Left atrial stiffness in women with ischemia and no obstructive coronary artery disease: Novel insight from left atrial feature tracking

Abstract Background Women with signs and symptoms of ischemia and no obstructive coronary artery disease (INOCA) are at risk of heart failure with preserved ejection fraction (HFpEF); however, the mechanism for HFpEF progression remains unclear. Studies in INOCA have largely focused on left ventricular function. The left atrium serves an important role in maintaining transmitral flow, and is impaired in HFpEF; however, it remains unclear if left atrial function is impaired in INOCA. Hypothesis Left atrial function is progressively worse in INOCA and HFpEF compared to controls. Methods We compared 39 reference control subjects to 64 women with INOCA and 22 subjects with HFpEF. Left atrial strain was assessed by feature tracking using magnetic resonance cine images. Results Peak left atrial strain was reduced in HFpEF compared to controls (22.9 ± 4.8% vs 25.9 ± 3.2%, P < .01), but similar in INOCA (24.8 ± 4.5%) compared to HFpEF and controls (P = .18). However, left ventricular end‐diastolic pressure (LVEDP) was elevated in 33% of INOCA participants, suggesting that left atrial stiffness (LVEDP/LA strain) is elevated in a large portion of women with INOCA. Conclusions Taken together, we interpret these data to support our working hypothesis that INOCA is a pre‐HFpEF state, with left atrial stiffness preceding overt left atrial dysfunction; representing a putative therapeutic target to prevent HFpEF progression in this at‐risk population.


| INTRODUCTION
Ischemic heart disease is the leading cause of death in women, with annual mortality rates exceeding all forms of cancer combined. 1 Work from the NHLBI-Sponsored Women's Ischemia Syndrome Evaluation (WISE) study suggests that most women presenting with signs and symptoms of ischemia have no obstructive coronary artery disease (INOCA), 2 and that these women are at increased risk of developing heart failure 3 ; confirmed to be almost exclusively heart failure with preserved ejection fraction (HFpEF). 4 However, despite our increased understanding, the pathophysiologic mechanism(s) driving heart failure progression in INOCA remains incompletely understood.
Consistent with the HFpEF phenotype, our group has shown that women with INOCA often have left ventricular diastolic dysfunction [5][6][7][8] ; however, these investigations have focused almost exclusively on left ventricular relaxation. The left atrium also contributes in generating the transmitral pressure gradient needed to fill the ventricle, and multiple investigations have shown that left atrial function provides important insight into the adaptive changes that contribute to ventricular filling. [9][10][11][12][13][14][15][16][17] To the best of our knowledge, no study has evaluated left atrial function in women with INOCA. Given that left atrial function is often impaired in HFpEF, 16 Obstructive coronary artery disease was ruled out in HFpEF subjects using cardiac-computed tomographic angiography. 23 Reference control women did not have any symptoms, risk factors for, or evidence of ischemic heart disease, confirmed by a standardized 12-lead treadmill stress test. 24 All study subjects gave written informed consent before undergoing evaluation and the study protocol was approved by the Institutional Review Board at Cedars-Sinai Medical Center.

| Cardiac magnetic resonance imaging
In the majority of subjects, cardiac magnetic resonance imaging was performed on a 3.0T scanner (Siemens Healthineers, Erlangen, Germany); however, a subset of participants (n = 27) underwent cardiac magnetic resonance imaging on a 1.5T scanner (Siemens Healthineers, Erlangen, Germany). In all cases, images were electrocardiogram-gated and a phase-array surface coil (CP Body Array Flex; Siemens Healthineers) was used. Long-axis views (ie, 4-, 3-, and 2-chamber) along with a series of short-axis cine images spanning the entire left ventricle were collected (steady-statefree-precession pulse sequence) for assessment of left ventricular and left atrial morphology and function.
Left ventricular mass and volumes were assessed using the method of disks, from a series of short-axis cine images spanning the entire left ventricle, as previously described. 25 Briefly, using the commercially available software (CVI 42 version 5. ject. Data were excluded if both reviewers were not satisfied with the tracking quality. Left atrial function was characterized by three distinct phases: reservoir, when the left atrium passively receives blood from the pulmonary circulation; conduit, when blood flows passively from the atrium to the ventricle along the transmitral pressure gradient; and booster, when the left atrium contracts, transferring blood into the ventricle. Intraobserver variability for reservoir, conduit, and booster strain, reported as a coefficient of variation, are 2%, 4%, and <1%, respectively. Left atrial stiffness index was defined by the ratio between LVEDP and peak reservoir strain, as previously reported. 27-31 31 To define the threshold of normal, all control subjects were assumed to have a LVEDP of 12 mmHg (the upper limit of normal), with the threshold for abnormal left atrial stiffness set two SD above the calculated mean.
Left atrial volume was measured using the same horizontal and vertical long-axis images, along with the left ventricular outflow tract view (ie, 3-chamber), as previously described, 32 at three specific time points: end of ventricular systole (immediately prior to mitral valve opening) to assess left atrial reservoir volume, early ventricular diastole (immediately prior the atrial contraction) to assess left atrial conduit volume, and at late ventricular diastole (immediately following mitral-valve closure) to determine left atrial booster volume. Left atrial ejection fraction was calculated as the difference between reservoir and booster volume divided by reservoir volume, expressed as a percentage.

| Statistical methods
Data were analyzed using IBM SPSS Statistics 24 (version 13.0). All data are reported as mean ± SD, unless otherwise specified. Group comparisons were made using either one-way analysis of variance or to controls, resulting in a small but significant difference in left ventricular ejection fraction between groups (Table 1).
Consistent with prior reports, left atrial reservoir strain was lower in HFpEF than reference controls, as was left atrial conduit strain ( Figure 1). In contrast, left atrial reservoir strain and conduit strain were similar between both controls and INOCA, and HFpEF and INOCA (Figure 1). A similar between-group pattern was also observed in left atrial volume, with HFpEF having the highest reservoir, conduit, and booster volume compared to controls, with INOCA in the middle ( Table 2). After adjusting for age and body mass index, group differences in left atrial strain were eliminated; however, the group differences in left atrial volume persisted (Table 2). Together, these data support left atrial dysfunction as a putative mechanism contributing to HFpEF progression in INOCA.
Previous work by our group has established impaired left ventricular early diastolic strain rate and peak ventricular untwisting rate, as hallmark features of INOCA. 7,8,33 Here we extend these observations by evaluating left atrial function for the first time in INOCA, and relate these data to HFpEF. Consistent with recent reports, we observed a marked reduction in left atrial reservoir strain in HFpEF compared to reference controls. 18,20,21,34 In contrast to our original hypothesis; however, we did not observe differences in left atrial strain between INOCA and controls; nor did the differences between HFpEF and These measures were only available in a subset of participants; however, and therefore not included in the body of the manuscript; despite supporting our overall interpretation ( Figure S1). The sample size particularly for the reference control group and HFpEF participants was admittedly low; however, the study was adequately powered to detect differences in our primary endpoint, left atrial reservoir strain. Moreover, the HFpEF participants were older and had a higher body mass index compared to the other two groups. While statistical adjustments were made to account for these group differences, this remains a limitation of this study. Finally, as with other image analysis, the data herein should be interpreted within the confines of both the image modality used to acquire the data (ie, MRI) and analysis software used to determine left atrial strain (ie, feature tracking).

| CONCLUSIONS
The data herein suggest that left atrial stiffness is elevated in a large cohort of women with INOCA. Future investigations are needed to define the mechanism contributing to left atrial stiffness in this patient cohort and to determine if lowering left atrial stiffness can prevent heart failure progression in this at risk population. Cipher for her assistance and expert advice on statistical analysis.