Association of Preterm Birth With Myocardial Fibrosis and Diastolic Dysfunction in Young Adulthood

Background Preterm birth affects about 10% of live births worldwide and is associated with cardiac alterations. Animal models of preterm birth suggest that left ventricular functional impairment may be due to an up-regulation of myocardial fibrosis. Objectives The aim of this study was to determine whether diffuse left ventricular fibrosis is evident in young adults born preterm. Methods One hundred one normotensive young adults born preterm (n = 47, mean gestational age 32.8 ± 3.2 weeks) and term (n = 54) were included from YACHT (Young Adult Cardiovascular Health sTudy). Left ventricular structure and function were quantified by cardiovascular magnetic resonance and echocardiography. Intravenous administration of a gadolinium-based contrast agent during cardiovascular magnetic resonance was used to quantify focal myocardial fibrosis on the basis of late gadolinium enhancement and, in combination with T1 mapping, to quantify diffuse myocardial fibrosis on the basis of assessment of myocardial extracellular volume fraction. Results Adults born preterm had smaller left ventricular end-diastolic and stroke volumes, with greater left ventricular mass and wall thickness (P < 0.001). In addition, longitudinal peak systolic strain and diastolic strain rate by both cardiovascular magnetic resonance and echocardiography, and E/A ratio measured by echocardiography, were lower in preterm-born compared to term-born adults (P < 0.05). Extracellular volume fraction was greater in preterm-born compared with term-born adults (27.81% ± 1.69% vs 25.48% ± 1.41%; P < 0.001) and was a significant mediator in the relationship between gestational age and both longitudinal peak diastolic strain rate and E/A ratio. Conclusions Preterm-born young adults have greater extracellular volume fraction in the left ventricle that is inversely related with gestational age and may underlie their diastolic functional impairments.

Large-scale birth registries investigating the impact of being born preterm have shown an increased risk for cardiovascular diseases by childhood and adulthood, including heart failure (4,5), ischemic heart disease (6), and early cardiovascular-related mortality (7).
In line with an increased cardiovascular risk, left ventricular (LV) morphologic and functional changes in people born preterm are first observed in the neonatal period and continue into infancy, childhood, and young adulthood (8). These potentially adverse alterations extend across gestational age categories of prematurity and include smaller internal cavity dimensions and volumes, reduced systolic and diastolic function, and an increased rate of hypertrophy from childhood to adulthood (8), though there is variability in LV mass findings across studies (9,10). This is consistent with findings from animal models of prematurity, which show smaller cardiac dimensions and greater myocardial interstitial fibrosis (11,12). Of note, reduced LV systolic and diastolic function measured using echocardiography in a preterm rat model relate to an up-regulation of cardiac fibrotic pathways (13,14).
To date, there have been no human studies investigating myocardial fibrosis in adults born preterm and how this relates to changes in systolic and diastolic function. Cardiovascular magnetic resonance (CMR) imaging offers the opportunity to assess focal myocardial fibrosis with late gadolinium enhancement (LGE) and diffuse myocardial fibrosis by preand postcontrast T1 mapping to measure myocardial extracellular volume fraction (ECV), without requiring direct access to tissue (15,16). We have therefore used CMR and echocardiography to determine whether preterm-born adults have greater LV myocardial fibrosis and whether this relates to the degree of prematurity and their known reductions in systolic and diastolic function.

METHODS
STUDY POPULATION. Subjects were recruited into YACHT (Young Adult Cardiovascular Health sTudy; NCT02103231), a cross-sectional, case-control study of preterm-born (<37 gestational weeks) and termborn ($37 gestational weeks) young adults (17,18).  clinical research investigators completed all measurements. As previously described (17,18), height was measured to the nearest centimeter and weight to the nearest 0.1 kg using an integrated height and weight measurement station (Seca). Twenty-fourhour ambulatory blood pressure monitoring was initiated at the end of the study visit using oscillometric, ambulatory devices (TM-2430, A&D Instruments). Venous blood samples were collected and full blood count was measured on the day of the study visit, with additional samples centrifuged and stored at À80 C for future blood biochemistry analysis.
Acquisition of pulsed-wave Doppler sample volumes between the mitral leaflet tips (E and A waves) was done using pulsed-wave Doppler with blood flow on apical 4-chamber views for later offline analysis of E/A ratio. Tissue Doppler imaging of the lateral and septal basal regions from apical 4-chamber views were also acquired for later offline analysis of average e 0 velocity to calculate E/e 0 ratio. C M R a c q u i s i t i o n . CMR was performed using a 3.0-T magnetic resonance imaging scanner (3T TIM Trio, Siemens Medical Solutions). Horizontal and vertical long-axis and LV outflow tract retrospective electrocardiographically gated steady-state free precession cine images were acquired, followed by short-axis steady-state free precession cine images (20). Data were acquired at end-expiratory breath-hold and digitally stored for offline analysis. LV short-axis images were obtained from the base to the apex of the heart. The thickness and gap between slices were 8.0 and 2.0 mm, respectively. A midventricular shortaxis T1 map was acquired for each participant prior to LGE imaging (native T1 mapping) using the shortened modified Look-Locker inversion recovery sequence (21). LGE imaging was acquired using a T1-weighted phase-sensitive inversion recovery sequence in multiple short-axis slices to match cine views and longaxis planes. Images were acquired approximately 8 to 10 minutes after intravenous administration of a gadolinium-based contrast agent (Gadovist, Bayer) at a dose of 0.15 mmol/kg body weight. The inversion time was adjusted for optimal nulling of remote normal myocardium (22). Fifteen minutes after intravenous administration of gadolinium-based contrast agent, a midventricular short-axis T1 map (identical to the one acquired prior to LGE) was acquired. The end-diastolic and end-systolic cardiac phases, as well as the basal and apical LV slices, were visually determined as previously described (20,24). LV wall thickness was measured on the midventricular shortaxis slice at end-diastole. Myocardial deformation analysis was performed using feature tracking (CVI42). Endocardial and epicardial borders of the LV long-axis cines and midventricular short-axis cine were manually contoured on the end-diastolic frame.
The deformation of the myocardium was then automatically tracked through the cardiac cycle to calculate systolic and diastolic longitudinal and circumferential strain parameters.   (25). To determine if the mediation effect was significant, bootstrapping was used, and the indirect effect with 95% CI was computed (26). Statistical significance was determined at P < 0.05. P values and 95% CIs presented have not been adjusted for multiplicity, and therefore inferences drawn from these statistics may not be reproducible.

RESULTS
PARTICIPANT CHARACTERISTICS. A total of 149 participants were recruited into YACHT. For this study, 48 of the 149 participants from YACHT were excluded from analysis to prevent the potential confounding effects of hypertensive cardiac remodeling on our results (n ¼ 32 from the hypertension clinic being treated pharmacologically and n ¼ 16 with high blood pressure from 24-hour blood pressure monitoring). Of the 101 normotensive young adult participants included in this analysis, 54 were born at term (39.5 AE 1.4 gestational weeks), and 47 were born preterm (32.8 AE 3.2 gestational weeks). The cohort size allowed us to detect a 0.65-SD difference between groups powered at 90% and P ¼ 0.05. There were no differences in age and body mass index between preterm-born and term-born groups (P > 0.05). The percentage of men was lower in the group of young Lewandowski et al.   Gestational age was also positively related with E/A ratio and longitudinal peak diastolic strain rate, with a 0.041-unit higher E/A ratio (P ¼ 0.020), 0.075-s À1 higher CMR-derived longitudinal peak diastolic strain rate (P ¼ 0.016), and 0.083-s À1 higher echocardiography-derived longitudinal peak diastolic strain rate (P ¼ 0.018) per 1-week greater gestational age.
Mediation analyses were performed to determine whether the relationships between the degree of prematurity and impaired diastolic function were mediated by diffuse myocardial fibrosis (ECV). Controlling for gestational age did not alter the association between ECV and E/A ratio (b ¼ À0.384; P ¼ 0.021), but gestational age was no longer a significant predictor of E/A ratio (b ¼ 0.179; P ¼ 0.357) (Figure 2A). Controlling for gestational age also did not alter the association between ECV and CMR-derived longitudinal peak diastolic strain rate (b ¼ À0.410; P ¼ 0.013), but gestational age was no longer a significant predictor of CMR-derived longitudinal peak diastolic strain rate Figure 2B). Similarly, controlling for gestational age did not alter the association between ECV and echocardiography-derived longitudi-  Values in bold are statistically significant (P < 0.05). B represents the difference in left ventricular structural or functional variables per 1% elevation in ECV (slope or unstandardized regression coefficient). b represents the standardized regression coefficient (À1 to 1). Table 2.

Abbreviations as in
Lewandowski et al.
Compared with their term-born peers (n ¼ 54), preterm-born young adults (n ¼ 47) had higher left ventricular extracellular volume fraction, which is a surrogate measure of diffuse myocardial fibrosis measured by cardiovascular magnetic resonance. They were also shown to have lower left ventricular diastolic function in young adulthood measured by both cardiovascular magnetic resonance and echocardiography. A greater degree of prematurity (lower gestational age) associated with elevated diffuse myocardial fibrosis and impaired diastolic function, with diffuse myocardial fibrosis shown to be a significant mediator in the association between gestational age and diastolic function. Further follow-up will be needed to determine the clinical significance of these left ventricular changes in structure and function in people born preterm as they reach middle to late adulthood.
Lewandowski et al. Lewandowski et al.