- Email: [email protected]
Mismatch between right- and left-sided filling pressures in heart failure patients with preserved ejection fraction
International Journal of Cardiology 257 (2018) 143–149
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Mismatch between right- and left-sided filling pressures in heart failure patients with preserved ejection fraction☆ Yu Horiuchi a,⁎, Shuzou Tanimoto a, Jiro Aoki a, Nozomi Fuse a, Kazuyuki Yahagi a, Keita Koseki a, Taishi Okuno a, Hiroyoshi Nakajima b, Kazuhiro Hara c, Kengo Tanabe a a b c
Division of Cardiology, Mitsui Memorial Hospital, Tokyo, Japan Division of General Medicine, Mitsui Memorial Hospital, Tokyo, Japan Division of Internal Medicine, Mitsui Memorial Hospital, Tokyo, Japan
a r t i c l e
i n f o
Article history: Received 1 February 2017 Received in revised form 26 October 2017 Accepted 2 November 2017
Keywords: Heart failure with preserved ejection fraction Right heart catheterization Pulmonary capillary wedge pressure
a b s t r a c t Background: Mismatch between right- and left-sided filling pressures is poorly understood in heart failure with preserved ejection fraction (HFpEF). Methods and results: We retrospectively analyzed 170 patients with HFpEF (EF ≥ 40%) who underwent right heart catheterization. Low match (right atrial pressure [RAP] b 10 mm Hg and pulmonary capillary wedge pressure [PCWP] b 10 mm Hg) was 76%, high match (RAP ≥ 10 mm Hg and PCWP ≥ 22 mm Hg) was 6.5%, high-R mismatch (RAP ≥ 10 mm Hg and PCWP b 22 mm Hg) was 12%, and high-L mismatch (RAP b 10 mmHg and PCWP ≥ 22 mm Hg) was 5.9%. Elevated PCWP was a significant predictor of the composite endpoint of death or HF hospitalization within 12 months (hazard ratio 5.40, 95% confidence interval 2.17–12.5, p b 0.001). Elevated RAP was not significantly associated with worse outcomes. Pulmonary artery systolic pressure (PASP) and diastolic pressure (PADP) showed strong correlations with PCWP (PASP, r = 0.738, p b 0.001; PADP, r = 0.834, p b 0.001; RAP, r = 0.638, p b 0.001, respectively). Conclusions: Discordance exists between right- and left-sided filling pressures in HFpEF. Physicians may utilize pulmonary artery pressure to evaluate left-sided filling pressure, which is a significant predictor of prognosis. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Decongestion is a crucial component of heart failure (HF) treatment, because many symptoms of HF arise from congestion, which is associated with elevated left-sided filling pressure. Jugular venous pressure (JVP) is one of the most reliable physical signs for elevated left-sided filling pressure [1–3]. The relationship between JVP and left-sided filling pressure is predicated upon concordance between right- and left-sided filling pressure, although dissociation between right- and left-sided pressure has been reported [4,5], in which right- and left-sided pressures are not tightly coupled and decongestion therapy guided by JVP can result in over- or under-treatment. This dissociation, termed right–left (R–L) mismatch, has been primarily investigated in severe grade HF patients, such as patients being considered for heart transplantation. However, little is known about R–L mismatch and its prognostic
☆ These authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation. ⁎ Corresponding author at: Division of Cardiology, Mitsui Memorial Hospital, KandaIzumi-cho 1, Chiyoda-ku, Tokyo 101-8643, Japan. E-mail addresses: [email protected] (Y. Horiuchi), [email protected] (H. Nakajima), [email protected] (K. Tanabe).
https://doi.org/10.1016/j.ijcard.2017.11.004 0167-5273/© 2017 Elsevier B.V. All rights reserved.
impact in the general HF population, especially in HF patients with preserved ejection fraction (HFpEF) [6]. The present study aimed to investigate the prevalence of R–L mismatch in HFpEF patients and to determine whether R–L mismatch is a predictor of worse clinical outcomes. 2. Methods 2.1. Study population Consecutive patients who underwent right heart catheterization (RHC) have been prospectively registered in our institutional database since January 2012. Patients without acute coronary syndrome (ACS) or hemodialysis were retrospectively analyzed from January 2012 to September 2015. The indications for RHC were patients with suspected HF who had typical HF symptoms, chest X-ray showing congestion, elevation of brain natriuretic peptide (BNP) or echocardiography abnormalities. All patients underwent RHC during hospitalization. Patients with BNP level b100 pg/ml and patients without echocardiography data were excluded. Of these, patients with EF ≥40% were included in the study. Patient characteristics and medical history were obtained on admission. Ischemic heart disease as the etiology of HF was defined as the presence of at least one of the following: prior myocardial infarction, prior percutaneous coronary intervention or prior coronary bypass grafting. Hypertension (blood pressure ≥ 140/90 mm Hg or the use of antihypertensive medications), diabetes mellitus (hemoglobin A1c ≥ 6.5% or the use of oral hypoglycemic agents or insulin) and dyslipidemia (fasting serum low-density lipoprotein
144
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
cholesterol ≥140 mg/dL, high-density lipoprotein cholesterol b40 mg/dL, triglycerides ≥150 mg/dL, or the use of medications for dyslipidemia) were recorded. Left ventricular ejection fraction (LVEF) was calculated by the modified Simpson method. Ultrasonographers performed echocardiography and specialists of the Japanese Society of Echocardiography approved the findings. Vital signs, laboratory data and medication at the time of RHC were also collected. All patients provided written informed consent and all data were anonymized throughout the study and analysis. The study was conducted in accordance with the Declaration of Helsinki.
2.2. Right heart catheterization and match/mismatch hemodynamic characteristics RHC was performed after the optimal treatment of diuresis, vasodilators, and other pharmacologic therapies based on the treating physician's discretion. RHC was performed in the supine position and was measured by a 6F balloon-tipped fluid-filled catheter (Swan Ganz Thermodilution Catheter, Edwards Lifesciences, California, USA). Transducers were zeroed at the mid-axilla and measured by callipers in each case. RHC was placed under fluoroscopic guidance through a femoral vein to a pulmonary artery. We confirmed the wedge position of the catheter by fluoroscopy and the presence of typical wave forms. Hemodynamic data were measured at the end of expiration and represent the mean of ≥3 beats. Cardiac output was measured by the thermodilution method, and indexed to body surface area (cardiac index, CI). Right ventricle stroke work index (RVSWI) was calculated as follows: (CI/heart rate) × (mean pulmonary artery pressure [PAP] − mean right atrium pressure [RAP]) × 13.6. Pulmonary artery resistance index (PARI) was determined as follows: PARI = (mean PAP − pulmonary capillary wedge pressure [PCWP])/CI. Patients were divided into four match or mismatch groups based on the following definitions: low match group, RAP b10 and PCWP b22 mm Hg; high match group, RAP ≥10 and PCWP ≥22 mm Hg; high-R mismatch group, RAP ≥10 and PCWP b22 mm Hg; high-L mismatch group, RAP b10 and PCWP ≥22 mm Hg.
2.3. Study endpoints The endpoint was the composite of death or first HF hospitalization within 12 months. HF hospitalization was defined as an unexpected hospitalization with at least one of the following symptoms: increasing dyspnea on exertion, worsening orthopnea, paroxysmal nocturnal dyspnea, increasing fatigue/worsening exercise tolerance, or altered mental status and at least two of the following symptoms: peripheral edema, elevated jugular venous pressure, radiologic signs of HF, increasing abdominal distension or ascites, pulmonary edema or crackles, rapid weight gain, hepatojugular reflux, S3 gallop or elevated BNP. These endpoints were observed by retrospective medical record review.
2.4. Statistical analysis Normally distributed continuous variables were described as mean ± standard deviation and non-normally distributed data were expressed as medians and interquartile ranges. Categorical variables were described as percentages. The prevalence of the four match/mismatch groups (low match, high match, high-R mismatch, and high-L mismatch) was investigated. To compare characteristics among the four match/mismatch groups, we used the one-way ANOVA test for continuous variables and the chi-square test for categorical variables. We conducted multiple comparisons using Tukey–Kramer method. Kaplan–Meier curves and 12-month event rates of the composite endpoint were estimated. The log-rank test was used for comparisons among the four match/ mismatch groups. A p value of b0.05 was considered significant. Cox proportional hazards regression analysis was performed to investigate whether PCWP ≥22 mm Hg or RAP ≥10 mm Hg predicted 12-month clinical outcomes. Factors with a p value b0.05 by univariate analysis were included in the multivariate analysis. Because the composite events occurred in 23 patients, PCWP ≥22 mm Hg or RAP ≥10 mm Hg was adjusted by each confounding factor separately to avoid over-fitting. All statistical analyses were performed using JMP version 12.0.1 for Windows (SAS, North Carolina, USA).
3. Results 3.1. Study population During the study period, 456 patients without ACS or hemodialysis underwent RHC in our institution. We excluded 12 patients without BNP data, 106 patients with BNP b 100 pg/ml, and 34 patients without echocardiography. We also excluded 134 patients with EF b 40%. A total of 170 HFpEF patients were included in the final analysis. Mean age was 73 ± 11 years, 63% were male, and 31% had ischemic HF etiology. Mean EF was 58 ± 12%, median creatinine was 0.93 (0.76, 1.26) mg/dl, and median BNP was 331 (174, 581) pg/ml. ACE-Is or ARBs were prescribed in 74% and β-blockers were prescribed in 49%.
3.2. Match and mismatch groups Fig. 1A shows the prevalence of match and mismatch groups. The match group represented 83% (low match, 76%; high match, 6.5%) and the mismatch group represented 18% (high-R mismatch, 12%; high-L mismatch, 5.9%). Table 1 reports patient characteristics of match/ mismatch groups. Demographic characteristics, including age, sex, HF etiology and past medical history were not different among the groups. BNP was higher in the high match and high-L mismatch groups than in the low match and high-R mismatch groups. Regarding echocardiography, the high match and high-L mismatch groups had higher E wave and tricuspid regurgitation peak gradient (TRPG) than the low match and high-R mismatch groups. Hemodynamic data, such as pulmonary artery systolic pressure (PASP), pulmonary artery diastolic pressure (PADP), and RVSWI were also higher in the high match and high-L mismatch groups than in the low match and high-R mismatch groups. 3.3. Clinical outcomes The composite endpoint of death or HF hospitalization occurred in 14 patients (11%) in the low match, 3 patients (27%) in the high match, 1 patient (5%) in the high-R mismatch, and 5 patients (50%) in the high-L mismatch groups (Table 2A). Kaplan–Meier estimates showed that the composite endpoint was more frequently observed in the high match and the high-L mismatch groups than in the low match and high-R mismatch groups (Fig. 1B). Patients with PCWP ≥22 mm Hg were more frequently associated with the composite endpoint than patients with PCWP b 22 mm Hg (Fig. 1C). These relationships were not observed in RAP (Fig. 1D). In multivariate analysis, PCWP ≥22 mm Hg was a significant predictor of the composite endpoint (Table 2B). 3.4. Relationships between PCWP and other characteristics Fig. 2 shows the relationships between PCWP and other factors. EF were not significantly related to PCWP (r = − 0.035, p = 0.652). BNP and E/e′ showed significant but weak correlations with PCWP (BNP, r = 0.256, p b 0.001 and E/e′ r = 0.199, p = 0.015, respectively). Hemodynamic characteristics including PASP and PADP showed strong correlations with PCWP (PASP, r = 0.738, p b 0.001; PADP, r = 0.834, p b 0.001; RAP, r = 0.638, p b 0.001, respectively). 4. Discussion The present study revealed that: 1) R–L mismatch exists in HFpEF patients; 2) elevated left-sided pressure was a significant predictor of worse clinical outcomes; and 3) hemodynamic characteristics, including PASP and PADP, strongly correlate with PCWP. Dranzer et al. analyzed hemodynamic data of 4079 HFrEF patients over 3 years (1993 to 1997, 1998 to 2002, and, 2003 to 2007) [4], in which the frequency of concordant hemodynamics were 74%, 72%, and 73%, respectively. Campbell et al. analyzed hemodynamic data of 537 patients with advanced HF [5]. Among these patients, the frequency of concordant hemodynamics was 72%. In another study of patients with HFpEF, 11 HF patients with EF N 50% underwent RHC at rest and under loading conditions by lower body negative pressure and saline infusion [6]. Match or mismatch of RAP and PCWP was investigated among 66 paired measurements. The frequency of concordant hemodynamics was 79% (low match was 67% and high match was 12%) and high-R mismatch was more prevalent than high-L mismatch (21% vs. 0%). In the present study with HF patients with EF ≥ 40%, concordant hemodynamics were present in 84%, and high-R mismatch was also more frequently observed than high-L mismatch (11% vs. 5.6%). We also analyzed patients with EF N 50% (117 patients). The match group was 85% (low match, 77%; high match, 7.7%) and the mismatch group was 15%
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
(high-R mismatch, 10%; high-L mismatch, 5.1%), suggesting the comparable results to the previous study [6]. The cutoffs of RAP ≥10 mm Hg and PCWP ≥22 mm Hg were applied to compare the prevalence of R–L mismatch to prior studies [4–6]. However, PCWP ≧ 22 mm Hg can be high as a cutoff value because patients with PCWP b 22 mm Hg may benefit from decongestion therapy. When the cutoff of PCWP ≥18 mm Hg were applied in our study, the match group represented 81% (low match, 68%; high match, 13%) and the mismatch group represented 19% (high-R mismatch, 5.3%; high-L mismatch, 14%). R–L mismatch do exist with different cutoff values. The predictive value of relative elevations in RAP versus PCWP was not determined in the present study, although these relationships have been reported previously [7–9]. In patients with advanced HF, continuous elevation of PCWP can lead to a rise in PAP. The elevated right ventricular after-load causes right ventricle dysfunction, which
145
results in elevation of RAP. Elevated RAP is reportedly associated with impaired renal function and diastolic ventricular interdependence [10,11]. Therefore, disproportionately elevated RAP to PCWP can be a result of disease progression and may suggest cardiorenal syndrome and unfavorable ventricular interdependence. Because the condition of HFpEF patients in our study was less severe than that of those evaluated in previous reports, RAP elevation as a result of HF progression might not have been observed. Obesity is another possible mechanism of relative elevations in right side versus left side pressures currently being investigated [12]. Compared with non-obese HFpEF and healthy control subjects, obese HFpEF had increased epicardial fat thickness and greater total epicardial heart volume. These increases were associated with greater epicardial restraint and heightened ventricular interdependence, which may result in an increased ratio of right to left side filling pressures. In the present study, BMI was numerically higher in
Fig. 1. A. R–L mismatch in the study population. B. Kaplan–Meier estimates for the composite endpoint of death or first HF hospitalization within 12 months in 4 hemodynamic categories. The composite endpoint was more frequently observed in the high match and the high-L mismatch groups than in the low match and high-R mismatch groups (log-rank p b 0.001). C. Kaplan–Meier estimates for the composite endpoint of death or first HF hospitalization within 12 months in patients with PCWP ≥22 mm Hg or b22 mm Hg. Patients with PCWP ≥22 mm Hg were more frequently associated with the composite endpoint than patients with PCWP b22 mm Hg (log-rank p b 0.001). D. Kaplan-Meier estimates for the composite endpoint of death or first HF hospitalization in patients with RAP ≥ 10 mm Hg or b10 mm Hg. The rates of the composite endpoint were not different between the two groups (log-rank p = 0.867).
146
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
Fig. 1 (continued).
the patients with high-R mismatch compared to the other groups. Although values of BMI in the present analysis of Japanese population were lower than those in the Western population, obesity with elevated epicardial restraint and ventricular interdependence might be related to high-R mismatch in HFpEF. In HFpEF patients, RV dysfunction, defined as RV fractional area change of b35%, was shown to be predictive of worse outcomes [7]. RV fractional area change correlated with both PASP and RAP. After adjusting for PASP, RAP was not significantly associated with RV function, suggesting that PASP is more strongly related to RV dysfunction than RAP. RVSWI is another surrogate marker of RV function [13]. In the present analysis, RVSWI was higher in the high match and high-L mismatch groups than in the low match and high-R groups, suggesting that RVSWI correlates with PCWP rather than RAP. These facts indicate that elevated RAP may not represent impaired RV function in HFpEF patients. Elevated PCWP was a significant predictor of worse outcomes after adjusting for previously established prognostic factors, such as age, BNP and diuretic use [14–17]. Dorfs et al. investigated PCWP at rest
and during exercise in 355 patients with unexplained dyspnea and suspected HFpEF [18]. Resting PCWP was strongly linked to long-term mortality. In addition, increase in PCWP during exercise improved the prediction of mortality with resting PCWP alone. Because elevated PCWP is significantly associated with worse outcomes, physicians are required to estimate PCWP precisely in daily practice. In the present study, demographic and clinical characteristics were of limited use in estimating PCWP. BNP and E/e′ showed significant but weak correlations with PCWP. RAP, which correlates with JVP, showed a relatively strong correlation with PCWP. However, in patients with elevated PCWP (the high match and high-L mismatch groups), 48% of patients had elevated RAP. Almost half of patients with elevated PCWP may lack signs of right side congestion, such as jugular vein distention, edema or ascites. These mismatches can result in under usage of diuretic therapy. In patients with elevated RAP (the high-R mismatch and the high match groups), only 35% had elevated PCWP. In these patients, over usage of diuretics may occur. The present study demonstrated the limitation of right-sided filling pressure as a surrogate marker of left-sided pressure. From et al. examined the accuracy of bedside
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
147
Table 1 Clinical and hemodynamic characteristics according to concordant/disconcordant hemodynamics.
Age (years) Male sex (%) Body mass index (kg/m2) Ischemic heart failure etiology (%) Medical history Prior myocardial infarction (%) Prior percutaneous coronary intervention (%) Prior coronary bypass grafting (%) Hypertension (%) Diabetes (%) Dyslipidemia (%) Atrial fibrillation (%) Vital signs Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Heart rate (/bpm) Laboratory data Creatinine (mg/dl) Sodium (mEq/L) BNP (pg/ml) Log BNP Medications ACE-I or ARB (%) β-blocker (%) Mineralocorticoid (%) Diuretic (%) Echocardiography LV end diastolic volume (ml) LV end systolic volume (ml) LV ejection fraction (%) LV mass (g) Left atrial diameter (mm) E (mm Hg) e E/e′ Deceleration time (ms) TRPG (mm Hg) Hemodynamic profiles Pulmonary capillary wedge pressure (mm Hg) Pulmonary artery systolic pressure (mm Hg) Pulmonary artery diastolic pressure (mm Hg) Right atrium pressure (mm Hg) Right ventricle stroke work index (g·m/m2 per beat) Pulmonary artery resistance index (dynes·s·cm−5/m2) Cardiac index (L/min/m2)
Concordant hemodynamics
Disconcordant hemodynamics
Low match
High match
High-R mismatch
High-L mismatch
(n = 129)
(n = 11)
(n = 20)
(n = 10)
73 ± 11 62 23 ± 4.1 33
74 ± 9.0 55 24 ± 3.3 30
75 ± 10 75 25 ± 5.0 30
77 ± 10 50 24 ± 4.6 9.1
0.514 0.512 0.084 0.451
21 23 16 84 37 60 33
9.1 9.1 9.1 64 45 36 45
15 10 15 85 45 60 55
20 30 10 70 50 70 40
0.755 0.353 0.915 0.821 0.762 0.414 0.235
125 ± 20 70 ± 14 69 ± 14
125 ± 24 64 ± 12 65 ± 20
131 ± 18 69 ± 10 65 ± 12
134 ± 41 70 ± 14 66 ± 6.6
0.450 0.597 0.521
0.93 (0.76, 1.21) 140 ± 3.2 305 (167, 567) 2.5 ± 0.4
0.91 (0.75, 1.29) 141 ± 1.3 535 (453, 1253) 2.8 ± 0.3
0.99 (0.82, 1.40) 139 ± 3.9 252 (169, 459) 2.5 ± 0.3
1.13 (0.60, 2.19) 137 ± 5.5 538 (447, 929) 2.8 ± 0.3
0.746 0.035⁎ 0.009 0.012⁎⁎⁎
77 50 26 43
45 27 9.1 36
70 60 35 65
70 40 10 60
0.149 0.336 0.284 0.196
100 ± 47 43 ± 27 59 ± 12 202 ± 84 33 ± 4.2 87 ± 29 5.5 ± 1.6 17 ± 6.5 215 ± 74 31 ± 13
97 ± 30 42 ± 19 58 ± 10 221 ± 66 33 ± 4.6 100 ± 20 5.9 ± 1.8 18 ± 7.1 205 ± 50 35 ± 7.9
105 ± 56 50 ± 36 56 ± 12 203 ± 82 33 ± 5.7 95 ± 26 5.9 ± 1.6 17 ± 4.7 180 ± 50 27 ± 8.1
99 ± 52 41 ± 33 61 ± 16 221 ± 53 34 ± 4.6 108 ± 29 5.4 ± 2.1 21 ± 6.8 195 ± 32 41 ± 15
0.966 0.737 0.715 0.801 0.873 0.070 0.611 0.277 0.211 0.043⁎⁎⁎⁎
11 ± 4.4 30 ± 8.4 12 ± 3.9 4.8 ± 2.2 8.8 ± 4.0 207 (143, 257) 3.0 ± 0.7
28 ± 4.8 51 ± 12 26 ± 6.8 13 ± 2.7 14 ± 4.2 252 (101, 350) 2.8 ± 0.7
17 ± 2.5 34 ± 5.6 17 ± 4.7 13 ± 2.7 7.9 ± 4.3 186 (142, 227) 3.1 ± 0.7
24 ± 2.5 53 ± 21 20 ± 6.9 6.6 ± 2.6 17 ± 8.7 242 (173, 384) 3.3 ± 1.1
b0.001⁎⁎⁎⁎⁎ b0.001⁎⁎⁎⁎⁎⁎ b0.001a b0.001⁎⁎ b0.001⁎⁎⁎⁎⁎⁎⁎ 0.301 0.597
p value
ACE-I, angiotensin converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BNP, brain natriuretic peptide; LV, left ventricle, TRPG, tricuspid regurgitation peak gradient. ⁎ p b 0.05 for high match vs. high-L mismatch. ⁎⁎ p b 0.05 for low match vs. high match, low match vs. high-R mismatch, high match vs. high-L mismatch and high-R mismatch vs. high-L mismatch. ⁎⁎⁎ p b 0.05 for high match vs. high-R mismatch. ⁎⁎⁎⁎ p b 0.05 for high-R mismatch vs. high-L mismatch. ⁎⁎⁎⁎⁎ p b 0.05 for all comparisons except high match vs. high-L mismatch. ⁎⁎⁎⁎⁎⁎ p b 0.05 for low match vs. high match, low match vs. high-L mismatch, high match vs. high-R mismatch and high-R mismatch vs. high-L mismatch. ⁎⁎⁎⁎⁎⁎⁎ p b 0.05 for low match vs. high match, low match vs. high-L mismatch, high match vs. high-R mismatch and high-R mismatch vs. high-L mismatch. a all comparisons except high-R mismatch vs. high-L mismatch.
examination assessment of right and left filling pressures in patients referred for catheterization [19]. Abnormal right side filling pressure was defined by RAP ≧ 10 mm Hg and abnormal left side filling pressure was defined by PCWP ≧ 15 mm Hg. The match group represented 80% (low
match, 33%; high match, 47%) and the mismatch group represented 20% (high-R mismatch, 3%; high-L mismatch, 17%). Right and left heart pressures were accurately predicted from examination alone in 71% and 60%, respectively. Echocardiographic data (E/e′ and inferior vena cava)
Table 2A Clinical outcomes. Concordant hemodynamics
Death or hear failure hospitalization, n (%) Death, n (%) Heart failure hospitalization, n (%)
Disconcordant hemodynamics
Low match
High match
High-R mismatch
High-L mismatch
(n = 129)
(n = 11)
(n = 20)
(n = 10)
14 (11%) 4 (3.1%) 10 (7.8%)
3 (27%) 0 (0%) 3 (27%)
1 (5%) 0 (0%) 1 (5%)
5 (50%) 1 (10%) 5 (50%)
148
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
Table 2B Prognostic factors of death or heart failure hospitalization. Univariate analysis 95%-CI
Multivariate analysis
Variables
HR
p value Adjusted HR 95%-CI
PCWP ≧ 22 mm Hg Age (per age)
5.40 2.17–12.5 b0.001 1.06 1.01–1.12 0.013
5.49 1.06
2.20–12.7 b0.001 1.01–1.12 0.012
p value
PCWP ≧ 22 mm Hg Log BNP
5.40 2.17–12.5 b0.001 4.06 1.12–12.1 0.015
4.56 3.35
1.82–10.6 0.002 0.99–10.9 0.052
PCWP ≧ 22 mm Hg Diuretic
5.40 2.17–12.5 b0.001 2.35 1.02–5.83 0.045
5.35 2.32
2.15–12.3 b0.001 1.01–5.75 0.049
PCWP ≧ 22 mm Hg 5.40 2.17–12.5 b0.001 TRPG (per mm Hg) 1.03 1.00–1.06 0.030
3.96 1.02
1.43–10.1 0.010 0.99–1.05 0.132
BNP, brain natriuretic peptide; CI, confidence interval; HR, hazard ratio; PCWP, Pulmonary capillary wedge pressure; TRPG, tricuspid regurgitation peak gradient.
and NT-pro-BNP did not improve accuracy beyond bed side examination alone. These findings suggest that physical assessment of PCWP or estimation based on RAP can be inaccurate, even if E/e′ and natriuretic peptide were incorporated. Direct assessment with RHC can be considered to measure PCWP precisely. Although direct assessment with RHC is golden standard for precise observation of PCWP and LV filling pressure, they are invasive and difficult to be performed repeatedly. In the current analysis, PASP and PADP were highly correlated with PCWP. In advanced HF, PASP was shown to relate closely to PCWP than RAP, suggesting PASP is less affected by the interposition of RV between PA and RA [5]. In patients with HFpEF and noncardiac dyspnea, Borlaug et al. examined hemodynamic response to exercise with RHC [20]. At rest and during exercise, patients with HFpEF had higher resting PASP and PCWP than those with noncardiac
dyspnea. In addition, PASP and PCWP were highly correlated. The relationship between elevated PADP and HF events has also been recently investigated using implantable hemodynamic monitors [21]. Patients with elevated PADP at baseline were more likely to experience future HF events. Increased PADP may suggest the presence of triggers of fluid retention and precede acute HF events [22]. Hemodynamicguided therapy for HF patients was proven to be superior to conventional strategies, which depend on monitoring of body weight and symptoms [23,24]. Because PASP and PADP are correlated to PCWP and can be estimated noninvasively by Doppler echocardiography [25–28], these parameters could be useful to predict PCWP.
5. Limitations There are several limitations of this investigation. First, the retrospective nature of this study could affect the results. Identified or unidentified confounding factors could influence the outcome. Second, we do not have sufficient information concerning patients' symptoms and physical examination findings. However, hemodynamic data, such as elevated PCWP or RAP, could substitute for this data. Third, we could not find out specific laboratory or hemodynamic findings which indicate either high-R or high-L mismatch and its pathophysiology. In patients with high match and high-L mismatch, E wave, BNP and TRPG were higher than in those with low match and high-R mismatch. However, these findings were associated with elevated PCWP rather than mismatch. Additional evaluation for RV function, ventricular interaction and pericardial restraint with echocardiography or cardiac magnetic resonance imaging would provide a better understanding of RV function in R–L mismatch and should be investigated in future studies.
Fig. 2. A. Relationship between PCWP and log BNP (r = 0.256, p b 0.001). B. Relationship between PCWP and EF (r = −0.035, p = 0.652). C. Relationship between PCWP and E/e′ (r = 0.199, p = 0.015). D. Relationship between PCWP and PASP (r = 0.738, p b 0.001). E. Relationship between PCWP and PADP (r = 0.834, p b 0.001). F. Relationship between PCWP and RAP (r = 0.638, p b 0.001).
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
6. Conclusions R–L mismatch exists in HFpEF patients. Elevated left-sided pressure was a significant predictor of worse clinical outcomes, while elevated right-sided pressure was not associated with worse outcomes. Physicians are required to estimate left-sided pressure precisely. Because clinical characteristics were of limited use in estimating leftsided pressure, utilization of additional hemodynamic parameters, such as PASP or PADP, could be useful. Conflict of interest Kengo Tanabe received remuneration from Abbott Vascular, Terumo, Kaneka, Zeon, Sanofi, Daiichi-Sankyo, and Tanabe-Mitsubishi. Acknowledgement This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. References [1] A.G. Vinayak, J. Levitt, B. Gehlbach, A.S. Pohlman, J.B. Hall, J.P. Kress, Usefulness of the external jugular vein examination in detecting abnormal central venous pressure in critically ill patients, Arch. Intern. Med. 166 (2006) 2132–2137. [2] A. Nohria, E. Lewis, L.W. Stevenson, Medical management of advanced heart failure, JAMA 287 (2002) 628–640. [3] M.H. Drazner, A.S. Hellkamp, C.V. Leier, et al., Value of clinician assessment of hemodynamics in advanced heart failure: the ESCAPE trial, Circ. Heart Fail. 1 (2008) 170–177. [4] M.H. Drazner, R.N. Brown, P.A. Kaiser, et al., Relationship of right- and left-sided filling pressures in patients with advanced heart failure: a 14-year multiinstitutional analysis, J Heart Lung Transplant 31 (2012) 67–72. [5] P. Campbell, M.H. Drazner, M. Kato, et al., Mismatch of right- and left-sided filling pressures in chronic heart failure, J. Card. Fail. 17 (2011) 561–568. [6] M.H. Drazner, A. Prasad, C. Ayers, et al., The relationship of right- and left-sided filling pressures in patients with heart failure and a preserved ejection fraction, Circ. Heart Fail. 3 (2010) 202–206. [7] V. Melenovsky, S.J. Hwang, G. Lin, M.M. Redfield, B.A. Borlaug, Right heart dysfunction in heart failure with preserved ejection fraction, Eur. Heart J. 35 (2014) 3452–3462. [8] M.H. Drazner, M. Velez-Martinez, C.R. Ayers, et al., Relationship of right- to left-sided ventricular filling pressures in advanced heart failure: insights from the ESCAPE trial, Circ. Heart Fail. 6 (2013) 264–270. [9] J.L. Grodin, M.H. Drazner, M. Dupont, et al., A disproportionate elevation in right ventricular filling pressure, in relation to left ventricular filling pressure, is associated with renal impairment and increased mortality in advanced decompensated heart failure, Am. Heart J. 169 (2015) 806–812. [10] A. Nohria, V. Hasselblad, A. Stebbins, et al., Cardiorenal interactions: insights from the ESCAPE trial, J. Am. Coll. Cardiol. 51 (2008) 1268–1274. [11] J.J. Atherton, T.D. Moore, S.S. Lele, et al., Diastolic ventricular interaction in chronic heart failure, Lancet 349 (1997) 1720–1724.
149
[12] M. Obokata, Y.N.V. Reddy, S.V. Pislaru, V. Melenovsky, B.A. Borlaug, Evidence supporting the existence of a distinct obese phenotype of heart failure with preserved ejection fraction, Circulation 136 (2017) 6–19. [13] J.C. Matthews, T.M. Koelling, F.D. Pagani, K.D. Aaronson, The right ventricular failure risk score a pre-operative tool for assessing the risk of right ventricular failure in left ventricular assist device candidates, J. Am. Coll. Cardiol. 51 (2008) 2163–2172. [14] W.T. Abraham, G.C. Fonarow, N.M. Albert, et al., OPTIMIZE-HF Investigators and Coordinators, Predictors of in-hospital mortality in patients hospitalized for heart failure: insights from the Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF), J. Am. Coll. Cardiol. 52 (2008) 347–356. [15] G.C. Fonarow, W.F. Peacock, C.O. Phillips, M.M. Givertz, M. Lopatin, ADHERE Scientific Advisory Committee and Investigators, Admission B-type natriuretic peptide levels and in-hospital mortality in acute decompensated heart failure, J. Am. Coll. Cardiol. 49 (2007) 1943–1950. [16] C.M. O'Connor, V. Hasselblad, R.H. Mehta, et al., Triage after hospitalization with advanced heart failure: the ESCAPE (Evaluation Study of Congestive Heart Failure and Pulmonary Artery Catheterization Effectiveness) risk model and discharge score, J. Am. Coll. Cardiol. 55 (2010) 872–878. [17] M. Domanski, J. Norman, B. Pitt, M. Haigney, S. Hanlon, E. Peyster, Studies of left ventricular dysfunction. Diuretic use, progressive heart failure, and death in patients in the Studies Of Left Ventricular Dysfunction (SOLVD), J. Am. Coll. Cardiol. 42 (2003) 705–708. [18] S. Dorfs, W. Zeh, W. Hochholzer, et al., Pulmonary capillary wedge pressure during exercise and long-term mortality in patients with suspected heart failure with preserved ejection fraction, Eur. Heart J. 35 (2014) 3103–3112. [19] A.M. From, C.S. Lam, S.R. Pitta, et al., Bedside assessment of cardiac hemodynamics: the impact of noninvasive testing and examiner experience, Am. J. Med. 124 (2011) 1051–1057. [20] B.A. Borlaug, R.A. Nishimura, P. Sorajja, C.S. Lam, M.M. Redfield, Exercise hemodynamics enhance diagnosis of early heart failure with preserved ejection fraction, Circ. Heart Fail. 3 (2010) 588–595. [21] L.W. Stevenson, M. Zile, T.D. Bennett, et al., Chronic ambulatory intracardiac pressures and future heart failure events, Circ. Heart Fail. 3 (2010) 580–587. [22] M.R. Zile, T.D. Bennett, M. St John Sutton, et al., Transition from chronic compensated to acute decompensated heart failure: pathophysiological insights obtained from continuous monitoring of intracardiac pressures, Circulation 118 (2008) 1433–1441. [23] W.T. Abraham, P.B. Adamson, R.C. Bourge, et al., CHAMPION Trial Study Group, Wireless pulmonary artery haemodynamic monitoring in chronic heart failure: a randomised controlled trial, Lancet 377 (2011) 658–666. [24] W.T. Abraham, L.W. Stevenson, R.C. Bourge, J.A. Lindenfeld, J.G. Bauman, P.B. Adamson, CHAMPION Trial Study Group, Sustained efficacy of pulmonary artery pressure to guide adjustment of chronic heart failure therapy: complete follow-up results from the CHAMPION randomised trial, Lancet 387 (2016) 453–461. [25] P.G. Yock, R.L. Popp, Noninvasive estimation of right ventricular systolic pressure by Doppler ultrasound in patients with tricuspid regurgitation, Circulation 70 (1984) 657–662. [26] M. Berger, A. Haimowitz, A. Van Tosh, R.L. Berdoff, E. Goldberg, Quantitative assessment of pulmonary hypertension in patients with tricuspid regurgitation using continuous wave Doppler ultrasound, J. Am. Coll. Cardiol. 6 (1985) 359–365. [27] B. Stephen, P. Dalal, M. Berger, P. Schweitzer, S. Hecht, Noninvasive estimation of pulmonary artery diastolic pressure in patients with tricuspid regurgitation by Doppler echocardiography, Chest 116 (1999) 73–77. [28] A.E. Abbas, F.D. Fortuin, N.B. Schiller, C.P. Appleton, C.A. Moreno, S.J. Lester, Echocardiographic determination of mean pulmonary artery pressure, Am. J. Cardiol. 92 (2003) 1373–1376.
Contents lists available at ScienceDirect
International Journal of Cardiology journal homepage: www.elsevier.com/locate/ijcard
Mismatch between right- and left-sided filling pressures in heart failure patients with preserved ejection fraction☆ Yu Horiuchi a,⁎, Shuzou Tanimoto a, Jiro Aoki a, Nozomi Fuse a, Kazuyuki Yahagi a, Keita Koseki a, Taishi Okuno a, Hiroyoshi Nakajima b, Kazuhiro Hara c, Kengo Tanabe a a b c
Division of Cardiology, Mitsui Memorial Hospital, Tokyo, Japan Division of General Medicine, Mitsui Memorial Hospital, Tokyo, Japan Division of Internal Medicine, Mitsui Memorial Hospital, Tokyo, Japan
a r t i c l e
i n f o
Article history: Received 1 February 2017 Received in revised form 26 October 2017 Accepted 2 November 2017
Keywords: Heart failure with preserved ejection fraction Right heart catheterization Pulmonary capillary wedge pressure
a b s t r a c t Background: Mismatch between right- and left-sided filling pressures is poorly understood in heart failure with preserved ejection fraction (HFpEF). Methods and results: We retrospectively analyzed 170 patients with HFpEF (EF ≥ 40%) who underwent right heart catheterization. Low match (right atrial pressure [RAP] b 10 mm Hg and pulmonary capillary wedge pressure [PCWP] b 10 mm Hg) was 76%, high match (RAP ≥ 10 mm Hg and PCWP ≥ 22 mm Hg) was 6.5%, high-R mismatch (RAP ≥ 10 mm Hg and PCWP b 22 mm Hg) was 12%, and high-L mismatch (RAP b 10 mmHg and PCWP ≥ 22 mm Hg) was 5.9%. Elevated PCWP was a significant predictor of the composite endpoint of death or HF hospitalization within 12 months (hazard ratio 5.40, 95% confidence interval 2.17–12.5, p b 0.001). Elevated RAP was not significantly associated with worse outcomes. Pulmonary artery systolic pressure (PASP) and diastolic pressure (PADP) showed strong correlations with PCWP (PASP, r = 0.738, p b 0.001; PADP, r = 0.834, p b 0.001; RAP, r = 0.638, p b 0.001, respectively). Conclusions: Discordance exists between right- and left-sided filling pressures in HFpEF. Physicians may utilize pulmonary artery pressure to evaluate left-sided filling pressure, which is a significant predictor of prognosis. © 2017 Elsevier B.V. All rights reserved.
1. Introduction Decongestion is a crucial component of heart failure (HF) treatment, because many symptoms of HF arise from congestion, which is associated with elevated left-sided filling pressure. Jugular venous pressure (JVP) is one of the most reliable physical signs for elevated left-sided filling pressure [1–3]. The relationship between JVP and left-sided filling pressure is predicated upon concordance between right- and left-sided filling pressure, although dissociation between right- and left-sided pressure has been reported [4,5], in which right- and left-sided pressures are not tightly coupled and decongestion therapy guided by JVP can result in over- or under-treatment. This dissociation, termed right–left (R–L) mismatch, has been primarily investigated in severe grade HF patients, such as patients being considered for heart transplantation. However, little is known about R–L mismatch and its prognostic
☆ These authors take responsibility for all aspects of the reliability and freedom from bias of the data presented and their discussed interpretation. ⁎ Corresponding author at: Division of Cardiology, Mitsui Memorial Hospital, KandaIzumi-cho 1, Chiyoda-ku, Tokyo 101-8643, Japan. E-mail addresses: [email protected] (Y. Horiuchi), [email protected] (H. Nakajima), [email protected] (K. Tanabe).
https://doi.org/10.1016/j.ijcard.2017.11.004 0167-5273/© 2017 Elsevier B.V. All rights reserved.
impact in the general HF population, especially in HF patients with preserved ejection fraction (HFpEF) [6]. The present study aimed to investigate the prevalence of R–L mismatch in HFpEF patients and to determine whether R–L mismatch is a predictor of worse clinical outcomes. 2. Methods 2.1. Study population Consecutive patients who underwent right heart catheterization (RHC) have been prospectively registered in our institutional database since January 2012. Patients without acute coronary syndrome (ACS) or hemodialysis were retrospectively analyzed from January 2012 to September 2015. The indications for RHC were patients with suspected HF who had typical HF symptoms, chest X-ray showing congestion, elevation of brain natriuretic peptide (BNP) or echocardiography abnormalities. All patients underwent RHC during hospitalization. Patients with BNP level b100 pg/ml and patients without echocardiography data were excluded. Of these, patients with EF ≥40% were included in the study. Patient characteristics and medical history were obtained on admission. Ischemic heart disease as the etiology of HF was defined as the presence of at least one of the following: prior myocardial infarction, prior percutaneous coronary intervention or prior coronary bypass grafting. Hypertension (blood pressure ≥ 140/90 mm Hg or the use of antihypertensive medications), diabetes mellitus (hemoglobin A1c ≥ 6.5% or the use of oral hypoglycemic agents or insulin) and dyslipidemia (fasting serum low-density lipoprotein
144
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
cholesterol ≥140 mg/dL, high-density lipoprotein cholesterol b40 mg/dL, triglycerides ≥150 mg/dL, or the use of medications for dyslipidemia) were recorded. Left ventricular ejection fraction (LVEF) was calculated by the modified Simpson method. Ultrasonographers performed echocardiography and specialists of the Japanese Society of Echocardiography approved the findings. Vital signs, laboratory data and medication at the time of RHC were also collected. All patients provided written informed consent and all data were anonymized throughout the study and analysis. The study was conducted in accordance with the Declaration of Helsinki.
2.2. Right heart catheterization and match/mismatch hemodynamic characteristics RHC was performed after the optimal treatment of diuresis, vasodilators, and other pharmacologic therapies based on the treating physician's discretion. RHC was performed in the supine position and was measured by a 6F balloon-tipped fluid-filled catheter (Swan Ganz Thermodilution Catheter, Edwards Lifesciences, California, USA). Transducers were zeroed at the mid-axilla and measured by callipers in each case. RHC was placed under fluoroscopic guidance through a femoral vein to a pulmonary artery. We confirmed the wedge position of the catheter by fluoroscopy and the presence of typical wave forms. Hemodynamic data were measured at the end of expiration and represent the mean of ≥3 beats. Cardiac output was measured by the thermodilution method, and indexed to body surface area (cardiac index, CI). Right ventricle stroke work index (RVSWI) was calculated as follows: (CI/heart rate) × (mean pulmonary artery pressure [PAP] − mean right atrium pressure [RAP]) × 13.6. Pulmonary artery resistance index (PARI) was determined as follows: PARI = (mean PAP − pulmonary capillary wedge pressure [PCWP])/CI. Patients were divided into four match or mismatch groups based on the following definitions: low match group, RAP b10 and PCWP b22 mm Hg; high match group, RAP ≥10 and PCWP ≥22 mm Hg; high-R mismatch group, RAP ≥10 and PCWP b22 mm Hg; high-L mismatch group, RAP b10 and PCWP ≥22 mm Hg.
2.3. Study endpoints The endpoint was the composite of death or first HF hospitalization within 12 months. HF hospitalization was defined as an unexpected hospitalization with at least one of the following symptoms: increasing dyspnea on exertion, worsening orthopnea, paroxysmal nocturnal dyspnea, increasing fatigue/worsening exercise tolerance, or altered mental status and at least two of the following symptoms: peripheral edema, elevated jugular venous pressure, radiologic signs of HF, increasing abdominal distension or ascites, pulmonary edema or crackles, rapid weight gain, hepatojugular reflux, S3 gallop or elevated BNP. These endpoints were observed by retrospective medical record review.
2.4. Statistical analysis Normally distributed continuous variables were described as mean ± standard deviation and non-normally distributed data were expressed as medians and interquartile ranges. Categorical variables were described as percentages. The prevalence of the four match/mismatch groups (low match, high match, high-R mismatch, and high-L mismatch) was investigated. To compare characteristics among the four match/mismatch groups, we used the one-way ANOVA test for continuous variables and the chi-square test for categorical variables. We conducted multiple comparisons using Tukey–Kramer method. Kaplan–Meier curves and 12-month event rates of the composite endpoint were estimated. The log-rank test was used for comparisons among the four match/ mismatch groups. A p value of b0.05 was considered significant. Cox proportional hazards regression analysis was performed to investigate whether PCWP ≥22 mm Hg or RAP ≥10 mm Hg predicted 12-month clinical outcomes. Factors with a p value b0.05 by univariate analysis were included in the multivariate analysis. Because the composite events occurred in 23 patients, PCWP ≥22 mm Hg or RAP ≥10 mm Hg was adjusted by each confounding factor separately to avoid over-fitting. All statistical analyses were performed using JMP version 12.0.1 for Windows (SAS, North Carolina, USA).
3. Results 3.1. Study population During the study period, 456 patients without ACS or hemodialysis underwent RHC in our institution. We excluded 12 patients without BNP data, 106 patients with BNP b 100 pg/ml, and 34 patients without echocardiography. We also excluded 134 patients with EF b 40%. A total of 170 HFpEF patients were included in the final analysis. Mean age was 73 ± 11 years, 63% were male, and 31% had ischemic HF etiology. Mean EF was 58 ± 12%, median creatinine was 0.93 (0.76, 1.26) mg/dl, and median BNP was 331 (174, 581) pg/ml. ACE-Is or ARBs were prescribed in 74% and β-blockers were prescribed in 49%.
3.2. Match and mismatch groups Fig. 1A shows the prevalence of match and mismatch groups. The match group represented 83% (low match, 76%; high match, 6.5%) and the mismatch group represented 18% (high-R mismatch, 12%; high-L mismatch, 5.9%). Table 1 reports patient characteristics of match/ mismatch groups. Demographic characteristics, including age, sex, HF etiology and past medical history were not different among the groups. BNP was higher in the high match and high-L mismatch groups than in the low match and high-R mismatch groups. Regarding echocardiography, the high match and high-L mismatch groups had higher E wave and tricuspid regurgitation peak gradient (TRPG) than the low match and high-R mismatch groups. Hemodynamic data, such as pulmonary artery systolic pressure (PASP), pulmonary artery diastolic pressure (PADP), and RVSWI were also higher in the high match and high-L mismatch groups than in the low match and high-R mismatch groups. 3.3. Clinical outcomes The composite endpoint of death or HF hospitalization occurred in 14 patients (11%) in the low match, 3 patients (27%) in the high match, 1 patient (5%) in the high-R mismatch, and 5 patients (50%) in the high-L mismatch groups (Table 2A). Kaplan–Meier estimates showed that the composite endpoint was more frequently observed in the high match and the high-L mismatch groups than in the low match and high-R mismatch groups (Fig. 1B). Patients with PCWP ≥22 mm Hg were more frequently associated with the composite endpoint than patients with PCWP b 22 mm Hg (Fig. 1C). These relationships were not observed in RAP (Fig. 1D). In multivariate analysis, PCWP ≥22 mm Hg was a significant predictor of the composite endpoint (Table 2B). 3.4. Relationships between PCWP and other characteristics Fig. 2 shows the relationships between PCWP and other factors. EF were not significantly related to PCWP (r = − 0.035, p = 0.652). BNP and E/e′ showed significant but weak correlations with PCWP (BNP, r = 0.256, p b 0.001 and E/e′ r = 0.199, p = 0.015, respectively). Hemodynamic characteristics including PASP and PADP showed strong correlations with PCWP (PASP, r = 0.738, p b 0.001; PADP, r = 0.834, p b 0.001; RAP, r = 0.638, p b 0.001, respectively). 4. Discussion The present study revealed that: 1) R–L mismatch exists in HFpEF patients; 2) elevated left-sided pressure was a significant predictor of worse clinical outcomes; and 3) hemodynamic characteristics, including PASP and PADP, strongly correlate with PCWP. Dranzer et al. analyzed hemodynamic data of 4079 HFrEF patients over 3 years (1993 to 1997, 1998 to 2002, and, 2003 to 2007) [4], in which the frequency of concordant hemodynamics were 74%, 72%, and 73%, respectively. Campbell et al. analyzed hemodynamic data of 537 patients with advanced HF [5]. Among these patients, the frequency of concordant hemodynamics was 72%. In another study of patients with HFpEF, 11 HF patients with EF N 50% underwent RHC at rest and under loading conditions by lower body negative pressure and saline infusion [6]. Match or mismatch of RAP and PCWP was investigated among 66 paired measurements. The frequency of concordant hemodynamics was 79% (low match was 67% and high match was 12%) and high-R mismatch was more prevalent than high-L mismatch (21% vs. 0%). In the present study with HF patients with EF ≥ 40%, concordant hemodynamics were present in 84%, and high-R mismatch was also more frequently observed than high-L mismatch (11% vs. 5.6%). We also analyzed patients with EF N 50% (117 patients). The match group was 85% (low match, 77%; high match, 7.7%) and the mismatch group was 15%
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
(high-R mismatch, 10%; high-L mismatch, 5.1%), suggesting the comparable results to the previous study [6]. The cutoffs of RAP ≥10 mm Hg and PCWP ≥22 mm Hg were applied to compare the prevalence of R–L mismatch to prior studies [4–6]. However, PCWP ≧ 22 mm Hg can be high as a cutoff value because patients with PCWP b 22 mm Hg may benefit from decongestion therapy. When the cutoff of PCWP ≥18 mm Hg were applied in our study, the match group represented 81% (low match, 68%; high match, 13%) and the mismatch group represented 19% (high-R mismatch, 5.3%; high-L mismatch, 14%). R–L mismatch do exist with different cutoff values. The predictive value of relative elevations in RAP versus PCWP was not determined in the present study, although these relationships have been reported previously [7–9]. In patients with advanced HF, continuous elevation of PCWP can lead to a rise in PAP. The elevated right ventricular after-load causes right ventricle dysfunction, which
145
results in elevation of RAP. Elevated RAP is reportedly associated with impaired renal function and diastolic ventricular interdependence [10,11]. Therefore, disproportionately elevated RAP to PCWP can be a result of disease progression and may suggest cardiorenal syndrome and unfavorable ventricular interdependence. Because the condition of HFpEF patients in our study was less severe than that of those evaluated in previous reports, RAP elevation as a result of HF progression might not have been observed. Obesity is another possible mechanism of relative elevations in right side versus left side pressures currently being investigated [12]. Compared with non-obese HFpEF and healthy control subjects, obese HFpEF had increased epicardial fat thickness and greater total epicardial heart volume. These increases were associated with greater epicardial restraint and heightened ventricular interdependence, which may result in an increased ratio of right to left side filling pressures. In the present study, BMI was numerically higher in
Fig. 1. A. R–L mismatch in the study population. B. Kaplan–Meier estimates for the composite endpoint of death or first HF hospitalization within 12 months in 4 hemodynamic categories. The composite endpoint was more frequently observed in the high match and the high-L mismatch groups than in the low match and high-R mismatch groups (log-rank p b 0.001). C. Kaplan–Meier estimates for the composite endpoint of death or first HF hospitalization within 12 months in patients with PCWP ≥22 mm Hg or b22 mm Hg. Patients with PCWP ≥22 mm Hg were more frequently associated with the composite endpoint than patients with PCWP b22 mm Hg (log-rank p b 0.001). D. Kaplan-Meier estimates for the composite endpoint of death or first HF hospitalization in patients with RAP ≥ 10 mm Hg or b10 mm Hg. The rates of the composite endpoint were not different between the two groups (log-rank p = 0.867).
146
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
Fig. 1 (continued).
the patients with high-R mismatch compared to the other groups. Although values of BMI in the present analysis of Japanese population were lower than those in the Western population, obesity with elevated epicardial restraint and ventricular interdependence might be related to high-R mismatch in HFpEF. In HFpEF patients, RV dysfunction, defined as RV fractional area change of b35%, was shown to be predictive of worse outcomes [7]. RV fractional area change correlated with both PASP and RAP. After adjusting for PASP, RAP was not significantly associated with RV function, suggesting that PASP is more strongly related to RV dysfunction than RAP. RVSWI is another surrogate marker of RV function [13]. In the present analysis, RVSWI was higher in the high match and high-L mismatch groups than in the low match and high-R groups, suggesting that RVSWI correlates with PCWP rather than RAP. These facts indicate that elevated RAP may not represent impaired RV function in HFpEF patients. Elevated PCWP was a significant predictor of worse outcomes after adjusting for previously established prognostic factors, such as age, BNP and diuretic use [14–17]. Dorfs et al. investigated PCWP at rest
and during exercise in 355 patients with unexplained dyspnea and suspected HFpEF [18]. Resting PCWP was strongly linked to long-term mortality. In addition, increase in PCWP during exercise improved the prediction of mortality with resting PCWP alone. Because elevated PCWP is significantly associated with worse outcomes, physicians are required to estimate PCWP precisely in daily practice. In the present study, demographic and clinical characteristics were of limited use in estimating PCWP. BNP and E/e′ showed significant but weak correlations with PCWP. RAP, which correlates with JVP, showed a relatively strong correlation with PCWP. However, in patients with elevated PCWP (the high match and high-L mismatch groups), 48% of patients had elevated RAP. Almost half of patients with elevated PCWP may lack signs of right side congestion, such as jugular vein distention, edema or ascites. These mismatches can result in under usage of diuretic therapy. In patients with elevated RAP (the high-R mismatch and the high match groups), only 35% had elevated PCWP. In these patients, over usage of diuretics may occur. The present study demonstrated the limitation of right-sided filling pressure as a surrogate marker of left-sided pressure. From et al. examined the accuracy of bedside
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
147
Table 1 Clinical and hemodynamic characteristics according to concordant/disconcordant hemodynamics.
Age (years) Male sex (%) Body mass index (kg/m2) Ischemic heart failure etiology (%) Medical history Prior myocardial infarction (%) Prior percutaneous coronary intervention (%) Prior coronary bypass grafting (%) Hypertension (%) Diabetes (%) Dyslipidemia (%) Atrial fibrillation (%) Vital signs Systolic blood pressure (mm Hg) Diastolic blood pressure (mm Hg) Heart rate (/bpm) Laboratory data Creatinine (mg/dl) Sodium (mEq/L) BNP (pg/ml) Log BNP Medications ACE-I or ARB (%) β-blocker (%) Mineralocorticoid (%) Diuretic (%) Echocardiography LV end diastolic volume (ml) LV end systolic volume (ml) LV ejection fraction (%) LV mass (g) Left atrial diameter (mm) E (mm Hg) e E/e′ Deceleration time (ms) TRPG (mm Hg) Hemodynamic profiles Pulmonary capillary wedge pressure (mm Hg) Pulmonary artery systolic pressure (mm Hg) Pulmonary artery diastolic pressure (mm Hg) Right atrium pressure (mm Hg) Right ventricle stroke work index (g·m/m2 per beat) Pulmonary artery resistance index (dynes·s·cm−5/m2) Cardiac index (L/min/m2)
Concordant hemodynamics
Disconcordant hemodynamics
Low match
High match
High-R mismatch
High-L mismatch
(n = 129)
(n = 11)
(n = 20)
(n = 10)
73 ± 11 62 23 ± 4.1 33
74 ± 9.0 55 24 ± 3.3 30
75 ± 10 75 25 ± 5.0 30
77 ± 10 50 24 ± 4.6 9.1
0.514 0.512 0.084 0.451
21 23 16 84 37 60 33
9.1 9.1 9.1 64 45 36 45
15 10 15 85 45 60 55
20 30 10 70 50 70 40
0.755 0.353 0.915 0.821 0.762 0.414 0.235
125 ± 20 70 ± 14 69 ± 14
125 ± 24 64 ± 12 65 ± 20
131 ± 18 69 ± 10 65 ± 12
134 ± 41 70 ± 14 66 ± 6.6
0.450 0.597 0.521
0.93 (0.76, 1.21) 140 ± 3.2 305 (167, 567) 2.5 ± 0.4
0.91 (0.75, 1.29) 141 ± 1.3 535 (453, 1253) 2.8 ± 0.3
0.99 (0.82, 1.40) 139 ± 3.9 252 (169, 459) 2.5 ± 0.3
1.13 (0.60, 2.19) 137 ± 5.5 538 (447, 929) 2.8 ± 0.3
0.746 0.035⁎ 0.009 0.012⁎⁎⁎
77 50 26 43
45 27 9.1 36
70 60 35 65
70 40 10 60
0.149 0.336 0.284 0.196
100 ± 47 43 ± 27 59 ± 12 202 ± 84 33 ± 4.2 87 ± 29 5.5 ± 1.6 17 ± 6.5 215 ± 74 31 ± 13
97 ± 30 42 ± 19 58 ± 10 221 ± 66 33 ± 4.6 100 ± 20 5.9 ± 1.8 18 ± 7.1 205 ± 50 35 ± 7.9
105 ± 56 50 ± 36 56 ± 12 203 ± 82 33 ± 5.7 95 ± 26 5.9 ± 1.6 17 ± 4.7 180 ± 50 27 ± 8.1
99 ± 52 41 ± 33 61 ± 16 221 ± 53 34 ± 4.6 108 ± 29 5.4 ± 2.1 21 ± 6.8 195 ± 32 41 ± 15
0.966 0.737 0.715 0.801 0.873 0.070 0.611 0.277 0.211 0.043⁎⁎⁎⁎
11 ± 4.4 30 ± 8.4 12 ± 3.9 4.8 ± 2.2 8.8 ± 4.0 207 (143, 257) 3.0 ± 0.7
28 ± 4.8 51 ± 12 26 ± 6.8 13 ± 2.7 14 ± 4.2 252 (101, 350) 2.8 ± 0.7
17 ± 2.5 34 ± 5.6 17 ± 4.7 13 ± 2.7 7.9 ± 4.3 186 (142, 227) 3.1 ± 0.7
24 ± 2.5 53 ± 21 20 ± 6.9 6.6 ± 2.6 17 ± 8.7 242 (173, 384) 3.3 ± 1.1
b0.001⁎⁎⁎⁎⁎ b0.001⁎⁎⁎⁎⁎⁎ b0.001a b0.001⁎⁎ b0.001⁎⁎⁎⁎⁎⁎⁎ 0.301 0.597
p value
ACE-I, angiotensin converting enzyme inhibitor; ARB, angiotensin II receptor blocker; BNP, brain natriuretic peptide; LV, left ventricle, TRPG, tricuspid regurgitation peak gradient. ⁎ p b 0.05 for high match vs. high-L mismatch. ⁎⁎ p b 0.05 for low match vs. high match, low match vs. high-R mismatch, high match vs. high-L mismatch and high-R mismatch vs. high-L mismatch. ⁎⁎⁎ p b 0.05 for high match vs. high-R mismatch. ⁎⁎⁎⁎ p b 0.05 for high-R mismatch vs. high-L mismatch. ⁎⁎⁎⁎⁎ p b 0.05 for all comparisons except high match vs. high-L mismatch. ⁎⁎⁎⁎⁎⁎ p b 0.05 for low match vs. high match, low match vs. high-L mismatch, high match vs. high-R mismatch and high-R mismatch vs. high-L mismatch. ⁎⁎⁎⁎⁎⁎⁎ p b 0.05 for low match vs. high match, low match vs. high-L mismatch, high match vs. high-R mismatch and high-R mismatch vs. high-L mismatch. a all comparisons except high-R mismatch vs. high-L mismatch.
examination assessment of right and left filling pressures in patients referred for catheterization [19]. Abnormal right side filling pressure was defined by RAP ≧ 10 mm Hg and abnormal left side filling pressure was defined by PCWP ≧ 15 mm Hg. The match group represented 80% (low
match, 33%; high match, 47%) and the mismatch group represented 20% (high-R mismatch, 3%; high-L mismatch, 17%). Right and left heart pressures were accurately predicted from examination alone in 71% and 60%, respectively. Echocardiographic data (E/e′ and inferior vena cava)
Table 2A Clinical outcomes. Concordant hemodynamics
Death or hear failure hospitalization, n (%) Death, n (%) Heart failure hospitalization, n (%)
Disconcordant hemodynamics
Low match
High match
High-R mismatch
High-L mismatch
(n = 129)
(n = 11)
(n = 20)
(n = 10)
14 (11%) 4 (3.1%) 10 (7.8%)
3 (27%) 0 (0%) 3 (27%)
1 (5%) 0 (0%) 1 (5%)
5 (50%) 1 (10%) 5 (50%)
148
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
Table 2B Prognostic factors of death or heart failure hospitalization. Univariate analysis 95%-CI
Multivariate analysis
Variables
HR
p value Adjusted HR 95%-CI
PCWP ≧ 22 mm Hg Age (per age)
5.40 2.17–12.5 b0.001 1.06 1.01–1.12 0.013
5.49 1.06
2.20–12.7 b0.001 1.01–1.12 0.012
p value
PCWP ≧ 22 mm Hg Log BNP
5.40 2.17–12.5 b0.001 4.06 1.12–12.1 0.015
4.56 3.35
1.82–10.6 0.002 0.99–10.9 0.052
PCWP ≧ 22 mm Hg Diuretic
5.40 2.17–12.5 b0.001 2.35 1.02–5.83 0.045
5.35 2.32
2.15–12.3 b0.001 1.01–5.75 0.049
PCWP ≧ 22 mm Hg 5.40 2.17–12.5 b0.001 TRPG (per mm Hg) 1.03 1.00–1.06 0.030
3.96 1.02
1.43–10.1 0.010 0.99–1.05 0.132
BNP, brain natriuretic peptide; CI, confidence interval; HR, hazard ratio; PCWP, Pulmonary capillary wedge pressure; TRPG, tricuspid regurgitation peak gradient.
and NT-pro-BNP did not improve accuracy beyond bed side examination alone. These findings suggest that physical assessment of PCWP or estimation based on RAP can be inaccurate, even if E/e′ and natriuretic peptide were incorporated. Direct assessment with RHC can be considered to measure PCWP precisely. Although direct assessment with RHC is golden standard for precise observation of PCWP and LV filling pressure, they are invasive and difficult to be performed repeatedly. In the current analysis, PASP and PADP were highly correlated with PCWP. In advanced HF, PASP was shown to relate closely to PCWP than RAP, suggesting PASP is less affected by the interposition of RV between PA and RA [5]. In patients with HFpEF and noncardiac dyspnea, Borlaug et al. examined hemodynamic response to exercise with RHC [20]. At rest and during exercise, patients with HFpEF had higher resting PASP and PCWP than those with noncardiac
dyspnea. In addition, PASP and PCWP were highly correlated. The relationship between elevated PADP and HF events has also been recently investigated using implantable hemodynamic monitors [21]. Patients with elevated PADP at baseline were more likely to experience future HF events. Increased PADP may suggest the presence of triggers of fluid retention and precede acute HF events [22]. Hemodynamicguided therapy for HF patients was proven to be superior to conventional strategies, which depend on monitoring of body weight and symptoms [23,24]. Because PASP and PADP are correlated to PCWP and can be estimated noninvasively by Doppler echocardiography [25–28], these parameters could be useful to predict PCWP.
5. Limitations There are several limitations of this investigation. First, the retrospective nature of this study could affect the results. Identified or unidentified confounding factors could influence the outcome. Second, we do not have sufficient information concerning patients' symptoms and physical examination findings. However, hemodynamic data, such as elevated PCWP or RAP, could substitute for this data. Third, we could not find out specific laboratory or hemodynamic findings which indicate either high-R or high-L mismatch and its pathophysiology. In patients with high match and high-L mismatch, E wave, BNP and TRPG were higher than in those with low match and high-R mismatch. However, these findings were associated with elevated PCWP rather than mismatch. Additional evaluation for RV function, ventricular interaction and pericardial restraint with echocardiography or cardiac magnetic resonance imaging would provide a better understanding of RV function in R–L mismatch and should be investigated in future studies.
Fig. 2. A. Relationship between PCWP and log BNP (r = 0.256, p b 0.001). B. Relationship between PCWP and EF (r = −0.035, p = 0.652). C. Relationship between PCWP and E/e′ (r = 0.199, p = 0.015). D. Relationship between PCWP and PASP (r = 0.738, p b 0.001). E. Relationship between PCWP and PADP (r = 0.834, p b 0.001). F. Relationship between PCWP and RAP (r = 0.638, p b 0.001).
Y. Horiuchi et al. / International Journal of Cardiology 257 (2018) 143–149
6. Conclusions R–L mismatch exists in HFpEF patients. Elevated left-sided pressure was a significant predictor of worse clinical outcomes, while elevated right-sided pressure was not associated with worse outcomes. Physicians are required to estimate left-sided pressure precisely. Because clinical characteristics were of limited use in estimating leftsided pressure, utilization of additional hemodynamic parameters, such as PASP or PADP, could be useful. Conflict of interest Kengo Tanabe received remuneration from Abbott Vascular, Terumo, Kaneka, Zeon, Sanofi, Daiichi-Sankyo, and Tanabe-Mitsubishi. Acknowledgement This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. References [1] A.G. Vinayak, J. Levitt, B. Gehlbach, A.S. Pohlman, J.B. Hall, J.P. Kress, Usefulness of the external jugular vein examination in detecting abnormal central venous pressure in critically ill patients, Arch. Intern. Med. 166 (2006) 2132–2137. [2] A. Nohria, E. Lewis, L.W. Stevenson, Medical management of advanced heart failure, JAMA 287 (2002) 628–640. [3] M.H. Drazner, A.S. Hellkamp, C.V. Leier, et al., Value of clinician assessment of hemodynamics in advanced heart failure: the ESCAPE trial, Circ. Heart Fail. 1 (2008) 170–177. [4] M.H. Drazner, R.N. Brown, P.A. Kaiser, et al., Relationship of right- and left-sided filling pressures in patients with advanced heart failure: a 14-year multiinstitutional analysis, J Heart Lung Transplant 31 (2012) 67–72. [5] P. Campbell, M.H. Drazner, M. Kato, et al., Mismatch of right- and left-sided filling pressures in chronic heart failure, J. Card. Fail. 17 (2011) 561–568. [6] M.H. Drazner, A. Prasad, C. Ayers, et al., The relationship of right- and left-sided filling pressures in patients with heart failure and a preserved ejection fraction, Circ. Heart Fail. 3 (2010) 202–206. [7] V. Melenovsky, S.J. Hwang, G. Lin, M.M. Redfield, B.A. Borlaug, Right heart dysfunction in heart failure with preserved ejection fraction, Eur. Heart J. 35 (2014) 3452–3462. [8] M.H. Drazner, M. Velez-Martinez, C.R. Ayers, et al., Relationship of right- to left-sided ventricular filling pressures in advanced heart failure: insights from the ESCAPE trial, Circ. Heart Fail. 6 (2013) 264–270. [9] J.L. Grodin, M.H. Drazner, M. Dupont, et al., A disproportionate elevation in right ventricular filling pressure, in relation to left ventricular filling pressure, is associated with renal impairment and increased mortality in advanced decompensated heart failure, Am. Heart J. 169 (2015) 806–812. [10] A. Nohria, V. Hasselblad, A. Stebbins, et al., Cardiorenal interactions: insights from the ESCAPE trial, J. Am. Coll. Cardiol. 51 (2008) 1268–1274. [11] J.J. Atherton, T.D. Moore, S.S. Lele, et al., Diastolic ventricular interaction in chronic heart failure, Lancet 349 (1997) 1720–1724.
149
[12] M. Obokata, Y.N.V. Reddy, S.V. Pislaru, V. Melenovsky, B.A. Borlaug, Evidence supporting the existence of a distinct obese phenotype of heart failure with preserved ejection fraction, Circulation 136 (2017) 6–19. [13] J.C. Matthews, T.M. Koelling, F.D. Pagani, K.D. Aaronson, The right ventricular failure risk score a pre-operative tool for assessing the risk of right ventricular failure in left ventricular assist device candidates, J. Am. Coll. Cardiol. 51 (2008) 2163–2172. [14] W.T. Abraham, G.C. Fonarow, N.M. Albert, et al., OPTIMIZE-HF Investigators and Coordinators, Predictors of in-hospital mortality in patients hospitalized for heart failure: insights from the Organized Program to Initiate Lifesaving Treatment in Hospitalized Patients with Heart Failure (OPTIMIZE-HF), J. Am. Coll. Cardiol. 52 (2008) 347–356. [15] G.C. Fonarow, W.F. Peacock, C.O. Phillips, M.M. Givertz, M. Lopatin, ADHERE Scientific Advisory Committee and Investigators, Admission B-type natriuretic peptide levels and in-hospital mortality in acute decompensated heart failure, J. Am. Coll. Cardiol. 49 (2007) 1943–1950. [16] C.M. O'Connor, V. Hasselblad, R.H. Mehta, et al., Triage after hospitalization with advanced heart failure: the ESCAPE (Evaluation Study of Congestive Heart Failure and Pulmonary Artery Catheterization Effectiveness) risk model and discharge score, J. Am. Coll. Cardiol. 55 (2010) 872–878. [17] M. Domanski, J. Norman, B. Pitt, M. Haigney, S. Hanlon, E. Peyster, Studies of left ventricular dysfunction. Diuretic use, progressive heart failure, and death in patients in the Studies Of Left Ventricular Dysfunction (SOLVD), J. Am. Coll. Cardiol. 42 (2003) 705–708. [18] S. Dorfs, W. Zeh, W. Hochholzer, et al., Pulmonary capillary wedge pressure during exercise and long-term mortality in patients with suspected heart failure with preserved ejection fraction, Eur. Heart J. 35 (2014) 3103–3112. [19] A.M. From, C.S. Lam, S.R. Pitta, et al., Bedside assessment of cardiac hemodynamics: the impact of noninvasive testing and examiner experience, Am. J. Med. 124 (2011) 1051–1057. [20] B.A. Borlaug, R.A. Nishimura, P. Sorajja, C.S. Lam, M.M. Redfield, Exercise hemodynamics enhance diagnosis of early heart failure with preserved ejection fraction, Circ. Heart Fail. 3 (2010) 588–595. [21] L.W. Stevenson, M. Zile, T.D. Bennett, et al., Chronic ambulatory intracardiac pressures and future heart failure events, Circ. Heart Fail. 3 (2010) 580–587. [22] M.R. Zile, T.D. Bennett, M. St John Sutton, et al., Transition from chronic compensated to acute decompensated heart failure: pathophysiological insights obtained from continuous monitoring of intracardiac pressures, Circulation 118 (2008) 1433–1441. [23] W.T. Abraham, P.B. Adamson, R.C. Bourge, et al., CHAMPION Trial Study Group, Wireless pulmonary artery haemodynamic monitoring in chronic heart failure: a randomised controlled trial, Lancet 377 (2011) 658–666. [24] W.T. Abraham, L.W. Stevenson, R.C. Bourge, J.A. Lindenfeld, J.G. Bauman, P.B. Adamson, CHAMPION Trial Study Group, Sustained efficacy of pulmonary artery pressure to guide adjustment of chronic heart failure therapy: complete follow-up results from the CHAMPION randomised trial, Lancet 387 (2016) 453–461. [25] P.G. Yock, R.L. Popp, Noninvasive estimation of right ventricular systolic pressure by Doppler ultrasound in patients with tricuspid regurgitation, Circulation 70 (1984) 657–662. [26] M. Berger, A. Haimowitz, A. Van Tosh, R.L. Berdoff, E. Goldberg, Quantitative assessment of pulmonary hypertension in patients with tricuspid regurgitation using continuous wave Doppler ultrasound, J. Am. Coll. Cardiol. 6 (1985) 359–365. [27] B. Stephen, P. Dalal, M. Berger, P. Schweitzer, S. Hecht, Noninvasive estimation of pulmonary artery diastolic pressure in patients with tricuspid regurgitation by Doppler echocardiography, Chest 116 (1999) 73–77. [28] A.E. Abbas, F.D. Fortuin, N.B. Schiller, C.P. Appleton, C.A. Moreno, S.J. Lester, Echocardiographic determination of mean pulmonary artery pressure, Am. J. Cardiol. 92 (2003) 1373–1376.