Aortic flow is abnormal in HFpEF

Aims Turbulent aortic flow makes the cardiovascular system less effective. It remains unknown if patients with heart failure with preserved ejection fraction (HFpEF) have disturbed aortic flow. This study sought to investigate advanced markers of aortic flow disturbances in HFpEF. Methods This case-controlled observational study used four-dimensional flow cardiovascular magnetic resonance derived, two-dimensional phase-contrast reformatted plane data at an orthogonal plane just above the sino-tubular junction. We recruited 10 young healthy controls (HCs), 10 old HCs and 23 patients with HFpEF. We analysed average systolic aortic flow displacement (FDsavg), systolic flow reversal ratio (sFRR) and pulse wave velocity (PWV). In a sub-group analysis, we compared old HCs versus age-gender-matched HFpEF (N=10). Results Differences were significant in mean age (P<0.001) among young HCs (22.9±3.5 years), old HCs (60.5±10.2 years) and HFpEF patients (73.7±9.7 years). FDsavg, sFRR and PWV varied significantly (P<0.001) in young HCs (8±4%, 2±2%, 4±2m/s), old HCs (16±5%, 7±6%, 11±8m/s), and HFpEF patients (23±10%, 11±10%, 8±3). No significant PWV differences existed between old HCs and HFpEF.HFpEF had significantly higher FDsavg versus old HCs (23±10% vs 16±5%, P<0.001). A FDsavg > 17.7% achieved 74% sensitivity, 70% specificity for differentiating them. sFRR was notably higher in HFpEF (11±10% vs 7±6%, P<0.001). A sFRR > 7.3% yielded 78% sensitivity, 70% specificity in differentiating these groups. In sub-group analysis, FDsavg remained distinctly elevated in HFpEF (22.4±9.7% vs 16±4.9%, P=0.029). FDsavg of >16% showed 100% sensitivity and 70% specificity (P=0.01). Similarly, sFRR remained significantly higher in HFpEF (11.3±9.5% vs 6.6±6.4%, P=0.007). A sFRR of >7.2% showed 100% sensitivity and 60% specificity (P<0.001). Conclusion Aortic flow haemodynamics namely FDsavg and sFRR are significantly affected in ageing and HFpEF patients.

contrast reformatted plane data at an orthogonal plane just above the sino-tubular junction.We recruited 10 young healthy controls (HCs), 10 old HCs and 23 patients with HFpEF.We analysed average systolic aortic flow displacement (FDsavg), systolic flow reversal ratio (sFRR) and pulse wave velocity (PWV).In a sub-group analysis, we compared old HCs versus age-gender-matched HFpEF (N=10).

Introduction
Heart failure with preserved ejection fraction (HFpEF) is a heterogeneous clinical syndrome in which patients have signs and symptoms of heart failure.The diagnosis of HFpEF has a significant negative impact on quality of life and a prognostic outcome comparable to those with heart failure with reduced ejection fraction (HFrEF) 1,2 .The hallmark pathophysiology of HFpEF is high left ventricular (LV) filling pressure (LVFP) despite normal or near normal LV ejection fraction (LVEF; ≥50 per cent) 3 .The epidemiology of heart failure with preserved ejection fraction (HFpEF) is constantly evolving as the condition continues to be a significant global health concern, with a prevalence of 2-3% in the general population and up to 50% in the elderly population 2,4 .There is a higher prevalence of hypertension, atrial fibrillation (AF), coronary artery disease (CAD), dyslipidemia, obesity, anaemia, diabetes mellitus, chronic kidney disease, and sleep-disordered breathing in these patients 2,5,6 .
The aorta is subject to unique and complex flow dynamics, characterised by high flow rates, extreme pressure variations, and intricate flow patterns in both physiological and pathological states 7 .The relationship between arterial stiffness and LV diastolic function is well established [8][9][10] .Numerous studies have demonstrated and established quantitative aortic flow parameters such as flow displacement (FD) and flow reversal ratio (FRR) in pathological states such as aortopathy and aortic valve disease 7,[11][12][13] .Elevated systolic flow displacement (FDsavg), a marker of aortic flow eccentricity, causes turbulence within the ascending aorta.Any increase in FD leads to a rise in energy dissipation, which in turn reduces the efficiency of the cardiovascular system.
Additionally, any increase in systolic flow reversal ratio (sFRR) in the ascending aorta causes a loss of forward flow which is detrimental to the aortic conduit function, directly resulting in reduced peripheral perfusion and tissue oxygenation 7 .This intricate interplay between aortic haemodynamics and left ventricular function, referred to as ventricular-arterial coupling [14][15][16][17] , remains insufficiently researched.Even though specific aetiological factors of HFpEF, for example, hypertension or diabetes, have been associated with aortic stiffness, it remains unknown if patients with HFpEF have aortic flow abnormalities described above, which can result in heightened cardiac workload, decreased cardiac output and compromised distal perfusion subsequently causing shortness of breath.
We hypothesise that patients with HFpEF have signs of abnormal aortic flow which compromises the aortic conduit function and results in more energy expenditure, making the cardiovascular system less efficient.Hence, the main objective of this study was to investigate aortic flow haemodynamics utilising four-dimensional (4D) flow cardiovascular magnetic resonance (CMR) imaging in patients with HFpEF and healthy controls.

Patient and Public Involvement
The engagement of patients and the public was initiated at the project's inception through Norfolk and Suffolk Primary and Community Care Research Office (https://nspccro.nihr.ac.uk/ working-with-us/public-patient-and-carer-voice-in-research).
The PPI panel helped to make the study protocol patient friendly.PPI group provided insight into design of patient information sheet for the study.PPI group were in agreement that the research will produce open access research papers available to all to read.

Study cohort
We identified patients from the PREFER-CMR registry (Clini-calTrials.gov:NCT05114785).We enrolled 20 subjects into the healthy control (HC) group.The main inclusion criteria for the HCs were: > 18 years of age (<30 years for young HCs and, >50 years for old HCs) and no prior history of cardiovascular disease.We enrolled 23 patients with HFpEF.The main inclusion criteria for patients were over 18 years of age and a confirmed clinical diagnosis of HFpEF by clinical history of symptoms and signs, including CMR features of HFpEF (mainly raised left ventricular filling pressure >15mmHg) 18 .
The exclusion criteria were limited to any CMR contraindication or fast atrial fibrillation (heart rate > 100 bpm) and high R-R variability.

Amendments from Version 1
The important changes made in this version include: 1. We have clarified how the diagnosis of HFpEF was made.
2. We have added internal validation of the agreement between 2D phase-contrast flow and 4D flow methods

3.
We have corrected some typos and grammar.
Our results have not been changed or impacted and the message of the paper remains the same.
Any further responses from the reviewers can be found at the end of the article

Ethics approval and consent to participate
This study was conducted according to the principles outlined in the Declaration of Helsinki -Version 2013.The collection and management of data were approved by the National Research Ethics Service in the United Kingdom (21/NE/0149).
A pragmatic opt-out informed consent was obtained from all subjects included in the study 19 .
Cardiovascular Magnetic Resonance protocol CMR study was performed on a 1.5 Tesla Magnetom Sola Siemens system with a superconducting magnet (Siemens Healthineers AG, Erlangen, Germany).All patients were examined in the supine position, headfirst, using a respiratory sensor and electrocardiogram gating.Additionally, the scanner was equipped with a Biometric body with 18 coils.

CMR analysis
All image analyses were post-processed with the in-house developed MASS research software (MASS; Version 2022-EXP, Leiden University Medical Center, Leiden, The Netherlands).A static reformatted plane was planned through the ascending aorta at the mid-main pulmonary artery level to generate a through-plane velocity encoded two-dimensional (2D) phase contrast (PC) data using 4D flow CMR data.This plane was treated as a two-dimensional phase contrast plane.Ascending aortic helical flow was defined as the flow swirling around the aortic centre line.Ascending aorta vortex flow was defined as any flow rotating on the long axis of the aorta near the inner curvature of the aortic root 26 (Figure 1).The following parameters we automatically derived based on the aortic contours: • Aortic Forward Flow: This refers to the stroke volume during a cardiac cycle.
• Aortic maximum and minimum area are the largest and smallest cross-sectional area (respectively) computed in the ascending aorta during a cardiac cycle.
• Flow displacement systolic average (FDsavg): This is the distance between the vessel's central point and the centre-of-velocity of the forward flow, normalised to the vessel size during systole.It is presented as a percentage.The centre-of-velocity of the forward flow is computed as the mean location of pixels weighted by the velocity values within a defined aortic contour on a 2D Phase Contrast (PC) image.
• Rotational Angle signifies the angle formed by the line connecting the entre of forward flow velocity and the radius pointing at 12 o'clock position within the ascending aorta cross-section on every 2D PC image during cardiac cycle.RA was established at zero when the vector pointed at 12 o'clock and progressed clockwise to 180 degrees when pointing backward (6 o'clock position).This is sensitive to errors in negligible flow displacement as the position of the centre-of-velocity and vessel are at a close proximity.Therefore, a slight modification of the aortic contour can displace the location of the centre-of-velocity flow and significantly alter RA.
To circumvent this effect, we chose a FD=12% threshold based on our bench testing.
• Systolic Forward and Backward Flows: These are obtained using per-pixel information.All positive velocities within the region of interest during systole were used to derive systolic forward flow.In contrast, all negative velocities within the same region were used to derive systolic negative flow.
• Systolic Flow Reversal Ratio Flow curves generated for each cardiac cycle.sFRR= systolic retrograde flow/systolic forward flow x 100%.
• Pulse Wave Velocity is the ratio of distance and transit time between ascending to descending aorta.A 3D image is created from parallel scout images of ascending, arch and descending aorta.2D PC aortic flow plane matched to 3D images noting exact velocity measure points.Arch length is traced between these points.Transit time, calculated as the time difference between the points where the ascending and descending aorta flow waveforms reached half maximum of their peak velocities.

Statistical analysis
Data analyses were performed using MedCalc® Statistical Software, version 20.215 (MedCalc Software Ltd., Ostend, Belgium) and OriginPro, version 2023 (OriginLab Corporation, Northampton, MA, USA).Continuous variables are presented as the median along with the interquartile range (IQR).We treated data as non-parametric.To compare variables between the two groups, we employed the Mann-Whitney U test.For testing across three groups, we utilised the Kruskal-Wallis one-way ANOVA and followed it with a post-hoc analysis using the Conover method.To explore the independent association of all CMR indices, we used a partial correlation factor generated via the multiple regression method using all other CMR variables as covariates.To evaluate the ability of these blood flow parameters to distinguish HFpEF, we conducted Receiver Operator Characteristic (ROC) analysis and used Youden Index to determine cut-off values.We deemed statistical significance at a threshold of P<0.05.

Study size
We used healthy young versus old systolic flow displacement data to derive the possible sample size to differentiate healthy old versus HFpEF cohorts.Factoring a mean difference of 8% and standard deviations of 4 and 5, to achieve an alpha of 0.05, we need to recruit at least 11 HFpEF patients and 6 elderly patients with a total sample size of 17 to achieve a power of beta=0.20.
To further improve the diagnostic range, we planned to recruit >20 HFpEF cases and at least 10 elderly healthy subjects.
Both the old controls and patients with HFpEF exhibited a notable decrease in aortic forward flow in comparison to young controls (P=0.02)independent of gender and despite no significant variance in the left ventricular ejection fraction (Table 2).There was a significant rise in FDsavg in old controls and HFpEF in contrast to young controls with marked differences observed between these two groups as well (16±5% vs 23±10%, P<0.001).Likewise, the sFRR significantly differed in all three groups (P<0.001).Importantly, individuals   with HFpEF exhibited significantly higher sFRR compared to old controls (11±10% vs 7±6%, P<0.001).These differences in FDsavg and sFRR in these two groups persisted regardless of the gender.Although the mean values of pulse wave velocity (PWV) in both old HCs and HFpEF were significantly higher than in controls, no significant differences were observed between these groups.2).
On multiple regression analysis using Enter method aortic minimal area (partial R=0.45, P=0.01) and FDsavg (partial R=0.41, P=0.03) were the only two independent variables associated with age while factoring in all other CMR indices as covariates (Figure 3).

Aortic flow indices in Old HCs vs Age-gender-matched HFpEF
The differences in FDsavg and sFRR remained consistent and significant between HFpEF patients and old HCs when matched by age and gender.We observed a noticeable increase in FDsavg in age-gender matched HFpEF patients (N=10) vs old HCs (22.4±9.7% vs 16±4.9%,P=0.029) (Figure 5).Furthermore, sFRR was significantly elevated in age-gendermatched HFpEF patients vs old HCs (11.3±9.5% vs 6.6±6.4%,P=0.007).However, there was no significant difference observed in the indexed aortic forward flow, PWV and RA among these two groups.(Table 3).Two-dimensional phase-contrast versus fourdimensional phase-contrast reformatted plane The coefficient of variation (CoV) between 4D flow derived and two-dimensional phase-contrast derived was 9.6% for FDsavg and for sFRR CoV was 10.6% with a with-in subject variation of only 0.7% (P=0.26).

Discussion
This study is one of the first to explore shifts in aortic flow hemodynamics within the ageing population and patients with HFpEF.A significant finding of this study is that both ageing and HFpEF demonstrate a rise in the systolic flow reversal ratio, which is a marker of aortic conduit function.This aortic flow biomarker, which also represents turbulent flow, exhibited marked disparities between old control group and agegender-matched HFpEF patients.Moreover, FDsavg, an indicator of flow eccentricity in the ascending aortic root, increased progressively from young to old controls to HFpEF with marked differences between old controls and agegender-matched HFpEF patients.Aortic forward flow also decreased in all respective groups despite no significant difference in left ventricular ejection fraction.We observed that the associations of sFRR and FDsavg were aligned with ageing rather than LV functional parameters.These CMR-derived aortic flow biomarkers have high sensitivity to detect HFpEF early in the disease process and contribute to improved phenotyping.

Mechanism underpinning Systolic Flow Displacement
Increased flow eccentricity in the aorta causes turbulence resulting in energy dissipation, requiring the heart to expend more energy to maintain sufficient blood flow making the cardiovascular system inefficient 23,24,27,28 .The clinical utility of FDsavg as an independent biomarker has been established in predicting the rate of aortic growth in patients with bicuspid aortic valve (BAV) and aortopathy (with and without aortic valve disease) 7,[28][29][30][31] .We speculate that flow eccentricity, measured by FDsavg, develops due to left ventricular outflow tract remodelling and aortic remodelling due to an age-associated increase in afterload conditions on the ventricle [32][33][34] .

Mechanism underpinning Systolic Flow Reversal
Elevated sFRR, a marker of reduced aortic conduit function and increased vorticity, has been independently linked to aortic root dilatation 35 and shows a linear relationship with the severity of aortic stenosis 36 .Barker et al. 36 demonstrated an FRR of >10% in patients with BAV and significant aortic stenosis.The systolic flow reversal ratio measures the retrograde flow of vorticity and represents areas of low-pressure development near the inner curvature of the aortic root due to pressure equalisation, which will plausibly happen if left ventricular filling pressures are high or if, during systolic contraction, the left ventricle is not able to generate enough mechanical force 37 , for example, after myocardial infarction.These underlying pathophysiological processes have been described in previous computational fluid dynamics simulation studies 38 .In our HFpEF cohort, 48% of patients had systemic hypertension, and 35% had a previous myocardial infarction.
Our study is the first to show that abnormal aortic hemodynamics correlates with age and HFpEF without significant aortic valve disease.We observed that an sFRR level greater than 7.3% was found to have a sensitivity of 78% and specificity of 70% in distinguishing between old controls and HFpEF.Moreover, an FDs avg level greater than 18% had a sensitivity of 74% and specificity of 70% for the same differentiation.This study highlights that the mechanism of both flow eccentricity and flow turbulence is not only associated with the stenotic aortic valve..
With increase in both this aortic flow abnormalities, PVO2 decreases.The findings from their research, along with our study, support the notion that breathlessness in patients with HFpEF is correlated with aortic flow abnormalities.With the growing literature, FD and sFRR analysis can be applied to 2D phasecontrast imaging, extending their applicability to wider and more diverse populations to define this causation better.

Clinical impact
Current echocardiogram models to estimate left ventricular filling pressure remain complex and factor in right heart haemodynamics, including tricuspid regurgitation velocity, which will get affected in the end-stage of HFpEF.An increase in rightsided pressure will result after the whole pulmonary vascular bed has been adversely remodelled, leading to a late diagnosis of HFpEF.The utility of 4D Flow CMR to quantify aortic flow hemodynamics (FDsavg and sFRR) could aid in the early diagnosis of HFpEF and sub-phenotyping patients with HFpEF.This is important as it will allow early pharmacological intervention and improve clinical outcomes.Moreover, FDsavg and FRR can be easily applied to 2D phase contrast CMR, making them widely applicable for more extensive studies.Future studies are required to evaluate the direct link between aortic flow physiology and exercise capacity.We must also establish how aortic flow influences left ventricular filling and central aortic pressures.

Limitations
While our observational and case-controlled study suggests that there may be a link between aortic flow abnormalities with ageing and HFpEF, it is important to note that the sample size used in this study was relatively small, which limits the generalisability and calls for further research with larger and more diverse cohort to draw definitive conclusions.There is a possibility of selection bias in the HFpEF cohort, and we may be identifying individuals who are much further down the temporal changes of ageing associated with HFpEF.

Conclusion
Aortic flow haemodynamics, namely FDsavg and sFRR, are significantly affected in ageing and patients with HFpEF.
Incorporating aortic flow haemodynamics assessment in routine clinical practice may allow early and improved detection of HFpEF.

Ethics approval and consent to participate
This study was conducted according to the principles outlined in the Declaration of Helsinki -Version 2013.The collection and management of data were approved by the National Research Ethics Service in the United Kingdom (21/NE/0149).

Consent
All healthy controls gave informed consent.All patients were provided with clear information about the study.A pragmatic opt-out informed consent was obtained from all patients included in the study 20 . https://doi.org/10.21956/wellcomeopenres.22355.r72475 © 2024 Elamin N.This is an open access peer review report distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Nadir Elamin
Sheffield Teaching Hospital, Sheffield, UK This study is exceptional in its structure and presentation.The authors effectively introduced the topic of heart failure with preserved ejection fraction (HFpEF) and the significance of aortic flow.
The hypotheses and objectives of the study were clearly outlined, making it easy to understand the purpose of the research.To enhance the introduction further, it would be beneficial to elaborate slightly on the methods used to evaluate aortic flow and provide illustrations of why echocardiography may not be suitable for intricate flow assessment.
The methodology section is well-organised and clearly defined, allowing for seamless navigation between sections.The authors provided a comprehensive description of the study cohort and the standardised cardiovascular magnetic resonance (CMR) protocol, including image acquisition and data analysis.The different types of aortic flow assessments were also clearly defined in the methods section, and the statistical methods were well described, contributing to the reproducibility of the study.
The presentation of the results was clear and concise, with appropriate tables and figures used to support the findings.Nevertheless, incorporating a video comparing the aortic flow in the three groups would be beneficial.The authors effectively interpreted the results and integrated them with existing literature, highlighting the significance of aortic flow in HFpEF patients and older healthy control cohort.This advances our knowledge in both areas and provides opportunities for future research.
Overall, this study is of the highest quality, as evidenced by its coherent and well-organised structure.The paragraphs effectively summarised the objectives, and it is easy to create a list of bullet points from them.There are no grammatical or language errors that could impact the quality of the paper.This study will make a significant contribution to the literature on HFpEF patients, the innovative technology of CMR, and the importance of aortic flow in the aging population and patients with cardiovascular disease.

Is the work clearly and accurately presented and does it cite the current literature? Yes
Is the study design appropriate and is the work technically sound?Yes

Are sufficient details of methods and analysis provided to allow replication by others? Yes
If applicable, is the statistical analysis and its interpretation appropriate?
Yes Are all the source data underlying the results available to ensure full reproducibility?Yes Are the conclusions drawn adequately supported by the results?Yes Competing Interests: No competing interests were disclosed.
Reviewer Expertise: Anti-thrombotic therapy, Cardiovascular disease, Percutaneous coronary and structural intervention, Heart failure, Hamodynamic of the heart and vascular system.
I confirm that I have read this submission and believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.

Figure 1 .
Figure 1.Central illustration of case examples from the study cohort.A and B - Flow mapping demonstrates normal predominantly laminar flow at peak systole in healthy individual with minimal flow reversal during systolic phases.C and D - Flow mapping a HFpEF patient demonstrating significant shift in flow displacement at peak systole.There is clear flow reversal observed (blue zone) at peak systole which is affecting aortic forward flow.There is quantified by FRR which is 17.5% during systole and can be noted on backward flow mapping (blue line) on the aortic flow curves.

Figure 2 .
Figure 2. A -Scatter plot with 95% confidence interval illustrating a direct correlation between age and systolic flow reversal ratio (P<0.001).B -Scatter plot with 95% confidence interval illustrating a linear correlation between age and flow displacement systolic average (P<0.001).

Figure 3 .
Figure 3. Radial bar plot illustrating a linear partial independent correlation of age with left ventricular and aortic flow parameters.Aortic minimum area (AO min area) and FDsavg have independent association to age of the whole study cohort and both increase with age.

Figure 4 .
Figure 4.A A -Bar charts demonstrating flow displacement systolic average trends in young vs old healthy cohorts vs HFpEF patients.B -Bar charts demonstrating systolic flow reversal ratio trends in young and old healthy cohorts and HFpEF patients.C and D -Receiver operating characteristic (ROC) with area under the curve demonstrating acceptable correlation in old HCs vs patients with HFpEF.

Figure 5 .
Figure 5.A -Bar charts demonstrating average systolic flow displacement trends in old healthy cohorts and age-gender-matched HFpEF patients.B -Bar charts demonstrating systolic flow reversal ratio trends in old healthy cohorts vs age-gender-matched HFpEF patients.C and D -Receiver operating characteristic (ROC) with area under the curve demonstrating acceptable correlation in old HCs vs patients with age-gender-matched HFpEF.

Table 2 . Aortic flow indices trend across the three groups. Young HCs Old HCs HFpEF P Number recruited 10 10 23 Aortic flow parameters
Data were represented as median ± IQR (%).AO aorta, HCs healthy cohorts, HFpEF heart failure with preserved ejection fraction.*P<0.05 compared young HCs versus old HCs; # P<0.05 compared old HCs versus HFpEF.
Nevertheless, by doing age-matched comparison, we still are able to demonstrate patho-physiological step-up changes in different ages and diseased state.With no intervention currently available to improve aortic flow hemodynamics, our study is limited in establishing causality and restricts observations to only correlations.It is possible that other prevalent systemic risk factors in HFpEF, such as hypertension, obstructive sleep apnoea (OSA), diabetes, and obesity, could also contribute to changes in flow displacement and sFRR.Therefore, caution must be exercised when interpreting these findings as they may not represent the boarder population with HFpEF and age-related changes in the cardiovascular system.