Tips and tricks for the non-invasive assessment of myocardial work: the good, the bad and the ugly

This editorial refers to ‘Impact of echocardiographic analyses of valvular event timing on myocardial work indices’, by F. J. Olsen et al. https://doi.org/10.1093/ehjci/jeac171. The non-invasive estimation of myocardial work (MW) by pressure-strain loops (PSLs) analysis is a new promising tool for the characterization of left ventricular (LV) performance.1 This method has the advantage of considering the impact of loading conditions on LV deformation without the need to perform left heart catheterization.2,3 Practically, as shown in Figure 1, the LV pressure is estimated starting from a referral curve that is scaled using aortic and valvular timings and then merged with strain data to obtain segmental LV PSLs and subsequent MW parameters.4 Description of the impact of the aortic valve (AVC) timing on the calculation of MW. Previous publications have shown the reliability and reproducibility of MW parameters in different clinical conditions, such as CRT, heart failure, and ischaemic heart disease.2,3,5–10 A sub-analysis of the NORRE study and one monocentric paper have provided the reference ranges for MW parameters in healthy subjects.11,12 Moreover, a ‘How to’ paper previously published in this journal underscored that the reliability of PSLs analysis relies essentially on three points: (i) the quality of the apical two-, three-, and four-chamber view, (ii) the careful speckle tracking analysis, and (iii) the assessment of ambulatory blood pressure, that should be obtained at the end on the echocardiographic examination, with the patients lying in a calm and comfortable position for at least 15 min.1 Interestingly, no specific indication has been provided concerning the best method for the measure of valvular event timing. Scrolling the available literature on this topic, we can find substantial heterogeneity among papers. Some used a visual approach reporting the opening and closing time points of the aortic and mitral valves assessed in an apical three-chamber view, or in a parasternal long axis view.5,8–10,13 Some assessed the PW Doppler of the LV outflow tract and mitral valve.5 Finally, in some other papers, the specific method for the estimation of valvular timing is not reported, portending the hypothesis that this was not considered as a potential issue for the reliability of results. In their very interesting work, Olsen et al.14 aimed at assessing the impact of the visual method vs. PW Doppler and colour tissue Doppler imaging M-mode for the identification of valvular events on the estimation of MW parameters. They found that the visual method, systematically overestimates valvular timings and that this ‘systematic error’ significantly influences the evaluation of all MW parameters.14 This can be attributable to the lack of specific ‘rules’ for the visual assessment of valvular events, but also to the higher temporal resolution of the Doppler-based methods (Figure 2). In a previous paper,3 we demonstrated that the estimation of the LV pressure through the referral curve proposed by Russell et al.15 was substantially imprecise, particularly during the diastolic phase when we observed the greater error between the measured and estimated LV pressure curve.3 The paradox is that, although the estimation of LV pressure is imprecise, the deducted estimation of LV-work was accurate because the temporal integration obtained through valvular timing induces a smoothing of the difference between measured and estimated works. Of course, in our study, we used always the same method (visual) for the estimation of valvular timings, which significantly reduced the source of error.3 Impact of the definition of the valve opening and closure (MVC, AVO, AVC, MVO) on the building of pressure curves (MVC, mitral valve closure; AVO, aortic valve opening; AVC, aortic valve closure; MVO, mitral valve opening; SYS, systole; IVC, isovolumic ventricular contraction). Therefore, the results of the paper of Olsen et al.14 show that a careful assessment of valvular timing has pivotal importance for the reliable assessment of MW. First, it is fundamental for scaling the referral pressure curve that is used to estimate LV performance (Figure 2). Second, it can affect the duration of the ejection and relaxation phases, which impacts the estimation of constructive and wasted work. In the last few years, an increasing amount of literature has been produced, applying the non-invasive estimation of MW to very different clinical scenarios. Nevertheless, as for any new technique, we have to avoid excessive simplification and proceed cautiously to prevent avoidable mistakes. Two recent applications of the non-invasive assessment of MW in the field of aortic stenosis16,17 and pulmonary hypertension underscore the undeniable potentialities of this approach for better stratifying LV and right ventricular (RV) function.18,19 On the other hand, they also emphasize the need for further investigation to increase the precision of the method. In the field of aortic stenosis, the visual assessment of aortic valve opening and closure can be particularly challenging, because of valve calcifications, so contributing to errors in measurement.16 In the case of the estimation of RV performance, the application of PSL analysis is even more challenging because the RV pressure curve does not exactly correspond to the LV pressure curve provided by Russell et al.15 and because the assessment of valvular timing might be less reliable due to hangout interval. Other methodological issues exist: wall thickness and ventricular radius can impact on the calculation of segmental PSLs.10 If we take into account the limitations listed above, the MW software is ‘imperfect’ but very probably sufficient in most clinical situations. As for any new techniques, users should be aware of the limitations, tips and tricks related to its application. The careful measurement of valvular timing is now another issue that we should consider when we assess MW by echocardiography. The higher reproducibility and reliability of the Doppler-based methods compared with the visual approach demonstrated by Olsen et al.14 seems to favour their application in clinical practice. If this is not possible, always the same technique for the estimation of valvular timing should be used when following-up a patient. The non-invasive estimation of MW in specific clinical conditions such as aortic valve stenosis, severe LV dilatation/hypertrophy and the application to the RV, requires improvements and further experimental and clinical validations.