Article Text
Abstract
The assessment of regional wall motion is useful to identify myocardial ischaemia because wall motion abnormalities occur relatively upstream in the ischaemic cascade. Echocardiography is widely used for this, but the subjectivity of visual observation may hamper accurate evaluation. The analysis of myocardial velocity and strain by tissue Doppler and speckle tracking echocardiography has allowed the quantitative assessment of regional wall motion and facilitated the detection of subtle myocardial deformation that is difficult to identify by conventional methods, such as post-systolic shortening (PSS). PSS is defined as myocardial shortening that occurs after end-systole (or aortic valve closure), and it is observed in the myocardium with regional contractile dysfunction. In experimental and clinical studies, it has been reported that the assessment of PSS is superior to that of conventional parameters such as wall thickening or peak systolic strain in detecting acute ischaemia and diagnosing coronary artery disease. Moreover, it has recently been found that PSS remains after recovery from brief ischaemia despite the rapid recovery of peak systolic strain. The assessment of PSS allows after-the-fact recognition of myocardial ischaemic insults and is expected to be used for ischaemic memory imaging. In this review, the usefulness of the assessment of PSS for the diagnosis of acute ischaemia and ischaemic memory is demonstrated, and issues that need to be resolved for the widespread use of this assessment in the echocardiographic laboratory are discussed.
Statistics from Altmetric.com
Introduction
Myocardial ischaemia is a pathophysiological state that occurs when myocardial blood flow cannot supply adequate oxygen for myocardial demands. The imbalance between myocardial oxygen supply and demand is followed by metabolic disorders, wall motion abnormalities, electrocardiographic changes, and finally chest pain. This sequence of events is called the ischaemic cascade.1 The downstream events in the cascade need a longer period of time for their occurrence and are not appropriate for early detection of an ischaemic event. The assessment of regional wall motion, for which echocardiography is widely used, is useful to identify myocardial ischaemia, because wall motion abnormalities induced by ischaemia occur relatively upstream in the cascade. Exercise and pharmacological (such as dobutamine or dipyridamole) stress echocardiography is well known to have excellent diagnostic accuracy for identifying ischaemia.2
There are some issues in the assessment of regional wall motion by echocardiography. The subjectivity in the assessment is the most important limitation. Because the assessment of regional wall motion is usually done by visual observation of systolic wall thickening, training and experience are needed for accurate diagnosis. To address this issue, analyses of myocardial velocity and strain derived from tissue Doppler or speckle tracking techniques have been developed. These analyses have allowed the quantitative assessment of regional wall motion and facilitated the detection of subtle myocardial deformation that is difficult to identify by conventional methods, such as post-systolic shortening (PSS).
During the last decade, many studies have demonstrated that the assessment of PSS using these analyses is useful for detecting acute ischaemia. However, various issues have prevented its widespread use in clinical settings. Since it has recently been reported that the assessment of PSS allows after-the-fact recognition of myocardial ischaemic insults (ie, ischaemic memory),3 ,4 this subtle myocardial deformation is attracting attention again. This review will attempt to demonstrate the usefulness of the assessment of PSS for the diagnosis of acute ischaemia and ischaemic memory, and it will highlight issues that need to be resolved for its widespread use in the echocardiographic laboratory.
Post-systolic shortening
Regional contraction of the myocardium is affected not only by inherent contractility of the concerned myocardium, but also by tension from the surrounding myocardium. Consequently, when regional contractility deteriorates because of ischaemia, the amplitude of shortening during ejection time decreases, and early systolic lengthening (ESL) and PSS are observed in the ischaemic myocardium.5
PSS is defined as myocardial shortening that occurs after end-systole (or aortic valve closure (AVC)) and is observed mainly during isovolumic relaxation (figure 1). PSS does not contribute to ejection of blood because it occurs after. In radial or transverse deformation, the term ‘post-systolic thickening’ is used. Various patterns can be observed in the strain profile of PSS. For example, biphasic shortening is often seen in the ischaemic myocardium, especially immediately after the onset of severe ischaemia. The latter shortening represents PSS.
There are several parameters for quantifying PSS. The post-systolic index, which is calculated from a ratio—([peak post-systolic strain]−[end-systolic strain])/(peak strain or maximum strain change during the cardiac cycle)—is widely used.6 ,7 This parameter shows the ratio of the amplitude of PSS to total shortening. The time from aortic valve closure to peak post-systolic strain is used as another parameter.
PSS is well known in the ischaemic myocardium through animal experiments with sonomicrometry; however, it has not been assessed by conventional echocardiography because of the difficulty of visual detection of subtle motion. In the late 1990s, Derumeaux et al8 reported that, in anesthetised pigs with graded reduction of coronary flow, peak systolic and early diastolic velocities derived from tissue Doppler echocardiography decreased, and a positive wave during isovolumic relaxation increased in the ischaemic myocardium. It is considered that the increase of the positive wave is due to PSS. Subsequently, the experimental and clinical studies related to PSS have been mainly conducted with tissue Doppler and speckle tracking echocardiography. The reason why there are few reports about PSS with other imaging modalities seems to be that high temporal resolution is needed to identify PSS.
Mechanisms of PSS
The potential mechanisms of PSS have been reported by some researchers. PSS was initially considered as a delayed but active contraction induced by ischaemia. As described above, however, shortening of the ischaemic myocardium is affected not only by intrinsic contractility, but also by tension arising from shortening of the surrounding non-ischaemic myocardium. Therefore, PSS can be caused by the interaction between the ischaemic and surrounding myocardium as passive recoil.5
It was controversial whether PSS represents active contraction or passive recoil. Skulstad et al9 measured LV pressure and segmental lengths by sonomicrometry and analysed LV pressure (or stress)-segmental length loops before and during coronary stenosis and occlusion in anesthetised dogs. In their results, PSS in the dyskinetic segment generated no active stress, but PSS in the hypokinetic and akinetic segments seemed to be due to active contraction.
In contrast, Akaishi et al5 calculated regional myocardial elastance as an index of contractility using a non-linear elastic model of contraction and found that peak elastance decreased in hypokinetic segments induced by ischaemia, but the timing of the peak was not delayed compared to the non-ischaemic condition. This result suggests that PSS occurs as a passive phenomenon.
In the study by Claus et al,10 which used a mathematical model to describe active force development, elasticity, and segmental interaction, PSS did not need active force development. This study concluded that PSS could be explained in a unified way as passive recoil that was the result of elastic segmental interaction (figure 2). In fact, it seems that most PSS observed during ongoing ischaemia is caused by passive recoil. In this mechanism, a relative decrease of myocardial strain during ejection time in the ischaemic region compared to that in the surrounding non-ischaemic regions is necessary for PSS. It is known that PSS increases as afterload increases.9 The theory of Claus et al can well explain this phenomenon, because the relative decrease of strain in the ischaemic region increases with afterload augmentation.
However, PSS is sometimes observed without a relative decrease of strain in the ischaemic region, especially after recovery from ischaemia.3 ,4 In simulations described above, the timing of electrical activation is not included. The delay of electrical activation induces ESL, which can cause PSS by regional preload augmentation.11 Such PSS might occur as active contraction. Further investigation is needed to address this controversy.
Diagnosis of acute ischaemia by PSS
The assessment of PSS is valuable in identifying acute ischaemia, because PSS occurs in the myocardium with regional contractile dysfunction. Tissue Doppler and speckle tracking echocardiography can clearly visualise PSS during ischaemia.12–14 The extent of PSS detected by these methods is consistent with the ischaemic area (figure 3).
It is important to determine whether the assessment of PSS can detect acute ischaemia more accurately than conventional parameters such as wall thickening or systolic strain. Previous animal experiments have indicated the superiority of PSS. Among the myocardial deformation parameters evaluated by strain and strain rate analyses, PSS parameters best differentiated ischaemic from non-ischaemic segments during acute coronary occlusion.15 Moreover, in graded coronary flow reduction, the post-systolic index increased significantly even in the flow reduction in which peak systolic strain did not decrease.16 ,17
The superiority of PSS to conventional parameters in detecting acute ischaemia has also been indicated in clinical studies. Kukulski et al 6 performed tissue Doppler echocardiography before and during elective coronary angioplasty in 73 patients with stable angina and analysed myocardial strain during coronary occlusion. In their results, PSS occurred in the risk area during occlusion, and the post-systolic index in that area increased. In diagnostic accuracy for identifying acute ischaemia, the post-systolic index was superior to peak systolic strain.
In a study by Celutkiene et al18 involving 60 patients who underwent dobutamine stress echocardiography, a velocity wave during isovolumic relaxation, which represents PSS, was observed in the ischaemic region and increased during stress. The peak velocity of PSS was the most accurate parameter of induced ischaemia (sensitivity 73–100% and specificity 82–97%) compared to systolic and early diastolic velocities (sensitivity 52–77% and 63–68%, specificity 63–77% and 59–81%, respectively).
Voigt et al7 evaluated strain and strain rate parameters during dobutamine stress in 44 patients with known and suspected coronary artery disease and compared the diagnostic accuracies of the deformation parameters with that of perfusion scintigraphy performed simultaneously. PSS was observed during stress in all ischaemic segments, and PSS parameters such as the post-systolic index were better for identifying stress-induced ischaemia than maximal strain and strain during ejection time (figures 4 and 5). The assessment of PSS by strain rate imaging improved the diagnostic accuracy compared to conventional visual wall-motion reading (sensitivity 81% [visual] vs 86% [strain rate], specificity 82% [visual] vs 90% [strain rate]).
In the other study assessed by analysis of tissue Doppler displacement, a PSS parameter at peak dobutamine stress predicted the presence of coronary artery disease with sensitivity of 89% and specificity of 77%.19 The sensitivity was 93% for single vessel disease and 78% for multivessel disease.
In contrast, there are studies showing PSS not to be superior to systolic stain in assessing ischaemia. For example, Bjork Ingul et al20 reported that the post-systolic index was inferior to end-systolic strain in the receiver operating characteristic curve analysis of dobutamine stress echocardiography for diagnosing coronary artery disease. Some potential reasons are considered.
Temporal resolution of the strain analysis affects the diagnostic accuracy of PSS. When echocardiographic images are acquired at a low frame rate, PSS often fails to be detected because of its relatively short duration. The frame rate of speckle tracking echocardiography is generally lower than that of tissue Doppler echocardiography. In the results of Bjork Ingul et al,20 the area under the curve of the post-systolic index derived from the speckle tracking method was smaller than that derived from the tissue Doppler method. Therefore, a high frame rate should be selected to enable accurate detection of PSS in the speckle tracking method.
Because the heterogeneity of contraction among myocardial regions induces PSS, as shown by Claus et al,10 it can be observed not only in ischaemic heart disease but also in non-ischaemic heart diseases such as hypertrophic and dilated cardiomyopathies, hypertensive heart disease, pulmonary hypertension, left bundle branch block, and even in healthy subjects. PSS has been reported to be observed in approximately one-third of myocardial segments in healthy subjects.21 The presence of physiologic PSS can complicate the accurate identification of pathologic PSS. This non-specificity hampers its widespread use in the echocardiographic laboratory. Thus, it is indispensable to distinguish between pathologic and physiologic PSS. In general, pathologic PSS has a larger amplitude and longer duration than that in healthy subjects.21 ,22 Criteria for pathologic PSS, proposed by Voigt et al,21 are shown in box 1. The relative decrease of myocardial strain during ejection time seen in concurrence with PSS is important for the differentiation in ongoing ischaemia. However, the cut-off value of the post-systolic index in each myocardial segment is still unclear, especially in speckle tracking echocardiography, and this needs to be determined for use in clinical settings.
Criteria for pathologic post-systolic shortening (PSS)
-
Transient PSS (occurrence during and resolution after ischaemia)
-
Clearly reduced systolic function or systolic bulging (εET >−7%)
-
Moderately reduced function (−7% >εET >−18%) and
-
PSS exceeding 20% of εtotal or
-
Shortening continues from systole or PSS peak occurs >90 ms after AVC
-
AVC, aortic valve closure; εET, ejection time strain; εtotal, total strain during heart cycle.21
Because the amplitude and duration of PSS become greater during ischaemia, even in the segment with physiologic PSS, comparison of the strain profile during the ischaemic episode and that before the episode is desirable for accurate diagnosis. Therefore, analysis of PSS is considered advantageous in stress echocardiography. In the analysis of myocardial strain, subtle change of the strain amplitude during ejection time is difficult to detect because its normal variation is relatively large; however, that of timing in peak strain is relatively easy to detect.
It may be difficult to detect ischaemia in the myocardium damaged previously using PSS. The simulation and experimental models of Claus et al10 demonstrated that PSS did not occur in chronic infarction. This result suggests that increased myocardial stiffness can diminish PSS caused by passive recoil.
The question remains about which direction of strain (ie, longitudinal, transverse, circumferential, and radial) should be analysed for the assessment of PSS in speckle tracking echocardiography. The answer may partly depend on vendor differences in the speckle tracking algorithm.23 Longitudinal strains in apical views are usually analysed in clinical settings. However, it is known empirically that the acquisition of reliable longitudinal strain profiles is often difficult in apical segments. This may hamper the diagnostic accuracy of ischaemia in which the left anterior descending coronary artery is a culprit vessel.
Not only PSS but also ESL occurs in the ischaemic myocardium. It has recently been reported that assessment of ESL is useful to detect myocardial ischaemia.24 The diagnostic accuracies for detecting ischaemia of PSS and ESL need to be compared in further studies.
Detection of ischaemic memory by PSS
Acute chest pain often resolves before a patient can present to the hospital, which makes a proper diagnosis difficult. In such patients, imaging technology that allows detection of ischaemic memory would be of tremendous utility. It has been reported that the suppression of fatty acid metabolism due to myocardial ischaemia persists after relief from ischaemia.25 ,26 Such metabolic imaging is promising for the assessment of ischaemic memory, but must be performed in a radiation-controlled area. In contrast, ischaemic memory imaging by echocardiography would have greater utility in the clinical setting and could even facilitate bedside examination.
When transient severe ischaemia occurs, myocardial dysfunction can persist despite normalisation of blood flow; this phenomenon is known as myocardial stunning. In myocardial stunning, the functional, biochemical, and microstructural abnormalities that occur following ischaemia are reversible, and contractile force is gradually restored. Systolic dysfunction in myocardial stunning can be used as a measure of ischaemic memory, but this phenomenon has not been considered to be induced by relatively mild ischaemia. However, a subtle abnormality such as PSS may persist after brief ischaemia, even when a contractile abnormality is not observed.
To verify this hypothesis, we performed tissue Doppler echocardiography in a dog model with 15 min or 5 min coronary occlusion followed by reperfusion and evaluated the chronological changes of peak systolic strain and the post-systolic index.3 In our results, peak systolic strain decreased significantly in the risk area during occlusion. This decrease in the peak systolic strain in the 15 min group did not recover completely to the baseline level even 120 min after reperfusion (ie, conventional myocardial stunning), whereas the decrease in the 5 min group recovered immediately after reperfusion. The post-systolic index increased significantly during occlusion, but the increased post-systolic index in the 5 min group remained until 30 min after reperfusion, despite the rapid recovery of peak systolic strain (figure 6). We also tested another model, in which ischaemia was induced by dobutamine stress during non-flow limiting stenosis (ie, demand ischaemia), and found that the increase in the post-systolic index persisted even 20 min after dobutamine stress. These results suggest that PSS remains longer than the decrease in peak systolic strain after recovery from ischaemia, and the assessment of PSS can be used to detect ischaemic memory. Moreover, we analysed strain and strain rate parameters derived from speckle tracking echocardiography in the 2 min occlusion/reperfusion model and evaluated which regional myocardial deformation parameters could demonstrate ischaemic memory.4 We found that only PSS-related parameters persisted after recovery from 2 min occlusion. Although the strain rate during early diastole, which is a parameter of regional diastolic function, was expected to be another parameter of ischaemic memory, it was not found to be so (figure 7).
Recently, myocardial layer-specific analysis became feasible. In the 2 min occlusion/reperfusion model, analysis of the circumferential endocardial layer was better than that of the epicardial layer to detect ischaemic memory.27
In clinical settings, Ishii et al28 investigated whether delayed relaxation could be detected after treadmill exercise in 117 patients with stable angina caused by significant coronary stenosis. They analysed the rate of change in the strain during the first one-third of diastole (strain imaging diastolic index (SI-DI)) using speckle tracking echocardiography. SI-DI decreased significantly in the risk area even at 10 min after the exercise, and the decrease of SI-DI was observed in 85% of the segments within the risk area. They also evaluated the chronological changes of SI-DI in 35 patients with stable angina in whom elective percutaneous coronary intervention was performed.29 Peak systolic strain decreased significantly during coronary occlusion and recovered to near-normal pre-occlusion values after reperfusion. In contrast, SI-DI decreased significantly during occlusion, but the significant decrease remained even at 24 h after reperfusion. In both studies, the decrease of SI-DI seemed to be caused mainly by PSS, as shown in their figures (figure 8). In the other study, PSS displayed by velocity vector imaging could be detected 20 min after exercise in patients with exercise-induced ischaemia.30
In the assessment of ischaemic memory using PSS, some issues need to be addressed for clinical use. The time period during which PSS can be detected after brief ischaemia is still unclear. Because the persistency of PSS depends on the duration and severity of ischaemia,3 remaining PSS could be detected for several hours after severe supply ischaemia. However, in cases of demand ischaemia induced by an exercise stress, it seems to persist only for <30 min.28 ,30 This duration may be considered too short as ischaemic memory, but it would still be valuable in stress echocardiography. A rapid heart rate during peak stress makes the detection of PSS difficult under the limited frame rate of speckle tracking echocardiography. Moreover, echocardiographic images are poor during peak stress, especially in exercise. The later assessment would facilitate the acquisition of robust strain profiles.
Differentiation between pathologic and physiologic PSS is still an important issue here. The cut-off value of the post-systolic index for ischaemic memory is unclear. We recommend that echocardiography for diagnosing ischaemic memory should be performed twice with 10 or 15 min intervals, because PSS as a sign of ischaemic memory would decrease over time, whereas normal PSS would remain at the same level.
The reason why PSS persists after brief ischaemia is unknown. The persistency of PSS may reflect subtle systolic stunning that cannot be detected by conventional systolic parameters, and a reversible metabolic disorder can be one possible mechanism for this phenomenon. PSS is considered a motion that reflects regional contractile dysfunction, but it is still controversial whether a regional relaxation disorder is related to the occurrence of PSS. Further investigation is needed to address these issues.
Conclusions
The analysis of myocardial velocity and strain by tissue Doppler and speckle tracking echocardiography has allowed the detection of subtle myocardial deformation, such as PSS, that is difficult to identify by conventional methods. Assessment of PSS is valuable in diagnosing acute ischaemia because it improves diagnostic accuracy, though there are some issues to be resolved. The detection of PSS that persists after brief ischaemia would be invaluable for ischaemic memory imaging.
References
Footnotes
-
Contributors TA, SN writing of the manuscript.
-
Competing interests None.
-
Provenance and peer review Commissioned; externally peer reviewed.