Impact of acute hypertension transients on diastolic function in patients with heart failure with preserved ejection fraction

Transcription

1 Cardiovascular Research (2017) 113, doi:101093/cvr/cvx047 Impact of acute hypertension transients on diastolic function in patients with heart failure with preserved ejection fraction Candelas Pérez del Villar 1, Konstantinos Savvatis 2,3, Bego~na Lopez 4, Mario Kasner 5, Pablo Martinez-Legazpi 1, Raquel Yotti 1, Arantxa Gonzalez 4,JavierDıez 4, Francisco Fernandez-Avilés 1, Carsten Tschöpe 3,6,7, and Javier Bermejo 1 * 1 Department of Cardiology, Hospital General Universitario Gregorio Mara~non, Facultad de Medicina, Universidad Complutense de Madrid, Instituto de Investigacion Sanitaria Gregorio Mara~non, and CIBERCV, Dr Esquerdo 46, Madrid, Spain; 2 Department of Cardiology, Barts Heart Centre, Queen Mary University of London, London EC1A 7BE, UK; 3 Berlin- Brandenburg Centre for Regenerative Therapies, Föhrer Str 15, Berlin, Germany; 4 Program of Cardiovascular Diseases, Centre for Applied Medical Research, University of Navarra, IdiSNA, Navarra Institute for Health Research, and CIBERCV, Pıo XII 55, Pamplona, Spain; 5 Department of Cardiology, Charité University Medicine Berlin, Campus Benjamin Franklin, Hindenburgdamm 30, Berlin, Germany; 6 Department of Cardiology, Charité-University Medicine Berlin, Campus Virchow Klinikum, Augustenburger Pl 1, Berlin, Germany; and 7 DZHK, German Center for Cardiovascular Research, Partner Site Berlin-Charité, Oudenarder Str 16, Berlin, Germany Received 9 September 2016; revised 6 December 2016; editorial decision 20 February 2017; accepted 8 March 2017; online publish-ahead-of-print 11 April 2017 Time for primary review: 52 days Aim To address the mechanisms responsible for the increase in LV filling pressures induced by acute hypertension transients in patients with heart failure with preserved ejection fraction (HFpEF) Methods Multiple-beat pressure volume loops were recorded during inferior vena cava occlusion in 39 HFpEF patients and and results 20 controls during handgrip and atrial pacing We measured the contribution of relaxation, elastic recoil, and stiffness to instantaneous diastolic pressure using a novel processing method Fibrosis was quantified from endomyocardial biopsies HFpEF patients showed higher diastolic pressures and stiffness constant than controls (P < 005 for all) As opposed to controls, all intrinsic global diastolic properties were sensitive to acute changes in systolic pressure in the HFpEF group In fact, the stiffness constant increased by more than 50% during handgrip in HFpEF patients (P < 005), tightly related to changes in systolic pressure (fixed-effect = 026 mm Hg per mm Hg [95% CI = ]; P < 00001) Incomplete relaxation contributed to increasing pressure before atrial contraction, but changes in end-diastolic pressure was mostly caused by the increase in stiffness The degree of pressure-sensitivity of stiffness correlated with myocardial collagen volume and crosslinking (R = 040 to 082 for all) Conclusion Acute chamber stiffening is the main mechanism responsible for rising late-diastolic pressures when HFpEF patients undergo hypertension transients This stiffening behaviour is related to impaired dynamic systolic diastolic interactions and correlates with matrix remodelling Ventricular-vascular relationships are a promising target in HFpEF and should be taken into account when assessing diastolic function Keywords Diastolic function Left ventricle Heart failure Hypertension Diastolic stiffness Introduction Heart failure with preserved ejection fraction (HFpEF) is a complex and heterogeneous syndrome that leads to a common final stage of functional impairment and frequent acute decompensations 1 Although a number of abnormal properties of LV diastolic and systolic function concur in patients with HFpEF, 2 8 impaired chamber properties at rest do not fully explain the acute symptoms of patients with HFpEF 9 Acute hypertension transients play an important role in pulmonary congestion episodes in HFpEF Intrinsic diastolic chamber properties are known to be highly dynamic, 3,6 8 but the biomechanical foundations by which acute hypertension increases filling pressures are not fully understood; clarifying these mechanisms is a necessary first-step to target pharmacological therapies efficiently We hypothesized that acute hypertension episodes * Corresponding author Tel: ; fax: , javierbermejo@saludmadridorg The first two authors contributed equally to the study Published on behalf of the European Society of Cardiology All rights reserved VC The Author 2017 For permissions, please journalspermissions@oupcom Downloaded from

2 Acute hypertension transients in HFpEF 907 Table 1 Clinical and demographic data Clinical and Echocardiography Variables CONTROL HFpEF P-value Subjects (n) P-value Age (years old) 41 ± 8 54 ± 8 <0001 Gender (female) [n (%)] 5 (25) 18 (46) 020 Body mass index (kgm -2 ) 257 ± ± NYHA class I/II/III [n] 20/0/0 0/28/11 <0001 NT-proBNP (pg/ml) 56 ± ± 162 < min walk test (m) 528 ± ± 58 <0001 Cardiovascular Risk Factors Hypertension [n (%)] 2 (10) 23 (59) <0001 Diabetes mellitus [n (%)] 1 (5) 4 (10) 065 Dyslipemia [n (%)] 10 (50) 17 (44) 064 Smoker [n (%)] 9 (45) 18 (46) 093 Doppler-Echocardiography LV mass index (gm -2 ) 100 ± ± Diastolic interventricular septal thickness (mm) 10 ± 2 12 ± 3 <0001 Diastolic LV posterior wall thickness (mm) 10 ± 1 11 ± 3 <0001 Relative wall thickness 039 ± ± 018 <0001 ED diameter (mm) 51 ± 3 50 ± End-systolic diameter (mm) 37 ± 6 30 ± Left atrial volume index (mlm -2 ) 16 ± 3 22 ± E wave velocity (ms -1 ) 084 ± ± A wave velocity (ms -1 ) 066 ± ± E/A wave velocity ratio 133 ± ± e wave velocity (cms -1 ) 114 ± ± E/e wave velocity ratio 83 ± ± Values show mean ± standard deviation or n (%) NYHA, New York Heart Association; NT-proBNP, N-terminal pro-brain natriuretic peptide Significance values obtained by unpaired t-tests are followed by acute changes of intrinsic diastolic properties and that these changes are amenable to be addressed combining pressure volume (PV) measurements with appropriate signal processing techniques The present study was therefore designed to precisely quantify the impact of acute hypertension transients on diastolic function of patients with HFpEF on intrinsic diastolic chamber properties and filling pressures For this purpose, we used state-of-the-art methods combining (i) tailored acute haemodynamic interventions of handgrip and pacing to isolate the confusion caused by concomitant tachycardia, (ii) multibeat PV measurements during preload manipulation and RV unloading to rule-out external forces, 10,11 (iii) new data processing techniques using realistic biomechanical models 12,13 to accurately characterize intrinsic diastolic chamber properties, and (iv) endomyocardial biopsies to address the implication of extracellular matrix (ECM) remodelling in the hemodynamic response to hypertension Methods Patients and study protocol We studied 59 patients in sinus rhythm referred for coronary angiography to the Department of Cardiology and Pneumology of the Charité Campus Benjamin Franklin (Berlin, Germany) Data from some of these patients have been previously published The HFpEF group consisted of 39 patients presenting HF symptoms, LV ejection fraction > 50% and echocardiographic criteria of diastolic dysfunction (Table 1) Non-cardiac causes of symptoms and cardiomyopathies were specifically ruled out Twenty patients with atypical angina or intermittent arrhythmias showing normal systolic and diastolic LV function by Doppler-echocardiography were enrolled as controls Patients with valvular heart disease or significant coronary artery disease verified by angiography in all patients, were excluded Right endomyocardial biopsies were obtained in 24 patients and 12 controls to rule out cardiomyopathy The research protocol complied with the Declaration of Helsinki and was approved by the ethics committee of the Charité-University Medicine Berlin, Campus Benjamin Franklin (approval reference number ) Written informed consent was obtained from all subjects Data acquisition and analysis High-fidelity pressure-conductance 7F pig-tail catheters were placed retrogradely in the LV and connected to a dual-field conductance processor (CD- Leycom CFL-512) via the femoral artery Arterial pressure was recorded from the femoral sheath and corrected for augmentation using simultaneous high fidelity central aortic pressure waveforms registered before entering the ventricle LV volume was calibrated with the thermodilution and hypertonic saline methods A temporary pacemaker lead was introduced into the right atrium, and PV data were obtained sequentially during sinus rhythm, handgrip, and atrial pacing at 80, 100, and 120 bpm At each phase, at least three full PV datasets were obtained during inferior vena cava balloon occlusion (IVC occlusion hemodynamic runs) Acute caval occlusion is recognized as the best clinically available method to induce acute RV unloading in order to avoid the confounding effect of external forces on LV diastolic properties and pressures 10,11 Downloaded from

3 908 C Pérez del Villar et al We used an alternative algorithm to analyse the PV data which has important advantages over the conventional methods 12,13 Conventional PV-data analyses are based on fitting pressure-time during the isovolumic relaxation period to measure the time-constant of relaxation (s), and fitting enddiastolic PV relationship (EDPVR) to measure stiffness 17,18 However, this conventional approach provides mathematically unstable metrics accounting for operational mechanics instead of indices of true intrinsic chamber diastolic properties Also, conventional methods are unsuitable for isolating the effect of elastic recoil, and are unable to completely uncouple stiffness from residual relaxation at end diastole All these limitations can be solved using a new signal processing algorithm We have previously shown that it is possible to uncouple LV relaxation from mechanical forces caused by elastic recoil and stiffness by applying optimization methods to the full diastolic PV-time dataset from aortic valve closure to mitral valve closure of every consecutive beat acquired during preload manipulation 12 Using this alternative algorithm, robust estimates of s, the LV volume at zero passive pressure (V 0 ), the constant of elastic recoil (S ), the stiffness constant (S þ ), as well as the dead-volume and maximal-volume asymptotes (V d and V m, respectively) are obtained with very narrow confidence intervals 12 The contribution of each of these diastolic properties to instantaneous diastolic pressure can be calculated as the sum of relaxation-mediated pressure and the two volumemediated pressures related, one, to elastic recoil (suction pressure, when ventricular volume is < V 0 ), and the other one, to chamber stiffness (when volume is > V 0 ; see Supplementary material online, for details) These components of pressure were measured during early (aortic valve closure, mitral valve opening and at the time of minimum LV pressure, P min ), and late diastolic periods (pre-a, visually identified as the instant immediately before atrial contraction from simultaneous time pressure and time volume tracings, and end-diastole, ED, automatically determined and visually validated from the bottom right corner of the PV curve) We used the iterative method to estimate the maximum of time-varying LV elastance 13 Pulse pressure, systemic vascular resistance, and effective arterial elastance (Ea) were obtained as systolic minus diastolic blood pressure, mean blood pressure divided by cardiac output, and end-systolic pressure divided by stroke volume, respectively LV end-systolic wall stress was calculated from measurements of end-systolic pressure, end-systolic volume and echocardiographic measurements of LV mass 19 Histological studies Endomyocardial biopsies were obtained from the right-ventricular septum using a flexible bioptome via the femoral vein Collagen volume fraction was determined in Sirius red-staining paraffin samples by quantitative morphometry 14,20 Crosslinked (insoluble) and non-crosslinked (soluble) collagen were distinguished by enzymatic and colorimetric procedures 21 We used a fast green/sirius red assay to identify and quantify total collagen, followed by a sircol based assay to quantify soluble collagen Insoluble collagen was calculated by subtracting soluble collagen to the total collagen, and the degree of collagen crosslinking was obtained as the ratio between insoluble and soluble fractions Statistical analysis We used linear mixed-effects (LME) models to analyse the effects of interventions following a repeated measurements factorial design (two groups three interventions: baseline, handgrip and 120 bpm pacing data) Least square means, standard errors, and the significance of relevant contrasts for phase (Dunnett method against baseline) and group (HFpEF vs controls) are reported for each one of the six cells The impact of LVP max, indices of vascular load, and wall-stress on main diastolic properties was also addressed by LME models, adjusted for changes in heart-rate; from these models, bias and corrected 95% confidence intervals for the fixed effects were obtained by 500 bootstrap replicates, as well as pseudo-r 2 goodness-of-fit statistics 22 The slopes of the S and S þ over LVP max relationships (DS /DLVP max and DS þ /DLVP max, respectively) were obtained as the handgrip minus baseline Figure 1 Intrinsic chamber properties determined LV diastolic pressures Data from control (n = 20) and HFpEF patients (n =39) are shown at baseline and during interventions Scatterplots of each patient (dots) and average fitting (lines and 95% confidence interval ribbons) estimated at the times of aortic valve closure, mitral valve opening, minimal pressure, immediately before atrial contraction (pre-a) and end-diastole Measured pressure value values are overlaid Pressure values have been plotted at average times for each event See Supplementary material online for the method of pressure decomposition increments of the stiffness constants divided by the simultaneous increments of LVP max Paired, and unpaired t-tests as well as Pearson correlation and robust linear regression methods were used where appropriate Statistical analysis was performed using R v32 and P-values < 005 were considered significant Results Global haemodynamics At baseline, as well as during handgrip and pacing interventions, diastolic pressures in HFpEF patients were significantly higher than in controls at all measured phases of diastole (P < 0001 for all; Figure 1, Table 2 and see Supplementary material online, Table S1) At baseline, HFpEF patients also showed significantly higher values of S þ (P < 0001; Table 3), whereas differences in s and S were not significant between groups Indices of vascular afterload, Ea and SVR, were higher in HFpEF patients at baseline Downloaded from

4 Acute hypertension transients in HFpEF 909 Table 2 Global hemodynamic response to interventions Control HFpEF Haemodynamic Variable Baseline Handgrip Pacing 120 bpm Baseline Handgrip Pacing 120 bpm Subjects (n) IVC-occlusion hemodynamic runs (n) Heart rate (bpm) 71 ± 2 91 ± 2 a 120 ± 2 a 72 ± 2 93 ± 2 a 120 ± 2 a End diastolic volume (ml) 147 ± ± ± 7 a 134 ± ± ± 6 a End systolic volume (ml) 57 ± 5 68 ± 5 a 49 ± 5 a 58 ± 3 69 ± 4 a 51 ± 4 a Stroke volume (ml) 91 ± 5 81 ± 5 a 66 ± 5 a 76 ± 4 b 66 ± 4 a,b 51 ± 4 a,b LV ejection fraction 063 ± ± ± 002 a 056 ± 002 a,b 048 ± 002 a,b 050 ± 002 a,b LV end diastolic pressure (mm Hg) 8 ± 1 15 ± 1 a 7±1 12±1 b 19 ± 1 a,b 11 ± 1 b LV peak systolic pressure (mm Hg) 113 ± ± 5 a 111 ± ± 3 b 171 ± 4 a,b 127 ± 4 b Systolic blood pressure (mm Hg) 118 ± ± 4 a 122 ± ± 3 b 171 ± 3 a,b 132 ± 3 b Diastolic blood pressure (mm Hg) 67 ± 2 70 ± 2 73 ± 2 a 69 ± 1 72 ± 1 a 75 ± 2 Mean blood pressure (mm Hg) 88 ± 2 98 ± 2 a 92 ± 2 a 93 ± ± 2 a 97 ± 2 a Systemic vascular resistance (dynscm -5 ) 1187 ± ± ± ± 74 b 1454 ± 80 b 1442 ± 83 b E a (mm Hg/mL) 09 ± ± 02 a 12 ± ± 01 b 20 ± 01 a,b 22 ± 01 a,b E max (mm Hg/mL) 089 ± ± ± ± ± ± 030 E max /E a 10 ± ± ± ± ± 02 a 07 ± 02 a IVC, inferior vena cava; E max, peak systolic LV elastance; E a, aortic elastance; E max /E a, ventriculo-arterial coupling a P < 005 vs Baseline b P < 005 vs Control Table 3 Diastolic chamber properties during hemodynamic interventions Control HFpEF Diastolic index Baseline Handgrip Pacing 120 bpm Baseline Handgrip Pacing 120 bpm s (ms) 44 ± 2 47 ± 2 36 ± 2 a 48 ± 1 51 ± 2 a 42 ± 2 ab S - (mm Hg) 80 ± ± ± ± ± 16 ab 125 ± 17 ab S þ (mm Hg) 114 ± ± ± ± 14 b 258 ± 28 ab 129 ± 17 V 0 (ml) 83 ± 8 91 ± 9 73 ± 7 80 ± 6 86 ± 7 69 ± 6 Values represent least-square means ± their standard error s: time-constant of relaxation; S - : constant of diastolic elastic recoil; S þ : constant of passive stiffness; V 0 : left ventricular (LV) volume at zero passive pressure; a P < 005 vs Baseline b P < 005 vs Control Handgrip increased blood pressure and Ea in both groups, whereas systemic vascular resistance remained unchanged during interventions (Table 2) Handgrip significantly increased LV pressure at every moment of diastole in both groups (Figure 1, Table 2 and see Supplementary material online, Table S1) ED volumes were similar between groups at rest, becoming significantly smaller during pacing (Table 2) Pacing shortened s and reduced prea pressures without any change in S þ and ED pressures (EDPs) in either group (Figure 1, Table 3 See Supplementary material online, Table S1) Determinants of diastolic pressures Early-diastolic pressures were determined in both groups by the equilibrium between positive relaxation-mediated pressure and negative volume related suction pressure mediated by S, the constant of elastic recoil (Figure 1 and Supplementary material online, Table S1) In patients with HFpEF, diastolic suction pressure at aortic valve closure, mitral valve opening, and P min was not different from control patients at baseline and was preserved during handgrip, regardless the increase in endsystolic volume (see Supplementary material online, Table S1) Late-diastolic filling pressures (pre-a and ED) were determined by the sum of positive relaxation-mediated pressure and positive volumerelated pressure mediated by S þ, the stiffness constant (Figures 1 and 2 and see Supplementary material online, Table S1) Irrelevant at baseline, relaxation-mediated pressure was a main determinant of pre-a pressure during handgrip and pacing in either population By ED, the effect of relaxation-pressure was negligible in either population (Figures 1 and 2 and see Supplementary material online, Table S1) Instead, EDPs were mostly stiffness-dependent pressures, mediated by S þ (Figures 1 and 2 and see Supplementary material online, Table S1) Effects of systolic pressure transients on diastolic properties In HFpEF patients, indices of diastolic chamber properties were highly sensitive to the changes in LVP max inducedbyhandgrip(tables 3 and 4), whereas they remained stable in the control population Particularly, the elastic recoil (S ) and stiffness constants (S þ ) were tightly related to LVP max in HFpEF patients (Table 4 and Figure 3) This relationship of diastolic properties with LVP max was much stronger than with end-systolic wall stress or indices of vascular load such as arterial elastance, pulse Downloaded from

5 910 C Pérez del Villar et al Figure 2 Chamber stiffening was a major determinant of late-diastolic pressures Incomplete relaxation was a significant contributor to pre-a pressure (immediately before atrial contraction), with a minor role in ED pressure during interventions Bars (mean and standard error bars) are stacked so that total height accounts for total estimated diastolic pressure Table 4 Relationship between peak LV pressure and diastolic chamber properties Control HFpEF b b 95% P-value Pseudo b b 95% P-value Pseudo Diastolic index (per mm Hg) Std CI R 2 (per mm Hg) Std CI R 2 s (ms) S - (mm Hg) < S þ (mm Hg) < V 0 (ml) Heart-rate adjusted linear-mixed models for repeated measures b, fixed-effect coefficient; b Std, standardized fixed-effect coefficient Pseudo-R 2, goodness-of-fit of multivariate models 95% CI, 95% Confidence interval Other abbreviations as in previous tables pressure, and systemic vascular resistance (see Supplementary material online, Table S2) The slopes of DS /DLVP max and DS þ /DLVP max were significantly higher in HFpEF patients (021 ± 029 and 036 ± 051 mm Hg per mm Hg) than in controls (-003 ± 024 and 004 ± 045 mm Hg per mm Hg, P =001andP = 002, HFpEF vs controls, respectively; Figures 3 and 4) Both slopes were non-significantly different from zero in the control population (P =066 and P = 074, for DS /DLVP max and DS þ / DLVP max, respectively) The significant changes in diastolic properties induced by interventions in HFpEF patients summarized in Table 3 became non-significant once adjusted for LVP max (P > 005 for all withinsubject effects of interventions when LVP max is entered as covariable) ECM and stiffness Compared with controls, HFpEF patients showed higher collagen volume fractions (16 ± 9% vs 5 ± 4%, P < 00001), and higher levels of crosslinking (17 ± 08% vs 09 ± 04%, P = 003) The DS /DLVP max slope significantly correlated with the degree of collagen crosslinking (R = 082, P = 0003), whereas the DS þ /DLVP max slope correlated with the collagen Downloaded from

6 Acute hypertension transients in HFpEF 911 Figure 3 Elastic recoil and chamber stiffness were sensitive to LV systolic pressure in HFpEF Acute changes in elastic recoil (S - )(Panels A and B) and stiffness constants (S þ )(Panels C and D) that take place in parallel to changes in LVP max during haemodynamic interventions in a representative control subject and an HFpEF patient volume fraction (R = 056, P =003; Figures 5 and 6) These correlations were higher than those observed with baseline values of S - (R = 030 and 008 against the collagen volume fraction and crosslinking, respectively) and S þ (R = 024 and 012 against the collagen volume fraction and crosslinking, respectively) Discussion In the present study, we demonstrate that formerly designated passive diastolic forces are highly dynamic in HFpEF patients, showing a tight interaction with systolic ventricular pressure Thus, beyond incomplete relaxation, acute chamber stiffening is a major mechanism responsible for the deleterious effects of acute hypertension in HFpEF Impact of blood pressure transients on diastolic function The fact that diastolic function is sensitive to acute changes in blood pressure was described decades ago studying the effects of angiotensin, 3 nitroprusside, 23 and handgrip 11 on the EDPVR However, due to different methodological limitations, a direct acute effect of hypertension on global chamber stiffness had never been unequivocally demonstrated as a major source of dynamic diastolic dysfunction In fact, single-beat studies have generally attributed acute modifications in the slope of the EDPVR to incomplete relaxation and/or increased external forces due to load mediated modifications of right-ventricular and intra-pericardial pressures 3,4,23 The few studies that used multiple-beat experiments during preload unloading to rule out the effect of external forces reached discordant conclusions Kamaguchi et al 6 and Wachter et al 24 Figure 4 The DS - /DLVP max (Panel A) andds þ /DLVP max (Panel B) slopes were steeper in the HFpEF group hypothesized that the slowing of relaxation was the most plausible mechanism of the increase in EDPs they induced by handgrip in HFpEF patients Westermann et al 15 failed to detect acute changes in chamber stiffness analysing PV data obtained during handgrip Remarkably, these two studies fitted the EDPVR assuming a non-physiological zero-volume asymptote In a murine model of HFpEF, the increase in filling pressures induced by isovolumic pressure transients was also ascribed to slowing relaxation 7 Closer to our findings, Shapiro et al 8 identified a reduction in ventricular distensibility taking place in parallel to the acute manipulation of arterial elastance in a dog model of acute over chronic pressure overload Albeit estimated non-invasively using ultrasound, an increase of stiffness has also been recently suggested as a mechanism of exercise intolerance in some HFpEF patients 25 Using state-of-the art techniques, we demonstrate that the strong relationship between peak systolic pressure and intrinsic global LV diastolic chamber stiffness is the main mechanism responsible for acute elevation of ED pressure caused by hypertension transients in patients with HFpEF Relaxation is slowed by hypertension Consequently, a remarkable increase of relaxation-mediated pressure significantly contributes to the increase in pre-a pressure during handgrip However, the raise in ED pressure driven by hypertension transients, was almost exclusively caused by acute chamber stiffening Thus, we believe that the strong interaction between LVP max and stiffness explains most of previously described phenomena of diastolic ventricular-vascular coupling 3,6,8,11,23,26 Potential mechanisms of acute changes in diastolic function At the tissue level, a number of molecular mechanisms are able to account for the acute changes in volume-mediated diastolic properties in HFpEFweobservedinourstudyTitinstructureanditsdynamicphosphorylation at low slack-lengths, 5 titin actin interactions, abnormal resurgence of cross bridge interactions during filling, 5,27 and ECM Downloaded from

7 912 C Pérez del Villar et al Figure 5 The DS - /DLVP max (Panels A and B)andDS þ /DLVP max (Panels C and D) slopes correlated with matrix remodeling Bivariate associations in the subgroup of subjects undergoing myocardial biopsies composition at high slack lengths, are all well known determinants of myocardial stiffness Importantly, the systolic pressure dependence of V 0 may reflect true changes in slack-length, unapparent from end diastolic volume measurements Because we did not undergo deep myocyte molecular analyses, the mechanisms responsible for the hypertension induced stiffening response deserve further clarification in well controlled animal experiments At the chamber level, stiffness is also known to be highly dynamic and sensitive to overall ventricular geometry and septal conformation We have recently shown in the right ventricle, that the equilibrium volume is related to the degree of septal bulging caused by ventricular interdependence during early diastole 13 A similar phenomenon of septal displacement at zero volume-mediated pressure might be responsible for the relationship between LVP max and equilibrium volume changes we found in HFpEF patients in the present study Hypothetically, the interventricular septum could act as a functional reservoir, lowering stiffness until it is fully distended towards the contralateral ventricle Once this buffering reserve disappears due to an increase in the equilibrium volume, chambers stiffness would rise rapidly, at this point becoming directly conditioned by the material properties of the ECM To our knowledge, this is the first clinical study to characterize elastic restoring forces in HFpEF As opposed to the stiffness constant, the constant of elastic recoil was not significantly different in HFpEF subjects than in controls at baseline Consequently, volume-mediated earlydiastolic pressures, were also similar in both groups However, as opposed to controls, HFpEF patients increased their constant of elastic recoil in response to an acute increase in blood pressure By intensifying diastolic suction, this positive DS /DLVP max slope facilitates early filling in HFpEF patients Unfortunately, parallel changes on the stiffness constant outweigh this initial beneficial effect on early diastolic function and are the main mechanism responsible for the elevation of late-diastolic pressures induced by hypertension Clinical implications Vigorous vasodilator treatment should be advocated to treat acute hypertensive exacerbations of HFpEF In addition, a new field of pharmacological research should be targeted towards this acute chamber stiffening response, whereas strategies to improve relaxation or slow heart rate are probably less useful The diagnostic workup of frequently coexistent conditions of borderline significance (eg low-gradient aortic valve stenosis) 28 should always rule-out acute hypertensive responses as a potential cause of exercise-related symptoms which can be readily recognized and treated Values of systolic pressure must be taken into account when measuring, comparing, and reporting indices of diastolic chamber properties Finally, the highly dynamic nature of intrinsic diastolic properties emphasizes the need for stress interventions to accurately characterize diastolic function in the clinical setting 25 Downloaded from

8 Acute hypertension transients in HFpEF 913 Figure 6 Relationships between matrix remodelling, hypertensive transients, and ED pressure Haemodynamic and histological data from a representative control subject (Panels A C) and an HFpEF patient (Panels D F) A higher hypertensive response in the HFpEF subject is followed by an immediate increase in the slope of the volume-dependent component of the PV curve and is associated to a higher collagen content of the ECM Panels A and D: diastolic pressures at baseline (red), during handgrip (green) and atrial pacing (blue); pressure time plots are shown in the insert Panels B and E: recoil/stiffness (volume-dependent) PV curves (solid lines) from end-systolic to ED volume overlaid on the full characteristic curve (dotted lines) Baseline PV loops are shown in the inserts Panels C and F: collagen-specific Sirius red staining from endomyocardial biopsies obtained in the same subjects Limitations A number of limitations related to our HFpEF group need to be taken into account The referral for coronary angiography probably biased our population towards a younger age and more frequently male than typically found in epidemiological studies In addition, although patients with significant epicardial coronary artery disease were excluded, referral for angiography may lead additional uncontrolled confusion factors such as small vessel-mediated myocardial ischaemia Low filling pressures in HFpEF patients were caused by intensive treatment before cardiac catheterization, frequently using intravenous diuretics For this reason, baseline EDPs were only mildly elevated in the HFpEF group In addition, anti-hypertensive medical treatment was not standardized, and data on medical treatment at the time of the examinations was unavailable Using our data-processing method, we obtained s values shorter than typically found in HFpEF cohorts, probably due to methodological issues or less advanced disease states We analysed haemodynamic responses to handgrip and pacing because dynamic exercise is incompatible with simultaneous left and right catheterization during inferior vena cava occlusion Since we did not analyse titin and microvascular function, the molecular mechanisms responsible for acute chamber stiffening could not be completely deciphered Conclusions Acute dependence of global LV chamber stiffness on systolic arterial pressure plays a major role in the pathophysiology of HFpEF Beyond incomplete relaxation, the deleterious effect of hypertensive episodes on late-diastolic pressures is mainly caused by acute chamber stiffening and not by external forces, simultaneous tachycardia, or chamber dilatation Downloaded from

9 914 C Pérez del Villar et al This acute stiffening response correlates with the degree of matrix remodelling in patients with HFpEF Supplementary material Supplementary material is available at Cardiovascular Research online Conflict of interest: none declared Funding This work was supported by the Instituto de Salud Carlos III, Ministerio de Economıa y Competitividad, Spain, and by the EU European Regional Development Fund [RD12/0042 (Red de Investigacion Cardiovascular), PI12/02885 to JB, CM12/00273 and JR15/00039 to CPV] CPV was partially supported by a grant from the Fundacion para la Investigacion Biomédica Gregorio Mara~non, Spain References 1 Paulus WJ, Tschope C A novel paradigm for heart failure with preserved ejection fraction: comorbidities drive myocardial dysfunction and remodeling through coronary microvascular endothelial inflammation J Am Coll Cardiol 2013;62: Kass DA, Bronzwaer JG, Paulus WJ What mechanisms underlie diastolic dysfunction in heart failure? Circ Res 2004;94: Alderman EL, Glantz SA Acute hemodynamic interventions shift the diastolic pressure-volume curve in man Circulation 1976;54: Borlaug BA, Jaber WA, Ommen SR, Lam CS, Redfield MM, Nishimura RA Diastolic relaxation and compliance reserve during dynamic exercise in heart failure with preserved ejection fraction Heart 2011;97: Bronzwaer JG, Paulus WJ Matrix, cytoskeleton, or myofilaments: which one to blame for diastolic left ventricular dysfunction? Prog Cardiovasc Dis 2005;47: Kawaguchi M, Hay I, Fetics B, Kass DA Combined ventricular systolic and arterial stiffening in patients with heart failure and preserved ejection fraction: implications for systolic and diastolic reserve limitations Circulation 2003;107: Leite S, Rodrigues S, Tavares-Silva M, Oliveira-Pinto J, Alaa M, Abdellatif M, Fontoura D, Falcao-Pires I, Gillebert TC, Leite-Moreira AF, Lourenco AP Afterload-induced diastolic dysfunction contributes to high filling pressures in experimental heart failure with preserved ejection fraction Am J Physiol Heart Circ Physiol 2015;309: H1648 H Shapiro BP, Lam CS, Patel JB, Mohammed SF, Kruger M, Meyer DM, Linke WA, Redfield MM Acute and chronic ventricular-arterial coupling in systole and diastole: insights from an elderly hypertensive model Hypertension 2007;50: Gandhi SK, Powers JC, Nomeir AM, Fowle K, Kitzman DW, Rankin KM, Little WC The pathogenesis of acute pulmonary edema associated with hypertension N Engl J Med 2001;344: Dauterman K, Pak PH, Maughan WL, Nussbacher A, Arie S, Liu CP, Kass DA Contribution of external forces to left ventricular diastolic pressure Implications for the clinical use of the Starling law Ann Intern Med 1995;122: Rommel KP, von Roeder M, Latuscynski K, Oberueck C, Blazek S, Fengler K, Besler C, Sandri M, Lucke C, Gutberlet M, Linke A, Schuler G, Lurz P Extracellular volume fraction for characterization of patients with heart failure and preserved ejection fraction J Am Coll Cardiol 2016;67: Bermejo J, Yotti R, Perez del Villar C, del Alamo JC, Rodriguez-Perez D, Martinez- Legazpi P, Benito Y, Antoranz JC, Desco MM, Gonzalez-Mansilla A, Barrio A, Elizaga J, Fernandez-Aviles F Diastolic chamber properties of the left ventricle assessed by global fitting of pressure-volume data: improving the gold standard of diastolic function J Appl Physiol 2013;115: Perez Del Villar C, Bermejo J, Rodriguez-Perez D, Martinez-Legazpi P, Benito Y, Antoranz JC, Desco MM, Ortuno JE, Barrio A, Mombiela T, Yotti R, Ledesma- Carbayo MJ, Del Alamo JC, Fernandez-Aviles F The role of elastic restoring forces in right-ventricular filling Cardiovasc Res 2015;107: Kasner M, Westermann D, Lopez B, Gaub R, Escher F, Kuhl U, Schultheiss HP, Tschope C Diastolic tissue Doppler indexes correlate with the degree of collagen expression and cross-linking in heart failure and normal ejection fraction J Am Coll Cardiol 2011;57: Westermann D, Kasner M, Steendijk P, Spillmann F, Riad A, Weitmann K, Hoffmann W, Poller W, Pauschinger M, Schultheiss HP, Tschope C Role of left ventricular stiffness in heart failure with normal ejection fraction Circulation 2008;117: Westermann D, Lindner D, Kasner M, Zietsch C, Savvatis K, Escher F, von Schlippenbach J, Skurk C, Steendijk P, Riad A, Poller W, Schultheiss HP, Tschope C Cardiac inflammation contributes to changes in the extracellular matrix in patients with heart failure and normal ejection fraction Circ Heart Fail 2011;4: Matsubara H, Takaki M, Yasuhara S, Araki J, Suga H Logistic time constant of isovolumic relaxation pressure-time curve in the canine left ventricle Better alternative to exponential time constant Circulation 1995;92: Burkhoff D, Mirsky I, Suga H Assessment of systolic and diastolic ventricular properties via pressure-volume analysis: a guide for clinical, translational, and basic researchers Am J Physiol Heart Circ Physiol 2005;289:H501 H Zile MR, Gaasch WH, Levine HJ Left ventricular stress-dimension-shortening relations before and after correction of chronic aortic and mitral regurgitation Am J Cardiol 1985;56: Querejeta R, Lopez B, Gonzalez A, Sanchez E, Larman M, Martinez Ubago JL, Diez J Increased collagen type I synthesis in patients with heart failure of hypertensive origin: relation to myocardial fibrosis Circulation 2004;110: Lopez B, Querejeta R, Gonzalez A, Beaumont J, Larman M, Diez J Impact of treatment on myocardial lysyl oxidase expression and collagen cross-linking in patients with heart failure Hypertension 2009;53: Nakagawa S, Schielzeth H A general and simple method for obtaining R2 from generalized linear mixed-effects models Methods Ecol Evol 2013;4: Brodie BR, Grossman W, Mann T, McLaurin LP Effects of sodium nitroprusside on left ventricular diastolic pressure-volume relations J Clin Invest 1977;59: Wachter R, Schmidt-Schweda S, Westermann D, Post H, Edelmann F, Kasner M, Luers C, Steendijk P, Hasenfuss G, Tschope C, Pieske B Blunted frequencydependent upregulation of cardiac output is related to impaired relaxation in diastolic heart failure Eur Heart J 2009;30: Kasner M, Sinning D, Lober J, Post H, Fraser AG, Pieske B, Burkhoff D, Tschope C Heterogeneous responses of systolic and diastolic left ventricular function to exercise in patients with heart failure and preserved ejection fraction ESC Heart Fail 2015;2: Chung CS, Strunc A, Oliver R, Kovacs SJ Diastolic ventricular-vascular stiffness and relaxation relation: elucidation of coupling via pressure phase plane-derived indexes Am J Physiol Heart Circ Physiol 2006;291:H2415 H Kouzu H, Miki T, Tanno M, Kuno A, Yano T, Itoh T, Sato T, Sunaga D, Murase H, Tobisawa T, Ogasawara M, Ishikawa S, Miura T Excessive degradation of adenine nucleotides by up-regulated AMP deaminase underlies afterload-induced diastolic dysfunction in the type 2 diabetic heart J Mol Cell Cardiol 2015;80: Perez Del Villar C, Yotti R, Espinosa MA, Gutierrez-Ibanes E, Barrio A, Lorenzo MJ, Sanchez Fernandez PL, Benito Y, Prieto R, Perez David E, Martinez-Legazpi P, Fernandez-Aviles F, Bermejo J The functional significance of paradoxical low gradient aortic valve stenosis: Hemodynamic findings during cardiopulmonary exercise testing JACC Cardiovasc Imaging 2017;10:29 39 Downloaded from