INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ASSESSMENT OF LEFT VENTRICULAR (LV) intracavitary filling by Doppler echocardiography is one of several noninvasive approaches for studying LV diastolic function. In the normal LV, the initial intraventricular filling wave propagates rapidly toward the apex, and, as demonstrated by color M-mode Doppler, there is near simultaneous onset of filling velocities along the entire LV inflow tract (7). However, in patients with reduced LV ejection fraction or impaired diastolic function, the early diastolic apical filling may be markedly delayed, and this can be measured as slowing of mitral-to-apical flow propagation (1, 13). This has been attributed to a decrease in rate of LV relaxation, which causes a decrease in the driving pressure for mitral-to-apical flow (1, 12, 13), to loss of "apical suction" (12), and enhanced vortex formations due to changes in LV and mitral valve geometry (10).

There is, however, very limited insight into how regional wall motion interacts with intraventricular flow. A common finding in ischemic myocardium is postsystolic (postejection) shortening and consequently delayed onset of diastolic lengthening (8, 16, 17). Therefore, in the ischemic ventricle, there may be substantial regional asynchrony in the onset of myocardial diastolic lengthening, and this in turn may have an impact on intraventricular flow. The paucity of data regarding interactions between regional function and intraventricular filling reflects the limited ability of current imaging modalities to quantify regional myocardial function throughout the different phases of the cardiac cycle. We have recently demonstrated that strain Doppler echocardiography (SDE) represents a noninvasive method to quantify regional function (18). In this regard, SDE appears superior to conventional tissue Doppler imaging, which is strongly influenced by function in neighboring segments and translational effects. Therefore, SDE in combination with color flow imaging may represent a method to study interactions between intraventricular flow and regional wall motion.

The present study was designed to investigate how wall motion abnormalities in the ischemic myocardium interact with intracavitary filling, and the main objective was to determine if postsystolic shortening of ischemic apical segments may lead to a reversal of early diastolic apical flow. The study was done in a dog model, and sonomicrometry was used as a reference method for regional myocardial function.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Fourteen dogs of either sex and an average body weight of 21.2 kg were given thiopentone (25 mg/kg body wt) and morphine (100 mg) intravenously, followed by an infusion of morphine (50-100 mg/h iv) and pentobarbital (50 mg iv) every hour. The animals were artificially ventilated through a cuffed endotracheal tube using room air with 20-50% oxygen. A limb lead was monitored. After a median sternotomy, the pericardium was split from the apex to base, and the edges of the pericardial incision were loosely resutured. Inflatable vascular occluders were placed around the proximal one-third of the left anterior descending coronary artery (LAD). The dog was placed in the right supine position during recordings. The study was approved by the National Animal Experimentation Board.

Pressure Measurements

Sonomicrometry

Strain Doppler Echocardiography

Two-Dimensional Color Doppler Echocardiography

The low-velocity filter was set in the range of 1-4 cm/s with a frame rate varying from 38 to 90 (mean value 58 frames/s). The digital color velocity data were transferred to an external computer (Macintosh, Apple Computer) and analyzed with the use of a dedicated analysis program (Echopac, GE Vingmed Ultrasound).

Experimental Protocol

At the end of each recording, we induced ventricular extrasystoles and used long diastoles after premature contractions to adjust absolute pressure levels for the micromanometers. Data were recorded with the respirator off and digitized at 200 samples/s. In four dogs, contrast was injected rapidly into the LA before and during LAD occlusion, and 2-D echocardiographic images were recorded immediately after the injection.

Calculations

Pressure-derived variables. We calculated LV peak systolic pressure, LV end-diastolic pressure, and the time derivative of LV pressure (dP/dt). The time course of the fall in LV pressure from peak dP/dt to 5 mmHg above end-diastolic pressure was characterized by the time constant () of an assumed exponential decay to zero pressure (19). For definition of end diastole and end systole, we used the peak R wave of the ECG and peak LV dP/dt, respectively. We calculated the time from end diastole to the first diastolic crossover of LV and LA pressures (onset of transmitral filling). In four dogs, we also calculated the difference between LV apical and LV outflow tract pressures.

Sonomicrometry-derived variables. We calculated the following variables: peak systolic shortening [as the percent change in length (%L)], as

systolic lengthening (ischemic segments; %L), as

and postsystolic shortening (%L), as

During ischemia, we calculated the time from the onset of late systolic segmental shortening to the peak early diastolic LV apex to the outflow tract pressure gradient.

Doppler-derived variables. Doppler measurements of shortening strains were represented by negative values and lengthening strains by positive values. However, to simplify comparison with sonomicrometry, we report strain values in terms of percent shortening: peak systolic shortening and lengthening were calculated as a percentage of end-diastolic dimensions. Postsystolic shortening was calculated as segment shortening after end systole, as a percentage of end-diastolic dimension.

Statistical Analysis

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Hemodynamic variables before and during ischemia are presented in Table 1 and Fig. 1. Peak systolic shortening by sonomicrometry and SDE decreased from 12.5 ± 2.4% to 12.5 ± 2.2% (P < 0.01) and from 12.6 ± 1.7% to 9.5 ± 1.1% (P < 0.01), respectively, indicating dyskinesis.

View larger version (27K): [in this window] [in a new window]   Fig. 1.   A: baseline left ventricular (LV) and left atrial (LA) pressures, the time derivative of LV pressure (LV dP/dt), myocardial longitudinal segment length and strain by Doppler in the left anterior descending coronary artery (LAD) region, and ECG. The first vertical dotted line indicates the time of peak LV dP/dt, which was used as a marker of end systole, and the second vertical dotted line indicates the time of LV and LA pressure crossover, which was used as a marker of onset of transmitral filling. B: recordings obtained during LAD occlusion. Both sonomicrometry and strain by Doppler indicate systolic lengthening. Postsystolic shortening is demonstrated by both methods as shortening after the time of peak LV dP/dt (first vertical dotted line). Note that postsystolic shortening continues after the onset of transmitral filling (second vertical dotted line).

Figure 2 demonstrates marked changes in intraventricular filling after LAD occlusion. During baseline, the early diastolic filling wave by 2-D color Doppler propagated rapidly towards the apex, as indicated by the uniformly red color throughout the LV cavity. During LAD occlusion, however, all experiments showed reversal of early diastolic flow in the apex, i.e., the early diastolic velocities were directed from the apex toward the base. The reversed apical velocities are represented by the blue-colored area in Fig. 2, bottom left. To quantify this change in filling, we measured peak velocities in the distal one-third of the LV at the onset of transmitral filling, defined as the time of first diastolic crossover of LV and LA pressures. The peak velocity decreased from 9.2 ± 1.6 cm/s at baseline to 2.3 ± 0.4 cm/s (P < 0.01) during LAD occlusion. The 2-D color pattern suggested that the transmitral filling wave propagated only a short distance into the LV and then was redirected toward the LV outflow tract, where it appeared to rotate in a large vortex. These interpretations were confirmed by observations with contrast that was injected into the LA (Fig. 2). Before ischemia, the contrast propagated rapidly toward the apex and filled the entire LV cavity immediately after the onset of transmitral filling. During LAD occlusion, the contrast moved toward the midportion of the LV cavity and then rotated toward the LV outflow tract, whereas the apex remained free of contrast. Later in diastole, the contrast moved toward the apex as part of a macrovortex.

View larger version (38K): [in this window] [in a new window]   Fig. 2.   Top: LV early diastolic filling by two-dimensional (2-D) color Doppler (left) and contrast echocardiography (right) from an apical view. At baseline, the dominant flow wave propagated rapidly toward the apex, as demonstrated by 2-D color Doppler (red-encoded velocities represent flow toward the apex) and by contrast echocardiography. Bottom: LV early diastolic flow by 2-D color Doppler (left) and contrast echocardiography (right). During LAD occlusion, the 2-D color image recorded immediately after mitral valve opening demonstrates a dramatic change of the flow pattern compared with baseline. A counterflow, coded blue, is directed toward the red-encoded transmitral inflow. The echo-contrast propagates toward the midcavity but does not reach the apex on this early diastolic frame.

The reversed apical flow during early LV filling was associated with postsystolic shortening of the ischemic apical segment (Fig. 1B). Postsystolic shortening by sonomicrometry and SDE changed from 1.3 ± 0.4% and 0.8 ± 0.4%, respectively, to 14.2 ± 2.4 (P < 0.01) and 9.7 ± 1.7% (P < 0.01) during LAD occlusion. Figure 1 also demonstrates the temporal relationships between myocardial segmental motion and transmitral driving pressure. During LAD occlusion, the ischemic segment was shortening at the time of LA and LV pressure crossover, and, as demonstrated in Fig. 3, the postsystolic shortening continued while the nonischemic segment was lengthening. The ischemic segment continued to shorten 47 ± 8 and 65 ± 16 ms after the onset of filling by sonomicrometry and SDE, respectively. The time constant of LV relaxation () increased from 47 ± 4 to 64 ± 7 ms during ischemia (P < 0.01).

View larger version (33K): [in this window] [in a new window]   Fig. 3.   LV and LA pressure tracings, LV dP/dt, segment lengths in the LAD and left circumflex (Cx) regions, and ECG before and during LAD occlusion. Before ischemia, both segments shorten during systole and lengthen during early diastole. During LAD occlusion, the ischemic segment lengthens during systole and shortens during early diastole. Line A indicates end diastole; line B indicates the time of peak LV dP/dt, which was used as a marker of end systole; and line C indicates the first diastolic crossover of LV and LA pressures, which was used as a marker of the onset of transmitral filling.

Figure 4 displays the intraventricular pressure gradients as calculated between the apex and LV outflow tract. During LV ejection, the pressure gradient was directed from the apex toward the LV outflow tract, and there was a small reversal of the pressure gradient in late systole. The systolic pressure gradient decreased after LAD occlusion. During most of the isovolumic relaxation period and during early transmitral filling, the pressure gradient at baseline was directed toward the apex (peak value 1.1 ± 0.3 mmHg). During LAD occlusion, however, the gradient was reversed during most of the isovolumic relaxation period (peak value 1.2 ± 0.3 mmHg), consistent with the observed reversal of early diastolic apical flow velocities. The time from onset of late systolic shortening of the ischemic segment by sonomicrometry to the peak early diastolic apex-to-outflow tract pressure gradient was 25 ± 3 ms.

View larger version (29K): [in this window] [in a new window]   Fig. 4.   A: Recordings at baseline showing LV pressures measured at the apex (PLVapex) and outflow tract (PLVoutflow tract), pressure difference between the apex and LV outflow tract, segment length in the LAD region, and ECG. During most of systole, the pressure gradient was positive (solid area), which means it directed from the apex toward the outflow tract. During the isovolumetric relaxation (IVR) period and the initial phase of transmitral filling, the pressure gradient was negative (hatched area) and was directed toward the apex. B: recordings obtained during LAD occlusion. The ischemic segment lengthens during systole and shortens during the IVR period and during early transmitral filling. The intraventricular pressure gradient during the IVR period has reversed and is directed from the apex toward the LV outflow tract (solid area). Line A indicates end diastole; line B indicates peak LV dP/dt, which was used as a marker of end systole; and line C indicates the first diastolic crossover of LV and LA pressures, which was used as a marker of onset of transmitral filling.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Slowing of mitral-to-apical flow propagation has been observed in patients with acute myocardial ischemia in different clinical settings (11, 13) and has been reproduced in experimental models (1, 12, 13). The etiology of the impaired apical filling is unclear, but several mechanisms have been proposed (10, 12-14). In the present study, we demonstrated that postsystolic shortening of the ischemic myocardium has a profound effect on intraventricular flow. The postsystolic shortening resulted in early diastolic ejection of blood from the apex toward the base, which resulted in markedly disturbed flow at the base of the ventricle, where flow from the apex encountered the transmitral filling wave. The postsystolic shortening was demonstrated by SDE and was confirmed by sonomicrometry. The present study therefore demonstrates the ability of Doppler echocardiography to quantify interactions between intraventricular flow and regional LV wall motion.

Stugaard et al. (13) reported delayed apical filling by color M-mode Doppler during LAD occlusion. In their study, no early diastolic apical flow was reported. This may be due to their velocity filter setting (12 cm/s), which was above the range of reversed apical velocities measured in the present study. The reversed velocities are very low, so the filter setting is critical to capture these velocities. Edvardsen et al. (2) reported reversed flow during the isovolumetric relaxation period in patients with anterior wall infarction. The velocity filter in that study was between 4 and 8 cm/s.

Mechanisms of Delayed Early Diastolic Mitral-to-Apical Flow Propagation

As demonstrated by sonomicrometry, the ejection of blood from the ischemic apical region started during isovolumic relaxation and persisted for some time after the onset of transmitral filling. Therefore, lengthening of the nonischemic segment (Fig. 3) during early filling may be ascribed not only to transmitral flow but also to redistribution of blood caused by postsystolic shortening. Blood that was ejected from the ischemic segment during early diastole encountered the transmitral jet, which was hindered from propagating toward the apex. As suggested by 2-D color flow and echocontrast, the transmitral flow appeared to propagate only a short distance in the ventricle and then formed macrovortexes that rotated toward the LV outflow tract.

The observed reversal of the intraventricular pressure gradient during ischemia supported the interpretation that postsystolic shortening caused early diastolic ejection of blood from the apical region. At baseline, the pressure gradient during isovolumic relaxation and early filling was directed from the base toward the apex, and intraventricular flow velocities by color flow imaging were directed toward the apex. During LAD occlusion, the pressure gradient during isovolumetric relaxation reversed and was directed toward the base, along with reversal of flow velocities. During ischemia, there was a small time delay of ~25 ms between the onset of late systolic segmental shortening and the peak apex-to-outflow tract pressure gradient, which is consistent with the notion that the gradient is caused by the segmental shortening.

Yellin et al. (20) described intraventricular vortexes that occurred during LV filling due to shear between blood and mitral leaflet surface. The flow reversal described in the present study, however, started during isovolumic relaxation and therefore cannot be attributed to a similar mechanism.

Apical Flow Reversal: Diagnostic Implications

The present study suggests that color flow imaging may be helpful as an additional diagnostic tool for identifying regions with suspected postsystolic shortening. The finding of early diastolic velocities that are directed from the apex toward the LV base is consistent with postsystolic shortening. Therefore, the two methods are complementary, and their combined use could increase the diagnostic power of echocardiography in the assessment of regional myocardial function. The clinical potential of this approach remains to be investigated.

Limitations

One significant limitation of SDE is a strong angle dependency (18). In the present study with open-chest animals, we could easily optimize transducer position and therefore minimized this problem. In a clinical setting, however, the problem of angle dependency of the echobeam relative to myocardial fiber orientation may be significant, in particular when studying apical segments.

The present study was limited to short-term ischemia. Potentially, with ischemia of longer duration, resulting in tissue necrosis and stiffening of the myocardium, one might see less passive systolic distension and subsequently less marked postsystolic shortening (5). This in turn may reduce the early diastolic flow reversal.

In conclusion, in the present experimental study, we demonstrated that delayed early diastolic LV apical filling during LAD occlusion may be due to postsystolic shortening of ischemic segments. During isovolumic relaxation and during the early phase of transmitral filling, blood was ejected from the apical region and redistributed toward the basal portions of the cavity. This implies that, in addition to global factors such as the rate of LV relaxation and ventricular geometry, one should consider asynchronous wall motion, and in particular postsystolic shortening, as a mechanism of disturbed intraventricular flow. Potentially, the use of color flow imaging along with SDE may represent a clinical tool for studying interactions between regional function and intraventricular flow. This combined imaging might increase the diagnostic power of echocardiography in the assessment of myocardial function.