Afterload-related reference values for myocardial work indices

The novel noninvasive pressure-strain loop (PSL) is a reliable tool to reflect myocardial work (MW) in a variety of hemodynamic states by comparison with invasive experimental and clinical studies. It has shown a strong correlation with regional myocardial glucose metabolism by positron emission tomography (PET) [1‐4].

The PSL algorithm has two advantages over strain measurements in evaluating left ventricular function. Firstly, the main limitation of strain imaging is load dependency [5]. An increase in afterload can lead to decreased strain giving rise to misinterpretation of the true contractile function, which in turn leads to false conclusions with regard to reduced myocardial function. However, MW takes into account deformation as well as afterload, potentially offering incremental value to myocardial function assessment. This advantage was proved by a canine experiment where a substantial decrease in longitudinal strain (LS) occurred, whereas myocardial work index (MWI) was unaltered during aortic constriction [6]. Secondly, MW parameters are calculated as integration over time of the strain rate obtained by differentiating the strain curve multiplied by the instantaneous left ventricular pressure (LVP). LVP was estimated by a surrogate of the LVP curve. Therefore, MW indices were used to address all the changes noted in the myocardial strain along with LVP changing during the cardiac cycle and provide more comprehensive information for evaluating left ventricular (LV) function. By contrast, LS, the most commonly used strain index, is used to assess the peak systolic strain value, but not the process of obtaining the peak strain. It is well known that LVP changes over time in the cardiac cycle. Therefore, the same strain occurring at different time periods during systole may achieve different MW measurements. The identical peak LS value, which occurs in early or mid-or end-systole may not correspond to identical myocardial work. However, this information cannot be represented by LS, nor can it be reflected in case of transient hypokinisis, akinesis or even paradoxical movement, which can affect the MW. In other words, different patients may present significantly different MW values even if they have the same LS and the same systolic blood pressure (SBP) (Fig. 1).

Therefore, theoretically, MW is the most comprehensive indicator of left ventricular myocardial function. Previous studies have reported reference values for MW indices grouped by gender or age [7‐9]. Despite these reference values, the wide variability of MW indices prevents the application of this technique in the routine clinical setting. Multivariable linear regression analysis for MW indices has shown that SBP rather than gender or age was the only independent factor following adjustment for confounders [7]. Therefore, it is not advisable to assess MW in the absence of SBP. In other words, a given MW level for a patient with SBP of 90 mmHg has a significantly different meaning than the same MW for a patient with SBP of 180 mmHg. Given the superiority of MW, an urgent need is required to establish reference values for MW parameters grouped by SBP. The present study recruited healthy subjects without hypertension and subjects in the early stage of hypertension without myocardial remodeling. To the best of our knowledge, the present study provided for the first time the reference values for MW indices based on SBP grading over a wide range to benefit the application of this new indicator in the routine clinical diagnosis.

Of the 248 subjects included in the present study, 18 were excluded from further analysis due to poor image quality. Finally, a total of 80 men and 150 women were included. The demographic data of the population listed in Table 1 are grouped by SBP in 10-mmHg subgroups.

EF, GLS and MW indices grouped by SBP are presented in Table 2. No significant differences were detected for EF and GLS among SBP-groups. When SBP ranged from 90 to 189 mmHg, GWI and GCW tended to increase linearly with the afterload. Significant differences were noted among the SBP-groups for GWI and GCW. For GWI, the majority of the statistical comparisons of the differences at each SBP-group were significant with the exception of the comparisons for every two adjacent groups. For GCW, the majority of the differences between groups were statistically significant with the exception of the comparisons between the SBP 130–139 mmHg and 140–149 mmHg subgroups, the SBP 140–149 mmHg and 150–159 mmHg subgroups and the SBP 150–159 mmHg and SBP ≥ 160 mmHg subgroups (Table 2, Fig. 2).

Overall, GWW tended to rise with the increase of SBP, but not all of the differences in GWW were significant at each SBP group. The differences among the SBP 90–99 mmHg, 100–109 mmHg and 110–119 mmHg subgroups were not statistically significant. Non-significant differences were also noted among the SBP 100–109 mmHg, 110–119 mmHg, 120–129 mmHg and 130–139 mmHg subgroups. Similarly, non-significant differences were noted among the SBP 120–129 mmHg, 130–139 mmHg, 140–149 mmHg and 150–159 mmHg subgroups. The GWE values remained stable among all SBP groups, with the exception of a slight drop noted when SBP exceeded 160 mmHg (Table 2, Fig. 3).

Intra-observer and inter-observer analyses demonstrated optimal repeatability and reproducibility with regard to the MW parameters (Table 3).

MW is derived from pressure–volume or pressure-length loops and it has been investigated for more than 40 years [14‐17]. MW assessment was initially measured invasively during cardiac catheterization, which limited its widespread use in clinical practice. Russell et al. introduced recently a method of PSL (in mmHg%) for calculating MW non-invasively, which involved the combination of STE with LVP as estimated from brachial artery cuff pressure [1]. The accuracy of the novel method has been validated by subsequent studies [2‐4]. A number of clinical researches have been performed within the last two years, since this technique was commercially available. To date, MW has been investigated with regard to cardiac resynchronization therapy (CRT) [18‐23], in the diagnosis of different categories of coronary heart disease [6, 24‐28] and in the evaluation of cardiomyopathy [29‐33]. In addition, MW can also be used to predict and evaluate therapeutic effects [34‐38].

However, due to the lack of accurate reference values used for the MW indices, this method is limited to clinical research but cannot be used for routine clinical examination. Previous studies have obtained reference values of MW grouped by gender or age [7‐9]. However, correlations between MW and demographic variables have been investigated, showing the absence of a strong dependence of MW indices on age or gender [7, 8]. Multivariable analysis revealed significant correlations only with SBP for MW parameters [7]. Morbach et al. observed an upward shift of GCW and GWW with advancing age [9]. This finding can be explained when considering the increase of SBP with increased age, even if the SBP remains in the normal range. The present study is the first to measure MW reference values using SBP as a grouping standard. According to our results, GWI and GCW varied greatly according to the different afterload conditions. Therefore, it is unreasonable to assess whether the MW of a given patient is in the normal range in the absence of SBP. For one thing, some patients with myocardial dysfunction will be incorrectly mistaken as normal subjects if the examination results are interpreted using the normal reference ranges previously provided by the EACVI NORRE study [7] for all the patients without taking SBP into account. For example, the normal lowest value of GWI in men was set as 1,270 mmHg%. However, according to our results, this standard was set too low for patients with SBP higher than 100 mmHg, which can lead to misinterpretation of the examinations and result in false negative conclusions. A similar finding was noted for GCW. For another, certain studies have used PSLs to predict myocardial dysfunction and obtained the optimal cut-off values but not considered the effect of SBP on MW [6, 26]. For example, the optimal cutoff for GWI was established as 1,810 mmHg% to predict significant CAD [26]. However, according to our results, it can be normal if GWI is less than 1,810 mmHg% as long as the SBP is low. Clearly, a low GWI due to low SBP can be mistaken for cardiac abnormality resulting in a false positive result. This reason may account for the low diagnostic specificity in that study. Therefore, the critical effect of afterload on MW cannot be ignored during clinical research or diagnosis and is required to make a reasonable judgment on the myocardial function.

Normal reference values have been reported for GWI and GCW and these indices correlated positively with SBP [7‐9, 39]. However, no studies have been conducted with regard to the reference values of MW when SBP is above normal. The changes in these indices following increased afterload have not been fully investigated. It is well known that a large percentage of heart disease patients present with hypertension. Therefore, the afterload-dependent reference values are required in this subset of patients for further interpretation. According to our results (SBP of 140–189 mmHg), LVEF and GLS values were preserved in hypertensive subjects without myocardial remodeling and GWI and GCW exhibited a linear association with SBP (Fig. 2). However, this finding may not apply to population with much higher blood pressure. The results demonstrated that the LV myocardium may function at higher energy levels against the increased afterload to preserve LV contractility during the compensatory phase. It is interesting to note that the results from the exercise stress echocardiography confirmed our findings. Healthy subjects demonstrated increased GWI with elevated SBP during exercise, whereas in patients with inducible ischemia, GWI did not increase and MWI was decreased in the affected segments [25, 40, 41]. True myocardial contractility was more likely to be detected under high afterload. The present study was the first to provide work reference for heart disease patients with hypertension. The potential ability to detect myocardial dysfunction under different loading conditions can be employed and assessed in future studies.

The present study demonstrated that GWE values remained constant across all SBP-groups since almost a proportional relationship was noted in both GCW and GWW. Since GWE was not affected by afterload [7, 40, 41] and had a stable reference value in all healthy subjects, it may be more suitable than other parameters as a diagnostic index of myocardial impairment. In some studies, it has been proved that GWE is the best predictor of LV myocardial contractile performance in all MW parameters [27, 28]. However, additional work is required to assess its efficacy in other types of heart disease.

This study has some limitations. Firstly, this was a single-center study including a limited sample size, which may not be sufficient to provide particularly accurate reference values for MW parameters. However, the present study highlighted the importance of afterload in evaluating MW by examining the association between these two parameters and emphasizing on the influence of afterload. This finding cannot be ignored in the clinical research or diagnosis and is required in order to make a reasonable judgment on the myocardial function. Secondly, the present study did not include subjects with SBP > 190 mmHg. Therefore, we could not provide the MW reference and establish the variation trend of MW with SBP in that population. Thirdly, it should be noted that the use of PSL did not provide a direct measure of MW, but rather an index of this parameter due to pressure rather than wall stress being assessed in the method.

The amount of MW but not the work efficiency varied greatly according to the different afterload. This finding cannot be ignored during clinical research or diagnosis and is required to make a reasonable judgment on the myocardial function.