Author + information
- Received January 15, 1997
- Revision received June 5, 1997
- Accepted June 21, 1997
- Published online October 1, 1997.
- Juerg Schwitter, MDA,1,1,
- Maythem Saeed, PhD, DVMA,
- Michael F. Wendland, PhDA,
- Nikita Derugin, MAA,
- Emmanuelle Canet, DVMA,
- Robert C. Brasch, MDA and
- Charles B. Higgins, MDA,*
- ↵*Dr. Charles B. Higgins, Department of Radiology, University of California San Francisco, 505 Parnassus Avenue, Room L-308, San Francisco, California 94143-0628.
Objectives. This study sought to 1) compare the distribution of extravascular (573 Da) and intravascular (92 kDa) magnetic resonance (MR) contrast agents in reperfused infarcted myocardium, and 2) investigate the effect of injury severity on these distribution patterns.
Background. Myocardial distribution of low and high molecular weight contrast agents depends on vascular permeability, diffusive/convective transport within the interstitium and accessibility of the intracellular compartment (cellular integrity).
Methods. To vary the severity of myocardial injury, 72 rats were subjected to 20, 30, 45 or 75 min (n = 18, respectively) of coronary artery occlusion. After 2 h of reflow, the animals received either 0.05 mmol/kg of gadolinium-diethylenetriaminepentaacetic acid-bismethylamide (Gd-DTPA-BMA) (n = 24), (Gd-DTPA)30-albumin (n = 24) or saline (control group, n = 24). Three minutes after injection, the hearts were excised and imaged (spin-echo imaging parameters: repetition time 300 ms, echo time 8 ms, 2-tesla system), followed by triphenyltetrazolium chloride staining for infarct detection and sizing.
Results. Histomorphometric and MR infarct size (expressed as percent of slice surface) correlated well: r = 0.96 for Gd-DTPA-BMA; r = 0.95 for (Gd-DTPA)30-albumin. On Gd-DTPA-BMA–enhanced images, reperfused myocardial infarctions were homogeneously enhanced. The ratio of signal intensity of infarcted/normal myocardium increased with increasing duration of ischemia (overall p < 0.0001, analysis of variance [ANOVA]), indicating an increase in the distribution volume of Gd-DTPA-BMA in postischemic myocardium. On (Gd-DTPA)30-albumin–enhanced images, reperfused infarctions consisted of a bright border zone and a less enhanced central core. The extent of the core increased with increasing duration of ischemia (overall p value < 0.0001, ANOVA).
Conclusions. At 2 h of reperfusion, the distribution of MR contrast agents in postischemic myocardium is 1) specific for extravascular and intravascular agents, and 2) modulated by the duration of ischemia.
Revascularization procedures to restore impaired myocardial blood supply in the setting of acute myocardial infarction find their rationale in the assumption that at least some viable myocardium is present in a flow-compromised territory. In this context, intense work is still directed toward the development of diagnostic techniques to assess myocardial viability.
On the basis of pathophysiologic studies, both myocardial cell and microvascular damage is invoked as a possible marker of myocardial infarction [1–5]. A number of studies [6–18]have assessed the capability of contrast-enhanced magnetic resonance imaging (MRI) for the characterization of reperfused myocardial infarctions. Unlike intravascular contrast agents, extravascular contrast agents distribute rapidly into the extravascular space . Therefore, their distribution volume may depend on 1) the size of the interstitial compartment, and 2) myocardial cell membrane integrity because myocytes with loss of membrane integrity fail to exclude the extravascular contrast agent from the intracellular compartment. In this regard, it is possible that myocardial cell injury might be evaluated by low molecular weight extravascular contrast agents [6, 20]. In contrast, microvascular damage may be probed by high molecular weight intravascular MR contrast agents. Various enhancement patterns of reperfused infarcted myocardium have been reported after the administration of extravascular [6, 7, 9, 11–13, 15, 21]and intravascular gadolinium (Gd) chelates [10, 16, 17, 22, 23]. To our knowledge, the use of both types of contrast agents to more specifically characterize ischemically damaged myocardium has not been attempted. Accordingly, the current study was intended to 1) directly compare the enhancement patterns induced by low molecular weight extravascular and high molecular weight intravascular MR contrast agents in reperfused injured myocardium, and 2) investigate the influence of injury severity on these different enhancement patterns.
1.1 Animal preparation.
The study was approved by the Committee on Animal Research at our institution. Seventy-two Sprague-Dawley rats (Simonsen Labs, Inc.) weighing 250 to 320 g were anesthetized with an intraperitoneal injection of sodium pentobarbital (50 mg/kg body weight). After tracheotomy, the animals were mechanically ventilated (Harvard Apparatus Rodent Respirator). After left thoracotomy, the anterior branch of the left main coronary artery was occluded using a snare ligature (5-0 suture). The presence of occlusion was confirmed by development of regional myocardial cyanosis. The tail vein was cannulated (using a 27-gauge scalp vein catheter) for administration of contrast agent and supplemental anesthetic.
1.2 Experimental protocol.
The rats were subjected to 20, 30, 45 or 75 min of coronary artery occlusion (n = 18 for each occlusion period) [3, 8, 21, 24]. After completion of the occlusion period, reperfusion was initiated by loosening the snare. Myocardial reperfusion continued in all animals for 2 h before imaging. Rats were divided into three groups receiving 1.0 ml/kg of either 1) extravascular contrast agent (Gd-diethylenetriaminepentaacetic acid-bismethylamide [Gd-DTPA-BMA], Omniscan at a dose of 0.05 mmol/kg); 2) intravascular contrast agent [(Gd-DTPA)30-albumin at a dose of 0.025 mmol/kg]; or 3) physiologic saline. Three minutes later, all animals were killed by intravenous injection of 0.3 ml of saturated solution of potassium chloride. The hearts were excised and dabbed dry with gauze, and a cotton swab was inserted into the left ventricular cavity through a small incision in the left atrium to expand the left ventricular chamber. Each heart was wrapped in clear plastic to minimize dehydration during imaging.
1.3 Contrast agents.
The longitudinal (T1) and transverse (T2) relaxation times of the nonionic low molecular weight chelate complex Gd-DTPA-BMA (molecular mass 573 Da) in water at 10 MHz (corresponding to a magnetic field strength of 0.25 tesla) at 37°C are 4.6 and 5.1 mmol/liter per s, respectively . (Gd-DTPA)30-albumin (molecular mass 92 kDa) was prepared according to Ogan et al. and contained an average of 30 Gd-DTPA chelates that are covalently conjugated to human serum albumin. For the (Gd-DTPA)30-albumin, the T1 and T2 relaxation times, as measured in water at 10 MHz at 37°C, were 15 and 22 mmol/liter per s per Gd3+ion, respectively .
1.4 MRI technique.
All images were acquired using a Bruker Omega 2.0-tesla system (Bruker Instruments, Inc.). Each heart was positioned with its long axis parallel to the main magnetic field such that axial images would represent short-axis views of the heart. The midventricular level was defined on a horizontal localizer as the midpoint of the long axis of the heart (between the base and apex of the heart). Two axial (short axis) slices, one below and one above the midventricular level, were acquired using T1-weighted spin-echo imaging (repetition time [TR] 300 ms, echo time [TE] 8 ms; slice thickness 2 mm; field of view 30 mm × 30 mm; data matrix 128 × 256 interpolated to 256 × 256 during Fourier transformation; four acquisitions; acquisition time 2.56 min). A previous report showed high correlation between infarction size as measured from all slices of the heart compared with measurements derived from only two midventricular slices in rats. Therefore, in the current study, two midventricular slices were acquired, and all results derived from these two slices are given as mean value of the two slices.
1.5 Histochemical staining.
After completion of the MRI session, the heart was removed from the magnet. The midventricular level was again defined (before removing the cotton swab) as the midpoint of the long axis (between the base and apex of the heart). The heart was then cut in half at the midventricular level, and from each half a 2-mm thick slice of the cut surface was obtained. The two slices were then soaked in 2% solution of triphenyltetrazolium chloride (TTC) at 37°C for 15 min. The TTC dye allows differentiation between viable myocardium (brick red) and nonviable myocardium (unstained). A prerequisite for accurate detection and sizing of myocardial infarction by TTC staining is sufficient time for washout of reduced coenzymes from the infarcted tissue . In pigs, where collateral circulation is as poor as that in rats, no significant differences in infarction size were found when measured with nitroblue tetrazolium after 2 and 24 h of reperfusion, respectively . Therefore, in the present study reperfusion was allowed for 2 h in all animals. The lower and upper surface of each slice was photographed and scanned into a MacIntosh computer. The sizes of red (viable) and unstained (infarcted) myocardium were measured by planimetry (NIH Image, version 1.59) and then expressed as a percentage of the entire slice surface area.
1.6 MR image analysis.
1.6.1 Measurement of infarction size.
All images were transferred to a SUN computer (SPARC 10, Sun Microsystems), and image analysis was performed using analysis software (MRVision Co.). On contrast-enhanced images, the extent of the enhanced region was determined by using a threshold/signal intensity of normal myocardium plus 3 SD. The signal intensity of normal myocardium was measured in a region of interest (ROI) that was placed over the septum of the left ventricle (with the right ventricle used as a landmark) and encompassed ∼25% of the slice surface. The pixels with intensities above the threshold were defined as representing the infarcted region on MRI. These pixels were summed and expressed as a percentage of the slice surface area.
1.6.2 Measurement of signal intensities in infarcted myocardium.
On the Gd-DTPA-BMA–enhanced images, the signal intensity of the infarcted myocardium was derived from the entire enhanced zone. On the (Gd-DTPA)30-albumin–enhanced images, the signal intensity was measured in two locations: the hyperintense border of the infarcted zone (80 to 120 pixels) and the core of the infarcted zone (50 to 100 pixels). In cases where no dark core was observed (in four cases with 20 min of occlusion time), an ROI (∼50 pixels) was arbitarily located in the center of the bright region. On unenhanced images (no contrast agent administered), the infarcted region was not identifiable as a hyperintense zone. To obtain a signal intensity value of infarcted myocardium on unenhanced images, an ROI (∼100 pixels) was placed in the center of the infarcted (unstained) myocardium according to TTC staining. In all rats, myocardial signal intensity values were normalized to that of an oil phantom, which was included in each imaging session.
1.6.3 Measurement of extent of central core.
The size of the less enhanced central core of the infarction on the (Gd-DTPA)30-albumin–enhanced images was measured using the threshold/signal intensity ratio of the hyperintense border zone minus 3 SD. All pixels of the infarcted region below this threshold were counted and expressed as a percentage of the infarct size. On images enhanced with Gd-DTPA-BMA, no less enhanced core was observed. However, for comparison, the Gd-DTPA-BMA–enhanced images were treated the same as the (Gd-DTPA)30-albumin–enhanced images by measuring signal intensity in the border of the enhanced zone to derive the threshold that allowed calculation of hypointense pixels in the infarcted zone.
1.7 Statistical analysis.
Results are expressed as mean value ± SEM. Between-group comparisons were performed using analysis of variance (ANOVA). If the analysis showed an overall p < 0.05 value, the Scheffé test was implemented. For within-heart comparisons (e.g., normal vs. infarcted myocardial regions), ANOVA for repeated measures was used. Linear regression analysis was used to compare infarction size defined by histomorphometric analysis with that by MRI. A value of p < 0.05 was considered significant.
2.1 Myocardial infarction size.
Infarction size as defined by TTC staining did not differ among the three experimental groups—Gd-DTPA-BMA, (Gd-DTPA)30-albumin and control rats—at any given duration of coronary occlusion (two-way ANOVA). However, infarction size increased significantly with increasing duration of ischemia from 20 to 45 min (Fig. 1). Longer ischemia (>45 min) did not produce a larger infarction but was associated with macroscopic intramyocardial hemorrhage preferentially located in the center of the infarction. Hemorrhagic areas were undetectable on unenhanced MR images (no significant differences in signal intensity between normal myocardium and the center of infarcted myocardium) (Table 1), indicating that potential reduction of signal intensity by deoxyhemoglobin was not important for the short TE setting used.
On Gd-DTPA-BMA–enhanced images, infarcted myocardium was observed as a homogeneous hyperintense zone (Fig. 2). On (Gd-DTPA)30-albumin–enhanced images, infarcted myocardium was observed as an inhomogeneously enhanced zone, consisting of a hyperintense border zone that surrounded a less enhanced core of variable extent depending on duration of ischemia (Fig. 2). For these two groups of rats, the regression lines comparing infarct size on MR images and on TTC stains were similar: slope 1.06 and 1.14, intercept 10.3% and 9.7% for Gd-DTPA-BMA–enhanced and (Gd-DTPA)30-albumin–enhanced images, respectively (Fig. 3). Infarction size was overestimated on contrast-enhanced MR images compared with histomorphometric analysis.
2.2 Severity of myocardial injury: Gd-DTPA-BMA–enhanced images.
The signal intensity of unenhanced normal myocardium (control group) ranged from 0.27 to 0.28 for the different occlusion times and was significantly less than that of Gd-DTPA-BMA–enhanced images, ranging from 0.41 to 0.42 (p < 0.0001, unpaired ttest for each occlusion time) (Table 1). Reperfused infarcted myocardium was visualized as a hyperintense region compared with normal myocardium (Fig. 2). The signal intensity of the infarcted region increased with increasing duration of ischemia (overall p value < 0.0001, ANOVA), reaching a maximal level of 0.76 ± 0.01 at 45 min (Table 1). Similarly, the ratio of the signal intensity of infarcted/normal myocardium increased significantly (overall p value < 0.0001, ANOVA) from 20 to 45 min of ischemia, indicating an increase in the distribution volume of Gd-DTPA-BMA in the infarcted region with increasing duration of ischemia (Fig. 4). Quantitative analysis using a threshold of 3 SD below the signal intensity of the border zone yielded only few hypointense pixels encompassing 3.5% to 5.5% of the entire hyperintense region (Fig. 5). These hypointense pixels were homogeneously scattered throughout the infarcted region and did not form a hypointense core as in the (Gd-DTPA)30-albumin–enhanced images (Fig. 2). The percent of hypointense pixels did not change with increasing occlusion times (Fig. 5). The homogeneous enhancement of the reperfused infarcted region on Gd-DTPA-BMA–enhanced images provides evidence that the contrast agent was delivered to the entire reperfused infarcted region, and homogeneous distribution was rapidly established (within the first 3 min after injection) in the entire reperfused infarction.
2.3 Severity of myocardial injury: (Gd-DTPA)30-albumin–enhanced images.
On (Gd-DTPA)30-albumin–enhanced images, the infarcted region was essentially homogeneously enhanced in hearts subjected to 20 min of ischemia (Fig. 2). In hearts subjected to longer duration of ischemia, the reperfused infarcted region showed a hypoenhanced core (compared with the hyperintense border zone). The extent of the hypoenhanced core increased from 2.6 ± 1.5% of the infarcted region in rats (n = 6) subjected to 20 min of coronary artery occlusion to 34.5 ± 3.5% in those (n = 6) subjected to 75 min of occlusion (overall p value <0.0001, ANOVA) (Fig. 5). The signal intensity of the hyperintense border zone ranged from 0.58 to 0.63 and did not differ with increasing duration of ischemia (Table 1). The signal intensities of both the border zone and the less enhanced central core were higher than that of normal myocardium at all occlusion times (Table 1), providing evidence that (Gd-DTPA)30-albumin was delivered to the entire region of reperfused infarction.
The principal findings of the current studycan be summarized as follows: 1) The extent of reperfused myocardial infarctions can be quantified using low molecular weight (extravascular) and high molecular weight (intravascular) MR contrast agents; 2) extravascular Gd-DTPA-BMA distributes homogeneously within the reperfused infarcted myocardium, and its distribution volume increases with increasing duration of ischemia; 3) intravascular (Gd-DTPA)30-albumin distributes inhomogeneously within reperfused infarcted myocardium with impaired distribution in the central core of the infarction. The extent of the central core expands with increasing duration of ischemia.
3.1 Extent of reperfused myocardial infarction.
As shown in previous studies [8, 14]using low molecular weight Gd chelates, the size of myocardial infarction defined on contrast-enhanced MR images was closely correlated with infarction size defined by TTC histomorphometric analysis. The present data are in agreement with these earlier reports because Gd-DTPA-BMA–enhanced zones overestimated true infarction size by ∼10% of left ventricular slice area (6% to 12% in earlier studies). This overestimation may be caused by interstitial edema surrounding the infarcted territory. Other investigators [30, 31]found a slight underestimation of infarct size of 2.9% to 4.4% of left ventricular slice area when measured with TTC compared with histological analysis. It is thought that the microscopic heterogeneity at the infarct margins can cause underestimation of the TTC unstained area. Between the TTC-defined infarction size and (Gd-DTPA)30-albumin–enhanced zones, a similar correlation was found that also overestimated true infarction size by ∼10%. In the present study the size of reperfused myocardial infarctions increased up to an occlusion time of 45 min. This finding is in accordance with results reported by Hale et al. . In rats subjected to different durations of coronary artery occlusion and 24 h of reperfusion, small infarctions were observed after 20 min of ischemia, and infarction size did not increase further with occlusion times >40 min. In another study , electron microscopy revealed irreversible cell injury in rat myocardium having had as little as 10 min of ischemia.
3.2 Severity of myocardial injury.
In myocardial infarction, myocyte necrosis is associated with the loss of microvascular integrity [1, 5]. Because low molecular weight (extravascular) Gd-based MR contrast agents are excluded from intact cells, an increase in the distribution volume of these agents may reflect loss of myocardial cell membrane integrity [6, 20]. In contrast, leakage of the high molecular weight (intravascular) (Gd-DTPA)30-albumin may indicate microvascular damage [16, 22]. To study the relation between myocardial cell and microvascular damage in postischemic myocardium, the distribution patterns of low and high molecular weight MR contrast agents were analyzed in reperfused infarcted myocardium subjected to increasing durations of ischemia.
In the absence of contrast agents (control rats), reperfused myocardial infarction was undetectable, and the entire myocardium (normal and infarcted regions) was homogeneous with respect to signal intensity. This finding is explained by the small increase in T1 of postischemic myocardium in rats and the fact that the short TE is relatively insensitive to potential changes in T2 . Thus, differences in the enhancement pattern between Gd-DTPA-BMA– and (Gd-DTPA)30-albumin–enhanced images are likely to be caused by differences in the physiologic distribution of the different sized contrast agents. Furthermore, the three experimental groups did not differ with respect to the size of infarcted areas, as defined by histomorphometric analysis (Fig. 1).
3.3 Distribution of Gd-DTPA-BMA in reperfused infarcted myocardium.
On Gd-DTPA-BMA–enhanced images, the injury zone showed greater signal intensity than normal myocardium, and this differential increased with increasing duration of myocardial ischemia. Greater signal intensity in postischemic myocardium implies that more contrast agent is present than in normal myocardium. Earlier studies reported higher concentrations of extravascular Gd chelates in reperfused infarcted myocardium of rats and dogs . The primary cause of elevated Gd-DTPA-BMA content in the injured region is an enlarged tissue distribution volume. Two causes of enlarged distribution volume are interstitial edema [21, 33]and cellular necrosis [6, 20, 33]. Gd-DTPA-BMA is excluded from viable myocytes, but damage to the myocyte membrane during ischemia/reperfusion results in loss of cellular exclusion of contrast agent.
The increased distribution volume in the injured region can be estimated from quantitative evaluation of signal increase if the contrast agent is in equilibrium phase distribution. Previous work with this model suggest that this is true within minutes of injection of contrast agent. Then, the relative distribution volume of contrast agent in the tissue can be estimated from the relative change (Δ) in the longitudinal relaxation rate (R1) ( ΔR1 = 1/T1 before contrast administration− 1/T1 after contrast administration). In that study, fractional distribution volumes in normal and reperfused infarcted myocardium were 0.24 and 0.98, respectively. The distribution volume fraction of 0.98 for reperfused infarcted myocardium indicates that essentially the entire population of myocytes in the infarcted tissue failed to exclude the contrast agent. In dogs subjected to 2 h of ischemia (and variable duration of reperfusion), Pereira et al. found a high negative correlation between a viability marker (201thallium uptake) and myocardial regions with elevated signal intensity after Gd-DTPA.
In the present study, this concept of change in fractional distribution volume of a low molecular weight contrast agent reflecting myocardial cell damage was tested in hearts subjected to different durations of ischemia. Because increasing duration of ischemia is associated with progressive myocardial cell damage [3, 5, 24, 33, 34], it can be speculated that increase in signal intensity reflects an increase in the distribution volume of Gd-DTPA-BMA. From the signal intensity data of the present study (Table 1), ΔR1 values for different myocardial regions can be calculated using the equation: where SIpreand SIpost= precontrast and postcontrast myocardial signal intensity, respectively. These calculations are based on an assumed precontrast myocardial T1 of 0.9 , and dipolar T2 enhancement is neglected.
The ratio ΔR1infarcted myocardium/ΔR1normal myocardiumwould then provide an estimate of the relative contrast agent concentration in the infarcted myocardium. These calculations yield ΔR1 ratios of approximately 2, 4, 5 and 5 for reperfused myocardium subjected to 20, 30, 45 and 75 min of ischemia, respectively, consistent with an increase in fractional distribution volume of contrast agent with increasing severity of ischemia. Histologic and electron microscopic studies in rats showed that irreversibly and reversibly damaged myocytes may reside next to each other in the same infarct territory [2, 34]and that each of these cells may vary with respect to degree and extent of alteration of its subcellular components . These findings, and the fact that these changes depend on the duration of ischemia and reperfusion , may have implications as to the interpretation of changes in the fractional distribution volume of Gd-DTPA-BMA in infarcted reperfused myocardium. The lower distribution volumes for the 20- and 30-min occlusion groups compared with the 45- and 75-min occlusion groups indicate that cell membrane damage is less severe in the short-occlusion groups. Because this statement is true only at 2 h after reperfusion, it does not necessarily imply a better prognosis for the short-occlusion groups with respect to the fraction of intact cells that ultimately will survive the ischemic challenge. It cannot be ruled out that subcellular components in already irreversibly damaged myocytes would still exhibit membrane integrity after 2 h of reperfusion, thereby limiting the access of Gd-DTPA-BMA to the entire intracellular space. Serial MRI in vivo studies during reperfusion would allow definition of the time point when disintegration of subcellular compartmentalization is completed and may provide prognostic information.
Furthermore, reperfused infarcted myocardium exhibited a homogeneous enhancement after administration of Gd-DTPA-BMA, irrespective of the duration of ischemia. Fig. 5illustrates this finding more quantitatively, showing the very low percentage of hypointense pixels (3.5% to 5.5% of the infarcted region, consistent with Gaussian distribution of pixel values) that did not change with increasing occlusion times. This finding provides evidence that the fraction of damaged myocytes is distributed homogeneously (at the imaging resolution used) rather than clumped within the infarcted zone. This apparent homogeneity seen with MRI is in agreement with a histologic study that found severely damaged myocytes homogeneously scattered throughout reperfused rat myocardium subjected to various durations of anoxic perfusion/reoxygenation.
3.4 Distribution of (Gd-DTPA)30-albumin in reperfused infarcted myocardium.
On (Gd-DTPA)30-albumin–enhanced images, a homogeneous enhancement of infarcted myocardium was seen only with an ischemia time of 20 min. With a longer duration of ischemia, an inhomogeneous enhancement pattern in the reperfused infarcted myocardium developed, consisting of a bright border zone and a less enhanced central core that showed progressive expansion with increasing duration of ischemia (Figs. 2 and 5). This observation is in concert with reports from Lim et al. [16, 17]who used another high molecular weight Gd chelate (Gd-DTPA-polylysine) in cats. They found a homogeneous enhancement of small myocardial infarctions and an inhomogeneous enhancement pattern in larger infarctions consisting of a bright border and a dark core.
Enhancement of reperfused infarcted myocardium on (Gd-DTPA)30-albumin–enhanced images can be explained by the loss of vascular integrity as a consequence of ischemia and reperfusion. Functional studies of microvascular permeability indicated that ischemia of as short as 15 to 20 min caused substantial increase in protein extravasation in rats and dogs . Increased microvascular leakage would result in higher concentrations of (Gd-DTPA)30-albumin in the infarcted tissue. Once the (Gd-DTPA)30-albumin molecules reside in the edematous extravascular space, they gain access to a large water pool that further contributes to signal enhancement.
The finding of a homogeneous enhancement of reperfused infarcted myocardium on Gd-DTPA-BMA–enhanced images for all occlusion times is indicative that the contrast agent is delivered and distributed homogeneously within the entire myocardial infarction, regardless of the duration of ischemia. As with Gd-DTPA-BMA, it seemed reasonable to assume that (Gd-DTPA)30-albumin would also be delivered to the entire myocardial infarction. However, enhancement of the infarcted myocardium was inhomogeneous, with a less enhanced core in the center of the infarction (Fig. 2, Table 1). This pattern of contrast agent distribution in reperfused infarcted myocardium was confirmed histologically using biotin-labeled (Gd-DTPA)30-albumin in combination with immunohistochemical staining . Distribution of contrast agent in the tissue depends not only on its delivery (perfusion), but also on the traversal of the contrast agent across the capillary wall and the mixing of the contrast agent within the interstitium. Transport of molecules across the capillary membrane and thereafter within the interstitium is due to random thermal motion (diffusion) and the motion of bulk fluid (convection) [37–40]. Transport of large molecules, such as albumin, is predominantly driven by convective forces that are governed by pressure gradients. After a relative short period of ischemia (e.g., 20 min) sufficient to cause functional vascular damage [4, 36], (Gd-DTPA)30-albumin, mainly driven by the transcapillary pressure gradient, permeates the capillary wall . After severe, long-term ischemia (e.g., 45 and 75 min), severe interstitial edema and elevated tissue pressure may develop in the center of reperfused infarctions [4, 21, 33, 36, 39]. The (Gd-DTPA)30-albumin delivered to that zone would encounter a low transcapillary pressure gradient, and leakage driven by convective forces would be reduced. Because transport of low molecular weight compounds is mainly driven by diffusion, Gd-DTPA-BMA would be expected to move across the capillary wall and mix within the interstitium, irrespective of variations in tissue pressure [37, 38, 41].
In the present study, animals were killed 3 min after contrast agent administration. Studies in the same animal model demonstrated that this time point is far beyond first-pass conditions and close to equilibrium phase distribution . Therefore, the enhancement patterns of reperfused infarcted myocardium observed 3 min after contrast agent injection mainly reflect tissue distribution of contrast agent rather than its delivery (perfusion). Our data provide evidence that Gd-DTPA-BMA is homogeneously distributed in the reperfused infarcted myocardium (3 min after injection), and its distribution is unaffected by possible perfusion inhomogeneities (no-reflow phenomenon) [42, 43]. Consequently, perfusion inhomogeneities are unlikely to cause the inhomogeneous distribution of (Gd-DTPA)30-albumin that was observed in the reperfused infarcted myocardium.
3.5 First-pass versus equilibrium phase contrast agent distribution.
MRI in combination with low molecular weight contrast agents has been used in patients to assess the extent and perfusion of myocardial infarction after intravenous thrombolysis or percutaneous transluminal coronary angioplasty [11, 12, 15]. Lima et al. used a dynamic MRI approach in combination with a low molecular weight Gd chelate to monitor different enhancement patterns over time in patients after revascularization procedures. A homogeneous enhancement pattern was observed in successfully reperfused infarctions, whereas in cases of a (re)occluded infarct-related artery, hypointense areas within the enhanced region of myocardial infarction were detected. In this context, the fundamental differences between the two currently used MRI approaches for the assessment of myocardial infarction should be emphasized. With a dynamic approach, myocardial transit of contrast agent during first-pass conditions is measured to assess myocardial perfusion [15, 21, 44–46]. With the approach presented in the present report, fractional distribution volumes for an extravascular contrast agent are measured to assess cell membrane integrity [6, 20, 47]. During the first pass of contrast agent through infarcted hypoperfused myocardium, this territory would exhibit hypoenhancement due to delayed delivery of the contrast agent (hypoperfusion), whereas the same territory during equilibrium phase distribution would exhibit hyperenhancement due to an increased fractional distribution volume of the extravascular contrast agent (loss of cell membrane integrity). In addition to perfusion information, the results of the current study and others [6, 20]suggest that MRI in combination with a low molecular weight contrast agent offers the possibility of assessing contrast agent distribution volumes that may reflect cell membrane integrity. Intravascular MR contrast agents are now under development for clinical usage. The findings described in the present report may be helpful in interpreting clinical studies in the future.
Both extravascular and intravascular contrast agents allow identification and quantification of reperfused infarcted myocardium. However, the mechanisms that govern contrast agent distribution are different for the two contrast agents. As a consequence, contrast medium distribution depends on the underlying tissue damage (i.e., myocyte or microvascular injury). Extravascular Gd-DTPA-BMA provides a homogeneous enhancement of the injury zone; increase in signal intensity with increasing severity of injury may be related to progressive myocardial cell damage. Intravascular (Gd-DTPA)30-albumin provides inhomogeneous enhancement of the injury zone; its distribution within the injury zone may be impaired by elevated tissue pressure. Increasing size of the less enhanced core may reflect increasing severity of injury. Further studies are needed to investigate the underlying mechanisms responsible for the different distribution patterns described in the present report.
↵1 Dr. Schwitter was supported by a grant from the Swiss National Science Foundation, Zurich.
☆ All editorial decisions for this article, including selection of referees, were made by a Guest Editor. This policy applies to all articles with authors from the University of California San Francisco.
- analysis of variance
- gadolinium-diethylenetriaminepentaacetic acid-bismethylamide
- magnetic resonance
- magnetic resonance imaging
- region of interest
- longitudinal relaxation rates
- echo time
- repetition time
- triphenyltetrazolium chloride
- T1 and T2
- longitudinal and transverse relaxation times
- Received January 15, 1997.
- Revision received June 5, 1997.
- Accepted June 21, 1997.
- The American College of Cardiology
- Kloner RA,
- Rude RE,
- Carlson N,
- Maroko PR,
- DeBoer LW,
- Braunwald E
- Hale SL,
- Kloner RA
- Dauber IM,
- VanBenthuysen KM,
- McMurtry IF,
- et al.
- Wolfe CL,
- Moseley ME,
- Wikstrom MG,
- et al.
- Kim RJ,
- Chen EL,
- Lima JA,
- Judd RM
- Judd RM,
- Lugo OC,
- Arai M,
- et al.
- Lima JA,
- Judd RM,
- Bazille A,
- Schulman SP,
- Atalar E,
- Zerhouni EA
- Saeed M,
- Wendland MF,
- Yu KK,
- et al.
- Saeed M,
- van Dijke CF,
- Mann J,
- Wendland MF,
- Rosenau W,
- Brasch RC
- Simor T,
- Chu WJ,
- Johnson L,
- et al.
- Ogan MD,
- Schmiedl U,
- Moseley ME,
- Grodd W,
- Paajanen H,
- Brasch RC
- Oh BH,
- Ono S,
- Rockman HA,
- Ross JJ
- Horneffer PJ,
- Healy B,
- Gott VL,
- Gardner TJ
- Lotan CS,
- Bouchard A,
- Cranney GB,
- Bishop SP,
- Pohost GM
- Farb A,
- Kolodgie FD,
- Jenkins M,
- Virmani R
- Sunnergren KP,
- Rovetto MJ
- Granger HJ,
- Glen AL,
- Barnes GE,
- Lewis RE
- Bassingthwaighte JB,
- Wang CY,
- Chan IS
- Harris TR,
- L BK,
- Rowlett RD
- Gavin JB,
- Thomson RW,
- Humphrey SM,
- Herdson PB