We studied a single-center retrospective cohort of patients who underwent transthoracic echocardiography from 2010 to 2018 and were found to have at least mild sMR and reduced LVEF (defined as <50%). Patients had to be ambulatory and have been on guideline-directed heart failure (HF) therapy for at least three months before being included. Patients aged <18 years, hospitalized for decompensated HF in the previous three months, with at least moderate aortic valve disease, previous valve intervention, hypertrophic cardiomyopathy, arrhythmogenic right ventricular cardiomyopathy, left ventricular non-compaction cardiomyopathy or renal transplant were excluded. The study protocol was reviewed and approved by the local ethics committee, which waived the need for informed consent.

Transthoracic echocardiographic examinations were conducted with GE-Healthcare (Chicago, IL, USA) equipment. Data were retrieved from EchoPAC software (GE-Healthcare) for retrospective evaluation. Two-dimensional, M-mode and Doppler measurements (including MR quantification) were used based on the criteria of the ASE.

During follow-up, patients had regular medical visits. Therapy adjustments were left at the discretion of the attending physician. The primary endpoint was all-cause mortality.

Two frameworks for the evaluation of MR disproportionality were assessed in this cohort.

According to Lopes et al., proportionality was determined based on calculation of an individualized theoretical EROA cutoff determined by the formula: 50%×LVEF×LVEDVMitralVTI.7 EROA measured using the PISA method was then compared with this theoretical cutoff.7 Briefly, if measured EROA>individualized theoretical EROA, the patient was considered to have disproportionate sMR.7 If not, sMR was considered proportionate/non-severe.7

In order to determine disproportionality according to Grayburn et al., EROALVEDV ratio was calculated.5 The proportionality cutoff used was 0.14, according to published data (line of proportionality).6 Even though this cutoff was established specifically for patients with a LVEF of 30% at a RF of 50%, we considered that a ratio >0.14 identified disproportionate sMR, whereas a ratio ≤0.14 was categorized as non-severe/proportionate.

Mitral regurgitation severity was classified according to the EROA cutoffs proposed by the ASE classification to allow meaningful comparisons between the two above mentioned formulas of sMR proportionality (grade I [mild] if EROA<0.2 cm2; grade II [moderate] if EROA 0.2-0.29 cm2; grade III [moderate-severe] if EROA 0.3-0.39 cm2; and grade IV [severe] if EROA ≥0.4 cm2).8

Categorical variables are presented as frequencies and percentages, and continuous variables as mean and standard deviation, or median and interquartile range for variables with skewed distributions. Differences among groups were assessed using Pearson's chi-squared test, Mann-Whitney U, and independent samples t-test, where appropriate. κ statistic was used to assess agreement between proportionality formulas. Kaplan-Meier survival curves were plotted for each group of patients. Patients were censured if mitral intervention or heart transplant/left ventricular assist device was performed. The log-rank test was used to assess for significant differences in time to endpoint between groups. Univariate and multivariate analysis with Cox regression were applied to assess the association between studied variables and all-cause mortality. Variables with a p<0.05 were included in the multivariable model. Net reclassification index (categorical NRI) was used to ascertain whether proportionality enhances the prognostic value of MR grade stratification according to ASE guidelines. All reported p-values are two-tailed, with a p=0.05 indicating statistical significance. Analysis was performed using IBM SPSS Statistics software, version 25 (2017).

A total of 572 patients (mean age 69±12 years; 76% male) were included. The majority of patients were in class II (55.8%) or III (33.4%) of New York Heart Association (NYHA). There were 526 (92.0%) patients treated with beta blockers, 525 (91.8%) with angiotensin converting enzyme inhibitor/angiotensin II receptor blocker (ACEi/ARB), 235 (41.1%) with mineralocorticoid receptor antagonist, 170 (29.7%) had an implantable cardiac defibrillator (ICD), and 167 (29.2%) were under cardiac resynchronization therapy (CRT). Mean LVEF was 33±9%, with a median LVEDV of 174 mL (interquartile range (IQR): 136-220), and a median EROA of 14 mm2 (IQR: 8-22). According to ASE guidelines sMR was classified as mild (grade I) in 418 (73%) patients, moderate (grade II) in 109 (19%), moderate-to-severe (grade III) in 33 (6%) and severe (grade IV) in 12 (2%). Other clinical characteristics and echocardiographic parameters are further described in Table 1.

According to Grayburn's model, 62 patients (11%) were classified as disproportionate, and 510 (89%) as non-severe/proportionate. On the other hand, Lopes’ model categorized 96 (17%) patients as disproportionate and 476 (83%) as non-severe/proportionate. Amongst the 96 patients with disproportionate sMR Lopes’ formula, 46 (48%) were considered proportionate using Grayburn's formula; and of the 62 patients with disproportionate sMR according to Grayburn's, 12 (19%) were considered proportionate by Lopes’ (meaning only a moderate agreement between the two frameworks as depicted by Cohen's κ=0.58; p<0.001).

During a mean follow-up of 4.1±2.7 years, there were 254 (44.4%) deaths. Kaplan-Meier survival curves for both models showed significant association between the presence of disproportionality and all-cause mortality (Figures 1 and 2). However, and after adjusting for several confounding variables associated with prognosis (namely, age, sex, creatinine, hypertension, atrial fibrillation, ischemic etiology, NYHA functional class, beta blockers, ACEi/ARB, diuretics, ICD, CRT, LVEDV, LVEF, E, E/e’, left atrial volume index, systolic pulmonary artery pressure, TAPSE and TR≥moderate), only Lopes’ formula maintained independent prognostic impact (adjusted HR 1.5 (95% CI 1.07-2.1), p=0.018; vs. Grayburn's formula: adjusted HR 1.0 (95% CI 0.67-1.5), p=0.998) – Table 2.

Figure 1.

Kaplan-Meier survival curves for Grayburn's framework subgroups.

(0.07MB).

Figure 2.

Kaplan-Meier survival curves for Lopes’ framework subgroups.

(0.08MB).

The reclassification of ASE SMR severity is presented in Figure 3 and Table 3. Only Lopes’ model was able to distinguish lower and higher risk subsets of patients according to the proportionality status when added to ASE SMR classification (NRI=0.129; p<0.001). Grayburn's model for disproportionality showed a non-significant NRI of 0.003 (p=0.455) when added to the ASE sMR classification and did not improve risk prediction.

Figure 3.

a) Distribution and mortality of American Society of Echocardiography mitral regurgitation stratified according to proportionality sub-groups for both frameworks.

ASE: American Society of Echocardiography; fMR: secondary/functional mitral regurgitation.

(0.24MB).

Table 3.

Net reclassification index of American Society of Echocardiography mitral regurgitation classification: a) Grayburn et al. formula; b) Lopes et al. formula.

ASE: American Society of Echocardiography; sMR: secondary/functional mitral regurgitation.

Our study has some limitations that should be acknowledged. This was a retrospective, single center study, and our conclusions should be viewed as hypothesis generating due to the possible presence of unmeasured confounding factors and selection bias, and should be further validated in prospective studies. Also, the method used for evaluation of MR severity was two dimensional (2D) echocardiography, which has various limitations and imperfect concordance of severity among measured parameters.15,16 The regurgitant orifice in sMR is frequently semilunar or elliptical, thus affecting measurements and leading to possible underestimation of the EROA using the 2D PISA method. Furthermore, since volume status and blood pressure were not readily available on electronic records, they were not taken into account in the interpretation of sMR severity, which is notoriously dynamic. Moreover, caution should be taken when this concept is applied to patients with mild MR where measurement errors may be present.