Journal of Cardiac Failure Vol. 20 No. 12 2014

Relationships Between Biomarkers and Left Ventricular Filling Pressures at Rest and During Exercise in Patients After Myocardial Infarction MADS J. ANDERSEN, MD, PhD,1,2 MADS ERSBØLL, MD, PhD,1 JOHN BRO-JEPPESEN, MD, PhD,1 JACOB E. MØLLER, MD, PhD, DMSc,1 CHRISTIAN HASSAGER, MD, DMSc,1 LARS KØBER, MD, DMSc,1 BARRY A. BORLAUG, MD,2 JENS P. GOETZE, MD, DMSc,3 AND FINN GUSTAFSSON, MD, PhD, DMSc1 Copenhagen and Aarhus, Denmark; and Rochester, Minnesota

ABSTRACT Background: Increased pulmonary capillary wedge pressure (PCWP) is an independent prognostic predictor after myocardial infarction (MI), but PCWP is difficult to assess noninvasively in subjects with preserved ejection fraction (EF). We hypothesized that biomarkers would provide information regarding PCWP at rest and during exercise in subjects with preserved EF after MI. Methods and Results: Seventy-four subjects with EF O45% and recent MI underwent right heart catheterization at rest and during a symptom-limited semisupine cycle exercise test with simultaneous echocardiography. Plasma samples were collected at rest for assessment of midregional proeA-type natriuretic peptide (MR-proANP), N-terminal proeB-type natriuretic peptide (NT-proBNP), galectin-3 (Gal-3), copeptin, and midregional pro-adrenomedullin (MR-proADM). Plasma levels of MR-proANP and PCWP were associated at rest (r 5 0.33; P 5 .002) and peak exercise (r 5 0.35; P 5 .002) as well as with changes in PCWP (r 5 0.26; P 5 .03). Plasma levels of NT-proBNP and PCWP were weakly associated at rest (r 5 0.23; P 5 .03) and peak exercise (r 5 0.28; P 5 .02) but not with changes in PCWP (r 5 0.20; P 5 .09). In a multivariable analysis, plasma levels of MR-proANP remained associated with rest and exercise PCWP (P ! .01), whereas NT-proBNP did not. Plasma levels of Gal-3, copeptin, and MR-proADM were not associated with PCWP at rest or peak exercise. Conclusions: In subjects recovering from an acute MI with preserved EF, plasma levels of natriuretic peptides, particularly MR-proANP, are associated with filling pressures at rest and during exercise. (J Cardiac Fail 2014;20:959e967) Key Words: Acute myocardial infarction, hemodynamics, exercise testing, biomarkers.

Increased filling pressures independently predict outcome after myocardial infarction (MI).1,2 However, invasive hemodynamic testing is expensive and carries a risk of complications. When left ventricular (LV) ejection

fraction (LVEF) is preserved (O45%), noninvasive demonstration of elevated filling pressure is particularly challenging. Recent recommendations suggest using the quotient of peak early mitral inflow velocity (E) and peak early diastolic tissue Doppler velocity in the mitral annulus (e0 ), which have shown modest association with invasive obtained filling pressure in most studies3,4 though not all.5 European guidelines suggest that filling pressures are increased when E/e0 is O15.6 Accordingly studies have demonstrated that echocardiographic indices suggestive of increased LV filling pressure and pulmonary hypertension are associated with worse outcome after MI,7e11 but w25% of subjects with preserved LVEF after MI have E/e0 values in the intermediate range,8e15 where the association with filling pressure is less clear.12,13 Subjects with heart failure and preserved EF (HFpEF)14 and post-MI subjects with preserved LVEF and diastolic

From the 1Department of Cardiology, Heart Center, Rigshospitalet and University of Copenhagen, Copenhagen, Denmark; 2Division of Cardiovascular Diseases, Department of Medicine, Mayo Clinic, Rochester, Minnesota and 3Department of Biochemistry, Rigshospitalet and University of Aarhus, Aarhus, Denmark. Manuscript received May 23, 2014; revised manuscript received September 19, 2014; revised manuscript accepted September 29, 2014. Reprint requests: Mads J. Andersen, MD, PhD, Mayo Clinic College of Medicine, 200 First Street SW, Rochester, MN 55905. Tel: 507 255 0334, Fax: 507 266 0228. E-mail: [email protected] See page 966 for disclosure information. 1071-9164/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.cardfail.2014.09.012

959

960 Journal of Cardiac Failure Vol. 20 No. 12 December 2014 dysfunction13,15 are prone to elevation in pulmonary capillary wedge pressure (PCWP) during exercise in relation to limitations in LV diastolic reserve, but identifying these vulnerable subjects requires invasive assessment. The ability to noninvasively identify subjects with compromised resting or exercise hemodynamics has gained increasing interest owing to development of experimental therapies aiming at reducing LV filling pressures.14,16 Natriuretic peptides (N-terminal proeB-type natriuretic peptide [NT-proBNP] and midregional proeA-type natriuretic peptide [MR-proANP]) are released in response to increases in wall stress,17,18 but little is known about these correlations with filling pressures during exercise. In addition, other candidate biomarkers have been related to filling pressures, including galectin-3, copeptin, and midregional pro-adrenomedullin (MR-proADM), but comparative data with directly measured filling pressures are lacking. We hypothesized that plasma concentrations of these biomarkers would identify subjects with increased filling pressures at rest or during exercise in post-MI subjects with preserved LVEF. Methods Study Design and Patient Population We enrolled 80 post-MI subjects with preserved LVEF (O45%) who all underwent right heart catheterization at rest and during symptom limited semisupine cycle exercise test with simultaneous echocardiography. Inclusion criteria were preserved LVEF and written informed consent. Subjects with permanent atrial fibrillation, known history of cardiomyopathy, more than mild valvular heart disease (more than mild stenosis or regurgitation), obstructive or restrictive pulmonary disease, or inability to perform exercise testing were excluded. The majority (70) of included subjects were post-MI patients with echocardiographic signs of diastolic dysfunction (MI þ DD; E/e0 O8 and left atrial (LA) volume O32 mL/m2, and 10 post-MI subjects had normal diastolic function as judged by echocardiography (MIDD; E/e0 !8 and LA ! 32 mL/m2). Hemodynamic and echocardiographic data for these subjects has been previously published.12,15 Subjects were studied on chronic medications in the fasted state. The subjects were stratified in a binary fashion with the use of peak exercise PCWP O25 mm Hg as cutoff for abnormal filling pressure with exercise.14 The Ethics Committee in Hovedstaden Region approved the study, and written informed consent was obtained from every subject. Invasive Hemodynamic Measurements Right heart catheterization was performed with the use of a standard 7.5-F triple-lumen Swan-Ganz thermistor and balloontipped catheter (Edwards Lifesciences, Irvine, California). The catheter was introduced, guided by ultrasound, into the right internal jugular vein and advanced to the pulmonary artery. PCWP, right atrial pressure (RAP), systolic pulmonary arterial pressure (PAP), diastolic PAP, mean PAP, blood pressure (BP), and cardiac output (CO; thermodilution technique) were measured at rest, at each level of exercise until exhaustion, and after 5 minutes of rest. PCWP at rest and after exercise was measured at endexpiration. During exercise, mean PCWP was used. We

considered resting PCWP O15 mm Hg and/or exercise PCWP O25 mm Hg to be abnormally increased.14 Exercise Protocol Subjects performed a multistage symptom-limited semisupine cycle exercise test with the use of an Echo Cardiac Stress Table (Lode, the Netherlands). Workload started at 0 W and was increased by 25 W every 2 minutes. Subjects were encouraged to maintain a fixed pedaling speed of 60 rpm for the duration of the exercise. They were also encouraged to exercise until exhaustion (Borg O18). Echocardiography All subjects underwent resting echocardiographic examinations obtained according to current guidelines.19,20 During exercise, 2-dimensional tissue Doppler images (TDI) and pulsed- (PW) and continuous-wave Doppler images were acquired in the apical 4-chamber view. All examinations were performed by an experienced echocardiographer using a Philips iE33 (Philips Healthcare, Best, the Netherlands) cardiac ultrasound system. Echocardiographic cine loops were obtained by recording a minimum of 3 consecutive heart cycles. Images were stored digitally for offline analysis with the use of Philips Xcelera analysis software version 3.1 (Philips Healthcare). LV volumes and LVEF were assessed with the use of the Simpson biplane method of discs from the apical 4- and 2-chamber views at rest. LA volume was measured from the apical 4- and 2-chamber views with the use of the area-length method at rest. Volumes were indexed to body surface area (BSA) when appropriate. With the use of PW Doppler, E velocities were measured with the sample volume placed at the tips of mitral leaflets during diastole. With the use of TDI and PW Doppler with the sample volume placed in the septal and lateral mitral annulus, e0 velocities were measured and averaged.4 For Doppler recordings, horizontal sweep was of 75 or 100 mm/ s and 3e5 consecutive beats were used and averaged. All analyses were performed blinded to hemodynamic and biomarker values. Biomarkers Plasma samples were collected at rest from the internal jugular vein after positioning of the Swan-Ganz catheter before exercise. Plasma and serum were collected in EDTA-primed glass tubes, centrifuged for 10 minutes at 3,000 rpm, and stored at 80 C until analysis. Samples underwent #2 freeze/thaw cycles before analysis. NT-proBNP was measured on the Modular E platform (Roche Diagnostics) with lower limit of detection (LOD) at 25 pg/mL and interassay coefficient of variations (CVs) of 12.6% at 29.2 pg/mL and 9.6% at 8.5 pg/mL.21 Plasma concentrations of copeptin were measured on the automated Kryptor Plus platform (Thermo-Fischer, Waltham, Massachusetts). The interassay CVs were 18.3% for 1.4 pmol/L, 6.8% for 9.3 pmol/L, and !3% for concentrations O18 pmol/L.22 The automated Kryptor Plus platform was also used to quantify the plasma levels of MR-proADM (LOD of 0.08 nmol/L and CV !20% for values O0.12 nmol/L) and MR-proANP (LOD of 6.0 pmol/L and CV 10%).22e24 Galectin-3 was measured on a Vidas platform (Biomerieux, Ballerup, Denmark) with an LOD of 1.13 ng/mL and interassay CV !10.4%.25

Biomarkers and PCWP in Post MI patients With LVEF Statistical Analysis Data are presented as mean 6 SD for gaussian-distributed or median (interquartile range [IQR]) for nongaussian-distributed variables unless otherwise indicated. Between-group differences were tested with the use of Student t test, c2, or nonparametric rank sum test as appropriate. Multivariable analysis was performed in a general linear model and included resting values of LA volume indexed to BSA, E/e0 , age, and LVEF as covariates. All biomarkers were log transformed, and PCWP O25 mm Hg at peak exercise and O15 mm Hg at rest were used as binary cutoffs to create a logistic regression models. Predictive capability was assessed by comparing C-statistics derived from the area under the receiver operating characteristic (ROC) curves with the use of the method proposed by deLong et al.26 The C-statistics were then compared with the use of a paired t test. All tests were 2 sided, a P value of !.05 was considered to be significant, and the explained variation of the general linear model was derived from the global R2 value. Statistical analyses were performed with the use of R version 3.0.1 (R Development Core Team 2013, http:// www.R-project.org; libraries: Hmisc, psych, pROC).

Results Of 80 subjects enrolled, 6 subjects (all MI þ DD) were excluded due to missing blood samples. Thus the total study population consisted of 74 subjects (mean age 62 6 8 years, 86.5% male) with a recent myocardial infarction (31 days [IQR 23e43 days]) before right heart catheterization. Baseline characteristics and demographics are presented in Table 1. Compared with subjects with exercise PCWP O25 mm Hg, subjects with exercise PCWP #25 mm Hg had lower use of beta-blockers, lower levels of natriuretic peptides, and better diastolic function but similar LVEF and plasma levels of copeptin, MR-proADM, and Gal-3 compared with subjects with exercise PCWP O25 mm Hg. The use of beta-blockers was not associated with increased filling pressure in a logistic regression analysis. Hemodynamic response to exercise is presented in Table 2. All subjects in both groups exercised to exhaustion (O18 Borg scale) and all significantly increased lactate levels from rest to peak exercise. During exercise, no subjects complained of chest pain, no significant ischemia was noted on the electrocardiogram, nor was any regional wall motion abnormality observed on the simultaneous echocardiogram. The hemodynamic response to exercise and work load achieved were similar in subjects with abnormal and normal exercise PCWP except for rightsided and pulmonary arterial pressures (Table 2). MR-proANP

The median MR-proANP plasma levels were significantly higher in MI þ DD compared to MIDD (136 [IQR 93e188] vs 89 [IQR 65e120]; P 5 .01). Even larger differences were seen when comparing subjects with normal and elevated PCWP at peak exercise (Table 1), where the MR-proANP plasma levels in subjects with normal PCWP were 68% (95% confidence interval [CI] 53e87%) of the MR-



Andersen et al

961

proANP plasma levels of subjects with increased PCWP at peak exercise. There was a moderate but significant association between MR-proANP and PCWP at rest (r 5 0.33; P 5 .002) and peak exercise (r 5 0.35; P 5 .002) as well as with the changes in PCWP from baseline to peak exercise (r 5 0.23; P 5 .03; Fig. 1). Similar association was found between MR-proANP and E/e0 (r 5 0.37; P 5 .0008) and between MR-proANP and LA volume index (LAVI; r 5 0.41; P 5 .0002). There was a significant association between MR-proANP and PCWP in the multivariable analysis at rest (r 5 0.60; P 5 .003 [entire model]) and peak exercise (r 5 0.39, P 5 .04 [entire model]), but not between changes in MRproANP and PCWP. Interestingly, MR-proANP was the only significant variable at peak exercise, superior to echocardiographic variables (LVEF: P 5 .71; E/e0 : P 5 .07; LAVI: P 5 .27) and age (P 5 .28). An ROC analysis of the ability of MR-proANP to predict elevated PCWP was evaluated by ROC curve analysis and revealed an area under the ROC curve (AUC) of 0.78 (95% CI 0.65e0.90) at rest and 0.73 (95% CI 0.60e0.87) at peak exercise (Fig. 2). Resting MR-proANP levels O140 pmol/L predicted high peak exercise PCWP with 43% sensitivity and 88% specificity, and levels !90 pmol/L predicted peak exercise PCWP #25 mm Hg with 78% sensitivity and 44% specificity. NT-proBNP

The median NT-proBNP concentration was significantly higher in MI þ DD compared with MIDD (59 [IQR 29e99] vs 27 [IQR 12e32]; P 5 .02). However, opposed to MR-proANP, plasma levels of NT-proBNP did not differ between subjects with peak PCWP #25 mm Hg and PCWP O25 mm Hg (Table 1). The NT-proBNP plasma levels in subjects with normal filling pressure were on average 63% (95% CI 35%e112%) of the plasma levels in subjects with elevated filling pressure at peak exercise. Despite no significant differences between groups, a weak association between NT-proBNP and PCWP was found at rest (r 5 0.26; P 5 .03) and peak exercise (r 5 0.28; P 5 .02), but not between changes in PCWP from baseline to peak exercise (r 5 0.20; P 5 .09; Fig. 3). In a multivariable analysis, NT-proBNP was no longer associated with PCWP at rest (P 5 .39) or peak exercise (P 5 .29). In an ROC analysis of the ability of NT-proBNP to predict elevated PCWP we found an AUC of 0.70 (95% CI 0.55e0.86) at rest and 0.64 (95% CI 0.47e0.82) at peak exercise (Fig. 4). Despite poorer AUC, there were no significant differences between MR-proANP and NT-proBNP as predictors of elevated PCWP at rest (0.78 vs 0.70; P 5 .24) or peak exercise (0.73 vs 0.64; P 5 .40). Copeptin

The median copeptin concentration was significantly higher in subjects with MI þ DD compared with MIDD (6.4 [IQR 4.4e15.5] vs 3.8 [IQR 3.1e5.3] pmol/L;

962 Journal of Cardiac Failure Vol. 20 No. 12 December 2014 Table 1. Baseline Demographic, Clinical Characteristics, and Cardiovascular Parameters for the Total Study Population and Stratified According to Peak Exercise Pulmonary Capillary Wedge Pressure (PCWP) Total n 5 74 Age, y Male, (%) BSA (m2) Comorbidities MI þ DD, n (%) Hypertension, n (%) NYHA OII, n (%) Diabetes, n (%) eGFR, (mL min1 1.73 m2) Ex- or current smokers, n (%) Medication Diuretics, n (%) Beta-blockers, n (%) ACEI/ARB, n (%) Statin, n (%) Biomarkers NT-proBNP, (pmol/L) MR-proANP, (pmol/L) Galectin-3, (pmol/L) MR-proADM, (pmol/L) Copeptin, (pmol/L) Echocardiography LVEF, (%) LAVOLi, (mL/m2) E/e0 Diastolic dysfunction, n (%)

PCWP #25 mm Hg (n 5 17)

PCWP O25 mm Hg (n 5 57)

P Value

62 6 8 64 (86%) 2.03 6 0.22

61 6 10 16 (94%) 2.08 6 0.2

62 6 8 48 (84%) 2.02 6 0.22

.48 .29 .32

64 (86%) 31 (42%) 0 (0%) 6 (8%) 88 6 24 46 (62%)

11 (65%) 6 (35%) 0 (0%) 0 87 6 28 11 (65%)

53 (93%) 25 (34%) 0 (0%) 6 (11%) 88 6 23 35 (61%)

.008 .75 .99 .38 .87 .95

4 64 20 73

1 12 2 16

3 52 18 57

(5%) (91%) (32%) (100%)

.92 .03 .11 .07

(29.1e91.8) (94.9e192.0) (8.7e12.7) (0.49e0.62) (4.5e10.4)

.09 .003 .09 .92 .13

51.9 131.9 11.2 0.55 5.6

(5%) (86%) (27%) (99%) (27.9e91.8) (89.4e165.2) (9.6e12.9) (0.48e0.62) (3.8e10.2)

29.4 94.8 11.8 0.56 4.7

56 6 6 41 6 12 10.4 6 2.8 64 (86%)

(6%) (71%) (12%) (94%) (15.4e71.9) (70.0e127.1) (11.2e14.2) (0.47e0.62) (3.3e7.1)

57.5 137.8 11.0 0.55 6.2

55 6 6 36 6 10 9.1 6 3.6 11 (65%)

56 6 6 43 6 12 10.8 6 2.5 53 (93%)

.46 .03 .03 .003

BSA, body surface area; MI, myocardial infarction; DD, diastolic dysfunction; NYHA, New York Heart Association functional class; eGFR, estimated glomerular filtration rate; ACEI/ARB, angiotensin converting enzyme inhibitor/aldosterone receptor blocker; NT-proBNP, N-terminal proeB-type natriuretic peptide; MR-proANP, midregional proeA-type natriuretic peptide; MR-proADM, midregional pro-adrenomedullin; LVEF, left ventricular ejection fraction; LAVOLi, left atrial volume indexed to body surface area; E/e0 , ratio of peak early mitral inflow velocity to peak early diastolic tissue velocity. P value calculated by unpaired t test or nonparametric test as appropriate.

P 5 .03). This seems to be unrelated to filling pressure, because plasma levels of copeptin did not significantly differ between subjects with peak PCWP #25 mm Hg and peak PCWP O25 mm Hg (Table 1). The copeptin

plasma levels in subjects with normal filling pressure were on average 73% (95% CI 49%e109%) of the plasma levels in subjects with elevated filling pressure at peak exercise. There was no association between copeptin and

Table 2. Hemodynamic Characteristics at Rest and at Peak Exercise PCPW #25 mm Hg (n 5 17) Resting CO (L/min) RAP (mm Hg) PCWP (mm Hg) mPAP (mm Hg) mBP (mm Hg) HR (beats/min) LVEF (%) E/e0 Lactate (mmol/L) Peak exercise CO (L/min) RAP (mm Hg) PCWP (mm Hg) mPAP (mm Hg) mBP (mm Hg) HR (beats/min) LVEF (%) E/e0 Lactate, (mmol/L) Peak work load (W)

PCWP O25 mm Hg (n 5 57)

P Value

5.2 5 10 16 89 61 55 9.1 0.93

6 6 6 6 6 6 6 6 6

1.4 2 2 2 10 9 6 3.6 0.34

5.3 7 12 19 91 61 56 10.8 0.92

6 6 6 6 6 6 6 6 6

1.2 2 3 4 11 10 6 2.5 0.41

.78 .05 .02 .003 .58 .85 .46 .03 .91

15.7 7 21 33 116 125 60 8.0 8.09 137

6 6 6 6 6 6 6 6 6 6

3.7 4 5 6 14 18 6 3.6 2.44 41

15.4 15 35 48 117 126 60 9.6 8.25 132

6 6 6 6 6 6 6 6 6 6

3.7 5 6 8 16 17 7 2.4 2.80 40

.81 !.0001 !.0001 !.0001 .77 .9 .77 .04 .83 .70

CO, cardiac output; RAP, right atrial pressure; mPAP, mean pulmonary arterial pressure; mBP, mean arterial blood pressure; HR, heart rate; other abbreviations as in Table 1.

Biomarkers and PCWP in Post MI patients With LVEF



Andersen et al

963

Fig. 1. Plot of correlation between log10(MR-proANP [pmol/L]) and pulmonary capillary wedge pressure (PCWP; mm Hg) (A) at rest (r 5 0.33; P 5 .002; Y 5 5.384X) and (B) at peak exercise (r 5 0.35; P 5 .002; Y 5 15.58X), and (C) changes from rest to peak exercise (r 5 0.23; P 5 .03; Y 5 10.20X). Subjects with myocardial infarction and no diastolic dysfunction (MI  DD; blue squares) and subjects with myocardial infarction with diastolic dysfunction (MI þ DD; red circles). Reported P values reflect testing for association between MRproANP plasma levels and PCWP by linear regression. Y is slope of the regression line. (Color version of figure is available online.)

PCWP at rest (P 5 .88) or peak exercise (P 5 .24), or with changes in PCWP (P 5 .22). MR-proADM

The median MR-proADM concentration did not differ between subjects with MI þ DD and MIDD (P 5 .11);

similarly, plasma concentrations did not differ between subjects with peak PCWP #25 mm Hg and peak PCWP O25 mm Hg (Table 1). The MR-proADM plasma levels in subjects with normal filling pressure on average were 102% (95% CI 90%e116%) of the plasma levels in subjects with elevated filling pressure at peak exercise. There was no association between plasma levels of

Fig. 2. Univariate logistic models of the ability to predict elevated filling pressure with receiver operating characteristic (ROC) curves and C-statistics (A) at rest (area under the ROC curve 5 0.78; 95% confidence interval [CI] 0.65e0.90) and (B) at peak exercise (AUC 5 0.73, 95% CI 0.60e0.87) for MR-proANP. The shaded areas represent 95% CI.

964 Journal of Cardiac Failure Vol. 20 No. 12 December 2014

Fig. 3. Plot of correlation between log10(NT-proBNP [pmol/L]) and pulmonary capillary wedge pressure (PCWP; mm Hg) (A) at rest (r 5 0.26; P 5 .03; Y 5 1.997X) and (B) at peak exercise (r 5 0.28; P 5 .02; Y 5 15.58X), and (C) changes from rest to peak exercise (r 5 0.20; P 5 .09; Y 5 4.157X). Subjects with myocardial infarction without diastolic dysfunction (MI  DD; blue squares) and subjects with myocardial infarction with diastolic dysfunction (MI þ DD; red circles). Reported P values reflect testing for association between NTproBNP plasma levels and PCWP by linear regression. (Color version of figure is available online.)

MR-proADM and PCWP at rest (P 5 .25) or peak exercise (P 5 .59), or with changes in PCWP (P 5 .88). Galectin-3

The median Gal-3 concentration did not differ between subjects with MI þ DD and MIDD (P 5 .09); similarly, there was no difference between subjects with peak

PCWP #25 mm Hg and peak PCWP O25 mm Hg (Table 1). The average Gal-3 plasma levels in subjects with normal filling pressure were 119% (95% CI 97% e147%) of the plasma levels in subjects with elevated filling pressure at peak exercise. There was no association between plasma levels of Gal-3 and PCWP at rest (P 5 .62) or peak exercise (P 5 .18), or with changes in PCWP (P 5 .11).

Fig. 4. Univariate logistic models of the ability to predict elevated filling pressure with receiver operating characteristic (ROC) curves and C-statistics at rest (area under the ROC curve 5 0.70; 95% confidence interval [CI] 0.55e0.86) and (B) at peak exercise (AUC 5 0.64, 95% CI 0.47e0.82) for NT-proBNP. The shaded areas represent 95% CI.

Biomarkers and PCWP in Post MI patients With LVEF

Discussion This is, to our knowledge, the 1st study to report associations or lack of association between MR-proANP, NTproBNP, Gal-3, copeptin, or MR-proADM and filling pressure at rest and during exercise in humans. We found that resting plasma levels of natriuretic peptides, particularly MRproANP, were associated with PCWP both at rest and at peak exercise in subjects with preserved LVEF after myocardial infarction, whereas copeptin, Gal-3, and MR-proADM were not predictive of PCWP. The increase in filling pressure is abrupt and abnormal even with minimal effort in post-MI subjects with diastolic dysfunction and in subjects suffering from HFpEF.14,15 The natriuretic peptides are secreted by cardiomyocytes in response to elevated wall tension, which varies directly with PCWP and chamber dimension.17,18 Earlier studies have shown tight correlations between diastolic wall stress and natriuretic peptide levels at rest.18 Because the majority of time is spent under some form of physical exercise, filling pressure would be elevated and accordingly increase wall stress. However, it has remained unclear how well natriuretic peptide levels might reflect PCWP during exercise. We show that natriuretic peptide plasma levels, apart from the association with resting PCWP, are reflective of peak exercise PCWP, that MR-proANP is associated with changes in PCWP, and that MR-proANP is a more robust marker than NT-proBNP. The fact that MR-proANP is a more robust marker of elevated filling pressure than NT-proBNP could be explained by differences in location of secretion of the natriuretic peptides, because MR-proANP is thought to be secreted primarily from the atria and NT-proBNP primarily from the ventricles. The relatively steeper and more rapid exercise-induced increase in atrial pressures in patients with elevated filling pressures could explain the better association between MR-proANP and PCWP. Similarly, the relative increase in mean blood pressure did not differ between groups, thus explaining the lesser association between NT-proBNP because only the LA pressures would differ between groups. Although correlations were significant, the AUC indicated that natriuretic peptide levels may be less helpful in the intermediate range. MR-proANP levels O140 pmol/L were indicative of high PCWP with exercise, but although the specificity was high (88%) the sensitivity was low (! 50%), which limits the ability of MR-proANP as a sole marker for increased filling pressure. The Lack of Association Between Filling Pressures and Gal-3, Copeptin, and MR-proADM

In recent studies, Gal-3 has been related to mortality in patients with acute and chronic heart failure and has been proposed as a novel marker of HFpEF.27 Galectin is thought to play an important role in cardiac fibrosis, and because cardiac fibrosis is an important contributor to the



Andersen et al

965

pathophysiology of diastolic dysfunction, we hypothesized that the marker would identify subjects with elevated filling pressure. However, this hypothesis was rejected by the data. One possible explanation might be that the subjects in the present study had recently suffered from an MI, so that remodeling with development of scar tissue (fibrosis) might not be completed. Furthermore, fibrosis is not the sole contributing factor to LV DD. MR-proADM is another notable biomarker in heart failure. MR-proADM is increased in hypertension, chronic renal disease, and chronic heart failure28 and has been shown to be an independent predictor of all-cause mortality in stable outpatients with stage AeD heart failure (HF).29 In accord with this evidence, we found a higher level of MR-proADM in subjects with MI þ DD. However, when comparing MR-proADM levels with peak exercise PCWP, we did not find any association with resting or peak exercise PCWP. Copeptin, a novel biomarker of arginine vasopressin, has antidiuretic properties and is a potent vasoconstrictor.30 Given the known association between filling pressures and outcome in heart failure and the fact that vasopressin blockade in acute heart failure reduces dyspnea and lowers PCWP,31 we speculated that copeptin might be associated with filling pressures. This hypothesis could not be confirmed in the present study population with early-stage heart failure, although this does not preclude an association in more advanced heart failure. There are several perspectives of the present study. Randomized clinical drug trials in HFpEF have generally not been able to show any beneficial effects of the tested drug on morbidity and mortality in HFpEF.32e36 Although lack of efficacy of the tested interventions may be a likely explanation, the inability to identify and select the optimal candidates for therapy might also contribute. Most therapies in one way or another have targeted the common denominator in HFpEFdelevated filling pressure (at rest or during exercise)dbut most studies, for practical reasons, did not include measurement of these parameters. Therefore, biomarkers that could provide reliable information regarding filling pressure and help in identifying suitable candidates for therapeutic interventions would be of considerable clinical significance. The present study suggests that MR-proANP deserves greater study in this context. Study Limitations

The small sample size increases the risk of a type II error; nevertheless, the present study is to our knowledge the largest to assess the association between biomarkers and exercise hemodynamics in subjects with ischemic heart disease and preserved LVEF. Furthermore, this limitation does not affect the main result of the study that natriuretic peptides are associated with filling pressure and that MR-proANP is a superior marker to NT-proBNP. Acknowledging the skewed inclusion of MI  DD and MI þ DD subjects we sought to reclassify subjects,

966 Journal of Cardiac Failure Vol. 20 No. 12 December 2014 by applying a physiologic approach, according to their filling pressure at peak exercise to have a more even distribution between groups. Despite being selected in 2 groups, all of the data were pooled to increase the strength of the data. Although all subjects were examined using the same protocol, one should always be careful when interpreting the results of pooled data, because of selection bias and the occurrence of spurious correlation. We cannot safely state that the significant association between natriuretic peptides and filling pressure is not partly caused by a spurious correlation. Owing to some variation in time from MI to enrollment we cannot safely state that this did not affect the results. However 32 MI þ DD subjects enrolled in the SIDAMI (Sildenafil and Diastolic Dysfunction After Acute Myocardial Infarction) trial13 as placebo group had the exact same invasive hemodynamic exercise test, including biomarker sampling, performed after 9 weeks of treatment. In those 32 subjects receiving placebo, NT-proBNP and copeptin decreased significantly, Gal-3 increased, and MR-proANP and MR-proADM did not change significantly over 9 weeks (Supplemental Table 1). Furthermore, the decrease was significant although the absolute changes over 63 days were small, so the variation in time from MI to enrollment would not significantly affect the overall results of the present study. Conclusion In subjects recovering from an acute MI with preserved EF, plasma levels of natriuretic peptides, particularly MRproANP, reflect the filling pressures of the LV at rest and with exercise. In contrast, no association between filling pressures (rest or exercise) and copeptin, Gal-3, or MRproADM was found. Further studies are required to determine if MR-proANP might be a useful noninvasive marker to identify subjects with elevated filling pressures, which potentially could have important implications for identification and selection of subjects for novel treatments. Disclosures None.

Supplementary Data Supplementary data related to this article can be found at http://dx.doi.org/10.1016/j.cardfail.2014.09.012.

References 1. Shell W, Peter T, Mickle D, Forrester JS, Swan HJ. Prognostic implications of reduction of left ventricular filling pressure in early transmural acute myocardial infarction. Am Heart J 1981;102(3 Pt 1): 335e40.

2. Planer D, Mehran R, Witzenbichler B, Guagliumi G, Peruga JZ, Brodie BR, et al. Prognostic utility of left ventricular end-diastolic pressure in patients with ST-segment elevation myocardial infarction undergoing primary percutaneous coronary intervention. Am J Cardiol 2011;108:1068e74. 3. Ommen SR, Nishimura RA, Appleton CP, Miller FA, Oh JK, Redfield MM, et al. Clinical utility of Doppler echocardiography and tissue Doppler imaging in the estimation of left ventricular filling pressures: a comparative simultaneous Doppler-catheterization study. Circulation 2000;102:1788e94. 4. Nagueh SF, Middleton KJ, Kopelen HA, Zoghbi WA, Quinones MA. Doppler tissue imaging: a noninvasive technique for evaluation of left ventricular relaxation and estimation of filling pressures. J Am Coll Cardiol 1997;30:1527e33. 5. Maeder MT, Thompson BR, BrunnereLa Rocca H-P, Kaye DM. Hemodynamic basis of exercise limitation in patients with heart failure and normal ejection fraction. J Am Coll Cardiol 2010;56:855e63. 6. Paulus WJ, Tschope C, Sanderson JE, Rusconi C, Flachskampf FA, Rademakers FE, et al. How to diagnose diastolic heart failure: a consensus statement on the diagnosis of heart failure with normal left ventricular ejection fraction by the Heart Failure and Echocardiography Associations of the European Society of Cardiology. Eur Heart J 2007;28:2539e50. 7. Hillis GS, Møller JE, Pellikka PA, Gersh BJ, Wright RS, Ommen SR, et al. Noninvasive estimation of left ventricular filling pressure by E/e0 is a powerful predictor of survival after acute myocardial infarction. J Am Coll Cardiol 2004;43:360e7. 8. Møller JE, Whalley GA, Dini FL, Doughty RN, Gamble GD, Klein AL, et al. Independent prognostic importance of a restrictive left ventricular filling pattern after myocardial infarction: an individual patient meta-analysis: Meta-Analysis Research Group in Echocardiography Acute Myocardial Infarction. Circulation 2008;117:2591e8. 9. Beinart R, Boyko V, Schwammenthal E, Kuperstein R, Sagie A, Hod H, et al. Long-term prognostic significance of left atrial volume in acute myocardial infarction. J Am Coll Cardiol 2004;44:327e34. 10. Whalley GA, Wright SP, Pearl A, Gamble GD, Walsh HJ, Richards M, et al. Prognostic role of echocardiography and brain natriuretic peptide in symptomatic breathless patients in the community. Eur Heart J 2008;29:509e16. 11. Moller JE, Pellikka PA, Hillis GS, Oh JK. Prognostic importance of diastolic function and filling pressure in patients with acute myocardial infarction. Circulation 2006;114:438e44. 12. Andersen MJ, Ersboll M, Gustafsson F, Axelsson A, Hassager C, Kober L, et al. Exercise-induced changes in left ventricular filling pressure after myocardial infarction assessed with simultaneous right heart catheterization and Doppler echocardiography. Int J Cardiol 2013;168:2803e10. 13. Andersen MJ, Ersboll M, Axelsson A, Gustafsson F, Hassager C, Kober L, et al. Sildenafil and diastolic dysfunction after acute myocardial infarction in patients with preserved ejection fraction: the Sildenafil and Diastolic Dysfunction After Acute Myocardial Infarction (SIDAMI) trial. Circulation 2013;127:1200e8. 14. Borlaug BA, Nishimura RA, Sorajja P, Lam CS, Redfield MM. Exercise hemodynamics enhance diagnosis of early heart failure with preserved ejection fraction. Circ Heart Fail 2010;3:588e95. 15. Andersen MJ, Ersboll M, Bro-Jeppesen J, Gustafsson F, Hassager C, Kober L, et al. Exercise hemodynamics in patients with and without diastolic dysfunction and preserved ejection fraction after myocardial infarction. Circ Heart Fail 2012;5:444e51. 16. Penicka M, Bartunek J, Trakalova H, Hrabakova H, Maruskova M, Karasek J, et al. Heart failure with preserved ejection fraction in outpatients with unexplained dyspnea: a pressure-volume loop analysis. J Am Coll Cardiol 2010;55:1701e10. 17. Flynn TG, de Bold ML, de Bold AJ. The amino acid sequence of an atrial peptide with potent diuretic and natriuretic properties. Biochem Biophys Res Commun 1983;117:859e65. 18. Iwanaga Y, Nishi I, Furuichi S, Noguchi T, Sase K, Kihara Y, et al. B-type natriuretic peptide strongly reflects diastolic wall stress in

Biomarkers and PCWP in Post MI patients With LVEF

19.

20.

21.

22.

23.

24.

25.

26.

27.

patients with chronic heart failure: comparison between systolic and diastolic heart failure. J Am Coll Cardiol 2006;47:742e8. Lang RM, Bierig M, Devereux RB, Flachskampf FA, Foster E, Pellikka PA, et al. Recommendations for chamber quantification: a report from the American Society of Echocardiography’s Guidelines and Standards Committee and the Chamber Quantification Writing Group, developed in conjunction with the European Association of Echocardiography, a branch of the European Society of Cardiology. J Am Soc Echocardiog 2005;18:1440e63. Nagueh SF, Appleton CP, Gillebert TC, Marino PN, Oh JK, Smiseth OA, et al. Recommendations for the evaluation of left ventricular diastolic function by echocardiography. J Am Soc Echocardiogr 2009;22:107e33. Sokoll LJ, Baum H, Collinson PO, Gurr E, Haass M, Luthe H, et al. Multicenter analytical performance evaluation of the Elecsys proBNP assay. Clin Chem Lab Med 2004;42:965e72. Terzic D, Johansson-Fallgren A, Ragnarsson O, Goetze J, Hammarsten O. Evaluation of a sensitive copeptin assay for clinical measurement. Open Clin Chem J 2012;5:21e6. Hunter I, Alehagen U, Dahlstrom U, Rehfeld JF, Crimmins DL, Goetze JP. N-terminal pro-atrial natriuretic peptide measurement in plasma suggests covalent modification. Clin Chem 2011;57:1327e30. Morgenthaler NG, Struck J, Alonso C, Bergmann A. Measurement of midregional proadrenomedullin in plasma with an immunoluminometric assay. Clin Chem 2005;51:1823e9. Food and Drug Administration. Galectin-3 assay. Available at: http: //www.accessdata.fda.gov/cdrh_docs/reviews/K093758.pdf 2010. Accessed September 19, 2014. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44:837e45. de Boer RA, Voors AA, Muntendam P, van Gilst WH, van Veldhuisen DJ. Galectin-3: a novel mediator of heart failure development and progression. Eur J Heart Fail 2009;11:811e7.



Andersen et al

967

28. Jougasaki M, Burnett JC Jr. Adrenomedullin: potential in physiology and pathophysiology. Life Sci 2000;66:855e72. 29. Xue Y, Taub P, Iqbal N, Fard A, Clopton P, Maisel A. Mid-region proadrenomedullin adds predictive value to clinical predictors and Framingham risk score for long-term mortality in stable outpatients with heart failure. Eur J Heart Fail 2013;15:1343e9. 30. Wenzel V, Krismer AC, Arntz HR, Sitter H, Stadlbauer KH, Lindner KH, et al. A comparison of vasopressin and epinephrine for out-of-hospital cardiopulmonary resuscitation. N Engl J Med 2004; 350:105e13. 31. Pang PS, Konstam MA, Krasa HB, Swedberg K, Zannad F, Blair JE, et al. Effects of tolvaptan on dyspnoea relief from the EVEREST trials. Eur Heart J 2009;30:2233e40. 32. Redfield MM, Chen HH, Borlaug BA, Semigran MJ, Lee KL, Lewis G, et al. Effect of phosphodiesterase-5 inhibition on exercise capacity and clinical status in heart failure with preserved ejection fraction: a randomized clinical trial. JAMA 2013;309: 1268e77. 33. Yusuf S, Pfeffer MA, Swedberg K, Granger CB, Held P, McMurray JJ, et al. Effects of candesartan in patients with chronic heart failure and preserved left-ventricular ejection fraction: the CHARM-Preserved trial. Lancet 2003;362:777e81. 34. Massie BM, Carson PE, McMurray JJ, Komajda M, McKelvie R, Zile MR, et al. Irbesartan in patients with heart failure and preserved ejection fraction. N Engl J Med 2008;359:2456e67. 35. Cleland JG, Tendera M, Adamus J, Freemantle N, Polonski L, Taylor J, et al. The perindopril in elderly people with chronic heart failure (PEP-CHF) study. Eur Heart J 2006;27:2338e45. 36. Edelmann F, Wachter R, Schmidt AG, Kraigher-Krainer E, Colantonio C, Kamke W, et al. Effect of spironolactone on diastolic function and exercise capacity in patients with heart failure with preserved ejection fraction: the Aldo-DHF randomized controlled trial. JAMA 2013;309:781e91.