- Email: [email protected]
Prevalence of, Associations With, and Prognostic Value of Tricuspid Annular Plane Systolic Excursion (TAPSE) Among Out-Patients Referred for the Evaluation of Heart Failure
Journal of Cardiac Failure Vol. 18 No. 3 2012
Prevalence of, Associations With, and Prognostic Value of Tricuspid Annular Plane Systolic Excursion (TAPSE) Among Out-Patients Referred for the Evaluation of Heart Failure THIBAUD DAMY, MD, PhD,1,2,3,4 ANNA KALLVIKBACKA-BENNETT,1 KEVIN GOODE, PhD,1 OLGA KHALEVA, MD,1 MD, PhD,2,3,4 CHRISTIAN LEWINTER, MD,1 JAMES HOBKIRK,1 NIKOLAY P. NIKITIN, MD,1 JEAN-LUC DUBOIS-RANDE, LUC HITTINGER, MD, PhD,2,3,4 ANDREW L. CLARK, MA, MD, FRCP,1 AND JOHN G. F. CLELAND, MD, FACC1 Kingston-upon-Hull, United Kingdom; and Creteil, France
ABSTRACT Background: Prevalence, predictors, and prognostic value of right ventricular (RV) function measured by the tricuspid annular plane systolic excursion (TAPSE) in patients with chronic heart failure (CHF) symptoms with a broad range of left ventricular ejection fraction (LVEF) are unknown. Methods and Results: Of 1,547 patients, mean (6SD) age was 71 6 11 years, 48% were women, median (interquartile range [IQR]) TAPSE was 18.5 (14.0e22.7) mm, mean LVEF was 47 6 16%, 47% had LVEF #45% and 67% were diagnosed with CHF, defined as systolic (S-HF) if LVEF was #45% and as heart failure with preserved ejection fraction (HFPEF) if LVEF was O45% and treated with a loop diuretic. During a median (IQR) follow-up of 63 (41e75) months, mortality was 34%. In multivariable analysis, increasing age, N-terminal proeB-type natriuretic peptide (NT-proBNP), New York Heart Association functional class, right atrial volume index, and transtricuspid pressure gradient; lower TAPSE, diastolic blood pressure, and hemoglobin; and atrial fibrillation (AF) or COPD were associated with an adverse prognosis. Receiver operating characteristic curve analysis identified a TAPSE of 15.9 mm as the best prognostic threshold (P 5 .0001); 47% of S-HF and 20% of HFPEF had a TAPSE of !15.9 mm. The main associations with a TAPSE !15.9 mm were higher NT-proBNP, presence of atrial fibrillation and presence of LV systolic dysfunction. Conclusions: In patients with CHF, low values for TAPSE are common, especially in those with reduced LVEF. TAPSE, unlike LVEF, was an independent predictor of outcome. (J Cardiac Fail 2012;18:216e225) Key Words: Heart failure, prognosis, right ventricle, echocardiography, TAPSE.
The pathogenesis of chronic heart failure (CHF) is complex and varied. Most attention has been paid to left ventricular (LV) dysfunction either with a reduced (systolic;
S-HF) or preserved (HFPEF) LV ejection fraction (LVEF). The pathogenesis of HFPEF may be especially complex and include abnormalities of long-axis systolic function, impaired myocardial relaxation, myocardial or pericardial constraint, or pulmonary hypertension resulting in predominant right-sided failure. Surveys suggest that the prognosis of heart failure (HF) is similar in patients with or without reduced LVEF.1 It is possible that comorbid disease drives the poor prognosis of heart failure patients regardless of LVEF.2 Alternatively, similar downstream hemodynamic effects of LV dysfunction on left atrial pressures, pulmonary artery pressures (PAP), and right ventricular (RV) function could cause the RV to fail with a similar adverse effect on prognosis in S-HF and HFPEF. This could be compounded by pathology of the RV myocardium. Several studies suggest that RV systolic function is a powerful prognostic marker in patients with CHF and in other cardiopulmonary disease.3e5 The prognostic value of RV dysfunction in HF due to systolic dysfunction has
From the 1Department of Cardiology, University of Hull, Castle Hill Hospital, Kingston-upon-Hull, United Kingdom; 2Federation de Cardiologie at the Assistance Publique des H^opitaux de Paris, Groupe HenriMondor Albert-Chenevier, Creteil, France; 3Unite U955, Institut National de la Sante Et de la Recherche Medicale, Creteil, France and 4Faculte de Medecine, Universite Paris 12, Creteil, France. Manuscript received May 27, 2011; revised manuscript received October 12, 2011; revised manuscript accepted December 14, 2011. Reprint requests: Dr Thibaud Damy, Department of Cardiology, Hull York Medical School, University of Hull, Castle Hill Hospital, Kingstonupon-Hull, East Riding, Yorkshire, UK. Tel: 0044 (0) 1 482 461 917; Fax: 0044 (0) 1 482 461 779. E-mail: [email protected] Funding: Dr Damy received a grant from the Societe Francaise de Cardiologie/Federation Francaise de Cardiologie. See page 225 for disclosure information. 1071-9164/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.cardfail.2011.12.003
216
Prognosis Value of RV Function in CHF
previously been studied.4e6 Lam et al recently showed that patients with HFPEF have raised PAP, which may impair RV function.7 However, the determinants of RV dysfunction are still unclear, and their clarification might help to develop new strategies to improve patient outcome. We describe herein the prevalence, determinants, and prognostic value of reduced of tricuspid annular plane systolic excursion (TAPSE), an echocardiographic measure of RV dysfunction,8 in patients presenting with symptoms suggestive of CHF. Methods Study Population and Procedures Patients were referred for the diagnosis and management of suspected CHF between August 2001 and June 2009 by primary and secondary care physicians to a community-based CHF program in Kingston-upon-Hull, UK. Patients were clinically evaluated and had an electrocardiogram, echocardiogram, and blood tests. We divided the patients into 3 diagnostic groups based on symptoms, echocardiography, and use of loop diuretics. Those with LVEF #45% were considered to have S-HF. Those with LVEF O45% and requiring a loop diuretic for treatment of symptoms or signs of congestion were defined as having HFPEF. The remainder, with LVEF O45% and not treated with a loop diuretic, were considered not to have HF (No-HF). Ischemic heart disease (IHD) was defined as a previous myocardial infarction, coronary bypass surgery, or coronary disease found at coronary angiography. Chronic obstructive pulmonary disease (COPD) was considered to be moderate or severe if both FEV1/FVC was !70% and FEV1 was !80% of predicted.9 Patients who were able had a 6-minute walk test.10 Patients with atrial fibrillation or atrial flutter were grouped as ‘‘AF.’’ Estimated glomerular filtration rate (eGFR) was calculated.11 Medications at the time of referral were recorded. Echocardiographic Measurements Echocardiograms were performed in accordance with the recommendations of the British Society of Echocardiography12 using a Vivid Five (GE Health Care, UK) system operating at 3.4 MHz. Echocardiographic methods have been previously published.5,6,13 Ethics The investigation conforms to the principles outlined in the Declaration of Helsinki. It was approved by the Hull and East Yorkshire Research Ethics Committee. All of the subjects gave their written informed consents. Statistical Analyses To explore the relation between RV function and outcomes, the population was divided into quartiles of TAPSE. Normality of distribution for continuous variables was checked using the Kolmogorov-Smirnov test. Values for other normally distributed variables are given as mean 6 SD for quantitative data and as numbers and percentages for categoric data. N-terminal proeB-type natriuretic peptide (NT-proBNP) was logarithmically transformed before entry into statistical models. Differences between continuous data were tested using the Kruskal-Wallis test for the 4 quartile comparisons and the Mann-Whitney test for the 2 group comparisons. Proportions were compared using Pearson chi-square test.
Damy et al
217
Cox proportional hazards analyses were used to assess prognostic associations . Univariable Cox proportional hazards modeling was used and 2 multivariable models were tested to compare the prognostic value of TAPSE. All the clinical and echocardiographic variables presented in Table 2 that had a P value of !.10 in the univariable model were included in model 1. The same variables as model 1 plus biologic variables (NT-proBNP, hemoglobin, and eGFR) were included in model 2. The prognostic value of TAPSE divided by quartiles was analyzed using Kaplan-Meier plots with the chi-square log-rank test. Predictors of TAPSE were assessed with the use of logistic regression with backward-elimination of independent variables based on likelihood-ratio tests. P values were considered to be significant if P ! .05. Analyses were performed using SPSS 16.0 (SPSS, Chicago, Illinois).
Results Characteristics of the Study Population
Of the 2,085 patients referred, 1,547 had a measurable TAPSE. TAPSE was not measured because of either failure to follow the departmental protocol or absence of adequate echocardiographic windows. Of the 1,547 patients, 47% had S-HF, 20% had HFPEF, and 33% were classified as No-HF (Fig. 1). The clinical characteristics of the patients overall and classified depending on their HF phenotype are presented in Table 1. Briefly, their mean 6 SD age was 71 6 11 years, 62% were men, and LVEF was 47 6 16%. Among patients with CHF, those with SHF were more likely to be men, have ischemic heart disease and COPD, were more breathless, and had more severe LV and RV dilatation, more mitral regurgitation (MR) and tricuspid (TR) regurgitation, and worse RV function compared with those with HFPEF or No-HF. Patients with HFPEF were older, heavier, and more likely to have AF. Clinical, Biologic, and LV Echocardiographic Characteristics by Quartile of TAPSE (Table 2)
The median (interquartile range [IQR]) of TAPSE was 18.5 (14.0e22.7) mm. Of the patients with TAPSE #14mm, 72% had S-HF, 15% HFPEF, and 13% No-HF. There were negative correlations between TAPSE and NT-proBNP (R 5 0.44; P 5 .0001), E/A ratio (R 5 0.23; P 5 .0001), and transtricuspid pressure gradient (TPG; R 5 0.22; P 5 .0001; Fig. 2A). There were positive correlations between TAPSE and body mass index (R 5 0.11; P 5 .0001), systolic blood pressure (R 5 0.21; P 5 .0001), distance walked during 6 minutes (R 5 0.14; P 5 .0001), eGFR (R 5 0.19; P 5 .0001), LVEF (R 5 0.41; P 5 .0001; Fig. 2B), and RV fractional area (RVFA; R 5 0.33; P ! .0001; Fig. 2C). Mortality and Prognostic Value of TAPSE and Interaction With NT-proBNP
During a median (IQR) follow-up of 63 (41e75) months, mortality in the overall population was 34%. Mortality was higher in patients with S-HF (38%) and HFPEF (45%) than
218 Journal of Cardiac Failure Vol. 18 No. 3 March 2012 Table 1. Characteristics of the Cohort Stratified According to TAPSE Quartiles Characteristic n Clinical Age, y Male, % BMI, kg/m2 IHD, % Diabetes, % History of hypertension, % COPD Omild class, % NYHA 3e4, % 6-min WT, m Unable/unwilling to do WT, % AF, % HR, beats/min Systolic BP, mm Hg Diastolic BP, mm Hg Treatment, %* ACEi ARB Beta-blockers Aldosterone ant. Digoxin Echo-RV TAPSE (mm) RVEDA, cm2 RVEDAI, cm2/m2 RVFA, % RAV, mL RAVI, mL/m2 TR class, % None Mild or less Moderate or severe TPG TPG, mm Hgy TPG class, % No measurable TR velocity, % #35 mm Hg, % O35 mm Hg, % Echo-LV LVEF, % LVEDVI, mL/m2 LAVI, mL/m2 E, m/s A, m/sz E/Az DT, ms MR class, % None Mild or less Moderate Severe Blood tests NT-proBNP, pg/mL Hemoglobin, g/dL eGFR, mL min1 1.73 m2
All Patients
S-HF
HFPEF
No HF
1,547
722
309
516
71 6 11 62 28 6 5 51 17 49 40 18 308 6 133 34 19 72 6 17 142 6 26 79 6 14
71 6 10 73 28 6 5 65 19 39 45 25 293 6 134 38 22 75 6 19 134 6 25 77 6 15
74 6 10 48 30 6 7 37 19 56 43 22 260 6 136 46 27 71 6 17 148 6 24 79 6 13
69 6 10 54 28 6 5 39 13 58 32 5 351 6 119 20 9 68 6 15 150 6 25 83 6 13
.0001 .0001 .0001 .0001 .001 .0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001
71 6 53 20 18
58 7 40 10 18
35 10 43 1 4
.0001 .057 .0001 .0001 .0001
16.1 6 5.3 17.3 6 6.3 10.2 6 3.6 40 6 12 54 6 39 32 6 23
19.6 6 6.0 15.3 6 5.6 9.2 6 3.4 46 6 11 45 6 30 27 6 18
21.5 6 5.7 15.1 6 6.3 8.9 6 3.5 47 6 11 38 6 23 23 6 13
.0001 .0001 .0001 .0001 .0001 .0001 .0001
64 25 11
74 16 10
81 15 4
28 6 15
26 6 13
22 6 10
69 22 10
67 26 8
79 20 1
33 6 9 79 6 31 32 6 23 80 6 30 70 6 29 1.4 6 1.0 197 6 85
59 6 9 42 6 16 33 6 22 80 6 29 84 6 25 1.0 6 0.5 217 6 81
60 6 9 42 6 14 27 6 14 72 6 23 81 6 21 0.9 6 0.4 226 6 76
28 12 42 18
54 8 33 5
63 9 24 4
1,509 (610e3678) 13.4 6 1.7 58 6 21
466 (186e1271) 13.3 6 1.7 59 6 20
161 (85e356) 13.6 6 1.4 70 6 18
56 8 47 12 13 18.6 6 6.1 15.2 6 6.2 9.6 6 3.4 44 6 12 38 6 34 28 6 20 71 20 9 26 6 14 72 22 6 47 6 16 59 6 30 33 6 19 77 6 28 77 6 26 1.1 6 0.8 211 6 82 45 10 34 11 588 (161e1910) 13.5 6 1.6 63 6 21
P Value
.001 .0001
.0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001
.0001 .0001 .0001
S-HF, systolic heart failure; HFPEF, heart failure with preserved ejection fraction; BMI, body mass index; IHD, ischemic heart disease; COPD, chronic obstructive pulmonary disease; NYHA, New York Heart Association functional class; WT, walking test; AF, atrial fibrillation/atrial flutter; HR, heart rate; BP, blood pressure; ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; ant., antagonist; RV, right ventricle; TAPSE, tricuspid annular plane systolic excursion; RVEDA, right ventricular end-diastolic area; RVEDAI, right ventricular end-diastolic area indexed to body surface area; RVFA, right ventricular fractional area; RAVI, right atrial volume; RAVI, right atrial volume indexed to body surface area; TR, tricuspid regurgitation; TPG, transtricuspid pressure gradient; LV, left ventricle; LVEF, left ventricular ejection fraction; LVEDVI, left ventricular end-diastolic volume indexed to body surface area; LAVI, left atrial volume indexed to body surface area; DT, deceleration time of the E-wave; MR, mitral regurgitation; NT-proBNP, N-terminal proeB-type natriuretic peptide; eGFR, estimated glomerular filtration rate. *Treatment at time of echo before optimization. y Systolic pulmonary artery pressure was measurable in 440 patients. z Only in patients in sinus rhythm.
Prognosis Value of RV Function in CHF All patients 1547
S-HF
_45% LVEF < N=722
HFPEF
No-HF
LVEF >45% With loop diuretic N=309
LVEF >45% Without loop diuretic N=516
Fig. 1. Distribution of patient phenotypes. S-HF, systolic heart failure; HFPEF, heart failure with preserved ejection fraction.
in the No-HF group (16%; Fig. 3). Mortality at 1 year for the population as a whole was 8% and for each of the TAPSE quartiles (from lowest to highest), respectively, 16%, 7%, 6% ,and 3% (P between quartiles: .0001). For S-HF, mortality at 1 year was 13%, and for each of the TAPSE quartiles, respectively, 18%, 11%, 16%, and 3%. For HFPEF, mortality was 6% overall and 9%, 4%, 6%, and 5%, respectively, for quartiles of TAPSE. For No-HF, mortality was 2% overall and 10%, 2%, 1%, and 2%, respectively, for quartiles of TAPSE. Patients who had died by 1 year were older, more symptomatic, had lower systolic and diastolic blood pressure, E-wave deceleration time, TAPSE, and hemoglobin by univariate analysis. They also had higher LV end-diastolic volume indexed to body surface area, left atrial volume indexed to body surface area, NT-proBNP, and E-wave on transmitral Doppler. Independent predictors of death are presented in Table 3. In both models, a lower TAPSE was an independent marker of a worse prognosis. Kaplan-Meier curves with the patient population divided by quartiles of TAPSE are shown for all patients (Fig. 4A), for patients with S-HF (Fig. 4B), HFPEF (Fig. 4C) and No-HF (Fig. 4D). In each group, those patients in the lowest TAPSE quartile (#14 mm) had the worst prognosis. Receiver operating characteristic (ROC) curve analysis of 1-year survival identified a value of 15.9 mm for TAPSE (area under the ROC curve: 0.69; 95% confidence interval 0.64e0.74; P 5 .0001) as the best cutoff point for predicting outcome when all patients were considered together as well as in each of the diagnostic categories. The prevalence of TAPSE #15.9mm was 50% in S-HF, 27% in HFPEF, and 17% in No-HF. Survival for patients on each side of the threshold was very different (c2 for log-rank: 67.3; P 5 .0001; data not shown). NT-proBNP was a strong predictor of mortality. Kaplan-Meier curve for event-free survival stratified according to the combination of TAPSE threshold (15.9 mm) and NT-proBNP (median 588 pg/mL) is shown in Figure 5. Variables Associated With TAPSE
The predictors of TASPE !15.9mm are presented in Table 4. Three patient groupings are presented: the overall population; patients in sinus rhythm; and patients with LVEF O45%.
Damy et al
219
In the overall population and excluding NT-proBNP from the model, the 3 strongest predictors of TAPSE !15.9 mm were the presence of AF, S-HF, and increasing TPG. In patients in sinus rhythm, lower TAPSE was most strongly associated with the presence of S-HF. In patients with LVEF O45%, lower TAPSE was most strongly associated with the presence of AF. Once NT-proBNP was included in the model, it became the dominant factor associated with TAPSE overall and in patients in sinus rhythm, but not in those with LVEF O45%, where AF remained dominant (Table 4). TAPSE was correlated with right atrial volume indexed to body surface area (R 5 0.19; P 5 .0001) and RVFA (R 5 0.33; P ! .0001; Fig. 2C) but not to RV size (RV enddiastolic area indexed to body surface area: P 5 .35; data not shown). The severity of tricuspid regurgitation increased with decreasing TAPSE (data not shown). Patients in AF had a lower TAPSE (14.2 6 5.2 mm) than those in sinus rhythm (19.6 6 5.8 mm; P 5 .0001).
Interaction of TAPSE With Systolic PAP and TR
Mortality was lowest in patients who had both TAPSE $15.9 mm and TPG that was either not measurable or !35 mm Hg (n 5 976). Patients with TPG !35 mm Hg and TAPSE !15.9 mm had an intermediate prognosis (n 5 471). Patients with TPG $35 mm Hg had a poor long-term prognosis whether TAPSE was decreased (n 5 62) or not (n 5 38) but, in the short-term, patients with TPG O35 mm Hg and TAPSE !15.9 mm had the poorest prognosis. TAPSE and TPG were independent predictors of death (Table 3). The interaction between TAPSE, TPG, and severity of TR is shown in Figure 6. In general, the combination of lower TAPSE and either higher TPG or greater severity of TR was associated with worse outcome.
Discussion We observed a high prevalence of RV dysfunction in patients with CHF, whether due to S-HF or to HFPEF. There was high concordance between LV and RV dysfunction, with 72% of patients with LVEF !45% having TAPSE in the lowest quartile. However, prognosis in the lowest quartile of TAPSE was similar regardless of LVEF. This suggests that poor RV function may be an important driver of an adverse prognosis regardless of LV function. Low TAPSE was less common but not absent in patients who were considered not to have HF. This reflects, in part, measurement errors in TAPSE, diagnostic misclassification, and perhaps some patients without heart disease at the extreme end of the normal spectrum. Low TAPSE was associated with an adverse outcome in patients with LVEF O45% whether or not HF was considered to be present, suggesting that low TAPSE in the absence of other features of HF should not be ignored.
220 Journal of Cardiac Failure Vol. 18 No. 3 March 2012 Table 2. Characteristics of the Cohort Stratified According to TAPSE Quartiles TAPSE Quartile, mm Characteristic
#14
O14 to #18.5
O18.5 to #22.7
O22.7
n Clinical Age, y Male, % BMI, kg/m2 IHD, % Diabetes, % History of hypertension, % COPD Omild class, % NYHA 3e4, % 6-min WT, m Unable/unwilling to do WT, % AF, % HR, beats/min Systolic BP, mm Hg Diastolic BP, mm Hg Treatment, %* ACEi ARB Beta-blockers Aldosterone ant. Digoxin Echo-RV RVEDA, cm2 RVEDAI, cm2/m2 RVFA, % RAV, mL RAVI, mL/m2 TR class, % None Mild or less Moderate or severe TPG TPG, mm Hgy TPG class, % No measurement, % #35 mm Hg, % O35 mm Hg, % Echo-LV LVEF, % LVEDVI, mL/m2 LAVI, mL/m2 E, m/s A, m/sz E/Az DT, ms MR class, % Mild or less, % Moderate, % Severe, % HF phenotype S-HF, % HFPEF, % No HF, % Blood tests NT-ProBNP, pg/mL Hemoglobin, g/dL eGFR, (mL min1 1.73 m2)
387
388
389
383
72 6 10 71 28 6 6 61 18 46 43 29 276 6 139 29 41 78 6 20 136 6 26 79 6 15
70 6 10 63 28 6 6 56 20 45 39 19 307 6 130 26 19 72 6 17 138 6 25 78 6 14
71 6 10 57 28 6 5 46 17 51 44 15 315 6 132 24 10 70 6 16 144 6 25 80 6 15
70 6 10 57 29 6 5 40 14 51 35 10 332 6 128 21 4 67 6 15 150 6 24 80 6 4
.006 .0001 .001 .0001 .28 .14 .047 .0001 .0001 .009 .001 .0001 .0001 .13
67 5 50 40 25
59 10 48 29 15
52 9 46 21 6
47 8 45 10 7
.0001 .06 .37 .0001 .0001
17.1 6 6.7 10.1 6 4.0 36 6 11 59 6 41 35 6 24
15.9 6 6.2 9.4 6 3.5 43 6 11 45 6 36 27 6 21
15.3 6 4.9 9.1 6 2.8 46 6 10 40 6 25 24 6 15
16.5 6 6.8 9.7 6 3.9 49 6 10 42 6 26 25 6 15
.003 .011 .0001 .0001 .0001 .0001
58 24 18
68 23 9
78 18 4
82 14 4
29 6 14
26 6 15
23 6 13
22 6 12
56 31 13
73 19 8
78 19 3
79 18 3
38 6 15 68 6 32 40 6 21 85 6 29 64 6 26 1.5 6 1.1 183 6 77
45 6 15 61 6 30 29 6 16 78 6 30 75 6 26 1.1 6 0.8 228 6 77
49 6 16 57 6 31 30 6 15 73 6 27 80 6 25 1.0 6 0.7 219 6 80
56 6 13 49 6 22 29 6 16 73 6 23 82 6 24 1.0 6 0.5 228 6 77
11 43 17
12 36 13
11 35 8
7 23 4
72 15 13
53 20 27
43 21 37
18 25 57
1849 (741e4191) 13.4 6 1.7 57 6 21
777 (232e1953) 13.6 6 1.6 63 6 21
352 (130e1453) 13.5 6 1.6 63 6 19
191 (100e474) 13.6 6 1.6 68 6 20
P Value
.0001 .0001
.0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001
.0001
.0001 .44 .0001
*Treatment at time of echo before optimization. y Systolic pulmonary artery pressure was measurable in 440 patients. z Only in patients in sinus rhythm.
Predictor of Low TAPSE
NT-proBNP, S-HF, and the presence of AF were the strongest associations with reduced TAPSE. Conceptually, the key determinants of impaired TAPSE should be low
preload, RV myocardial dysfunction, and increased RV afterload. We have no information on preload, but NTproBNP levels were generally elevated in patients with low TAPSE, and therefore low preload is unlikely to have played a major role in this population. The strength of the
Prognosis Value of RV Function in CHF
Damy et al
221
HF class (Overall population; N=1547) 1.0
No-HF N=516
Cumulative Survival
0.8
0.6
HFPEF N=309 S-HF N=722
0.4
0.2
for log-rank test : 113.8; p=0.0001
0
N:
0
20
40
1547
1335
1184
60
80
887
204
Time (months)
Fig. 3. Kaplan-Meier survival curves for all the out-patients according to their LV function. Abbreviations as in Figure 1.
Fig. 2. Correlation between tricuspid annular plane systolic excursion (TAPSE) and (A) transtricuspid pressure gradient (TPG), (B) left ventricular ejection fraction (LVEF), and (C) right ventricular fractional area (RVFA). (B) Regression line 1: all patients (n 5 1,514; R 5 0.41; P ! .0001); regression line 2: patients in sinus rhythm (open squares; n 5 1,238; R 5 0.41; P ! .0001); regression line 2: patients in atrial fibrillation (solid squares; n 5 276; R 5 0.31; P ! .0001).
association with AF could reflect common factors, such as pulmonary hypertension, causing RV dysfunction and RA dilation or loss of atrial contraction, which reduces RV filling and could reduce TAPSE.14 RV systolic function may improve after cardioversion of AF.15,16 RV myocardial dysfunction, which reflects RV damage due to CAD,17e19 is also an important potential determinant of low TAPSE. Alam et al. observed an acute decrease in TAPSE during exercise in patients with proximal right
coronary artery stenosis.20 Patients may have a decrease in RV function for at least a year after coronary artery bypass surgery, presumably reflecting inadequate RV myocardial protection.21 The role of RV afterload is more complex. We observed an inverse association between TAPSE and TPG, as previously noted by others, indicating that pulmonary hypertension is an important contributor.22e24 More severe MR25,26 and higher transmitral E/A ratio in patients in sinus rhythm and low TAPSE may reflect left atrial hypertension causing increases in RV afterload. RV dysfunction is also associated with more severe TR, which is somewhat paradoxic because TR should reduce RV afterload and attenuate the reduction in TAPSE. It is possible that TR conceals the severity of RV dysfunction as described by TAPSE. This is consistent with our finding that for a given value of TAPSE, patients with moderate or severe TR had a worse prognosis. Predictors of Mortality in Patients With CHF
Despite high levels of implementation of guidelineindicated therapy,27 mortality was high and similar in patients with either S-HF or HFPEF, as reported by others (Fig. 3).1,28 Age and NT-proBNP were the strongest predictors of an adverse prognosis. The association between age and TAPSE was weak, suggesting that RV dysfunction is not an important link between age and an adverse prognosis. On the other hand, the association between NTproBNP and TAPSE was strong, suggesting that some of the adverse prognostic signal of NT-proBNP is mediated through RV dysfunction. However, measures of RV dysfunction, unlike measures of LV dysfunction, retained independent prognostic significance even after the inclusion
222 Journal of Cardiac Failure Vol. 18 No. 3 March 2012 Table 3. Independent Prognostic Variables on Multivariable Analysis in the Overall Population Wald c2
P Value
Category
Hazard Ratio
95% CI
1.88 1.96 0.99 1.01 1.54 1.01 0.66 1.21 0.97 1.01 1.34 1.26 1.41 0.99 0.66
1.66e2.13 1.56e2.46 0.98e0.99 1.00e1.01 1.26e1.88 1.01e1.02 0.47e0.93 0.90e1.61 0.95e0.99 1.00e1.01 1.08e1.65 1.08e1.48 1.04e1.89 0.70e1.39 0.50e1.03
2.01 1.68 1.80 1.48 1.22 0.78 0.99 0.91 0.74 1.00 0.98 1.37 1.33
1.64e2.61 1.46e1.93 1.40e2.24 1.20e1.83 0.85e1.84 0.53e1.14 0.98e0.99 0.86e0.98 0.60e0.92 1.00e1.01 0.96e0.99 1.02e1.83 1.02e1.73
1. TAPSE and clinical and treatment variables and HF class and echo-LV and PAP variables* Age, class 97.3 .0001 Per 10 y NYHA, class 31.1 .0001 3e4 vs 1e2 Diastolic BP, mm Hg 22.1 .0001 Per unit 19.7 .0001 e LVEDVI, mL/m2 COPD, class 16.6 .0001 Omild vs mild or less 2 15.2 .001 Per unit RAVI, mL/m HF, phenotype 12.0 .002 No-HF vs S-HF HFPEF vs S-HF TAPSE, mm 11.6 .001 Per unit HR, beats/min 8.6 .003 e History of hypertension 8.8 .003 Yes vs no AF, class 8.2 .004 e Digoxin, class 6.7 .025 Per unit TPG, class 6.2 .04 Not measured vs O35 mm Hg #35 mm Hg vs O35 mm Hg 2. Same variables as model 1 including NT-proBNP, hemoglobin, and eGFR* Age , class 54.0 .0001 Per 10 y NT-proBNP, log 36.8 .0001 Per unit NYHA, class 22.0 .0001 3e4 vs 1e2 COPD, class 13.9 .0001 Omild vs mild or less TPG, class, 12.6 .001 Not measured vs O35 mm Hg #35 mm Hg vs O35 mm Hg Diastolic BP, mm Hg 12.4 .0001 Per unit 8.9 .006 e Hemoglobin, g/dL1 Beta-blocker, class 7.7 .010 Yes vs no 2 5.8 .016 Per Unit RAVI, mL/m TAPSE, mm 5.3 .022 e Digoxin, class 4.8 .029 Yes vs no AF, class 5.0 .021 e
Abbreviations as in Table 1. *All of the clinical, treatment, HF class and echo-LV-RV and PAP variables of Table 2 with P ! .10 in univariate analysis were included in model 1, except LVEF, because we used HF class which allowed the comparison between patients with S-HF, No-HF, and HFPEF; E/A and A, to not exclude patients in AF. RVEDAI measurement; and 6-minute walk distances, because they were obtained in fewer patients. In univariate analysis, the variables with P ! .10 were all of the variables presented in model 1 and ischemic heart disease.
of NT-proBNP, which was consistent with earlier reports.29 Different prognostic thresholds for TAPSE have been proposed, based mainly on studies of patients with
TAPSE
0.8
Q4 (N=383)
0.6
Q3 (N=389) Q2 (N=388)
0.4
Q1 (N=387)
Overall population (N=1547)
0.2
for log-rank test : 85.9 p=0.0001
0 0 N: 1547
20 1335
40 1184
60 887
80 204
TAPSE Q4 (N=95) Q3 (N=81) Q1 (N=57) Q2 (N=76)
0.6 0.4
HFPEF (N=309)
0.2
for log-rank test : 5.8 p=0.12
0 0 N: 309
20 268
40 223
60 171
80 39
TAPSE
0.8
Q4 (N=70)
0.6 0.4
Time (Months)
Q2 (N=166) Q3 (N=206) Q1 (N=280)
S-HF (N=722)
0.2
for log-rank test : 23.0 p=0.0001
0 0 N: 722
D
0.8
1.0
Time (Months)
C 1.0 Cumulative Survival
Cumulative Survival
B
1.0
Cumulative Survival
Cumulative Survival
A
S-HF.5,6 Consistent with earlier reports, 25% of our patients had TAPSE !14 mm,5 but we found that 15.9 mm was a better prognostic threshold value in the overall population and when restricted to patients with S-HF.
20 576
40 491
60 351
80 81
Time (Months)
TAPSE Q4 (N=218) Q3 (N=142) Q2 (N=106)
1.0 0.8
Q1 (N=50)
0.6 0.4
No heart failure 0.2
(N=516)
for log-rank test : 11.7, p=0.008
0 0 N: 516
20 491
40 470
60 365
80 84
Time (Months)
Fig. 4. Kaplan-Meier survival curves according to quartiles of TAPSE for all patients (A), for patients with S-HF (B), for patients with HFPEF (C) and no heart failure patients (D). Abbreviations as in Figures 1 and 2.
Prognosis Value of RV Function in CHF
Damy et al
223
HF class (Overall population; N=1547) 1.0
0.6
TAPSE ≥ 15.9mm, NT-proBNP<588pg.ml-1 N=635 TAPSE<15.9mm, NT-proBNP<588pg.ml-1 N=138
0.4
TAPSE ≥ 15.9mm, NT-proBNP ≥588pg.ml-1 N=379
Cumulative Survival
0.8
TAPSE<15.9mm, NT-proBNP ≥ 588pg.ml-1 N=395
0.2
χ² for log-rank test : 237.9; p=0.0001
0 0
20
40
60
80
Time (months)
Fig. 5. Kaplan-Meier survival curves for all patients according to TAPSE (divided by its threshold: 15.9 mm) and N-terminal proeB-type natriuretic peptide (NT-proBNP; divided by its median: 588 pg/mL). Abbreviations as in Figures 1 and 2. Table 4. Determinants of TAPSE !15.9 mm by Multinomial Logistic Regression Analysis Without NT-proBNP Category 1. In all patients AF, class HF, phenotype TPG, class IHD, class MR, class HR, beats/min Diastolic BP, beats/min eGFR, mL min1 1.73 m2 BMI, kg/m2 Systolic BP, beats/min NT-proBNP, log COPD, class LVEVDI, mL/m2
Yes vs no No-HF vs S-HF HFPEF vs S-HF Not measured vs O35 mm Hg #35 mm Hg vs O35 mm Hg Yes vs no $moderate vs none $moderate vs #mild Per unit e e e e Per unit Yes vs no Per unit
2. In patients with sinus rhythm HF, phenotype TPG, class COPD, class Diabetes, class BMI, kg/m2 Diastolic BP, mm Hg Hemoglobin, g/dL eGFR, mL min1 1.73 m2 E/A, ratio LVEDVI, mL/m2 NTproBNP, log Aldosterone antagonist Systolic BP, mm Hg
No-HF vs S-HF HFPEF vs S-HF Not measured vs O35 mm Hg #35 mm Hg vs O35 mm Hg Yes vs no e Per unit e e e Per unit e e Yes vs no Per unit
3. In patients with LVEF O45% AF, class Aldosterone ant. IHD, class NT-proBNP, log Hemoglobin, g/dL Abbreviations as in Table 1.
Yes vs no e e Per unit e
With NT-proBNP
Wald c P Value Hazard Ratio 95% CI Wald c P Value Hazard Ratio 95% CI 2
2
50.1 35.8 e 10.4 e 8.6 6.7 e 6.6 5.2 5.0 5.0 4.3 e e e
0.0001 0.0001 e 0.006 e 0.003 0.035 e 0.010 0.022 0.025 0.025 0.039 e e e
3.92 0.38 0.39 0.44 0.64 1.57 0.66 0.78 1.01 1.02 0.99 0.97 0.99 e e e
2.68e5.72 0.26e0.55 0.26e0.59 0.25e0.78 0.35e1.15 1.16e2.12 0.48e0.91 0.50e1.22 1.00e1.02 1.00e1.03 0.99e1.00 0.94e1.00 0.99e1.00 e e e
24.9 21.1 e e e 7.6 e e e e e e 3.9 32.6 6.3 4.9
0.0001 0.0001 e e e 0.006 e e e e e e 0.048 0.0001 0.012 0.03
2.76 0.43 0.41 e e 1.53 e e e e e e 0.99 2.46 0.68 0.99
1.85e4.12 0.28e0.66 0.26e0.64 e e 1.13e2.08 e e e e e e 0.99e1.00 1.81e3.36 0.51e0.92 0.99e1.00
38.1 e 6.5 e 6.3 5.4 4.8 4.7 4.6 4.2 4.4 e e e e
0.0001 e 0.039 e 0.012 0.021 0.028 0.030 0.033 0.040 0.036 e e e e
0.32 0.31 0.39 0.45 0.64 1.66 0.96 1.02 1.13 0.99 1.23 e e e e
0.21e0.49 0.19e0.52 0.19e0.81 0.21e1.00 0.45e0.90 1.08e2.56 0.93e1.00 1.00e1.04 1.01e1.27 0.98e1.00 1.01e1.49 e e e e
23.5 e e e 9.5 e e e 7.5 e e 9.9 45.7 4.7 3.9
0.0001 e e e 0.002 e e e 0.006 e e 0.002 0.0001 0.03 0.049
0.36 0.32 e e 0.57 e e e 1.17 e e 0.99 3.12 1.71 0.99
0.23e0.59 0.19e0.56 e e 0.40e0.82 e e e 1.05e1.31 e e 0.98e1.00 2.24e4.33 1.05e2.78 0.98e1.00
65.0 7.0 5.3 e e
0.0001 0.008 0.02 e e
7.85 3.26 1.67 e e
4.75e12.95 1.36e7.82 1.08e2.60 e e
18.6 6.5 4.0 8.8 5.1
0.0001 0.011 0.046 0.003 0.024
4.03 3.18 4.03 2.1 1.20
2.14e7.57 1.30e7.74 1.01e2.52 1.28e3.39 1.02e1.40
224 Journal of Cardiac Failure Vol. 18 No. 3 March 2012
A
1.0
Cumulative Survival
0.8 TAPSE ≥15.9 and without TPG or <35mmHg N=976 0.6
TAPSE <15.9 and without TPG or <35mmHg N=471 TAPSE ≥15.9 and TPG ≥35mmHg N=38
0.4
TAPSE <15.9 and TPG ≥35mmHg N=62
TAPSE and sPAP (N=1547)
0.2
χ² for log-rank test : 104.8 p=0.0001
0
Time (months) 0
B
20
40
60
80
1.0
Cumulative Survival
0.8 TAPSE ≥15.9 and without TR or ≤mild N=968 TAPSE ≥15.9 and with TR ≥ moderate N=446
0.6
TAPSE <15.9 and without TR or ≤mild N=46
0.4
TAPSE <15.9 and TR ≥ moderate N=87
TAPSE and TR (N=1547)
0.2
χ² for log-rank test : 84.0 p=0.0001
0
Time (months) 0
20
40
60
80
Fig. 6. Kaplan-Meier survival curves for all patients according to TAPSE (divided by its threshold: 15.9 mm) and (A) TPG above or below 35 mm Hg and (B) presence or absence of moderate or severe tricuspid regurgitation (TR). Abbreviations as in Figures 1 and 2.
TPG determined by echocardiography was also an independent predictor of prognosis. Prognosis was worst in those with TPG O35 mm Hg and best in those with normal TPG and TAPSE. In a selected group of patients, Ghio et al. found that most patients (215/377) with severe HF had both high PAP and RV dysfunction, and those patients had a poor prognosis. Few patients had only RV dysfunction (68) or only raised PAP (21). When neither or only 1 abnormality was present, the prognosis was more favorable. We also found, in a much larger population, that measures of both PH and RV dysfunction alone had a worse outcome although normal RV function with PH was rare (2.5% of patients). Patients with either PH or RV dysfunction alone had a poor outcome in our study, but those with both low TAPSE and raised TPG did worst of all. The high late mortality among patients with high TAPSE but raised TPG may reflect a later decline in RV function. In other words, we may have observed patients at
different stages in the progression of disease with raised PAP preceding a decline in RV function. Study Limitations
Mitral tissue Doppler measurements were not recorded in a daily routine basis in our department in 2001, so it was not possible to assess LV filling pressure. Echocardiographic estimation of right atrial pressure is still a matter of debate and leads to errors in estimating PAP. We take the view that the unadjusted raw data on TPG is the best approach to report estimation of PAP in the current state of knowledge. Conclusion RV dysfunction, as reflected by reduced TAPSE, is common in HF. It may contribute both to the symptoms and signs of
Prognosis Value of RV Function in CHF
HF and to an adverse prognosis whether or not LVEF is reduced. TAPSE is easy to measure in most patients and should be considered as part of the routine echocardiographic examination in patients with suspected HF. Whether RV dysfunction is a specific target for therapy, either by reducing PAP or improving RV myocardial function, is uncertain. Disclosures None. References 1. Owan TE, Hodge DO, Herges RM, Jacobsen SJ, Roger VL, Redfield MM. Trends in prevalence and outcome of heart failure with preserved ejection fraction. N Engl J Med 2006;3:251e9. 2. Cleland JG, Taylor J, Tendera M. Prognosis in heart failure with a normal ejection fraction. N Engl J Med 2007;8:829e30. 3. Kreit JW. The impact of right ventricular dysfunction on the prognosis and therapy of normotensive patients with pulmonary embolism. Chest 2004;4:1539e45. 4. Kjaergaard J, Akkan D, Iversen KK, Kober L, Torp-Pedersen C, Hassager C. Right ventricular dysfunction as an independent predictor of short- and long-term mortality in patients with heart failure. Eur J Heart Fail 2007;6e7:610e6. 5. Ghio S, Recusani F, Klersy C, Sebastiani R, Laudisa ML, Campana C, et al. Prognostic usefulness of the tricuspid annular plane systolic excursion in patients with congestive heart failure secondary to idiopathic or ischemic dilated cardiomyopathy. Am J Cardiol 2000;7:837e42. 6. Damy T, Viallet C, Lairez O, Deswarte G, Paulino A, Maison P, et al. Comparison of four right ventricular systolic echocardiographic parameters to predict adverse outcomes in chronic heart failure. Eur J Heart Fail 2009;9:818e24. 7. Lam CS, Roger VL, Rodeheffer RJ, Borlaug BA, Enders FT, Redfield MM. Pulmonary hypertension in heart failure with preserved ejection fraction: a community-based study. J Am Coll Cardiol 2009; 13:1119e26. 8. Kaul S, Tei C, Hopkins JM, Shah PM. Assessment of right ventricular function using two-dimensional echocardiography. Am Heart J 1984; 3:526e31. 9. Gold Global Initiative for Chronic Obstructive Lung Disease. Global Strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: NHLB/WHO workshop report 2006. Available at: http://www.goldcopd.com/. 10. Alkan T, Sarioglu A, Samanli UB, Sarioglu T, Akcevin A, Turkoglu H, et al. Atrial natriuretic peptide: could it be a marker for postoperative recurrent effusions after Fontan circulation in complex congenital heart defects? Asaio J 2006;5:543e8. 11. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;1:31e41. 12. British Society of Echocardiography. Guidelines and statements. Available at: http://www.bsecho.org/. 13. Damy T, Goode K, Kallvikbacka-Bennet A, Lewinter C, Hobkirk J, Nikitin N, et al. Determinants and prognostic value of pulmonary artery pressure in patients with chronic heart failure. Eur Heart J 2010; 18:2280e90.
Damy et al
225
14. Bazaz R, Edelman K, Gulyasy B, Lopez-Candales A. Evidence of robust coupling of atrioventricular mechanical function of the right side of the heart: insights from M-mode analysis of annular motion. Echocardiography 2008;6:557e61. 15. Alam M, Samad BA, Hedman A, Frick M, Nordlander R. Cardioversion of atrial fibrillation and its effect on right ventricular function as assessed by tricuspid annular motion. Am J Cardiol 1999; 10:1256e8. 16. Khan MN, Ja€ıs P, Cummings J, di Biaise L, Martin DO, Kautzner J, et al. Pulmonary-vein isolation for atrial fibrillation in patients with heart failure. N Engl J Med 2008;359:1778e85. 17. Kjaergaard J, Iversen KK, Akkan D, Moller JE, Kober LV, TorpPedersen C, Hassager C. Predictors of right ventricular function as measured by tricuspid annular plane systolic excursion in heart failure. Cardiovasc Ultrasound 2009;4:51. 18. Zornoff LA, Skali H, Pfeffer MA, St John Sutton M, Rouleau JL, Lamas GA, et al. Right ventricular dysfunction and risk of heart failure and mortality after myocardial infarction. J Am Coll Cardiol 2002; 9:1450e5. 19. Samad BA, Alam M, Jensen-Urstad K. Prognostic impact of right ventricular involvement as assessed by tricuspid annular motion in patients with acute myocardial infarction. Am J Cardiol 2002;7: 778e81. 20. Alam M, Samad BA. Detection of exercise-induced reversible right ventricular wall motion abnormalities using echocardiographic determined tricuspid annular motion. Am J Cardiol 1999;1:103e5. A8. 21. Alam M, Hedman A, Nordlander R, Samad B. Right ventricular function before and after an uncomplicated coronary artery bypass graft as assessed by pulsed wave Doppler tissue imaging of the tricuspid annulus. Am Heart J 2003;3:520e6. 22. Morrison DA, Lancaster L, Henry R, Goldman S. Right ventricular function at rest and during exercise in aortic and mitral valve disease. J Am Coll Cardiol 1985;1:21e8. 23. Morrison D, Goldman S, Wright AL, Henry R, Sorenson S, Caldwell J, Ritchie J. The effect of pulmonary hypertension on systolic function of the right ventricle. Chest 1983;3:250e7. 24. Ghio S, Gavazzi A, Campana C, Inserra C, Klersy C, Sebastiani R, et al. Independent and additive prognostic value of right ventricular systolic function and pulmonary artery pressure in patients with chronic heart failure. J Am Coll Cardiol 2001;1:183e8. 25. Enriquez-Sarano M, Rossi A, Seward JB, Bailey KR, Tajik AJ. Determinants of pulmonary hypertension in left ventricular dysfunction. J Am Coll Cardiol 1997;1:153e9. 26. Dini FL, Conti U, Fontanive P, Andreini D, Banti S, Braccini L, de Tommasi SM. Right ventricular dysfunction is a major predictor of outcome in patients with moderate to severe mitral regurgitation and left ventricular dysfunction. Am Heart J 2007;1:172e9. 27. Shelton R, Rigby A, Cleland JG, Clark A. Effect of a community heart failure clinic on uptake of betablocker by patients with obstructive airways disease and heart failure. Heart 2006;3:331e6. 28. Bursi F, Weston SA, Redfield MM, Jacobsen SJ, Pakhomov S, Nkomo VT, et al. Systolic and diastolic heart failure in the community. Jama 2006;18:2209e16. 29. Bistola V, Parissis JT, Paraskevaidis I, Panou F, Nikolaou M, Ikonomidis I, et al. Prognostic value of tissue Doppler right ventricular systolic and diastolic function indexes combined with plasma B-type natriuretic peptide in patients with advanced heart failure secondary to ischemic or idiopathic dilated cardiomyopathy. Am J Cardio 2010; 2:249e54.
Prevalence of, Associations With, and Prognostic Value of Tricuspid Annular Plane Systolic Excursion (TAPSE) Among Out-Patients Referred for the Evaluation of Heart Failure THIBAUD DAMY, MD, PhD,1,2,3,4 ANNA KALLVIKBACKA-BENNETT,1 KEVIN GOODE, PhD,1 OLGA KHALEVA, MD,1 MD, PhD,2,3,4 CHRISTIAN LEWINTER, MD,1 JAMES HOBKIRK,1 NIKOLAY P. NIKITIN, MD,1 JEAN-LUC DUBOIS-RANDE, LUC HITTINGER, MD, PhD,2,3,4 ANDREW L. CLARK, MA, MD, FRCP,1 AND JOHN G. F. CLELAND, MD, FACC1 Kingston-upon-Hull, United Kingdom; and Creteil, France
ABSTRACT Background: Prevalence, predictors, and prognostic value of right ventricular (RV) function measured by the tricuspid annular plane systolic excursion (TAPSE) in patients with chronic heart failure (CHF) symptoms with a broad range of left ventricular ejection fraction (LVEF) are unknown. Methods and Results: Of 1,547 patients, mean (6SD) age was 71 6 11 years, 48% were women, median (interquartile range [IQR]) TAPSE was 18.5 (14.0e22.7) mm, mean LVEF was 47 6 16%, 47% had LVEF #45% and 67% were diagnosed with CHF, defined as systolic (S-HF) if LVEF was #45% and as heart failure with preserved ejection fraction (HFPEF) if LVEF was O45% and treated with a loop diuretic. During a median (IQR) follow-up of 63 (41e75) months, mortality was 34%. In multivariable analysis, increasing age, N-terminal proeB-type natriuretic peptide (NT-proBNP), New York Heart Association functional class, right atrial volume index, and transtricuspid pressure gradient; lower TAPSE, diastolic blood pressure, and hemoglobin; and atrial fibrillation (AF) or COPD were associated with an adverse prognosis. Receiver operating characteristic curve analysis identified a TAPSE of 15.9 mm as the best prognostic threshold (P 5 .0001); 47% of S-HF and 20% of HFPEF had a TAPSE of !15.9 mm. The main associations with a TAPSE !15.9 mm were higher NT-proBNP, presence of atrial fibrillation and presence of LV systolic dysfunction. Conclusions: In patients with CHF, low values for TAPSE are common, especially in those with reduced LVEF. TAPSE, unlike LVEF, was an independent predictor of outcome. (J Cardiac Fail 2012;18:216e225) Key Words: Heart failure, prognosis, right ventricle, echocardiography, TAPSE.
The pathogenesis of chronic heart failure (CHF) is complex and varied. Most attention has been paid to left ventricular (LV) dysfunction either with a reduced (systolic;
S-HF) or preserved (HFPEF) LV ejection fraction (LVEF). The pathogenesis of HFPEF may be especially complex and include abnormalities of long-axis systolic function, impaired myocardial relaxation, myocardial or pericardial constraint, or pulmonary hypertension resulting in predominant right-sided failure. Surveys suggest that the prognosis of heart failure (HF) is similar in patients with or without reduced LVEF.1 It is possible that comorbid disease drives the poor prognosis of heart failure patients regardless of LVEF.2 Alternatively, similar downstream hemodynamic effects of LV dysfunction on left atrial pressures, pulmonary artery pressures (PAP), and right ventricular (RV) function could cause the RV to fail with a similar adverse effect on prognosis in S-HF and HFPEF. This could be compounded by pathology of the RV myocardium. Several studies suggest that RV systolic function is a powerful prognostic marker in patients with CHF and in other cardiopulmonary disease.3e5 The prognostic value of RV dysfunction in HF due to systolic dysfunction has
From the 1Department of Cardiology, University of Hull, Castle Hill Hospital, Kingston-upon-Hull, United Kingdom; 2Federation de Cardiologie at the Assistance Publique des H^opitaux de Paris, Groupe HenriMondor Albert-Chenevier, Creteil, France; 3Unite U955, Institut National de la Sante Et de la Recherche Medicale, Creteil, France and 4Faculte de Medecine, Universite Paris 12, Creteil, France. Manuscript received May 27, 2011; revised manuscript received October 12, 2011; revised manuscript accepted December 14, 2011. Reprint requests: Dr Thibaud Damy, Department of Cardiology, Hull York Medical School, University of Hull, Castle Hill Hospital, Kingstonupon-Hull, East Riding, Yorkshire, UK. Tel: 0044 (0) 1 482 461 917; Fax: 0044 (0) 1 482 461 779. E-mail: [email protected] Funding: Dr Damy received a grant from the Societe Francaise de Cardiologie/Federation Francaise de Cardiologie. See page 225 for disclosure information. 1071-9164/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved. doi:10.1016/j.cardfail.2011.12.003
216
Prognosis Value of RV Function in CHF
previously been studied.4e6 Lam et al recently showed that patients with HFPEF have raised PAP, which may impair RV function.7 However, the determinants of RV dysfunction are still unclear, and their clarification might help to develop new strategies to improve patient outcome. We describe herein the prevalence, determinants, and prognostic value of reduced of tricuspid annular plane systolic excursion (TAPSE), an echocardiographic measure of RV dysfunction,8 in patients presenting with symptoms suggestive of CHF. Methods Study Population and Procedures Patients were referred for the diagnosis and management of suspected CHF between August 2001 and June 2009 by primary and secondary care physicians to a community-based CHF program in Kingston-upon-Hull, UK. Patients were clinically evaluated and had an electrocardiogram, echocardiogram, and blood tests. We divided the patients into 3 diagnostic groups based on symptoms, echocardiography, and use of loop diuretics. Those with LVEF #45% were considered to have S-HF. Those with LVEF O45% and requiring a loop diuretic for treatment of symptoms or signs of congestion were defined as having HFPEF. The remainder, with LVEF O45% and not treated with a loop diuretic, were considered not to have HF (No-HF). Ischemic heart disease (IHD) was defined as a previous myocardial infarction, coronary bypass surgery, or coronary disease found at coronary angiography. Chronic obstructive pulmonary disease (COPD) was considered to be moderate or severe if both FEV1/FVC was !70% and FEV1 was !80% of predicted.9 Patients who were able had a 6-minute walk test.10 Patients with atrial fibrillation or atrial flutter were grouped as ‘‘AF.’’ Estimated glomerular filtration rate (eGFR) was calculated.11 Medications at the time of referral were recorded. Echocardiographic Measurements Echocardiograms were performed in accordance with the recommendations of the British Society of Echocardiography12 using a Vivid Five (GE Health Care, UK) system operating at 3.4 MHz. Echocardiographic methods have been previously published.5,6,13 Ethics The investigation conforms to the principles outlined in the Declaration of Helsinki. It was approved by the Hull and East Yorkshire Research Ethics Committee. All of the subjects gave their written informed consents. Statistical Analyses To explore the relation between RV function and outcomes, the population was divided into quartiles of TAPSE. Normality of distribution for continuous variables was checked using the Kolmogorov-Smirnov test. Values for other normally distributed variables are given as mean 6 SD for quantitative data and as numbers and percentages for categoric data. N-terminal proeB-type natriuretic peptide (NT-proBNP) was logarithmically transformed before entry into statistical models. Differences between continuous data were tested using the Kruskal-Wallis test for the 4 quartile comparisons and the Mann-Whitney test for the 2 group comparisons. Proportions were compared using Pearson chi-square test.
Damy et al
217
Cox proportional hazards analyses were used to assess prognostic associations . Univariable Cox proportional hazards modeling was used and 2 multivariable models were tested to compare the prognostic value of TAPSE. All the clinical and echocardiographic variables presented in Table 2 that had a P value of !.10 in the univariable model were included in model 1. The same variables as model 1 plus biologic variables (NT-proBNP, hemoglobin, and eGFR) were included in model 2. The prognostic value of TAPSE divided by quartiles was analyzed using Kaplan-Meier plots with the chi-square log-rank test. Predictors of TAPSE were assessed with the use of logistic regression with backward-elimination of independent variables based on likelihood-ratio tests. P values were considered to be significant if P ! .05. Analyses were performed using SPSS 16.0 (SPSS, Chicago, Illinois).
Results Characteristics of the Study Population
Of the 2,085 patients referred, 1,547 had a measurable TAPSE. TAPSE was not measured because of either failure to follow the departmental protocol or absence of adequate echocardiographic windows. Of the 1,547 patients, 47% had S-HF, 20% had HFPEF, and 33% were classified as No-HF (Fig. 1). The clinical characteristics of the patients overall and classified depending on their HF phenotype are presented in Table 1. Briefly, their mean 6 SD age was 71 6 11 years, 62% were men, and LVEF was 47 6 16%. Among patients with CHF, those with SHF were more likely to be men, have ischemic heart disease and COPD, were more breathless, and had more severe LV and RV dilatation, more mitral regurgitation (MR) and tricuspid (TR) regurgitation, and worse RV function compared with those with HFPEF or No-HF. Patients with HFPEF were older, heavier, and more likely to have AF. Clinical, Biologic, and LV Echocardiographic Characteristics by Quartile of TAPSE (Table 2)
The median (interquartile range [IQR]) of TAPSE was 18.5 (14.0e22.7) mm. Of the patients with TAPSE #14mm, 72% had S-HF, 15% HFPEF, and 13% No-HF. There were negative correlations between TAPSE and NT-proBNP (R 5 0.44; P 5 .0001), E/A ratio (R 5 0.23; P 5 .0001), and transtricuspid pressure gradient (TPG; R 5 0.22; P 5 .0001; Fig. 2A). There were positive correlations between TAPSE and body mass index (R 5 0.11; P 5 .0001), systolic blood pressure (R 5 0.21; P 5 .0001), distance walked during 6 minutes (R 5 0.14; P 5 .0001), eGFR (R 5 0.19; P 5 .0001), LVEF (R 5 0.41; P 5 .0001; Fig. 2B), and RV fractional area (RVFA; R 5 0.33; P ! .0001; Fig. 2C). Mortality and Prognostic Value of TAPSE and Interaction With NT-proBNP
During a median (IQR) follow-up of 63 (41e75) months, mortality in the overall population was 34%. Mortality was higher in patients with S-HF (38%) and HFPEF (45%) than
218 Journal of Cardiac Failure Vol. 18 No. 3 March 2012 Table 1. Characteristics of the Cohort Stratified According to TAPSE Quartiles Characteristic n Clinical Age, y Male, % BMI, kg/m2 IHD, % Diabetes, % History of hypertension, % COPD Omild class, % NYHA 3e4, % 6-min WT, m Unable/unwilling to do WT, % AF, % HR, beats/min Systolic BP, mm Hg Diastolic BP, mm Hg Treatment, %* ACEi ARB Beta-blockers Aldosterone ant. Digoxin Echo-RV TAPSE (mm) RVEDA, cm2 RVEDAI, cm2/m2 RVFA, % RAV, mL RAVI, mL/m2 TR class, % None Mild or less Moderate or severe TPG TPG, mm Hgy TPG class, % No measurable TR velocity, % #35 mm Hg, % O35 mm Hg, % Echo-LV LVEF, % LVEDVI, mL/m2 LAVI, mL/m2 E, m/s A, m/sz E/Az DT, ms MR class, % None Mild or less Moderate Severe Blood tests NT-proBNP, pg/mL Hemoglobin, g/dL eGFR, mL min1 1.73 m2
All Patients
S-HF
HFPEF
No HF
1,547
722
309
516
71 6 11 62 28 6 5 51 17 49 40 18 308 6 133 34 19 72 6 17 142 6 26 79 6 14
71 6 10 73 28 6 5 65 19 39 45 25 293 6 134 38 22 75 6 19 134 6 25 77 6 15
74 6 10 48 30 6 7 37 19 56 43 22 260 6 136 46 27 71 6 17 148 6 24 79 6 13
69 6 10 54 28 6 5 39 13 58 32 5 351 6 119 20 9 68 6 15 150 6 25 83 6 13
.0001 .0001 .0001 .0001 .001 .0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001
71 6 53 20 18
58 7 40 10 18
35 10 43 1 4
.0001 .057 .0001 .0001 .0001
16.1 6 5.3 17.3 6 6.3 10.2 6 3.6 40 6 12 54 6 39 32 6 23
19.6 6 6.0 15.3 6 5.6 9.2 6 3.4 46 6 11 45 6 30 27 6 18
21.5 6 5.7 15.1 6 6.3 8.9 6 3.5 47 6 11 38 6 23 23 6 13
.0001 .0001 .0001 .0001 .0001 .0001 .0001
64 25 11
74 16 10
81 15 4
28 6 15
26 6 13
22 6 10
69 22 10
67 26 8
79 20 1
33 6 9 79 6 31 32 6 23 80 6 30 70 6 29 1.4 6 1.0 197 6 85
59 6 9 42 6 16 33 6 22 80 6 29 84 6 25 1.0 6 0.5 217 6 81
60 6 9 42 6 14 27 6 14 72 6 23 81 6 21 0.9 6 0.4 226 6 76
28 12 42 18
54 8 33 5
63 9 24 4
1,509 (610e3678) 13.4 6 1.7 58 6 21
466 (186e1271) 13.3 6 1.7 59 6 20
161 (85e356) 13.6 6 1.4 70 6 18
56 8 47 12 13 18.6 6 6.1 15.2 6 6.2 9.6 6 3.4 44 6 12 38 6 34 28 6 20 71 20 9 26 6 14 72 22 6 47 6 16 59 6 30 33 6 19 77 6 28 77 6 26 1.1 6 0.8 211 6 82 45 10 34 11 588 (161e1910) 13.5 6 1.6 63 6 21
P Value
.001 .0001
.0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001
.0001 .0001 .0001
S-HF, systolic heart failure; HFPEF, heart failure with preserved ejection fraction; BMI, body mass index; IHD, ischemic heart disease; COPD, chronic obstructive pulmonary disease; NYHA, New York Heart Association functional class; WT, walking test; AF, atrial fibrillation/atrial flutter; HR, heart rate; BP, blood pressure; ACEi, angiotensin-converting enzyme inhibitor; ARB, angiotensin II receptor blocker; ant., antagonist; RV, right ventricle; TAPSE, tricuspid annular plane systolic excursion; RVEDA, right ventricular end-diastolic area; RVEDAI, right ventricular end-diastolic area indexed to body surface area; RVFA, right ventricular fractional area; RAVI, right atrial volume; RAVI, right atrial volume indexed to body surface area; TR, tricuspid regurgitation; TPG, transtricuspid pressure gradient; LV, left ventricle; LVEF, left ventricular ejection fraction; LVEDVI, left ventricular end-diastolic volume indexed to body surface area; LAVI, left atrial volume indexed to body surface area; DT, deceleration time of the E-wave; MR, mitral regurgitation; NT-proBNP, N-terminal proeB-type natriuretic peptide; eGFR, estimated glomerular filtration rate. *Treatment at time of echo before optimization. y Systolic pulmonary artery pressure was measurable in 440 patients. z Only in patients in sinus rhythm.
Prognosis Value of RV Function in CHF All patients 1547
S-HF
_45% LVEF < N=722
HFPEF
No-HF
LVEF >45% With loop diuretic N=309
LVEF >45% Without loop diuretic N=516
Fig. 1. Distribution of patient phenotypes. S-HF, systolic heart failure; HFPEF, heart failure with preserved ejection fraction.
in the No-HF group (16%; Fig. 3). Mortality at 1 year for the population as a whole was 8% and for each of the TAPSE quartiles (from lowest to highest), respectively, 16%, 7%, 6% ,and 3% (P between quartiles: .0001). For S-HF, mortality at 1 year was 13%, and for each of the TAPSE quartiles, respectively, 18%, 11%, 16%, and 3%. For HFPEF, mortality was 6% overall and 9%, 4%, 6%, and 5%, respectively, for quartiles of TAPSE. For No-HF, mortality was 2% overall and 10%, 2%, 1%, and 2%, respectively, for quartiles of TAPSE. Patients who had died by 1 year were older, more symptomatic, had lower systolic and diastolic blood pressure, E-wave deceleration time, TAPSE, and hemoglobin by univariate analysis. They also had higher LV end-diastolic volume indexed to body surface area, left atrial volume indexed to body surface area, NT-proBNP, and E-wave on transmitral Doppler. Independent predictors of death are presented in Table 3. In both models, a lower TAPSE was an independent marker of a worse prognosis. Kaplan-Meier curves with the patient population divided by quartiles of TAPSE are shown for all patients (Fig. 4A), for patients with S-HF (Fig. 4B), HFPEF (Fig. 4C) and No-HF (Fig. 4D). In each group, those patients in the lowest TAPSE quartile (#14 mm) had the worst prognosis. Receiver operating characteristic (ROC) curve analysis of 1-year survival identified a value of 15.9 mm for TAPSE (area under the ROC curve: 0.69; 95% confidence interval 0.64e0.74; P 5 .0001) as the best cutoff point for predicting outcome when all patients were considered together as well as in each of the diagnostic categories. The prevalence of TAPSE #15.9mm was 50% in S-HF, 27% in HFPEF, and 17% in No-HF. Survival for patients on each side of the threshold was very different (c2 for log-rank: 67.3; P 5 .0001; data not shown). NT-proBNP was a strong predictor of mortality. Kaplan-Meier curve for event-free survival stratified according to the combination of TAPSE threshold (15.9 mm) and NT-proBNP (median 588 pg/mL) is shown in Figure 5. Variables Associated With TAPSE
The predictors of TASPE !15.9mm are presented in Table 4. Three patient groupings are presented: the overall population; patients in sinus rhythm; and patients with LVEF O45%.
Damy et al
219
In the overall population and excluding NT-proBNP from the model, the 3 strongest predictors of TAPSE !15.9 mm were the presence of AF, S-HF, and increasing TPG. In patients in sinus rhythm, lower TAPSE was most strongly associated with the presence of S-HF. In patients with LVEF O45%, lower TAPSE was most strongly associated with the presence of AF. Once NT-proBNP was included in the model, it became the dominant factor associated with TAPSE overall and in patients in sinus rhythm, but not in those with LVEF O45%, where AF remained dominant (Table 4). TAPSE was correlated with right atrial volume indexed to body surface area (R 5 0.19; P 5 .0001) and RVFA (R 5 0.33; P ! .0001; Fig. 2C) but not to RV size (RV enddiastolic area indexed to body surface area: P 5 .35; data not shown). The severity of tricuspid regurgitation increased with decreasing TAPSE (data not shown). Patients in AF had a lower TAPSE (14.2 6 5.2 mm) than those in sinus rhythm (19.6 6 5.8 mm; P 5 .0001).
Interaction of TAPSE With Systolic PAP and TR
Mortality was lowest in patients who had both TAPSE $15.9 mm and TPG that was either not measurable or !35 mm Hg (n 5 976). Patients with TPG !35 mm Hg and TAPSE !15.9 mm had an intermediate prognosis (n 5 471). Patients with TPG $35 mm Hg had a poor long-term prognosis whether TAPSE was decreased (n 5 62) or not (n 5 38) but, in the short-term, patients with TPG O35 mm Hg and TAPSE !15.9 mm had the poorest prognosis. TAPSE and TPG were independent predictors of death (Table 3). The interaction between TAPSE, TPG, and severity of TR is shown in Figure 6. In general, the combination of lower TAPSE and either higher TPG or greater severity of TR was associated with worse outcome.
Discussion We observed a high prevalence of RV dysfunction in patients with CHF, whether due to S-HF or to HFPEF. There was high concordance between LV and RV dysfunction, with 72% of patients with LVEF !45% having TAPSE in the lowest quartile. However, prognosis in the lowest quartile of TAPSE was similar regardless of LVEF. This suggests that poor RV function may be an important driver of an adverse prognosis regardless of LV function. Low TAPSE was less common but not absent in patients who were considered not to have HF. This reflects, in part, measurement errors in TAPSE, diagnostic misclassification, and perhaps some patients without heart disease at the extreme end of the normal spectrum. Low TAPSE was associated with an adverse outcome in patients with LVEF O45% whether or not HF was considered to be present, suggesting that low TAPSE in the absence of other features of HF should not be ignored.
220 Journal of Cardiac Failure Vol. 18 No. 3 March 2012 Table 2. Characteristics of the Cohort Stratified According to TAPSE Quartiles TAPSE Quartile, mm Characteristic
#14
O14 to #18.5
O18.5 to #22.7
O22.7
n Clinical Age, y Male, % BMI, kg/m2 IHD, % Diabetes, % History of hypertension, % COPD Omild class, % NYHA 3e4, % 6-min WT, m Unable/unwilling to do WT, % AF, % HR, beats/min Systolic BP, mm Hg Diastolic BP, mm Hg Treatment, %* ACEi ARB Beta-blockers Aldosterone ant. Digoxin Echo-RV RVEDA, cm2 RVEDAI, cm2/m2 RVFA, % RAV, mL RAVI, mL/m2 TR class, % None Mild or less Moderate or severe TPG TPG, mm Hgy TPG class, % No measurement, % #35 mm Hg, % O35 mm Hg, % Echo-LV LVEF, % LVEDVI, mL/m2 LAVI, mL/m2 E, m/s A, m/sz E/Az DT, ms MR class, % Mild or less, % Moderate, % Severe, % HF phenotype S-HF, % HFPEF, % No HF, % Blood tests NT-ProBNP, pg/mL Hemoglobin, g/dL eGFR, (mL min1 1.73 m2)
387
388
389
383
72 6 10 71 28 6 6 61 18 46 43 29 276 6 139 29 41 78 6 20 136 6 26 79 6 15
70 6 10 63 28 6 6 56 20 45 39 19 307 6 130 26 19 72 6 17 138 6 25 78 6 14
71 6 10 57 28 6 5 46 17 51 44 15 315 6 132 24 10 70 6 16 144 6 25 80 6 15
70 6 10 57 29 6 5 40 14 51 35 10 332 6 128 21 4 67 6 15 150 6 24 80 6 4
.006 .0001 .001 .0001 .28 .14 .047 .0001 .0001 .009 .001 .0001 .0001 .13
67 5 50 40 25
59 10 48 29 15
52 9 46 21 6
47 8 45 10 7
.0001 .06 .37 .0001 .0001
17.1 6 6.7 10.1 6 4.0 36 6 11 59 6 41 35 6 24
15.9 6 6.2 9.4 6 3.5 43 6 11 45 6 36 27 6 21
15.3 6 4.9 9.1 6 2.8 46 6 10 40 6 25 24 6 15
16.5 6 6.8 9.7 6 3.9 49 6 10 42 6 26 25 6 15
.003 .011 .0001 .0001 .0001 .0001
58 24 18
68 23 9
78 18 4
82 14 4
29 6 14
26 6 15
23 6 13
22 6 12
56 31 13
73 19 8
78 19 3
79 18 3
38 6 15 68 6 32 40 6 21 85 6 29 64 6 26 1.5 6 1.1 183 6 77
45 6 15 61 6 30 29 6 16 78 6 30 75 6 26 1.1 6 0.8 228 6 77
49 6 16 57 6 31 30 6 15 73 6 27 80 6 25 1.0 6 0.7 219 6 80
56 6 13 49 6 22 29 6 16 73 6 23 82 6 24 1.0 6 0.5 228 6 77
11 43 17
12 36 13
11 35 8
7 23 4
72 15 13
53 20 27
43 21 37
18 25 57
1849 (741e4191) 13.4 6 1.7 57 6 21
777 (232e1953) 13.6 6 1.6 63 6 21
352 (130e1453) 13.5 6 1.6 63 6 19
191 (100e474) 13.6 6 1.6 68 6 20
P Value
.0001 .0001
.0001 .0001 .0001 .0001 .0001 .0001 .0001 .0001
.0001
.0001 .44 .0001
*Treatment at time of echo before optimization. y Systolic pulmonary artery pressure was measurable in 440 patients. z Only in patients in sinus rhythm.
Predictor of Low TAPSE
NT-proBNP, S-HF, and the presence of AF were the strongest associations with reduced TAPSE. Conceptually, the key determinants of impaired TAPSE should be low
preload, RV myocardial dysfunction, and increased RV afterload. We have no information on preload, but NTproBNP levels were generally elevated in patients with low TAPSE, and therefore low preload is unlikely to have played a major role in this population. The strength of the
Prognosis Value of RV Function in CHF
Damy et al
221
HF class (Overall population; N=1547) 1.0
No-HF N=516
Cumulative Survival
0.8
0.6
HFPEF N=309 S-HF N=722
0.4
0.2
for log-rank test : 113.8; p=0.0001
0
N:
0
20
40
1547
1335
1184
60
80
887
204
Time (months)
Fig. 3. Kaplan-Meier survival curves for all the out-patients according to their LV function. Abbreviations as in Figure 1.
Fig. 2. Correlation between tricuspid annular plane systolic excursion (TAPSE) and (A) transtricuspid pressure gradient (TPG), (B) left ventricular ejection fraction (LVEF), and (C) right ventricular fractional area (RVFA). (B) Regression line 1: all patients (n 5 1,514; R 5 0.41; P ! .0001); regression line 2: patients in sinus rhythm (open squares; n 5 1,238; R 5 0.41; P ! .0001); regression line 2: patients in atrial fibrillation (solid squares; n 5 276; R 5 0.31; P ! .0001).
association with AF could reflect common factors, such as pulmonary hypertension, causing RV dysfunction and RA dilation or loss of atrial contraction, which reduces RV filling and could reduce TAPSE.14 RV systolic function may improve after cardioversion of AF.15,16 RV myocardial dysfunction, which reflects RV damage due to CAD,17e19 is also an important potential determinant of low TAPSE. Alam et al. observed an acute decrease in TAPSE during exercise in patients with proximal right
coronary artery stenosis.20 Patients may have a decrease in RV function for at least a year after coronary artery bypass surgery, presumably reflecting inadequate RV myocardial protection.21 The role of RV afterload is more complex. We observed an inverse association between TAPSE and TPG, as previously noted by others, indicating that pulmonary hypertension is an important contributor.22e24 More severe MR25,26 and higher transmitral E/A ratio in patients in sinus rhythm and low TAPSE may reflect left atrial hypertension causing increases in RV afterload. RV dysfunction is also associated with more severe TR, which is somewhat paradoxic because TR should reduce RV afterload and attenuate the reduction in TAPSE. It is possible that TR conceals the severity of RV dysfunction as described by TAPSE. This is consistent with our finding that for a given value of TAPSE, patients with moderate or severe TR had a worse prognosis. Predictors of Mortality in Patients With CHF
Despite high levels of implementation of guidelineindicated therapy,27 mortality was high and similar in patients with either S-HF or HFPEF, as reported by others (Fig. 3).1,28 Age and NT-proBNP were the strongest predictors of an adverse prognosis. The association between age and TAPSE was weak, suggesting that RV dysfunction is not an important link between age and an adverse prognosis. On the other hand, the association between NTproBNP and TAPSE was strong, suggesting that some of the adverse prognostic signal of NT-proBNP is mediated through RV dysfunction. However, measures of RV dysfunction, unlike measures of LV dysfunction, retained independent prognostic significance even after the inclusion
222 Journal of Cardiac Failure Vol. 18 No. 3 March 2012 Table 3. Independent Prognostic Variables on Multivariable Analysis in the Overall Population Wald c2
P Value
Category
Hazard Ratio
95% CI
1.88 1.96 0.99 1.01 1.54 1.01 0.66 1.21 0.97 1.01 1.34 1.26 1.41 0.99 0.66
1.66e2.13 1.56e2.46 0.98e0.99 1.00e1.01 1.26e1.88 1.01e1.02 0.47e0.93 0.90e1.61 0.95e0.99 1.00e1.01 1.08e1.65 1.08e1.48 1.04e1.89 0.70e1.39 0.50e1.03
2.01 1.68 1.80 1.48 1.22 0.78 0.99 0.91 0.74 1.00 0.98 1.37 1.33
1.64e2.61 1.46e1.93 1.40e2.24 1.20e1.83 0.85e1.84 0.53e1.14 0.98e0.99 0.86e0.98 0.60e0.92 1.00e1.01 0.96e0.99 1.02e1.83 1.02e1.73
1. TAPSE and clinical and treatment variables and HF class and echo-LV and PAP variables* Age, class 97.3 .0001 Per 10 y NYHA, class 31.1 .0001 3e4 vs 1e2 Diastolic BP, mm Hg 22.1 .0001 Per unit 19.7 .0001 e LVEDVI, mL/m2 COPD, class 16.6 .0001 Omild vs mild or less 2 15.2 .001 Per unit RAVI, mL/m HF, phenotype 12.0 .002 No-HF vs S-HF HFPEF vs S-HF TAPSE, mm 11.6 .001 Per unit HR, beats/min 8.6 .003 e History of hypertension 8.8 .003 Yes vs no AF, class 8.2 .004 e Digoxin, class 6.7 .025 Per unit TPG, class 6.2 .04 Not measured vs O35 mm Hg #35 mm Hg vs O35 mm Hg 2. Same variables as model 1 including NT-proBNP, hemoglobin, and eGFR* Age , class 54.0 .0001 Per 10 y NT-proBNP, log 36.8 .0001 Per unit NYHA, class 22.0 .0001 3e4 vs 1e2 COPD, class 13.9 .0001 Omild vs mild or less TPG, class, 12.6 .001 Not measured vs O35 mm Hg #35 mm Hg vs O35 mm Hg Diastolic BP, mm Hg 12.4 .0001 Per unit 8.9 .006 e Hemoglobin, g/dL1 Beta-blocker, class 7.7 .010 Yes vs no 2 5.8 .016 Per Unit RAVI, mL/m TAPSE, mm 5.3 .022 e Digoxin, class 4.8 .029 Yes vs no AF, class 5.0 .021 e
Abbreviations as in Table 1. *All of the clinical, treatment, HF class and echo-LV-RV and PAP variables of Table 2 with P ! .10 in univariate analysis were included in model 1, except LVEF, because we used HF class which allowed the comparison between patients with S-HF, No-HF, and HFPEF; E/A and A, to not exclude patients in AF. RVEDAI measurement; and 6-minute walk distances, because they were obtained in fewer patients. In univariate analysis, the variables with P ! .10 were all of the variables presented in model 1 and ischemic heart disease.
of NT-proBNP, which was consistent with earlier reports.29 Different prognostic thresholds for TAPSE have been proposed, based mainly on studies of patients with
TAPSE
0.8
Q4 (N=383)
0.6
Q3 (N=389) Q2 (N=388)
0.4
Q1 (N=387)
Overall population (N=1547)
0.2
for log-rank test : 85.9 p=0.0001
0 0 N: 1547
20 1335
40 1184
60 887
80 204
TAPSE Q4 (N=95) Q3 (N=81) Q1 (N=57) Q2 (N=76)
0.6 0.4
HFPEF (N=309)
0.2
for log-rank test : 5.8 p=0.12
0 0 N: 309
20 268
40 223
60 171
80 39
TAPSE
0.8
Q4 (N=70)
0.6 0.4
Time (Months)
Q2 (N=166) Q3 (N=206) Q1 (N=280)
S-HF (N=722)
0.2
for log-rank test : 23.0 p=0.0001
0 0 N: 722
D
0.8
1.0
Time (Months)
C 1.0 Cumulative Survival
Cumulative Survival
B
1.0
Cumulative Survival
Cumulative Survival
A
S-HF.5,6 Consistent with earlier reports, 25% of our patients had TAPSE !14 mm,5 but we found that 15.9 mm was a better prognostic threshold value in the overall population and when restricted to patients with S-HF.
20 576
40 491
60 351
80 81
Time (Months)
TAPSE Q4 (N=218) Q3 (N=142) Q2 (N=106)
1.0 0.8
Q1 (N=50)
0.6 0.4
No heart failure 0.2
(N=516)
for log-rank test : 11.7, p=0.008
0 0 N: 516
20 491
40 470
60 365
80 84
Time (Months)
Fig. 4. Kaplan-Meier survival curves according to quartiles of TAPSE for all patients (A), for patients with S-HF (B), for patients with HFPEF (C) and no heart failure patients (D). Abbreviations as in Figures 1 and 2.
Prognosis Value of RV Function in CHF
Damy et al
223
HF class (Overall population; N=1547) 1.0
0.6
TAPSE ≥ 15.9mm, NT-proBNP<588pg.ml-1 N=635 TAPSE<15.9mm, NT-proBNP<588pg.ml-1 N=138
0.4
TAPSE ≥ 15.9mm, NT-proBNP ≥588pg.ml-1 N=379
Cumulative Survival
0.8
TAPSE<15.9mm, NT-proBNP ≥ 588pg.ml-1 N=395
0.2
χ² for log-rank test : 237.9; p=0.0001
0 0
20
40
60
80
Time (months)
Fig. 5. Kaplan-Meier survival curves for all patients according to TAPSE (divided by its threshold: 15.9 mm) and N-terminal proeB-type natriuretic peptide (NT-proBNP; divided by its median: 588 pg/mL). Abbreviations as in Figures 1 and 2. Table 4. Determinants of TAPSE !15.9 mm by Multinomial Logistic Regression Analysis Without NT-proBNP Category 1. In all patients AF, class HF, phenotype TPG, class IHD, class MR, class HR, beats/min Diastolic BP, beats/min eGFR, mL min1 1.73 m2 BMI, kg/m2 Systolic BP, beats/min NT-proBNP, log COPD, class LVEVDI, mL/m2
Yes vs no No-HF vs S-HF HFPEF vs S-HF Not measured vs O35 mm Hg #35 mm Hg vs O35 mm Hg Yes vs no $moderate vs none $moderate vs #mild Per unit e e e e Per unit Yes vs no Per unit
2. In patients with sinus rhythm HF, phenotype TPG, class COPD, class Diabetes, class BMI, kg/m2 Diastolic BP, mm Hg Hemoglobin, g/dL eGFR, mL min1 1.73 m2 E/A, ratio LVEDVI, mL/m2 NTproBNP, log Aldosterone antagonist Systolic BP, mm Hg
No-HF vs S-HF HFPEF vs S-HF Not measured vs O35 mm Hg #35 mm Hg vs O35 mm Hg Yes vs no e Per unit e e e Per unit e e Yes vs no Per unit
3. In patients with LVEF O45% AF, class Aldosterone ant. IHD, class NT-proBNP, log Hemoglobin, g/dL Abbreviations as in Table 1.
Yes vs no e e Per unit e
With NT-proBNP
Wald c P Value Hazard Ratio 95% CI Wald c P Value Hazard Ratio 95% CI 2
2
50.1 35.8 e 10.4 e 8.6 6.7 e 6.6 5.2 5.0 5.0 4.3 e e e
0.0001 0.0001 e 0.006 e 0.003 0.035 e 0.010 0.022 0.025 0.025 0.039 e e e
3.92 0.38 0.39 0.44 0.64 1.57 0.66 0.78 1.01 1.02 0.99 0.97 0.99 e e e
2.68e5.72 0.26e0.55 0.26e0.59 0.25e0.78 0.35e1.15 1.16e2.12 0.48e0.91 0.50e1.22 1.00e1.02 1.00e1.03 0.99e1.00 0.94e1.00 0.99e1.00 e e e
24.9 21.1 e e e 7.6 e e e e e e 3.9 32.6 6.3 4.9
0.0001 0.0001 e e e 0.006 e e e e e e 0.048 0.0001 0.012 0.03
2.76 0.43 0.41 e e 1.53 e e e e e e 0.99 2.46 0.68 0.99
1.85e4.12 0.28e0.66 0.26e0.64 e e 1.13e2.08 e e e e e e 0.99e1.00 1.81e3.36 0.51e0.92 0.99e1.00
38.1 e 6.5 e 6.3 5.4 4.8 4.7 4.6 4.2 4.4 e e e e
0.0001 e 0.039 e 0.012 0.021 0.028 0.030 0.033 0.040 0.036 e e e e
0.32 0.31 0.39 0.45 0.64 1.66 0.96 1.02 1.13 0.99 1.23 e e e e
0.21e0.49 0.19e0.52 0.19e0.81 0.21e1.00 0.45e0.90 1.08e2.56 0.93e1.00 1.00e1.04 1.01e1.27 0.98e1.00 1.01e1.49 e e e e
23.5 e e e 9.5 e e e 7.5 e e 9.9 45.7 4.7 3.9
0.0001 e e e 0.002 e e e 0.006 e e 0.002 0.0001 0.03 0.049
0.36 0.32 e e 0.57 e e e 1.17 e e 0.99 3.12 1.71 0.99
0.23e0.59 0.19e0.56 e e 0.40e0.82 e e e 1.05e1.31 e e 0.98e1.00 2.24e4.33 1.05e2.78 0.98e1.00
65.0 7.0 5.3 e e
0.0001 0.008 0.02 e e
7.85 3.26 1.67 e e
4.75e12.95 1.36e7.82 1.08e2.60 e e
18.6 6.5 4.0 8.8 5.1
0.0001 0.011 0.046 0.003 0.024
4.03 3.18 4.03 2.1 1.20
2.14e7.57 1.30e7.74 1.01e2.52 1.28e3.39 1.02e1.40
224 Journal of Cardiac Failure Vol. 18 No. 3 March 2012
A
1.0
Cumulative Survival
0.8 TAPSE ≥15.9 and without TPG or <35mmHg N=976 0.6
TAPSE <15.9 and without TPG or <35mmHg N=471 TAPSE ≥15.9 and TPG ≥35mmHg N=38
0.4
TAPSE <15.9 and TPG ≥35mmHg N=62
TAPSE and sPAP (N=1547)
0.2
χ² for log-rank test : 104.8 p=0.0001
0
Time (months) 0
B
20
40
60
80
1.0
Cumulative Survival
0.8 TAPSE ≥15.9 and without TR or ≤mild N=968 TAPSE ≥15.9 and with TR ≥ moderate N=446
0.6
TAPSE <15.9 and without TR or ≤mild N=46
0.4
TAPSE <15.9 and TR ≥ moderate N=87
TAPSE and TR (N=1547)
0.2
χ² for log-rank test : 84.0 p=0.0001
0
Time (months) 0
20
40
60
80
Fig. 6. Kaplan-Meier survival curves for all patients according to TAPSE (divided by its threshold: 15.9 mm) and (A) TPG above or below 35 mm Hg and (B) presence or absence of moderate or severe tricuspid regurgitation (TR). Abbreviations as in Figures 1 and 2.
TPG determined by echocardiography was also an independent predictor of prognosis. Prognosis was worst in those with TPG O35 mm Hg and best in those with normal TPG and TAPSE. In a selected group of patients, Ghio et al. found that most patients (215/377) with severe HF had both high PAP and RV dysfunction, and those patients had a poor prognosis. Few patients had only RV dysfunction (68) or only raised PAP (21). When neither or only 1 abnormality was present, the prognosis was more favorable. We also found, in a much larger population, that measures of both PH and RV dysfunction alone had a worse outcome although normal RV function with PH was rare (2.5% of patients). Patients with either PH or RV dysfunction alone had a poor outcome in our study, but those with both low TAPSE and raised TPG did worst of all. The high late mortality among patients with high TAPSE but raised TPG may reflect a later decline in RV function. In other words, we may have observed patients at
different stages in the progression of disease with raised PAP preceding a decline in RV function. Study Limitations
Mitral tissue Doppler measurements were not recorded in a daily routine basis in our department in 2001, so it was not possible to assess LV filling pressure. Echocardiographic estimation of right atrial pressure is still a matter of debate and leads to errors in estimating PAP. We take the view that the unadjusted raw data on TPG is the best approach to report estimation of PAP in the current state of knowledge. Conclusion RV dysfunction, as reflected by reduced TAPSE, is common in HF. It may contribute both to the symptoms and signs of
Prognosis Value of RV Function in CHF
HF and to an adverse prognosis whether or not LVEF is reduced. TAPSE is easy to measure in most patients and should be considered as part of the routine echocardiographic examination in patients with suspected HF. Whether RV dysfunction is a specific target for therapy, either by reducing PAP or improving RV myocardial function, is uncertain. Disclosures None. References 1. Owan TE, Hodge DO, Herges RM, Jacobsen SJ, Roger VL, Redfield MM. Trends in prevalence and outcome of heart failure with preserved ejection fraction. N Engl J Med 2006;3:251e9. 2. Cleland JG, Taylor J, Tendera M. Prognosis in heart failure with a normal ejection fraction. N Engl J Med 2007;8:829e30. 3. Kreit JW. The impact of right ventricular dysfunction on the prognosis and therapy of normotensive patients with pulmonary embolism. Chest 2004;4:1539e45. 4. Kjaergaard J, Akkan D, Iversen KK, Kober L, Torp-Pedersen C, Hassager C. Right ventricular dysfunction as an independent predictor of short- and long-term mortality in patients with heart failure. Eur J Heart Fail 2007;6e7:610e6. 5. Ghio S, Recusani F, Klersy C, Sebastiani R, Laudisa ML, Campana C, et al. Prognostic usefulness of the tricuspid annular plane systolic excursion in patients with congestive heart failure secondary to idiopathic or ischemic dilated cardiomyopathy. Am J Cardiol 2000;7:837e42. 6. Damy T, Viallet C, Lairez O, Deswarte G, Paulino A, Maison P, et al. Comparison of four right ventricular systolic echocardiographic parameters to predict adverse outcomes in chronic heart failure. Eur J Heart Fail 2009;9:818e24. 7. Lam CS, Roger VL, Rodeheffer RJ, Borlaug BA, Enders FT, Redfield MM. Pulmonary hypertension in heart failure with preserved ejection fraction: a community-based study. J Am Coll Cardiol 2009; 13:1119e26. 8. Kaul S, Tei C, Hopkins JM, Shah PM. Assessment of right ventricular function using two-dimensional echocardiography. Am Heart J 1984; 3:526e31. 9. Gold Global Initiative for Chronic Obstructive Lung Disease. Global Strategy for the diagnosis, management, and prevention of chronic obstructive pulmonary disease: NHLB/WHO workshop report 2006. Available at: http://www.goldcopd.com/. 10. Alkan T, Sarioglu A, Samanli UB, Sarioglu T, Akcevin A, Turkoglu H, et al. Atrial natriuretic peptide: could it be a marker for postoperative recurrent effusions after Fontan circulation in complex congenital heart defects? Asaio J 2006;5:543e8. 11. Cockcroft DW, Gault MH. Prediction of creatinine clearance from serum creatinine. Nephron 1976;1:31e41. 12. British Society of Echocardiography. Guidelines and statements. Available at: http://www.bsecho.org/. 13. Damy T, Goode K, Kallvikbacka-Bennet A, Lewinter C, Hobkirk J, Nikitin N, et al. Determinants and prognostic value of pulmonary artery pressure in patients with chronic heart failure. Eur Heart J 2010; 18:2280e90.
Damy et al
225
14. Bazaz R, Edelman K, Gulyasy B, Lopez-Candales A. Evidence of robust coupling of atrioventricular mechanical function of the right side of the heart: insights from M-mode analysis of annular motion. Echocardiography 2008;6:557e61. 15. Alam M, Samad BA, Hedman A, Frick M, Nordlander R. Cardioversion of atrial fibrillation and its effect on right ventricular function as assessed by tricuspid annular motion. Am J Cardiol 1999; 10:1256e8. 16. Khan MN, Ja€ıs P, Cummings J, di Biaise L, Martin DO, Kautzner J, et al. Pulmonary-vein isolation for atrial fibrillation in patients with heart failure. N Engl J Med 2008;359:1778e85. 17. Kjaergaard J, Iversen KK, Akkan D, Moller JE, Kober LV, TorpPedersen C, Hassager C. Predictors of right ventricular function as measured by tricuspid annular plane systolic excursion in heart failure. Cardiovasc Ultrasound 2009;4:51. 18. Zornoff LA, Skali H, Pfeffer MA, St John Sutton M, Rouleau JL, Lamas GA, et al. Right ventricular dysfunction and risk of heart failure and mortality after myocardial infarction. J Am Coll Cardiol 2002; 9:1450e5. 19. Samad BA, Alam M, Jensen-Urstad K. Prognostic impact of right ventricular involvement as assessed by tricuspid annular motion in patients with acute myocardial infarction. Am J Cardiol 2002;7: 778e81. 20. Alam M, Samad BA. Detection of exercise-induced reversible right ventricular wall motion abnormalities using echocardiographic determined tricuspid annular motion. Am J Cardiol 1999;1:103e5. A8. 21. Alam M, Hedman A, Nordlander R, Samad B. Right ventricular function before and after an uncomplicated coronary artery bypass graft as assessed by pulsed wave Doppler tissue imaging of the tricuspid annulus. Am Heart J 2003;3:520e6. 22. Morrison DA, Lancaster L, Henry R, Goldman S. Right ventricular function at rest and during exercise in aortic and mitral valve disease. J Am Coll Cardiol 1985;1:21e8. 23. Morrison D, Goldman S, Wright AL, Henry R, Sorenson S, Caldwell J, Ritchie J. The effect of pulmonary hypertension on systolic function of the right ventricle. Chest 1983;3:250e7. 24. Ghio S, Gavazzi A, Campana C, Inserra C, Klersy C, Sebastiani R, et al. Independent and additive prognostic value of right ventricular systolic function and pulmonary artery pressure in patients with chronic heart failure. J Am Coll Cardiol 2001;1:183e8. 25. Enriquez-Sarano M, Rossi A, Seward JB, Bailey KR, Tajik AJ. Determinants of pulmonary hypertension in left ventricular dysfunction. J Am Coll Cardiol 1997;1:153e9. 26. Dini FL, Conti U, Fontanive P, Andreini D, Banti S, Braccini L, de Tommasi SM. Right ventricular dysfunction is a major predictor of outcome in patients with moderate to severe mitral regurgitation and left ventricular dysfunction. Am Heart J 2007;1:172e9. 27. Shelton R, Rigby A, Cleland JG, Clark A. Effect of a community heart failure clinic on uptake of betablocker by patients with obstructive airways disease and heart failure. Heart 2006;3:331e6. 28. Bursi F, Weston SA, Redfield MM, Jacobsen SJ, Pakhomov S, Nkomo VT, et al. Systolic and diastolic heart failure in the community. Jama 2006;18:2209e16. 29. Bistola V, Parissis JT, Paraskevaidis I, Panou F, Nikolaou M, Ikonomidis I, et al. Prognostic value of tissue Doppler right ventricular systolic and diastolic function indexes combined with plasma B-type natriuretic peptide in patients with advanced heart failure secondary to ischemic or idiopathic dilated cardiomyopathy. Am J Cardio 2010; 2:249e54.