Journal of Clinical Medicine Article Right Ventricular Dysfunction Staging System for Mortality Risk Stratification in Heart Failure with Preserved Ejection Fraction Enrique Santas 1 , Rafael De la Espriella 1 , Francisco Javier Chorro 1 , Patricia Palau 2 , Gema Miñana 1 , Raquel Heredia 1 , Martina Amiguet 1 , Héctor Merenciano 1 , Juan Sanchis 1 , Josep Lupón 3 , Antoni Bayés-Genís 3 and Julio Núñez 1, * 1 Cardiology Department, Hospital Clínico Universitario, Universitat de València, INCLIVA, CIBERCV, avenida Blasco Ibáñez 17, 46010 Valencia, Spain; ensantas@gmail.com (E.S.); rdelaespriella@gmail.com (R.D.l.E.); Francisco.J.Chorro@uv.es (F.J.C.); gemineta@gmail.com (G.M.); rhcambra@gmail.com (R.H.); martinna1990@gmail.com (M.A.); hectormeren@gmail.com (H.M.); sanchis_juafor@gva.es (J.S.) 2 Cardiology Department, Hospital General de Castellón, Universitat Jaume I, avenida de Benicassim 128, 12004 Castellón, Spain; patripalau@gmail.com 3 Cardiology Department, Hospital Germans Trias i Pujol, Universitat Autònoma de Barcelona, CIBERCV, carretera de Canyet s/n, 08196 Badalona, Spain; jlupon.germanstrias@gencat.cat (J.L.); abayesgenis@gmail.com (A.B.-G.) * Correspondence: yulnunez@gmail.com; Tel.: +34-963862658  Received: 22 February 2020; Accepted: 16 March 2020; Published: 18 March 2020  Abstract: Right ventricular dysfunction (RVD) parameters are increasingly important features in heart failure with preserved ejection fraction (HFpEF). We sought to evaluate the prognostic impact of a progressive RVD staging system by combining the tricuspid annular plane systolic excursion (TAPSE) to pulmonary artery systolic pressure (TAPSE/PASP) ratio with functional tricuspid regurgitation (TR) severity. We prospectively included 1355 consecutive HFpEF patients discharged for acute heart failure (HF). Of them, in 471 (34.7%) patients, PASP could not be accurately measured, leaving the final sample size to be 884 patients. Patients were categorized as Stage 1: TAPSE/PASP ≥ 0.36 without significant TR; stage 2: TAPSE/PASP ≥ 0.36 with significant TR; stage 3: TAPSE/PASP < 0.36 without significant TR; and stage 4: TAPSE/PASP < 0.36 with significant TR. By the 1 year follow-up, 207 (23.4%) patients had died. We found a significant and graded association between RVD stages and mortality rates (15.8%, 25%, 31.2%, and 45.4% from stage 1 to stage 4, respectively; log-rank test, p < 0.001). After multivariable adjustment, and compared to stage 1, stages 3 and 4 were independently associated with mortality risk (HR: 1.8219; 95% CI 1.308–2.538; p < 0.001 and HR = 2.2632; 95% CI 1.540–3.325; p < 0.001, respectively). A RVD staging system, integrating TAPSE/PASP and TR, provides a comprehensive and widely available tool for risk stratification in HFpEF. Keywords: heart failure with preserved ejection fraction; right ventricular; risk stratification 1. Introduction Heart failure with preserved ejection fraction (HFpEF) is becoming the dominant form of heart failure (HF). No evidence-based treatments are available and the morbimortality burden remains high [1,2]. While attention in HFpEF has traditionally focused on changes affecting the left heart [3], recent studies have highlighted that right ventricular (RV) dysfunction (RVD) and pulmonary hypertension (PH) are common features that contribute importantly to the pathophysiology and prognosis of this syndrome [1,3–5]. J. Clin. Med. 2020, 9, 831; doi:10.3390/jcm9030831 www.mdpi.com/journal/jcm J. Clin. Med. 2020, 9, 831 2 of 14 Beyond other potentially involved mechanisms, the development of RV dysfunction is mainly related to elevated pulmonary pressures [3,4]. In HFpEF, the RV is markedly dependent on afterload, and the coupling of RV systolic function for a given pressure overload is commonly impaired, so RV contractility gets worse with progressively higher vascular loading [6,7]. RV–pulmonary artery (PA) coupling has emerged as a global index of RV performance and right length–force relationships [8,9]. The ratio of tricuspid annular plane systolic excursion/systolic pulmonary artery pressure (TAPSE/PASP) measured by echocardiography has been proposed as a noninvasive index of RV–PA coupling, showing excellent correlations with PA compliance and distributing accordingly in the PA compliance and pulmonary vascular relationship by invasive data [10,11]. Functional tricuspid regurgitation (TR) is commonly the final consequence of right-side pressure overload and RV–PA uncoupling, leading to further RV remodeling and dysfunction, perpetuating the process [4,5,12]. All of these RV dysfunction (RVD) parameters have been shown to be strongly related to outcomes in HFpEF in isolation [8,10,13–15], but little is known of the prognostic value of the interplay between them. Our aim was to create a comprehensive noninvasive staging system of progressive RVD features in HFpEF, integrating both TAPSE/PASP and TR, and to assess the impact of these stages on survival after a hospitalization for acute HF. 2. Experimental Section 2.1. Study Group and Protocol We prospectively included a consecutive observational cohort of 1355 HFpEF patients who were discharged alive after a hospitalization for acute HF in the cardiology department of a tertiary-care teaching hospital in Valencia (Spain). During index hospitalization, a comprehensive set of demographics, medical history, vital signs, 12-lead electrocardiogram, standard laboratory (72 ± 12 hours after admission), echocardiographic parameters, and treatments at discharge were routinely recorded using pre-established registry questionnaires. Most of the patients exhibited on admission acute decompensated HF (82.7%), followed by acute pulmonary edema (12.6%). Patients were enrolled from 1 January 2007 to 1 August 2016. Patients with new-onset or acutely decompensated HF were enrolled in the registry. HFpEF was defined according to the European Society of Cardiology Clinical Practice Guidelines [2]. We retrospectively evaluate this cohort for the aim of this study. In 471 (34.7%) patients, PASP could not be measured accurately due to the lack of a proper Doppler TR signal, leaving a final sample size of 884 patients. Treatment strategies were individualized following established guidelines that were operating at the time the patient was included in the registry. 2.2. Echocardiography In all patients, a 2-dimensional transthoracic echocardiogram was performed during index hospitalization (96 ± 24 hours after admission) using the left lateral decubitus position. Three commercially available systems were used throughout the study: Agilent Sonos 5500, iE33, and EPIQ 7 (Philips, Massachusetts, USA). Patients were clinically stable and free from intravenous therapies by the time of the examination. All images were recorded with the second harmonic at the time of end-expiration. Left ventricular ejection fraction was assessed by the biplane Simpson method. Left ventricular diastolic dysfunction was evaluated using transmitral flow Doppler velocities and tissue Doppler imaging-derived mitral annulus velocities. PASP was estimated by measuring the maximum continuous Doppler-derived velocity of the TR jet, following established recommendations [16]. Right atrial pressure was estimated in the subcostal view according to inferior vena cava (IVC) size and its breathing-related collapsibility, following a normal sniff as follows: 3 mmHg if IVC < 21 mm that collapses > 50%, 15 mmHg if IVC > 21 mm that collapses < 50%, and an intermediate value of 8 mmHg in the situations in which IVC diameter and collapse did not fit this paradigm [16]. TAPSE was tracked in the four chamber view per M-mode, as recommended [16]. RV–PA coupling was J. Clin. Med. 2020, 9, 831 3 of 14 J. Clin. Med. 2020, 9, 831 3 of 13 evaluated by calculating the ratio of TAPSE to echocardiographic-derived PASP. This is a noninvasive [16]. RV–PA coupling was evaluated by calculating the ratio of TAPSE to echocardiographic-derived index PASP. of global This is RV performance a noninvasive against index a given of global RVdegree of afterload, performance againstproposed by Guazzi a given degree et al. [8], of afterload, that has been proposed bystrongly Guazzi et related al. [l8],tothat outcomes has beenever sincerelated strongly as a noninvasive to outcomesmeasure of as ever since pulmonary a noninvasiveartery distensibility [6,10,11]. Previous studies have shown that the lower the TAPSE/PASP measure of pulmonary artery distensibility [6,10,11]. Previous studies have shown that the lower the ratio, the worse theTAPSE/PASP outcome, but a prognostic ratio, the worse cut-off of <0.36 the outcome, butwas identified a prognostic in most cut-off of the of <0.36 wasstudies [8,9,11] identified andofhas in most been theacknowledged studies [8,9,11] in aand recent has statement by the HeartinFailure been acknowledged a recentAssociation statementof bythe the European Heart Society Failure of Cardiology [4]. So, the variable was categorized according to this cut-off value Association of the European Society of Cardiology [4]. So, the variable was categorized according for statistical analyses. to In this addition, cut-offwe assessed value the adequacy for statistical analyses.of this cut-off inwe In addition, our cohort the assessed by exploring adequacy of thethis functional cut-off inform our of thiscohort ratio as bya exploring continuous variable the functionalin theform multivariate model of this ratio as aforcontinuous all-cause mortality. variable inAccording to current the multivariate model for all-cause recommendations, mortality. the severity of According TR was assessedto current recommendations, by using the severity of TR an integrated multi-parametric was score that assessed by using an integrated multi-parametric score that includes qualitative and semi-quantitative parameters [17]. Color-flow Doppler was examined in theincludes qualitative and semi-quantitative apical parameters 4-chamber, parasternal short[17].andColor-flow long axis,Doppler was examined and subcostal in the apical views. Quantitative 4-chamber, measures of the TRparasternal shortisovelocity jet by proximal and long axis, surfaceandareasubcostal methodviews. Quantitative were available measures in less than 10%of the TR jet by of patients, so it proximal isovelocity surface area method were available in less than 10% was not used in this analysis. TR severity grade was assigned by the cardiac sonographer and was of patients, so it was not used in this confirmed andanalysis. TR severityby finally established grade was investigator another assigned by the cardiac of the studysonographer and was who was blinded to confirmed clinical data. and finally established by another investigator of the study who was blinded to clinical data. The TR The TR severity score comprises four categories: (1) mild, (2) moderate, (3) moderate-to-severe, and (4) severity score comprises four categories: 1) mild, 2) moderate, 3) moderate-to-severe, and 4) severe. severe. All regurgitation jets with a vena contracta ≥ 7 mm and/or systolic flow reversal in hepatic All regurgitation jets with a vena contracta ≥ 7 mm and/or systolic flow reversal in hepatic veins veins were considered to be severe (grade 4). Grades 3 or 4 were labeled as significant TR. RVD stages were considered to be severe (grade 4). Grades 3 or 4 were labeled as significant TR. RVD stages were created combining both TAPSE/PASP and significant TR, as follows: Stage 1: TAPSE/PASP > 0.36 were created combining both TAPSE/PASP and significant TR, as follows: Stage 1: TAPSE/PASP > with no significant TR; stage 2: TAPSE/PASP > 0.36 with significant TR; stage 3: TAPSE/PASP < 0.36 0.36 with no significant TR; stage 2: TAPSE/PASP > 0.36 with significant TR; stage 3: TAPSE/PASP < with 0.36nowith significant TR; and no significant TR;stage 4: TAPSE/PASP and stage 4: TAPSE/PASP < 0.36< 0.36 withwithsignificant TR (Figure significant 1). 1). TR (Figure Figure Figure 1. 1.Proposed Proposedechocardiographic echocardiographic right rightheart heartdysfunction dysfunctionstaging classification staging system classification basedbased system on right ventricular–pulmonary artery coupling and functional tricuspid regurgitation severity. HFpEF: on right ventricular–pulmonary artery coupling and functional tricuspid regurgitation severity. heart failure with preserved ejection fraction; RV–PA: right ventricular–pulmonary artery; HFpEF: heart failure with preserved ejection fraction; RV–PA: right ventricular–pulmonary artery; TAPSE/PASP: tricuspid annular plane systolic excursion to pulmonary artery systolic pressure ratio; TAPSE/PASP: tricuspid annular plane systolic excursion to pulmonary artery systolic pressure ratio; TR: tricuspid regurgitation. TR: tricuspid regurgitation. 2.3.2.3. Outcomes Outcomes The The primaryendpoint primary endpointofofthe thestudy study was was 1-year 1-year all-cause all-cause mortality. mortality. The Thesurvival survivalstatus was status was ascertained by reviewing the electronic medical records, corroborated by an investigator who was ascertained by reviewing the electronic medical records, corroborated by an investigator who was blinded to the RVD staging system. Death due to cardiovascular (CV) causes and death due to HF blinded to the RVD staging system. Death due to cardiovascular (CV) causes and death due to HF were were evaluated as secondary outcomes. CV death included sudden death, death due to HF evaluated as secondary outcomes. CV death included sudden death, death due to HF progression, progression, myocardial infarction, stroke, arrhythmia, or thromboembolic disease. Unknown myocardial infarction, stroke, arrhythmia, or thromboembolic disease. Unknown deaths were also deaths were also considered as CV deaths. considered as CV deaths. J. Clin. Med. 2020, 9, 831 4 of 14 2.4. Ethical Concerns The observational registry was prospectively designed, conformed to the principles outlined in the 1975 Declaration of Helsinki, and approved by the institutional local review ethical committee. All patients gave informed consent. 2.5. Statistical Analysis Continuous variables were expressed as mean ± standard deviation (SD) or median (interquartile range (IQR)), as appropriate. Discrete variables were summarized as percentages. Baseline characteristics were compared among categories with Pearson’s chi-square and p values for trend tests for categorical or continuous variables, respectively. The association between the RVD staging system and 1-year all-cause mortality was evaluated using multivariable Cox proportional hazard models, and the results were expressed as hazard ratios (HR) with 95% confidence intervals (CI). Candidate covariates were initially chosen based on previous medical knowledge; then, a backward stepwise selection was performed. The linearity assumption for all continuous variables was simultaneously tested and the variable transformed, if appropriate, with fractional polynomials. The proportionality assumption for the hazard function over time was tested by means of the Schoenfeld residuals. Final multivariate analysis included all the following covariates assessed during index admission: age, gender, systolic blood pressure at admission, heart rate at admission, Charlson comorbidity index, bundle branch block, left atrial size, blood urea nitrogen, and plasma N-terminal pro-B-type natriuretic peptide (NT-proBNP). The detailed multivariate model including all the covariates and their estimates is presented in the Supplementary Materials (Table S1). The mean variance inflation factor of the covariates in the multivariable model was 1.2. For evaluating CV and HF-related mortality, Cox regression models adjusting for non-CV and non-HF-related death as competing events were used, respectively [18]. The discriminative abilities of the models were assessed by Harrell’s C-statistics. A 2-sided p value of <0.05 was considered to be statistically significant for all analyses. All statistical analyses were performed using STATA 14.1 (StataCorp. 2014. Stata Statistical Software: Release 14.1. College Station, TX, USA). 3. Results 3.1. Study Population Characteristics Baseline clinical characteristics for the overall study cohort (mean age 76.4 ± 9.6 years, 67% female) are listed in Table 1. Mean TAPSE, PASP, and TAPSE/PASP were 18.8 ± 3.5 mm, 48.8 ± 20.3 mmHg, and 0.44 ± 0.17, respectively. The proportion of patients with TR grades 1, 2, 3, and 4 were 48.3%, 28.3%, 12.7%, and 4.3%, respectively. The distribution of patients according to the RVD stages were as follows: 532 patients (60.2%) were in stage 1 (TAPSE/PASP ≥ 0.36 and no significant TR); 40 patients (4.5%) were in stage 2 (TAPSE/PASP ≥ 0.36 and significant TR); 202 patients (22.9%) were in stage 3 (TAPSE/PASP < 0.36 and no significant TR); and 110 patients (12.4%) were in stage 4 (TAPSE/PASP < 0.36 and significant TR) (Figure 2). Table 1 summarizes the baseline characteristics of the study cohort according to RVD stages. Patients with advanced RVD were older, more likely to be female and suffered from a poor functional class prior to admission. Likewise, they showed a higher prevalence of atrial fibrillation, more renal dysfunction, higher NT-proBNP, more surrogates of congestion, greater mitral regurgitation, and more advanced features of diastolic dysfunction. J. Clin. Med. 2020, 9, 831 5 of 14 Table 1. Patient characteristics across right ventricular dysfunction stages. TAPSE/PASP TAPSE/PASP TAPSE/PASP TAPSE/PASP Total ≥ 0.36 without ≥ 0.36 with < 0.36 without < 0.36 with Population p Value Significant Significant Significant Significant (n = 884) TR (n = 532) TR (n = 40) TR (n = 202) TR (n = 110) Age, years 76.1 ± 9.7 75.9 ± 9.9 74.0 ± 11.1 77.7 ± 8.9 77.2 ± 8.3 0.036 Women 569 (64.4) 339 (63.7) 30 (75.0) 144 (71.3) 82 (74.5) 0.041 First HF 494 (55.7) 332 (62.4) 23 (57.5) 96 (47.5) 48 (43.6) <0.001 admission Prior NYHA 159 (17.9) 63 (11.2) 9 (22.5) 55 (27.2) 32 (29.1) <0.001 class III/IV Hypertension 700 (79.2) 429 (80.6) 27 (67.5) 163 (80.7) 81 (73.6) 0.097 Diabetes 347 (39.2) 208 (39.1) 13 (32.5) 87 (43.1) 39 (35.4) 0.445 Mellitus Dyslipidemia 452 (51.1) 284 (53.4) 18 (45.0) 100 (49.5) 50 (45.4) 0.346 Current 59 (6.7) 48 (9.0) 4 (10.0) 5 (2.5) 2 (1.8) 0.002 smoker Ischemic 185 (20.9) 112 (21.0) 7 (17.5) 51 (25.2) 15 (13.6) 0.107 heart disease Charlson 274 (31) 160 (30.1) 9 (22.5) 71 (35.1) 34 (30.9) 0.363 index >2 Heart rate 96.4 ± 30.7 98.8 ± 31.7 94.7 ± 29.1 94.5 ± 29.2 89.4 ± 27.4 0.005 (beats/min) SBP (mmHg) 144.5 ± 29.6 146.8 ± 29.8 133.0 ± 17.1 145.4 ± 32.3 135.6 ± 24.4 <0.001 DBP 77.7 ± 18.9 79.2 ± 19.5 74.8 ± 15.9 76.7 ± 18.1 73.5 ± 16.8 0.017 (mmHg) QRS >120 202 (22.8) 122 (22.9) 7 (17.5) 43 (21.3) 30 (27.3) 0.541 ms Atrial 547 (61.9) 297 (55.8) 31 (77.5) 139 (68.8) 80 (72.7) <0.001 fibrillation Hemoglobin 11.9 ± 1.8 11.9 ± 1.9 11.9 ± 1.8 11.8 ± 1.6 12.0 ± 1.9 0.604 (g/dL) Sodium 138.2 ± 5.0 138.5 ± 4.7 138.2 ± 4.2 138.0 ± 5.5 137.3 ± 5.7 0.173 (mEq/l) NT-proBNP 3653 (5628) 3088 (4351) 2026 (3280) 5282 (7017) 5635 (6364) <0.001 (pg/mL) * Serum Creatinine 1.22 ± 0.59 1.19 ± 0.59 1.10 ± 0.53 1.26 ± 0.58 1.31 ± 0.61 0.016 (mg/dL) BUN 60.9 ± 33.1 56.8 ± 29.1 56.5 ± 30.8 65.7 ± 35.7 73.7 ± 41.0 <0.001 (mg/dL) GFR ( mL/min/1.73 61.9 ± 29.8 63.2 ± 25.9 65.8 ± 25.3 59.9 ± 30.5 58.3 ± 43.9 <0.001 m2 ) CA125 56 (103) 48 (83) 79 (126) 60 (98) 99 (146) 0.001 (U/mL) * LVEF (%) 61.7 ± 7.5 61.8 ± 7.2 62.1 ± 7.6 61.8 ± 8.2 61.2 ± 7.7 0.864 LAD (mm) 45.7 ± 8.2 43.9 ± 7.4 48.3 ± 11.2 46.5 ± 7.3 51.5 ± 9.1 <0.001 J. Clin. Med. 2020, 9, 831 6 of 14 Table 1. Cont. TAPSE/PASP TAPSE/PASP TAPSE/PASP TAPSE/PASP Total ≥ 0.36 without ≥ 0.36 with < 0.36 without < 0.36 with Population p Value Significant Significant Significant Significant (n = 884) TR (n = 532) TR (n = 40) TR (n = 202) TR (n = 110) LAVI 44.6 ± 12.1 41.3 ± 11.5 43.9 ± 12.9 46.4 ± 15.7 50.0 ± 16.9 <0.001 (mL/m2 ) DT (ms) 222.5 ± 62.4 225.3 ± 59.2 241.4 ± 77.2 214.4 ± 60.1 216 ± 72.0 0.026 E/e’ ratio 19.3 ± 11.3 18.7 ± 11.8 19.3 ± 10.5 21.3 ± 10.6 19.1 ± 9.4 0.359 TAPSE (mm) 19.2 ± 3.3 19.9 ± 3.2 20.6 ± 4.0 16.8 ± 3.0 16.8 ± 2.8 <0.001 S’ tricuspid 11.3 ± 3.1 11.8 ± 3.2 12.0 ± 3.0 9.9 ± 2.5 9.9 ± 2.6 <0.001 wave (cm/s) PASP 47.5 ± 16.0 38.8 ± 38.4 45.7 ± 9.2 63.7 ± 26.7 70.4 ± 16.4 <0.001 (mm Hg) Functional 188 (21.3) 79 (14.8) 14 (35.0) 48 (23.8) 47 (42.7) <0.001 MR III/IV Data given as n (%), mean ± standard deviation or median (IQR) *. BUN: blood urea nitrogen; CA125: carbohydrate antigen 125; DBP: diastolic blood pressure; DT: deceleration time; GFR: glomerular filtration rate; HF: heart failure; MR: mitral regurgitation; NT-proBNP: amino-terminal pro-brain natriuretic peptide; NYHA: New York Heart Association; LAD: left atrial diameter; LVAI: left atrium volume index; LVEF: left ventricular ejection fraction; PASP: pulmonary arterial systolic pressure; PH: pulmonary hypertension; SBP: systolic blood pressure; TAPSE: tricuspid annular plane systolic excursion; TR: tricuspid regurgitation. J. Clin. Med. 2020, 9, 831 5 of 13 Figure 2. Distribution of right ventricular dysfunction stages in the total study population. Figure 2. Distribution TAPSE/PASP: of annular tricuspid right ventricular dysfunction plane systolic excursion stages in the artery to pulmonary total study systolicpopulation. pressure ratio; TAPSE/PASP: tricuspid annular TR: tricuspid regurgitation. plane systolic excursion to pulmonary artery systolic pressure ratio; TR: tricuspid regurgitation. 3.2. All-cause Mortality Across RVD Stages Table 1 summarizes At the the baseline 1 year follow-up, characteristics 207 (23.4%) of the patients had died.study cohorta according We found to RVDinstages. stepped increase mortality Patients rates when moving from lower to higher RVD stages (15.8%, 25%, 31.2%, and 45.4%from with advanced RVD were older, more likely to be female and suffered fromastage poor1 to functional class prior to admission. Likewise, they showed a higher prevalence of atrial fibrillation, stage 4, respectively (log-rank test, p < 0.001)). Kaplan–Meier curves show how trajectories of survival more renal across RVD dysfunction, higher stages separated NT-proBNP, early morecontinued and differences surrogates of congestion, increasing along thegreater mitral entire follow-up regurgitation, and more advanced features of diastolic dysfunction. (Figure 3). It is noteworthy that nearly half of the patients in stage 4 died at the 1 year follow-up. Table 1. Patient characteristics across right ventricular dysfunction stages. TAPSE/PASP TAPSE/PASP TAPSE/PASP Total TAPSE/PASP ≥ 0.36 ≥ 0.36 with < 0.36 without < 0.36 with p population without significant significant TR significant TR significant TR (n value (n = 884) TR (n = 532) (n = 40) (n = 202) = 110) rates when moving from lower to higher RVD stages (15.8%, 25%, 31.2%, and 45.4% from stage 1 to stage 4, respectively (log-rank test, p < 0.001)). Kaplan–Meier curves show how trajectories of survival across RVD stages separated early and differences continued increasing along the entire follow-up (Figure 3). It is noteworthy that nearly half of the patients in stage 4 died at the 1 year follow-up. J. Clin. Med. 2020, 9, 831 7 of 14 J. Clin. Med. 2020, 9, 831 7 of 13 Figure 3. Kaplan–Meier estimates for the cumulative event rates of all-cause mortality according to the stage of right ventricular dysfunction. After multivariable adjustment, and compared to stage 1, stage 3 (TAPSE/PASP < 0.36 without Figure 3. significant Kaplan–Meier TR), estimates for the cumulative and particularly event rates of0.36 all-cause withmortality according towere After multivariable adjustment,stage 4 (TAPSE/PASP and compared to stage 1, 0.36) and stage 4 were independently (Table 2). For HF-mortality, related when compared to the outcome. However,to the reference category, a significant excess ofstage 2 (significant risk was not reachedTR for withstage TAPSE/PASP > 3 (Table 2). 0.36) All and stage estimates 4 were of risk for independently all-cause CV and related to the outcome. HF-mortality However, risk are shown a significant in Table 2. excess of risk was not reached for stage 3 (Table 2). All estimates of risk for all-cause CV and HF-mortality risk are shown 4. in Table 2. Discussion In this cohort of patients with HFpEF with a recent hospitalization for acute HF, a simple 4. Discussion and comprehensive RVD echo-derived staging system, integrating TAPSE/PASP and functional TR, In this provided cohort useful of patientsfor information with HFpEF 1-year riskwith a recent hospitalization stratification. This RVD score for showed acute HF,a apositive simple andand comprehensive RVD echo-derived staging system, integrating TAPSE/PASP graded association with mortality. Of note, about half of the patients in stage 4 (RV–PA uncoupling and and functional TR, provided useful significant TR) died information during thefor 1-year first year risk afterstratification. discharge, and This afterRVD score showed a thorough a positive adjustment and including graded association with mortality. Of note, about half of the patients well-established prognosticators and potential confounders, patients at higher stages exhibited anin stage 4 (RV–PA uncoupling and significant increased risk ofTR) died during all-cause, CV, and the first year death. HF-related after discharge, As far as we andknow, after this a thorough adjustment is the first study to includingthe evaluate well-established additive prognostic prognosticators value of RV–PA andcoupling potentialand confounders, TR in HFpEF. patients at higher stages exhibited an increased Left heart dysfunction riskhas of all-cause, CV, and considered been traditionally HF-relatedthe death. As farof landmark asHFpEF, we know, but this this is the first paradigm study to evaluate the additive prognostic value of RV–PA coupling and has changed over the years [1,3–5]. Currently, HFpEF is recognized as a complex and heterogeneous TR in HFpEF. Left heart syndrome in which dysfunction has been different cardiac and traditionally considered the extracardiac abnormalities play landmark a relevantofpathophysiological HFpEF, but this paradigm has changed over the years [1,3–5]. Currently, HFpEF role, beyond diastolic dysfunction [1,2,5]. Several studies have highlighted that RVD is recognized as a complex and PH and are heterogeneous common features syndrome in HFpEF,inpresent whichin different up to 30% cardiac and 80% andofextracardiac abnormalities the patients [4,7,14], play aand respectively, relevant both pathophysiological of role, beyond them contribute importantly todiastolic the clinicaldysfunction expression[1,2,5]. Several studies and prognosis have highlighted of the syndrome, that as has been RVD and PH are common features in HFpEF, present in up to 30% and recently recognized in a position statement on behalf of the Heart Failure Association of the European80% of the patients [4,7,14], respectively, Society and both[4]. of Cardiology of them contribute importantly to the clinical expression and prognosis of the syndrome, as has been recently recognized Several right heart parameters have been inevaluated a positioninstatement recent years on behalf of the in patients Heart with HFpEFFailure [4]. Association of the European Society of Cardiology [4]. However, some of them are invasively obtained or are still experimental [3,4]. In contrast, noninvasive Severalsuch parameters rightasheart TAPSE, parameters PASP, or TR have been evaluated severity in recent years are well-established proxies in of patients withdysfunction right heart HFpEF [4]. However, some of them are invasively obtained or are still experimental and are widely available. They all have a strong prognostic value and are closely interrelated. However, [3,4]. In contrast, noninvasive their prognosticparameters implicationssuch as haveTAPSE, been PASP, or TR severity often evaluated are well-established separately and there is little proxies dataofofright the heart dysfunction and are widely available. They all have a strong prognostic prognostic value of the interplay between them [4–11,14]. Characterization of HFpEF should ideally value and are closely interrelated. integrate However, metrics of RV andtheirPAprognostic implications function, including have been tricuspid valveoften evaluated evaluation separately [3,19]. Given theand factthere that is little data functional TRofis the prognostic usually value of the the consequence interplay of both RVD and between PH, and them its [4–11,14]. Characterization presence causally contributes of HFpEF should ideally integrate metrics of RV and PA function, including to further RV remodeling, dysfunction, and progression of the disease [12,15], we believe a simple tricuspid valve evaluation [3,19]. Given method the fact integrating thethat functional continuum of RVDTR is usually process the consequence in HFpEF, by combining of both RVD metrics ofand RV, PAPH,function, and its presence and causally tricuspid valvecontributes evaluationto further seems RV remodeling, pertinent. All of these dysfunction, parametersand progression should be used of in the disease evaluation [12,15], we believe a simple method integrating the continuum of RVD process in HFpEF, by combining metrics of RV, PA function, and tricuspid valve evaluation seems pertinent. All of these J. Clin. Med. 2020, 9, 831 10 of 14 and risk stratification, but also potentially as therapeutic targets in HFpEF, but ideally they should not be evaluated in isolation. Many years ago, Guazzi et al. proposed a noninvasive global marker of RVperformance and RV–PA coupling, namely the TAPSE/PASP ratio [8]. This is an index that can be easily obtained by echocardiography and can be taken as an in vivo marker of changes in RV length versus developed forces. Data confirming its correlation with invasive RV Ees/ea in HFpEF is lacking. However, the TAPSE/PASP ratio has been shown to have an excellent correlation with invasive PA compliance, distributing along the exponential relationship between PA compliance and pulmonary circulation resistance [10,11], and showing an optimal correlation with the pre-capillary component of PH in HFpEF [20]. This index is a strong predictor of mortality and rehospitalizations in HFpEF beyond its components in isolation [8,10,11,20,21], and its value has been expanded to other conditions beyond HF, such as arterial PH or severe aortic stenosis [22,23]. However, for a global assessment of the RVD severity, it seems mandatory to evaluate the competency of the tricuspid valve, which should no longer be considered as an epiphenomenon. TR has a central role in RVD. It is the pathophysiological consequence of both RV remodeling and PH, leading to further valve tethering and RV enlargement, resulting in more TR in a vicious cycle. Thus, TR is the consequence of RV–PA uncoupling and maladaptive RV remodeling [12,19,24]. Its clinical and prognostic value in HF should not be neglected. TR causes fluid retention and systemic congestion, resulting in inflammation, neuro-hormonal activation, and multi-organ damage [24]. In HFpEF, TR is a strong predictor of death and rehospitalizations [13,25] and increases as RVD progresses over time [26]. In a recent meta-analysis including >30,000 patients with significant TR in different cardiac scenarios, TR was associated with an increased mortality risk independent of RV failure or PH [27]. On the other hand, when a significant functional TR is present, patients with RV failure (but no RV enlargement) have worse clinical outcomes [28]. This is in line with our data, in which TAPSE/PASP prognostic value is not significantly affected by the severity of TR, whereas significant TR provides additive and complementary prognostic value to the RV–PA coupling index. So, we hypothesize that including TR in a staging model could improve risk stratification beyond TAPSE/PASP in patients with HFpEF. As has been shown in our study, patients with reduced TAPSE/PASP and significant TR showed the highest risk of mortality, representing a very high-risk advanced phenotype. In line with our proposed staging system, Bax et al. recently proposed a new staging system for evaluating patients with ischemic heart failure [29]. In this stepwise imaging assessment, the most advanced stage belongs to patients with significant TR as a result of RV remodeling and PH, as was proposed in our study. In the setting of aortic stenosis, another echo-derived staging was recently recommended [30,31]. In this scenario, RV failure was considered to be the most advanced stage. We believe that in the natural history of RVD, RV failure and PH usually precede the development of severe functional TR. This situation belongs to the stage 3 of our proposed schema and is not uncommon (23% of our cohort). On the other hand, sometimes the degree of TR can exceed the magnitude of RV failure, but it is quite uncommon that functional TR may be present without structural RVD (4% of patients in our cohort, stage 2). This stage may include patients with structural valve disease that could have been skipped in the cardiac imaging evaluation. The final step of RVD is determined by the development of significant functional TR, reflecting the failure of the RV. Significant TR is a tipping point in the clinical evolution of patients with HFpEF [4,12,19,24]. In our opinion, the ominous prognosis of patients with both RV failure and significant TR in our cohort (stage 4) confirms the adequacy of this proposed staging system in HFpEF. Left atrial enlargement is an independent risk factor for mortality in our study, beyond RVD features. Left atrial cardiomyopathy is a key mechanism in the pathophysiology and clinical expression of HFpEF [32]. However, at this stage, as a result of RV damage and pulmonary circulation remodeling, RVD becomes the dominant phenotype in many patients, overwhelming left heart dysfunction features. Right-sided dysfunction and failure, rather than left heart disease, seems to become a crucial factor of systemic multi-organ dysfunction and prognosis in HF [33]. J. Clin. Med. 2020, 9, 831 11 of 14 In summary, for a global assessment of the RVD severity in HFpEF, it seems mandatory to evaluate the competency of the tricuspid valve, which should no longer be considered as an epiphenomenon. A simple and widely available staging system, including TAPSE/PASP and TR severity, was significantly and strongly associated to 1-year mortality risk. These data support the clinical use of both parameters in combination for risk stratification in HFpEF, and suggest considering right heart features, such as TAPSE/PASP and TR, as potential therapeutic targets in this complex syndrome Limitations First, this is a single-center observational study in which hidden bias might be operating. Second, echocardiographic studies were not reviewed by an independent core laboratory external to the sonographers. Third, TAPSE has caveats as an index of RV systolic function [16]. Nonetheless, it is a widely available parameter recommended in HFpEF given its simplicity and prognostic data [4]. Advanced imaging techniques, such as RV strain, do not seem to increase the prognostic value of the TAPSE/PASP index [34]. Fourth, a different distribution of stages and/or RV parameters might be designed Others parameters such as S’ or fractional area change were available in a small subset of patients (26% of the cohort) and were not evaluated. However, from our point of view, this stepwise assessment was pathophysiologically considered to be the most adequate and simple tool for an RVD staging in HFpEF. Additionally, this staging system may be easily incorporated into clinical daily practice. Fifth, in patients with very severe TR, Doppler PASP estimation may be inaccurate due to an early equalization of RV and right atrial pressures [16]. Beyond the simplicity of this noninvasive index, hemodynamic evaluation should be required for better characterization in selected patients. Finally, the lack of longitudinal assessment of these RVD parameters precluded inferring the influence of their changes over time on prognosis. 5. Conclusions An RVD staging system, integrating the TAPSE/PASP ratio and functional TR severity, is independently related to a reduced survival in patients with HFpEF. Patients with RV-PA uncoupling and significant TR are a very high risk subgroup. Further studies should confirm these findings and consider RVD stage severity for risk stratification, monitoring, and tailoring treatments in this complex syndrome. Supplementary Materials: The following are available online at http://www.mdpi.com/2077-0383/9/3/831/s1, Figure S1. Estimates of risk of TAPSE/PASP evaluated as a continuous variable in patients with or without significant TR in the multivariable model for all-cause mortality, Table S1. All the covariates and their estimates for all-cause mortality in the final multivariable, Table S2. Multivariable model estimates testing the interaction between right ventricular dysfunction stages and different clinical features for all-cause mortality risk prediction: peripheral edema (yes or no), radiological pleural effusions (yes or no), natriuretic peptides (above vs. below median values), and clinical presentation (acute decompensated heart failure vs. other clinical presentations), Table S3. Univariable and multivariable estimates of risk for all-cause mortality including the absence of a tricuspid regurgitation jet as a variable. Author Contributions: Conceptualization, E.S. and J.N.; methodology, E.S. and J.N.; software, E.S. and J.N.; validation, E.S., R.D.l.E., and J.N.; formal analysis, E.S. and J.N.; investigation, E.S., R.D.l.E., P.P., G.M., R.H., M.A., and H.M., writing—original draft preparation, E.S.; writing—review and editing, E.S., R.D.l.E., A.B.-G., J.L., J.S., F.J.C., and J.N.; visualization, E.S.; supervision, A.B.-G., J.L., J.S., F.J.C., and J.N.; project administration, J.N.; funding acquisition, J.N. All authors have read and agreed to the published version of the manuscript Funding: This work is supported in part by grants from the Ministry of Economy and Competitiveness, Instituto Carlos III (PI17/01426) and CIBER Cardiovascular (16/11/00420 and 16/11/00403). Conflicts of Interest: The authors declare no conflict of interest. J. Clin. Med. 2020, 9, 831 12 of 14 References 1. Lam, C.S.P.; Voors, A.A.; de Boer, R.A.; Solomon, S.D.; van Veldhuisen, D.J. Heart failure with preserved ejection fraction: From mechanisms to therapies. Eur. Heart J. 2018, 39, 2780–2792. [CrossRef] [PubMed] 2. Ponikowski, P.; Voors, A.A.; Anker, S.D.; Bueno, H.; Cleland, J.G.; Coats, A.J.; Falk, V.; González-Juanatey, J.V.; Harjola, V.P.; Jankowska, E.A.; et al. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure of the European Society of Cardiology (ESC). Developed with the special contribution of the Heart Failure Association of the ESC. Eur. Heart J. 2016, 37, 2129–2200. [CrossRef] [PubMed] 3. Chaterjee, N.A.; Steiner, J.; Lewis, G.D. It is time to look at heart failure with preserved ejection fraction from the right side. Circulation 2014, 130, 2272–2277. [CrossRef] 4. Gorter, T.M.; van Veldhuisen, D.J.; Bauersachs, J.; Borlaug, B.A.; Celutkiene, J.; Coats, A.J.S.; Crespo-Leiro, M.G.; Guazzi, M.; Harjola, V.P.; Heymans, S.; et al. Right heart dysfunction and failure in heart failure with preserved ejection fraction: Mechanisms and management. Position statement on behalf of the Heart Failure Association of the European Society of Cardiology. Eur. Heart J. 2018, 20, 16–37. [CrossRef] 5. Parikh, K.S.; Sharma, K.; Fuizat, M.; Surks, H.K.; George, J.T.; Honarpour, N.; Depre, C.; Desvigne-Vickens, P.; Nkulikinyinka, R.; Lewis, G.D.; et al. Heart failure with preserved ejection fraction expert panel report. Current controversies and implications for clinical trials. J. Am. Coll. Cardiol. HF 2018, 6, 619–632. 6. Borlaug, B.; Kane, G.C.; Melenovsky, V.; Olson, T.P. Abnormal right ventricular-pulmonary artery coupling with exercise in heart failure with preserved ejection fraction. Eur. Heart J. 2016, 37, 3294–3302. [CrossRef] [PubMed] 7. Guazzi, M.; Naeije, M. Pulmonary hypertension in heart failure. J. Am. Coll. Cardiol. 2017, 69, 1718–1734. [CrossRef] [PubMed] 8. Guazzi, M.; Bandera, F.; Pelissero, G.; Castelvecchio, S.; Menicanti, L.; Ghio, S.; Temporelli, P.L.; Arena, R. Tricuspid annular plane systolic excursion and pulmonary arterial systolic pressure relationship in heart failure: An index of right ventricular contractile function and prognosis. Am. J. Physiol. Heart Circ. Physiol. 2013, 305, 1373–1381. [CrossRef] [PubMed] 9. Kaye, D.M.; Marwick, T.H. Impaired right heart and pulmonary vascular function in HFpEF. Time for more risk markers? JACC Cardiovasc. Imaging 2017, 10, 1222–1224. [CrossRef] 10. Guazzi, M.; Dixon, D.; Labate, V.; Beussink-Nelson, L.; Bandera, F.; Cuttica, M.J.; Shah, S.J. RV contractile function and its coupling to pulmonary circulation in heart failure with preserved ejection fraction. Stratification of clinical phenotypes and outcomes. J. Am. Coll. Cardiol. Img. 2017, 10, 1211–1221. [CrossRef] 11. Gerges, M.; Gerges, C.; Pristitto, A.M.; Lang, M.B.; Trip, T.; Jakowistch, J.; Binder, T.; Lang, I.M. Pulmonary hypertension in heart failure, epidemiology, right ventricular function and survival. Am. J. Respir. Crit. Care Med. 2015, 192, 1234–1246. [CrossRef] [PubMed] 12. Hahn, R.; Waxman, A.B.; Denti, P.; Delhaas, T. Anatomic relationship of the complex tricuspid valve, right ventricle and pulmonary vasculature. A review. JAMA Cardiol. 2019, 4, 478–487. [CrossRef] [PubMed] 13. Santas, E.; Chorro, F.J.; Miñana, G.; Méndez, J.; Muñoz, J.; Escribano, D.; García-Blas, S.; Valero, E.; Bodí, V.; Núñez, E.; et al. Tricuspid regurgitation and mortality risk across left ventricular systolic function in acute heart failure. Circ. J. 2015, 79, 1526–1533. [CrossRef] [PubMed] 14. Gorter, T.M.; Hoendermis, E.S.; van Valdhuisen, D.J.; Voors, A.A.; Lam, C.S.P.; Geelhoed, B.; Willems, T.P.; van Melle, J.P. Right ventricular dysfunction in heart failure with preserved ejection fraction: A systematic review and meta-analysis. Eur. J. Heart Fail. 2016, 18, 1472–1487. [CrossRef] [PubMed] 15. Lam, C.S.P.; Roger, V.L.; Rodeheffer, R.J.; Borlaug, B.A.; Enders, F.T.; Redfield, M.M. Pulmonary hypertension in heart failure with preserved ejection fraction: A community-based study. J. Am. Coll. Cardiol. 2009, 53, 1119–1126. [CrossRef] [PubMed] 16. Rudski, L.G.; Lai, W.W.; Afilalo, J.; Hua, L.; Handschumacher, M.D.; Chandrasekaran, K.; Solomon, S.D.; Louie, E.K.; Schiller, N.B. Guidelines for the Echocardiographic Assessment of the Right Heart in Adults: A report from the American Society of Echocardiography; endorsed by the European Association of Echocardiography, a registered branch of the European Society of Cardiology, and the Canadian Society of Echocardiography. J. Am. Soc. Echocardiogr. 2010, 23, 685–713. [PubMed] J. Clin. Med. 2020, 9, 831 13 of 14 17. Zohgbi, W.A.; Enriquez-Sarano, M.; Foster, E.; Grayburn, P.A.; Kraft, C.D.; Levine, R.A.; Nihoyannopoulos, P.; Otto, C.M.; Quinones, M.A.; Rakowski, H.; et al. Recommendations for Evaluation of the Severity of Native Valvular Regurgitation with Two-dimensional and Doppler Echocardiography. A report from the American Society of Echocardiography’s Nomenclature and Standards Committee and The Task Force on Valvular Regurgitation, developed in conjunction with the American College of Cardiology Echocardiography Committee, The Cardiac Imaging Committee Council on Clinical Cardiology, the American Heart Association, and the European Society of Cardiology Working Group on Echocardiography. J. Am. Soc. Echocardiogr. 2003, 16, 777–802. 18. Fine, J.P.; Gray, R.J. A proportional hazard model for the subdistribution of a competing risk. J. Am. Stat. Assoc. 1999, 94, 496–509. [CrossRef] 19. Taramaso, M.; Gavazzoni, M.; Pozzoli, A.; Dreyfus, G.D.; Bolling, S.F.; George, I.; Kapos, I.; Tanner, F.C.; Zuber, M.; Maisano, F.; et al. Tricuspid regurgitation. Predicting the need for intervention, procedural success, and recurrence of disease. J. Am. Coll. Cardiol. Imaging 2019, 12, 605–621. 20. Gorter, T.M.; van Veldhuisen, D.J.; Voors, A.A.; Hummel, Y.M.; Lam, C.S.P.; Berger, R.M.F.; van Melle, J.P.; Hoendermis, E.S. Right ventricular-vascular coupling in heart failure with preserved ejection fraction and pre- vs. post-capillary pulmonary hypertension. Eur. J. Cardiovasc. Imaging 2018, 19, 425–432. [CrossRef] 21. Santas, E.; Palau, P.; Guazzi, M.; de la Espriella, R.; Miñana, G.; Sanchis, J.; Bayés-Genís, A.; Lupón, J.; Chorro, F.J.; Núñez, J. Uselfulness of right ventricular to pulmonary circulation coupling as an indicator of risk for recurrent admissions in heart failure with preserved ejection fraction. Am. J. Card. 2019, 124, 567–572. [CrossRef] [PubMed] 22. Tello, K.; Wan, J.; Dalmer, A. Validation of the tricuspid annular plane systolic excursion/systolic pulmonary artery pressure ratio for the assessment of right ventricular-arterial coupling in severe pulmonary hypertension. Circ. Cardiovasc. Imaging 2019, 12, e009047. [CrossRef] [PubMed] 23. Sultan, I.; Cardounel, A.; Abdelkarim, I.; Kilic, A.; Althouse, A.D.; Sharbaugh, M.S.; Gupta, A.; Xu, J.; Fukui, M.; Simon, M.A.; et al. Right ventricular to pulmonary artery coupling in patients undergoing transcatheter aortic valve implantation. Heart 2019, 105, 117–121. [CrossRef] [PubMed] 24. Taramaso, M.; Gavazzoni, M.; Maisano, F. Is tricuspid regurgitation a prognostic interventional target or is it just an indicator of worst prognosis in heart failure patients? Eur. Heart J. 2019, 40, 485–487. [CrossRef] 25. De la Espriella, R.; Santas, E.; Chorro, F.J.; Miñana, G.; Soler, M.; Bodí, V.; Valero, E.; Núñez, E.; Bayés-Genís, A.; Lupón, J.; et al. Functional tricuspid regurgitation and recurrent admissions in patients with acute heart failure. Int. J. Cardiol. 2019, 291, 83–88. [CrossRef] 26. Obokata, M.; Reddy, Y.N.V.; Melenovsky, V.; Pislaru, S.; Borlaug, B.A. Deterioration in right ventricular structure and function over time in patients with heart failure and preserved ejection fraction. Eur. Heart J. 2019, 40, 689–698. [CrossRef] 27. Wang, N.; Fulcher, J.; Abeysuriya, N.; McGrady, M.; Wilcox, I.; Celermajer, D.; Lal, S. Tricuspid regurgitation is associated with increased mortality independent of pulmonary pressures and right heart failure: A systematic review and meta-analysis. Eur. Heart J. 2019, 40, 476–484. [CrossRef] 28. Dietz, M.F.; Prihadi, E.A.; van der Bijl, P.; Goedemans, L.; Mertens, B.J.A.; Gursoy, E.; van Genderen, O.S.; AjmonAjmone Marsane, N.; Delgado, V.; Bax, J.J. Prognostic implications of right ventricular remodeling and function in patients with significant secondary tricuspid regurgitation. Circulation 2019, 140, 836–845. [CrossRef] 29. Bax, J.J.; Di Carli, M.; Narula, J.; Delgado, V. Multimodality imaging in ischemic heart failure. Lancet 2019, 393, 1056–1070. [CrossRef] 30. Généreux, P.; Pibarot, P.; Redfors, B.; Mack, M.J.; Makkar, R.R.; Jaber, W.A.; Svensson, L.G.; Kapadia, S.; Tuzcu, E.M.; Thourani, V.H.; et al. Staging classification of aortic stenosis based on the extent of cardiac damage. Eur. Heart J. 2017, 38, 3351–3358. [CrossRef] 31. Vollema, E.M.; Amanullah, M.R.; Ng, A.C.T.; van der Bijl, P.; Prevedello, F.; Sin, Y.G.; Prihadi, E.A.; Marsan, N.A.; Ding, Z.P.; Généreux, P.; et al. Staging cardiac damage with symptomatic aortic valve stenosis. J. Am. Coll. Cardiol. 2019, 74, 538–549. [CrossRef] [PubMed] 32. Bisbal, F.; Baranchuk, A.; Braundwald, E.; Bayés de Luna, A.; Bayés-Genís, A. Atrial failure as a clinical entity. J. Am. Coll. Cardiol. 2020, 75, 222–232. [CrossRef] [PubMed] J. Clin. Med. 2020, 9, 831 14 of 14 33. Rosenkranz, S.; Howard, L.S.; Gomberg-Maitland, M.; Hoeper, M.M. Systemic consequences of pulmonary hypertension and right-sided heart failure. Circulation 2020, 141, 678–693. [CrossRef] [PubMed] 34. Bosch, L.; Lam, C.S.P.; Gong, L.; Chang, S.P.; Sim, D.; Yeo, D.; Jaufeerally, F.; Long, K.T.G.; Ong, H.Y.; Ng, T.P.; et al. Right ventricular dysfunction in left-sided heart failure with preserved vs reduced ejection fraction. Eur. J. Heart Fail. 2017, 19, 1664–1671. [CrossRef] [PubMed] © 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).