graft was also performed during mitral valve surgery. The rate of cardiac surgery was %, %, and % at 1, 5, and 10 years, respectively, after diagnosis. During follow-up with medical and surgical treatment, 162 deaths were recorded. Overall survival rates for 1, 5, and 10 years were %, %, and 69 2%, respectively. The 10-year survival rate was higher for patients with LVESD 40 mm than for patients with LVESD 40 mm (Fig. 1B). Older age, higher New York Heart Association functional class, coronary artery disease, history of infective endocarditis, comorbidity index, AF at baseline, EF, and diuretic use were also univariate predictors of overall mortality (all p 0.03). End-diastolic LV diameter was not a predictor of overall mortality (p 0.76). In multivariable analysis, LVESD was independently predictive of mortality after diagnosis (adjusted HR: 1.86, 95% CI: 1.24 to 2.80 for LVESD 40 mm vs. LVESD 40 mm, and adjusted HR: 1.04, 95% CI: 1.02 to 1.07 per 1-mm LVESD increment) (Table 2). Patients with LVESD 22 mm/m 2 had poorer survival with medical and Figure 3 Survival Free of Cardiac Death According to LVESD in Patients With Organic MR Survival free of cardiac death according to LVESD in patients with organic MR: (A) conservative management (B) medical and surgical treatment. Abbreviations as in Figure 1.

6 1966 Tribouilloy et al. JACC Vol. 54, No. 21, 2009 LV Size in Mitral Regurgitation November 17, 2009: surgical management (adjusted HR: 1.73, 95% CI: 1.14 to 2.60, p 0.003). The relationship between LVESD 40 mm and mortality was still significant after excluding patients with moderate MR (p 0.003) or patients with history of infective endocarditis (p 0.02). During follow-up, cardiac mortality was recorded in 103 cases. Cardiac mortality rates for 1, 5, and 10 years were 2 0.5%, 10 1%, and 21 2%, respectively, for the entire cohort. The 10-year cardiac mortality was significantly lower for patients with LVESD 40 mm (Fig. 3B). In multivariable analysis, LVESD as continuous and categorical variable was independently predictive of cardiac death (Table 2). Of the 552 patients in whom mitral valve surgery was performed, 88 died during post-operative follow-up ( years). In multivariable analysis, LVESD was an independent predictor of post-operative overall death and death from cardiac causes (Fig. 4, Table 2). Cox proportional hazards analysis with surgery as a time-dependent variable showed that surgery was associated with reduced subsequent mortality (adjusted HR: 0.62, 95% CI: 0.45 to 0.86, p ). There was no significant interaction between LVESD 40 mm and the magnitude of survival benefit after surgery (p 0.20), which was observed in patients with LVESD 40 mm (adjusted HR: 0.42, 95% CI: 0.24 to 0.73, p 0.002) and with LVESD 40 mm (adjusted HR: 0.65, 95% CI: 0.44 to 0.96, p 0.028). Discussion This study is the first report of the relationship between LVESD and survival from diagnosis under medical management and after mitral surgery in MR due to flail leaflets. Figure 4 Adjusted Post-Operative Overall Survival According to LVESD in Operated Patients With Organic MR Adjusted post-operative overall survival according to LVESD in operated patients with organic MR. Abbreviations as in Figure 1. Our results show that LVESD is independently predictive of survival in patients with organic MR under conservative management in routine clinical practice. The effect of LVESD was powerful because after adjustment for age, sex, comorbidity, symptoms, AF, and EF, each 1-mm LVESD increment was associated with impressive 7% increase in overall mortality and 13% increase in cardiac mortality. Thus, LVESD 40 mm ( 22 mm/m 2 ) was associated with approximately doubling of the risk of overall mortality and tripling of the risk of cardiac death under medical management, irrespective of characteristics that are now considered class I indications for surgery (symptoms, EF 60%). Moreover, our results show that LVESD 40 mm ( 22 mm/m 2 ) is not just predictive of excess mortality under medical management with mortality increasing linearly above the cut-off of 40 mm, but also is an independent determinant of lower survival after surgical correction of MR despite the fact that surgery was associated with marked mortality reduction. Thus, whereas surgery is required in patients with LVESD 40 mm, it is preferable to indicate mitral surgery before this threshold is reached to avoid the excess post-operative mortality associated with LVESD 40 mm. The management of MR is disputed. The general agreement is that patients with overt symptoms should undergo prompt mitral surgery (7,8) because of their high risk under medical management (3). Symptomatic patients incur excess mortality after surgery (9,14). The other general consensus (7,8) is that asymptomatic patients with severe organic MR and overt LV dysfunction should be considered for mitral surgery. Consequently, in routine practice, simple and reproducible echocardiographic parameters of LV systolic function with definite prognostic value are useful to discuss the best timing for mitral surgery. Interpretation of systolic function parameters in MR is complex (7,14,24). It is widely agreed that LV dysfunction may be concealed behind a normal EF because the loading conditions are profoundly modified (18). Over time, patients with severe chronic MR develop an irreversible impairment in LV systolic function and reduced EF is a sign of overt LV dysfunction. With reduced EF, it has been observed that mortality under medical management (3) and post-operative occurrence of LV dysfunction, congestive heart failure, and death are all increased in patients with organic MR (12 15). End-systolic LV characteristics are considered less preload-dependent (7,19) than EF. Data on the prognostic importance of LVESD are scarce (7), generally limited to small surgical studies without data on mortality (15 18). These reports indicated that LVESD was associated with pre-operative (18) or post-operative LV function (15,17,19). However, all these surgical series remained small or mingled organic and functional MR (15). Only a small cohort of patients with severe MR suggested that LVESD was associated with progression of symptoms or LV dysfunction during conservative follow-up (20). The guidelines of 1998 endorsed a threshold of LVESD 45 mm for surgery based

7 JACC Vol. 54, No. 21, 2009 November 17, 2009: Tribouilloy et al. LV Size in Mitral Regurgitation 1967 on a series of 61 patients (19). However, this study group with rheumatic MR was extremely young and experienced few post-operative events (19) so that applicability of the 45-mm threshold is questionable in older patients with degenerative MR (17,18,20). U.S. guidelines recently endorsed a surgical threshold of LVESD 40 mm (7) whereas European guidelines maintain a threshold of 45 mm (8). Currently, mitral valve repair is the preferred surgical procedure for MR. This approach is supported by its low risk and excellent long-term results of the procedure, but it requires the precise identification of subgroups of patients at high risk under conservative management. The population included in the current study has a homogenous diagnosis of flail leaflet. Although the presence of a flail leaflet might not be systematically associated with severe MR (21), the ventricular dilation observed in this study is consistent with that observed in previous surgical series and denotes marked volume overload. The relative risk of mortality under medical management increased linearly with LVESD above the cut-off of 40 mm. Consistent with previous studies, mitral surgery was associated with a significant reduction in the risk of death. However, among operated patients, LVESD 40 mm remained independently predictive of adverse outcome. The judgment on severe MR was relatively uniform (95% of patients). There was no interaction between MR severity and outcome prediction of LVESD (p for interaction 0.49). Despite the size of this series, we cannot assume that each possible subgroup (i.e., subgroup with moderate MR) follows the same rules. Although in this study there was no interaction between sex and LVESD 40 mm for predicting survival (p 0.21), recent data show that among patients with severe organic MR, women have higher mortality and lower surgery rates than men (25). This is probably due to the fact that patients with small body size are allowed to develop greater relative chamber dilation. Our data show that a LVESD 22 mm/m 2 is associated with excess mortality. This threshold can be used, particularly when discussing surgery for severe organic MR in patients with small body size. Study strengths and limitations. A limitation of the present study was that whereas echocardiographic data were prospectively collected, clinical and nonechocardiographic data were obtained by review of medical records. In multivariable analysis, the center in which patients were enrolled did not affect survival (p 0.16) whereas LVESD remained predictive of outcome. There was no interaction between the origin of patients (Europe vs. U.S.) and LVESD 40 mm for predicting survival after diagnosis (p 0.14). Among included cases, there were patients who presented with characteristics now known as class I surgical indications by guidelines. Surgery was delayed because physicians judged that patients were well initially with medical management. Stratification by class I characteristics does not affect the results of our study and shows that LVESD remains independently predictive of survival. This study (and the MIDA database) used flail leaflet as a surrogate for severe MR. Although prominent flail usually is associated with severe MR, not all flail leaflets are associated with severe MR, and not all patients in this study were felt, at the time of echocardiography interpretation, to have severe MR. However, we believe that these findings can be extrapolated to most patients with chronic severe MR. We acknowledge that during the last 2 decades 2-dimensional transthoracic echocardiographic imaging has undergone substantial evolution allowing more accurate detection of flail leaflets. The inclusion period (1980 to 1991 vs to 2004) had no influence on the prognostic impact of LVESD on outcome (p for interaction 0.29). There was no change in measured LVESD over the years of the study (R , p 0.69). Clinical Implications Our multicenter study demonstrates that LVESD is a powerful predictor of survival in patients with pure organic MR due to flail leaflets. Therefore, it is essential to measure LV diameters in all patients with MR and to use LVESD for clinical decision making. Patients with LVESD 40 mm ( 22 mm/m 2 ) do not have a benign outcome, exhibit increased risk of death under conservative management, and should be promptly considered for mitral surgery, because surgery considerably reduces mortality. The pejorative effect of LVESD 40 mm ( 22 mm/m 2 ) is also observed after the surgical correction of MR. Therefore, in our opinion, in patients with severe MR due to flail leaflets valve surgery should be considered, even in the absence of symptoms, before the end-systolic diameter exceeds 40 mm. Conversely, patients with smaller end-systolic dimensions and no symptoms incur low mortality risk under medical management and may be initially followed medically if there are no other markers of high risk. In this context, serial LVESD measurements represent an objective, valuable, and easily measurable tool for discussing the optimal timing for surgery. Reprint requests and correspondence: Prof. Christophe Tribouilloy, INSERM, ERI 12, and University Hospital, Department of Cardiology, Avenue René Laënnec, Amiens Cedex 1, France. tribouilloy.christophe@chu-amiens.fr. REFERENCES 1. Iung B, Baron G, Butchart EG, et al. A prospective survey of patients with valvular heart disease in Europe: the Euro Heart Survey on Valvular Heart Disease. Eur Heart J 2003;24: Nkomo VT, Gardin JM, Skelton TN, Gottdiener JS, Scott CG, Enriquez-Sarano M. Burden of valvular heart diseases: a populationbased study. Lancet 2006;368: Ling LH, Enriquez-Sarano M, Seward JB, et al. Clinical outcome of mitral regurgitation due to flail leaflet. N Engl J Med 1996;335: Grigioni F, Tribouilloy C, Avierinos JF, et al. Outcomes in mitral regurgitation due to flail leaflets. J Am Coll Cardiol Img 2008;1: Rosen SE, Borer JS, Hochreiter C, et al. Natural history of the asymptomatic/minimally symptomatic patient with severe mitral re-

8 1968 Tribouilloy et al. JACC Vol. 54, No. 21, 2009 LV Size in Mitral Regurgitation November 17, 2009: gurgitation secondary to mitral valve prolapse and normal right and left ventricular performance. Am J Cardiol 1994;74: Enriquez-Sarano M, Avierinos JF, Messika-Zeitoun D, et al. Quantitative determinants of the outcome of asymptomatic mitral regurgitation. N Engl J Med 2005;352: Bonow RO, Carabello BA, Chatterjee K, et al. ACC/AHA 2006 practice guidelines for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1998 Guidelines for the Management of Patients with Valvular Heart Disease). J Am Coll Cardiol 2006;48:e Vahanian A, Baumgartner H, Bax J, et al. Guidelines on the management of valvular heart disease. Eur Heart J 2007;28: Tribouilloy CM, Enriquez-Sarano M, Schaff HV, et al. Impact of preoperative symptoms on survival after surgical correction of organic mitral regurgitation: rationale for optimizing surgical indications. Circulation 1999;99: Braunberger E, Deloche A, Berrebi A, et al. Very long-term results (more than 20 years) of valve repair with Carpentier s techniques in nonrheumatic mitral valve insufficiency. Circulation 2001;104 Suppl 1:I Rosenhek R, Rader F, Klaar U, et al. Outcome of watchful waiting in asymptomatic severe mitral regurgitation. Circulation 2006;113: Crawford MH, Souchek J, Oprian CA, et al. Determinants of survival and left ventricular performance after mitral valve replacement. Department of Veterans Affairs Cooperative Study on Valvular Heart Disease. Circulation 1990;8: Phillips HR, Levine FH, Carter JE, et al. Mitral valve replacement for isolated mitral regurgitation: analysis of clinical course and late postoperative left ventricular ejection fraction. Am J Cardiol 1981;48: Enriquez-Sarano M, Tajik AJ, Schaff HV, Orszulak TA, Bailey KR, Frye RL. Echocardiographic prediction of survival after surgical correction of organic mitral regurgitation. Circulation 1994;90: Enriquez-Sarano M, Tajik AJ, Schaff HV, et al. Echocardiographic prediction of left ventricular function after correction of mitral regurgitation: results and clinical implications. J Am Coll Cardiol 1994;24: Zile MR, Gaasch WH, Carroll JD, Levine HJ. Chronic mitral regurgitation: predictive value of preoperative echocardiographic indexes of left ventricular function and wall stress. J Am Coll Cardiol 1984;3: Matsumura T, Ohtaki E, Tanaka K, et al. Echocardiographic prediction of left ventricular dysfunction after mitral valve repair for mitral regurgitation as an indicator to decide the optimal timing of repair. J Am Coll Cardiol 2003;42: Flemming MA, Oral H, Rothman ED, Briesmiester K, Petrusha JA, Starling MR. Echocardiographic markers for mitral valve surgery to preserve left ventricular performance in mitral regurgitation. Am Heart J 2000;140: Wisenbaugh T, Skudicky D, Sareli P. Prediction of outcome after valve replacement for rheumatic mitral regurgitation in the era of chordal preservation. Circulation 1994;89: Krauss J, Pizarro R, Oberti PF, Falconi M, Cagide A. Prognostic implication of valvular lesion and left ventricular size in patients with chronic organic mitral regurgitation and normal left ventricular performance. Am Heart J 2006;152:1004.e Zoghbi WA, Enriquez-Sarano M, Foster E, et al. Recommendations for evaluation of the severity of native valvular regurgitation with two-dimensional and Doppler echocardiography. J Am Soc Echocardiogr 2003;16: Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987;40: Eilers PH, Marx BD. Flexible smoothing with B-splines and penalties. Stat Sci 1996;11: Gaasch WH, Roy MJ, Aurigemma GP. Managing asymptomatic patients with chronic mitral regurgitation. Chest 1995;108: Avierinos JF, Inamo J, Grigioni F, Gersh B, Shub C, Enriquez-Sarano M. Sex differences in morphology and outcomes of mitral valve prolapse. Ann Intern Med 2008;149: Key Words: mitral regurgitation y left ventricular size y survival y surgery. APPENDIX MIDA Investigators University of Amiens, France: C. Tribouilloy, D. Rusinaru, C. Szymanski, F. Trojette, G. Touati, J. P. Remadi, H. Poulain, T. Caus. University of Bologna, Italy: F. Grigioni, M. Bigliardi, A. Russo, E. Biagini, G. Piovaccari, M. Ferlito, A. Branzi, C. Savini, G. Marinelli, R. Di Bartolomeo. University of Marseille, France: J. F. Avierinos, L. Tafanelli, G. Habib, F. Collard, A. Riberi, D. Metras. University of Modena, Italy: A. Barbieri, F. Bursi, T. Grimaldi, A. Nuzzo, M. G. Modena. Mayo Clinic, Rochester Minnesota: M. Enriquez- Sarano, R. Suri, D. W. Mahoney. Go to to take the CME quiz for this article.