JACC: Heart Failure Vol. 1, No. 4, 2013� 2013 by the American College of Cardiology Foundation ISSN 2213-1779/$36.00Published by Elsevier Inc. http://dx.doi.org/10.1016/j.jchf.2013.05.001

Clinical Features, Hemodynamics, and Outcomes ofPulmonary Hypertension Due to Chronic Heart FailureWith Reduced Ejection Fraction

Pulmonary Hypertension and Heart Failure

Wayne L. Miller, MD, PHD,* Diane E. Grill, MSC,y Barry A. Borlaug, MD*

Rochester, Minnesota

JACC: HEART FAILURE CME

This article has been selected as the month’s JACC: Heart FailureCME activity.

Accreditation and Designation Statement

The American College of Cardiology Foundation (ACCF) isaccredited by the Accreditation Council for Continuing MedicalEducation (ACCME) to provide continuing medical education forphysicians.

The ACCF designates this Journal-based CME activity fora maximum of 1 AMA PRA Category 1 Credit(s). Physicians shouldonly claim credit commensurate with the extent of their participationin the activity.

Method of Participation and Receipt of CME Certificate

To obtain credit for JACC: Heart Failure CME, you must:1. Be an ACC member or JACC: Heart Failure subscriber.2. Carefully read the CME-designated article available online and

in this issue of the journal.3. Answer the post-test questions. At least 2 out of the 3 questions

provided must be answered correctly to obtain CME credit.4. Complete a brief evaluation.

From the *Division of Cardiovascular Diseases, Mayo Clinic, Rochester, Minnesota;

and the yDepartment of Biomedical Statistics and Informatics,Mayo Clinic, Rochester,

Minnesota. This study was supported by the Division of Cardiovascular Diseases, Mayo

5. Claim your CME credit and receive your certificate electroni-cally by following the instructions given at the conclusion ofthe activity.

CME Objective for This Article: After reading this article, thereader should be able to: 1) discuss the causes of pulmonaryhypertension in patients with systolic heart failure; 2) have anunderstanding of the differences between passive and mixedpulmonary hypertension; and 3) discuss the impact of pulmonaryhypertension subtypes on disease severity and risk of death.

CME Editor Disclosure: Deputy Managing Editor Mona Fiuzat,PharmD, FACC, reports that she has equity interest or stock optionsin ARCA Biopharma, consults for CCA, and receives researchsupport from ResMed, GE Healthcare, Gilead, Critical Diagnostics,BG Medicine, Otsuka, Astellas, and Roche Diagnostics.

Author Disclosures: The authors report that they have no rela-tionships relevant to the contents of this paper to disclose.

MediumofParticipation: Print (article only); online (article andquiz)

CME Term of Approval:Issue date: August 2013Expiration date: July 31, 2014

Clinic and Foundation, Rochester,Minnesota. The authors have reported that they have

no relationships relevant to the contents of this paper to disclose.

Manuscript received April 29, 2013; accepted May 3, 2013.

JACC: Heart Failure Vol. 1, No. 4, 2013 Miller et al.August 2013:290–9 Pulmonary Hypertension Due to Chronic HF

291

Clinical Features, Hemodynamics, and Ou

tcomes of Pulmonary Hypertension

Due to Chronic Heart Failure With Reduced Ejection Fraction

Pulmonary Hypertension and Heart Failure

Objectives T

he purpose of this study was to assess the clinical, functional, and hemodynamic characteristics of passive andmixed pulmonary hypertension (PH), compare outcomes, and contrast conventional and novel hemodynamicpartition values in patients with chronic heart failure of reduced left ventricular ejection fraction (HFREF).

Background P

H in HFREF may develop from left-sided venous congestion (passive PH) or the combination of pulmonary arterialdisease and venous congestion (mixed PH). Subgroup outcomes are not well defined, and the partition values usedto define risk are based largely on consensus opinion rather than outcome data.

Methods A

mbulatory patients referred for hemodynamic catheterization were analyzed retrospectively (N ¼ 463).

Results C

omparingpatientswithnoPH to thosewithpassivePHandmixedPH,aprogressive gradient ofmore severely derangedhemodynamics, diastolic dysfunction, and mitral regurgitation was observed. In multivariate analysis, the presence ofany PHormixedPHwas associatedwith older age, diuretic use, atrial fibrillation, and lower pulmonary artery compliance(PAC). Over a median follow-up of 2.1 years, patients with PH displayed greater risk of death (hazard ratio [HR]: 2.24;confidence limits [95%CL]: 1.39, 3.98; p< 0.001)withmixedPHdemonstrating greater risk than passive PH (HR: 1.55;95% CL: 1.11, 2.20; p< 0.001). Partition values identifying highest risk were pulmonary vascular resistance>4Woodunits, systolicpulmonary arterypressure>35mmHg, pulmonarywedgepressure>25mmHg,andPAC<2.0ml/mmHg.

Conclusions A

mong stable HFREF outpatients, PH was associated with markers of greater disease severity and risk of death.However, the presence of pulmonary arterial disease (mixed PH) carries incremental risk. Abnormalities inpulmonary vascular resistance and compliance may serve as novel therapeutic targets. (J Am Coll Cardiol HF2013;1:290–9) ª 2013 by the American College of Cardiology Foundation

The development of pulmonary hypertension (PH) is animportant marker in the progression of heart failure withreduced left ventricular (LV) ejection fraction (HFREF) (1–4).Initially, an increase in LV filling pressure causes an elevationin pulmonary venous pressures (5–7), resulting in post-capillaryor passive PH (8–10). Some patients also develop abnormali-ties in pulmonary arterial structure and function resulting ina superimposed pulmonary arterial (pre-capillary) vasocon-strictor component that produces a “mixed” PH characterizedby high LV filling pressure and elevated pulmonary vascularresistance (11–13). While PH is well established as animportant risk factor for poor outcome in HFREF (8–10), theavailable data regarding passive and mixed PH subgroups arevariable and limited by relatively short-term follow-up,incomplete characterization of left and right ventricular func-tion, and limited data on coexisting valvular pathology (11–15).Accordingly, the aim of this study was to comprehensively

assess the clinical, functional, and hemodynamic characteristicsof passive andmixedPH in a large cohort of ambulatory patientswith chronic HFREF. Additional aims were to compare long-term outcomes in these subgroups of PH, contrast conventionaland novel hemodynamic parameters of PH severity, and topropose partition values that might better stratify mortality riskand potentially serve as novel therapeutic targets in HFREF.

Methods

Ambulatory outpatients with HFREF referred by theirprimary cardiologist to the Mayo Clinic Rochester

Cardiovascular Catheterization Laboratory for resting rightheart hemodynamic catheterization for the period January1, 2002, to December 31, 2008, were studied retrospec-tively. Inclusion criteria were age >18 years, LVEF �40%,and measurable LV diastolic and mitral valve function byechocardiography-Doppler evaluation at the time of theindex catheterization. Exclusion criteria were primary pa-renchymal lung disease; chronic obstructive pulmonarydisease other than mild as defined by pulmonary spirometrytesting (forced expiratory volume in 1 s [FEV1] �80% ofexpected normal and the FEV1/forced vital capacity [FVC]ratio <70%) or clinical records description; prior valvularsurgery; infiltrative, constrictive, or hypertrophic cardiomy-opathy; myocardial infarction within 6 months; any historyof risk factors associated with group 1 or group 3 PH(idiopathic, familial or pulmonary thromboembolic disease);congenital heart disease; tachycardia-related dysrhythmia;serum creatinine level �2.5 mg/dl; history of chest radiationtherapy, collagen vascular disease, or cardiac or lung trans-plantation; consent for use of patient information for clinicalresearch not available or consent refused. Indications forcatheterization based upon referring physician (more thanone could be provided) were abstracted from the database.

Patient survival or all-cause mortality status was confirmedthrough June 30, 2010 (censor date), using Mayo Clinic,Rochester, electronic medical records, Olmsted County,Minnesota, medical record linkage system (RochesterEpidemiology Project) and Social Security mortality index.No patients hospitalized for acute decompensated HF were

Abbreviationsand Acronyms

ADHF = acute

decompensated heart failure

HFREF = heart failure of

reduced left ventricular

ejection fraction

PAC = pulmonary artery

compliance

PH = pulmonary

hypertension

Miller et al. JACC: Heart Failure Vol. 1, No. 4, 2013Pulmonary Hypertension Due to Chronic HF August 2013:290–9

292

included in the analysis. Thestudy was approved by theMayo Foundation InstitutionalResearch Review Board andincluded only those patients whoprovided informed consent asrequired by Minnesota Statute144.335/CRF 21 (Part 50).Invasive hemodynamic charac-terization. Right heart cathe-terization was performed byflow-directed pulmonary artery

catheter using hemodynamic and fluoroscopic guidance.Hemodynamic data were abstracted from the computerizedchart records and included the following: pulmonary arterialsystolic, diastolic, and mean pressures (sPAP, dPAP, mPAP,respectively); pulmonary capillary wedge pressure (PCWP);right atrial pressure (RAP); transpulmonary gradient (TPG¼mPAP � PCWP); cardiac output (CO), pulmonary vascularresistance (PVR) in Wood units (WU) ¼ TPG/CO);systemic blood pressure (systolic/diastolic), and heart rate.Cardiac output was measured by the direct Fick method (i.e.,measured oxygen consumption) or by the thermodilutionmethod if Fick method was not performed. Pulmonary ar-tery compliance (PAC) was defined as stroke volume (SV)/[sPAP � dPAP].Noninvasive hemodynamic characteristics. Parameters ofLV diastolic function including early transmitral flowvelocity (E), late transmitral flow velocity (A), early diastolicmitral annular velocity (e0), the ratio of peak E to peak e0 (E/e0)and to peak A (E/á), and the deceleration time (DT) of themitral E-wave were abstracted from the Doppler echo-cardiography evaluations obtained within 3 months of theindex catheterization. The presence and extent of mitral valveinsufficiency were quantified from the effective mitral regur-gitant orifice area (EROA) as previously described (16–18).EROA was considered to be zero in patients with no or tracemitral regurgitation by color flow imaging. Glomerularfiltration rate (eGFR [ml/min/1.73 m2]) was estimated usingthe modification of diet in renal disease equation (19).Hemodynamic definitions. The passive PH cohort wasdefined as mPAP �25 mm Hg, PCWP �15 mm Hg, andPVR <3 WU. The mixed PH cohort was defined asmPAP �25 mm Hg, PCWP �15 mm Hg, and PVR �3WU. The patient subgroup with no PH was defined asmPAP <25 mm Hg. All patients had LVEF �40%.Statistical methods. Data are presented as mean � SD,medians with interquartile ranges for continuous data, and asnumber/percentages for categorical data. To assess univariatedifferences, baseline and hemodynamic variables were com-pared using Student’s t-test or nonparametric Wilcoxsonsigned-rank test when normality assumptions were notachieved. Categorical variables were tested for significanceusing a likelihood ratio chi-square test. Linear regressionanalyses were used to assess associations between hemody-namic variables. Logistics regression analysis was performed to

identify variables associated with the presence of correlates ofPH. Results are presented as odds ratios with correspondingc-statistics (measure of predictive accuracy) and p values; theodds ratios for continuous variables are the odds of spanningthe range of the data. Survival was estimated using theKaplan-Meier method with log-rank analysis used to comparethe end point of all-cause mortality among subgroups.Mortality was evaluated in univariate andmultivariate analysesusing Cox proportional hazards models adjusting for relevantbaseline differences that would impact mortality. Risk factorsthat were significantly different among the hemodynamicgroups at baseline and factors that were clinically importantwere evaluated for these models. Cox proportional hazardsregression was used to assess the contribution to unadjustedhazards analysis for mortality based upon hemodynamicparameters using different hemodynamic variables to defineabnormality. Results are presented as forest plots with corre-sponding p values and concordance indices. The association ofrelative risk of death using hemodynamic parameters ascontinuous variables was investigated using cubic splines with3 knots within the Cox model. With this approach, we wereable to verify the proportional hazards assumption androbustly estimate the functional form of each variable with therisk of death. Analyses were done using SAS statistical soft-ware version 9.2 (SAS Institute, Cary, North Carolina) andJMP 8.

Results

Demographic and clinical characteristics. The studycohort (N¼ 463) consisted of 126 patients with no PH (27%of total cohort), 151 patients with passive PH (45% of patientswith PH, 33% of total), and 186 with mixed PH (55% ofpatients with PH, 40% of total). Patients with and withoutPH were of comparable age and sex. An ischemic origin ofHFREF was 46% in the patients with passive PH, 54%in mixed PH, and 45% in patients with no PH. The in-dications for catheterization (as selected by the referringphysician) in descending order were: heart failure/dilatedcardiomyopathy in 35%, coronary disease in 28%, hearttransplant evaluation in 21%, valvular disease in 15%, pul-monary hypertension in 7%, and other in 18%.

Compared to patients without PH, patients with PHdemonstrated slightly greater body mass index (BMI) andbody surface area, a higher prevalence of diabetes andatrial fibrillation, higher brain natriuretic peptide (BNP)levels, slightly lower hemoglobin levels, and slightly worserenal function. Medications were comparable across the 3subgroups, with the exception of more diabetic, nitrate,diuretic, and anticoagulation therapies in the PH patients.Patients with passive and mixed PH were distinguishableonly by slightly higher body mass, higher BNP level, andslightly worse renal function in mixed PH (Table 1).Invasive hemodynamic characteristics. Both sPAP andPCWP were highly correlated with mPAP (Fig. 1A). UsingmPAP as the dependent and sPAP as the independent

Table 1 Demographic and Clinical Characteristics PH defined by PVR </�3 WU

VariablePassive PH(n ¼ 151)

Mixed PH(n ¼ 186) p Value*

No PH(n ¼ 126) p Valuey

Age (yrs) 57 � 13 60 � 15 0.054 57 � 14 0.070

Males 79% 72% 0.161 69% 0.243

BMI (kg/m2) 29.9 � 6.7 28.4 � 6.1 0.033 27.9 � 5.4 0.043

BSA (m2) 2.06 � 0.26 1.96 � 0.24 <0.001 1.94 � 0.25 0.001

Comorbidities

Hypertension 63% 68% 0.299 55% 0.063

Diabetes 41% 40% 0.911 37% 0.025

Hyperlipidemia 59% 64% 0.265 76% 0.345

CAD 47% 51% 0.584 43% 0.300

Prior MI 30% 32% 0.723 26% 0.431

Atrial fibrillation 45% 46% 0.947 37% 0.002

Obesity (BMI >30 kg/m2) 36% 38% 0.619 34% 0.674

Hemoglobin (g/dl) 12.5 � 2.0 12.8 � 2.0 0.173 13.2 � 1.8 0.058

Creatinine (mg/dl) 1.6 � 1.1 1.6 � 1.2 1.00 1.4 � 1.1 0.134

GFR (ml/min/1.73 m2) 54 � 22 53 � 20 0.663 59 � 19 0.047

Sodium (mEq/l) 138 � 4.2 138 � 4.1 0.998 138 � 3.4 0.589

Potassium (mEq/l) 4.3 � 0.5 4.3 � 0.6 1.00 4.3 � 0.5 0.998

BNP (pg/ml) 674 (249, 1,280) 1,073 (491, 1,670) 0.003 295 (131, 679) <0.001

Medications

Beta blocker 92% 91% 0.642 95% 0.318

ARB 30% 31% 0.832 22% 0.224

ACEI 85% 80% 0.211 74% 0.071

Diuretic 94% 97% 0.289 86% 0.011

CCB 24% 27% 0.546 22% 0.567

Digoxin 68% 69% 0.967 57% 0.089

Insulin 28% 34% 0.295 21% 0.022

Oral diabetic medication 22% 26% 0.444 15% 0.036

Antiplatelet 74% 70% 0.343 65% 0.240

Nitrate 26% 29% 0.714 15% 0.004

Statin 71% 72% 0.843 65% 0.322

Warfarin 67% 66% 0.908 55% 0.014

Values are mean � SD, % of cohort, or median (25th, 75th percentile). Mixed PH was defined as mPAP �25 mm Hg; PCWP �15 mm Hg; and PVR �3 WU. Post-capillary passive PH was defined asmPAP �25 mm Hg; PCWP �15 mm Hg; and PVR <3 WU. *Mixed PH compared with passive PH. yNo PH compared with passive PH-Mixed PH.ACEI ¼ angiotensin converting enzyme inhibitor; ARB ¼ angiotensin receptor blocker; BMI ¼ body mass index; BNP ¼ brain natriuretic peptide; BSA ¼ body surface area; CAD ¼ coronary artery disease;

CCB ¼ calcium channel blocker; GFR ¼ glomerular filtration rate; MI ¼ myocardial infarction; mPAP ¼ mean pulmonary artery pressure; PCWP ¼ pulmonary capillary wedge pressure; PH ¼ pulmonaryhypertension; PVR ¼ pulmonary vascular resistance; WU ¼ Wood units.

JACC: Heart Failure Vol. 1, No. 4, 2013 Miller et al.August 2013:290–9 Pulmonary Hypertension Due to Chronic HF

293

variable, the regression equation was mPAP ¼ 0.615,sPAP þ 3 (r ¼ 0.93, p < 0.001). Lesser, but still significantcorrelations, were also demonstrated for mPAP with PVR(r ¼ 0.669, p < 0.001), and TPG (r ¼ 0.696, p < 0.001).There was a much weaker correlation between sPAP andsystemic arterial pressure (Fig. 1B).

As expected, PA pressures, PVR, and TPG were higher,while PAC was lower in patients with PH than in thosewithout PH (Table 2). Other hemodynamic markers ofincreasing HF severity, including HR, RAP, and PCWP,were higher, while CO and SV were lower in patients withPH than in patients without PH.

Compared to patients with passive PH, those with mixedPH had higher PA pressures, PVR, and TPG, with lowerPAC, CO, and SV (Table 2). However, PCWP and RAPwere similar in both subgroups, indicating that the presenceof pulmonary vascular disease was the principle determinantof mixed PH rather than a difference in downstream leftheart filling pressures.

Noninvasive hemodynamic characteristics. Compared toHF patients without PH, patients with PH had similar LVsize and mass, slightly lower LVEF, more profound diastolicdysfunction (higher E/A and E/e0 ratio, shorter DT, largerleft atrial volume), more severe mitral valve regurgitation,and more right ventricular dysfunction (Table 2). Comparedto patients with passive PH, patients with mixed PH hadmore severe LV diastolic dysfunction and more severe mitralvalve regurgitation (higher EROA) and tended to displaymore severe RV dysfunction.Correlates of pulmonary hypertension in HFREF. Clini-cal and hemodynamic variables associated with PH (passiveor mixed) by unadjusted univariate and multivariateanalyses are shown in Table 3. In univariate analysis, dia-betes, atrial fibrillation, diuretic use, hemoglobin, renalfunction, LVEF, and PAC were predictors of the presenceof PH. Age was a borderline univariate predictor but wasincluded in the multivariate analysis because of its commonassociation with PH. After multivariate adjustment, age,

Figure 1 Correlation of Hemodynamic Parameters

Correlation of mean pulmonary artery pressure (mPAP) with systolic pulmonary artery pressure (sPAP) and PCWP (A) and systemic blood pressure with pulmonary capillary wedge

pressure (sPAP) (B) for the cohort as a whole.

Miller et al. JACC: Heart Failure Vol. 1, No. 4, 2013Pulmonary Hypertension Due to Chronic HF August 2013:290–9

294

diuretic use, atrial fibrillation, and PAC remained indepen-dent predictors of PH. After multivariate adjustment, age,atrial fibrillation, and PAC remained predictors of mixedPH as compared to passive PH.Relationship of PH subtypes to mortality. There were171 deaths (37%) over a median follow-up of 2.1 (95%confidence interval [CI]: 0.1 to 3.8) years. Using PVR todefine the PH subgroups, we found a gradient of increasingrisk of death from patients with no PH to those with passivePH and to those with mixed PH (Fig. 2A). However, whenpassive and mixed PH were defined according to TPG(passive, TPG <12 mm Hg; mixed, TPG �12 mm Hg),there were no differences in mortality between mixed andpassive PH (Fig. 2B).

Table 4 shows univariate and multivariate clinical andhemodynamic predictors of mortality for the patient cohort.Both passive and mixed PH, defined using PVR, werepredictors of death in addition to the previously describedclinical variables. PAC was also associated with increasedrisk of death. In multivariate models including CO andmPAP, mixed PH remained an independent predictor ofmortality, along with age, renal function, and hemoglobin.Hemodynamic predictors of mortality. Increasing mPAPwas associated with greater mortality risk than compared withno PH, although there were no statistically significant differ-ences from one cutoff point to the other (Fig. 3). Similarly,pulmonary venous hypertension was associated with increasedmortality with no differences between mild-moderate or severeelevations (PCWP �15 mm Hg and �22 mm Hg, re-spectively). A TPG �12 mm Hg also is often used to definemixed PH, but a partition value of 18 mm Hg provided morerobust identification of risk relative to TPG <12 mm Hg(HR: 1.81, 95% CI: 1.23 to 2.65, p ¼ 0.002 vs. HR: 1.22,95% CI: 0.06 to 1.73, p ¼ 0.257, respectively).

PAC, which has been shown to predict mortality in group1 PH, demonstrated a separation of mortality risk inHFREF (Fig. 3). A PAC �2.0 ml/mm Hg conveyed a 3.5-fold increased risk (95% CL: 1.74, 7.35, p < 0.0001) relativeto PAC values >5.0 ml/mm Hg. Using the median PACvalue (2.3 ml/mm Hg) to define subgroups (i.e., mPAP�25 mm Hg, PCWP �15 mm Hg, and PAC �/>2.3)revealed PAC values less than the median identified thesubgroup of PH patients with highest risk (HR: 2.28, 95%CL: 1.64, 3.15, p < 0.001) (Fig. 4). PVR �3.5 WUidentified greater risk of death (HR: 1.97, 95% CI: 1.42 to2.74, p < 0.0001) (Fig. 3), while partition values <3.5 WUwere statistically less predictive.

Figure 5 shows the hazard ratio (95% CL) of death asa function of PVR (Fig. 5A), sPAP (Fig. 5B), PCWP(Fig. 5C), and PAC (Fig. 5D) as continuous variables ofrisk. The inflection points of these curves suggest that PVR>w4.0 WU, sPAP >w35 mm Hg, PCWP >w25 mmHg, and PAC <w2.0 ml/mm Hg might identify patients athighest risk for all-cause mortality. Hazard ratios crossingunity signifying threshold increases in risk were demon-strated for pulmonary vascular resistance at 1.5 WU, systolicpulmonary artery pressure at 25 mm Hg, pulmonary wedgepressure at 14 mm Hg, and PAC at 2.5 ml/mm Hg.

Discussion

The findings of this study address important knowledgegaps regarding the prevalence, definition, and prognosticsignificance of PH subgroups in HFREF (20–22) bypresenting a comprehensive description of clinical, hemo-dynamic, and echocardiographic features in a large, well-described cohort of patients with HFREF. We found thatapproximately 75% of the catheterization laboratory referral

Table 2 Invasive and Noninvasive Hemodynamic Features PH Defined by PVR < or � 3 WU

VariablePassive PH(n ¼ 151)

Mixed PH(n ¼ 186) p Value*

No PH(n ¼ 126) p Valuey

Invasive hemodynamics

Systemic blood pressure(S/D, mm Hg)

110 � 22/64 � 12 115 � 24/64 � 12 0.051/0.999 110 � 22/62 � 13 0.338/0.294

Heart rate (beats/min) 78 � 18 77 � 16 0.591 73 � 17 0.035

Mean PAP 34 � 6 42 � 8 <0.001 19 � 4 <0.001

PASP (mm Hg) 48 � 9 62 � 14 <0.001 30 � 6 <0.001

PCWP (mm Hg) 24.0 � 5.9 24.5 � 5.7 0.432 11.2 � 4.1 <0.001

TPG (mm Hg) 9.4 � 3.7 17.3 � 6.1 <0.001 8.1 � 3.2 0.001

mRAP (mm Hg) 14.6 � 7.2 15.0 � 6.7 0.181 6.6 � 3.8 <0.001

PVR (Wood units) 2.0 � 0.7 5.0 � 1.9 <0.001 1.8 � 0.8 0.011

PAC (ml/mm Hg) 2.8 � 1.0 1.6 � 0.7 <0.001 4.4 � 2.1 <0.001

SVR (Wood units) 18 � 5 24 � 9 <0.001 17 � 5 <0.001

CO (l/min) 4.8 � 1.2 3.7 � 1.1 0.001 4.9 � 1.4 0.001

Stroke volume (ml) 64 � 22 50 � 17 0.001 72 � 25 <0.001

Echocardiography

LVEDD (mm) 65 � 11 65 � 12 1.00 65 � 12 1.00

LV mass index (g/m2) 74 � 28 78 � 29 0.139 75 � 21 0.239

LVEF (%) 24 � 9 23 � 8 0.989 26 � 8 0.011

LA volume index (ml/m2) 56 � 26 60 � 22 0.151 42 � 17 <0.001

E/A ratio 2.3 � 1.3 2.7 � 1.2 0.003 1.4 � 1.1 <0.001

E/e0 ratio 22.1 � 10.0 26.0 � 11.0 0.001 15.8 � 7.9 <0.001

MV DT (ms) 148 � 38 146 � 64 0.707 190 � 52 <0.001

MV ERO area (mm2) 28 � 15 32 � 16 0.023 23 � 12 0.009

Qualitative mitralregurgitation severity

Moderate, 36% Severe, 15% Moderate, 37% Severe, 28% 0.091 Moderate, 21% Severe, 2% 0.007

RVSP (mm Hg) 48 � 12 58 � 16 <0.001 35 � 10 <0.001

Qualitative RVsystolic dysfunction

Moderate, 35% Severe, 12% Moderate, 25% Severe, 21% 0.064 Moderate, 13% Severe, 3% 0.001

Values are mean � SD or % of cohort. Mixed PH was defined as mPAP �25 mm Hg; PCWP �15 mm Hg; and PVR � 3 WU. Post-capillary passive PH was defined as mPAP �25 mm Hg; PCWP �15 mm Hg;and PVR 3 WU. *Mixed PH compared with passive PH. yNo PH compared with passive PH-mixed PH.CO ¼ cardiac output; LA ¼ left atrial; LV ¼ left ventricular; LVEDD ¼ left ventricular end-diastolic dimension; LVEF ¼ left ventricular ejection fraction; mPAP ¼ mean pulmonary artery pressure; MV DT ¼

mitral valve deceleration time; MV ERO ¼ mitral valve effective regurgitant orifice; PAC ¼ pulmonary artery compliance (stroke volume/pulmonary artery pulse pressure); PAP ¼ pulmonary artery pressure;PASP ¼ pulmonary artery systolic pressure; RVSP ¼ right ventricular systolic pressure; S/D ¼ systolic/diastolic; SVR ¼ systemic vascular resistance; TPG ¼ transpulmonary gradient; other abbreviations as inTable 1.

JACC: Heart Failure Vol. 1, No. 4, 2013 Miller et al.August 2013:290–9 Pulmonary Hypertension Due to Chronic HF

295

population meeting study criteria displayed PH and thatamong these patients over half showed evidence of pul-monary vascular disease (mixed PH) superimposed onelevations in left heart filling pressures. Patients with anyPH and particularly those with mixed PH displayedprogressively more severe hemodynamic derangements withgreater burden of diastolic dysfunction, mitral regurgitation,

Table 3 Variables Associated With the Presence

VariableUnivariate Odds Ratio

(95% CL) p Value

Age 1.12 (0.97–1.28) 0.098

Diabetes 1.67 (1.08–2.60) 0.020

Atrial fibrillation 2.22 (1.47–3.39) 0.001

Diuretic use 2.55 (1.28–4.99) 0.008

Hemoglobin 2.03 (1.16–3.57) 0.012

eGFR 1.10 (1.02–1.21) 0.035

LVEF 1.24 (1.09–1.41) 0.001

PAC 8.58 (5.64–13.70) <0.001

*Model was adjusted for all univariate variables. C-statistics for adjusted muCL ¼ confidence limits; eGFR ¼ estimated glomerular filtration rate; othe

and right ventricular dysfunction despite grossly similarclinical profiles. Sensitivity analyses were performed toevaluate partition values of hemodynamic parameters thatidentified greater risk of death, and we further show thatPAC refines risk assessment in group 2 PH and may bea novel therapeutic target in addition to PVR and mPAP.These data reinforce the important role of PH in the

of PH

C StatMultivariate* Odds Ratio

(95% CL) p Value

0.540 1.40 (1.08–1.82) 0.011

0.568 2.02 (1.04–4.07) 0.037

0.598 2.85 (1.48–5.62) 0.002

0.539 1.49 (0.48–4.46) 0.486

0.584 1.59 (0.68–3.73) 0.283

0.595 1.01 (0.87–1.18) 0.885

0.598 1.13 (0.92–1.39) 0.258

0.868 8.98 (5.47–15.73) <0.001

ltivariate model ¼ 0.812.r abbreviations as in Tables 1 and 2.

Figure 2 Survival of HFREF Patients With PH

(A) Kaplan-Meier estimates of survival in patients with heart failure of reduced left ventricular ejection fraction (HFREF) relative to passive pulmonary hypertension (PH) and

mixed PH as defined by pulmonary vascular resistance (PVR) < or �3.0 Wood units (WU) and no PH. (B) Kaplan-Meier estimates of survival in patients with HFREF relative to

passive PH and mixed PH as defined by transpulmonary gradient (TPG) < and �12 mm Hg and no PH.

Miller et al. JACC: Heart Failure Vol. 1, No. 4, 2013Pulmonary Hypertension Due to Chronic HF August 2013:290–9

296

pathophysiology and progression of HF with reduced LVEFand support the need for trials of agents targeting passiveand mixed PH in HF.

Numerous studies have shown that PH carries a greaterrisk of death in HF than LV dysfunction alone, but therehave been few studies in HFREF patients that have evalu-ated the prognostic implications of the different subtypes ofPH (11,15). These studies were performed in hospitalizedpatients with acute decompensated HF (ADHF), withheterogeneous clinical and PH origin status, relatively short-term follow-up (6 months), and conflicting results. Usingdata from the ESCAPE (Evaluation Study of CongestiveHeart Failure and Pulmonary Artery CatheterizationEffectiveness) trial, Khush et al. (11) found no differencesin clinical outcomes between patients with passive andthose with mixed PH and also no differences in mortalitycompared with patients without PH. In contrast, using datafrom the VMAC (Vasodilation in the Management ofAcute Congestive Heart Failure) trial, Aronson et al. (12)demonstrated a significant increased risk of death

Table 4 Unadjusted and Adjusted Cox Proportion

VariableUnadjusted Hazard R

(95% CL)

Mixed PH defined as PVR �3.0 WU 3.06 (1.98–4.89)

Passive PH defined as PVR <3.0 WU 1.97 (1.27–3.14)

Age 1.43 (1.28–1.61)

Atrial fibrillation 1.36 (1.02–1.84)

eGFR 1.24 (1.15–1.34)

Hemoglobin 2.77 (1.82–4.23)

PA compliance 1.85 (1.43–2.45)

PA ¼ pulmonary artery; other abbreviations as in Tables 1 and 3.

advancing from no PH to passive and mixed PH, similar toour current findings. A key difference between these studieswas in the timing of the assessment of PH subgroups: priorto treatment of ADHF (11) and after treatment (15). Bycomparison, our data reflect the largest and most compre-hensively phenotyped cohort of outpatients with chronicHFREF where causes of PH other than LV systolicdysfunction were excluded, and baseline right heart hemo-dynamic catheterization data were collected under stabletherapeutic conditions and with long follow-up.

The partition values currently used to identify “abnormal”hemodynamics have been based largely on consensusopinion, and few studies have provided objective sensitivityanalyses to identify which cutoff values best stratify risk. Theadoption of empirically derived partition values may betteralign the magnitude of abnormalities with objective risk.Our sensitivity analyses suggest that cutoff points of PVRof �3.5 WU and PAC of �2.0 ml/mm Hg (unadjusted) aremost strongly associated with risk separation, although riskof death was suggested to be increased above unity, even at

al Hazards Model of All-Cause Mortality

atiop Value

Adjusted Hazard Ratio(95% CL) p Value

<0.001 2.31 (1.24–4.52) 0.008

0.002 1.48 (0.82–2.78) 0.190

<0.001 1.31 (1.14–1.51) <0.001

0.038 1.06 (0.72–1.51) 0.773

<0.001 1.16 (1.06–1.28) 0.002

<0.001 1.94 (1.21–3.14) 0.006

<0.001 1.11 (0.78–1.62) 0.586

Figure 3 Forest Plot Predictors of Mortality

PH hemodynamic parameter cut-points and their relative contributions to the

prediction of all- cause mortality. HR relative to reference values is denoted by an

asterisk (*mPAP<25 mm Hg; PCWP<15 mmHg; TPG<12 mm Hg; PVR<2.5 WU;

and PA compliance >5.0 ml/ mm Hg). Abbreviations as in Figures 1 and 2.

JACC: Heart Failure Vol. 1, No. 4, 2013 Miller et al.August 2013:290–9 Pulmonary Hypertension Due to Chronic HF

297

PVR values less than 2.5 WU (Fig. 5). If these risk markersare validated in other HFREF populations, they may servevaluable roles in both risk stratification and potentially asmodifiable endpoints in treatment trials.

We found that PH and particularly mixed PH were quitecommon in this referral population. While the cause ofmixed PH cannot be determined from this analysis (i.e.,structural vs. functional changes), the current and previous-ly published data (23,24) suggest that it may relate to in-creasing severity and chronicity of the underlying HFREF,as there was progressively more severe diastolic

Figure 4 Survival of HFREF Patients With PH

Kaplan-Meier estimates of survival in patients with HFREF relative to passive and

mixed PH as defined by pulmonary artery compliance (PAC) � and >2.3 ml/mm Hg

and no PH. Abbreviations as in Figure 2.

dysfunction, mitral regurgitation, and right ventriculardysfunction identified progressing from no PH to passiveand mixed PH. It is also notable that these groups werenearly indistinguishable on clinical grounds, with similarchamber remodeling, right- and left-sided filling pressures,and LVEF. The concept that mixed PH is associated withgreater anatomic pulmonary artery remodeling is supportedby the above-cited small study in transplant patients (23),where relative medial thickness was greater in the group withmixed PH than in those with passive PH. Future studiescorrelating clinicopathologic findings with pulmonaryhemodynamics might offer further insight into the mecha-nism of disease.

Prior studies in healthy controls and in patients withgroup 1 PH have shown that PA systolic, mean, anddiastolic pressures are very highly correlated, and weshow that this relationship also extends to group 2 PH.Interestingly, the linear least squares regression equationrelating mPAP to sPAP in the current sample was re-markably similar to that reported by Chemla et al. (25). Apractical implication of these findings is that an assessmentof sPAP is as valid as mPAP for identifying PH, withsPAP values of �35 mm Hg corresponding to mPAPof �25 mm Hg, based upon current and prior data. OftensPAP is reported in the context of systemic blood pressure,implying that there is some mechanistic relationshipbetween the 2 parameters. However, the poor correlationobserved between sPAP and systolic systemic BP in thecurrent study (Fig. 1B) suggests that this is not a validconclusion.

Previous studies have variably used PVR and TPG inthe nomenclature of PH. An advantage of PVR is that inaddition to assessing the hydraulic pressure drop across thepulmonary vascular bed, it accounts for flow (CO), whichvaries directly with TPG. The current findings support theuse of PVR rather than TPG to identify mixed PH.Indeed, outcomes were similar between mixed and passivePH when the subgroups were defined using TPG values</�12 mm Hg. However, as shown in our sensitivityanalyses, defining mixed PH by PVR �3.5 WU orPAC �2.0 ml/ mm Hg better identified patients at higherrisk who might benefit from more aggressive or noveltherapies. Intriguingly, these objectively defined partitionsare in line with prior publications (26,27). The fact that themixed PH phenotype remained a significant predictor ofmortality even after adjusting for CO and mPAP furthersupports the concept that mixed PH represents a moremalignant phenotype which may respond differently totherapies. These issues require further study.Study limitations. In interpreting these data, several issuesshould be considered. One issue is the retrospective design ofthe study with its associated inherit limitations includingpossible referral bias. Additionally, exercise and vasodilatorhemodynamic data which might provide further informationon the prevalence and prognostic significance of the subgroupsof PH are not available for this sample. Prior studies have

Figure 5 All-Cause Mortality and Hemodynamic Parameters

(A) Unadjusted risk (hazard ratio [HR]: 95% confidence limits [CL]) of all-cause mortality for sPAP analyzed as continuous variables. (B) Unadjusted risk (HR: 95% CL) of all-cause

mortality for PVR analyzed as continuous variables. (C) Unadjusted risk (HR: 95% CL) of all-cause mortality for PAC analyzed as continuous variables. (D) Unadjusted risk (HR:

95% CL) of all-cause mortality for PCWP analyzed as continuous variables. PA ¼ pulmonary artery; PASP ¼ pulmonary artery systolic pressure; PCWP ¼ pulmonary capillary

wedge pressure; PVR ¼ pulmonary vascular resistance.

Miller et al. JACC: Heart Failure Vol. 1, No. 4, 2013Pulmonary Hypertension Due to Chronic HF August 2013:290–9

298

shown that the RV response to PH is as important or moreimportant than the extent of pulmonary vascular diseasepresent (13). While we provide qualitative assessment of RVfunction from echocardiographic evaluations, more quantita-tive assessment might further refine risk stratification inaddition to the hemodynamic parameters. Cardiopulmonaryexercise testing is a validated prognostic tool in HFREF, andthis study does not provide data regarding exercise capacity.Further studies arewarranted to explore potential relationshipsbetween pulmonary hemodynamics and exercise physiology.The cubic spline analyses (Fig. 5) were unadjusted, andcomparisons were made only with the reference levels.

Conclusions

Group 2 PH is commonly identified in patients with chronicstable HFREF and carries a significantly increased mortalityrisk. The development of mixed PH with structural and/or

functional pulmonary vascular disease is related to othermarkers of increased HF severity and chronicity, and carriesincrementally greater risk of death. While several parameterscan be used to distinguish the subgroups of PH, PVR andPAC appear most robust in separating patients at higher riskand may serve as novel targets for therapy.

Reprint requests and correspondence: Dr. Wayne L. Miller,Cardiovascular Division, Mayo Clinic, 200 First Street SW, GondaBuilding 5S 130, Rochester, Minnesota 55905. E-mail: miller.wayne@mayo.edu.

REFERENCES

1. Guazzi M, Borlaug BA. Pulmonary hypertension due to left heartdisease. Circulation 2012;126:975–90.

JACC: Heart Failure Vol. 1, No. 4, 2013 Miller et al.August 2013:290–9 Pulmonary Hypertension Due to Chronic HF

299

2. Guglin M, Khan H. Pulmonary hypertension in heart failure. J CardiacFail 2010;16:461–74.

3. Butler J, Chomsky DB, Wilson JR. Pulmonary hypertension andexercise intolerance in patients with heart failure. J Am Coll Cardiol1999;34:1802–6.

4. Fang JC, DeMarco T, Givertz MM, et al., for World Health Organi-zation Pulmonary HypertensionGroup 2. Pulmonary hypertension due toleft heart disease in the adult. J Heart Lung Transplant 2012;31:913–33.

5. Miller WL, Mahoney DW, Michelena HI, Pislaru SV, Topilsky Y,Enriquez–Sarano M. Contribution of ventricular diastolic dysfunctionto pulmonary hypertension complicating chronic systolic heart failure.J Am Coll Cardiol Img 2011;4:946–54.

6. Hoeper MM, Barbera JA, Channick RN, et al. Diagnosis, assessment,and treatment of non-pulmonary arterial hypertension. Pulmonaryhypertension. J Am Coll Cardiol 2009;54:S85–96.

7. Zile MR, Bennett TD, St John Sutton M, et al. Transition fromchronic compensated to acute decompensated heart failure: patho-physiological insights obtained from continuous monitoring of intra-cardiac pressures. Circulation 2008;118:1433–41.

8. Kjaergaard J, Akkan D, Karmark Iverson K, et al. Prognostic impor-tance of pulmonary hypertension in patients with heart failure. Am JCardiol 2007;99:1146–50.

9. Grigioni F, Potena L, Galie N, et al. Prognostic implications of serialassessments of pulmonary hypertension in severe chronic heart failure.J Heart Lung Transplant 2006;25:1241–6.

10. Cappola TP, Felker GM, Kao WHL, Hare JM, Baughman KL,Kasper EK. Pulmonary hypertension and risk of death in cardiomy-opathy: patients with myocarditis are at higher risk. Circulation 2002;105:1663–8.

11. Khush KK, Taissa G, Butler J, McGlothin D, DeMarco T. Effect ofpulmonary hypertension on clinical outcomes in advanced heart failure:Analysis of the Evaluation Study of Congestive Heart Failure andPulmonary Artery Catheterization Effectiveness (ESCAPE) database.Am Heart J 2009;157:1026–34.

12. Aronson D, Eitan A, Dragu R, Burger AJ. Relationship betweenreactive pulmonary hypertension and mortality in patients with acutedecompensated heart failure. Circ Heart Fail 2011;4:644–50.

13. Tumminello G, Lancellotti P, Lempereur M, D’Orio V, Pierard LA.Determinants of pulmonary artery hypertension at rest and duringexercise in patients with heart failure. Eur Heart J 2007;28:569–74.

14. Mancini D, Katz S, Donchez L, Aaronson K. Coupling of hemody-namic measurements with oxygen consumption during exercise doesnot improve risk stratification in patients with heart failure. Circulation1996;94:2492–6.

15. Ghio S, Gavazzi A, Campana C, et al. Independent and additiveprognostic value of right ventricular systolic function and pulmonary

artery pressure in patients with chronic heart failure. J Am Coll Cardiol2001;37:183–8.

16. Artega RB, Hreybe H, Patel D, Landolfo C. Derivation and validationof a diagnostic model for the evaluation of left ventricular fillingpressures and diastolic function using mitral annulus tissue Dopplerimaging. Am Heart J 2008;155:924–9.

17. Zoghbi WA, Enriquez–Sarano M, Foster E, et al. Recommendationsfor evaluation of the severity of native valvular regurgitation with two-dimensional and Doppler echocardiography. J Am Soc Echocardiogr2003;16:777–802.

18. Enriquez-Sarano M, Rossi A, Seward JB, Bailey KR, Tajik AJ.Determinants of pulmonary hypertension in left ventricular dysfunc-tion. J Am Coll Cardiol 1997;29:153–9.

19. Levey AS, Coresh J, Green T, et al., for the chronic kidney diseaseepidemiology collaboration. Using standardized serum creatinine valuesin the modification of diet in renal disease study equation for estimatingglomerular filtration rate. Ann Intern Med 2006;145:247–54.

20. McGoon MD, Kane GC. Pulmonary hypertension: diagnosis andmanagement. Mayo Clin Proc 2009;84:191–207.

21. McLaughlin VV, Archer SL, Badesch DB, et al. ACC/AHA 2009Expert Consensus document on pulmonary hypertension: a report of theACC Foundation Task Force. J Am Coll Cardiol 2009;53:1573–619.

22. McGoon M, Gutterman D, Steen V, et al. Screening, early detection,and diagnosis of pulmonary arterial hypertension: ACCP evidence-based clinical practice guidelines. Chest 2004;126:14S–34S.

23. Delgado JF, Conde E, Sanchez V, et al. Pulmonary vascular remod-eling in pulmonary hypertension due to chronic heart failure. Eur JHeart Fail 2005;7:1011–6.

24. Moraes DL, Colucci ES, Givertz MM. Secondary pulmonary hyper-tension in chronic heart failure: the role of the endothelium in path-ophysiology and management. Circulation 2000;102:1718–23.

25. Chemla D, Castelain V, Provencher S, Humbert M, Simonneau G,Herve P. Evaluation of various empirical formulas for estimating meanpulmonary artery pressure using systolic pulmonary artery pressure inadults. Chest 2009;135:760–8.

26. Mehra MR, Kobashigawa J, Starling R, et al. Listing criteria for hearttransplantation: International Society for Heart and Lung Trans-plantation Guidelines for the care of cardiac transplant candidates –

2006. J Heart Lung Transplant 2006;25:1024–42.27. Miller LW. Listing criteria for solid organ transplantation. Trans-

plantation 1998;66:946–7.

Key Words: heart failure - mixed pulmonary hypertension - outcomes -

passive pulmonary hypertension - risk prediction.