Global longitudinal strain to determine optimal timing for surgery in primary mitral regurgitation: A systematic review

Primary mitral regurgitation (PMR) results in adverse remodeling changes and left ventricular (LV) dysfunction. Assessing LV function has prognostic value in predicting morbidity and mortality. Indications for surgery include parameters such as LV ejection fraction (LVEF) and systolic dimensions. Current guidelines are limited in identifying patients at optimal time for surgery. Impaired postoperative LVEF indicates poor prognostic outcomes and subsequent heart failure. Global longitudinal strain (GLS) via speckle tracking echocardiography (STE) presents as a promising parameter to detect subclinical dysfunction in asymptomatic patients.


| INTRODUCTION
Mitral regurgitation (MR) is defined as retrograde blood flow from the left ventricle (LV) through the mitral valve (MV) into the left atrium (LA). Primary MR (PMR), also known as organic, intrinsic, or degenerative MR, is due to organic or structural dysfunction of the MV. 1 In the Euro Heart Survey, severe MR was the second most common valvular abnormality. 2 With PMR disease progression, contractile function can be preserved, which can mask subclinical LV dysfunction. This may progress to adverse remodeling changes in the absence of clinical symptoms. These changes are often unmasked by changes in hemodynamic and loading conditions after surgical intervention, leading to heart failure (HF). Tribouilloy et al. 3 observed a higher long-term survival with lower operative mortality in asymptomatic patients compared to patients with severe symptoms. The health burden of PMR necessitates identification of early subclinical LV dysfunction in asymptomatic patients, before irreversible remodeling changes occur. In doing so, clinicians can better define the optimal timing for surgery. Surgery is recommended for asymptomatic patients with severe PMR and LV dysfunction (Table 1). 4 LV dysfunction is defined as LV ejection fraction (LVEF) less than 60% and/or LV end-systolic diameter (LVESD) ≥ 45 mm. 4 Surgery is also indicated for symptomatic patients with LVEF > 30%. 4 In addition, surgery is recommended in patients with new-onset atrial fibrillation or pulmonary hypertension (systolic pulmonary arterial pressure [SPAP] > 50 mmHg). Recent work concluded that triggers for Class-I were associated with doubling of long-term death/ HF risk. 5 When triggers present, they necessitate rescue surgery and are not the triggers for preferred surgical timing. 4 LVEF is a load-dependent measure of systolic function. 6 Hemodynamic compensation in chronic volume overload, can preserve LVEF despite decline in myocardial contractile function. 6 Therefore, LVEF is a relatively late marker in detecting myocardial dysfunction. In addition, echocardiographic measurements of LVEF have considerable intraobserver and interobserver variability. 7,8 Therefore, there is a need for reproducible, accurate, and load-independent echocardiographic parameters that can detect early subclinical LV dysfunction in asymptomatic patients.
Cardiac contraction causes a change in myocardial length, described as deformation. Strain defines the intrinsic deformation by application of a force. 9 Myocardial strain represents the percentage Surgery should be considered in asymptomatic patients with preserved LV function (LVEF > 60%) and LVESD 40-44 mm when durable repair is likely, surgical risk is low, the repair is performed in a heart valve center, and at least one of the following findings is present: -Flail leaflet -Presence of significant LA dilatation in sinus rhythm (volume index ≥ 60 ml/m 2 BSA)

IIa
C Mitral valve repair should be considered in symptomatic patients with severe LV dysfunction (LVEF < 30% and/or LVESD > 55 mm) refractory to medical therapy when the likelihood of successful repair is high and comorbidity low IIb C Mitral valve replacement may be considered in symptomatic patients with severe LV dysfunction (LVEF < 30% and/or LVESD > 55 mm) refractory medical therapy when the likelihood of successful repair is low and comorbidity low IIb C Percutaneous edge-to-edge procedure may be considered in patients with symptomatic severe PMR who fulfill the echocardiographic criteria of eligibility and are judged inoperable or at high surgical risk by the Heart Team, avoiding futility Note:  is a novel imaging modality for the assessment of myocardial strain in determining LV global and regional function. It is commonly hypothesized that in many cardiac pathologies, the first change is the loss of function in endocardial and epicardial longitudinal fibers. 11 This is compensated by augmenting circumferential fiber shortening in the mid-wall to preserve LV function. 11 Global longitudinal strain (GLS) appears to be a strong predictor of long-term prognosis due to its potential to identify early LV dysfunction in patients with preserved LVEF ( Figure 2B).
The study of GLS in PMR is a growing interest, indicting diagnostic and prognostic value. GLS by STE has a strong positive correlation with LVEF, especially in LV systolic impairment. 12 Santoro et al. 13

| METHODS
A literature search was performed to organize relevant primary citations to answer the following research question: "Can GLS be used to determine optimal timing for surgery for asymptomatic patients with chronic severe PMR?" In aid of answering this study question the following outcomes were set: 1) To establish a correlation between GLS and LVEF.
2) To assess whether preoperative GLS can detect and predict contractile impairment. Review Board is not applicable. Data sharing not applicable to this article as no data sets were generated or analyzed during the study.

| RESULTS
The key findings from the 12 included primary studies are summarized in Table 2. 20-31

| Correlation between GLS and LVEF
Mascle et al. 23 25 When adjusted for other significant predictors using multivariate analysis, GLS remained the strongest independent predictor of long-term LV dysfunction, followed by LVESD ≥ 40 mm. 25

| Associated clinical outcomes
Magne et al. 24 identified the determinants and impact on outcome of brain natriuretic peptide (BNP) in asymptomatic patients.

| DISCUSSION
Impaired postoperative LVEF after surgery is associated with worse longterm prognosis and greater incidence of HF. 5,32 However, in patients who do not meet Class I guidelines for surgery, there remains equipoise on whether an early surgery versus watchful waiting approach is superior.
Hence, more sensitive and accurate imaging may provide more data on the optimum timing of surgery. The desired outcome of a successful surgery is a functional decrease in LV size and preserved LVEF, after surgery and at long-term follow-up. 33 Many studies adopted the early MV surgery strategy, where recruited patients undergoing surgery had preserved mean LVEF ( > 60%) and impaired GLS at baseline (Table 2) | 2463 dysfunction in PMR, further investigation is required to determine a predictive GLS cut-off that demonstrates both reproducibility and efficacy.

| GLS as a clinical tool
The literature agrees there is a significant correlation between GLS and LVEF. Impaired preoperative GLS was negatively correlated with postoperative LVEF. 22,23 LVEF has long been considered a robust parameter of LV systolic function and a predictor of late survival. 32 The established correlation between these two parameters translates that GLS can be a reliable clinical tool to determine LV systolic function and have similar prognostic utility as LVEF.
A Presently, the Class I recommendation for surgery for patients with PMR requires LV dysfunction to be present. 4 The issue that lies is that current preoperative methods that assess LV impairment later corresponds with poor postoperative outcomes. 5 The literature conclusively highlights GLS has sensitivity to detect subclinical LV impairment, despite patients being asymptomatic with preserved LVEF. 20,21,[25][26][27] In addition, an association was seen between baseline GLS and postoperative outcomes, where impaired GLS was associated with higher mortality rates and increased risk of cardiac events. 24,[28][29][30] Importantly, reproducible findings between studies were seen at short-, mid-and long-term follow-up. This demonstrates the  Although many of the included studies used similar vendors, (e.g. GE Vivid) and processing software, (e.g. EchoPAC), majority of the echocardiographic data collected was before 2015. Therefore, intervendor variability is likely bias.
In addition, the accuracy of GLS calculated is largely dependent on standardized image acquisition and quality. Guidelines stipulate that when regional tracking is suboptimal in more than two myocardial segments in a single view, the calculation of GLS should be avoided. 43 Furthermore, the entire LV myocardium must be discernible on the gray-scaled images, which can pose a particular challenge in the setting of severe LV dilatation. Judgment of suboptimal image acquisition is subjective and may be prone to operator-related errors.
The clinical application of 2D strain imaging to quantify GLS in pathological states requires the definition of a normal range. Reported normal ranges largely vary across studies. 44

| Determining optimal timing for MV surgery
Findings from the reviewed studies conclude that GLS is a sensitive marker to detect subclinical LV dysfunction in asymptomatic patients with chronic severe PMR with additional prognostic value in determining postoperative outcomes. However, there is lack of a standardized baseline threshold value to stipulate when surgery is indicated within this cohort of patients. This poses a challenge in determining the optimal time for surgical intervention. Nonetheless, GLS shows promise in being the parameter of choice to identify patients at the preferred time. Serial measurements can assess relative changes in GLS from baseline and may aid risk stratification particularly during the preoperative active surveillance stage. This has been adopted in other cardiac fields. In cardio-oncology, a relative reduction in GLS more than 15% from baseline can define chemotherapy-related cardiac dysfunction. 50