Coarctation of Aorta is Associated with Left Ventricular Stiffness, Left Atrial Dysfunction and Pulmonary Hypertension

Abstract

Background:

Brachial systolic blood pressure (BP) is the most commonly used metric for monitoring hypertension. However, recent studies suggest that brachial systolic BP underestimates LV systolic load in patients with coarctation of aorta (COA). Since brachial systolic BP is used as a surrogate of arterial afterload in clinical practice, it is important to determine how well it correlates with LV remodeling and stiffness in patients with COA as compared to patients with idiopathic hypertension.

Methods:

This is cross-sectional study of COA patients with hypertension (COA group) and adults with idiopathic hypertension (control group). Both groups were matched 1:1 based on age, sex, BMI and systolic BP. We hypothesized that the COA group will have higher LV systolic and diastolic stiffness, and more advanced left atrial (LA) remodeling and pulmonary hypertension. We assessed LV systolic stiffness using end-systolic elastance (Ees), and diastolic stiffness using LV stiffness constant (β) and chamber capacitance (LV-end-diastolic volume at an end-diastolic pressure of 20mmHg, LVEDVI₂₀)

Results:

There were 112 patients in each group. Although both groups had similar systolic BP, the COA group had a higher Ees (2.37±0.74 vs 2.11±0.54 mmHg/ml, p=0.008), higher β (6.91±0.81 vs 5.93±0.79, p=0.006) and lower LVEDVI₂₀ (58±9 vs 67±11 ml/m², p<0.001). Additionally, the COA group had more advanced LA remodeling and higher pulmonary artery pressures which is corroborating evidence of high LV filling pressure.

Conclusions:

COA patients have more LV stiffness and abnormal hemodynamics compared to non-COA patients with similar systolic BP, suggesting that systolic BP may underestimate LV systolic load in this population. Further studies are required to determine whether the observed LV stiffness and dysfunction translates to more cardiovascular events during follow-up, and whether adopting a stricter systolic BP target in clinical practice or changing threshold for COA intervention will lead to less LV stiffness and better clinical outcomes.

INTRODUCTION

Systemic hypertension is the leading cause of cardiovascular death in the United States.^{1, 2} There is a direct correlation between brachial systolic blood pressure (BP) and cardiovascular death, and a reduction in the risk of cardiovascular death per unit reduction in brachial systolic BP during therapy.^{1, 2} As a result, the American College of Cardiology/American Heart Association guidelines for the management of hypertension recommend initiation and titration of antihypertensive therapy to achieve a BP target of <140/90 mmHg in most patients, and a lower BP target of <130/80 mmHg in high-risk patients such as those with atherosclerotic cardiovascular disease (ASCVD) risk factors.³

Coarctation of aorta (COA) is the most common cause of hypertension among patients with congenital heart disease, and up to 50% of patients with repaired COA have persistent hypertension.⁴ In COA patients without hemodynamically significant recoarctation, the medical management of hypertension (screening, monitoring and titration of antihypertensive therapy) is based on the same guideline recommendations used in other patients with idiopathic hypertension.^{5, 6} A potential pitfall of this management paradigm is the implicit assumption that the pathophysiology of hypertension in patients with COA (in the absence residual coarctation) should be similar to that of patients with idiopathic hypertension. Although this paradigm has not been rigorously tested, emerging data question this assumption. A recent study from our group demonstrated that COA patients had worse arterial stiffening, higher systolic arterial afterload, and left ventricle (LV) hypertrophy as compared to non-COA patients with similar systolic BP.^{7, 8} Based on these data, we will expect that COA patients with hypertension will have more advanced LV remodeling leading to elevated LV systolic and diastolic stiffness, both of which are risk factors for incident heart failure and cardiovascular adverse events.^{9, 10} If this concept is validated by empirical data, it will challenge the current management paradigm with its implicit assumptions, and also provide justification for development of evidenced-based paradigm for the management of hypertension in this unique population. The purpose of this study was to compare LV diastolic and systolic stiffness between COA patients with hypertension and patients with idiopathic hypertension.

METHODS

Study Population and Study Design

The current study was supported by Dr. Egbe’s National Heart, Lung, and Blood Institute (NHLBI) grant K23 HL141448. The authors are solely responsible for the design and conduct of this study, all study analyses, the drafting and editing of the paper and its final contents

In this retrospective case-control study, we reviewed the Mayo Adult Congenital Heart Disease (MACHD) registry and identified adults (age ≥18 years) with repaired COA that received care at Mayo Clinic Rochester Minnesota between January 1, 2000 and December 31, 2018. The inclusion criteria were: (1) prior surgical or transcatheter COA repair; (2) absence of hemodynamically significant residual coarctation defined as aortic isthmus continuous wave Doppler peak velocity of <2.5 m/s in the absence of collaterals; (3) brachial systolic BP measured in the right arm in the absence of aberrant origin of the right subclavian artery; (4) a diagnosis of hypertension treated with at least one of the antihypertensive drugs listed in the guidelines.^{3, 11} From this cohort, we excluded patients with the following conditions: (1) atrial arrhythmia or paced rhythm at the time of echocardiogram because both conditions will interfere with accurate assessment of diastolic function using Doppler echocardiography; (2) significant mitral valve disease defined as a native mitral valve mean gradient >3 mmHg or >mild mitral regurgitation, severe mitral annular calcification based on qualitative assessment, or mitral valve prosthesis; (3) significant aortic valve disease defined as native aortic valve peak velocity >2 m/sec or >mild aortic regurgitation, or prosthetic aortic valve.

For the control group, we identified patients with idiopathic hypertension treated with at least one antihypertensive drug therapy and had no structural heart disease on echocardiogram. We then performed 1:1 matching of the patients with COA and hypertension (COA group) and patients with idiopathic hypertension (control group) using propensity score method based on age, sex, body mass index, and systolic BP at the time of echocardiogram. We hypothesized that the COA group will have higher LV systolic and diastolic stiffness, and more advanced left atrial (LA) remodeling and pulmonary hypertension.

Echocardiography

Assessment of LV mass and volumes

All patients underwent comprehensive echocardiograms based on contemporary guidelines,¹² and offline analysis of echocardiographic images were performed in all cases. LV volumes and ejection fraction were measured using the biplane Simpson’s method, and LV mass and relative wall thickness were assessed using standard techniques.¹³ Based on these measurements and criteria,¹³ we categorized the patients into 4 patterns of LV remodeling: normal, concentric remodeling, concentric hypertrophy, and eccentric hypertrophy. All measurements were index to body surface area.

Assessment of LV Systolic Stiffness

For the purpose of this study, we used LV end-systolic elastance (Ees) as the measure of LV systolic stiffness. LV Ees was assessed using the modified single beat method based on the following indices: brachial systolic BP using arm blood pressure cuff, Doppler derived LV stroke volume, pre-ejection time and total ejection time based on continuous wave Doppler of aortic flow.^{10, 14} Effective arterial elastance index (EaI) which is a lumped measure of the total stiffness of the arterial system was calculated as 0.9 x brachial systolic BP/stroke volume index, and the ratio of Ea/Ees was used as the measure of ventricular-vascular coupling.^{7, 10, 14}

Assessment of LV Diastolic Stiffness

Based on previous work by Lam et al¹⁰ and Anand et al,⁹ we assessed LV diastolic stiffness using these 2 indices: (1) LV stiffness constant (β) and LV capacitance which is the predicted LV end-diastolic volume at an LV end-diastolic pressure (LVEDP) of 20 mmHg indexed to the body surface area (LVEDVI₂₀). These LV diastolic stiffness indices were derived using the methods described below.

First, we measured the septal mitral annular tissue Doppler early diastolic velocity (e’) and mitral pulsed wave Doppler early velocity (E), and calculated E/e’ ratio using standard techniques.¹⁵ LVEDP was then calculated as previously described: LVEDP=11.96 + 0.596 x E/e’.¹⁶ The time constant of isovolumic relaxation (tau, τ) which is inversely related to e’, was calculated using this formula: relaxation τ = (14.70 – 100e’)/0.15.^{10, 16}

The single-beat approach proposed by Klotz et al¹⁷ assumes that the relationship between LV end-diastolic volume (LVEDV) and LVEDP, otherwise known as the LV end-diastolic pressure-volume relationship (EDPVR) share a common shape thus can be derived from a single pressure-volume point. Based on this, the LV EDPVR relationship was estimated as “LVEDP= αLVEDVβ” where α is a curve-fitting constant and β is a diastolic stiffness constant.¹⁰ Similar to a previous study from our institution,^{9, 10} we used the α and β derived from each subject to predict the LVEDV at an LVEDP of 20 mmHg indexed to the body surface area (LVEDVI₂₀) which is a measure of LV capacitance. We assessed LA remodeling and pulmonary hypertension using the LA volume index, LA reservoir strain and tricuspid regurgitation velocity.¹⁵

Exploratory Analyses

We performed exploratory analysis to determine the predictors of LV stiffness, in order to identify potentially modifiable risk factors that can potentially be targeted with medical therapy in order to improve LV stiffness. We performed an exploratory analysis to determine whether the ASCVD risk profile was related to LV stiffness.

Statistical Analysis

Propensity matching was used to balance the between-group differences in baseline characteristics. We calculated the propensity score, which is the probability of having COA, using logistic regression based on age, sex, body mass index, and systolic BP at the time of echocardiogram. One-to-one nearest neighbor caliper matching was used to match patients based on the logit of the propensity score using a caliper equal to 0.2 of the standard deviation of the logit of the propensity score.²⁰ Between-group comparisons were performed using Fisher’s exact test and unpaired t-tests as appropriate. Multivariate linear regression models were developed to determine the predictors of LV stiffness, and these models were based on variables that are known to be associated with clinical outcomes in this population.^{21, 22} The variables used in the model were current age, age at the time COA repair, type of COA repair (surgical vs transcatheter therapy), sex, body mass index, systolic BP, aortic valve morphology (bicuspid/unicuspid aortic valve vs trileaflet aortic valve).

For the calculation of ASCVD risk score, we handled missing data using single conditional imputation method.²³ We also performed sensitivity analyses to determine whether the exclusion of patients with missing data (complete case analysis) resulted in statistically significant difference in the observed outcomes and correlations. A p<0.05 was considered statistically significant. All statistical analyses were performed with JMP software (version 14.1.0; SAS Institute Inc, Cary NC).

RESULTS

Baseline Characteristics

After propensity matching based on pre-defined criteria, we identified 112 COA patients (COA group) and 112 patients with idiopathic hypertension without COA (control group) that met the inclusion criteria for the study. In the COA group, 66 (59%) had bicuspid aortic valve. All patients had prior COA repair/intervention (inclusion criteria) and 14 (13%) had re-intervention for recurrent coarctation after the initial COA repair. The age at the time of initial COA repair/intervention was 5±3 years, and the procedures were resection and end-to-end anastomosis (n=39, 35%), subclavian flap repair (n=19, 17%), patch aortoplasty (n=12, 11%), interposition graft repair (n=29, 26%), balloon aortic dilation (n=9, 8%), and stent implantation (n=4, 4%). Among the 14 patients that had re-interventions, the age at the time of re-intervention was 22±7 years, and the procedures were patch aortoplasty (n=1, 7%), interposition graft repair (n=2, 14%), and stent implantation (n=11, 79%).

The baseline clinical characteristics of both groups are shown in Table 1. By design, both groups had similar age, sex distribution, body mass index and systolic BP. However, the COA group had lower diastolic BP, higher pulse pressure and higher resting heart rate.

Table 1:

Baseline Characteristics

	COA (n=112)	Control (n=112)	p
Age, years	42±8	42±8	0.9
Male	66 (59%)	66 (59%)	0.9
Body mass index, kg/m2	29±5	29±5	0.9
Body surface area, m2	1.9±0.3	1.9±0.2	0.5
Systolic BP, mmHg	134±18	132±17	0.4
Diastolic BP, mmHg	71±11	82±13	0.009
Pulse pressure, mmHg	61±12	50±10	<0.001
Heart rate, bpm	73±7	66±8	0.03
ASCVD risk profile
White/Caucasian	81 (73%) [n=104)	73 (75%) [n=98]	0.6
Current smoker	10 (9%) [n=108)	12 (12%) [n=101]	0.7
Diabetes	5 (5%)	3 (3%)	0.8
Hyperlipidemia	18 (16%)	12 (12%)	0.2
Total cholesterol, mg/dl	168±25	163±31	0.1
HDL cholesterol, mg/dl	48±11	44±9	0.2
LDL cholesterol, mg/dl	102±22	98±14	0.1
Statin therapy	12 (11%)	5 (5%)	0.08
Aspirin therapy	9 (8%)	2 (2%)	0.2
10y-ASCVD risk	4.2±3.7	3.9±2.4	0.2
10y-ASCVD risk ≥10%	6 (5%)	4 (4%)	0.8
Medications
Thiazide diuretics	16 (14%)	19 (17%)	0.6
Beta blockers	25 (22%)	17 (15%)	0.2
Calcium channel blockers	22 (20%)	29 (26%)	0.3
RAAS antagonist	49 (44%)	46 (41%)	0.6
Hydralazine	6 (5%)	2 (2%)	0.9
2 BP medications	4 (4%)	1 (1%)	0.9
≥3 BP medications	2 (2%)	---	---
Laboratory data
Hemoglobin, g/dl	13.6±1.1	13.7±1.0	0.5
Estimated GFR, ml/min/1.73m2	84±26	98±29	0.02

COA: coarctation of aorta; ASCVD: atherosclerotic cardiovascular disease; HDL: high density lipoprotein; LDL: low-density lipoprotein; RAAS: renin-angiotensin-aldosterone system; BP: blood pressure; GFR: glomerular filtration rate

[n] denotes number of patients with available data

LV Systolic and Diastolic Stiffness

LV Systolic Stiffness and Vascular Coupling

Although both groups had similar LV ejection fraction, the COA group had higher LV systolic stiffness as evidence by a higher Ees (2.37±0.74 vs 2.11±0.54 mmHg/ml, p=0.008), Table 2, Figure 1. The COA group also had higher arterial stiffness (EaI 3.3±0.8 vs 2.9±0.6 mmHg/ml*m², p<0.001) but similar ventricular-vascular coupling (Ea/Ees: 0.72±0.23 vs 0.68±0.19, p=0.4) as compared to the control group (Table 2, Figure 1). Consistent with the higher arterial stiffness in the COA group, the patients with COA had higher LV mass index, relative wall thickness and higher prevalence of concentric LV hypertrophy and remodeling (Table 2).

Table 2:

Echocardiography

	COA (n=112)	Control (n=112)	p
LV mass and volumes
LV end-diastolic volume index, ml/m2	52±10	58±11	0.03
LV ejection fraction, %	63±8	60±5	0.1
LV stroke volume index, mL/m2	41±11	46±8	0.008
Heart rate, bpm	73±7	66±8	0.03
Cardiac index, L/min/m2	3.0±0.6	3.1±0.5	0.1
LV mass index, g/m2	110±22	94±29	<0.001
Relative wall thickness*	0.43±0.06	0.40±0.05	<0.001
LV hypertrophy	47 (42%)	30 (27%)	0.02
Concentric hypertrophy	38 (34%)	16 (14%)	<0.001
Concentric remodeling	38 (34%)	21 (19%)	0.02
Eccentric hypertrophy	9 (8%)	14 (13%)	0.3
Normal	29 (26%)	61 (55%)	<0.001
LV systolic stiffness and coupling
Ea, mmHg/ml	1.7±0.4	1.4±0.3	<0.001
Eal, mmHg/ml*m2	3.3±0.8	2.9±0.6	<0.001
Ees, mmHg/ml	2.37±0.74	2.11±0.54	0.008
Ea/Ees	0.72±0.23	0.68±0.19	0.4
LV diastolic stiffness and function
E, m/s	1.1±0.4	0.9±0.1	0.06
Septal e’, m/s	0.08±0.04	0.10±0.04	0.04
E/e’ ratio	12±4	9±4	0.008
τ, ms	47.6±22.1	38.2±19.4	0.008
β	6.91±0.81	5.93±0.79	0.006
LVEDP mmHg	15±6	12±5	<0.001
LVEDVI20, ml/m2	58±9	67±11	<0.001
LA remodeling and pulm hypertension
LA volume index, ml/m2	30±11	24±7	0.007
LA reservoir strain, %	33±9	39±12	0.004
Tricuspid regurgitation velocity, m/s	2.7±0.4	2.5±0.3	0.06

COA: coarctation of aorta; LV: left ventricle; Ees: end-systolic elastance; Ea: effective arterial elastance; E: mitral inflow pulsed wave Doppler early velocity; e’: tissue Doppler early velocity; τ. tau; β: LV stiffness constant; LVEDP: LV end-diastolic pressure; LVEDVI: LV end-diastolic volume index at an LVEDP of 20 mmHg; LA: left atrium;

Figure 1:

Comparison of left ventricular (LV) stiffness indices between patients with coarctation of aorta (red) and patients with idiopathic hypertension (black).

Top panel: COA patients had higher LV systolic stiffness indices as evidenced by higher LV end-systolic elastance (Ees) and arterial elastance (Ea), but similar ventricular-arterial coupling (Ea/Ees)

Middle panel: COA patients had higher LV diastolic stiffness indices as evidence by higher LV stiffness index (β) and lower LV capacitance (LV end-diastolic volume index at LV end-diastolic pressure of 20 mmHg)

Bottom panel: COA patients had higher left atrial (LA) volume index, lower LA reservoir strain and higher tricuspid regurgitation velocity suggestive of more advance LA remodeling and pulmonary hypertension

LV Diastolic Stiffness and Filling Pressures

Compared to the control group, the COA group had higher LV diastolic stiffness as evidenced by a higher β (6.91±0.81 vs 5.93±0.79, p=0.006) and lower LV capacitance (LVEDVI₂₀ 58±9 vs 67±11 ml/m², p<0.001). Consistent with the higher LV diastolic stiffness in the COA group, the COA patients had more impaired LV relaxation (lower e’ and longer τ), higher LV filling pressures (E/e’ and LVEDP), and more advanced LA remodeling (higher LA volume index and lower LA reservoir strain), and higher pulmonary artery pressures (higher tricuspid regurgitation velocity) Table 2, Figure 1

Exploratory Analyses

Determinants of LV stiffness

There was a correlation between systolic BP and Ees (r=0.42, p <0.001), and a correlation between systolic BP and LV diastolic stiffness (r=0.37, p =0.001). Systolic BP was inversely related to LV capacitance (r= −0.32, p=0.03). Table 3 shows the multivariate models of the relationship between systolic BP and the different LV stiffness indices after adjustments for age, male sex, body mass index, type of COA repair, and bicuspid aortic valve

Table 3:

Determinants of LV Systolic and Diastolic Stiffness

Determinants of LV systolic stiffness (Ees)	Beta ± SE	p
Current age, years	---	---
Age at COA repair, years	---	---
Systolic BP, mmHg	0.18±0.04	<0.001
Male sex	---	---
Body mass index, kg/m2	---	---
Transcatheter COA repair	---	---
Bicuspid/unicuspid aortic valve	---	---
Determinants of LV chamber stiffness (β stiffness constant)	Beta ± SE	p
Age, years	0.09±0.02	<0.001
Age at COA repair, years	---	---
Systolic BP, mmHg	0.12±0.01	<0.001
Male sex	---	---
Body mass index, kg/m2	0.08±0.03	0.02
Transcatheter COA repair	---	---
Bicuspid/unicuspid aortic valve	---	---
Determinants of LV capacitance (LVEDVI20)	Beta ± SE	p
Age, years	−0.02.1±0.01	0.01
Age at COA repair, years	---	---
Systolic BP, mmHg	−0.07.1±0.02	0.002
Male sex	----	---
Body mass index, kg/m2	−0.03±0.0.01	0.007
Transcatheter COA repair	---	---
Bicuspid/unicuspid aortic valve	---	---

COA: coarctation of aorta; LV: left ventricle; Ees: end-systolic elastance; BP: blood pressure; SE: standard error; LVEDVI: LV end-diastolic volume index at an LVEDP of 20 mmHg

---: denotes non-significant beta estimates and p values.

Predicted LV stiffness indices

The mean systolic BP for the control was 132±17 mmHg, and at this systolic BP, the measured LV stiffness indices were: Ees (2.11±0.54 mmHg/ml), LV diastolic stiffness constant (β) (5.93±0.79), and LV capacitance (LVEDVI₂₀) (67±11 ml/m²), Table 2. Using these values as reference points, we performed an exploratory analysis to determine which systolic BP in the COA group corresponded to the LV stiffness indices observed in the control group (Table 4). A systolic BP of 120 mmHg in the COA group corresponded to predicted LV stiffness indices that were similar to the values observed in the control group: Ees (2.08±0.66 vs 2.11±0.54 mmHg/ml, p=0.7); β stiffness constant (5.87±0.71 vs 5.93±0.79, p=0.8) and LVEDVI₂₀ (69±10 vs 67±11 ml/m², p=0.5).

Table 4:

Predicted LV Systolic and Diastolic Stiffness Indices at Different Systolic BP

Systolic BP	Ees (mmHg/ml)	β	LVEDVI20 (ml/m2)
Measured
134 ±18 mmHg	2.37±0.74	6.91±0.81	58±9
Predicted
140, mmHg	2.42±0.81	7.14±0.68	53±8
130 mmHg	2.21±0.74	6.22±0.84	64±11
120 mmHg	2.08±0.66	5.87±0.71	69±10

Ees: end-systolic elastance; LVEDVI: LV end-diastolic volume index at an LVEDP of 20 mmHg; BP: blood pressure

ASCVD Risk Score and LV Stiffness Indices

Although the COA group had significantly worse LV stiffness indices, we observed comparable ASCVD risk scores in both groups (Table 1). There was no correlation between ASCVD risk score and LV stiffness indices in either group.

DISCUSSION

Systemic hypertension is common in patients with repaired COA.⁴ In the absence of significant residual coarctation, COA patients with hypertension are managed using the same treatment criteria as patients with idiopathic hypertension.^{3, 6} In this study, we performed a comprehensive echocardiographic analysis of LV remodeling and function between propensity-matched cohorts of hypertensive patients with and without COA. We observed that although both groups had similar brachial systolic BP, the COA patients had higher LV systolic and diastolic stiffness and more advanced LA remodeling and pulmonary hypertension. The predicted severity of LV stiffness in a hypothetical cohort of COA patients with an average systolic BP of 120 mmHg was comparable to the observed LV stiffness in the control groups with an average systolic BP of 132 mmHg. Collectively, these data suggest that COA patients experienced higher arterial load, LV pressure overload and remodeling, LA dysfunction and pulmonary hypertension as compared to non-COA patients with similar brachial systolic BP.

Chronic LV pressure overload leads to cardiomyocyte hypertrophy and fibrosis which then leads to LV hypertrophy, diastolic dysfunction, and systolic dysfunction.^{24, 25} Consistent with these previous data, the current study showed that the COA group had a higher LV mass and concentric hypertrophy, impaired LV relaxation (lower e’ and longer τ), increased LV diastolic stiffness and higher filling pressures (E/e’ and LVEDP). The COA group also had other corroborating evidence of high LV filling pressure (higher LA volume index, lower LA reservoir strain, and higher tricuspid regurgitation velocity). These structural and functional changes are known risk factors for heart failure and cardiovascular mortality.^{9, 10} Additionally, the COA patients had elevated LV systolic stiffness (Ees) and vascular stiffness (Ea) but a normal ventricular-vascular coupling (Ea/Ees), similar to what has been described in patients with heart failure with preserved ejection fraction.²⁶ While a normal ventricular-vascular coupling suggests an efficient hemodynamic performance, having an elevated Ees is associated with impaired cardiac output reserve, reduced aerobic capacity, and hypertensive response during exercise.²⁶ Since exercise intolerance and exercise-induced hypertension are relatively common in the COA population,^{21, 27} further studies are required to determine whether these abnormal exercise hemodynamics are mechanistically linked to high LV systolic stiffness

Pathophysiologic Mechanisms for High Arterial Afterload in COA

Since LV remodeling and stiffness is driven by pressure overload, the observed increase in LV stiffness and dysfunction in the COA group suggests that these patients experienced LV pressure overload, that is out of proportion to the severity of hypertension measured by the brachial systolic BP. We postulated that this mismatch between brachial systolic BP and LV pressure overload/remodeling in the COA group is likely related to endothelial dysfunction and increased arterial stiffness which is prevalent in this population.²⁸ In health, the proximal aorta is a compliant vessel, and helps to buffer pressure rise during the ejection phase of the cardiac cycle.^{24, 25} In the setting of increased arterial stiffness such as seen in COA, the aorta is less compliant resulting in isolated systolic hypertension, and an increase in LV systolic load which is a trigger for LV remodeling and stiffness.

Another potential factor contributing to the advanced LV stiffness in the COA group is the impact of wave propagation and reflection. In a healthy arterial system with normal pulse wave velocity, the wave reflections arrive at the proximal aorta during diastole, thereby augmenting diastolic pressure, and in turn, coronary artery perfusion.^{24, 25} In the setting increased arterial stiffness (which is often present in COA patients), there is an increase in the velocity of wave propagation and wave reflection resulting in the wave reflections arriving in the proximal aorta in mid-to-late systole producing systolic pressure augmentation, which is a primary driver of LV remodeling, stiffness and dysfunction.²⁹ Although the patients is in the current study did have not residual coarctation, the presence of a surgical scar, prosthetic material or stent at the site of COA repair creates an impedance mismatch which further promotes wave reflection and systolic pressure augmentation. We did not observe any correlation between type of COA repair and LV remodeling which is consistent with other studies.³⁰

Clinical implications

We observed that a brachial systolic BP of 120 mmHg in the COA group correlated with severity of LV systolic and diastolic stiffness comparable to that of the control group (mean systolic BP of 132 mmHg). This suggests that the severity of LV remodeling in COA patients corresponds to that of non-COA patients with systolic BP that is >10 points higher, and this has potential clinical implications with regards to BP targets for medical therapy. We observed that ASCVD risk score (the guideline directed metric for intensification of antihypertensive therapy) was not associated with the severity of LV stiffness. However, it is important to highlight that ASCVD risk has not been validated for use in patients younger 40 years of age such as in the current study. Collectively, these findings highlight the potential pitfalls of tailoring medical therapy for hypertension in COA patients based on the general guidelines for patient with idiopathic hypertension. Since we use the brachial systolic BP for decision making since the central aortic pressure cannot be easily measured in routine clinical practice, this study adds to the emerging body of evidence demonstrating that brachial systolic BP underestimates LV pressure overload in this population. Perhaps aiming for a lower brachial systolic BP target might reduce systolic arterial load on the LV, and potentially slow the pace of LV remodeling. Further studies are needed to provide empirical validation for these postulates.

In addition to the harmful effects on the LV, isolated systolic hypertension and wide pulse pressure (which was present in the COA group) also results in excessive penetration of pulsatile energy into the microvasculature of target organs that operate at low vascular resistance such as the brain and the kidney.^{24, 25} We did observe in lower glomerular filtration rate in the COA group, but a mechanistic link between arterial stiffness and renal function cannot be deduced from the current study.

Limitations

We did not measure pulse wave velocity and endothelial function in the current study, and hence we can only postulate that the higher arterial elastance and LV stiffness in the COA group was due to increased arterial stiffness. Since the severity of LV remodeling is dependent on both the severity and duration of pressure overload, it is plausible that the higher LV stiffness and dysfunction in the COA group may have been related to a lead-time bias since COA patients tend to develop hypertension at a young age. Notwithstanding, the advanced LV stiffness and dysfunction observed in the COA group and the concordant evidence of LA dysfunction and pulmonary hypertension suggest that these patients are at higher risk for cardiovascular adverse events. If LV stiffness is shown to be prognostic in the COA population, it will provide strong justification to reconsider the current management paradigm that assumes similar disease pathophysiology between patients with COA and those with idiopathic hypertension.

Conclusion

COA patients had higher LV systolic and diastolic stiffness and more advanced LA remodeling and pulmonary hypertension as compared to patients with idiopathic hypertension with similar brachial systolic BP. These data suggest that brachial systolic BP underestimates LV systolic load in this population, and highlights potential differences in disease pathophysiology. Further studies are required to determine whether advanced LV stiffness and dysfunction observed in this study translates to more cardiovascular events during follow-up, and whether adopting a stricter systolic BP target in clinical practice will lead to less LV stiffness and better clinical outcomes.

Highlights.

• COA was associated with LV stiffness, LA remodeling, and pulmonary hypertension

• COA patients had higher arterial elastance for a given systolic BP

• Systolic BP underestimate LV systolic load in COA patients

Abbreviations:

COA

Coarctation of aorta

LV

left ventricle

ASCVD

atherosclerotic cardiovascular disease

BP

blood pressure

LVEDP

LV end-diastolic pressure

LVEDV

LV end-diastolic volume

LV-EDPVR

LV end-diastolic pressure-volume relationship

LVEDVI₂₀

LV end-diastolic volume index at an LVEDP of 20 mmHg

LA

left atrium