Impact of ascending aortic prosthetic grafts on early postoperative descending aortic biomechanics on cardiac magnetic resonance imaging

Abstract

OBJECTIVES

Among patients with ascending thoracic aortic aneurysms, prosthetic graft replacement yields major benefits but risk for recurrent aortic events persists for which mechanism is poorly understood. This pilot study employed cardiac magnetic resonance to test the impact of proximal prosthetic grafts on downstream aortic flow and vascular biomechanics.

METHODS

Cardiac magnetic resonance imaging was prospectively performed in patients with thoracic aortic aneurysms undergoing surgical (Dacron) prosthetic graft implantation. Imaging included time resolved (4-dimensional) phase velocity encoded cardiac magnetic resonance for flow quantification and cine-cardiac magnetic resonance for aortic wall distensibility/strain.

RESULTS

Twenty-nine patients with thoracic aortic aneurysms undergoing proximal aortic graft replacement were studied; cardiac magnetic resonance was performed pre- [12 (4, 21) days] and postoperatively [6.4 (6.2, 7.2) months]. Postoperatively, flow velocity and wall shear stress increased in the arch and descending aorta (P < 0.05); increases were greatest in hereditary aneurysm patients. Global circumferential strain correlated with wall shear stress (r = 0.60–0.72, P < 0.001); strain increased postoperatively in the native descending and thoraco-abdominal aorta (P < 0.001). Graft-induced changes in biomechanical properties of the distal native ascending aorta were associated with post-surgical changes in descending aortic wall shear stress, as evidenced by correlations (r = −0.39–0.52; P ≤ 0.05) between graft-induced reduction of ascending aortic distensibility and increased distal native aortic wall shear stress following grafting.

CONCLUSIONS

Prosthetic graft replacement of the ascending aorta increases downstream aortic wall shear stress and strain. Postoperative increments in descending aortic wall shear stress correlate with reduced ascending aortic distensibility, suggesting that grafts provide a nidus for high energy flow and adverse distal aortic remodelling.

Prosthetic aortic graft replacement is widely used to treat patients with thoracic aortic aneurysms (TAAs).

INTRODUCTION

Prosthetic aortic graft replacement is widely used to treat patients with thoracic aortic aneurysms (TAAs). Grafts eliminate dissection risk within replaced regions, but risks persist in the native distal aorta—especially in patients with genetic aortopathies [1–6]. Nearly half of aortic dissections in genetic TAA patients occur after prophylactic TAA graft surgery [4, 5]. Prior TAA grafting has been associated with risk for type B dissection in Marfan syndrome independent of aortic size [1]. Given the seriousness of such events, insight into mechanism is of substantial importance.

One reason for clinical events after TAA surgery may stem from biomechanical properties of prosthetic grafts, which are stiffer than the native aorta. Prior echocardiography research by our group showed that descending aortic strain increases after proximal grafting [7, 8]. Whereas this suggests that grafts increase energy transmission to the descending aorta, data are lacking as to whether graft-induced changes in the ascending aorta parallel postoperative changes in descending aortic flow and wall strain.

Cardiac magnetic resonance (CMR) imaging enables unique insights into aortic physiology. This exploratory study used CMR to test the impact of prosthetic TAA grafts on the downstream (native) aorta. The goals were to test changes in downstream aortic flow and vessel wall dynamics after proximal grafting, magnitude of graft-induced effects in relation to TAA aetiology, and whether postoperative alterations in descending aortic biomechanics parallel graft-induced changes within the ascending aorta.

PATIENTS AND METHODS

Ethics statement

Enrolment of patients undergoing TAA graft replacement surgery occurred at Weill Cornell Medicine (New York, NY, USA) between February 2018 and June 2020. The Cornell Institutional Review Board (IRB) approved this study protocol (#1710018681 approved 11/27/18); participants provided written informed consent for study participation. Normative controls also provided written informed consent under auspices of an established protocol approved by the Weill Cornell IRB (#1104011660 approved 26 April 2021).

Data availability

Data, analytic methods and study materials will be made available to other researchers for purposes of reproducing the results or replicating the procedure, upon request (contingent on approval of the Weill Cornell Institutional Review Board and assurance of data de-identification).

Study population

The population comprised TAA patients undergoing surgical ascending aortic prosthetic graft [polyethylene terephthalate (Dacron)] replacement. CMR was prospectively performed pre- (within 1 month) and post- (target 6 months) operatively. Clinical data, obtained via standardized questionnaires and chart review, were used to classify TAA aetiology into 3 groups—heritable (predisposing gene variant or TAA onset <50 years old), bicuspid valve or degenerative [4, 9]. Patients with advanced (≥moderate) aortic stenosis or regurgitation, prior aortic surgery (pre-existing graft replacement) or contraindications to CMR were excluded, as were patients undergoing concomitant aortic arch graft replacement.

Normative controls (used as a comparator for hereditable TAA) comprised age/gender matched volunteers without clinically reported coronary artery disease, cardiovascular symptoms or self-reported (or imaging evidenced) TAA at time of CMR.

Imaging protocol

Data acquisition

CMR (3 Tesla) was performed via a standardized protocol, including cine-CMR for aortic deformation and 4-dimensional (4D) flow CMR for aortic flow indices. Cine-CMR employed a steady-state free precession pulse sequence acquired in short axis at pre-specified locations (ascending, arch, descending, thoraco-abdominal aorta), and contiguous Left ventricle (LV) short- and (2, 3, 4-chamber) long-axis imaging. After cine-CMR, gadolinium (0.2 mmol/kg) was infused to patients without contra-indications [90% (26/29)]: Magnetic resonance angiography and 4D flow CMR were acquired in matched (sagittal) orientation; velocity encoding limits (typically ∼1.5 m/s) were adjusted to avoid aliasing.

Data analysis

Aortic flow indices were quantified on 4D flow CMR at landmarks shown in the Central Illustration, inclusive of velocity and wall shear stress [7]. Additional analyses included aortic size (area, linear cross-sectional dimensions) and regurgitation.

Aortic deformation (in TAA patients) was quantified on cine-CMR in locations co-registered to 4D flow using a semi-automated algorithm developed in MATLAB (MathWorks, Natick, MA, USA) for which details have been previously reported [10]. Seed points were placed around the aortic wall, automatically tracked throughout the cardiac cycle, and manually adjusted to optimize border tracking. Global circumferential strain (GCS) was calculated [GCS = (Cs-Cd)/Cd × 100] based on aortic end-systolic and end-diastolic circumference (Cs, Cd). Fractional area change was derived from end-systolic [6] and end-diastolic (EDA) area [FAC= (ESA—EDA)/ESA × 100]; distensibility was calculated using a formula [(ESA—EDA)/(ESA × PP)] incorporating pulse pressure [11]. LV volumes and function were quantified via cine-CMR border planimetry [12, 13].

Pre- and postoperative CMR measurements were made in co-registered locations (blinded to one another).

Statistical methods

Continuous variables are reported as median, interquartile range (given small sample size and non-normality) and were compared between pre- and postoperative time points using paired non-parametric Wilcoxon Signed Rank tests. Multiple group comparisons (between TAA subgroups) were performed using unpaired non-parametric Kruskal–Wallis tests. Correlation coefficients were used to test magnitude of association between continuous variables. Two-sided P < 0.05 was considered indicative of statistical significance. Statistical calculations were performed using SPSS 26.0 (SPSS Inc., Chicago, IL, USA).

RESULTS

Population characteristics

The population comprised 29 TAA patients undergoing proximal aortic surgical graft implantation, in whom CMR was performed pre- [12 (4, 21) days] and post- [6.4 (6.2, 7.2) months] operatively. Table 1 details population characteristics. As shown, 59% (n = 16) of patients had heritable, 28% (n = 8) bicuspid and 17% (n = 5) degenerative TAA. Among heritable TAA patients, aetiology included Marfan syndrome (n = 5), other predisposing genes (n = 1), familial TAA or dissection (n = 4) and idiopathic/onset <50 years old (n = 5). Age increased in relation to TAA aetiology (P = 0.001) and was youngest in patients with heritable TAA; groups were otherwise similar with respect to clinical characteristics.

Table 1:

Population characteristics

	Overall (n = 29)	Heritablea (n = 16)	Bicuspid (n = 8)	Degenerative (n = 5)	P-valueb
Clinical
Age [years]	53 [48, 61]	48 [42, 53]	56 [50, 60]	68 [64, 72]	0.001
Body surface area [m2]	2.2 [1.9, 2.3]	2.2 [1.9, 2.3]	2.2 [1.9, 2.4]	1.9 [1.8, 2.2]	0.68
Male gender	83% (24)	88% (14)	88% (7)	60% [6]	0.33
Coronary artery disease	10% (6)	6% (1)	0% (0)	40% (2)	0.051
Diabetes mellitus	3% (1)	0% (0)	0% (0)	20% (1)	0.08
Tobacco use	14% (4)	6% (1)	13% (1)	40% (2)	0.16
Medication regimen
Beta-blocker	86% (25)	94% (15)	75% (6)	80% (4)	0.42
Angiotensin receptor blocker	28% (8)	13% (2)	63% (5)	20% (1)	0.76
Surgical/proceduralc
Prosthetic graft size (mm)	28 [26,30]	30 [28,30]	27 [25,30]	28 [25,28]	0.09
Valve-sparing type root replacementd	59% (17)	81% (13)	25% (2)	40% (2)	0.14
Aortic regurgitation (mild/moderate)	45% (13)/21% (6)	56% (9)/13% (2)	50% (4)/38% (6)	0% (0)/17% (1)	0.13

^aAmong heritable patients, aetiology was as follows: Marfan (n = 5), Familial TAA or Dissection (n = 4), predisposing gene variants (n = 1: MYH11) and age < 50 (n = 5).

^bKruskal–Wallis test used for between-group comparisons.

^cTen per cent (n = 3) underwent coronary revascularization at the time of TAA graft replacement surgery.

^dTwenty-seven per cent (n = 8) underwent concomitant aortic valve replacement.

TAA: thoracic aortic aneurysm.

Bold values represent statistical significance, as defined by a P value < 0.05.

Table 2 compares haemodynamic and LV indices between pre- and postoperative time points within the overall population, and between patients grouped by TAA aetiology. As shown, LV end-diastolic and end-systolic volumes decreased postoperatively (both P < 0.01), although changes in LV ejection fraction and stroke volume did not attain significance. Regarding sub-groups, LV stroke volume declined in patients with bicuspid TAA and increased in degenerative TAA patients (both P = 0.04). Notably, beta-blocker and angiotensin receptor blocker regimens were unchanged between baseline and follow-up CMR.

Table 2:

Haemodynamic and cardiac indices

	Overall (n = 29)			Heritable (n = 16)			Bicuspid (n = 8)			Degenerative (n = 5)
	Pre	Post	P-valuea	Pre	Post	P-value*	Pre	Post	P-value*	Pre	Post	P-valuea
Haemodynamics
Systolic BP (mmHg)	124 [116, 136]	123 [113, 138]	0.50	124 [119, 133]	120 [113, 136]	0.44	116 [112, 133]	119 [113, 141]	0.44	153 [132, 156]	128 [127, 161]	0.50
Diastolic BP (mmHg)	78 [72, 87]	76 [69, 87]	0.49	79 [70, 88]	73 [68, 88]	0.28	74 [70, 83]	76 [71, 86]	0.48	81 [79, 97]	83 [74, 105]	0.89
Mean arterial pressure (mmHg)	95 [86, 102]	90 [84, 103]	0.48	94 [86, 101]	90 [83, 99]	0.24	90 [82, 96]	89 [85, 106]	0.40	103 [97, 117]	98 [91, 123]	0.14
Heart rate	63 [58, 71]	63 [61, 71]	0.64	61 [53, 63]	63 [61, 74]	0.009	69 [61, 71]	64 [50, 66]	0.24	71 [62, 90]	65 [58, 68]	0.89
Left ventricular structure/function
End-diastolic volume (ml)	182 [145, 210]	159 [148, 189]	0.009	182 [158, 223]	164 [142, 191]	0.04	202 [134, 210]	152 [149, 186]	0.05	161 [106, 202]	157 [127, 195]	0.69
End-systolic volume (ml)	67 [51, 87]	60 [49, 73]	0.006	65 [54, 89]	60 [48,76]	0.09	75 [48, 91]	58 [47, 71]	0.05	66 [43, 100]	60 [43, 87]	0.35
Ejection fraction (%)	62 [58, 65]	64 [59, 69]	0.07	61 [58, 70]	63 [59, 70]	0.83	63 [61, 65]	65 [61, 69]	0.16	58 [49, 63]	62 [52, 70]	0.08
Stroke volume (ml)	112 [94, 132]	101 [92, 124]	0.09	115 [102, 138]	105 [93, 125]	0.10	114 [86, 136]	101 [93, 121]	0.04	89 [63, 105]	97 [76, 116]	0.04

^aWilcoxon Signed Rank Test used for comparisons.

BP: blood pressure.

Bold values represent statistical significance, as defined by a P value < 0.05.

Aortic flow physiology pre- and post-thoracic aortic aneurysm graft implantation

Table 3 reports aortic geometric and flow indices within the prosthetically grafted aorta and the arch, inclusive of comparisons between pre- and postoperative timepoints. As shown, graft implantation decreased ascending aortic size across TAA sub-groups (all P < 0.05), paralleled by increased flow velocity and wall shear stress (all P < 0.05). Similar magnitudes of difference were evident in regions immediately past grafted territories, as shown by increased aortic arch flow velocity and wall shear stress (all P < 0.05) in the overall population, and in analyses limited to heritable TAA patients.

Table 3:

Pre- and postoperative proximal aortic geometry and flow

	Overall (n = 29)			Heritable (n = 16)			Bicuspid (n = 8)			Degenerative (n = 5)
	Pre	Post	P-valuea	Pre	Post	P-valuea	Pre	Post	P-valuea	Pre	Post	P-valuea
Ascending aorta
Maximal diameter (cm)	5.0 [4.8, 5.2]	3.3 [3.2, 3.6]	<0.001	5.0 [4.8, 5.2]	3.3 [3.2, 3.3]	<0.001	5.0 [4.9, 5.2]	3.4 [3.1,3.8]	0.01	4.9 [4.6, 5.7]	3.5 [3.4, 3.7]	0.04
Root maximal diameter	4.8 [4.2, 5.0]	3.3 [3.2, 3.6]	<0.001	4.9 [4.7, 5.1]	3.2 [3.2, 3.3]	<0.001	4.3 [4.0, 5.0]	3.4 [3.1,3.8]	0.01	3.8 [3.5, 4.5]	3.5 [3.4, 3.7]	0.04
Orthogonal diameter	4.5 [3.8, 4.7]	3.2 [3.0, 3.4]	<0.001	4.7 [4.3, 4.8]	3.1 [3.0, 3.2]	<0.001	4.0 [3.3, 4.7]	3.3 [3.0,3.5]	0.02	3.6 [3.4, 4.1]	3.4 [3.3, 3.5]	0.07
Ascending maximal diameter	4.6 [3.7, 5.0]	3.0 [2.8, 3.1]	<0.001	3.8 [3.1, 4.7]	3.1 [2.8, 3.2]	<0.001	4.9 [4.7, 5.2]	3.1 [2.9,3.1]	0.01	4.6 [4.1, 5.6]	2.8 [2.75 ,2.8]	0.04
Orthogonal diameter	4.3 [3.6, 4.8]	2.9 [2.8, 3.1]	<0.001	3.7 [3.0, 4.5]	3.0 [2.7, 3.1]	0.001	4.8 [4.5, 5.0]	3.0 [2.8,3.1]	0.01	4.3 [3.9, 5.5]	2.8 [2.75, 2.8]	0.04
Flow indices
Velocity max (cm/s)	83 [64, 101]	118 [100, 146]	<0.001	83 [65, 92]	125 [103, 137]	<0.001	124 [88, 147]	122 [95,168]	0.21	59 [47, 66]	103 [76, 145]	0.04
Velocity mean (cm/s)	38 [32, 52]	64 [51, 76]	<0.001	41 [36, 51]	65 [56, 78]	<0.001	49 [36, 60]	61 [51,74]	0.07	27 [20, 33]	56 [42, 71]	0.04
WSS maximal (Pa)	1.0 [0.7, 1.6]	1.5 [1.1, 2.1]	0.002	0.95 [0.81, 1.37]	1.5 [1.1, 2.4]	0.005	1.6 [1.4, 1.8]	1.5 [1.4,1.7]	0.48	0.55 [0.43, 0.67]	1.11 [0.91, 1.63]	0.04
WSS mean (Pa)	0.50 [0.40, 0.76]	0.83 [0.66, 1.05]	<0.001	0.50 [0.43, 0.79]	0.92 [0.65, 1.34]	0.002	0.64 [0.51, 0.83]	0.82 [0.67,0.94]	0.01	0.31 [0.20, 0.36]	0.82 [0.49, 0.90]	0.04
Aortic arch
Arch maximal diameter (cm)	2.9 [2.7, 3.1]	2.8 [2.7, 3.2]	0.048	2.8 [0.5, 3.1]	2.7 [2.5, 3.0]	0.13	3.0 [2.7, 3.3]	2.9 [2.7,3.2]	0.21	3.0 [2.8, 3.4]	3.0 [2.8, 3.4]	>0.99
Orthogonal diameter	2.8 [2.6, 3.1]	2.7 [2.6, 3.1]	0.02	2.7 [2.5, 3.0]	2.7 [2.4, 2.8]	0.07	2.9 [2.6, 3.1]	2.8 [2.6,3.1]	0.39	2.9 [2.8, 3.4]	2.9 [2.7, 3.3]	0.08
Flow indices
Velocity maximal (cm/s)	73 [60, 88]	83 [74, 101]	0.002	76 [69, 97]	84 [78, 103]	0.04	67 [51, 90]	76 [62,91]	0.21	62 [39, 69]	79 [63, 91]	0.08
Velocity mean (cm/s)	46 [36, 53]	53 [45, 58]	<0.001	50 [42, 55]	54 [51, 65]	0.01	43 [33, 57]	46 [39,58]	0.26	39 [25, 43]	41 [39, 51]	0.04
WSS maximal (Pa)	1.1 [0.9, 1.3]	1.4 [1.2, 1.9]	<0.001	1.2 [0.8, 1.4]	1.5 [1.3, 2.2]	<0.001	1.2 [1.0, 1.3]	1.2 [1.0,1.4]	0.67	0.90 [0.73, 1.03]	1.2 [0.9, 1.4]	0.14
WSS mean (Pa)	0.75 [0.59, 0.89]	0.82 [0.70, 1.06]	0.001	0.81 [0.59,1.00]	1.0 [0.8,1.2]	0.007	0.75 [0.66, 0.85]	0.78 [0.67,0.94]	0.40	0.59 [0.41, 0.69]	0.63 [0.61, 0.80]	0.08

^aWilcoxon Signed Rank Test used for comparisons.

WSS: wall shear stress.

Bold values represent statistical significance, as defined by a P value < 0.05.

Table 4 reports pre- and postoperative aortic geometry and flow in the native descending aorta. Whereas descending aortic size was unchanged between time points (P = NS), mean wall shear stress in the mid-descending aorta and thoraco-abdominal aorta increased among patients with heritable or bicuspid TAA (both P ≤ 0.05). Descending aortic flow indices as measured postoperatively did not differ (P = NS) between subgroups partitioned based on Aortic Valve Replacement or Valve Sparing Root Replacement at time of TAA grafting. Figure 1 illustrates wall shear stress in native aortic regions prior to and following proximal graft implantation, demonstrating results at each time point to be higher among patients with heritable TAA compared to those with degenerative TAA (all P < 0.05).

Table 4:

Pre- and postoperative distal aortic geometry and flow

	Overall (n = 29)			Heritable (n = 16)			Bicuspid (n = 8)			Degenerative (n = 5)
	Pre	Post	P-valuea	Pre	Post	P-valuea	Pre	Post	P-valuea	Pre	Post	P-valuea
Mid-descending aorta
Maximal diameter (cm)	2.6 [2.3, 2.8]	2.6 [2.4, 2.8]	0.82	2.6 [2.3, 2.8]	2.5 [2.3, 2.8]	0.53	2.7 [2.5, 2.9]	2.7 [2.5, 2.9]	0.71	2.6 [2.5, 2.9]	2.7 [2.5, 3.0]	0.10
Orthogonal diameter	2.5 [2.3, 2.7]	2.5 [2.3, 2.7]	0.98	2.5 [2.2, 2.8]	2.5 [2.3, 2.7]	0.41	2.7 [2.4, 2.7]	2.6 [2.4, 2.7]	0.48	2.6 [2.4, 2.8]	2.6 [2.4, 2.8]	0.56
Flow indices
Velocity max (cm/s)	73 [53, 91]	71 [64, 91]	0.16	81 [64, 95]	82 [69, 94]	0.28	63 [48, 88]	65 [62, 95]	0.33	55 [50, 68]	57 [52, 68]	0.89
Velocity mean (cm/s)	51 [39, 58]	52 [45, 60]	0.02	54 [47, 61]	57 [51, 61]	0.04	43 [34, 56]	48 [44, 65]	0.21	40 [33, 50]	42 [35, 48]	0.69
WSS maximal (Pa)	1.2 [0.9, 1.5]	1.3 [1.1, 1.7]	0.001	1.3 [1.0, 1.6]	1.5 [1.13, 1.8]	0.09	1.1 [0.9, 1.3]	1.3 [1.1, 1.7]	0.02	0.86 [0.74, 1.14]	1.0 [0.9, 1.5]	0.08
WSS mean (Pa)	0.82 [0.69, 1.04]	1.0 [0.8, 1.2]	<0.001	0.93 [0.74, 1.13]	1.07 [1.00, 1.24]	0.007	0.80 [0.65, 0.85]	0.96 [0.78, 1.14]	0.05	0.61 [0.58, 0.88]	0.75 [0.58, 0.99]	0.69
Thoraco-abdominal aorta
Maximal diameter (cm)	2.4 [2.2, 2.6]	2.3 [2.3, 2.5]	0.41	2.3 [2.0, 2.6]	2.3 [2.1, 2.4]	0.13	2.4 [2.2, 2.5]	2.4 [2.3, 2.4]	>0.99	2.4 [2.3, 2.7]	2.4 [2.3, 2.8]	0.32
Orthogonal diameter	2.3 [2.2, 2.5]	2.3 [2.2, 2.4]	0.76	2.3 [2.0, 2.5]	2.3 [2.0, 2.4]	0.45	2.3 [2.2, 2.5]	2.3 [2.3, 2.4]	0.28	2.4 [2.3, 2.7]	2.4 [2.3, 2.7]	0.32
Flow indices
Velocity maximal (cm/s)	72 [55, 87]	80 [63, 91]	0.06	78 [60, 93]	84 [73, 92]	0.049	79 [56, 88]	77 [60, 96]	0.67	54 [52, 58]	60 [49, 69]	0.35
Velocity mean (cm/s)	50 [38, 58]	54 [41, 62]	0.03	53 [43, 65]	58 [48, 64]	0.04	52 [39, 57]	53 [41, 70]	0.48	37 [34, 40]	41 [34, 50]	0.35
WSS maximal (Pa)	1.2 [1.0, 1.5]	1.4 [1.2, 1.7]	0.002	1.4 [1.0, 1.7]	1.5 [1.2, 1.8]	0.02	1.2 [1.0, 1.6]	1.4 [1.2, 1.6]	0.07	0.94 [0.80, 1.08]	1.1 [0.9, 1.2]	0.23
WSS mean (Pa)	0.88 [0.73, 1.04]	1.1 [0.8, 1.3]	<0.001	0.96 [0.78, 1.25]	1.2 [0.9, 1.4]	0.02	0.87 [0.77, 1.03]	1.1 [0.9, 1.3]	0.04	0.70 [0.56, 0.75]	0.78 [0.58, 0.89]	0.35

^aWilcoxon Signed Rank Test used for comparisons.

WSS: wall shear stress.

Bold values represent statistical significance, as defined by a P value < 0.05.

Figure 1:

Pre- and postoperative descending aortic wall shear stress. Wall shear stress quantified by 4D flow CMR at landmarks in the aortic arch (top), mid-descending aorta (middle) and thoraco-abdominal aorta (bottom) partitioned by TAA aetiology (data shown as median [interquartile range]). Note that native wall shear stress differed in relation to TAA aetiology, with heritable patients manifesting higher wall stress compared to patients with degenerative aneurysms [asterisks refer to significant comparisons (P < 0.05) between heritable or BAV groups with degenerative TAA]. BAV: bicuspid aortic valve; TAA: thoracic aortic aneurysm.

Further analyses were performed to test whether native aortic wall shear stress and flow, as were increased after proximal graft implantation among patients with heritable TAA, differed from physiologic norms. To do so, postoperative 4D flow indices among heritable TAA patients were compared to equivalent indices in normative controls (n = 10): Groups were matched for age [48 (42, 53) vs 47 (33, 52); P = 0.65] and gender (88% vs 80% male; P = 0.61), and did not differ in LV stroke volume (P = 0.14). Figure 2 compares postoperative wall shear stress and velocity (in aortic regions distal to grafts) between heritable TAA patients and controls, as well as aortic size in co-registered regions: As shown, wall shear stress in the aortic arch (P < 0.001), mid-descending (P = 0.002) and thoraco-abdominal aorta (P = 0.02) were higher in postoperative heritable TAA patients, despite non-significant differences in aortic size (all P = NS). Similarly, flow velocity measured in the aortic arch and descending aorta were higher (both P < 0.05) among heritable TAA patients who had undergone proximal grafting.

Figure 2:

Downstream aortic flow and geometry in postoperative heritable TAA patients and normative controls. Aortic flow physiology [wall shear stress, velocity (left)] and geometric indices [area, maximum diameter (right)] as quantified in co-registered landmarks among normative controls and postoperative heritable TAA patients (blue = flow indices, grey = geometry). Note that wall shear stress and velocity were higher among heritable TAA patients who had undergone proximal grafting despite non-significant differences in aortic size (data shown as median [interquartile range]). TAA: thoracic aortic aneurysm.

Aortic wall biomechanics before and after thoracic aortic aneurysm grafting

Aortic vessel wall strain and distensibility were tested in relation to flow, and used to assess postoperative changes in native aortic biomechanics after proximal grafting. On preoperative CMR, GCS correlated with wall shear stress in the mid-descending (r = 0.72, P < 0.001) and thoraco-abdominal aorta (r = 0.60, P < 0.001). Similar magnitude of correlation between strain and wall shear stress were evident on postoperative CMR measurements in the mid-descending (r = 0.59, P = 0.001) and thoraco-abdominal aorta (r = 0.72, P < 0.001).

Table 5 compares aortic biomechanics (on cine-CMR) between pre- and postoperative timepoints. As shown, strain and fractional area change decreased approximately three-fold within grafted territories (P < 0.001) in the overall population, with similar patterns within each TAA subgroup. Graft-associated decrements in ascending aortic indices were accompanied by inverse changes in the descending thoracic aorta: GCS and fractional area change in the mid-descending both increased irrespective of whether assessed in the overall population (P < 0.001), or among patients with heritable or bicuspid TAA (all P < 0.05). Among patients with heritable TAA, graft-associated increments in native aortic circumferential strain and Fractional Area Change were also demonstrable in the thoraco-abdominal aorta (P < 0.05), although magnitude of increase versus preoperative assessment was less than that of the mid-descending aorta.

Table 5:

Pre- and postoperative aortic wall biomechanics

	Overall (n = 29)			Heritable (n = 16)			Bicuspid (n = 8)			Degenerative (n = 5)
	Pre	Post	P-valuea	Pre	Post	P-valuea	Pre	Post	P-valuea	Pre	Post	P-valuea
Ascending aorta
Circumferential strain (%)	3.5 [2.3, 4.5]	0.75 [0.34, 0.99]	<0.001	4.1 [2.7, 4.8]	0.64 [0.33, 1.00]	<0.001	2.9 [1.4, 3.8]	0.60 [0.13, 0.95]	0.01	3.5 [1.0, 4.6]	1.0 [0.6, 1.3]	0.07
Fractional area change (%)	7.5 [4.7, 9.1]	1.6 [0.4, 2.0]	<0.001	8.5 [5.4, 10.1]	1.0 [0.3, 2.2]	<0.001	5.9 [1.5, 7.6]	1.77 [0.45, 1.85]	0.07	7.1 [2.0, 9.4]	2.0 [0.5, 2.6]	0.07
Distensibility (10−3 mmHg−1)	1.6 [1.1, 1.9]	0.24 [0.06, 0.41]	<0.001	1.8 [1.5, 2.2]	0.16 [0.07, 0.36]	<0.001	1.2 [0.4, 1.7]	0.29 [0.06, 0.45]	0.04	1.2 [0.3, 1.6]	0.35 [0.09, 0.43]	0.07
Aortic arch
Circumferential strain (%)	6.1 [4.7, 7.7]	7.7 [6.3, 9.1]	0.02	6.8 [5.6, 8.4]	8.4 [6.9, 9.5]	0.33	5.0 [3.2, 7.3]	7.6 [5.7, 8.6]	0.07	4.9 [3.5, 5.3]	6.1 [6.1, 7.8]	0.11
Fractional area change (%)	12 [10, 15]	16 [13, 19]	0.02	14 [12, 17]	17 [14, 20]	0.29	10 [7, 15]	16 [12, 18]	0.07	10 [7, 11]	12.61 [12.58, 12.61]	0.11
Distensibility (10−3 mmHg−1)	2.9 [1.8, 3.3]	3.4 [2.7, 4.4]	0.04	3.0 [2.5, 3.9]	4.2 [3.2, 5.0]	0.24	2.7 [1.1, 3.5]	3.0 [2.3, 3.4]	0.27	1.6 [1.4, 1.7]	2.9 [2.4, 4.0]	0.11
Descending aorta
Circumferential strain (%)	6.7 [5.2, 7.6]	9.4 [7.6, 10.5]	<0.001	7.1 [6.3, 8.1]	9.7 [8.9, 10.5]	<0.001	5.8 [4.9, 6.8]	8.3 [7.1, 13.1]	0.02	6.7 [4.3, 7.0]	7.7 [7.0, 8.5]	0.07
Fractional area change (%)	14 [11, 16]	19 [16, 22]	<0.001	15 [13, 17]	20 [19, 22]	<0.001	12 [10, 14]	17 [15, 28]	0.02	14 [9, 15]	16 [15, 18]	0.07
Distensibility (10−3 mmHg−1)	2.7 [2.2, 3.8]	4.4 [3.8, 5.1]	<0.001	3.2 [2.5, 4.1]	4.4 [4.3, 5.0]	0.004	2.6 [2.1, 3.6]	4.0 [3.0, 5.9]	0.03	2.0 [1.5, 3.0]	3.1 [2.4, 4.0]	0.07
Thoraco-abdominal aorta
Circumferential strain (%)	9.0 [6.7, 10.4]	10.7 [9.3, 11.9]	0.003	8.9 [7.4, 11.8]	11 [10, 12]	0.01	9.3 [7.0, 11.2]	11 [10, 12]	0.23	6.6 [5.1, 9.9]	8.7 [7.1, 10.1]	0.47
Fractional area change (%)	19 [15, 22]	23 [19, 25]	0.004	19 [16, 25]	23 [21, 26]	0.02	20 [16, 24]	23 [21, 25]	0.14	14 [11, 21]	18 [15, 21]	0.47
Distensibility (10−3 mmHg−1)	3.9 [2.9, 5.5]	4.8 [3.8, 5.7]	0.051	4.5 [3.1, 5.7]	5.2 [4.4, 5.9]	0.13	3.9 [3.5, 5.5]	4.7 [4.2, 5.7]	0.04	2.2 [1.8, 3.9]	3.5 [2.5, 4.3]	0.47

^aWilcoxon Signed Rank Test used for comparisons.

WSS: wall shear stress.

Bold values represent statistical significance, as defined by a P value < 0.05.

Impact of thoracic aortic aneurysm biomechanics on postoperative downstream wall shear stress

Analyses were performed to test whether graft-induced changes in biomechanical properties of the native ascending aorta were associated with postoperative changes in native aortic wall shear stress. As shown in Fig. 3, graft-induced reduction of ascending aortic distensibility correlated with increase in downstream aortic wall shear stress in the arch (r = −0.54, P = 0.004); lesser correlations were evident in the descending aorta (r = −0.39, P = 0.05)—consistent with geometric distance between graft termination and downstream measurement location.

Figure 3:

Ascending aortic distensibility in relation to downstream wall shear stress. Scatterplots demonstrating correlations between change in ascending aortic distensibility (pre- vs post-thoracic aortic aneurysm graft implantation) in relation to change in native aortic wall shear stress as measured in co-registered landmarks in the aortic arch (top) and mid-descending aorta (bottom).

DISCUSSION

This pilot study yields new insights regarding the impact of proximal aortic graft implantation on the downstream native aorta. Key findings are as follows: First, among TAA patients undergoing CMR, proximal aortic graft-induced alterations in ascending aortic flow physiology were accompanied by changes in alterations in native distal flow—including increased wall shear stress in downstream regions (P < 0.01): Magnitude of increase was greatest in patients with hereditary TAA, who demonstrated higher wall shear stress than age/gender matched normal controls (P ≤ 0.02). Second, native aortic wall strain also changed following TAA graft implantation: GCS decreased approximately three-fold in grafted territories, accompanied by increased strain in the native descending thoracic aorta (both P < 0.01). Third, graft-induced changes in biomechanical properties of the ascending aorta were associated with post-surgical changes in descending aortic wall shear stress, as evidenced by correlations (r = −0.39–0.54; P ≤ 0.05) between graft-induced reduction of ascending aortic distensibility and increased native aortic wall shear stress.

Our findings are consistent with the notion that prosthetic grafts are stiffer than the native aorta, and thus provide a conduit to propagate high velocity laminar flow to the downstream aorta. Consistent with this, computational fluid dynamics simulations have shown proximal grafts to increase proximal aortic pulse pressure through augmentation of forward wave flow propagation [11]. Similarly, using an ex vivo circulatory model, Scharfschwerdt et al. [14] demonstrated that prosthetic replacement of the ascending aorta increased wall tension in the descending thoracic aorta. Whereas these ex vivo studies support the notion that physiologic effects of grafts extend beyond surgically intervened upon territories, uncertainty persisted as to whether such changes were observable in patients—providing a rationale for our study.

Regarding greater downstream aortic wall shear stress and velocity among patients with hereditary (compared to degenerative) TAA, we speculate this may be attributable to differential aortic resistance which increases with age. In our current study, greater increments in wall shear stress and flow velocity among heritable TAA patients paralleled younger age in this group, suggesting that vascular ageing may attenuate impact of proximal grafts on downstream aortic remodelling after TAA surgery. Consistent with this, prior research by our group reported that, among patients with degenerative TAA, attenuated increments in descending aortic strain after proximal grafting was associated with increased wall thickness and atherosclerotic plaque [8].

Our findings extend upon prior work that has used 4D flow CMR to yield insights regarding impact of prosthetic grafts on aortic biomechanics, and altered flow physiology among patients with different TAA aetiologies. In a prior study by our group, 4D flow CMR demonstrated that surgical re-creation of the sinuses of Valsalva resulted in lower ascending aortic wall shear stress than did tubular grafts, and that differential effects of aortic root reconstruction techniques were evident in the descending aorta [15]. Others have shown aortic flow to be altered with genetic aortopathies [16], and vascular stiffness to be higher with Marfan syndrome—in whom increased aortic stiffness has been associated with accelerated growth [17]. Regarding the descending aorta, Guala et al. [18] reported pulse wave velocity to be higher in Marfan patients (P < 0.001), and showed increased wall shear stress in the descending aorta to be associated with increased aortic size. Whereas our study did not test aortic remodelling in relation to clinical events, it showed that proximal aortic grafts increased downstream wall shear stress and strain, which may increase risk for dissection and aneurysmal degeneration. These findings support the notion that graft-induced changes in aortic biomechanics may contribute to adverse postoperative remodelling—especially in patients with hereditary aortopathies.

This study tested the impact of Dacron grafts on the downstream aorta, so as to provide generalizable findings relevant to current clinical practice. Our findings of increased wall shear stress after grafting highlight the need for research to test new graft materials that can be paired to patient-specific aortic physiology, including more compliant grafts in patients at greatest risk for dissection after TAA graft repair. Whereas our study focused on the downstream aorta, prior research by Kreibich et al. has reported left and right ventricular systolic function to decline after endovascular aortic repair [19]—suggesting that current day grafts can have delirious upstream effects on the heart itself and further highlighting the need to test novel graft technologies. To this end, prior research by Scharfschwerdt tested a novel elastic graft in an ex vivo model, in which it improved root distensibility and ‘windkessel’ function [20]. Given these prior data and our current findings, further research is warranted to investigate in vivo effects and durability of new graft materials designed to replicate native aortic geometry and compliance.

Several limitations should be noted. First, while CMR employed a standardized longitudinal protocol among TAA patients undergoing proximal graft implantation, the population was limited in size which impaired our ability to discern differences between subgroups. Second, whereas all patients underwent proximal TAA grafting, additional surgical interventions varied as evidence by the fact that 24% of patients underwent concomitant aortic valve replacement. Whereas it is possible that aortic valve interventions and changes in LV performance may have affected our findings, it is worth noting that descending aortic flow and vessel wall biomechanical indices markedly increased among patients with hereditary TAA (P < 0.001)—in whom in LV stroke volume was unchanged (P = 0.10). Finally, this protocol entailed brief follow-up (median 6.4 months); longer follow-up is needed to test sustained impact of grafts on downstream remodelling and events.

In conclusion, this pilot study supports the notion that proximal aortic grafting alters distal aortic physiology, including increased wall shear stress and strain, for which alterations are most marked in patients with hereditary aortopathies. Future larger-scale research is necessary to confirm current findings, test modifiers of graft-induced remodelling such as valve replacement/repair strategy, and determine whether graft-induced changes in descending aortic biomechanics predict clinical events including dissection.

Funding

Support: National Institutes of Health R01-HL128278 (Jonathan W. Weinsaft and Jiwon Kim), K23-HL140092 (Jiwon Kim); Marfan Foundation Faculty Grant (Jonathan W. Weinsaft).

Conflict of interest: none declared.

Author contributions

Maria C. Palumbo: Data curation; Formal analysis; Investigation; Software; Writing—original draft; Writing—review & editing. Alberto Redaelli: Formal analysis; Methodology; Writing—original draft; Writing—review & editing. Matthew Wingo: Data curation; Writing—review & editing. Katherine A. Tak: Formal analysis; Writing—review & editing. Jeremy R. Leonard: Data curation; Writing—review & editing. Jiwon Kim: Data curation; Formal analysis; Writing—review & editing. Lisa Q. Rong: Data curation; Writing—review & editing. Christine Park: Data curation; Writing—review & editing. Hannah W. Mitlak: Data curation; Formal analysis; Writing—review & editing. Richard B. Devereux: Data curation; Investigation; Methodology; Writing—review & editing. Mary J. Roman: Conceptualization; Data curation; Writing—review & editing. Arindam RoyChoudhury: Formal analysis; Writing—review & editing. Christopher Lau: Data curation; Investigation; Writing—review & editing. Mario F.L. Gaudino: Data curation; Investigation; Writing—review & editing. Leonard N. Girardi: Data curation; Funding acquisition; Resources; Supervision; Writing—review & editing. Jonathan Weinsaft: Conceptualization; Data curation; Formal analysis; Funding acquisition; Methodology; Project administration; Supervision; Writing—original draft; Writing—review & editing.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks Roman Gottardi, David Schibilsky, Clarence Pienteu Pingpoh and the other anonymous reviewers for their contribution to the peer review process of this article.

ABBREVIATIONS

4D

4-Dimensional

CMR

Cardiac magnetic resonance

GCS

Global circumferential strain

LV

Left ventricle

TAA

Thoracic aortic aneurysm

WSS

Wall shear stress