Though there are many measures available to characterise the performance of LV, to the best of our knowledge, this is the first study to report a comprehensive comparison of measures of LV systolic performance that includes strain energy density in patients with hypertension, DCM, and amyloid heart disease. GLASED provides 4 important advantages over most previous methods. GLASED is: i) based on sound engineering principles; ii) calculates the fundamental measure of energy production (i.e. work done) per unit volume of myocardium; iii) accounts for the known confounding variables such as haemodynamic load and ventricular remodelling; iv) matches the anticipated symptom severity, brain natriuretic peptide (BNP) and mortality in our cohorts.
The synopsis of results for each cohort was as follows:
The hypertension cohort had a normal EF, a mildly reduced EFc, a moderately reduced myocardial contraction fraction and longitudinal strain with a normal longitudinal stress. The greater stroke volume/height2.7 was consistent with a high-output state. Both the TIN and longitudinal generated forces were moderately increased. Pressure-strain loop was unchanged. Stroke work and CASE were moderately increased, and GLASE was mildly elevated, consistent with the extra work needed to generate the elevated blood pressure. A significant reduction in GLASED would suggest early hypertensive heart disease that was driven by the fall in longitudinal shortening of 30%. CASED, however, was not reduced as midwall circumferential shortening did not fall significantly. The explanation for difference between CASED and GLASED is uncertain but may reflect subendocardial ischaemia in the hypertension cohort due to microvascular disease.
The EF, EFc, myocardial contraction fraction, longitudinal and midwall shortening were all severely reduced in the DCM cohort. The cohort also had a normal mural thickness, with remarkably high longitudinal and TIN force production. The EFc was higher than the EF. All the stresses in dilated cardiomyopathy were higher than healthy controls. Stroke work, pressure-strain loop, GLASE, GLASED, CASE and CASED were all moderately reduced, consistent with both a reduced LV energy production and reduced energy production per unit muscle volume.
Amyloid heart disease
The EF was mildly reduced in the amyloid cohort, but the EFc, myocardial shortening, proportional radial thickening and all the stresses were severely decreased. The peak longitudinal force and TIN force were increased compared to the control group. Myocardial contraction fraction in the amyloid and DCM cohorts were the same. Stroke work per myocardial mass was also markedly decreased. Pressure-strain loop was reduced to a similar degree as DCM. In contrast, the GLASE, GLASED, CASE, CASED and SSP were extremely low. The amyloid cohort had the lowest ASE and ASED of the four groups, suggesting the worst systolic dysfunction and poorest contractile function. This suggests that the notion of amyloid heart disease being the archetypal form of diastolic dysfunction with relatively preserved “systolic function”, as defined by EF, needs re-evaluation.
Strain energy density vs alternative methods
The EFc was better at quantifying abnormalities compared with the EF. GLASED provided a sensitive means of detecting mild hypertensive heart disease and was different between the amyloid and DCM cohorts. A finding that contrasted with the alternative methods. Clear distinctions existed between the EF and GLASED in the four cohorts. The EF is emergent from, and contingent upon, known variables (Table 1).10–12, 14,16–18 ASED the same input variables as EF, namely myocardial strain, mural thickness, and internal dimensions but also includes pressure-generation information. A decrease in longitudinal or midwall shortening decreases EF, EFc, GLASED and CASED (Table 1). In contrast, increasing LV diameter decreases EF but raises ASED. A greater mural thickness increases EF (without an improvement in pump performance) but decreases ASED (Table 1).
It can be appreciated why EF is not an ideal method of measuring LV performance since it does not allow for either the effect of haemodynamic load or geometric changes. The EF, under high-loading conditions, is maintained because of the adaptation of the myocardium and the structural changes such as parallel cardiomyocyte hypertrophy. Of the thirty-two individuals with a preserved EF, just under one-fifth of our overall cohort, had significantly reduced GLASED values. This suggests that the use of EF would miss a sizeable proportion of patients with important myocardial dysfunction. In contrast, the EFc, by removing the confounding effects of LV mural thickness and dimensions, 17,19 only misses nine individuals, or one-twentieth, of those with a reduced GLASED.
Work performed by the left ventricle was calculated using both the stroke work (pressure-volume loop) and the pressure-strain loop methods25. Neither of these techniques distinguish between normal and hypertension or between DCM and amyloid because they do not correct for wall thickness or ventricular size by calculating wall stress.
Myocardial forces, stresses and strains
Myocardial forces such as TIN and longitudinal forces were increased in each of the disease cohorts especially so in DCM. Correcting these forces for LV mass improved their utility although they remained high in DCM. Using myocardial stress alone missed the abnormalities in hypertension and was misleading in assessing the contractile abnormalities of DCM. Myocardial strains were reduced in all the disease cohorts apart from midwall shortening in the hypertensive cohort. Importantly, myocardial strains alone did not differentiate amyloid from DCM.
GLASED vs Laplace stress-strain product
The GLASED values provide a more exact assessment for ASED compared with SSP, since it uses the more appropriate Lamé equation to calculate the stress for a thick-walled chamber. The simpler SSP gave comparable results, albeit with a slight underestimate in the amyloid cohort and an overestimate in the DCM cohort. SSP, however, is an alternative with a simpler equation to use.
Symptoms and prognosis
We did not assess either the severity of symptoms, BNP or prognosis in our cohorts. The reduced indexed stroke volume in the amyloid and DCM cohorts would suggest our patients had mild low-output states and heart failure. Symptom severity is usually marked in amyloid heart disease, moderate in treated DCM, minor or absent in hypertensive heart disease. Plasma BNP levels are correlated with both symptoms and prognosis. BNP is highest in amyloid heart disease (~ 690 pg/mL), followed by DCM (~ 170 pg/mL), hypertension (~ 26 pg/mL) and normal individuals (~ 10 pg/mL).26,27 This trend is reflected by our GLASED data (Fig. 4A). The expected 10-year mortality in amyloid heart disease, DCM, and hypertension is 95%, 65% and 1.5% respectively (Fig. 4B).28–30 Reduced myocardial strain,4,31,32 higher internal diastolic diameter,33 low blood pressure in heart failure34 and concentric LV hypertrophy (i.e. increased wall thickness) 35,36 are individually associated with an increased risk, and each of these risk factors reduces GLASED (Table 1), indicating that GLASED has a highly promising role in risk assessment.
Implications for heart failure
A key physiological requirement of the heart is to maintain adequate tissue perfusion both at rest and with exercise, even in the presence of myocardial disease.37–39 The left ventricle achieves this through remodelling.37–39 Tissue and organ perfusion is, however, independent of the EF, since resting stroke volume and cardiac output are often preserved in heart failure.39 There is a high incidence of hypertension, concentric hypertrophy, reduced wall stress and reduced myocardial strain in heart failure with a preserved EF.40 Assessment of suspected HFpEF patients using GLASED would provide additional insight to the complicated mechanisms of this enigmatic and heterogenous condition by establishing the contribution of myocardial contractile abnormalities in HFpEF subgroups.
Myocardial function and contractility
The phrase “myocardial contractility” is used ubiquitously in both clinical and experimental studies. Despite this widespread use, contractility remains ill-defined 125 years after being introduced. GLASED provides a quantifiable definition of myocardial function by combining its known features, namely force generation, active stress, and deformation. Furthermore, changes in contractile function so defined would allow for the definition of inotropy to be clarified and, importantly, quantified. To date, there are no current measures of contractility that are applicable to in vitro, ex vivo and in vivo studies. Myocardial ASED, however, is applicable to both types and provides a method to compare in vivo and in vitro studies.41 We recently introduced the term contractance as a measure of myocardial function, which is defined and quantified by myocardial ASED.41 The definitions of ventricular performance and myocardial function are simplified and unified by the term contractance. We suggest that myocardial ASED could be used in clinical practice to assess the health of the myocardial muscle, since it appears to identify myocardial systolic and contractile dysfunction more accurately than current methods.
Our cohorts were of limited size and were not age or sex matched as our aim was to explore distinct disease cohorts with different, yet near symmetrical, LV geometric patterns. Details of their cardioactive medication were not available. We did not use the current recommended criteria for diagnosis of amyloid as this was a historical cohort. These limitations would not alter the conclusions of this study as the analyses are centred around the structural and haemodynamic changes. Indeed, we think that this is a strength of the methods used as they consider the structural changes that may occur with age, sex or previous hypertension as well as the haemodynamic effects of any drug therapy. We were unable to provide information from the time-volume curves as the software was unavailable when the study was performed.
Numerical methods derived from an anisotropic visco-hyperelastic non-linear constitutive model could provide more exact results but would only be possible in a research environment and performed on a small number of cases using methods such as finite element modelling.42,43 The calculations for ASE and ASED ignored the small amount of work performed in deforming myocardium alone and the minor component of kinetic energy imparted on the ejected blood’s motion. This is also a limitation of other methods such as stroke work calculations. Our study lacks external validation and direct prognostic data although the indirect evidence is highly promising (Figs. 4A and 4B). Future studies looking at GLASED potential role in assessing prognosis are required prior to widespread implementation outside of a research environment. Calculation of ASED may be prone to propagation error as several of the terms are squared. This emphasises the need for accurate measurements of the input variables in any future studies. Though we make these limitations explicit, however, we believe that those limitations do not affect our conclusion on the potential advantage of GLASED as a promising new marker of the LV performance.