The Impact of Surgical Ventricular Restoration on Ischemic Mitral Regurgitation

Summary. Objective. The aim of current study was to evaluate and compare the changes in clinical status and mitral regurgitation (MR) grade and long-term postoperative survival after left ventricle surgical restoration (SVR) operations. Material and Methods. We retrospectively analyzed the data of 139 patients suffering from ischemic heart disease and left ventricular aneurysms or large akinesia, who underwent SVR and coronary artery bypass grafting without MR surgical correction between 1999 and 2006. The mean long-term postoperative follow-up was 3.6 years (SD, 3.0). Nine patients (6.5%) died during the first 30 postoperative days. Results. The mean MR grade during the long-term period increased significantly. The univariate logistic regression analysis showed that factors for the long-term mortality were age (P=0.002), decompensation signs before SVR (P=0.03), treatment with diuretics (P=0.01), NYHA functional class IV (P=0.008), and moderate and severe MR (P=0.04); however, multivariate logistic regression analysis demonstrated that only patient’s age was an independent predictive factor (P=0.004). MR correction was found to be a significant prognostic factor of borderline significance for perioperative mortality (P=0.05). The analysis of MR grade (mild versus moderate versus severe) impact on long-term survival failed to demonstrate any association (P=0.22). Conclusions. Remodeling continued during the long-term period after SVR and CABG: there was an increase in the mean MR degree. Left ventricular remodeling with moderate and severe MR decreased survival rates during long-term period after surgical ventricular restoration; however, mitral regurgitation was not found to be an independent predictor of poor outcome.


Introduction
The lack of information has caused a number of discussions about the treatment strategies of ischemic mitral valve insuffi ciency, typical for remodeled left ventricle (LV) (1,2). There is no consensus on the treatment of patients with ischemic heart disease (IHD) with LV dysfunction and mitral regurgitation (MR). The effectiveness of surgical ventricular restoration (SVR) in patients with depressed LV systolic function and mitral regurgitation is still being discussed (3,4). According to Menicanti et al. (5), SVR reduces the size of the MV annulus, the distance between papillary muscles, and diastolic and systolic LV sphericity: when SVR was done without MV repair, in 87% of the cases, moderate and severe MR reduced to trivial or mild, and in 13% of the cases, it remained the same or increased later on. In a study by Shapira et al. (6), grade 1 and 2 MR was left without surgical correction with expectations that after the revascularization of myocardium, it will remain stable or decrease; however, the results showed that untreated grade 1 and 2 MR had a signifi cant nega-tive impact on long-term survival. Yotsumoto et al. (7) also recommended surgical repair for mild MR. However, there are contrary opinions. A study by Paparella et al. (8) showed that grade 2 and 3 MR was not a predictor of poor outcome. Paparella et al. and Mickleborough et al. (8,9) reported that grade 2 and 3 MR did not increase mortality risk during long-term follow-up after myocardial revascularization operations. These results were supported by Ryden et al. (10). According to Kaza et al. (11), an improvement in the degree of mitral insuffi ciency was documented in 60% of cases when SVR and coronary artery bypass graft (CABG) operations were performed without MR surgical correction. According to the authors, SVR results in a narrower confi guration of papillary muscles and reduces MR. A study by Sartipy et al. (12) showed that only grade 3 and 4 MR had a negative impact on longterm survival rates. The aim of our study was to evaluate how the grade of ischemic MR changed after SVR, to analyze long-term survival rates after SVR, and to investigate the relation between survival rates and MR grade.

Materials and Methods
We analyzed the data of 139 consecutive patients suffering from IHD and LV aneurysms or large akinesia; these patients underwent SVR with or without CAGB and with or without MV surgical correction during the period from February 1999 to May 2006. The data were collected retrospectively, using medical case histories, outpatient records, and operation protocols. The patients' clinical condition and functional status, drug treatment, cardiovascular complications were evaluated on interview and physical examination. Echocardiographic examinations were performed preoperatively, during early period, after 6 months, and once a year during long-term follow-up after operation. This series included 119 male and 20 female patients (85.6% and 14.4%). The mean age was 60.5 years (SD, 10.6; range, 40-80 years). NYHA FC III-IV was diagnosed in 129 patients (92.8%) preoperatively; the mean FC was 3.4 (SD, 0.6). Analysis of morphometry and function showed that the mean LV ejection fraction (EF) before SVR was 32.0% (SD, 8.8%, range, 12%-56%); mean LV end systolic volume (ESV), 123.5 mL (SD, 52.6); mean end-systolic volume index (ESVI), 64.5 mL/m 2 (SD, 27.6); and mean LV end-diastolic dimension (EDD), 6.4 cm (SD, 0.8). The signs of heart decompensation were present in 53 patients (39.6%) The mean angina pectoris (AP) score according to the Canadian classifi cation was 3.1 (SD, 0.6). All the patients had a history of myocardial infarction (MI) (mean, 1.4; SD, 0.7). The mean duration from MI to operation was 67.1 months (SD, 69.6 months; range, 1 month to 26 years). Acute MI was diagnosed for 11 patients (7.9%). The data of coronary angiography showed that the mean number of coronary artery (CA) affected was 2.4 (SD, 0.9) per patient. Eighty patients (57.5%) had signifi cant stenosis (≥75% diameter) of three CA; stenosis of the main left CA was observed in 19 operated patients (13.7%). The aneurysms of anterior wall of LV and apex were present in 128 patients (92.1%); in 11 (7.9%) of the cases, there were aneurysms of inferior-posterobasal wall. Mild (grade 1) mitral regurgitation was diagnosed in 71 patients (51.1%), moderate (grade 2) in 43 (30.9%), and severe (grade 3) in 9 patients (6.5%). Only 16 patients (11.5%) had no mitral regurgitation. The mean MR grade was 1.3 (SD, 0.7). Primary arterial hypertension was diagnosed for 75 patients (54.0%), and 25 patients (18.0%) suffered from diabetes mellitus. Rhythm disturbances were documented in 33 study subjects (23.7%): 18 (12.9%) had paroxysmal ventricular tachycardia and 15 (10.8%) suffered from paroxysmal atrial fi brillation or fl utter.
The long-term postoperative follow-up was performed for 111 (79.1%) of the 139 patients. The follow-up duration of patients discharged for outpatient treatment ranged from 3 months to 7 years (mean follow-up, 3.6 years; SD, 3.0). Nine patients (6.5%) died during the fi rst 30 postoperative days. Twenty-six patients (23.4%) discharged from the hospital had died during long-term follow-up. The fact of death was checked-up for all study subjects (100%) and was confi rmed by the Resident's Register Service at the Ministry of Interior of the Republic of Lithuania.
The patients after SVR and CABG were treated with ACE inhibitors, β-adrenoblockers, aspirin or indirect anticoagulants, diuretics, and statins. Mitral regurgitation was evaluated using color fl ow Doppler and continuous wave Doppler methods defi ning width, size, and velocity of the regurgitant jet in the left atrium (13), and the proximal isovelocity surface area (PISA) method was used (14). Regurgitant jet area (RJA) was obtained and divided by left atria area (LAA) measured in the same frame: RJA/ LAA <20% (mild MR, grade 1), 20%-40% (moderate MR, grade 2), and >40% (severe MR, grade 3). Using the PISA method, MR was graded as mild and signifi cant <20 mm 2 and >20 mm 2 , respectively.
Data were analyzed using the SPSS 16.0 (version for Windows) software. Results were considered statistically signifi cant when P value was ≤0.05.

Results
After one year following SVR and concomitant CABG (95.0% of the cases), 108 patients (97.3%) had no symptoms of AP. The mean AP class was 0.1 (SD, 0.3) one year after the operation (P<0.001, in comparison with the preoperative one), and there was an improvement during long-term follow-up (P<0.001). The mean New York Heart Association (NYHA) functional class (FC) one year after surgery was 2.3 (SD, 0.4) (improved, P<0.001). The improvement in the mean NYHA FC was also recorded during the long-term follow-up (2.4 [SD, 0.5] after 5 years; P<0.001, in comparison with the preoperative one). An improvement in LV EF was recorded as well: LV EF was 37.5% (SD, 7.9%) one year after the operation (P<0.001). The decrease in LV EDD was observed after SVR. The mean LV EDD was 6.1 cm (SD, 0.5) one year after the operation (P<0.02); however, an increase in LV EDD was observed with time, and 3 years after the operation did not differ from the preoperative one (6.3 cm [SD, 0.5]; P=0.52). There was a decrease in the mean LV ESVI after SVR: one year after the operation, it was 53.7 mL/m 2 (SD, 17.4) (P=0.015). The mean ESVI increased during the long-term followup and was 57.7 mL/m 2 (SD, 23.4) after 3 years (P=0.15, in comparison with the preoperative one).
The changes in MR grade after SVR were assessed. The mean preoperative MR grade was 1.3 (SD, 0.7); one year after SVR, it was 1.0 (SD, 0.5) (P=0.003). The mean MR grade during the longterm follow-up increased to 1.1 (SD, 1.4) (after 2 years), 1.2 (SD, 0.4) (after 3 years), and 1.3 (SD, 0.4) (after 4 years) and did not differ from the preoperative one (P=0.48). The changes in MR grade in patients who underwent SVR without surgical correction of MV insuffi ciency were analyzed during the one-year period after the surgery. In order to examine the relationship between patient's functional class during one-year period after SVR and preoperative MR grade, the Fisher exact test was used. It was demonstrated that MR degree before the operation and patients postoperative NYHA FC were not related (P=0.53). The mean MR grade decreased during one-year period after SVR (P=0.003), and there was a decrease of the number of patients who had moderate or severe MR (P<0.01). However, the mean LV EDD, ESVI, and MR decreased during the one-year period after SVR, but thereafter, a gradual increase in these parameters was documented: 3-4 years after the operation, the mean LV EDD, LV ESVI, and MR became similar to the preoperative ones.
The 1-, 3-, 4-, 5-, and 6-year survival in this group was 90.6% (SD, 2.5%), 81.4% (SD, 3.5%), 76.0% (SD, 3.9%), 69.8% (SD, 4.5%), and 62.8% (SD, 6.5%), respectively. The univariate logistic regression model with one independent variable was used to predict the perioperative and long-term mortality by means of evaluating preoperative clinical, operative, and echocardiographic data (  of borderline signifi cance for perioperative mortality (P=0.05). Signifi cant factors for prediction of long-term mortality were as follows: age (P=0.002), decompensation signs before SVR (P=0.03), treatment with diuretics before operation (P=0.01), NYHA FC IV (P=0.008), and moderate and severe MR (P=0.04) ( Table 3). It is important to note that analysis of MR impact on long-term survival using a logistic regression model (when comparison of odds ratio in patients without MR or with mild MR versus patients with moderate and severe MR was performed) showed that the assessment of odds ratio was statistically signifi cant (P=0.04). However, the analysis of MR grade (mild versus moderate versus severe) impact on long-term survival using the chisquare or Fisher exact tests failed to demonstrate any association (P=0.22). To evaluate MR impact on long-term survival, the association between MR grade one year after SVR (grade 1 versus grade 2 and 3) and long-term survival using the Cox regression method was examined. No signifi cant association was found (Exp(B)=2.48, P=0.13). In order to identify an independent predictor for worse outcome, a multivariate logistic regression model was developed (Table 4). Multivariate analysis revealed that only patient's age was an independent predictor of poor 6-year outcome (P=0.004).

Discussion
Functional mitral regurgitation is a signifi cant complication of highly developed left ventricular remodeling and heart failure, whose pathogenesis is multifactorial. It develops due to changes in the MV annulus -LV apparatus shape and volume geometry -leading to incomplete leafl et cooptation. In case of ischemic heart failure, it may also develop due to papillary muscles or lateral LV wall myocardial dysfunction. MV regurgitation in turn increases the volume overload of dilated LV, which results in MV annular dilatation, increasing LV wall tension, increasing MR degree, and progression of HF (15). Higher MV degree is associated with larger LV volumes, worse LV function, and lower patient survival rates: MR of any, even a mild, degree increased the mortality risk, and the increasing MR degree significantly raised the risk of cardiovascular events (16). Grigioni et al. (17) reported that ischemic MR, irrespective of LV volumes and dysfunction degree, was found to be an independent predictor of mortality during 5 years. However, Prucz et al. (4) provided evidence that the preoperative degree of MR was not related to long-term survival after SVR. Trichon et al. (18) showed that patients with HF often also had MR: of the 2057 patients with HF symptoms and LV EF <40%, 56.2% were diagnosed with MR. MV insuffi ciency of any degree increased the mortality risk. MR can be eliminated by decreasing LV endsystolic volume and sphericity -SVR, reconstructing the MV annulus (by undersizing annuloplasty), by myocardial revascularization, and removing myocardial ischemia surrounding the MV apparatus. According to Yotsumuto et al. (7), the development   It seems that LV remodeling, associated with MR, has a greater impact on long-term survival than MR degree. During the late postoperative period, mean MR grade, LV EDD, and LV ESVI were increased, and this confi rms the development of further remodeling.

Conclusions
Remodeling continued during the long-term period after surgical ventricular restoration and coronary artery bypass grafting: there was an increase in the mean mitral regurgitation degree. Left ventricular remodeling with moderate and severe mitral regurgitation decreased survival rates during the long-term periods after surgical ventricular restoration; however, mitral regurgitation was not found to be an independent predictor of poor outcome.

Statement of Conflict of Interest
The authors state no confl ict of interest.