1. INTRODUCTION

Mitral regurgitation is the most common valvular heart disease, which may require surgical intervention. The feasibility of mitral valve surgery majorly depends on the etiology of severe mitral regurgitation,1 while subjects with primary mitral regurgitation (PMR) but not secondary mitral regurgitation (SMR) may get truly survival benefits after surgery.2–9 The indication of surgical intervention for severe PMR involves the presence of heart failure symptoms or left ventricular remodeling.10 In contrast, subjects with severe SMR would be suggested mitral valve surgery as a concomitant open-heart procedure.10 Moreover, the surgical risks related to age, comorbidities and left ventricular function may further affect the surgeon’s willingness to do the operations.11

Goel et al4 have demonstrated the dismal clinical outcomes of a total of 1095 unoperated patients with severe symptomatic mitral regurgitation. The majority of unoperated patients had SMR and the mortality rate elevated during the follow-up. In the meantime, the proportion of the survival patients hospitalized for heart failure increased from 41% in the first year to 90% by 5 years, which implied these patients were at greater risks of mortality.4,12 However, the comparison of clinical impact of surgery between different mechanisms of severe MR and the determinants of long-term survival who were unwilling to receive surgery has yet been elucidated. We therefore investigated the clinical outcomes of patients with either severe PMR or SMR, hospitalized for heart failure in a tertiary medical center. We further analyzed the prognostic factors of the unoperated patients in the present study.

2. METHODS 2.1. Study population

The study population was drawn from the HARVEST registry that patients hospitalized for heart failure in Taipei Veterans General Hospital between 2002 and 2012 were eligible. All the participants have received comprehensive echocardiographic studies, and those with severe mitral regurgitation, defined by an effective regurgitation orifice area (EROA) of ≥40 mm2 were enrolled in this analysis.13 Patients who were younger than 18 years old have prior open-heart surgery, active infectious diseases, active malignancies, or concomitant cardiac or coronary abnormalities required surgery were excluded. Data of demographic characteristics, comorbidities, biochemistry, and hemogram were obtained. Estimated glomerular filtration rate (eGFR) was calculated using the Chinese Modification of Diet in Renal Disease equation.14 An eGFR of <60 mL/min/1.73 m2 was defined as chronic kidney disease. Medications, including renin-angiotensin system inhibitors, β-blockers, and mineralocorticoid receptor antagonists, were recorded. The investigation was conformed to the principles outlined in the Declaration of Helsinki. The institutional review committee of Taipei Veterans General Hospital approved the use of the registry data for research purposes, and the informed consent was waived.

2.2. Echocardiography

The transthoracic echocardiographic study was conducted by experienced technicians according to the recommendations from the American Society of Echocardiography.15 Left atrial (LA) dimension, left ventricular internal diameter at end-diastole (LVIDd), and end-systole (LVIDs) were measured by M-mode. LVEF was calculated from the LV end-diastolic volume and end-systolic volume estimates by bi-plane Simpson’s method. Right ventricular systolic pressure (RVSP) was estimated using Doppler echocardiography by calculating trans-tricuspid pressure gradient during systole and right atrial pressure by the dimension and collapsibility of inferior vena cava.

The quantification of mitral regurgitation was measured by adopting the proximal isovelocity hemispheric surface area of the flow convergence on the ventricular side in apical four chamber view with an aliasing velocity of color Doppler of 30 cm/s. Severe mitral regurgitation was defined by an EROA of ≥40 mm.2,10 PMR is referred to the presence of excessive motions of mitral leaflets with or without ruptured chordae, causing significant regurgitation of blood during systolic phase. Otherwise, mitral leaflet morphology is normal in SMR, while leaflet tethering and/or annulus dilation hinder leaflet coaptation due to papillary muscle displacement or dysfunction.10

2.3. Statistical analysis

Categorical variables were expressed as percentages, and descriptive continuous variables were expressed as means ± standard deviations. Student’s t test was used to compare continuous variables and the χ2 test was used to compare categorical data.

Because of the heterogenous baseline characteristics between the operated and unoperated subjects, we performed 1:1 and 3:1 propensity score matching for PMR and SMR, respectively. A multivariate logistic regression analysis was conducted to evaluate the independent associations of the variables, which were significantly different between the operated and the unoperated subjects, with the decision of surgery. Propensity score was then estimated using a multivariate logistic regression model, including age, eGFR, LVIDs for PMR, and age, eGFR, LVEF and LA dimension for SMR. The calipers of 0.1 and 0.01, without replacement, were applied for PMR and SMR matching, respectively. The Kaplan–Meier survival curve analysis was used to assess the prognostic difference between the operated and unoperated subjects. Cox proportional hazards models were used to evaluate the determinants of the clinical outcomes.

All statistical analyses were performed on SPSS software (SPSS, version 24.0.0.0, IBM Corporation, Armonk, NY, USA). All tests were two-sided and a p value of <0.05 was considered statistically significant.

3. RESULTS 3.1. Baseline characteristics

A total of 1143 patients (mean age 66.4 ± 16.6 years; 65.1% men) with severe MR were analyzed, while 682 subjects (59.7%) of them had PMR and 461 had SMR. The baseline characteristics were demonstrated in Table 1. Patients with severe SMR were older, more male predominantly, and had lower eGFR and more comorbidities, including hypertension, diabetes, prior heart failure, coronary artery disease, stroke, and chronic obstructive pulmonary disease (COPD) than those with PMR. Patients with SMR also have greater LA dimension, LVIDd, LVIDs, and RVSP, but lower LVEF and EROA than those with PMR. In addition, subjects with SMR were treated with more renin-angiotensin system inhibitors, β-blockers, and mineralocorticoid antagonists. During a mean follow-up duration of 3.3 ± 2.7 years, 359 subjects with PMR (31.4%) and 236 subjects with SMR (55.1%) died. The Kaplan survival curve analysis suggested a better survival in patients with PMR than those with SMR (Fig. 1).

Table 1 - Characteristics of the study population, stratified by primary or secondary cause PMR SMR Overall (n = 1143) Subtotal (n = 682) Operated (n = 326) Unoperated (n = 356) Subtotal (n = 461) Operated (n = 32) Unoperated (n = 429) Age 66.4 ± 16.6 62.4 ± 17.2a 58.3 ± 15.6b 66.2 ± 17.7 72.4 ± 13.7 63.5 ± 12.5b 73.1 ± 13.5 Male, % 745 (65.1) 440 (64.5)a 212 (65.0) 228 (64.0) 305 (66.2) 26 (81.3) 279 (65.0) Creatinine, mg/dL 1.75 ± 1.75 1.38 ± 1.35a 1.3 ± 1.4 1.49 ± 1.2 2.22 ± 2.06 2.15 ± 2.85 2.22 ± 2.0 eGFR, mL/min/1.73 m2 59.8 ± 29.2 69.4 ± 27.6a 74.6 ± 26.2b 63.6 ± 28.1 46.7 ± 26.0 57.1 ± 24.9b 45.9 ± 25.9 Comorbidities, n (%)  Hypertension 419 (36.7) 213 (31.2)a 82 (25.2)b 131 (36.8) 206 (44.7) 5 (15.6)b 201 (46.9)  Diabetes mellitus 185 (16.2) 69 (10.1)a 27 (8.3) 42 (11.8) 116 (25.2) 3 (9.4)b 113 (26.3)  Coronary artery disease 331 (29.0) 137 (20.1)a 59 (18.1) 78 (21.9) 194 (42.1) 7 (21.9)b 187 (43.6)  Chronic kidney disease 111 (9.7%) 38 (5.6%)a 16 (4.9) 22 (6.2) 73 (15.8) 2 (6.3)b 71 (16.6)  Atrial fibrillation 125 (10.9) 73 (10.7) 29 (8.9) 44 (12.4) 52 (11.3) 3 (9.4) 49 (11.4)  Stroke 69 (6.0) 30 (4.4)a 12 (3.7) 18 (5.1) 39 (8.5) 1 (3.1) 38 (8.9)  COPD 204 (17.8) 95 (13.9) + 34 (10.4)b 61 (17.1) 109 (23.6) 5 (15.6) 104 (24.2) Echocardiogram  LVEF, % 56.4 ± 18 65.8 ± 11a 65.9 ± 11 65.8 ± 11 42.5 ± 17 48.8 ± 16b 42.0 ± 17   EF <50%, % 352 (30.8) 54 (7.9)a 27 (8.3) 27 (7.6) 298 (64.5) 19 (59.4) 279 (65.0)  LAD, mm 50.2 ± 10.5 49.9 ± 10.9a 50.5 ± 10.9 49.3 ± 11.0 50.6 ± 9. 58.8 ± 12.9b 50.0 ± 9.2  LVIDd, mm 59.1 ± 9.5 56.5 ± 8.0a 57.9 ± 8.2b 55.1 ± 7.6 63.0 ± 10.1 65.2 ± 10.3 62.8 ± 10.1  LVIDs, mm 39.0 ± 12.3 33.0 ± 7.7a 34.0 ± 8.1b 32.1 ± 7.2 47.8 ± 12.5 46.7 ± 18.2 47.9 ± 12.5  RVSP, mmHg 47.0 ± 19.9 45.4 ± 20.7a 46.2 ± 21.5 44.7 ± 19.9 49.5 ± 18.5 46.6 ± 18.2 49.7 ± 18.5  EROA, mm2 64.8 ± 40.8 72.5 ± 43.2a 78.8 ± 44.9b 66.9 ± 41.0 47.8 ± 28.3 50.6 ± 19.8 47.5 ± 29.0 Medication, n (%)  RAS inhibitors 573 (50.2) 315 (46.3)a 140 (42.9) 175 (49.2) 258 (56.0) 14 (43.8) 244 (56.9)  Beta-blockers 31 (2.7) 22 (3.2)a 10 (3.1) 12 (3.4) 9 (2.0) 0 (0.0) 9 (2.1)  MRA 267 (23.4) 116 (17.0)a 56 (17.2) 60 (16.9) 151 (32.8) 9 (28.1) 142 (33.1)

COPD = chronic obstructive pulmonary disease; eGFR = estimated glomerular filtration rate; EROA = effective regurgitant orifice area; LAD = left atrial dimension; LVEF = left ventricular ejection fraction; LVIDd = ventricular internal diameter at end-diastole; LVIDs = left ventricular internal diameter at end-systole; MRA = mineralocorticoid antagonist; PMR = primary mitral regurgitation; RAS = renin-angiotensin system; RVSP = right ventricular systolic pressure; SMR = primary mitral regurgitation.

ap value < 0.05, compared with SMR group.

bp value < 0.05, compared with the unoperated subjects.

Fig. 1:

The Kaplan–Meier survival curve analysis of the study population, stratified by PMR or SMR. PMR = primary mitral regurgitation; SMR = secondary mitral regurgitation.

There were 47.8% of PMR patients (n = 326) and 6.9% of SMR patients (n = 32) receiving mitral valve surgery. Among patients with PMR, the operated subjects were younger, had better eGFR, less hypertension or COPD, and greater LVIDs and EROA than the unoperated subjects (Table 1). Majority of the operated PMR subjects underwent mitral valve replacement, and only 88 patients (27%) underwent mitral valve repair (Table 2). The operated PMR patients deemed to have better survival than their counterpart (Fig. 2A). In the subpopulation with SMR, the operated subjects were younger, had better eGFR, and less hypertension, diabetes or coronary artery disease, and better LVEF than the unoperated subjects (Table 1). The distribution of the operated patients stratified by LVEF is illustrated in Fig. 3. Among subjects with preserved LVEF, there was only 7.9% of them would receive surgery. Among the operated SMR patients, 19 (59.3%) of them received mitral valve repair, 9 (21.9%) underwent mitral valve replacement, and 4 (12.5%) accepted heart transplantation (Table 2). The mortality rates were similar in SMR subjects with or without surgical intervention (Fig. 2B).

Table 2 - Types of surgical interventions Primary MR (n = 326) Secondary MR (n = 32) p Valve repairment, % 88 (27.0) 19 (59.3) 0.001 Valve replacement, % 238 (73.0) 9 (21.9) <0.001 Heart transplantation, % 0 (0.0) 4 (12.5) <0.001

MR = mitral regurgitation.

Fig. 2:

The Kaplan–Meier survival curve analysis of patients with primary mitral regurgitation (A) or secondary mitral regurgitation (B), stratified by surgery. PMR = primary mitral regurgitation; SMR = secondary mitral regurgitation.

Fig. 3:

The distribution of patients with secondary mitral regurgitation and mitral valve surgery, stratified by LVEF. LVEF = left ventricular ejection fraction; SMR = secondary mitral regurgitation.

The propensity score matching cohort is demonstrated in Table 3. Among the matched PMR subpopulation, the operated subjects remained a bit younger, has less hypertension or COPD, larger EROA, and more prescriptions of RAS inhibitors. Mitral valve surgery was associated with better long-term survival in PMR patients (Fig. 4A, log rank p = 0.024). The Cox proportional hazards model showed mitral valve surgery was significantly associated with less mortality (hazard ratio, 0.63; 95% confidence intervals, 0.40-0.99), after accounting for age, EROA, and the presence of hypertension and COPD in the matched PMR cohort.

Table 3 - Characteristics of the propensity score matching cohorts, stratified by primary or secondary cause PMR SMR Operated (n = 233) Unoperated (n = 233) Operated (n = 32) Unoperated (n = 81) Age, y 63.7 ± 13.9a 66.6 ± 15.7 63.1 ± 12.6 67.9 ± 14.5 Male, % 163 (70.0) 159 (68.2) 27 (84.4) 57 (70.4) Creatinine, mg/dL 1.26 ± 1.31 1.33 ± 1.20 1.99 ± 2.61 2.02 ± 2.28 eGFR, mL/min/1.73 m2 71.1 ± 24.8 68.4 ± 27.6 57.2 ± 24.9 54.0 ± 27.8 Comorbidities, n (%)  Hypertension 67 (28.8)a 97 (41.6) 5 (15.6)a 36 (44.4)  Diabetes mellitus 26 (11.2) 30 (12.9) 3 (9.4) 20 (24.7)  Coronary artery disease 50 (21.5) 54 (23.2) 7 (21.9) 30 (37.0)  Chronic kidney disease 10 (4.3) 16 (6.9) 2 (6.3) 9 (11.1)  Atrial fibrillation 25 (10.7) 25 (10.7) 3 (9.4) 10 (12.3)  Stroke 9 (3.9) 14 (6.0) 1 (3.1) 5 (6.2)  COPD 30 (12.9)a 47 (20.2) 5 (15.6) 17 (21.0) Echocardiogram  LVEF, % 66.4 ± 10.7 65.7 ± 10.8 48.8 ± 16.8 46.4 ± 17.6   EF< 50%, % 18 (7.7) 18 (7.8) 19 (59.4) 43 (53.1)  LAD, mm 51.0 ± 10.9 49.6 ± 11.4 58.9 ± 13.0 55.6 ± 8.4  LVIDd, mm 57.5 ± 8.3 55.4 ± 7.8 65.1 ± 10.3 63.3 ± 10.5  LVIDs, mm 33.6 ± 8.0 32.7 ± 7.8 46.3 ± 13.4 46.5 ± 13.4  RVSP (mmHg) 51.3 ± 20.5 47.7 ± 18.1 51.6 ± 15.3 54.7 ± 16.2  EROA, mm2 74.2 ± 34.6a 62.7 ± 34.1 50.7 ± 19.9 45.7 ± 14.8 Medication, n (%)  RAS inhibitors 102 (43.8)a 130 (55.8) 14 (43.8) 47 (58.0)  Beta-blockers 7 (3.0) 9 (3.9) 0 (0.0) 2 (2.5)  MRA 37 (15.9) 46 (19.7) 9 (28.1) 27 (33.3)

ACEi or ARB = angiotensin converting enzyme inhibitor or angiotensin II receptor blocker; COPD = chronic obstructive pulmonary disease; eGFR = estimated glomerular filtration rate; EROA = effective regurgitant orifice area; LAD = left atrial diameter; LVEF = left ventricular ejection fraction; LVIDd = ventricular internal diameter at end-diastole; LVIDs = left ventricular internal diameter at end-systole; PMR = primary mitral regurgitation; RVSP = right ventricular systolic pressure; SMR = secondary mitral regurgitation.

ap value < 0.05, compared with the unoperated subjects.

Fig. 4:

The Kaplan–Meier survival curve analysis of the propensity score matching cohorts with primary mitral regurgitation (A) or secondary mitral regurgitation (B), stratified by surgery. PMR = primary mitral regurgitation; PSM = propensity score matching; SMR = secondary mitral regurgitation.

In the matched SMR cohort, the operated subjects still had less hypertension than the unoperated patients. However, the Kaplan–Meier survival curve analysis demonstrated similar survival probabilities of them (Fig. 4B, log rank p = 0.097). The multivariate Cox regression analysis suggested the surgical intervention in the matched SMR cohort did not correlate with better long-term survival (HR, 0.67, 95% CIs, 0.35-1.28), after accounting for the presence of hypertension.

Among the 356 unoperated PMR patients, age, RVSP, and presence of hypertension, diabetes and COPD were positively and eGFR was negatively associated with all-cause mortality (Table 4). The multivariate Cox regression analysis indicated the increased age and RVSP, decreased eGFR, and presence of diabetes were independent risk factors of mortality. In contrast, age, eGFR, coronary artery disease, LVEF, and RVSP were all related to the survival in the unoperated SMR patients (Table 4). In the multivariate Cox regression analysis, increased age and RVSP, and decreased eGFR and LVEF remained associated with all-cause mortality.

Table 4 - Predictors of mortality in the unoperated subjects PMR (n = 356) SMR (n = 429) Univariate Multivariate Univariate Multivariate Age, y 1.07 (1.05-1.09)a 1.05 (1.02-1.08)a 1.02 (1.01-1.03)a 1.02 (1.01-1.03)a Gender 0.98 (0.62-1.55) 1.07 (0.80-1.43) eGFR, mL/min/1.73 m2 0.97 (0.96-0.98)a 0.98 (0.96-0.99)a 0.98 (0.98-0.99)a 0.98 (0.98-0.99)a Hypertension 1.96 (1.27-3.03)a 1.17 (0.70-1.95) 1.26 (0.97-1.65) Diabetes mellitus 2.04 (1.20-3.49)a 2.19 (1.21-3.95)a 1.25 (0.94-1.67) Coronary artery disease 1.15 (0.69-1.91) 1.52 (1.16-1.98)a 1.15 (0.86-1.55) Atrial fibrillation 1.53 (0.86-2.73) 0.80 (0.52-1.24) Stroke 1.72 (0.79-3.74) 1.30 (0.84-2.03) COPD 1.70 (1.02-2.82)a 0.76 (0.42-1.36) 1.25 (0.92-1.69) LVEF, % 0.81 (0.14-4.55) 0.36 (0.17-0.76)a 0.33 (0.14-0.77)a LAD, mm 1.02 (0.99-1.04) 0.99 (0.98-1.01) LVIDd, mm 0.98 (0.95-1.01) 1.00 (0.99-1.02) LVIDs, mm 0.99 (0.96-1.03) 1.01 (0.99-1.02) RVSP, mmHg 1.03 (1.02-1.04)a 1.02 (1.00-1.03)a 1.02 (1.01-1.03)a 1.01 (1.00-1.02)a EROA, mm2 0.98 (0.96-1.00) 0.99 (0.98-1.01)

The values are displayed as hazard ratio and 95% confidence intervals.

COPD = chronic obstructive pulmonary disease; eGFR = estimated glomerular filtration rate; EROA = effective regurgitant orifice area; LAD = left atrial diameter; LVEF = left ventricular ejection fraction; LVIDd = ventricular internal diameter at end-diastole; LVIDs = left ventricular internal diameter at end-systole; PMR = primary mitral regurgitation; RVSP = right ventricular systolic pressure; SMR = secondary mitral regurgitation.

ap < 0.05.

4. DISCUSSION

The present single-center study suggested patients with severe MR and heart failure were associated with dismal clinical outcomes, and the median survival of the study population was 33.6 months. While half of the PMR patients have undergone mitral valve surgery, only 6.9% of FMR patients would do so. When age, renal function, and lung disease were the determinants of operation among subjects with PMR, diabetes, coronary artery disease, and LA and ventricular sizes also affected the surgical decisions. However, mitral valve surgery might