Evaluating the characteristics of myocardial work by two-dimensional echocardiography during pre-eclampsia pregnancy

This study aimed to analyze the changes in myocardial work (MyW) properties and the correlation of MyW with cardiovascular and clinical indices during the pre-eclampsia (PE) pregnancy. Standard two-dimensional and speckle-tracking echocardiography were sequentially performed on 77 women with PE and 89 with normal pregnancy. Four components of MyW: global myocardial work index (GWI), constructive work (GCW), wasted work (GWW), and work efficiency (GWE) were measured. The significant increased GWI, GCW and GWW were observed, while GWW elevated more than GCW with consequently resulting the decline in GWE among PE cases. Although there was a diverse relationship between MyW components and LV morphological as well as functional indices, MyW parameters were significantly correlated with the grades of arterial hypertension and the incidence of adverse outcome of PE. With the hypertension stages, GWI, GCW and GWW gradually increased but GWE decreased. Meanwhile, the higher GWI and GCW and the lower GWE, the more adverse events occurred in PE group. In conclusion, during the PE pregnancy, GWI, GCW and GWW increase, while GWW elevates more than GCW, which leads to the decrease in GWE. Moreover, the changes in MyW are associated with the hypertension grades and the poor prognosis in PE. The non-invasive manner for MyW assessment provides a new perspective on the myocardial biomechanics, cardio-metabolic conditions and pathophysiological changes in the condition of PE.


Introduction
Acute pre-eclampsia (PE) is a pregnancy complication that occurs in 5-8% of pregnancies with the characterizations of new-onset of elevation in blood pressure (BP) and proteinuria [1]. The pathological process in this disorder is mainly ischemic changes in nature and is known to injury maternal heart, liver, kidney and brain [2]. Previous studies have shown that PE is associated with higher morbidity and mortality of cardiopulmonary manifestations when compared to normotensive pregnancy [2][3][4].
There are many reports about asymptomatic left ventricular (LV) abnormal function, structure remodeling and myocardial impairment during PE [5][6][7]. LV ejection fraction (LVEF) and stroke volume which reflect global performance, and myocardial velocity as well as strain which are regional function indices, are commonly used as the substitutes for cardiac function. Although able to be acquired relatively easily and non-invasively, all those shortening measurements are still limited to the load dependency and not able to reflect myocardial work and oxygen consumption.
Myocardial work (MyW) that incorporating longitudinal strain and cardiac load within a given time interval, emerged as an alternative tool for evaluating LV systolic function more than 40 years ago [8]. However, because of its invasive pressure measurement, the widespread application of MyW in clinic has been limited. Recently, Russell et al. proposed a non-invasive manner for MyW assessment, which based on non-invasive LV pressure acquired from brachial BP and strain data derived from speckle tracking echocardiography [9]. Due to integrating afterload and strain together, MyW is considered to be a less load-dependent index when compared to LVEF or myocardial deformation parameters [10,11]. To date, the technique has been reported in cardiac synchronization therapy and other cardiac disorders with good results [10][11][12][13].
Notwithstanding these documents, there is a paucity of reports on the changes in MyW during the pregnancy complicated by PE. The objective of this study was thus to: (1) investigate the characteristics of MyW indices between normal and PE pregnancy; (2) differentiate myocardial constructive from wasted work components in the pregnant women with PE; (3) analyze the correlation of MyW measurements with cardiovascular and clinical properties during the PE pregnancy.

Study subjects
This was a case -control study carried out from May 2020 to May 2022. The Affiliated Hospital of Qingdao University Ethics Committee approved the study protocol. After informed consent, all singleton pregnant women with on-set PE from the single tertiary center were consecutively enrolled as cases. PE was defined as the following criteria: hypertension arising de novo after 20 weeks of gestation (systolic BP ≥ 140 mmHg or diastolic BP ≥ 90 mmHg measured on two occasions at least 4 h apart) and significant proteinuria (> 0.3 g/24 h or urinary albumin to creatinine ratio ≥ 30 mg/ mmol) [1]. All the PE patients were hypertensive arising de novo without receiving any therapy preceding the study. Moreover, using body mass index (BMI) ≥ 30 kg/m 2 , PE women were divided to two subgroups: normal and high BMI [14]. Normotensive healthy pregnant women matched for maternal age and gestation were recruited as control subjects from the routine antenatal clinic. The exclusion criteria were pre-existing chronic hypertension, multiple pregnancy, known heart disease, pulmonary hypertension, renal disease, diabetes, systemic lupus erythematosus, any connective tissue disease and antiphospholipid syndrome. Were considered as complications: eclampsia, heart failure, cerebrovascular accident, HELLP syndrome, placental abruption, maternal or fetal death, fetal growth restriction, fetal distress, hospitalization of more than 7 days for the newborn due to tension disorder, neonatal weight < 2500 g. In addition, all were Han-Chinese and no-smokers.

Blood pressure and hemodynamic assessment
Before ultrasound scanning, brachial BP was measured on the right arm using an oscillometric device (OMRON 9020, OMRON Healthcare, Dalian, China) and the casual BP level was measured by the average of three readings 1 min apart. Mean BP was calculated as (systolic BP + 2*diastolic BP)/3. On the basis of the present guidelines [15], the PE patients were classified as 1,2 and 3 hypertension grades.

Echocardiographic examination
According to the current recommendations [16], all participants underwent an extensive transthoracic echocardiography using a Vivid E95 ultrasound system equipped with a M5S 3.5 mHz transducer (GE Vingmed Ultrasound, Horten, Norway). All image recordings consisting of three consecutive cardiac cycles with a frame rate of 60 to 80 s − 1 were obtained and saved in DICOM format for offline analysis (EchoPAC, GE Healthcare).
Following the American Society of Echocardiography guidelines [16], we measured interventricular septum, posterior wall, LV diameters at end-diastole and end-systole via M-Mode in the parasternal long-axis view. Then, LVEF, stroke volume, relative wall thickness (RWT) and sphericity index as well as LV mass were calculated. Left atrial volumes were measured by the biplane method. LV end-diastolic volume (LVEDVI), stroke volume (SVI), LV mass (LVMI) and atrial volume (LAVI) were normalized by body surface area, respectively. Subsequently, transmitral inflow peak early (E) -and late-wave velocities were obtained via pulsed-wave Doppler. After that, the ratio of transmitral flow E peak to mitral annular velocity (e) by tissue pulsed Doppler was calculated for estimating LV filling pressure [17]. As suggested previously, Tei index was calculated as the sum of isovolumetric contraction and isovolumetric relaxation divided by total ejection time [18]. Using two-dimensional speckle tracking echocardiography, global longitudinal strain (GLS) was acquired in a 17-segment LV model from the apical four, two and three views.

Myocardial work measurements
The measurements of MyW components were performed using an off-line vendor-specific module (EchoPAC Version 202, GE Vingmed Ultrasound). As proposed by Russell et al. [9], after inserting values of brachial BP, quantifying data on GLS and indicating the timing of aortic and mitral valve events by echocardiography, the software constructed non-invasive LV pressure-strain loop. Then, the area of the loop, which was calculated from mitral valve closure to opening, served as the index of global MyW. The values of segmental and global work index (GWI) were shown in a bull's eye. Moreover, additional elements of MyW were obtained. Altogether, the parameters of MyW were derived as follows [19,20]: GWI, total work with in the area of the LV pressure-strain loop during mitral valve closure to opening; global constructive work (GCW), MyW performed during LV shortening in systole adding negative work during lengthening in the isovolumic relaxation stage; global wasted work (GWW), MyW performed during lengthening in systole adding LV shortening during the isovolumic relaxation stage; global work efficiency (GWE), GCW divided by the sum of constructive and wasted work (Fig. 1).

Statistical analysis
Statistical analysis was performed using SPSS version 26 (SPSS Inc., Chicago, IL, USA). Normality of the distribution of continuous variables was examined using the Kolmogorov-Smirnov test. All data were expressed as mean ± standard deviation (SD) or median (interquartile range) as appropriate. The 95% confidence interval was calculated as ± 1.96 SDs from the mean. Differences between groups were analyzed for statistical significance by the unpaired t-test for normally distributed variables and the Mann-Whitney U test for non-normally distributed variables. Correlation between continuous variables was performed by Pearson's or Spearman's correlation coefficient. Multivariable linear regression analyses were done to examine the independent correlates between cardiovascular and clinical parameters and MyW components, and a stepwise selection was used. Additionally, the variables (BP, GLS) used for the calculation of MyW were excluded from the multivariable model. The inter-and intra-observer agreement for MyW indices were assessed in 20 randomly selected participants using Bland-Altman plot analysis. P < 0.05 was considered as statistically significant.

Clinical and hemodynamic characteristics
Clinical and hemodynamic data of all subjects are presented in Table 1. The age and gestational week were closely matched in these women. There were no differences in the maternal weight, body surface area and heart rate between the two groups. The PE cases had higher BMI, BP, and SVI than the normal pregnant individuals (P < 0.001). Meanwhile, due to a slight decrease in heart rate, lower cardiac index was detected in the PE group. Furthermore, the number of PE patients with grade 1, 2 and 3 hypertension was 41, 25 and 11, respectively, with 53 patients in the normal BMI and 24 ones in the high BMI subgroup. Additionally, there were 54 cases with adverse outcomes in PE group and 1 in the controls (P = 0.000).

LV morphological and functional remodeling
Results of the echocardiographic assessment are listed in Table 2. When compared to control subjects, the women with PE were characterized by the larger chamber of LV and LA with higher sphericity index, higher LV mass, even after indexed by body surface area (P = 0.000). Although within normal value, LVEF was slightly lower in the PE group accompanied by a remarkable decline in GLS (P = 0.000). Furthermore, the prolonged Tei index, decreased mitral flow E peak and tissue Doppler e velocity with the significantly higher value in E/e ratio were recorded in the PE patients (P = 0.000).

Myocardial work assessment
Evaluation of MyW components of the study population during normal pregnancy and PE are summarized in Table 3. The measurements of GWI, GCW, GWW in the PE pregnant women were higher than those in normal group (P < 0.001), while the increased proportion were 16.60%, 24.37% and

Study subjects
Initially, 185 subjects were recruited in the study: 90 with PE and 95 with normal pregnancy. Thirteen individuals (5 with poor quality images, 5 with pregnancy-related diabetes mellitus, 2 with systemic lupus erythematosus and 1 with sicca syndrome) in PE group and six participants (3 poor quality images, 1 gestational hypertension and 2 pregnancyrelated diabetes mellitus) in controls were excluded from the analysis. Finally, a total of 77 cases with PE and 89 healthy pregnant women were included in the study.   145.38%, respectively. Meanwhile, a significant reduction of GWE was observed in the pregnancies complicated by PE (Fig. 2). Additionally, there was no remarkable difference in the MyW indices between the normal BMI to high BMI subgroups during PE pregnancy (Table 4).

Repeatability and reproducibility
Bland-Altman plots analysis for intra-and inter-observer variability exhibited good repeatability and reproducibility in MW indices (Fig. 3).

Fig. 3
Bland-Altman analysis for inter-and intra-observer variability showed good repeatability and reproducibility in all four myocardial work indices. A-D, analysis for inter-observer variability in GWI, GCW, GWW and GWE, respectively. E-F, analysis for intra-observer variability in GWI, GCW, GWW and GWE, respectively. GWI, global work index; GCW, global constructive work; GWW, global work waste; GWE, global work efficiency to both systolic and diastolic function. Similar results have been obtained from other studies [11].
In the present study, the significantly higher BMI in PE cases than controls were observed, which could independently impact the indices measured. Therefore, with BMI ≥ 30 kg/m 2 as the cut-off value, PE individuals were divided to normal and high BMI subgroups. Statistical test showed no significant difference in MyW measures between BMI ≥ 30 kg/m 2 to BMI < 30 kg/m 2 PE women. Furthermore, stepwise regression analysis demonstrated no association between MyW indices and body surface area as well as BMI.
In addition, although BP and GLS were excluded from the analysis for being used to calculate MyW, the relationship between the grades of arterial hypertension and MyW parameters was detected, which with the increasing of hypertensive stage, GWI, GCW and GWW became gradually higher but GWE lower. Moreover, there was remarkable difference in the incidence of poor prognosis between PE patients versus normal pregnancy. In the condition of PE, the higher GWI and GCW and the lower GWE, the more adverse events occurred.

How to explain the changes in MyW components during PE pregnancy
Since arterial pressure increases during the pregnancy complicated by PE, stroke work performs in a shortening time interval, compensated by the LV raising myocardial oxidative metabolism, improving energy level to enhance the ability of the LV to pump against the elevated afterload [23]. Except for taking into account the increased cardiac loading, higher GWI and GCW reflect the changes of cardiac contraction and metabolism in PE gestation, but not detected by LVEF and GLS.
The element of GWW is the work done by the ventricle but does not generate stroke volume. It has been reported that the increase in LV end-systolic stiffness, a parameter of myocardial contractility indicating LV pumping against a given pressure, is related to the increase in arterial afterload. As such, in the present study, higher GWW may also be associated with higher wall stress of myocardium against the elevated arterial pressure in the women with PE. In addition, because metabolic substrates cannot be recycled by the myocytes during ventricular wall stretching, the energy will be converted to heat and storage in the myocardium, which is so-called internal cardiac work and partially accounts for the ventricle suction during the early diastolic phase [24,25]. And the higher ventricular mass, the higher values of GWW in the study. Therefore, the upward GWW should indicate an increase in the contractile reserve in the condition of PE. the higher GWI and GCW and the lower GWE, the more adverse events occurred in PE group.
To our knowledge, this is the first report about the characteristics of MyW indices during the pregnancy complicated by PE. In the present study, the measurements of GWI, GCW, GWW, and GWE derived from the new non-invasive way exhibited excellent feasibility and a good intra-and inter-observer variability, thus, rendering MyW a reliable tool to assess cardiac function in the PE pregnant women. Furthermore, the new method allows us to distinguish the changes in each MyW component, therefore providing a new perspective on the myocardial biomechanics, cardiometabolic conditions and pathophysiological changes during PE pregnancy.

Characteristics of MyW during PE pregnancy
Early works reported the declined longitudinal deformation in the PE cases despite LVEF value within the normal ranges [3,21]. The data in the current study confirmed their findings. However, LVEF and strain represent myocardial shortening but neglect the load condition. It has been proven that the increase in afterload can result in the decrease in contractile shortening and lead to misinterpreting the true systolic function [10,22]. Instead, MyW takes into account afterload as well as myocardial deformation together and is therefore considered to be a less load-dependent parameter for evaluating cardiac performance. Using the new non-invasive method, the current study presented a significant increase in GWI, GCW and GWW in the PE pregnancies versus normal gestations. Further, among those MyW components, the increment in the median GWW values far exceeded the increment in GCW (145.38% versus 24.37%, respectively), which causing the remarkable decrease in GWE.

Correlation of MyW measurement with cardiovascular and clinical indices in PE
Despite weak association, the study demonstrated the variable relationship of MyW properties with cardiac structural and performative indices. Higher GCW and GWI were correlated with higher SVI, larger LA volume, higher sphericity index and higher RWT. GWW was positively related to LVMI and inversely to tissue Doppler e velocity. Meanwhile higher GWE was linked to LVEF and higher tissue Doppler e. Furthermore, Tei index, a time-interval index for comprehensive evaluation of diastolic and systolic function, showed a negative correlation with GCW, GWI, and GWE, and a positive correlation with GWW. Thus, among MyW components, probably, GWI and GCW are related to myocardial contractility, GWW to diastolic function, and GWE 145.38%, respectively. Moreover, a significant reduction of GWE was detected in the pre-eclamptic women. GWI, global work index; GCW, global constructive work; GWW, global work waste; GWE, global work efficiency. Figure 3. Bland-Altman analysis for inter-and intraobserver variability showed good repeatability and reproducibility in all four myocardial work indices. A-D, analysis for inter-observer variability in GWI, GCW, GWW and GWE, respectively. E-F, analysis for intra-observer variability in GWI, GCW, GWW and GWE, respectively. GWI, global work index; GCW, global constructive work; GWW, global work waste; GWE, global work efficiency.
However, the findings could not be observed by the strain alone for not taking into account pressure.
Finally, GWE represents work efficiency of myocardium and is calculated indirectly from the ratio of GCW and GWW. Since an unproportional increase in constructive and wasted work properties, GWE decreased significantly in the PE group, which might imply the imbalance between cardiac stroke work and myocardial energy metabolism during the PE pregnancy.

Limitation
There were some limitations in the study as follows: Firstly, selection bias may be unavoidable in this study because it is a cross-sectional, single-center study with a limited number of participants. Secondly, BP readings derived from sphygmomanometer in the brachial artery, for measuring LV pressure non-invasively, can be imprecise. Central BP should be lower than the brachial values. However, this was also a limit in the study of Russell et al. [9] and it have been considered reliable in the health subjects [10,11]. Thirdly, owing to the software supplied by a single vendor, the calculation of MyW may not independent of vendors.

Conclusion
Derived from the new non-invasive method, there were increased GWI, GCW and GWW during the pregnancy with PE versus normal controls. Meanwhile, the wasted component elevating more than the constructive led to the remarkable decrease in GWE among PE patients. e Those findings of MyW may indicate the changes in cardiac contraction and myocardial metabolism as well as the imbalance between them in PE, which not detected by LVEF and GLS alone. In addition, MyW elements were related to the hypertension grades and the occurrence of poor prognosis in the pregnancy complicated by PE. Figure 1. Using a commercially available software, analyzing LV myocardial work parameters in pre-eclamptic (upper plate) and normal pregnant (under plate) women. A, bull's eye for GWI; B, LV pressure-strain loop; C, the histogram of GCW (green) and GWW (blue); and D, the values of myocardial work analysis. GWI, global work index; GCW, global constructive work; GWW, global work waste; GWE, global work efficiency; LV, left ventricle. Figure 2. Comparison of myocardial work components between pre-eclamptic and normal pregnant individuals. The measurements of GWI, GCW, GWW in the pre-eclamptic group were remarkably higher than those in normal pregnancy, the degrees of elevation were 16.60%, 24.37% and