Article Analysis

heartdiseaseexperimentalresearch.pdf

Home >Nursing homework help >Article Analysis

European Journal of Heart Failure (2017) 19, 88 – 97 RESEARCH ARTICLE doi:10.1002/ejhf.675

Determinants and prognostic implications of the negative diastolic pulmonary pressure gradient in patients with pulmonary hypertension due to left heart disease Anikó Ilona Nagy1*, Ashwin Venkateshvaran2,3, Béla Merkely1, Lars H. Lund4,5, and Aristomenis Manouras4,5

1Heart and Vascular Center, Semmelweis University, Budapest, Hungary; 2School for Technology and Health, Royal Institute of Technology, Stockholm, Sweden; 3Sri Sathya Sai Institute of Higher Medical Sciences, Bangalore, India; 4Department of Cardiology, Karolinska University Hospital, Stockholm, Sweden; and 5Department of Medicine, Karolinska Institutet, Stockholm, Sweden

Received 19 March 2016; revised 14 August 2016; accepted 8 September 2016 ; online publish-ahead-of-print 17 October 2016

Aims The diastolic pulmonary pressure gradient (DPG) has recently been introduced as a specific marker of combined pre-capillary pulmonary hypertension (Cpc-PH) in left heart disease (LHD). However, its diagnostic and prognostic superiority compared with traditional haemodynamic indices has been challenged lately. Current recommendations explicitly denote that in the normal heart, DPG values are greater than zero, with DPG ≥7 mmHg indicating Cpc-PH. However, clinicians are perplexed by the frequent observation of DPG <0 mmHg (DPGNEG), as its physiological explanation and clinical impact are unclear to date. We hypothesized that large V-waves in the pulmonary artery wedge pressure (PAWP) curve yielding asymmetric pressure transmission might account for DPGNEG and undertook this study to clarify the physiological and prognostic implications of DPGNEG.

.....................................................................................................................................................................

Methods and results

Right heart catheterization and echocardiography were performed in 316 patients with LHD due to primary myocardial dysfunction or valvular disease. A total of 256 patients had PH-LHD, of whom 48% demonstrated DPGNEG. The V-wave amplitude inversely correlated with DPG (r = −0.45, P < 0.001) in patients with low pulmonary vascular resistance (PVR), but not in those with elevated PVR (P > 0.05). Patients with large V-waves had negative and lower DPG than those without augmented V-waves (P < 0.001) despite similar PVR (P >0.05). Positive, but normal DPG (0 – 6 mmHg) carried a worse 2-year prognosis for death and/or heart transplantation than DPGNEG (hazard ratio 2.97; P < 0.05).

.....................................................................................................................................................................

Conclusion Our results advocate against DPGNEG constituting a measurement error. We propose that DPGNEG can partially be ascribed to large V-waves and carries a better prognosis than DPG within the normal positive range.

..........................................................................................................

Keywords Diastolic pressure gradient • Pulmonary hypertension • V-wave

Introduction Pulmonary hypertension (PH) is a common complication of left heart disease (LHD). In isolated post-capillary PH, the pul- monary arterial pressure (PAP) elevation is governed solely by the upstream-transmitted left atrial pressure (LAP). Long-standing

*Corresponding author. Semmelweis University, Heart and Vascular Center, 68 Városmajor utca, Budapest, H-1122, Hungary. Tel: +36 20 8259738, Fax: +36 1 4586818, Email: [email protected]

.. ..

.. .. post-capillary PH may, however, lead to pathological alterations of

the pre-capillary vasculature, contributing to further PAP increase, a state denoted as combined post- and pre-capillary PH (Cpc-PH). Although this latter condition is clearly associated with worse prognosis,1,2 the optimal method to distinguish these two cohorts haemodynamically remains controversial.

Negative DPG in pulmonary hypertension 89

Traditionally, pulmonary vascular resistance (PVR) and transpul- monary gradient (TPG) have been employed for discerning Cpc-PH, both metrics bearing an established prognostic value in PH due to LHD (PH-LHD).3,4 However, as both these markers are influenced by LAP and stroke volume,5 their specificity has been questioned. In recent times, the diastolic pulmonary pres- sure gradient (DPG), considered less affected by heart failure (HF)-induced haemodynamic changes,5 has been introduced as a more reliable Cpc-PH index. Based on the above rationale and study results demonstrating prognostic superiority of the DPG,6,7

the Fifth World Symposium on PH proposed that a DPG ≥7 mmHg alone should define Cpc-PH.5 However, the failure of two recent large-scale studies to confirm the prognostic value of DPG8,9 raised concerns regarding its use in PH-LHD.8,10 Despite the significant prevalence of negative DPG values (DPGNEG), reportedly varying between 10% and 50%,8,11 the physiological background and the potential prognostic implications of DPGNEG have not yet been investigated; rather, DPGNEG has arbitrarily been considered to represent a measurement error.12 We hypothesized that prominent V-waves in the pulmonary artery wedge pressure (PAWP) recordings might account for the DPGNEG by causing ‘asymmetrical’ pressure transmission through the pulmonary capillaries, i.e. a backward LAP wave reflection characterized by disproportionate phasic pressure changes. We therefore under- took the present study in order to (i) investigate the impact of V-waves on the DPG and particularly on the occurrence of DPGNEG; (ii) elucidate the influence of PAWP as compared with direct LAP measurements on the DPG; and (iii) assess the prognostic significance of DPGNEG compared with positive but normal DPG.

Methods Study population The study population consisted of 316 patients. A total of 192 patients were enrolled prospectively; 86 consecutive patients with PH due to heart failure (HF) (denoted as PH-LHD in the following) referred for right heart catheterization (RHC) for HF assessment between January and December 2014 were enrolled prospectively at Karolinska Uni- versity Hospital, while 106 consecutive patients with severe rheumatic mitral valve stenosis (denoted as MS in the following) referred for per- cutaneous transvenous mitral commissurotomy (PTMC) between Jan- uary and June 2012 were enrolled again prospectively at the Sri Sathya Sai Institute (Bangalore, India). In addition, 124 consecutive patients with PH-LHD referred for RHC at the Karolinska University Hospital were studied retrospectively. In all PH-LHD cases, medical treatment had been titrated and haemodynamic stabilization achieved at the time of examination. None of the patients included in the study presented with acute coronary syndrome or had undergone cardiac surgery within 1 year before enrolment. In the case of the MS cohort, sub- jects with >1 grade mitral regurgitation, aortic valve disease, ischaemic heart disease, AF, or hypertension were not included in the study. In the PH-LHD cohort, no specific exclusion criteria were applied, apart from the fact that patients with pressure tracings of inadequate quality (i.e. that would not have allowed reliable and reproducible identification of waveforms) were not included. A flowchart describ- ing patient enrolment and haemodynamic grouping is provided in the .

.. ..

.. . Supplementary material online, Figure S1. Follow-up data were col-

lected from the Karolinska University Hospital database that is updated centrally; patients were followed until death, cardiac transplantation, or the end of the study period (mean time: 15.6 months). The prog- nostic value of DPGNEG vs. positive but normal DPG was assessed. The study was approved by the local ethics committee (registration number 2013/1991-32). All prospectively enrolled subjects provided written informed consent. All subjects underwent transthoracic echocardiog- raphy and RHC.

Catheterization Right heart catheterization was performed using a 6 F balloon-tipped fluid-filled Swan – Ganz catheter (Edwards Lifesciences, Irvine, CA, USA) through jugular or femoral vein access. Mean right atrial pressure (RAPM), diastolic (PAPD), mean pulmonary artery pressure (PAPM), mean pulmonary artery wedge pressure (PAWPM), and right ventric- ular systolic pressure (RVSP) were recorded under fluoroscopy after calibration with the zero level set at the mid-thoracic line. All pressure tracings were stored in a connected haemodynamic recorder and anal- ysed offline with commercially available software (Xper Information Management, Philips Medical Systems, The Netherlands). Importantly, in order to ensure the uniformity of data acquisition and the standard- ization of the study, the same investigator (A.M.) participated in RHC for all MS and the majority of PH-LHD patients and performed the analysis of all waveforms at both sites. From the PAWP recordings, the peak V- and A-wave and the PAWPM were obtained. All pressure measurements were averaged from a minimum of five heart cycles at end-expiration. Cardiac output (CO) was measured using Fick’s principle. The oxygen consumption was measured breath by breath by a dedicated gas analysis system. In 15 cases, thermodilution was employed.

The PVR, TPG, and DPG were calculated as: PVR = (PAPM – PAWPM)/CO; TPG = PAPM – PAWPM; and DPG = PAPD – PAWPM, respectively. The difference between TPG and DPG (ΔPG), which equals PAPM – PAPD, was analysed in order to investigate diagnos- tic discrepancies by the two measures. The right ventricular stroke work index was calculated as RVSWi = (PAPM – RAPM) × SVi × 0.0136, where SVi denotes the stroke volume index measured as: CO/heart rate (HR)/body surface area. In MS patients, measurements were per- formed prior to PTMC. For full details of methods, please see the Supplementary material online.

Simultaneous left atrial pressure and pulmonary artery wedge pressure assessment In 51 MS patients, simultaneous, beat-to-beat, LAP and PAWP tracings were obtained concurrently with RHC. Interatrial septal puncture was performed with an 8 F Mullins’ sheath, dilator and a Brockenbrough needle. The LAP was measured directly through the Mullins’ sheath used during valvuloplasty. Both transducers were zeroed after careful calibration, pressures were recorded during a 10 sec period and stored for offline analysis.

Statistical analysis The IBM SPSS statistics version 23.0 was used. Normality was tested by the Kolmogorov – Smirnov test. Continuous variables were

90 A.I. Nagy et al.

expressed as mean ± SD or median and interquartile range. Cat- egorical variables were expressed as absolute values and percent- age. Comparisons of groups were performed with Mann – Whitney rank-sum test. Correlations were tested by the Pearson’s two-tailed test. All tests were performed at 95% confidence intervals (CIs). A P-value of <0.05 was considered statistically significant. Receiver operator characteristic (ROC) curve analysis was performed. Sur- vival was analysed in 127 PH-LHD patients (115 retrospective, 12 prospective) with Kaplan – Meier non-parametric test and compared using a log-rank test.† Univariate and multiple Cox proportional haz- ards regression models were used to examine the effects of the DPG on patients’ survival. Age-, creatinine-, and sex-adjusted sur- vival curve estimates of the DPG were derived from stratified Cox models.

Results Study population Of the 316 patients enrolled, 269 (84.5%) demonstrated PH (PAPM ≥25 mmHg). Of these, 256 (MS: 37%) had PH-LHD (PAPM ≥25 mmHg and PAWPM > 15 mmHg). Demographics are presented in Table 1. Due to the different underlying pathology, the MS and PH-LHD groups were analysed separately. MS patients had higher PAPM, A- and V-waves, and RVSWi compared with the PH-LHD group. However, DPG did not differ between the two groups (Table 2).

V-wave influence on the diastolic pulmonary pressure gradient To evaluate the effect of V-waves on the DPG, we subgrouped the cohort based on the presence of large V-waves, defined as the V-wave exceeding the PAWPM by the arbitrary limit of >10 mmHg as previous investigators have performed.13 In the 69 cases (45%) with large V-waves (43 MS and 26 PH-LHD patients), the DPG was on average negative and lower (P < 0.05) compared with those with smaller V-waves, despite similar levels of TPG, PVR, PAP, and cardiac index (P > 0.05, for all comparisons; Table 3; Supplementary material online, Figure S2).

A significant inverse correlation between the V-wave and DPG was evident in patients with PVR <3 Wood Units (WU) (r = −0.45, P < 0.001), both in the MS (r = −0.34, P = 0.03) and in the PH-LHD group (r = −0.46, P < 0.001). A weaker, yet statistically significant inverse correlation (r = −0.36; P = 0.01) between the V-wave and DPG was found in patients with a PVR of 3 – 7 WU. However, this relationship disappeared at higher PVR values (P > 0.05; Figure 1A). Conversely, no association between the V-wave and TPG was observed (P > 0.05; Figure 1B). The modest overall correlation between the V-wave and DPG might be ascribed to the diver- gent association of V-waves with PAPD at higher PAPM and PVR (Figure 1D), whereas the association between V-waves and PAWPM was essentially unaltered throughout the examined PAPM and PVR range (Figure 1C).

†Correction added on November 23, 2016, after first online publication: the patient count given in this sentence was corrected. ..

.. ..

.. Importantly, in patients with PVR <3 WU, the V-wave showed the strongest correlation with the ΔPG (r = 0.45, P < 0.001 for the whole cohort, r = 0.36, P = 0.005 for PH-LHD; r = 0.6, P = 0.003 for the MS group, Figure 1E), with a weaker yet significant association of both the absolute and relative V-wave value with ΔPG (r = 0.26 and r = 0.19, respectively; P < 0.05). Conversely, neither the A-wave nor the CO correlated with ΔPG (P > 0.05, in all cases).

The puzzling finding of normal DPG with concomitantly ele- vated TPG (>12 mmHg) is not unusual. Indeed, in our study 59 patients (23%, MS: 29%), TPG and DPG demonstrated incongru- ent diagnostics (TPG >12 mmHg, DPG <7 mmHg). Furthermore, DPGNEG with concomitantly elevated TPG (>12 mmHg) occa- sionally occurs. In our study, we decided to quantify this discrep- ancy by calculating ΔPG (ΔPG = TPG – DPG). The ΔPG value that leads to discrepant Cpc-PH diagnostics between TPG and DPGNEG is 12 mmHg. In order to examine whether the V-wave amplitude impacted on this discrepancy, we employed ROC analy- sis in patients with PVR <3 WU. The association between ΔPG and V-wave amplitude is presented in Figure 1E. At an optimal cut-off limit of 30.5 mmHg, V-wave yielded a sensitivity of 85% and a specificity of 70% [area under the curve (AUC) 0.80, 95% CI 0.72 – 0.88; P < 0.001) for the identification of ΔPG >12 mmHg (Supplementary material online, Figure S3). For the whole cohort of patients with PVR <7 WU, the corresponding figures were: AUC 0.73, P < 0.003; 95% CI 0.61 – 0.84 at an optimal cut-off limit of the V-wave of 31.5 mmHg.

In an attempt to investigate potential non-invasive and clinical determinants of the V-wave amplitude, left atrial end-systolic volume index, LV mass index, internal LV dimensions, as well as the available clinical variables were tested. None of the tested variables, however, was associated with the V-wave (P > 0.05 in all cases).

Negative diastolic pulmonary pressure gradient values In total, 123 patients (48%) demonstrated DPGNEG (median −3 mmHg; interquartile range −5 to −2 mmHg) with higher preva- lence in the MS compared with the PH-LHD group (55% vs. 44%, P < 0.05). MS patients had significantly higher V-waves (P < 0.001, Table 2). When the whole study population was considered, patients with DPGNEG showed significantly larger V-waves and lower PAPM, RAPM, PVR, and TPG values, whereas the PAWPM and cardiac index levels were comparable with those with positive DPG (Table 4).

Assuming that pre-capillary changes differ between positive DPG and DPGNEG patients, we compared the two groups within a pre-defined PVR range (3 – 7 WU) in order to ensure a compar- atively equivalent degree of pre-capillary alterations between the two groups. Patients with DPGNEG demonstrated higher V-waves in both the MS and PH-LHD groups, and a less prominent right heart dilatation along with better RV function (P < 0.001) as com- pared with the positive DPG cohort, despite similar PAPM (P > 0.05, Table 4; Supplementary material online, Table S1). Interestingly, the V-wave amplitude was similar in MS and PH-LHD patients in the DPGNEG group.

Negative DPG in pulmonary hypertension 91

Table 1 Demographic and echocardiographic data of the study population

All patients (n = 256) MS (n = 94) PH-LHD (n = 162) P-value PH-LHD R (n = 124) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Demographics Age 50 ± 19 31 ± 9 61 ± 15 <0.001 61 ± 15 Female (%) 51% 72% 39% <0.001 40% BSA (m2) 1.8 ± 0.3 1.4 ± 0.2 2.0 ± 0.2 <0.001 1.9 ± 0.2 HT (%) 0% 85% 51% DM (%) 0% 60% 45% Aetiology of HF

IHD (n, %) 0% 36 (22%) 32 (26%) Idiopathic 68 (42%) 48 (39%) Myocarditis 21 (13%) 6 (5%) Other 37 (23%) 38 (31)

AF (n, %) 53 (21%) 0 53 (33%) 43 (35%) Functional class

NYHA II – IIIa 60 (64%) 84 (52%) <0.001 70 (56%) NYHA IIIb 34 (36%) 49 (30%) <0.001 29 (23%) NYHA IV – 29 (18%) 25 (20%)

Medication Diuretics 100% 81% 78% ACE inhibitor 85% 81% Beta-blocker 100% 98% 93% CCA 25% 18% MRA 31% 34%

Echo data EF ≤45% 69 (27%) 5 (5%) 62 (38%) <0.001 55 (44%) LVEDD (mm) 44 ± 7 52 ± 13 <0.001 54 ± 14 LVESD (mm) 29 ± 0.4 41 ± 15 <0.001 43 ± 16 LVMi (g/m2) 64 ± 18 105 ± 50 <0.001 114 ± 55 LA-ESVi (mL/m2) 68 ± 19 50 ± 21 <0.001 58 ± 20 MVA (cm2) 0.8 ± 0.2 MVG (mmHg) 19 ± 9 RVEDD (mm) 36 ± 5 40 ± 8 <0.001 41 ± 7 TAPSE (mm) 18 ± 3 14 ± 5 <0.001 14 ± 4

MR grade Mild 163 (63%) 64 (68%) 99 (61%) <0.001 82 (66%) Moderate 23 (9%) – 23 (14%) 14 (11%) Severe 17 (6%) – 17 (10.5%) 11 (9%)

AS grade Moderate 3 (1%) – 3 (2%) 4 (3%)

AR grade Mild 32 (13%) – 32 (20%) 31 (25%) Moderate 3 (1%) – 3 (2%) 6 (5%)

Data are expressed as mean ± SD or number (%). P-values indicate the difference between the two prospective cohorts, i.e. MS and LHD. AR, aortic valve regurgitation; AS, aortic valve stenosis; BSA, body surface area; CCA, calcium channel blocker; DM, diabetes mellitus; IHD, ischaemic heart disease; MS, mitral valve stenosis; PH-LHD, pulmonary hypertension due to left heart disease; PH-LHD R, retrospective arm of the PH-LHD group; HT, hypertension; LA-ESVi, left atrial end-systolic volume index; LVEDD, left ventricular end-diastolic diameter; LVESD, left ventricular end-systolic diameter; LVMi, LV mass index; MRA, mineralocorticoid receptor antagonist; MVA, mitral valve area; MVG, mitral valve mean diastolic gradient; RVEDD, right ventricular end-diastolic diameter; TAPSE, tricuspid annular plane systolic excursion; MR, mitral valve regurgitation.

Determinants of the diastolic pulmonary pressure gradient Left atrial pressure vs. pulmonary artery wedge pressure in diastolic pulmonary pressure gradient assessment

In the 51 MS patients with simultaneous PAWP and LAP recordings,

the DPG was calculated from PAWP (DPGPAWP) and LAP (DPGLAP)

.. ..

.. . separately. DPGPAWP was negative in 28 cases while DPGLAP was

negative in 22 cases, due to a slightly yet not significantly lower (mean bias: −2 mmHg) LAP (24.1 ± 8.0 mmHg) as compared with PAWP (26.0 ± 8.1 mmHg; P > 0.05). However, in only three cases with negative DPGPAWP was the corresponding DPGLAP positive, while in one case reclassification occurred in the opposite direction.

92 A.I. Nagy et al.

Table 2 Haemodynamics of the entire cohort

All patients (n = 256) MS (n = 94) PH-LHD (n = 162) P-value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PAPM (mmHg) 35 (29 to 44) (256) 38 (30 to 50) (94) 34 (29 to 43) (162) 0.024 PAPD (mmHg) 24 (20 to 31) (255) 27 (19 to 36) (94) 23 (20 to 29) (161) 0.026 RVSP (mmHg) 24 (21 to 29) (256) 59 (47 to 83) (94) 40 (49 to 63) (162) <0.001 PAWPM (mmHg) 24 (21 to 29) (256) 25 (23 to 32) (94) 23 (20 to 27) (162) 0.026 A-wave (mmHg) 26 (22 to 32) (229) 31 (26 to 37) (91) 24 (21 to 28) (138) <0.001 V-wave (mmHg) 31 (27 to 37) (235) 35 (31 to 44) (94) 28 (25 to 33) (141) <0.001 CI (L/min/m2) 1.9 (1.6 to 2.4) (256) 1.7 (1.4 to 2.1) (94) 2 (1.7 to 2.5) (162) <0.001 RAPM (mmHg) 10 (6 to 15) (255) 6 (3.8 to 8) (94) 12 (9 to 17) (161) <0.001 RVSWi (g/m2/beat) 9 (6.6 to 13) (255) 10.4 (7.8 to 14.8) (94) 8.2 (6 to 12.2) (161) <0.001 AV (mL/L) 54 (45 to 65) (241) 50 (42 to 57) (94) 57 (45 to 17) (147) <0.001 DPG (mmHg) 0 (−3 to 4) (255) −1 (−4 to 5) (94) 0 (−3 to 3) (161) 0.327 DPG <7 −1 (−4 to 1) (83%) −2 (−5 to 0) (79%) −1 (−3 to 1) (85%) DPG ≥7 13 (9 to 15) (17%) 14 (10 to 18) (21%) 12 (9 to 14) (14%) TPG (mmHg) 10 (7 to 18) (256) 9 (6 to 21) (94) 11 (7 to 16) (162) 0.72

TPG ≤12 8 (5.5 to 9) (61%) 7 (5 to 9) (62%) 8 (6 to 10) (61%) TPG >12 20 (16 to 27) (39%) 25 (18 to 34) (38%) 19 (15 to 23) (39%)

PVR (WU) 3 (1.8 to 5.2) (256) 4 (2.5 to 8.8) (94) 2.6 (1.7 to 4.5) (162) <0.001 PVR <3 1.8 (1.4 to 2.5) (51%) 1.9 (1.3 to 2.6) (36%) 1.8 (1. 3 to 2.4) (59%) PVR ≥3 5.3 (3.8 to 7.8) (49%) 7.1 (4.1 to 11.6) (64%) 4.8 (3.8 to 6.1) (41%)

Values are expressed as the median and interquartile range. P-values report the statistical difference between MS and PH-LHD. AV, arteriovenous difference of oxygen saturation; CI, cardiac index; DPG, diastolic pulmonary pressure gradient; MS, mitral valve stenosis; PAPM and PAPD, mean and diastolic pulmonary artery pressure, respectively; PAWPM , mean pulmonary artery wedge pressure; PH-LHD, pulmonary hypertension due to left heart disease; PVR, pulmonary vascular resistance; RAPM, mean right atrial pressure; RVSP, right ventricular systolic pressure; RVSWi, right ventricular stroke work index; TPG, transpulmonary pressure gradient; V- and A-wave, the maximal amplitude of the V- and A-wave of the PAWP waveform, respectively; WU, Wood Units.

Table 3 Haemodynamics stratified according to V-wave amplitude

Small V-waves n = 166 (51 MS) Large V-waves n = 69 (43 MS) P-value . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

PAPM (mmHg) 34 (29 to 44) 35 (30 to 45) 0.36 PAPD (mmHg) 24 (20 to 30) 23 (19 to 32) 0.77 PAWPM (mm Hg) 23 (20 to 27) 25 (22 to 31) 0.001 V-wave (mmHg) 28 (25 to 32) 39 (34 to 46) <0.001 V-waveabs (mmHg) 5 (3 to 7) 13 (11 to 17) <0.001 PVR (WU) 2.9 (1.9 to 5.6) 3.1 (1.7 to 5.2) 0.73 TPG (mmHg) 11 (7 to 19) 9 (7 to 15) 0.39 DPG (mmHg) 0 (−2 to 5) −2 (−4 to 1) 0.002 CI (L/min/m2) 1.9 (1.6 to 2.4) 1.8 (1.6 to 2.5) 0.26

Values are expressed as the median and interquartile range. Small V-wave signifies a difference between maximal amplitude of the V-wave of the PAWP waveform (PAWPv) and the mean pulmonary artery wedge pressure (PAWPM ), i.e. V-waveabs of <10 mmHg. Large V-wave signifies a V-waveabs ≥10 mmHg. CI, cardiac index; DPG, diastolic pulmonary pressure gradient; MS, mitral valve stenosis; PAPM and PAPD , pulmonary artery mean and diastolic pressure, respectively; PAWPM , mean pulmonary artery wedge pressure; PVR, pulmonary vascular resistance; TPG, transpulmonary pressure gradient; WU, Wood Units.

Heart rhythm

When the analysis was confined to the 192 patients with HR < 85 b.p.m., 52% demonstrated DPGNEG. Similarly, when only the 53 patients in AF were considered, DPGNEG was measured in 50%.

Alternative pulmonary artery wedge pressure measurements

As detailed in the Supplementary material online (Table S2), when the DPG was calculated using the PAWP value measured at the .

.. ..

.. z-point of the PAWP curve, instead of using PAWPM in patients with DPGNEG, this resulted in significantly higher DPG values. Still, the prevalence of DPGNEG was not significantly reduced.

Prognostic value of the diastolic pulmonary pressure gradient Two-year outcome for the combined endpoint of death or cardiac transplantation was significantly better for PH-LHD patients with

Negative DPG in pulmonary hypertension 93

10 DeltaPG [mmHg]

201550

V -w

a ve

[ m

m H

g ]

PH-LHD; r= 0.36

MS; r=0.60

806040200

P A

W P

M [m

m H

g ]

r = 0.88; p < 0.001

r = 0.76; p < 0.001

PVR < 3 WU PVR ≥ 3 WU

V-wave [mmHg]

PVR < 3 WU PVR ≥ 3 WU

806040200

P A

P D [ m

m H

g ]

r = 0.55; p < 0.001

r = 0.46; p < 0.001

V-wave [mmHg]

806040200

T P

G [ m

m H

g ]

-10

r = - 0.14; p: NS

r = 0.12; p: NS

V-wave [mmHg]

PVR < 3 WU PVR ≥ 3 WU

806040 V-wave [mmHg]

200

D P

G [ m

m H

g ]

-10

-20

r = - 0.45; p < 0.001

r = - 0.1; p = NS

PVR < 3 WU PVR ≥ 3 WU

A B

Figure 1 (A) Correlation between the diastolic pulmonary pressure gradient (DPG) and V-wave amplitude in patients with low (<3 WU) and high (≥ 3 WU) pulmonary vascular resistance (PVR). (B) Correlation between the transpulmonary pressure gradient (TPG) and V-wave amplitude in patients with low (<3 WU) and high (≥3 WU) PVR. (C) Correlation between mean pulmonary artery wedge pressure (PAWPM) and V-wave amplitude in patients with low (<3 WU) and high (≥3 WU) PVR. (D) Correlation between diastolic pulmonary artery pressure (PAPD) and V-wave amplitude in patients with low (<3 WU) and high (≥3 WU) PVR. (E) Correlation between V-wave amplitude and ΔPG in patients with mitral valve stenosis (MS) and pulmonary hypertension due to left heart disease (PH-LHD). WU, Wood Units.

94 A.I. Nagy et al.

Table 4 Comparison of negative and positive diastolic pulmonary pressure gradient groups within the entire study population and in patients with a predefined pulmonary vascular resistance range of 3 – 7 Wood Units

All patients PVR 3 – 7 WU . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

DPG <0 (n) DPG ≥0 (n) DPG <0 (n) DPG ≥0 (n) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

MS patients (n) 52 (42%) 42 (32%) 18 (64%) 11 (19%) PAPM (mmHg) 31 (28 to 37) (123) 41 (33 to 49) (132) (P < 0.001) 38 (30 to 43) (28) 40 (34 to 45) (57) (P = 0.128) PAPD (mmHg) 20 (17 to 26) (123) 28 (23 to 35) (132) (P < 0.001) 23 (18 to 30) (28) 27 (24 to 31) (57) (P = 0.013) V-wave (mmHg) 33 (28 to 39) (112) 29 (25 to 36) (123) (P < 0.001) 37 (32 to 42) (26) 28 (24 to 33) (52) (P < 0.001) PAWPM (mmHg) 24 (21 to 29) (123) 24 (20 to 28) (132) (P = 0.06) 25 (21 to 32) (28) 24 (20 to 28) (57) (P = 0.071) RVSP (mmHg) 49 (41 to 59) (123) 62 (47 to 78) (132) (P < 0.001) 51 (46 to 32) (28) 61(47 to 71) (56) (P = 0.67) RAPM (mmHg) 9 (5 to 13.5) (123) 11 (7 to 15) (132) (P = 0.004) 7.5 (4 to 10) (28) 11 (7 to 15) (57) (P = 0.005) PVR (WU) 2.2 (1.4 to 3.0) (123) 4.7 (2.6 to 7.6) (132) (P < 0.001) 4 (3.4 to 4.8) (28) 4.7 (3.7 to 5.6) (57) (P = 0.09) DPG (mmHg) −3 (−5 to −2) (123) 3 (1 to 9) (132) (P < 0.001) −2.5 (−4 to −1) (28) 3.0 (1 to 5) (57) (P < 0.001) TPG (mmHg) 7 (5 to 9) (123) 16 (11 to 24) (132) (P < 0.001) 9 (8 to 14) (28) 15 (12 to 21) (57) (P < 0.001) CI (L/min/m2) 1.9 (1.6 to 2.5) (123) 1.9 (1.6 to 2.3) (132) (P = 0.392) 1.7 (1.3 to 1.9) (28) 1.8 (1.6 to 2.2) (57) (P = 0.034) RVSWi (g/m2/beat) 8.2 (6.4 to 11) (123) 10.5 (6.8 to 15) (P = 0.004) 8.4 (6 to 12.6) (28) 10.3 (6.3 to 14) (57) (P = 0.24) A – V (mL/L) 49 (42 to 59) (115) 58 (48 to 69 (126) (P < 0.001) 49 (41 to 63) (28) 62 (49 to 71) (53) (P = 0.04) TAPSE (mm) 17 (12 to 19) (123) 15 (12 to 18) (132) (P = 0.025) 18 (15 to 21) (28) 14 (11 to 17) (57) (P = 0.004) RA area (cm2) 18 (12 to 24) (123) 22 (15 to 27) (132) (P = 0.002) 12 (10 to 24) (28) 23 (18 to 29) (57) (P < 0.001) RVEDD (mm) 36 (33 to 41) (123) 38 (34 to 46) (132) (P < 0.003) 34 (33 to 43) (28) 40 (36 to 48) (57) (P = 0.005)

Values are expressed as the median and interquartile range. A – V, arteriovenous difference in oxygen saturation; CI, cardiac index; DPG, diastolic pulmonary pressure gradient; MS, mitral valve stenosis; PAPM and PAPD , mean and diastolic pulmonary artery pressure, respectively; PAWPM and V-wave, mean pulmonary artery wedge pressure and the maximal amplitude of the V-wave of the PAWP waveform, respectively; PVR, pulmonary vascular resistance; RA, right atrial; RAPM , mean right atrial pressure; RVEDD, right ventricular end-diastolic diameter; RVSP, right ventricular systolic pressure; RVSWi, right ventricular stroke work index; TAPSE, tricuspid annular plane systolic excursion; TPG, transpulmonary pressure gradient; WU, Wood Units.

DPGNEG as compared with those with positive but normal DPG (0 ≤ DPG <7 mmHg) (Figure 2A). In the DPGNEG group (n = 57), the combined endpoint was documented in 16 cases (10 deaths and 6 transplantations), while in the 0 ≤ DPG <7 mmHg group (n = 53) the corresponding figures were 24 (14 deaths and 10 transplanta- tions). Finally, in the DPG ≥7 mmHg group (n = 17), eight combined endpoint events were recorded (5 deaths and 3 transplantations).

The occurrence of the combined endpoint of death or trans- plantation was significantly higher for 0 ≤ DPG <7 mmHg both in unadjusted analysis (P < 0.005) and when adjusted for age, creatinine, and ischaemic heart disease (Figure 2B). Conversely, neither TPG (cut-off 12 mmHg) nor PVR (cut-off 3 WU) pro- vided significant prognostic information (P = 0.522 and P = 0.718, respectively). Furthermore, combining DPG and TPG (DPGNEG and TPG ≤12 mmHg vs. 0 ≤ DPG <7 mmHg and TPG >12 mmHg) also failed to provide prognostic information (P = 0.223).

Discussion In the present study, we (i) confirm the high prevalence of DPGNEG in PH-LHD patients; (ii) demonstrate that DPGNEG does not always represent a measurement error, but instead may be ascribed to high V-wave amplitude in patients with relatively low resistance in the pulmonary vascular bed; and (iii) show that DPGNEG is associated with lower mortality as compared with the corresponding group of positive yet not elevated DPG.

In healthy subjects and in patients without significant pre-capillary alterations, PAPD is closely related to LAP, with .

.. ..

.. .. DPG values ranging between 0 and 5 mmHg.5 DPGNEG values

have so far been regarded as a measurement bias, ascribed to overwedging or inaccurate PAPD recordings.

5 However, the high DPGNEG prevalence, ranging from 20% in critically ill patients

11,14

to 35%8 and up to 50%15 in PH-LHD patients, calls for a reappraisal of its pathophysiological origin. DPGNEG was found in 44% of our PH-LHD cohort, most probably reflecting the higher proportion of PH (95%) compared with that (45%) reported in a recent study.8

V-wave influence on the diastolic pulmonary pressure gradient During systole, the second phase of LA filling occurs, yielding the most prominent positive deflection of the PAWP waveform desig- nated as the V-wave. The volume and the rate of blood entering the left atrium as well as this chamber’s compliance determine the V-wave amplitude,16,17 which in healthy subjects averages 12 mmHg, ranging between 4 and 19 mmHg, being at most 6 mmHg higher than LAPM.

18 Importantly, the LA volume – pressure rela- tionship follows an exponential rather than a linear pattern, so that at lower LAP a certain volume entering the left atrium yields minor pressure elevation, whereas at higher LAP an equal inflow- ing volume results in a greater pressure rise.13,16 Conceivably, large V-waves arise not only in the presence of severe acute mitral regurgitation19 but also in conditions such as MS20 and longstanding LV dysfunction, when LA distensibility is impaired, resulting in an upward shift of the LA volume – pressure curve. In our study, large V-waves were present in 20% of the PH-LHD group and in 46% of

Negative DPG in pulmonary hypertension 95

Figure 2 (A) Kaplan – Meier analysis for the three diastolic pulmonary pressure gradient (DPG) groups. Group I, DPG <0 mmHg; Group II, 0 ≥ DPG <7 mmHg; Group III, DPG ≥7 mmHg. (B) Hazard ratio for death and/or transplantation for patients with positive normal DPG (0 ≤ DPG <7 mmHg) and negative DPG. Due to few patients in Group III, only the statistical comparison between Group I and II is presented. CI, confidence interval; HTX, heart transplantation; IHD, ischaemic heart disease.

the MS cohort, similar to the findings of Wang and colleagues.20

It should be emphasized that the augmented V-waves in these two cohorts represent distinct haemodynamic conditions; in MS it reflects increased LA stiffness due to obstructed mitral valve ori- fice, whereas in PH-LHD it is mainly secondary to a rise in LV end-diastolic pressure (LVEDP). It has been shown that the dis- torted LAP waveform in the presence of large V-waves leads to overestimation of the LVEDP.21 Furthermore, there is evidence of retrograde superimposition of prominent V-waves on the PAP contour.22 Caro and colleagues demonstrated that at high LAP, the ratio of pulmonary arterial to pulmonary venous compliance changes, promoting an asymmetrical backward transmission of the phasic LAP.23 Although studies concomitantly reporting V-wave amplitude and PAPD are infrequent, the existing data on large V-waves in the context of increased LA stiffness reveal DPGNEG in essentially all cases.17 Importantly, we demonstrate that the inverse correlation between the V-wave and DPG was confined to patients with relatively low PVR, in accordance with the findings of Falicov and colleagues.15 Under physiological conditions, at end-diastole, the pulmonary vascular bed allows pressure equilibration24 which is .

.. ..

.. . otherwise hindered by the presence of vascular remodelling. Taken

together, our results indicate that in PH-LHD the V-wave ampli- tude significantly influences the DPG calculation unless significant pre-capillary remodelling is present. However, with progressive maladaptive pre-capillary alterations, the V-wave no longer acts as an important determinant of the DPG, which might be explained by increased stiffening of the pulmonary arteries and thus dampening of the backward LAP transmission. Previous investigations suggest that large V-waves inversely correlate with the ratio between sys- tolic and diastolic pulmonary inflow velocities.25 In accordance with previous investigators, LA volume was not associated with V-wave amplitude.26 As echocardiography plays a key role in the initial PH assessment in HF, further studies are warranted to address poten- tial incremental value of this modality.

Methodological considerations The current findings argue against the notion that DPGNEG represents merely an inaccurate measurement. First, the PAWP and PAP waveforms were assessed manually at end-expiration by a

96 A.I. Nagy et al.

single investigator, limiting the possibility of erroneous computer- ized PAPD measurements and preventing potential PAWPM under- estimation due to pressure averaging throughout the respiratory cycle.27 Experimental studies have shown that HR impacts on DPG; at higher HR, DPG rises due to lower LVEDP and a concomi- tant PAPD elevation.

28 Our results reveal that even when confining the analysis to patients with normal HR or patients with AF, the incidence of DPGNEG was unaltered. Finally, our simultaneously performed PAWP and LAP measurements partly contradict the opinion that DPG would be a result of erroneous PAWP record- ings. Direct LAP measurements yielded slightly higher DPG values as compared with PAWP. In ∼11% cases with negative DPGPAWP, the corresponding DPGLAP was positive, while in one case reclas- sification occurred in the opposite direction (4.5%). This finding points to the fact that due to its low absolute value, even a small measurement error will affect the DPG value; however, it also demonstrates that measurement error accounts for only a minority of DPGNEG cases. Taken together, although the slight discrepancy between LAP and PAWP might account for a minor portion of the DPGNEG, our findings suggest that DPGNEG values can for the most part be ascribed to the augmented V-waves.

Prognostic significance The prognostic impact of DPGNEG is as yet unknown. It has been suggested that patients with DPGNEG, instead of being a subclass of the isolated post-capillary PH (DPG <7 mmHg) group, in fact represent a cohort with worse haemodynamics.8 Our findings contradict this hypothesis. We demonstrate that when compar- ing DPGNEG patients with those with 0 ≤ DPG <7 mmHg, within a pre-defined range of PVR (3 – 7 WU), the DPGNEG cohort is characterized by lower RAP, and higher tricuspid annular plane sys- tolic excursion (TAPSE), reflecting a state of less pronounced right heart loading and remodelling advocating for milder haemodynamic derangements in the DPGNEG group. This, together with the lower event rate in the DPGNEG as compared with the DPG 0 – 7 mmHg cohort further supports the concept that DPGNEG in large part results from high V-waves shifting the DPG towards lower values, and suggests limited pre-capillary changes.

In our study, neither PVR nor TPG was associated with worse outcome. Furthermore, combining TPG and PVR with DPG failed to demonstrate significant prognostic value (P = 0.223 and P = 0.195, respectively). This observation stands in contrast to pre- vious results and might be partly related to differences in patient profile. Indeed, as compared with the report by Tampakakis et al., the occurrence of ischaemic heart disease was much higher in our study;8 additionally, our patient cohort comprised older patients than those studied by Tampakakis et al. or Tedford et al.8,9 Finally, the follow-up period was shorter in our study. The constellation of the aforementioned issues as well as the fact that our study comprised fewer patients might account for this discrepancy.

Limitations Heterogeneity might be considered as comprising a limitation of the current study as catheterizations were performed in two .

.. ..

.. . different centres. However, all studies in India were performed in

the presence of A.M. who was responsible for the standardization of the studies in the two centres; additionally, the same technical equipment and catheters were used at both sites. Patient char- acteristics as well as haemodynamics of the two studied cohorts are also rather divergent, as demonstrated in Table 1 (e.g. patients with AF, hypertension, or ischaemic heart disease were excluded from the MS but not the PH-LHD group); however, as the objec- tive of the present study was not to assess the influence of AF or other co-morbidities on the DPG, but rather to assess the effect of V-wave amplitude on DPG measurement, we believe that despite the patients’ heterogeneity, the haemodynamic essence of our hypothesis is still addressed. Our cohort comprised patients with PH-LHD (including both preserved and reduced EF) and MS, in which respect it is different from previous comparable studies. Indeed, pre-capillary involvement as defined by DPG ≥7 mmHg was more frequent in MS patients (20.2%). However, the preva- lence of Cpc-PH in the PH-LHD group was 13.6% that is com- parable with previous studies (8 – 16%).6,8,9 Finally, the current study was performed on haemodynamically stable patients, imply- ing that our findings might not be valid in a state of decompensated acute HF.

Conclusion The present study verifies the recently observed high frequency of DPGNEG. We propose an applicable physiological explanation for this haemodynamic finding demonstrating a significant inverse association of V-wave amplitude in the PAWP waveform with the DPG in patients with low PVR. Using direct LAP measure- ments, we show that the occurrence of DPGNEG is clearly not reflecting methodological inaccuracies; rather it largely represents the augmented disproportionate phasic LAP transmission. Finally, DPGNEG in patients with PH-LHD appears to be associated with milder haemodynamic derangements and better 2-year progno- sis compared with patients with DPG within the normal positive range.

Supplementary Information Additional Supporting Information may be found in the online version of this article: Supplementary Methods and Results. Figure S1. Flowchart demonstrating the patient enrolment pro- cess and haemodynamic classification. Figure S2. Representative pressure tracings illustrating the influ- ence of V-waves on the DPG value. Figure S3. Receiver operator characteristics (ROC) analysis of the prognostic ability of the V-wave (PAWPV) for identifying a ΔPG >12 mmHg in patients with pulmonary vascular resistance (PVR) <3 Wood Units. Table S1. Comparison of negative and positive DPG groups in MS and LHD patients with a pre-defined PVR range of 3 – 7 WU. Table S2. Alternative PAWP measurements and DPG calculation.

Negative DPG in pulmonary hypertension 97

Acknowledgement This project was supported by the János Bolyai Scholarship of the Hungarian Academy of Sciences. Conflict of interest: none to declare. Correction added on November 23, 2016, after first online publi- cation: Acknowledgement section was added.

References 1. Guazzi M, Borlaug BA. Pulmonary hypertension due to left heart disease.

Circulation 2012;126:975 – 990. 2. Mehra MR, Jessup M, Gronda E, Costanzo MR. Rationale and process: Interna-

tional Society for Heart and Lung Transplantation guidelines for the care of cardiac transplant candidates — 2006. J Heart Lung Transplant 2006;25:1001 – 1002.

3. Khush KK, Tasissa G, Butler J, McGlothlin D, De Marco T, ESCAPE Investiga- tors. Effect of pulmonary hypertension on clinical outcomes in advanced heart failure: analysis of the Evaluation Study of Congestive Heart Failure and Pul- monary Artery Catheterization Effectiveness (ESCAPE) database. Am Heart J 2009;157:1026 – 1034.

4. Miller WL, Grill DE, Borlaug BA. Clinical features, hemodynamics, and outcomes of pulmonary hypertension due to chronic heart failure with reduced ejection fraction: pulmonary hypertension and heart failure. JACC Heart Fail 2013;1: 290 – 299.

5. Naeije R, Vachiery JL, Yerly P, Vanderpool R. The transpulmonary pressure gradient for the diagnosis of pulmonary vascular disease. Eur Respir J 2013;41: 217 – 223.

6. Gerges C, Gerges M, Lang MB, Zhang Y, Jakowitsch J, Probst P, Maurer G, Lang IM. Diastolic pulmonary vascular pressure gradient: a predictor of prognosis in ‘out-of-proportion’ pulmonary hypertension. Chest 2013;143:758 – 766.

7. Sibbald WJ, Paterson NA, Holliday RL, Anderson RA, Lobb TR, Duff JH. Pul- monary hypertension in sepsis: measurement by the pulmonary arterial diastolic – pulmonary wedge pressure gradient and the influence of passive and active fac- tors. Chest 1978;73:583 – 591.

8. Tampakakis E, Leary PJ, Selby VN, De Marco T, Cappola TP, Felker GM, Russell SD, Kasper EK, Tedford RJ. The diastolic pulmonary gradient does not predict survival in patients with pulmonary hypertension due to left heart disease. JACC Heart Fail 2015;3:9 – 16.

9. Tedford RJ, Beaty CA, Mathai SC, Kolb TM, Damico R, Hassoun PM, Leary PJ, Kass DA, Shah AS. Prognostic value of the pre-transplant diastolic pulmonary artery pressure-to-pulmonary capillary wedge pressure gradient in cardiac trans- plant recipients with pulmonary hypertension. J Heart Lung Transplant 2014;33: 289 – 297.

10. Gerges C, Gerges M, Lang IM. Characterization of pulmonary hypertension in heart failure using the diastolic pressure gradient: the conundrum of high and low diastolic pulmonary gradient. JACC Heart Fail 2015;3:424 – 425.

11. Wilson RF, Beckman SB, Tyburski JG, Scholten DJ. Pulmonary artery diastolic and wedge pressure relationships in critically ill and injured patients. Arch Surg 1988;123:933 – 936. ..

.. ..

. 12. Chatterjee NA, Lewis GD. Characterization of pulmonary hypertension in heart failure using the diastolic pressure gradient: limitations of a solitary measurement. JACC Heart Fail 2015;3:17 – 21.

13. Pichard AD, Kay R, Smith H, Rentrop P, Holt J, Gorlin R. Large V waves in the pulmonary wedge pressure tracing in the absence of mitral regurgitation. Am J Cardiol 1982;50:1044 – 1050.

14. Heinonen J, Salmenpera M, Takkunen O. Increased pulmonary artery diastolic – pulmonary wedge pressure gradient after cardiopulmonary bypass. Can Anaesth Soc J 1985;32:165 – 170.

15. Falicov RE, Resnekov L. Relationship of the pulmonary artery end-diastolic pressure to the left ventricular end-diastolic and mean filling pressures in patients with and without left ventricular dysfunction. Circulation 1970;42:65 – 73.

16. Morrow AG, Braunwald E, Haller JA Jr, Sharp EH. Left heart catheterization by the transbronchial route: technic and applications in physiologic and diagnostic investigations. Circulation 1957;16:1033 – 1039.

17. Mehta S, Charbonneau F, Fitchett DH, Marpole DG, Patton R, Sniderman AD. The clinical consequences of a stiff left atrium. Am Heart J 1991;122:1184 – 1191.

18. Braunwald E, Brockenbrough EC, Frahm CJ, Ross J Jr. Left atrial and left ventricular pressures in subjects without cardiovascular disease: observations in eighteen patients studied by transseptal left heart catheterization. Circulation 1961;24:267 – 269.

19. Baxley WA, Kennedy JW, Feild B, Dodge HT. Hemodynamics in ruptured chor- dae tendineae and chronic rheumatic mitral regurgitation. Circulation 1973;48: 1288 – 1294.

20. Wang A, Harrison JK, Pieper KS, Kisslo KB, Bashore TM. What does the left atrial V wave signify during balloon commissurotomy of mitral stenosis? Am J Cardiol 1998;82:1388 – 1393.

21. Haskell RJ, French WJ. Accuracy of left atrial and pulmonary artery wedge pressure in pure mitral regurgitation in predicting left ventricular end-diastolic pressure. Am J Cardiol 1988;61:136 – 141.

22. Levinson DC, Wilburne M, Meehan JP Jr, Shubin H. Evidence for retrograde transpulmonary propagation of the V (or regurgitant) wave in mitral insufficiency. Am J Cardiol 1958;2:159 – 169.

23. Caro CG, Bergel DH, Seed WA. Forward and backward transmission of pressure waves in the pulmonary vascular bed of the dog. Circ Res 1967;20:185 – 193.

24. Kaltman AJ, Herbert WH, Conroy RJ, Kossmann CE. The gradient in pressure across the pulmonary vascular bed during diastole. Circulation 1966;34:377 – 384.

25. Klein AL, Stewart WJ, Bartlett J, Cohen GI, Kahan F, Pearce G, Husbands K, Bailey AS, Salcedo EE, Cosgrove DM. Effects of mitral regurgitation on pulmonary venous flow and left atrial pressure: an intraoperative transesophageal echocardiographic study. J Am Coll Cardiol 1992;20:1345 – 1352.

26. Pizzarello RA, Turnier J, Padmanabhan VT, Goldman MA, Tortolani AJ. Left atrial size, pressure, and V wave height in patients with isolated, severe, pure mitral regurgitation. Cathet Cardiovasc Diagn 1984;10:445 – 454.

27. Ryan JJ, Rich JD, Thiruvoipati T, Swamy R, Kim GH, Rich S. Current practice for determining pulmonary capillary wedge pressure predisposes to serious errors in the classification of patients with pulmonary hypertension. Am Heart J 2012;163: 589 – 594.

28. Enson Y, Wood JA, Mantaras NB, Harvey RM. The influence of heart rate on pulmonary arterial – left ventricular pressure relationships at end-diastole. Circulation 1977;56:533 – 539.

Copyright of European Journal of Heart Failure. Supplements is the property of John Wiley & Sons, Inc. and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.