Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Aug 1.
Published in final edited form as: Clin Imaging. 2025 May 29;124:110524. doi: 10.1016/j.clinimag.2025.110524

4D Flow MRI Derived Aortic Pulse Wave Velocity and Wall Shear Stress in Covid-19

Ozden Kilinc 1, Jay Bharat Bisen 1, Anthony Maroun 1, Ethan Johnson 1, Sandra Quinn 1, Daniel Kim 1,2, Daniel C Lee 3, Michael Markl 1,2, Bradley D Allen 1
PMCID: PMC12255154  NIHMSID: NIHMS2087244  PMID: 40460705

Abstract

Purpose:

This study aimed to investigate the changes in aortic pulse wave velocity (PWV) and wall shear stress (WSS) in COVID-19 using 4D Flow MRI.

Methods:

Thirty-seven COVID-19 patients and 37 healthy controls underwent thoracic cardiovascular MRI. The PWV and WSS comparisons were performed using independent t-test. Peak velocity (PV)-peak WSS correlations in patients; aortic dimension-regional WSS correlations; PWV-age correlations were reported using Pearson correlation coefficient (r) analysis.

Results:

The global aortic PWV was higher in the patient group (p = 0.007). There was a positive correlation between patient age and PWV values (r = 0.650, p = 0.000). The patient ascending aorta (AAo) WSS levels were lower in the entire cohort, in the subgroup of ages between 50–70, and in the age/gender matched subgroup (p < 0.05 for all). Voxelwise 5% PV was lower in the patient group (p = 0.005) and showed strong correlation with the 5% peak WSS (r = 0.957). In the patient group there was a negative correlation between the maximal aortic dimension and AAo WSS (r= −0.398, p = 0.014) and aortic arch WSS (r = −0.388, p = 0.017).

Conclusion:

The alterations to aortic stiffness in COVID-19 might be a late effect of the disease and should be confirmed in larger studies with longer follow-ups. The reasons behind the low AAo WSS levels in the COVID-19 group appears to be multifactorial and further work in larger cohorts eliminating the baseline aortic diameter and preexisting atherosclerotic risk factor differences is needed to validate our results and to establish reproducibility of the technique.

Keywords: Covid-19, 4D Flow MRI, Pulse Wave Velocity, Wall Shear Stress

INTRODUCTION

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the causative agent of coronavirus disease 2019 (COVID-19) and COVID-19 infection has been shown not to be restricted to respiratory tract infection only; conversely, it is considered as a multisystemic pathophysiological process and cardiovascular system involvement with possible long-term complications can also be seen (1–3). The clinical manifestations of the cardiovascular system involvement in COVID-19 are diverse and they can occur in the acute phase of the infection and may progress during the chronic phase secondary to the prolonged endothelial cell activation, enhanced thrombin generation and persistent inflammation which may result in increased vascular stiffness, lower systemic vascular function, large vessel vasculitis, and reduced vascular dilation (1,3–5). Vascular endothelial damage and increased inflammation have been shown to contribute to the pathophysiological process of the cardiovascular manifestations. However, the underlying pathogenic mechanism(s) of cardiac and vascular system manifestations are likely to be multifactorial in COVID-19 and are not well understood. Additionally, the severity of COVID-19 infection also correlates with the possibility of cardiovascular system involvement, therefore it is critically important to understand the interaction between the COVID-19 infection and pathogenic mechanism(s) of cardiovascular system complications including aortic involvement and the extent of vascular wall structural changes (1,3,6).

Arterial system aging accompanied by several structural vessel wall changes results in a gradual stiffening of the vasculature. This arterial stiffening is one of the major determinants of vascular health, is closely associated with cardiovascular diseases and can be accelerated by associated cardiovascular comorbidities (2,7). This gradual stiffening in the arterial system is accompanied by an increase in the speed of the arterial pressure wave traveling along arteries such as the aorta and can be measured by pulse wave velocity (PWV) which is the most widely recognized parameter to evaluate arterial stiffness (2,7). Several invasive or noninvasive techniques can be used to measure arterial PWV such as direct, invasive aortic PWV measurement using pressure catheter recordings, carotid-femoral PWV measurements, aortic PWV measurements using 2D phase-contrast magnetic resonance imaging (MRI) or three-dimensional, time-resolved phase contrast MRI with 3-directional velocity encoding (4D Flow MRI) (2,8,9). 4D Flow MRI enables measurement through retrospective analysis of different imaging planes and 3D volumetric quantification of aortic blood flow velocity. This technique has several advantages over other techniques to quantify PWV as reliable PWV quantifications require flow wave measurements over multiple planes along the 3D aorta which is unfeasible with a single 2D acquisition and would require several separate 2D acquisitions which may cause spatio-temporal misregistration and effect the accuracy of the PWV measurements (9).

The prevalent features of atherosclerosis such as hyperinflammation and endothelial damage also contribute disease severity and adverse outcomes in patients infected with COVID-19 linking these two pathological processes (10,11). Another hemodynamic parameter that can be measured through 4D Flow MRI is arterial wall shear stress (WSS) which is defined as frictional force generated by the flowing blood on the vessel walls, and which has been considered as an important influencer of the susceptibility to the atherosclerotic diseases (12–14). In this study we aimed to conduct comprehensive investigation of the changes to both aortic global PWV and aortic global and regional WSS in COVID-19 infection using 4D Flow MRI. We assess if the 4D Flow MRI derived global and regional aortic WSS is altered in patients after COVID-19 infection and if COVID-19 infection is associated with accelerated aortic stiffening using 4D Flow MRI derived aortic PWV.

MATERIALS AND METHODS

Study Cohort

Institutional Review Board approval was obtained for this Health Insurance Portability and Accountability Act (HIPAA) compliant study and all subjects including the patients and the volunteers provided written informed consent for their participation. Thirty-seven patients recently diagnosed with COVID-19 infection (56.2 +/− 14.6 year-old, 16 female) and 37 healthy controls with no known COVID-19 infection symptoms/diagnosis (46.7 +/− 11.4 year-old, 16 female) were prospectively recruited. The median age of the patients was 56 ranging from 22- to 83-year-old and the median age of the volunteers was 46 ranging from 21- to 82-year-old. Out of 37 COVID-19 patients, 36 patients were admitted in the hospital including 7 intensive care unit (ICU) admission and 1 patient was clinically managed as an outpatient. Of the 7 patients admitted in the ICU, 1 patient required endotracheal intubation secondary to respiratory failure. Six patients were diagnosed with diabetes mellitus, 2 patients were current smokers, 1 patient was a former smoker, and 15 patients were previously diagnosed with hypertension. Out of 37 patients, 1 patient was previously diagnosed with ascending aortic dilation, 1 with aortic ectasia, 1 with ascending aortic aneurysm and lastly 2 patients were previously diagnosed with heart failure. Average time elapsed between the initial COVID-19 diagnosis and 4D Flow MRI scan date was 184 +/− 7 days in the patient group (The median was 113 days ranging from 11 days to 791 days between the initial COVID-19 diagnosis and 4D Flow MRI scaan date).

Image Acquisition

All participants underwent thoracic cardiovascular MRI at 1.5 T MR-systems (Aera, Avanto, or Sola, Siemens Healthineers, Germany). Patients were scanned with a compressed sensing (acceleration factor, R = 7.7) accelerated coronal whole-chest 4D Flow MRI sequence using retrospective electrocardiogram gating without respiratory navigator gating. Volunteers were scanned with the traditional GRAPPA accelerated (R = 2) free-breathing 4D Flow MRI in sagittal oblique orientation covering the entire thoracic aorta and using prospective electrocardiogram and respiratory navigator-gating. The following 4D Flow MRI parameters were used for patient scans: mean spatial resolution: 2.49 × 2.49 × 2.55 mm3, field of view 160 mm2, mean slice thickness: 2.55 mm, echo time: 10.3 msec, velocity encoding (VENC): 120 cm/sec and flip angle: 15°. The scan parameters for the volunteers were as follows: mean spatial resolution: 2.61 × 2.61 × 2.89 mm3, mean field of view 136 mm2, mean slice thickness: 2.89 mm, mean echo time: 10.2 msec, velocity encoding (VENC): 150 cm/sec and flip angle: 15°.

Image Processing and Segmentation

The offline postprocessing steps of the 4D Flow MRI data for all acquisitions included automated phase-offset correction, noise masking of the areas outside of the flow regions, and velocity un-aliasing (MATLAB; TheMathWorks, Natick, MA) (15,16). This preprocessed data was used to generate time-averaged 3D phase contrast MR angiogram (PC-MRA) and time-averaged magnitude images to identify the aortic vessel anatomy and used for the manual 3D segmentation of the global and regional thoracic aorta based on the specific regions of interest including the ascending aorta (AAo) and aortic arch by one independent observer (AM, 3 years of experience in imaging research) on a designated commercial software (Mimics Innovation Suite; Materialise, Leuven, Belgium) in the volunteer group (17). The AAo was defined by placing a plane proximal to the brachiocephalic trunk and the aortic arch branch vessels were excluded from the segmentations. In the patient group, the global thoracic aorta segmentations were performed automatically on the designated software (MATLAB; TheMathWorks, Natick, MA) using an automated 3D U-Net network-based segmentation method as previously described (17) while regional segmentations for the AAo and the aortic arch were performed manually using the same method as in the volunteer group.

Aortic Global Pulse Wave Velocity Quantification

The aortic global PWV was quantified using an in-house automated PWV analysis algorithm (MATLAB; TheMathWorks, Natick, MA) (18). The aortic centerline generated from the aortic volume was used to place the orthogonal analysis planes every 4 mm along the aorta. The time-delay between flow waveforms was calculated by cross-correlating flow-time curves from these planes along the aortic centerlines. All analysis planes were used as a reference plane in the cross-correlation analysis repeating the PWV (1/slope) estimation multiple times for each scan to obtain the PWV estimate distribution. The global PWV estimate was taken as the mean distance/time value found from all analysis planes (Figure 1). (9,18,19). Analysis was performed by a clinical researcher (OK) with 4 years of imaging analysis experience.

Figure1.

Figure1.

Open in a new tab

Pulse wave velocity (PWV) quantification. The aortic centerline generated from the aortic volume was used to place the orthogonal analysis planes every 4 mm along the vessel. All analysis planes were used as a reference plane in the cross-correlation analysis repeating the PWV (1/slope) estimation multiple times for each scan to obtain the PWV estimate distribution. The global PWV estimate was taken as the mean distance/time value found from all analysis planes.

Aortic Global and Regional Wall Shear Stress and Peak Velocity Quantification

3D volumetric velocity maps were generated using in-house analysis tools (MATLAB; The MathWorks, Natick, MA) as previously described (20–22). Using spline interpolation, 4D flow velocity data were interpolated to 1 mm3 voxels and the 5th-percentile peak velocity (PV) was calculated for each voxel. The average of the maximum top 5% velocities was reported as global voxelwise PV in the entire thoracic aorta. The global thoracic aorta WSS was interpolated from the masked local velocities and reported as the average value of the top 5% percent during five systolic phases for each subject (MATLAB; TheMathWorks, Natick, MA) (12) (Figure 2). Similar WSS analysis was also performed regionally in the AAo and aortic arch separately and WSS was reported as the average number of the top 5% percent during five systolic phases.

Figure 2.

Figure 2.

Open in a new tab

Wall shear stress (WSS) quantification. The global thoracic aorta WSS was interpolated from the masked local velocities and reported as the average value of the top 5% percent during five systolic phases. Similar WSS analysis was also performed regionally in the ascending aorta and aortic arch separately and WSS was reported as the average number of the top 5% percent during five systolic phases.

Aortic Dimension Measurements

Mid-ascending aortic diameter was measured on the clinical images of the patients and volunteers obtained on the day of 4D Flow MRI scan by an experienced cardiothoracic radiology clinical fellow using a designated visualization and multiplanar reformation software (Visage 7, Visage Imaging, Inc., San Diego, CA, United States). The longest (maximal) mid-ascending aortic diameter was used in the correlation analysis with the WSS values.

Statistical Analysis

The cohort was analyzed as a whole and then stratified by specific age groups including the ages of 20–50 and 50–70, and age- and gender-matched subgroups. In the 20–50 subgroup, only patients between the ages of 20 and 50 were compared with healthy subjects between the ages of 20 and 50. The same method was applied to patients between the ages of 50 and 70 and healthy subjects between the ages of 50 and 70. In the last subgroup, patients of the same age and gender were matched with the healthy individuals and analyses were performed between them. For these groupwise PWV and WSS comparisons between the patients and volunteers in the entire cohort and for the specific subgroups, independent t-test was used. The correlation analysis between the 5% PV and 5% peak WSS in the whole thoracic aorta was also performed in the patient group using Pearson correlation coefficient (r) analysis. Similarly, the correlation levels between the aortic dimensions and the regional WSS results were also reported using the Pearson correlation coefficient analysis. Pearson correlation coefficient analysis results between aortic global PWV and patient age were also reported. A p-value < 0.05 was considered statistically significant.

RESULTS

Aortic Global Pulse Wave Velocity

In the overall cohort; aortic global PWV was significantly higher in the patient group (8.57 +/− 2.22 m/s) compared to the volunteers (7.29 +/− 1.57 m/s) (p = 0.007). However, no difference was observed between patients and volunteers in the specific age groups including 20–50 (6.85 +/− 1.03 vs. 6.82 +/− 1.05 m/s respectively), 50–70 (8.98 +/− 2.09 vs. 8.20 +/− 1.86 m/s respectively) and in the age/gender matched subgroup (7.54 +/− 1.60 vs. 7.20 +/− 1.30 m/s respectively) (p > 0.05 for all) (Table 1). There was also no significant difference in the PWV levels between patients with hypertension (n = 15, mean PWV +/− SD = 9.34 +/− 2.16 m/s) and without hypertension (n = 22, mean PWV +/− SD = 8.04 +/− 2.10 m/s) (p=0.084) and between ICU admitted patients (n = 7, mean PWV +/− SD = 8.68 +/− 2.44 m/s) and no-ICU admitted patients (n = 30, mean PWV +/− SD = 8.54 +/− 2.16 m/s) (p = 0.888) There was a significant positive correlation between the patient age and PWV results (r = 0.650, p = 0.000).

TABLE 1.

Global aortic pulse wave velocity comparisons using independent t-test

Age Group Patients Volunteers p value
n (Male) Age +/− SD PWV +/− SD (m/s) n (Male) Age +/− SD PWV +/− SD (m/s)
Entire Cohort 37 (21) 56.2 +/− 14.6 8.57 +/− 2.22 37 (21) 46.7 +/− 11.4 7.29 +/− 1.57 0.007
50–70 19 (11) 58.8 +/− 5.9 8.98 +/− 2.09 12 (6) 55.5 +/− 4.4 8.20 +/− 1.86 0.320
20–50 11 (6) 39.0 +/− 9.3 6.85 +/− 1.03 26 (15) 41.5 +/− 8.2 6.82 +/− 1.05 0.955
Age/Gender Matched 19 (11) 48.7 +/− 12.9 7.54 +/− 1.60 19 (11) 48.7 +/− 12.7 7.20 +/− 1.30 0.489
Open in a new tab

Aortic Global and Regional Wall Shear Stress and Peak Velocity

The WSS levels in the entire aorta and aortic arch were not significantly different between the patients and volunteers for the entire cohort or any of the three age-stratified subgroups (p > 0.05 for all). However, in the AAo, the WSS levels of the patients (0.916 +/− 0.303 Pa) were significantly lower in the entire cohort compared to the volunteers (1.195 +/− 0.212 Pa) (p < 0.001). Similarly, patients showed significantly lower WSS levels in the AAo in the subgroup of ages between 50 and 70 years (0.86 +/− 0.19 Pa) compared to the volunteers (1.210 +/− 0.127) (p < 0.001), and also in the age/gender matched subgroup (1.038 +/− 0.295 vs. 1.233 +/− 0.118 respectively, p = 0.013). In the subgroup of ages between 20 to 50, patients again had lower WSS values in AAocompared to the volunteers (1.107 +/− 0.362 Pa vs. 1.188 +/− 0.238 Pa respectively); however, the difference was not statistically significant (p > 0.05) (Table 2). In the whole cohort, voxelwise 5% PV in the entire aorta was significantly lower in the patient group compared to the volunteers (1.061 +/− 0.346 m/sec vs. 1.237 +/− 0.138 m/sec) (p = 0.005) (Figure 3A). Lastly the global aortic 5% PV and 5% peak WSS showed strong correlation in the whole patient group (r = 0.957) (Figure 3B).

TABLE 2.

Global and regional aortic wall shear stress level comparisons using independent t-test

Region Age Group Patients Volunteers p value
n (Male) Age +/− SD WSS +/− SD (Pa) n (Male) Age +/− SD WSS +/− SD (Pa)
Entire Aorta Entire Cohort 37 (21) 56.2 +/− 14.6 1.123 +/− 0.411 37 (21) 46.7 +/− 11.4 1.239 +/− 0.176 0.117
50–70 19 (11) 58.8 +/− 5.9 1.019 +/− 0.234 12 (6) 55.5 +/− 4.4 1.129 +/− 0.113 0.152
20–50 11 (6) 39.0 +/− 9.3 1.424 +/− 0.466 26 (15) 41.5 +/− 8.2 1.280 +/− 0.172 0.192
Age/Gender Matched 19 (11) 48.7 +/− 12.9 1.191+/− 0.331 19 (11) 48.7 +/− 12.7 1.248+/− 0.156 0.517
Aortic Arch Entire Cohort 37 (21) 56.2 +/− 14.6 0.956 +/− 0.425 37 (21) 46.7 +/− 11.4 0.958 +/− 0.217 0.973
50–70 19 (11) 58.8 +/− 5.9 0.887 +/− 0.299 12 (6) 55.5 +/− 4.4 0.893 +/− 0.105 0.949
20–50 11 (6) 39.0 +/− 9.3 1.242 +/− 0.533 26 (15) 41.5 +/− 8.2 1.003 +/− 0.232 0.073
Age/Gender Matched 19 (11) 48.7 +/− 12.9 1.020 +/− 0.321 19 (11) 48.7 +/− 12.7 0.967 +/− 0.125 0.514
Ascending Aorta Entire Cohort 37 (21) 56.2 +/− 14.6 0.916 +/− 0.303 37 (21) 46.7 +/− 11.4 1.195 +/− 0.212 0.000*
50–70 19 (11) 58.8 +/− 5.9 0.862 +/− 0.189 12 (6) 55.5 +/− 4.4 1.210 +/− 0.127 0.000*
20–50 11 (6) 39.0 +/− 9.3 1.107 +/− 0.362 26 (15) 41.5 +/− 8.2 1.188 +/− 0.238 0.436
Age/Gender Matched 19 (11) 48.7 +/− 12.9 1.038 +/− 0.295 19 (11) 48.7 +/− 12.7 1.233 +/− 0.118 0.013*
Open in a new tab
*

indicates statistical significance (p < 0.05)

Figure3.

Figure3.

Open in a new tab

A: Boxplot represents the global aortic voxelwise 5% peak velocity comparisons between the patient group and the volunteers B: Scatter plot represents the Pearson correlation coefficient (r) analysis results between the global aortic 5% peak velocity and global aortic 5% peak wall shear stress in the whole patient group (p < 0.05 indicates statistical significance).

Correlations Between the Regional Aortic Dimensions and Wall Shear Stress

In the patient group there was a negative correlation between the maximal mid-ascending aortic dimension and AAo WSS and aortic arch WSS (r= −0.398, p = 0.014 and r = −0.388, p = 0.017 respectively). The same negative correlations both for the AAo and aortic arch in the volunteer group were not statistically significant (r = −0.266 and −0.262 respectively and p > 0.05 for both). The mean maximal mid-ascending aortic dimension in the patient group was significantly higher compared to the volunteers (3.47 +/− 0.42 cm and 3.13 +/− 0.48 cm respectively, p = 0.02). The median maximal mid-ascending aortic dimension in the patient group was 3.50 cm ranging from 2.20 to 4.20 cm and the median maximal mid-ascending aortic dimension in the volunteer group was 3.20 cm ranging from 2.40 to 4.80 cm.

DISCUSSION

COVID-19 infection is known to contribute vascular stiffness and atherosclerotic process, however its impact on the aorta has not been well-described. We compared aortic PWV and WSS between COVID-19 infected patients and healthy controls. The AAo WSS levels were significantly lower in the patient population as a whole and also in the specific age, and age/gender matched subgroups. While the aortic global PWV was significantly higher in the entire patient group compared to the volunteers, we found no significant difference in PWV subgroup comparisons.

A considerable number of clinical studies have investigated the relationship between the COVID-19 infection and arterial stiffness using different techniques and several of these studies supported the importance of the arterial stiffening as a short-term prognostic factor or an independent risk factor for clinical deterioration in COVID-19 infection (23–26). Ratchford et al. used carotid-femoral pulse wave velocity as a marker of arterial stiffness and observed a 0.75-m/s higher PWV in the patient group, however they concluded that elevated level of the carotid-femoral PWV was within the expected range for the age-group in their cohort and may not be clinically relevant to cardiovascular risks (23). Stamatelopoulos et al. evaluated the prognostic role of the estimated PWV in the risk stratification of hospitalized patients with COVID-19 in their retrospective, longitudinal cohort study and found that higher estimated PWV provided incremental prognostic value for 28-day mortality in hospitalized COVID-19 patients and indicated that estimated PWV may be a cumulative death marker of the disease jointly reflecting the severity of a combination of multiple risk factors (25). Schnaubelt et al. compared brachial-ankle and carotid-femoral PWV of acutely ill patients with and without COVID-19 infection and concluded that COVID-19 was independently associated with higher carotid-femoral and brachial-ankle PWV as they found PWV values were higher in COVID-19 nonsurvivors, and correlations between the PWV levels and the length of hospital stay in the survivor group (26). Since there was a considerable difference in the mean ages when comparing the overall cohorts in our study, subgroup analyses were also performed. Even though the global PWV was significantly higher in the overall patient group and significant positive correlation between the patient age and PWV results was observed, no difference was seen in the specific age subgroups and in the age/gender matched subgroup. However, our patients were scanned an average of 184 +/− 7 days after the initial COVID-19 infection diagnosis and several of them had other comorbidities such as diabetes mellitus and high blood pressure. The subanalysis based on one of these comorbidities was performed between the patients with and without hypertension and also between the ICU-admitted patients and patients not admitted to the ICU, and no significant difference between the PWV values was observed. Even though the available literature data shows the relationships between the arterial stiffness and COVID-19 infection, the alterations to vascular physiology in COVID-19 disease might be a long-term effect of the disease. Additionally, the relationship between COVID-19 infection and arterial stiffness has been investigated using different techniques by many studies, no study used 4D Flow MRI based aortic PWV measurements to explore this relation. Therefore, future studies using our method should focus on long-term follow-up to determine if aortic stiffening is a long-term effect of COVID-19 infection and correlation reports between other cardiovascular disease comorbidities such as diabetes mellitus, hyperlipidemia and PWV levels would also be needed.

Arterial WSS has been studied in different cardiovascular disease processes, and the association between the WSS and atherosclerotic diseases has been well-established. However, the role of aortic WSS alterations quantified using 4D Flow MRI in COVID-19 infection has not been studied so far (12,27). It has long been considered that the low magnitude and/or high oscillatory WSS secondary to the cardiac cycle variations plays a critical role in the initial atherosclerotic plaque formation (27–29). Gibson et al. studied the local shear stress levels in human coronary arteries and found that low shear stress levels were significantly correlated with increased atherosclerosis progression in human coronary arteries (30). Similarly, Stone et al. studied the effect of WSS on human coronary plaque growth and found that low WSS is an independent predictor of increased plaque load with the lumen obstruction (31). On the other hand, Sastry et al. reviewed available studies on COVID-19 related thrombotic events and hypothesized that COVID-19 induced hyperviscosity leads to high WSS levels in arteries such as aorta, coronary arteries, and cerebral arteries and correspondingly disturbed endothelial homeostasis contributes to greater platelet activation and speeds up the arterial thrombosis (5). While in general, the low magnitude and high oscillatory WSS is considered proatherogenic and high WSS prevents the plaque growth in the initial atherosclerotic disease process, after the atherosclerotic plaques further progress, the high WSS may lead to the plaque phenotype that is vulnerable to rupture (14,28). However, the interactive relation between the hemodynamic factors involved in the atherosclerotic disease pathophysiology such as nonphysiological WSS levels and biological factors such as endothelial cell adaptation mechanisms to different WSS levels both play critical role and should be taken into consideration when interpreting the development of atherosclerotic disease process (14,27).

Additionally, WSS levels are closely linked to different factors such as age, body mass index, vessel wall diameter, blood pressure levels and may also vary along the vessel and around the different points of the vessel circumference. Even though, the mean maximal aortic dimension was significantly higher in the patient group compared to the volunteers in our study, the difference was in a clinically insignificant range.

Additionally, the significantly reduced levels of AAo WSS in the patient group were not only seen in the overall cohort but also in the subgroup of ages between 50 and 70 years and in the age/gender matched subgroup. The underlying mechanism for this difference is not clear. However, the blood flow pattern in the AAo is complex, and the curvature of the aortic arch and high flow pulsatility contribute to the helical flow pattern and flow reversal which is also affected by the disturbed WSS levels (27). It is clear that nonphysiological levels of the WSS are involved in the pathophysiology of the atherosclerotic disease process. However, it only acts as a link between disturbed biological vessel wall response to the atherosclerotic plaque formation, structural and functional integrity of endothelial cells and other systemic risk factors involved in the development of the atherosclerotic cardiovascular disease. Therefore, the complex and multifactorial etiopathogenesis of the atherosclerotic disease process can be explained only partially by the alterations to the physiological WSS levels. Further work exploring the differences in preexisting risks between COVID-19 patients and healthy controls, such as blood pressure at the time of scan, medication at the time of scan, smoking status is needed to better understand the atherosclerotic process and impacts of the disturbed WSS levels on this process seen in COVID-19 infection.

One limitation of our study is relatively small cohorts in both groups. Confirmation of our results in larger cohorts by also ensuring the average age consistency between the overall comparison cohorts with longer follow-ups and establishing reference normal ranges of AAo and aortic arch WSS levels in healthy controls using 4D Flow MRI are needed to explore COVID-19 impact on aortic growth or cardiovascular adverse events and to better understand how applicable and effective method 4D Flow MRI would be in the acute and long-term atherosclerotic disease prediction and diagnosis in COVID-19. There was also a difference in the methods used for the global thoracic aorta segmentations between the two cohorts in our study. Future studies using the same methodology for all segmentations would eliminate potential errors. While no impact on the PWV and WSS levels is expected depending on the image orientation and acceleration level differences, further studies using identical orientations and acceleration factors would increase the reliability and robustness of the study results.

In conclusion, our study highlights the feasibility of the 4D Flow MRI to quantify aortic PWV and WSS in COVID-19 infected patients. tFurther work using the same method in larger cohorts with longer follow-up is needed to validate these results and to establish reproducibility of the technique. Thus, 4D Flow MRI derived WSS measurements may offer a unique method in the screening of COVID-19 patients predisposed to the development of aortic stiffness and atherosclerosis and may be applied to other cardiovascular risk factors involved in the atherosclerosis etiopathogenesis.

Highlights:

  • This study investigated the applicability of the aortic pulse wave velocity and wall shear stress measurements using 4D Flow MRI in the assessment of altered aortic stiffness and increased atherosclerosis risk in COVID-19.

  • The ascending aorta wall shear stress levels were lower in the patient population. In the overall cohort; aortic global pulse wave velocity was significantly higher in the patient group compared to the volunteers; however, no difference was observed between patients and volunteers in the specific age groups including 20–50, 50–70 and in the age/gender matched subgroup.

  • The alterations to aortic stiffness in COVID-19 might be a late effect of the disease and should be confirmed in larger studies with longer follow-ups.

  • The reasons behind the low ascending aorta wall shear stress levels in the COVID-19 group appears to be multifactorial and further work in larger cohorts eliminating the baseline aortic diameter and preexisting atherosclerotic risk factor differences is needed to validate our results and to establish reproducibility of the technique.

ACKNOWLEDGEMENTS

Funding for this study was provided by the National Institutes of Health, USA (Grant Number: 3R01HL151079-01S1) and National Institutes of Health, USA/ National Heart, Lung, and Blood Institute (Grant Number: R01HL168700).

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Daniel Kim reports financial support was provided by National Institutes of Health. Bradley D. Allen reports financial support was provided by National Heart Lung and Blood Institute. Michael Markl reports a relationship with Third Coast Dynamics that includes: equity or stocks. Michael Markl reports a relationship with Siemens that includes: funding grants. Michael Markl reports a relationship with Circle Cardiovascular Imaging Inc that includes: funding grants. Bradley D. Allen reports a relationship with Third Coast Dynamics that includes: equity or stocks. Bradley D. Allen reports a relationship with Circle Cardiovascular Imaging Inc that includes: consulting or advisory. Bradley D. Allen reports a relationship with Siemens that includes: consulting or advisory. Bradley D. Allen reports a relationship with MRI Online that includes: consulting or advisory. Bradley D. Allen reports a relationship with National Institutes of Health that includes: funding grants. Bradley D. Allen reports a relationship with National Heart Lung and Blood Institute that includes: funding grants. Bradley D. Allen reports a relationship with The Ryan Family Acceleration Fund that includes: funding grants. Bradley D. Allen reports a relationship with Guerbet that includes: funding grants. Bradley D. Allen reports a relationship with Dixon Foundation that includes: funding grants. Bradley D. Allen reports a relationship with Siemens that includes: travel reimbursement. Bradley D. Allen reports a relationship with Burns White LLC that includes: paid expert testimony. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

ABBREVIATION LIST

AAo

Ascending Aorta

COVID-19

Coronavirus Disease 2019

ICU

Intensive Care Unit

MRI

Magnetic Resonance Imaging

PV

Peak Velocity

PWV

Pulse Wave Velocity

WSS

Wall Shear Stress

4D Flow MRI

Three-dimensional, time-resolved phase contrast MRI with 3-directional velocity encoding

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

References

  • 1.Martínez-Salazar B, Holwerda M, Stüdle C, et al. COVID-19 and the Vasculature: Current Aspects and Long-Term Consequences. Front Cell Dev Biol 2022;10:824851. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Zota IM, Stătescu C, Sascău RA, et al. Acute and Long-Term Consequences of COVID-19 on Arterial Stiffness—A Narrative Review. Life 2022;12(6):781. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Satterfield BA, Bhatt DL, Gersh BJ. Cardiac involvement in the long-term implications of COVID-19. Nat Rev Cardiol 2022;19(5):332–341. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Chung MK, Zidar DA, Bristow MR, et al. COVID-19 and Cardiovascular Disease: From Bench to Bedside. Circ Res 2021;128(8):1214–1236. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Sastry S, Cuomo F, Muthusamy J. COVID-19 and thrombosis: The role of hemodynamics. Thromb Res 2022;212:51–57. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Kang Y, Chen T, Mui D, et al. Cardiovascular manifestations and treatment considerations in COVID-19. Heart 2020;106(15):1132–1141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Sutton-Tyrrell K, Najjar SS, Boudreau RM, et al. Health ABC Study. Elevated aortic pulse wave velocity, a marker of arterial stiffness, predicts cardiovascular events in well-functioning older adults. Circulation 2005;111(25):3384–3390. [DOI] [PubMed] [Google Scholar]
  • 8.Hametner B, Wassertheurer S, Mayer CC, et al. Aortic Pulse Wave Velocity Predicts Cardiovascular Events and Mortality in Patients Undergoing Coronary Angiography: A Comparison of Invasive Measurements and Noninvasive Estimates. Hypertension 2021;77(2):571–581. [DOI] [PubMed] [Google Scholar]
  • 9.Jarvis K, Scott MB, Soulat G, et al. Aortic Pulse Wave Velocity Evaluated by 4D Flow MRI Across the Adult Lifespan. J Magn Reson Imaging 2022;56(2):464–473. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Sagris M, Theofilis P, Antonopoulos AS, et al. Inflammatory Mechanisms in COVID-19 and Atherosclerosis: Current Pharmaceutical Perspectives. Int J Mol Sci 2021;22(12):6607. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Vilaplana-Carnerero C, Giner-Soriano M, Dominguez À, et al. Atherosclerosis, Cardiovascular Disease, and COVID-19: A Narrative Review. Biomedicines 2023;11(4):1206. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Soulat G, Scott MB, Allen BD, et al. Association of Regional Wall Shear Stress and Progressive Ascending Aorta Dilation in Bicuspid Aortic Valve. JACC Cardiovasc Imaging 2022;15(1):33–42. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Kojima K, Hiro T, Koyama Y, et al. High Wall Shear Stress Is Related to Atherosclerotic Plaque Rupture in the Aortic Arch of Patients with Cardiovascular Disease: A Study with Computational Fluid Dynamics Model and Non-Obstructive General Angioscopy. J Atheroscler Thromb 2021;28(7):742–753. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Zhou M, Yu Y, Chen R, et al. Wall shear stress and its role in atherosclerosis. Front Cardiovasc Med 2023;10:1083547. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Bernstein MA, Zhou XJ, Polzin JA, et al. Concomitant gradient terms in phase contrast MR: Analysis and correction. Magn Reson Med 1998;39(2):300–308. [DOI] [PubMed] [Google Scholar]
  • 16.Walker PG, Cranney GB, Scheidegger MB, et al. Semiautomated method for noise reduction and background phase error correction in MR phase velocity data. J Magn Reson Imaging 1993;3(3):521–530. [DOI] [PubMed] [Google Scholar]
  • 17.Berhane H, Scott M, Elbaz M, et al. Fully automated 3D aortic segmentation of 4D flow MRI for hemodynamic analysis using deep learning. Magn Reson Med 2020;84(4):2204–2218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Jarvis K, Soulat G, Scott M, et al. Investigation of aortic wall thickness, stiffness and flow reversal in patients with cryptogenic stroke: A 4D flow MRI study. J Magn Reson Imaging 2021;53(3):942–952. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Johnson EMI, Scott MB, Jarvis K, et al. Global Aortic Pulse Wave Velocity is Unchanged in Bicuspid Aortopathy With Normal Valve Function but Elevated in Patients With Aortic Valve Stenosis: Insights From a 4D Flow MRI Study of 597 Subjects. J Magn Reson Imaging 2023;57(1):126–136. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Kilinc O, Chu S, Baraboo J, et al. Hemodynamic Evaluation of Type B Aortic Dissection Using Compressed Sensing Accelerated 4D Flow MRI. J Magn Reson Imaging 2023;57(6):1752–1763. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Cooper M, Hicks C, Ratchford EV, et al. Diagnosis and treatment of uncomplicated type B aortic dissection. Vasc Med 2016;21(6):547–552. [DOI] [PubMed] [Google Scholar]
  • 22.Chu S, Kilinc O, Pradella M, et al. Baseline 4D flow-derived in vivo hemodynamic parameters stratify descending aortic dissection patients with enlarging aortas. Front Cardiovasc Med 2022;9:905718. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Ratchford SM, Stickford JL, Province VM, et al. Vascular Alterations among Young Adults with SARS-CoV-2. Am J Physiol Heart Circ Physiol 2021;320: H404–H410. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Rodilla E, López-Carmona MD, Cortes X, et al. Impact of Arterial Stiffness on All-Cause Mortality in Patients Hospitalized With COVID-19 in Spain. Hypertension 2021;77:856–867. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Stamatelopoulos K, Georgiopoulos G, Baker KF, et al. Estimated Pulse Wave Velocity Improves Risk Stratification for All-Cause Mortality in Patients with COVID-19. Sci Rep 2021;11:20239. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Schnaubelt S, Oppenauer J, Tihanyi D et al. Arterial Stiffness in Acute COVID-19 and Potential Associations with Clinical Outcome. J Intern Med 2021;290:437–443. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Shaaban AM, Duerinckx AJ. Wall shear stress and early atherosclerosis: a review. AJR Am J Roentgenol 2000;174(6):1657–1665. [DOI] [PubMed] [Google Scholar]
  • 28.Malek AM, Alper SL, Izumo S. Hemodynamic shear stress and its role in atherosclerosis. JAMA 1999;282(21):2035–2042. [DOI] [PubMed] [Google Scholar]
  • 29.Zhou J, Li YS, Chien S. Shear stress-initiated signaling and its regulation of endothelial function. Arterioscler Thromb Vasc Biol 2014;34(10):2191–2198. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Gibson CM, Diaz L, Kandarpa K, et al. Relationship of vessel wall shear stress to atherosclerosis progression in human coronary arteries. Arterioscler Thromb 1993;13:310–315. [DOI] [PubMed] [Google Scholar]
  • 31.Stone PH, Saito S, Takahashi S, et al. Prediction of progression of coronary artery disease and clinical outcomes using vascular profiling of endothelial shear stress and arterial plaque characteristics: the PREDICTION study. Circulation 2012;126(2):172–181. [DOI] [PubMed] [Google Scholar]

RESOURCES