Predictors and Clinical Impact of Functional Mitral Stenosis Induced by Restrictive Annuloplasty for Ischemic and Functional Mitral Regurgitation

Abstract

Background: There are few reports of the determinants of “functional” mitral stenosis in terms of a residual mitral valve (MV) pressure gradient >5 mmHg following restrictive mitral annuloplasty (RMA) or the effect on long-term outcome in patients with functional mitral regurgitation (MR).

Methods and Results: Serial cardiac catheterization and echocardiographic studies were performed in 55 patients with functional MR who underwent RMA using a 24/26-mm semi-rigid complete ring. The mean postoperative (1 month) catheter-measured MV gradient was 3.4±1.6 mmHg, which was independently associated with corresponding cardiac output [standardized partial regression coefficient (SPRC)=0.59] and indexed effective orifice area (SPRC=−0.25). Body surface area (BSA) had the greatest contribution to MV gradient (SPRC=0.38), followed by use of a 24-mm ring (SPRC=0.33) and hemodialysis (SPRC=0.26). Receiver-operating characteristic curve analysis demonstrated an optimal BSA cutoff value of 1.86 m² to predict post-MV stenosis (21% for <1.86 m² vs. 86% for ≥1.86 m², P=0.002). During follow-up (75±32 months), freedom from adverse events did not differ between patients with (n=16) and without (n=39) an MV gradient ≥5 mmHg (log-rank P=0.24).

Conclusions: Post-RMA MV gradient was determined not only by the degree of annular reduction but also by patients’ hemodynamic factors (e.g., cardiac output). Implantation of a 24/26-mm annuloplasty ring for patients with BSA ≥1.86 m² indicated a high likelihood of post-MV stenosis. However, mild MV stenosis did not adversely affect late outcome after RMA.

Functional mitral regurgitation (MR) is a common complication of both ischemic and non-ischemic advanced cardiomyopathy, and its presence in the setting of severe systolic left ventricular (LV) dysfunction is strongly associated with poor outcome.^1,2 Currently, a restrictive mitral annuloplasty (RMA) technique utilizing an undersized prosthetic ring is the preferred surgical option to treat moderate to severe functional MR.^3–8

Recently, there has been increasing interest in iatrogenic “functional” mitral stenosis by implantation of an undersized annuloplasty ring, which leads to an increased gradient across the mitral valve (MV) after valvular repair.^9–14 Magne et al noted a high incidence of functional MV stenosis, as well as higher pulmonary artery pressure (PAP) and worse functional capacity, in patients who underwent RMA.⁹ In addition, Kubota et al demonstrated that subvalvular tethering caused by LV remodeling is the primary mechanism of functional MV stenosis at the leaflet tip level.¹⁰ These findings led us to examine which patients will be placed at risk for developing an increased MV gradient following RMA. Furthermore, it remains unknown whether postoperative functional MV stenosis following RMA has an effect on long-term clinical outcome. The purpose of the present study was to elucidate the preoperative risk factors and hemodynamic determinants of an increase in the postoperative MV pressure gradient created by RMA, as well as its clinical effect on long-term outcome in patients with functional MR and advanced cardiomyopathy.

Methods

Patients

Between 2003 and 2009, pre- and postoperative (1 month) cardiac catheterization procedures were performed for 55 patients (46 males, 9 females; mean age 64±9 years) with functional MR secondary to advanced cardiomyopathy who underwent an RMA procedure using a 24/26 mm semi-rigid complete ring. All patients met the following inclusion criteria: (1) history of at least 1 hospitalization for heart failure in the previous 6 months despite maximal medical treatment, (2) global LV dysfunction (LV ejection fraction <40%) with a significantly enlarged LV, and (3) severe or moderate to severe MR caused by restrictive leaflet motion (type IIIb according to Carpentier’s classification). We excluded patients with recent myocardial infarction (<3 months), organic MR, or rheumatic mitral disease, as well as those who underwent RMA with a partial or semi-rigid complete ring >26 mm, concomitant surgical ventricular reconstruction, or aortic valve replacement. Prior to surgical referral, all had been treated with an optimized medical regimen by an attending cardiologist, including angiotensin-converting enzyme inhibitors or angiotensin-receptor blockers, β-blockers, and diuretics.

Baseline patient characteristics and relevant surgical data are summarized in Table 1. The final study protocol was approved by the institutional ethical committee and all patients provided informed consent.

Table 1. Clinical Characteristics and Surgical Data for Patients With Functional Regurgitation Undergoing Restrictive Mitral Annuloplasty

Variables	All patients (n=55)	MV gradient <5 mmHg (n=39)	MV gradient ≥5 mmHg (n=16)	P value
Clinical characteristics
Age (years)	64±9	65±9	62±8	0.30
Male	46 (84%)	31 (79%)	15 (94%)	0.26
BSA (m2)	1.64±0.18	1.61±0.14	1.71±0.23	0.05
Ischemic etiology	30 (55%)	22 (56%)	8 (50%)	0.66
NYHA class
II	7 (13%)	4 (10%)	3 (19%)	0.42
III	41 (75%)	31 (79%)	10 (63%)
IV	7 (13%)	4 (10%)	3 (19%)
Comorbidities
Hypertension	25 (45%)	16 (41%)	9 (56%)	0.30
Diabetes	28 (51%)	19 (49%)	9 (56%)	0.61
Hyperlipidemia	18 (33%)	13 (33%)	5 (31%)	0.88
Hemodialysis	7 (13%)	3 (7.7%)	4 (25%)	0.18
Cerebral vascular accident	16 (29%)	11 (28%)	5 (31%)	0.82
Atrial fibrillation	27 (49%)	18 (46%)	9 (56%)	0.50
Medications
β-blockers	35 (64%)	23 (59%)	12 (75%)	0.26
ACEI	11 (20%)	7 (18%)	4 (25%)	0.71
ARB	17 (31%)	12 (31%)	5 (31%)	0.97
Diuretics	42 (76%)	29 (74%)	13 (81%)	0.73
Echocardiographic data
LVEDD (mm)	68±8	68±8	67±6	0.60
LVESD (mm)	59±10	59±10	57±7	0.54
LVEF (%)	29±8	28±9	29±7	0.70
Systolic PAP (mmHg)	46±14	43±12	53±15	0.006
MR grade, 0/1+/2+/3+/4+	0/0/1/22/32	0/0/1/17/21	0/0/0/5/11	0.53
TR grade, 0/1+/2+/3+/4+	2/8/28/13/4	1/8/20/7/3	1/0/8/6/1	0.23
Surgical data
MV ring
24-mm Physio ring	30 (55%)	17 (44%)	13 (81%)	0.016
26-mm Physio ring	25 (45%)	22 (56%)	3 (19%)
Concomitant procedures
Coronary artery bypass grafting	25 (45%)	19 (49%)	6 (38%)	0.45
Tricuspid annuloplasty	41 (75%)	27 (69%)	14 (88%)	0.19

ACEI, angiotensin-converting enzyme inhibitor; ARB, angiotensin-receptor blocker; BSA, body surface area; LVEDD, left ventricular end-diastolic dimension; LVEF, left ventricular ejection fraction; LVESD, left ventricular end-systolic dimension; MR, mitral regurgitation; MV, mitral valve; NYHA, New York Heart Association; PAP, pulmonary artery pressure; TR, tricuspid regurgitation.

Cardiac Catheterization

Right- and left-heart catheterization procedures were performed using standard techniques before and 1 month (within 1 day of echocardiography) after surgery. None of the patients had complications related to pre- or postoperative catheterization. Prior to performance of left ventriculography, standard pressure measurements were obtained to evaluate LV systolic pressure, LV end-diastolic pressure (LVEDP), pulmonary capillary wedge pressure (PCWP), mean PAP, and right atrial pressure (RAP). Right-sided pressure findings were obtained using a Swan-Ganz catheter and LV volumes were determined on the basis of left ventriculography findings. Cardiac output (CO) was determined using the thermodilution method. The gradient across the MV was calculated as the pressure difference between mean PCWP and LVEDP, though it is generally considered better to determine the trans-MV gradient by use of simultaneous recordings.

Echocardiography

2D and Doppler transthoracic echocardiography examinations were performed before and 1 month after surgery. All echocardiographic studies were done using commercially available 3.75-MHz transducers (Toshiba, Tokyo, Japan; Hewlett-Packard Sonos) by expert echocardiographic examiners blinded to the clinical status of the patient and their operative data. Each transthoracic echocardiographic evaluation consisted of a standard echocardiography examination, including grading of the severity of regurgitation [semi-quantitative color-flow Doppler: none (0), trivial (1+), mild (2+), moderate (3+), severe (4+)], LV end-diastolic dimension, LV end-systolic dimension, LA dimension, and LV ejection fraction. The mean MV diastolic gradient was calculated from continuous-wave Doppler echocardiography findings using the Bernoulli equation. The MV effective orifice area (EOA) was determined using the pressure half-time method and indexed for body surface area (BSA).

Systolic PAP was calculated by adding the value for the systolic pressure gradient across the tricuspid valve, derived from tricuspid regurgitation, to the estimated RAP value.^15,16 RAP was estimated from the diameter and degree of collapse of the inferior vena cava during inspiration.

Surgical Procedures

All procedures were performed through a standard median sternotomy with standard cardiopulmonary bypass under mild systemic hypothermia. Myocardial protection was achieved with intermittent antegrade plus retrograde hyperkalemic cold blood cardioplegia. All patients underwent a RMA procedure using a semi-rigid complete ring (Carpentier-Edwards Physio Ring; Edwards Lifesciences, Irvine, CA, USA). The ring size was determined after careful measurement of the height of the anterior leaflet and inter-trigonal distance, and then downsizing by 2–3 sizes. Consequently, 30 (55%) patients received a 24-mm Physio ring and 25 (45%) had a 26-mm Physio ring.

Clinical Follow-up Examinations

Every 6–12 months, each patient was assessed by us as well as by their primary cardiologist. A retrospective review of the medical records of these patients was performed to obtain preoperative and postoperative data. Current information was obtained by contacting the patient or the referring cardiologist. We also reviewed postoperative adverse cardiac events, which were defined as death and readmission for heart failure. Clinical follow-up examinations were completed for all 55 patients (100%), with a mean duration of 75±32 months.

Statistical Analysis

Continuous variables are summarized as the mean±standard deviation, and categorical variables as frequencies and proportions. For continuous variables, comparisons between groups were made using an unpaired t-test. Categorical variables were compared using chi-square analysis or Fisher’s exact test, as appropriate. Pre- and postoperative hemodynamic variable measurements obtained over time were compared using repeated measures analysis of variance. Correlations between variables were tested with Spearman’s correlation coefficient (rho).

Stepwise multiple linear regression analysis was performed to identify determinants of postoperative catheter-measured MV gradient. Factors with a P-value <0.1 in univariate analysis based on Spearman’s correlation coefficient were entered into the multivariate model. The results are summarized as correlation coefficient (rho) and standardized partial regression coefficient (SPRC) values. The optimal cutoff value for the preoperative covariate to predict postoperative MV stenosis (catheter-measured MV gradient ≥5 mmHg) was determined by receiver-operating characteristic (ROC) curve analysis. In addition, associations of preoperative and surgical variables with postoperative MV stenosis were examined using stepwise logistic regression analysis. Factors with a P-value <0.1 were entered into the multivariate model. The results are summarized as odds ratio and 95% confidence interval values.

The Kaplan-Meier method was used to assess time-related outcomes (freedom from adverse cardiac events, survival). Statistical significance was defined as a P<0.05. Statistical significance was defined as a P value less than 0.05. Statistical analyses were performed using the JMP 7.0 (SAS Institute, Cary, NC, USA).

Results

Patients’ Background Data (Table 1)

The postoperative MV pressure gradient value, calculated from the difference in pressure between mean PCWP and LVEDP, was 3.4±1.6 mmHg (range 1–7 mmHg) and demonstrated a positive correlation with Doppler-measured mean MV gradient (rho=0.65, P<0.001). Prevalence for an MV gradient ≥5 mmHg was identified in 16 (29%) patients (95% confidence interval 19–42%).

Patients were classified into 2 groups based on the postoperative catheter-measured MV gradient, which comprised 39 patients with low (<5 mmHg) and 16 with high (≥5 mmHg) MV gradient values. There were no differences in age, sex, heart failure severity, heart failure etiology, and severity of comorbidities between the groups, although patients with an MV gradient ≥5 mmHg were more likely to have a larger BSA. There were also no differences in the LV function parameters and severity of valvular disease, though the group with an MV gradient ≥5 mmHg showed increased systolic PAP. Furthermore, those patients were likely to undergo RMA with the smallest sized mitral ring (24 mm), despite the prevalence of similar concomitant surgical procedures.

Pre- and Postoperative Hemodynamic Data (Table 2)

From baseline to 1 month after surgery, LV volumes significantly decreased and the ejection fraction improved in patients with and without an MV gradient ≥5 mmHg, with no significant intergroup differences (time effect P<0.001, group and interaction effects P>0.05 for all). In addition, LVEDP, PCWP, and mean PAP values decreased, while CO significantly improved in both groups (time effect P<0.001 for all), accompanied by higher postoperative values for mean PAP and CO in patients with an MV gradient ≥5 mmHg (group effects P<0.05 for both). The prevalence of residual pulmonary hypertension (defined as mean PAP ≥25 mmHg) following RMA tended to be greater in patients with an MV gradient ≥5 mmHg (44% vs. 26%, P=0.19), though the difference was not statistically significant.

Table 2. Hemodynamic Changes After Restrictive Mitral Annuloplasty

Variables	MV gradient <5 mmHg		MV gradient ≥5 mmHg		Group effect P value	Time effect P value	Interaction effect P value
	Pre-op (n=39)	Post-op (n=39)	Pre-op (n=16)	Post-op (n=16)
LVESVI (mL/m2)	111±40	89±39	105±31	83±33	0.57	<0.001	0.64
LVEDVI (mL/m2)	147±42	121±39	140±38	119±37	0.65	<0.001	0.50
LVEF (%)	25±8	30±11	25±7	32±10	0.73	<0.001	0.68
LVEDP (mmHg)	19±6	11±5	19±7	11±4	0.56	<0.001	0.90
PCWP (mmHg)	21±6	14±4	22±7	16±4	0.14	<0.001	0.79
Mean PAP (mmHg)	31±6	20±6	36±9	25±6	0.003	<0.001	0.83
RAP (mmHg)	7±3	8±2	8±5	9±3	0.32	0.20	0.90
Cardiac output (L/min)	4.0±0.9	4.3±0.6	4.5±1.6	5.7±1.7	<0.001	<0.001	0.01
Cardiac index (L/min/m2)	2.5±0.6	2.6±0.4	2.7±0.8	3.4±0.8	0.002	<0.001	0.02
Heart rate (beats/min)	76±12	77±10	78±13	85±14	0.12	0.08	0.24

LVEDVI, left ventricular end-diastolic volume index; LVEDP, left ventricular end-diastolic pressure; LVESVI, left ventricular end-systolic volume index; PCWP, pulmonary capillary wedge pressure; RAP, right atrial pressure. Other abbreviations as in Table 1.

Postoperative Echocardiographic Findings (Table 3)

Following surgery, the majority of patients showed a substantial reduction in MR severity. The mean values for MV EOA and indexed MV EOA were 2.5±0.3 cm² (range 1.8–3.3 cm²) and 1.5±0.3 cm²/m² (0.9–2.2 cm²/m²), respectively, while that for Doppler-derived mean MV gradient was 3.5±1.5 mmHg (1.1–8.6 mmHg). When we compared those values between patients with and without an MV gradient ≥5 mmHg, there were no differences in MR severity. In patients with an MV gradient ≥5 mmHg, the measurements for MV EOA (2.3±0.4 vs. 2.5±0.3 cm², P=0.014) and indexed EOA (1.4±0.3 vs. 1.6±0.2 cm²/m², P=0.016) were significantly smaller. Also, patients with an MV gradient ≥5 mmHg were likely to show a higher level of systolic PAP (37±11 vs. 31±8 mmHg, P=0.047).

Table 3. Postoperative Echocardiographic Measurements (1 Month After Surgery)

Variables	All patients (n=55)	MV gradient <5 mmHg (n=39)	MV gradient ≥5 mmHg (n=16)	P value
MR grade, 0/1+/2+/3+/4+	36/13/5/1/0	27/8/4/0/0	9/5/1/1/0	0.323
TR grade, 0/1+/2+/3+/4+	11/33/10/1/0	10/22/6/1/0	1/11/4/0/0	0.333
MV EOA (cm2)	2.5±0.3	2.5±0.3	2.3±0.4	0.014
Indexed MV EOA (cm2/m2)	1.5±0.3	1.6±0.2	1.4±0.3	0.016
Mean MV gradient (mmHg)	3.5±1.5	2.9±1.1	5.0±1.5	<0.001
Systolic PAP (mmHg)	33±9	31±8	37±11	0.047

EOA, effective orifice area. Other abbreviations as in Table 1.

Notably, there were positive correlations between implanted annuloplasty ring size and postoperative (rho=0.63, P<0.001) and indexed (rho=0.36, P=0.009) MV EOA. Patients who underwent RMA with a 24-mm annuloplasty ring had a smaller postoperative MV EOA measurement as compared with those who received a 26-mm annuloplasty ring (2.3±0.3 vs. 2.7±0.3 cm², P<0.001).

Hemodynamic Determinants of Postoperative Catheter-Measured MV Gradient

The postoperative catheter-measured MV pressure gradient was positively correlated with postoperative mean PAP (rho=0.30, P=0.027) and CO (rho=0.59, P<0.001), and negatively with echo-derived MV EOA indexed by BSA (rho=−0.27, P=0.044) (Table 4). Stepwise linear regression analysis showed that CO had the greatest contribution to catheter-measured MV gradient (SPRC=0.59, P<0.001), followed by indexed EOA (SPRC=−0.25, P=0.019).

Table 4. Postoperative Data Correlating With Postoperative Catheter-Measured MV Gradient

Variables	Univariate		Multivariate
	rho	P value	SPRC	P value
Post-op echo data (n=55)
LVEDD (mm)	0.01	0.946
LVESD (mm)	−0.05	0.739
LVEF (%)	0.12	0.371
LA dimension (mm)	0.26	0.061
Systolic PAP (mmHg)	0.26	0.061
MV EOA (cm2)*	−0.26	0.054
Indexed MV EOA (cm2/m2)	−0.27	0.044	−0.25	0.019
Post-op volume data (n=55)
LVESVI (mL/m2)	−0.04	0.804
LVEDVI (mL/m2)	−0.01	0.921
LVEF (%)	0.10	0.457
Post-op hemodynamic data (n=55)
LVEDP (mmHg)	–0.15	0.271
PCWP (mmHg)	0.20	0.134
Mean PAP (mmHg)	0.30	0.027
RAP (mmHg)	0.16	0.258
Cardiac output (L/min)	0.59	<0.001	0.59	<0.001
Cardiac index (L/min/m2)*	0.58	<0.001
Heart rate (beats/min)	0.16	0.252

*MV EOA and cardiac index were not entered into multivariate analysis because of strong correlations between MV EOA and indexed MV EOA (rho=0.75), and cardiac index and cardiac output (rho=0.83). LA, left atrial; SPRC, standardized partial regression coefficient. Other abbreviations as in Tables 1–3.

Risk Factors and Predictors of Postoperative Functional MV Stenosis

Several preoperative and surgical factors showed correlations with the postoperative catheter-measured MV gradient, including male sex (rho=0.41, P=0.003), BSA (rho=0.32, P=0.019), hemodialysis (rho=0.56, P<0.001), and use of a 24-mm mitral ring (rho=0.44, P=0.001) (Table 5). Stepwise linear regression analysis showed that BSA had the greatest contribution (SPRC=0.38, P=0.002), followed by use of a 24-mm ring (SPRC=0.33, P=0.007) and hemodialysis (SPRC=0.26, P=0.029).

Table 5. Preoperative and Surgical Data Correlating With Postoperative Catheter-Measured MV Gradient

Variables	Univariate
	rho	P value
Demographics (n=55)
Age (years)	−0.16	0.251
Male	0.41	0.003
Ischemic etiology	0.13	0.332
BSA (m2)	0.32	0.019
Hypertension	0.29	0.032
Diabetes	0.31	0.024
Hyperlipidemia	0.15	0.287
Hemodialysis	0.56	<0.001
Atrial fibrillation	0.27	0.050
Medications (n=55)
β-blockers	0.21	0.121
ACEI	0.21	0.121
ARB	0.27	0.050
Diuretics	0.17	0.215
Pre-op echo data (n=55)
LVEDD (mm)	−0.09	0.508
LVESD (mm)	−0.05	0.740
LVEF (%)	0.20	0.149
LA dimension (mm)	0.30	0.028
Systolic PAP (mmHg)	0.22	0.113
Pre-op volume data (n=55)
LVESVI (mL/m2)	−0.11	0.427
LVEDVI (mL/m2)	−0.05	0.730
LVEF (%)	0.21	0.132
Pre-op hemodynamic data (n=55)
LVEDP (mmHg)	0.01	0.932
PCWP (mmHg)	0.10	0.447
Mean PAP (mmHg)	0.18	0.182
RAP (mmHg)	0.19	0.166
Cardiac output (L/min)	0.12	0.391
Cardiac index (L/min/m2)	0.09	0.493
Heart rate (beats/min)	0.08	0.543
Surgical data (n=55)
Use of 24-mm mitral ring	0.44	0.001
	Multivariate
	SPRC	P value
Demographics (n=55)
BSA (m2)	0.38	0.002
Hemodialysis	0.26	0.029
Surgical data (n=55)
Use of 24-mm mitral ring	0.33	0.007

Abbreviations as in Tables 1,2,4.

ROC curve analysis demonstrated an optimal cutoff value for BSA of 1.86 m² to predict a postoperative a catheter-measured MV gradient ≥5 mmHg (21% for <1.86 m² vs. 86% for ≥1.86 m², P=0.002). The prevalence of a postoperative catheter-measured MV gradient ≥5 mmHg in patients with BSA <1.86 m² who received a 24-mm annuloplasty ring was 33% (8 of 24), while it was 8.3% (2 of 24) in those who received a 26-mm annuloplasty ring (P=0.072). In patients with BSA ≥1.86 m², these values were 83% (5 of 6) and 100% (1 of 1), respectively (P=0.99).

Univariate analysis showed that large BSA value (≥1.86 m²), echo-derived systolic PA pressure, catheter-measured mean PA pressure, and use of a 24-mm ring were associated with postoperative MV stenosis (catheter-measured MV gradient ≥5 mmHg). Stepwise logistic regression analysis revealed BSA ≥1.86 m² as an independent risk factor (adjusted odds ratio=17, 95% confidence interval=1.8–173, P=0.01).

Effect of Postoperative Increase in MV Gradient on Late Outcome After RMA

During the follow-up period, there were 28 deaths and 31 readmissions for heart failure, with actuarial survival rates at 1, 3, and 5 years of 98±1%, 89±3%, and 71±5%, respectively. When comparing late outcomes according to postoperative MV gradient, there were 9 mortalities and 11 late readmissions due to recurrent heart failure in patients with an MV gradient ≥5 mmHg, while there were 19 mortalities and 20 readmissions due to recurrent heart failure in those with a gradient <5 mmHg. There was no statistical difference in regard to actuarial survival rate at 3 years after surgery between patients with and without an MV gradient ≥5 mmHg (94±6% vs. 85±6%, P=0.94), or for freedom from composite adverse events (56±12% vs. 62±8%, P=0.24).

Discussion

The major findings of our hemodynamic study can be summarized as follows. In patients with functional MR secondary to advanced cardiomyopathy who underwent an RMA procedure using a 24/26-mm semi-rigid complete ring; 1) there was a high likelihood of postoperative “functional” MV stenosis in terms of catheter-measured MV gradient ≥5 mmHg, especially for those with a 24-mm MV ring, 2) the catheter-measured MV gradient independently correlated with postoperative catheter-measured CO and echo-derived MV EOA indexed by BSA, (3) patients with BSA ≥1.86 m² were more likely to have functional MV stenosis as compared with those with BSA <1.86 m², and (4) postoperative functional MV stenosis did not adversely affect long-term freedom from death and/or readmission for heart failure.

Bertrand and colleagues¹¹ used Doppler echocardiography to study hemodynamic response at rest and during exercise in patients with ischemic MR who underwent a standardized RMA. They found that patients with an MV gradient ≥5 mmHg had greater CO at rest than those with <5 mmHg. They also noted that the trend of higher CO at rest in patients with an MV gradient ≥5 mmHg was more remarkable during exercise, along with a significant increase in the gradient. Likewise, Magne and co-workers used echocardiography to demonstrate an increase in Doppler-derived MV gradient in parallel with increased CO during exercise, and also reported that patients who underwent an RMA had a higher MV gradient value after surgery as compared with a control group (isolated-CABG population).⁹ The present findings provide further evidence that patient factors related to higher CO are the greatest contributors to catheter-measured MV gradient, followed by indexed MV EOA. Our results also suggested that an increase in the MV gradient following RMA may reflect increased CO, which can be attributed to correction of the underlying pathology (i.e., reduction in functional MR). Therefore, the MV gradient value should be interpreted in regard to post-RMA patient hemodynamics and is not necessarily detrimental for functional capacity.

It is reasonable that we detected a significant correlation of postoperative MV gradient with such preoperative factors as male sex, BSA, and hemodialysis, as those variables are likely correlated with high CO (data not shown). We found that downsizing the mitral annulus with use of a 24-mm mitral ring was also associated with the postoperative MV gradient after RMA. That finding was also not surprising because functional MV stenosis might be caused, at least in part, by utilization of a small annuloplasty ring. However, Kubota et al observed the presence of functional MV stenosis even in patients whose annular size was not aggressively reduced during RMA (i.e., those who did not undergo downsizing of the annulus).¹⁰ Likewise, Rubino et al reported no significant difference in size of the annuloplasty ring implanted during the RMA procedure between patients with and without a postoperative mean mitral gradient ≥5 mmHg (25.8±1.1 vs. 25.4±1.4 mm; P=0.06). Furthermore, despite a higher frequency of 24-mm annuloplasty rings implanted in our series, the MV gradient value observed in this study was quite consistent with that noted in previous studies.^6–12 This finding might be explained by the substantially lower BSA in our cohort as compared with other reports. It is worth noting that in our study, patients with BSA ≥1.86 m² had a high possibility for development of postoperative functional MV stenosis after RMA that utilized a 24- or 26-mm annuloplasty ring. Together, these results suggested that the degree of downsizing should be determined not only by the annular geometry of the native valve but also according to the patient’s BSA.

The effect of increased MV gradient after RMA on clinical outcome remains controversial. Magne et al provided evidence of a potential risk of iatrogenic mitral stenosis following RMA, which was associated with hemodynamic impairment (higher PAP) and worse functional capacity.⁹ Kubota et al noted that surgical annuloplasty frequently causes functional MV stenosis in association with persistent subvalvular tethering, which may have a deleterious effect on heart failure symptoms.¹⁰ On the other hand, another study conducted by Duke University that used echocardiographic assessments of 222 consecutive patients with ischemic MR who underwent MV repair found that an increased MV gradient did not adversely affect long-term survival or hospitalization for heart failure.¹³ In addition, Rubino et al revealed that mild MV stenosis (i.e., Doppler-derived MV gradient >5 mmHg) had no effects on survival, hospitalization for heart failure, NYHA functional class, or echocardiographic outcomes in 125 consecutive patients with chronic ischemic MR who underwent undersized ring MV annuloplasty procedures.¹⁴ Interestingly, in their study, the amount of improvement in the LV end-diastolic dimension was greater in patients with Doppler-derived MV gradient >5 mmHg, indicating that mild MV stenosis can play a protective role against LV filling pressure. In the present study, we also failed to find a statistical difference in regard to long-term clinical outcome (freedom from adverse events) between patients with and without an MV gradient ≥5 mmHg, which is consistent with the latter reports. Conflicting results are likely caused by different patient backgrounds, surgical procedures (i.e., implanted ring size), and study endpoints, such as a relatively short-term follow-up period or lack of findings from hemodynamic monitoring techniques in regard to postoperative functional MV stenosis. Additional studies that include greater ranges of recorded gradients and longer follow-up periods will be helpful to clarify the role of an increased MV gradient in terms of hemodynamic function and long-term clinical outcomes following the RMA procedure.

Clinical Implications

The strong correlation between MV gradient and CO seen in the present study indicates that such gradient measurements should be interpreted with respect to patient hemodynamics, and are not necessarily considered to be detrimental to functional capacity.¹² Nevertheless, our hemodynamic data clearly showed that patients with an MV gradient ≥5 mmHg were more likely to show elevated PAP following surgery as compared with those with a gradient <5 mmHg. Given that residual pulmonary hypertension (PH) may be a predictor of postoperative adverse events after an RMA procedure and a surrogate of exercise capacity,⁵ it is important to not place patients at risk for an increased MV gradient, which can be attributed, at least in part, to postoperative residual PH. Our findings also suggest that use of a 24-mm ring might not be recommended for patients with large BSA or those dependent on hemodialysis, who usually show an elevated CO caused by an arteriovenous shunt.

Study Limitations

This study has several major limitations. Since it was retrospective in nature and we only investigated a small number of subjects, the results should be cautiously interpreted until verified by an independent prospective study. Also, inclusion of patients with different cardiomyopathy etiologies as well as those who underwent concomitant surgical procedures might also have influenced the results. However, concomitant procedures are usually required for this population of very sick patients, who present a similar clinical and pathophysiologic status despite an etiology of LV dysfunction.

To reduce prosthetic-related bias, we restricted the present study to patients who received a 24/26-mm semi-rigid complete ring, which has been used for many years, with well-documented durability and reliability. Thus, our results may not be applicable for patients who received another type of prosthetic ring.

Lack of echocardiographic assessments of systolic PA pressure or MV pressure gradient during dobutamine administration or exercise stress did not allow us to determine whether an RMA procedure induces functional MV stenosis, especially with increased preload. In addition, diastolic subvalvular tethering parameters, such as anterior and posterior leaflet opening angle, annular dimension, and mitral leaflet tip opening dimension, previously shown to account for the mechanism of functional MV stenosis after RMA,¹⁰ were not evaluated for all of the patients in this series. Those sophisticated echocardiographic assessments would be informative to comprehensively understand the cause of iatrogenic MV stenosis following an RMA procedure.

Conclusions

The post-RMA MV gradient was determined not only by the degree of annular reduction but also by patients’ hemodynamic factors (e.g., CO). Use of a 24/26-mm annuloplasty ring for patients with BSA ≥1.86 m² was related to a high likelihood of post-MV stenosis. However, mild MV stenosis did not adversely affect late outcomes after RMA.

Acknowledgments

The authors thank Mr. Kiyoshi Yoshida (Section of Clinical Engineering, the Japan Organization of Occupational Health and Safety Osaka Rosai Hospital) for excellent management of the cardiopulmonary bypass procedures and Ms. Mariko Yamashita for helpful assistance with clinical data collection.

Conflicts of Interest

None to declare.

Funding Source

This study was partially supported by research funds to promote hospital function by the Japan Organization of Occupational Health and Safety Osaka Rosai Hospital.

References

• 1.   Grigioni F, Enriquez-Sarano M, Zehr KJ, Bauley KR, Tajik J. Ischemic mitral regurgitation: Long-term outcome and prognostic implications with quantitative Doppler assessment. Circulation 2001; 103: 1759–1764.
• 2.   Koelling TM, Aaronson KD, Cody RJ, Bach DS, Armstrong WF. Prognostic significance of mitral regurgitation and tricuspid regurgitation in patients with left ventricular systolic dysfunction. Am Heart J 2002; 144: 524–529.
• 3.   Chan KM, Punjabi PP, Flather M, Wage R, Symmonds K, Roussin I, et al; RIME Investigators. Coronary artery bypass surgery with or without mitral valve annuloplasty in moderate functional ischemic mitral regurgitation: Final results of the Randomized Ischemic Mitral Evaluation (RIME) trial. Circulation 2012; 126: 2502–2510.
• 4.   Kainuma S, Taniguchi K, Toda K, Funatsu T, Kondoh H, Nishino M, et al. Restrictive mitral annuloplasty for functional mitral regurgitation: Acute hemodynamics and serial echocardiography. Circ J 2011; 75: 571–579.
• 5.   Kainuma S, Taniguchi K, Daimon T, Sakaguchi T, Funatsu T, Kondoh H, et al; Osaka Cardiovascular Surgery Research (OSCAR) Group. Does stringent restrictive annuloplasty for functional mitral regurgitation cause functional mitral stenosis and pulmonary hypertension? Circulation 2011; 124: 97–106.
• 6.   Bax JJ, Braun J, Somer ST, Klautz R, Holman ER, Versteegh MI, et al. Restrictive annuloplasty and coronary revascularization in ischemic mitral regurgitation results in reverse left ventricular remodeling. Circulation 2004; 110(Suppl II): II-103–II-108.
• 7.   Braun J, van de Veire NR, Klautz RJ, Versteegh MI, Holman ER, Westenberg JJ, et al. Restrictive mitral annuloplasty cures ischemic mitral regurgitation and heart failure. Ann Thorac Surg 2008; 85: 430–437.
• 8.   De Bonis M, Taramasso M, Grimaldi A, Maisano F, Calabrese MC, Verzini A, et al. The GeoForm annuloplasty ring for the surgical treatment of functional mitral regurgitation in advanced dilated cardiomyopathy. Eur J Cardiothorac Surg 2011; 40: 488–495.
• 9.   Magne J, Sénéchal M, Mathieu P, Dumesnil JG, Dagenais F, Pibarot P. Restrictive annuloplasty for ischemic mitral regurgitation may induce functional mitral stenosis. J Am Coll Cardiol 2008; 51: 1692–1701.
• 10.   Kubota K, Otsuji Y, Ueno T, Koriyama C, Levine RA, Sakata R, et al. Functional mitral stenosis after surgical annuloplasty for ischemic mitral regurgitation: Importance of subvalvular tethering in the mechanism and dynamic deterioration during exertion. J Thorac Cardiovasc Surg 2010; 140: 617–623.
• 11.   Bertrand PB, Gutermann H, Smeets CJ, Van Kerrebroeck C, Verhaert D, Vandervoort P, et al. Functional impact of transmitral gradients at rest and during exercise after restrictive annuloplasty for ischemic mitral regurgitation. J Thorac Cardiovasc Surg 2014; 148: 183–187.
• 12.   Fino C, Iacovoni A, Ferrero P, Senni M, Merlo M, Cugola D, et al. Restrictive mitral valve annuloplasty versus mitral valve replacement for functional ischemic mitral regurgitation: An exercise echocardiographic study. J Thorac Cardiovasc Surg 2014; 148: 447–453.
• 13.   Williams ML, Daneshmand MA, Jollis JG, Horton JR, Shaw LK, Swaminathan M, et al. Mitral gradients and frequency of recurrence of mitral regurgitation after ring annuloplasty for ischemic mitral regurgitation. Ann Thorac Surg 2009; 88: 1197–1201.
• 14.   Rubino AS, Onorati F, Santarpia G, Achille F, Lorusso R, Santini F, et al. Impact of increased transmitral gradients after undersized annuloplasty for chronic ischemic mitral regurgitation. Int J Cardiol 2012; 28: 71–77.
• 15.   Pepi M, Tamborini G, Galli C, Barbier P, Doria E, Berti M, et al. A new formula for echo-Doppler estimation of right ventricular systolic pressure. J Am Soc Echocardiogr 1994; 7: 20–26.
• 16.   Kircher BJ, Himelman RB, Schiller NB. Noninvasive estimation of right atrial pressure from the inspiratory collapse of the inferior vena cava. Am J Cardiol 1990; 66: 493–496.