Long-term outcomes of left ventricular posterior wall plication for ischemic mitral regurgitation

Abstract

Purpose

To evaluate the early and long-term outcomes of left ventricular posterior wall plication for ischemic mitral regurgitation.

Methods

Patients with ischemic mitral regurgitation who underwent left ventricular posterior wall plication via right-sided left atriotomy at our institution between 2010 and 2020 were retrospectively reviewed. Cases with normal cardiac function, left ventricular end-systolic diameter < 50 mm, and left ventriculotomy approach were excluded.

Results

The mean follow-up period was 5.3 years [standard deviation (SD) = 3.5], with a maximum of 10 years. Among the 21 patients enrolled, 9 had New York Heart Association (NYHA) class ≥ III. Three patients required preoperative inotrope support, while two preoperative ventilator support. The mean left ventricular ejection fraction was 31.4% (SD: 8.6), and 16 patients had mitral regurgitation grade ≥ III. All patients underwent coronary artery bypass grafting and mitral annuloplasty. Concomitant surgeries included 11 chordae cutting and 3 tricuspid annuloplasties. One in-hospital death occurred due to sepsis. At the follow-up, echocardiographic data showed significant improvement in cardiac dilation and function and good control of mitral regurgitation. The serum brain natriuretic peptide level was significantly reduced, and 85% of patients improved to NYHA class I. Four deaths occurred later due to sudden, unknown causes. The 5- and 8-year survival rates were 60.2% and 46.8%, respectively, and the 5- and 8-year hospitalization rates due to heart failure were 14.9% and 21.3%, respectively.

Conclusion

The long-term outcomes of left ventricular posterior wall plication were satisfactory for controlling heart failure and improving survival rate and patient prognosis.

Supplementary Information

The online version contains supplementary material available at 10.1007/s12055-023-01527-2.

Introduction

Ischemic mitral regurgitation (IMR), a well-known complication caused by left ventricular remodeling after myocardial infarction (MI), causes heart failure (HF) and is associated with a poor prognosis [1]. Although medical therapies have been developed for severe HF, even when caused by ischemic heart disease, treatment options for IMR remain limited. Alternative surgical procedures, such as undersized annuloplasty, chordae cutting, and papillary muscle (PM) relocation or plication, have been explored, but no surgical methods have been thoroughly established. Left ventricular remodeling associated with MI changes the shape of the left ventricle (LV) from cylindrical to spherical due to the distension of the LV wall in ischemic myopathy. The PM is then dislocated, causing leaflet tethering and coaptation mismatch, which leads to IMR [2]. To repair this, a ventricular or subvalvular approach during surgical intervention has been suggested. In our hospital, a surgical procedure called left ventricular posterior wall plication (LVPWP) for IMR treatment has been performed for a long time, and we have previously reported a case using this technique [3]. This study evaluates the early and long-term outcomes of LVPWP as an IMR treatment.

Methods

Patients

This single-center, retrospective study was approved by the Institutional Review Board (IRB) of our hospital (no. 985, dated 8/6/2021). The requirement for informed consent was waived because of the retrospective and observational nature of the study.

Between June 2010 and March 2020, 38 patients underwent LVPWP at our hospital. Five patients with dilated cardiomyopathy, ten with a left ventricular end-systolic diameter (LVESD) < 50 mm with preoperative echocardiography, and two who underwent left ventriculotomy were excluded from this study. Therefore, 21 patients were enrolled and retrospectively reviewed for early outcomes, late mortality (cardiac-related death, including sudden death due to unknown cause), control of mitral regurgitation (MR), and hospitalization rate due to HF.

Patients diagnosed with IMR were carefully examined by cardiologists by using multiple modalities to diagnose and evaluate the severity of MR, LV contraction, coronary artery stenosis, myocardial viability, and HF degree. Patients were then treated with the optimal medical therapy to control HF, followed by device therapy and/or correction of any atrial arrhythmias. Surgical indications were considered by an expert panel of cardiovascular surgeons and cardiologists.

Assessment of cardiac geometry, function, mitral regurgitation, and viability

Transthoracic echocardiography (TTE) and transesophageal echocardiography were performed to evaluate preoperative cardiac function and wall motion, geometry, MR severity, valvular morphology, and subvalvular apparatus. Geometric parameters, such as ejection fraction (EF), left ventricular end-diastolic diameter (LVEDD), and LVESD, were measured and calculated using the modified Simpson method. The plication site was determined based on the preoperative TTE showing akinesis or severe hypokinesis at the myocardial wall. Myocardial viability at the site was evaluated with scintigraphy or enhanced magnetic resonance imaging.

Surgical techniques (Supplementary Video 1)

The surgical technique has been previously described by us [3]. The approach to the LV was right-sided left atriotomy through the mitral valve (MV) via the left atrium. Using 4–0 polyvinylidene fluoride sutures with a small felt strip, plication was performed by suturing the trabeculae carneae with a horizontal mattress stitch to lightly plicate the endocardial side of the posterior wall of the LV from the inside. The number of sutures requiring plication was dependent on the cases, and each suture was mostly arranged vertically in the direction of the apex.

Medical treatment and follow-up

Optimal medical therapy was resumed after surgery with angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, β-blockers, angiotensin receptor neprilysin inhibitors, sodium-glucose cotransporter 2 inhibitors, and diuretics for HF. TTE was conducted post-surgery, and patients were monitored every 6–12 months.

Statistical analysis

The distribution of continuous variables was examined using the Shapiro–Wilk test. Parametric continuous data are expressed as the mean (standard deviation [SD]), whereas non-parametric data are expressed as the median (interquartile range [IQR]). Categorical variables are presented as numbers and percentages. Perioperative data of continuous variables were compared using paired t-test for normal distribution and the Wilcoxon rank-sum test for non-normal distribution. Cumulative survival was calculated using the Kaplan–Meier method to estimate long-term survival, and the competing risks regression model was developed using the Fine-Gray method. All statistical analyses were performed using SPSS version 25 (IBM Corp., Armonk, NY, USA). Differences were considered statistically significant at p < 0.05.

Results

The mean follow-up period was 5.3 years (SD = 3.5), with a maximum of 10 years, and the follow-up rate was 100%.

Table 1 summarizes the patient characteristics. There were 19 (90%) males (mean age, 68.3 years). Nine patients (43%) were in New York Heart Association (NYHA) classes III and IV. Preoperative inotropic and ventilatory support was required for three (14%) and two (10%) patients, respectively. Three patients (14%) were insulin users, and one (5%) had a history of cardiac surgery. Two patients (10%) needed an emergency operation, and the preoperative median brain natriuretic peptide (BNP) was 479.6 pg/dL. The median of surgical risk, which was calculated using the European System for Cardiac Operative Risk Evaluation II risk model, was 9 points.

Table 1.

Characteristics of the study population

Total (n = 21)
Age (years), mean (SD)	68.3 (9.2)
Male sex, n (%)	19 (90)
NYHA class II, n (%)	8 (38)
NYHA class III, n (%)	3 (14)
NYHA class IV, n (%)	6 (29)
Comorbidities
Hypertension, n (%)	16 (76)
Dyslipidemia, n (%)	16 (76)
Diabetes on insulin, n (%)	3 (14)
CKD, n (%)  Serum creatinine (median, IQR) (mg/dL)	15 (71) 1.66 (1.37–3.07)
Chronic lung disease, n (%)	4 (19)
History of ventricular arrhythmia, n (%)	2 (10)
Previous PCI, n (%)	5 (24)
Previous cardiac surgery, n (%)	1 (5)
Preoperative inotrope support, n (%)	3 (14)
Preoperative ventilatory support, n (%)	2 (10)
Operative status: emergency, n (%)	2 (10)
Euro SCORE II, median (IQR)	9.0 (6.6–15.3)
BNP (pg/mL), median (IQR)	479.6 (258.0–732.7)

BNP brain natriuretic peptide; CKD chronic kidney disease; Euro SCORE II European System for Cardiac Operative Risk Evaluation II; IQR interquartile range; HD hemodialysis; NYHA New York Heart Association functional class; PCI percutaneous coronary intervention; SD standard deviation

Table 2 shows the preoperative echocardiographic data. The mean left ventricular ejection fraction (LVEF) was 31.4%, while the mean LVEDD and LVESD were 63.4 and 54.4 mm, respectively. Sixteen patients (76%) had an MR grade ≥ III, whereas ten (48%) had a tricuspid regurgitation grade ≥ II.

Table 2.

Preoperative echocardiographic parameters

Variables	(n = 21)
LVEF (%), mean (SD)	31.4 (8.6)
LVEDD (mm), mean (SD)	63.4 (6.3)
LVESD (mm), mean (SD)	54.4 (4.5)
EDV (ml), mean (SD)	171.6 (32.5)
ESV (ml), mean (SD)	120.8 (28.0)
Mitral regurgitation grade ≥ III, n (%)	16 (76)
Tenting height (mm), mean (SD)	9.4 (3.1)
Tricuspid regurgitation grade ≥ II, n (%)	10 (48)
Right ventricular systolic pressure (mmHg), median (IQR)	29.7 (25.1–55.4)

EDV end-diastolic volume; ESV end-systolic volume; IQR interquartile range; LVEDD left ventricular end-diastolic diameter; LVEF left ventricular ejection fraction; LVESD left ventricular end-systolic diameter; SD standard deviation

Table 3 summarizes the concomitant procedures and early outcomes. One patient (5%) needed mechanical support with Impella (Abiomed, Danvers, MA, USA) before surgery, and two (10%) needed intra-aortic balloon pumping for weaning from cardiopulmonary bypass (CPB). The number of median LVPWP sutures was 4 (IQR = 3–5). The mean CPB and median aortic cross-clamping times were 281.3 and 143 min, respectively. MV annuloplasty and coronary artery bypass grafting (CABG) were performed in all patients. The median size of the annuloplasty ring was 30 mm, including 18 cases (86%) of semirigid ring and 3 cases (14%) of flexible ring, and 19 cases (91%) of total ring, and the median number of revascularized vessels was 4. Chordae cutting was performed in 11 patients (52%), and no patient required MV replacement. Aortic valve replacement was performed in two patients (10%), tricuspid valve annuloplasty in three (14%), and left atrial appendage resection in six (29%). One patient (5%) died within 30 days postoperatively due to sepsis, and seven (33%) needed over 72 h of ventilatory support. The median duration of intensive care unit stay and postoperative hospitalization was 8 and 23 days, respectively. Postoperative complications in patients included reoperation for bleeding in five (24%), stroke in one (5%), new dialysis in four (19%), and pneumonia in three (14%).

Table 3.

Concomitant procedures and early outcomes

Total (n = 21)
LVPWP sutures, median (IQR)	4 (3–5)
Mechanical support (IABP/Impella), n (%)	Preoperative: 1 (5) Scheduled: 10 (48) Intraoperative: 2 (10)
Cardiopulmonary bypass time (minutes), mean (SD)	281.3 (51)
Aortic cross-clamping time (minutes), median (IQR)	143 (122.5–179)
Concomitant procedure, n (%)
MAP	21 (100)
Chordae cutting	11 (52)
CABG, n (%), number of revascularized vessels, median (IQR)	21 (100), 4 (3–5)
AVR	2 (10)
TAP	3 (14)
LAAR	6 (29)
PVI	1 (5)
LV lead implantation	1 (5)
Prolonged ventilation (> 72 h), n (%)	7 (33)
Tracheotomy, n (%)	3 (14)
ICU stay (days), median (IQR)	8 (4.5–14.5)
Postoperative hospitalization (days), median (IQR)	23 (15.5–36.5)
Complications, n (%)
Reoperation for bleeding	5 (24)
Stroke	1 (5)
New dialysis	4 (19)
Mediastinitis	0 (0)
NOMI	0 (0)
Pneumoniae	3 (14)
Early mortality, n (%)	1 (5)

AVR aortic valve replacement; CABG coronary artery bypass grafting; IABP intra-aortic balloon pumping; ICU intensive care unit; IQR interquartile range; LAAR left atrial appendage resection; LV left ventricular; LVPWP left ventricular posterior wall plication; MAP mitral annuloplasty; NOMI non-occlusive mesenteric ischemia; PVI pulmonary vein isolation; SD standard deviation; TAP tricuspid annuloplasty

The results of the follow-up data of LVPWP are summarized in Fig. 1 as echocardiographic data and in Fig. 2 as a change in BNP levels and NYHA classification. The mean echocardiographic follow-up times were 5.3 years. LVEF did not change significantly at discharge compared with that of the preoperative state; however, it showed significant improvement at the follow-up. LVEDD (63.4 mm [SD = 6.3] to 54.2 mm [SD = 6.3], p < 0.0001), LVESD (54.4 mm [SD = 4.5] to 45.5 mm [SD = 7.1], p < 0.0001), end-diastolic volume (171.6 ml [SD = 32.5] to 134.0 ml [SD = 30.4], p < 0.0001), and end-systolic volume (120.8 ml [SD = 28.0] to 94.0 ml [SD = 28.3], p < 0.0001) significantly decreased after surgery. At the follow-up, the LV diameter continued to shrink, and right ventricular systolic pressure significantly decreased (34.0 mmHg [IQR = 27.2–54.3] to 29.1 mmHg [IQR = 21.0–34.0], p = 0.0107). During the follow-up, BNP significantly decreased (479.6 pg/mL [IQR = 258.0–732.7] to 160.8 pg/mL [IQR = 84.3–202.4], p = 0.0002), 89% of patients had a less than mild level of MR, and 85% of patients were NYHA class I and in good condition.

Fig. 1.

Follow-up echocardiographic data of LVPWP. LVEDD/LVESD (a) and their respective volumes (b) significantly decreased after the surgery and did not worsen at follow-up. LVEF (c) and RVSP (d) significantly improved during follow-up EDV, end-diastolic volume; ESV, end-systolic volume; LVEDD, left ventricular end-diastolic diameter; LVEF, left ventricular ejection fraction; LVESD, left ventricular end-systolic diameter; RVSP, right ventricular systolic pressure

Fig. 2.

Changes in HF indicators. BNP levels decreased significantly at follow-up (a), and 85% of patients had NYHA class I and were in good condition during follow-up (b) BNP, brain natriuretic peptide; NYHA, New York Heart Association

For long-term outcomes, the 5- and 8-year survival rates were 60.2 ± 11.0% (95%CI: 0.36—0.78) and 46.8 ± 12.0% (95%CI: 0.12 – 0.68), respectively (Fig. 3a). A total of ten (50%) late deaths occurred, including four (40%) sudden deaths due to unknown causes and six (60%) due to non-cardiac death (three of pneumonia, one of renal failure, senility, and accidental death). No deaths were recorded in which the cause could be clearly identified as cardiogenic. The cumulative incidence of cardiac-related death, including sudden death due to unknown cause and MR degree ≥ III at the 8-year, were 21.6% (95%CI: 0.03 – 0.09) and 9.5% (95%CI: 0.02 – 0.26), respectively (Fig. 3b, c). The 5- and 8-year cumulative incidence of hospitalization rates due to heart failure were 14.9% (95%CI: 0.04 – 0.33) and 21.3% (95%CI: 0.07 – 0.42), respectively (Fig. 3d). Four late hospitalizations due to HF were observed, one of which was triggered by atrial fibrillation, one by severe aortic stenosis, and two by poor adherence to medication or salt restriction. The late cardiac event, including cardiac and sudden death due to unknown causes, and hospitalization due to HF are summarized in Table 4.

Fig. 3.

Kaplan–Meier curves of cumulative survival rate (a), cumulative incidence of cardiac-related death rate including death due to unknown cause (b), the degree of MR ≥ III, and hospitalization rate due to heart failure (c, d). The 5- and 8-year survival rates were 60.2% and 46.8%, respectively. The cumulative incidence of cardiac-related death, including death due to unknown cause and the degree of MR ≥ III at the 8-year were 21.6% and 9.5%. The 5- and 8-year cumulative incidence of hospitalization rates due to HF were 14.9% and 21.3%, respectively. HF, heart failure; MR, mitral regurgitation

Table 4.

Cardiac-related events in the late period

Patient number	Cardiac-related event	Cause of event	Duration from operation (days)
No. 1	Death	Unknown	3,255
No. 2	Death	Unknown	127
No. 3	Death	Unknown	763
No. 4	Death	Unknown	1,238
No. 5	HF hospitalization	Atrial fibrillation	2,785
No. 6	HF hospitalization	Atrial stenosis	1,291
No. 7	HF hospitalization	Poor adherence	789
No. 8	HF hospitalization	Poor adherence	112

HF heart failure

Discussion

The pathophysiology and mechanisms of IMR have gradually become clearer. MI causes left ventricular remodeling; the shape of the LV changes from cylindrical to spherical, and the distance of bilateral PMs is dislocated, causing a coaptation mismatch of the bilateral leaflet and IMR [2]. Thus, undersized annuloplasty alone can cause restriction of the antero-posterior distance and result in functional mitral stenosis [4, 5]. Therefore, PM approximation using two U-shaped pledgeted sutures [6, 7] or relocation [8] and chordae cutting of the subvalvular apparatus [9–11] were reported as the preferred subvalvular approach. Furthermore, considering the mechanism of IMR mentioned above, we hypothesized that it would be effective to approach the enlarged LV to repair IMR and the MR without LV incision.

As a ventricular approach, Isomura et al. [12] first reported the long-term results of the surgical ventricular reconstruction of the posterior wall (posterior restoration procedures [PRPs]); results showed significant improvements in EF, excellent control of MR, and a 5-year survival rate of 66%, despite severe cardiac insufficiency with a median preoperative EF of 25% and a median LVEDD of 74 mm. PRPs seem to be effective in directly shaping the LV from spherical to cylindrical and restoring the PM location; however, they carry the risk of incising the LV, and the plication site is limited between the two PMs. The 5-year survival rate of 60.2% in our study was acceptable compared with the rate of 66% reported by Isomura et al. [12]. The survival rate is comparable, and the recurrence of moderate or severe MR is lower to the surgical technique of papillary muscle relocation or chordae cutting in combination with downsized ring annuloplasty [13, 14]. Moreover, LVPWP does not require LV incision, thus eliminating these anatomical limitations. LVPWP allows plication inside the LV at any site of the dilated wall with severe hypokinesis or akinesis—without LV incision. This procedure directly plicates a posterior bulged LV, restores the shape, corrects PM displacement, promotes reverse remodeling, decreases leaflet tethering, and regulates MR. PRPs pose a risk of postoperative bleeding owing to LV incision, whereas the incision of the normal myocardium can cause cardiac dysfunction. Additionally, this procedure involves suturing the trabeculae carneae with horizontal mattress stitches to gently plicate the endocardial side of the LV's posterior wall from the inside. This is based on the identification of scar lesions through preoperative echocardiography, and helps to avoid injury to the coronary artery. If the artery is damaged, it would not impact cardiac function since it is scar tissue.

In the surgical treatment of IMR, reducing MR is important, but the procedure alone is insufficient to improve long-term outcomes. To plicate LV, ventricular wall stress at the site of the plication site will decrease, the LV will change shape, and the ejection efficiency will improve (Fig. 4a, b), thus leading to good long-term HF control. A surgical procedure for subvalvular tissue alone, such as PM approximation or relocation, cannot manage wall stress for the enlarged LV. In that regard, LVPWP is thought to be superior to other techniques.

Fig. 4.

a Illustration of the change in wall stress (yellow arrow) before and after LVPWP. After performing LVPWP, the wall stress is expected to decrease at the site of the posterior or inferior wall since ventricular wall stress is proportional to the left ventricular diameter and inversely proportional to the left ventricular wall thickness. b Example of preoperative and postoperative echocardiographic images showing that the posterior wall regained its shape and appeared thickened (white arrow). LVPWP, left ventricular posterior wall plication

Previously, some studies have reported that preoperative MR grade is significantly associated with patients’ prognosis [15]. MR severity is sometimes underestimated in patients with severe cardiac insufficiency [12]. Despite conflicting results on whether to perform intervention on the mitral valve in addition to CABG [16], some studies have reported benefits [17, 18]. Additionally, the main purpose of LVPWP is to increase the efficacy of blood ejection in the LV and to regulate MR to restore LV shape. Therefore, LVPWP should be widely indicated in surgical interventions, regardless of MR severity, since the regulation of MR is also an important prognostic factor. In the current study, five cases with a mild MR level underwent LVPWP, and their clinical course and the regulation of MR in the late period were favorable (Fig. 3c).

In addition, performing additional procedures to LVPWP, such as chordae cutting or annuloplasty, is also important due to IMR’s multifactorial nature.

Chordae cutting has been thought to aggravate LV function, but this has proved otherwise, and nowadays, it is one of the likely many methods that can offer an improvement of IMR [19, 20]. Regarding performing annuloplasty, we selected a relatively larger ring (the median size: 30 mm (IQR = 30–30), to prevent functional mitral stenosis.

Concomitant CABG for areas in which myocardial viability remained is also important in IMR surgery. After LVPWP, wall stress to the plication site will decrease, but it would relatively increase in other areas of the heart; therefore, revascularization by CABG must be performed to improve ischemia on the other sites with residual myocardial viability. Thus, in this study, concomitant CABG was performed in all cases.

Optimal medical therapy after an operation is also essential to treat IMR. In this study, four late hospitalizations due to HF were observed, and two patients had poor adherence to medication or salt restriction. Other causes of hospitalization were valvular or arrhythmia induced. Lastly, 11% of patients had ≥ III MR in the late period, relating to the occurrence of HF. However, this procedure has a possibility to cause the deformity of both papillary muscles, which can cause asymmetric tethering or even prolapse after annuloplasty; therefore, careful follow-up will be needed.

Two patterns have been reported concerning leaflet tethering causing IMR [21]: asymmetric tethering and symmetric tethering. Asymmetric tethering involving the posterior wall akinesis or severe hypokinesis is suitable for LVPWP. Moreover, plication, other than in the posterior or inferior wall or in patients with dilated cardiomyopathy or right-sided cardiac insufficiency, is ineffective because it can cause tricuspid regurgitation or LV dysfunction.

Limitations

This study has some limitations. First, this was a single-center, retrospective study with a small sample size. Next, the study lacked a control or comparison group—a major limitation. In addition, several cross-over factors could not be fully excluded because of the patients’ background, preoperative conditions, and concomitant procedures. The echocardiographic techniques to calculate LV volumes are notoriously inaccurate and observer dependent. Moreover, the plication site is decided by each surgeon’s judgment and cannot be generalized or visualized. Finally, our study was not adequately powered to detect a clinically relevant conclusion; however, LVPWP remains theoretically reasonable to treat IMR, and we believe our study will help clarify a clinical process or clue of surgical treatment that has been unclear until now. Therefore, further research in this area is warranted.

Conclusion

Despite the low EF and severe chronic cardiac insufficiency, the long-term surgical outcomes of LVPWP for treating IMR associated with ischemic myocardial damage of the posterior LV wall were satisfactory in most cases. Patients who underwent LVPWP for IMR had a low rate of hospitalization due to HF and good control of MR in the follow-up. This procedure may improve the prognosis and quality of life of patients with IMR. However, IMR is a multifactorial disease, and performing LVPWP alone may not resolve it completely. An optimal treatment plan that includes surgical and medical therapies for each patient should be considered and discussed by an expert panel of cardiologists.

Supplementary Information

Below is the link to the electronic supplementary material.

Authors’ contribution

Conceptualization: KN, TO. Data collection: KN, HM, MH. Manuscript draft: KN. Verification of analytical methods: KT, II, TS. Supervision of findings: TO, HU. Writing–review & editing: all authors.

Funding

No funds, grants, or other support was received.

Data Availability

The data underlying this article are available in the article and in its online supplementary material.

Declarations

Ethics approval

This single-center, retrospective study was approved by the Institutional Review Board (IRB) of our hospital (no. 985, dated 8/6/2021).

Consent

The requirement for informed consent was waived because of the retrospective and observational nature of the study.

Competing interests

The authors have no competing interests to declare that are relevant to the content of this article.