© 1999 American Heart Association, Inc.
Original Contribution |
Physiological Basis for Potassium (39K) Magnetic Resonance Imaging of the Heart
From the Feinberg Cardiovascular Research Institute and the Departments of Biomedical Engineering and Medicine, Northwestern University Medical School, Chicago, Ill.
Correspondence to Robert M. Judd, Feinberg Cardiovascular Research Institute, Northwestern University Medical School, 303 East Chicago Ave, Tarry 12-723 Chicago, IL 60611-3008. E-mail rjudd{at}nwu.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Abstract—The potassium cation (K+) is fundamentally involved in myocyte metabolism. To explore the potential utility of direct MRI of the most abundant natural isotope of potassium, 39K, we compared 39K magnetic resonance (MR) image intensity with regional myocardial K+ concentrations after irreversible injury. Rabbits were subjected either to 40 minutes of in situ coronary artery occlusion and 1 hour of reperfusion (n=26) or to 24 hours of permanent occlusion (n=4). The hearts were then isolated and imaged by 39K MRI (n=10), or tissue samples were analyzed for regional 39K content by MR spectroscopy (n=9), K+ and Na+ concentrations by atomic emission spectroscopy (inductively coupled plasma atomic emission spectroscopy; n=5), or intracellular K+ content by electron probe x-ray microanalysis (n=6). Three-dimensional 39K MR images of the isolated hearts were acquired in 44 minutes with 3x3x3–mm resolution. 39K MR image intensity was reduced in infarcted regions (51.7±4.8% of remote; P<0.001). The circumferential extent and location of regions of reduced 39K image intensity were correlated with those of infarcted regions defined histologically (r=0.97 and r=0.98, respectively). Compared with remote regions, tissue analysis revealed that infarcted regions had reduced 39K concentration (by MR spectroscopy, 40.5±9.3% of remote; P<0.001), reduced potassium-to-sodium ratio (by inductively coupled plasma atomic emission spectroscopy, 20.7±2.1% of remote; P<0.01), and reduced intracellular potassium (by electron probe x-ray microanalysis, K+ peak-to-background ratio 0.95±0.32 versus 2.86±1.10, respectively; P<0.01). We acquired the first 39K MR images of hearts subjected to infarction. In the pathophysiologies examined, potassium (39K) MR image intensity primarily reflects regional intracellular K+ concentrations.
Key Words: K+ • MRI • infarction • viability • myocyte
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Previous studies have demonstrated the fundamental role of the potassium cation (K+) to myocardial physiology. During ischemia, potassium efflux from myocytes1 has a time course similar to that of the loss of contractile function,1 2 3 4 and the resulting increase in extracellular K+ concentration may underlie ventricular arrhythmia.5 Irreversible injury to myocytes is characterized by a profound decrease in tissue K+,6 7 8 and the magnitude of intracellular K+ loss after irreversible injury is closely related to the increase in serum creatine kinase enzyme.9 10 Therefore, it has been established that cellular level K+ distributions closely relate to myocardial metabolism.
Because of the fundamental role of K+ in myocardial physiology, radiopharmaceuticals of potassium (38K, 42K, and 43K) and its analogues (81Rb, 82Rb, 86Rb, and 201Tl) have been developed11 12 and used to obtain diagnostic images. Indeed, 201Tl single-photon-emission CT is currently in widespread clinical use.13 Systematic study of the myocardial kinetics of these radiopharmaceuticals has revealed that image intensities are strongly influenced by both active K+ transport14 and regional delivery of the tracer by myocardial perfusion.15 16 Because active transport and perfusion both measurably contribute to image intensities, a number of injection and imaging protocols have developed depending on whether the clinical question is regional perfusion or viability.17 18 19
In principle, image intensity in a 39K magnetic
resonance (MR) image would be fundamentally different from that
obtained using radiopharmaceuticals. 39K is a
naturally occurring, nonradioactive isotope that constitutes 93% of
whole-body potassium20 and can be directly detected by
nuclear MR (NMR).21 Because no exogenous tracer is
injected, 39K MR image intensity would not be
dependent on tracer delivery by perfusion but rather would reflect
"static" potassium distributions. In practice, potassium MRI is
difficult, because the 39K MR signal is very
small, and, to date, no 39K MR images of the
heart have been reported. For example, compared with the proton
(1H) signal routinely used for MRI, the
39K signal is 
2 million times lower given that
in vivo concentrations of 39K are 1000 times
lower and the NMR sensitivity is 2000 times lower.22 On
the other hand, recent studies suggest that the application of fast
imaging techniques originally developed for 1H
MRI are well suited to imaging of nuclei such as
23Na and 39K because of the
much shorter NMR relaxation times of these nuclei.22 When
the advantages of fast imaging techniques are considered, we previously
reported that the quality of in vivo 23Na MR
images is significantly improved.23 More recently, we
found that 3-dimensional 23Na images of the heart
can be acquired in a few minutes in rabbits and dogs24 and
in humans at 1.5 T.22 On the basis of the
23Na results and theoretical
considerations,22 it is conceivable that some combination
of larger voxel sizes, longer imaging times, specialized radio
frequency (RF) receiver coils,25 26 and higher magnetic
field strengths27 may also make 39K
MRI of the heart feasible.
A definitive statement regarding the achievable quality of 39K MR images would require extensive study and, even then, the answer would likely evolve as MRI technology advances. The ultimate utility of 39K MRI, however, will be determined by the relationship of 39K MR image intensity to the underlying physiology, which will not be influenced by technological developments. To examine this relationship, we explored 39K MRI of isolated, nonbeating hearts previously subjected to infarction. We then compared image intensities to total (tissue-level) K+ concentrations, intracellular K+ levels, and histological changes. The goal of this study was to test the hypothesis that 39K MR image intensities reflect electrolyte imbalances secondary to cellular injury.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Thirty male New Zealand White rabbits (1.9 to 3.1 kg) were studied. All animals were cared for in accordance with the Position of the American Heart Association on Research Animal Use, adopted November 15, 1984. For each animal, the heart was studied by 39K MRI and 2,3,5-triphenyltretrazolium chloride (TTC) staining (n=10), 39K MR spectroscopy (MRS; n=9), inductively coupled plasma atomic emission spectroscopy (ICPAES; n=5), or electron probe x-ray microanalysis (EPXMA) and hematoxylin and eosin (H&E) staining (n=6). To explore 39K MRI in a large-animal model, a similar experimental protocol was performed on a 24-kg mongrel dog.
Animal Preparation and Isolated Heart Protocol
All animals were anesthetized, intubated, and ventilated
using a mechanical respirator. In 26 rabbits, reperfused infarction was
produced by performing a thoracotomy at the fourth left intercostal
space, opening the pericardium, and occluding an anterior branch of the
left coronary artery for 40 minutes using a reversible snare
ligature. The ligature was then removed to allow reperfusion for 60
additional minutes.23 The hearts were then rapidly excised
from the chest, placed momentarily in a 4°C Krebs-Henseleit solution,
and perfused in a retrograde manner. Perfusate composition
(in mmol/L) was as follows: Na+ 104,
K+ 15, Mg2+ 1,
Cl– 130,
HCO3– 25, and
Ca2+ 5. Adenosine and glucose were also
added to the perfusate in 1 and 11 mmol/L concentrations,
respectively. A 95% O2/5%
CO2 gas mixture was bubbled through the
perfusate to maintain a slightly alkaline pH.28 A
hyperkalemic cardioplegic solution was used to eliminate cardiac
motion. We have demonstrated in previous work that hearts perfused in
this manner remain viable.23 29 30
In 4 additional rabbits, nonreperfused infarction was produced by opening the chest under sterile conditions and permanently occluding the coronary artery. The chest was then closed, and the animals were allowed to recover for 24 hours before study by 39K MRI. In addition, to explore 39K MRI in a larger animal, a dog heart was also studied after occlusion of the left anterior descending coronary artery in situ for 90 minutes followed by 2 hours of reperfusion. In these 4 rabbits and 1 dog, the hearts were rapidly excised and cooled to 4°C but were not perfused with a hyperkalemic solution, to test whether the results would be similar in the absence of hyperkalemic arrest.
MR Image Acquisition and Analysis
Images were acquired with a spectrometer (4.7-T Omega CSI;
GE/Bruker) using a solenoid-type RF coil tuned to the
39K frequency (f=9.34 MHz). The
circuit used tune, match, and balance 1- to 24-pF variable
capacitors. Insulated copper wire was used to construct a
10-loop inductive coil that was 2.5 cm in diameter and 1.4 cm in
height. One 100-pF chip capacitor was added across the tune capacitor
to drive the resonant peak to 9.34 MHz. Using a spectrum
analyzer (HP 4195A, Hewlett Packard), the unloaded and
loaded 50-
quality factors, Q, of this coil were
determined to be 2300 and 1500, respectively. The RF coil used to image
the dog heart was larger (9 cm in diameter, 6 cm in height) but of
similar design.
Before MRI, gauze was inserted into the left ventricle to fill the
chamber. The exact position and orientation of each heart within the
coil was documented by making a small incision in the epicardium. The
isolated heart was centered in the magnet bore, and 3-dimensional
39K MR image data sets were acquired with a
gradient echo pulse sequence.31 The following
parameters were used for the rabbit hearts: voxel
size=3x3x3 mm (0-filled from 6 mm in phase and slice
directions), repetition time (TR)=40 ms, echo time (TE)=3 ms,
Navg=128, pulse width (pw)=200 µs, matrix
size=64x32x16, digital sampling interval=100 µs, and echo offset=20
data points. 39K MRI times were 44
minutes. For the dog heart, imaging parameters were
similar, except that voxel size was 5x5x5 mm and imaging time
was 60 minutes.
After 39K MRI data were collected, the hearts were sliced into 3-mm-thick short-axis sections and histologically stained with TTC to reveal viable and nonviable regions. TTC forms a red precipitate in the presence of intact dehydrogenase enzyme systems and therefore causes viable areas of myocardium to stain brick red, whereas nonviable areas do not stain and appear yellowish-white.32 For each heart, 4 slices of TTC-stained myocardium were photographed and the location of the incision marking the orientation of the heart was recorded. The photographs were scanned into a computer and analyzed using the software package NIH Image.
Circumferential extent and circumferential location of infarction were determined by manually tracing TTC negative areas. In the case of nontransmural infarction, a region was considered "infarcted" if >50% of wall thickness was TTC negative. Circumferential extent and circumferential location of reduced 39K MR image intensity were determined as follows. First, the mean and SD of 39K image intensity were measured in a remote (normal) region. Next, the 39K images were thresholded at a level defined as 2 SDs below the mean of the remote region. Myocardial regions with 39K image intensities below this threshold were defined as having reduced 39K image intensity. The circumferential extent of these regions was measured as the angle of the pie slice–shaped region of reduced 39K image intensity. The circumferential location was defined as the angular location of the center of the pie slice–shaped region where 0 degrees was defined to the right. Also measured were the mean 39K MR signal intensities in remote and reduced intensity regions and the SD of the background signal. Myocardial signal-to-noise ratios were determined using Henkelman's33 method for magnitude images.
MRS
Potassium concentrations of individual tissue samples were
determined with 39K MRS to examine the difference
in potassium content in remote and infarcted tissue samples using a
previously described technique.23 Tissue sections 0.75
g in weight were removed and weighed in labeled test tubes. Each sample
was then placed in the same RF coil used for imaging next to 1 mL of a
47.7 mmol/L KCl standard solution containing 43.5 mmol/L
dysprosium triethylene-tetraamine-hexaacetic acid, which caused the
precession frequency of the standard solution to shift such that 2
peaks appeared in the 39K MR spectrum, 1 from the
tissue sample and 1 from the test tube containing the known amount of
K+. Spectra were obtained using the following MR
parameters: Navg=4096 and 8192
(remote and infarcted samples, respectively), pulse width=140 µs,
digital sampling interval=200 µs, block size=1024 samples, and TR=250
ms. The free induction decays were Fourier transformed, phased, and
analyzed to determine areas under the sample and standard
peaks. From these data, the millimolar potassium concentration in each
tissue sample was calculated. The accuracy of this technique was
verified by measuring [K+] of 6 solutions with
known potassium concentrations. The correlation coefficient relating
the known to measured [K+] was 0.99.
ICPAES
Tissue Na+/K+ ratios
were measured by using ICPAES (Jarrell Ash Autoscan, model 25). Tissue
samples weighing 200 mg were dissolved in 1 mL of 3N HCl at 80°C
for 4 hours. The resulting solution was then filtered, diluted
volumetrically to 10 mL, and analyzed for potassium and
sodium content.
EPXMA and Morphometric Analysis
Intracellular sodium and potassium contents were examined using
EPXMA. The techniques used for tissue preparation and analysis
were similar to those previously described and validated in detail by
other investigators.34 35 36 In brief, 1- to 2-mm-thick
remote and infarcted myocardial tissue samples were obtained within 3
minutes of heart excision and "slam" frozen by clamping the tissue
between 2 copper blocks precooled to the liquid nitrogen temperature
(77 K). Samples were transported to a cryomicrotome in a
polystyrene container surrounded by dry ice and kept frozen
during sectioning (–20°C). Frozen 1- to 3-µm-thick sections were
transferred in the microtome to precooled aluminum stubs previously
covered by double-adhesive carbon tape (to eliminate the aluminum x-ray
peak). The frozen sections were then placed in an airtight chamber
surrounded by dry ice (–70°C) and freeze-dried by evacuating the
chamber to a pressure of <20x10-3 mm Hg for 4
hours.
Freeze-dried samples were then analyzed in a scanning electron
microscope (either Hitachi S-4500-II or Hitachi S-570 ) equipped with
an energy-dispersive x-ray spectrometer (Voyager, Noran Instruments,
Inc). On each tissue section, individual cells were clearly visible at
a magnification of x500, and the x-ray detector was engaged.
Individual myocytes lying <100 µm from the freezing surface
were selected at random for x-ray analysis. Intracellular x-ray
spectra were recorded for 300 seconds from a 5x5-µm area
positioned within the myocyte using an accelerating voltage of 10 keV,
an emission current of 10 to 100 µA, a dead time of 20% to 30%, and
a working distance of 18 mm. For each tissue sample, 3 to 10
spectra from within myocytes were acquired. As is commonly done for
EPXMA,35 for each spectra the magnitudes of the sodium and
potassium peaks were expressed as the ratio of the area under the
characteristic peak divided by the area under the continuum
(peak-to-background [P/B] ratio). The results from multiple spectra
in the same area (remote or infarcted) and from the same animal were
pooled.
In addition to taking 1- to 3-µm sections for analysis with
EPXMA, sections of the frozen remote and infarcted tissue samples were
used to measure extracellular volumes. In the same animals studied with
EPXMA, adjacent slices from the microtome were adhered to
room-temperature glass slides and fixed for 3 minutes in ethanol.
The slides were then stained with H&E, examined under a light
microscope using x400 magnification, and photographed at 5 random
locations per slide, where 1 infarct and 1 remote slide were taken for
each animal (n=6). The photographs were then scanned into a computer
for threshold analysis to measure extracellular area. In Adobe
Photoshop, a binary image was generated in which extracellular regions
were white (gray level 255) and all other areas were black (gray level
0). The average pixel intensity over the entire microscopic field was
then measured, and extracellular volume, expressed as percentage
volume, was calculated as extracellular space=(average pixel
value/255)x100.
Statistical Analysis
Results are expressed as mean±SEM. Circumferential location and
extent of TTC negative regions were compared with those of reduced
39K image intensity by linear regression and by
the method of Bland and Altman.37 Image intensities,
potassium contents, P/B ratios, and sodium-to-potassium ratios were
compared using 2-tailed paired t tests. Values of
P<0.05 were considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In all animals, visualization of necrosis by irreversible injury was confirmed by lack of TTC staining and characteristic features such as contraction band necrosis in H&E–stained sections.
MRI and Spectroscopy
Figure 1
shows 4 short-axis
39K MR images extracted from the 3-dimensional
image data with corresponding histological short-axis
slices from a heart subjected to reperfused infarction. A direct
spatial concordance seemed to exist between infarcted areas of
myocardium, identified histologically, and
areas of decreased pixel intensity in the MR images. Figures 2 and 3
show similar results in a rabbit 24 hours after nonreperfused
infarction and in a dog after reperfused infarction, respectively. For
all animals, circumferential extent and center of infarct using the TTC
and the MR image data are shown in Figure 4. The correlation coefficients relating
circumferential extent and location of infarction to regions of reduced
39K image intensity were r=0.97 and
r=0.98, respectively.
|
|
|
|
Figure 5 shows 39K
MR image intensities and K+ concentrations
determined by MRS. Infarcted areas had 51.7±4.8% of the MR image
intensity in remote areas (P<0.001, n=11). Figure 5
also reveals that the potassium concentration, determined by
39K MRS, of infarcted myocardium was
40.5±9.3% of the concentration in remote regions
(P<0.001, n=9).
|
ICPAES and EPXMA
ICPAES revealed that the potassium-to-sodium concentration ratio
of infarcted myocardial tissue was 20.7±2.1% of the remote tissue
(P<0.01, n=5). Potassium and sodium contents within
individual myocytes were determined using EPXMA. The top panels of
Figure 6 show typical scanning electron
micrographs of an unfixed, unstained, freeze-dried cryosection of
control myocardium at both high and low magnifications. The
immediately adjacent section showing the same myocytes was stained with
H&E and is displayed below the electron micrographs for comparison.
Although scanning electron microscopy showed general structural
features of myocardium such as individual myocytes, many
intracellular features were not readily identified. Although the use of
transmission rather than scanning electron microscopy of freeze-dried
cryosections may have improved image contrast and resolution, this
scanning electron micrograph is shown to demonstrate that image quality
was sufficient to identify myocytes and acquire intracellular EPXMA
spectra. Figure 7 shows typical
intracellular spectra from remote and infarcted myocytes. Compared with
the remote region, the spectrum from the infarcted region exhibited a
lower potassium peak and a higher sodium peak. Figure 8 summarizes the EPXMA data for all
animals. Each point represents the average of 3 to 10 spectra
acquired for each animal for a total of 72 spectra analyzed.
From Figure 8, the K+ P/B ratio of
infarcted myocytes was 0.95±0.32 compared with the value inside remote
myocytes, which was 2.86±1.10 (P<0.01, n=6). EPXMA further
revealed that the potassium-to-sodium ratio, in which both quantities
were expressed as P/B, of infarcted myocytes was 0.40±0.17 compared
with the value for remote myocytes, which was 3.07±0.97
(P<0.002, n=6), also shown in Figure 8. These data
demonstrate that the regions of reduced 39K MR
image intensity were associated with a loss of intracellular
K+.
|
|
|
Extracellular Volume
The extracellular volumes in remote and infarcted tissues
determined histologically were 14.3±5.3% and
17.2±7.6%, respectively (P<0.02, n=6).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To our knowledge, this is the first study in which 39K MRI of the heart has been described. We found that 39K MR image intensities were significantly reduced in regions subjected to infarction compared with remote regions. The results of tissue analysis by MRS, ICPAES, and EPXMA provide insight into the mechanisms that govern 39K MR image intensities.
Mechanisms Governing Image Intensity
MR image intensities in our experiments were primarily dependent
on 3 parameters: spin density (concentration),
T1, and T2. Direct
information regarding 39K spin densities can be
found in the MRS data of Figure 5
. The magnitudes of the
differences in tissue-level 39K concentrations in
remote and infarcted regions, 24.3±2.8 versus 9.8±4.5 mmol/L,
respectively, were similar to the differences in
39K image intensities (Figure 5), which
suggests that image intensities are primarily determined by
tissue-level K+ concentrations. Although in
principle regional differences in longitudinal relaxation time constant
(T1) and transverse relaxation time constant
(T2) in remote versus infarcted regions could affect image
intensities, in our experiments differences in T1
would not be expected to influence the results because the repetition
time for our imaging pulse sequence (TR=40 ms) was 4 times longer
than the 39K T1 (10
ms).38 Regional differences in T2
(or T2*), however, may have influenced our
results, and in principle this issue could have been addressed by
directly measuring T2 values in tissue samples.
In practice, we were unable to perform these experiments because the
39K spin densities in infarcted regions were so
low that MRS signal-to-noise ratios were not sufficient to accurately
measure multiple echo amplitudes with increasing echo times. Even
without the T2 data, however, our results (Figure 5) strongly suggest that differences in
39K image intensities were primarily determined
by regional differences in spin density.
Insight into the relationship of tissue-level spin densities to
cellular-level distributions of K+ can be
obtained by consideration of the ICPAES and EPXMA data. By ICPAES, we
found that the ratio of
K+/Na+ was much lower in
infarcted regions demonstrating an imbalance of tissue electrolytes.
Because K+ concentrations are normally far higher
in the intracellular compared with extracellular space, one would
expect that tissue K+ would be greatly affected
by changes in intracellular K+. The EPXMA data
(Figures 7 and 8) confirmed that the loss of tissue-level
K+ demonstrated by MRS and ICPAES was associated
with large losses of intracellular K+. In view of
these data, we conclude that 39K MR image
intensities are reduced in regions subjected to irreversible injury
primarily because of a loss of intracellular
K+.
Effects of Myocardial Injury on K+ Content
After irreversible injury, we found that tissue
K+ content in infarcted myocardium
constituted 40.5±9.3% of the remote region by MRS (Figure 5).
This value is similar to the finding of Jennings et al,7
who reported K+ concentrations of 41.5 versus
21.0 mmol per 100 g fat-free dry weight for remote versus
infarcted myocardium (infarcted myocardium was
51% of remote) in a canine model with 40 minutes of occlusion followed
by 50 minutes of reperfusion. In another study by this same group,
Whalen et al6 demonstrated that K+
concentration of infarcted canine myocardium decreased from
40.6 to 31.3 mmol per 100 g fat free dry weight (infarcted
myocardium was 77% of remote), corresponding to 155 versus
66l.6 mmol/L tissue water (infarcted myocardium was
58% of remote) after 40 minutes of occlusion and 10 minutes of
reperfusion. Jennings et al7 also reported that, 1580
minutes (1 day) after permanent occlusion, the
K+ concentration of infarcted tissue decreased
from 38.3 to 5.4 mmol per 100 g fat-free dry weight
(infarcted was 14% of remote). In our study, 39K
image intensity in infarcted regions was 61.7±3.5% of that of remote
regions 1 hour after the infarct had been reperfused (Figure 5, 
) and 33.9±7.6% of that of remote regions 1 day after permanent
occlusion (Figure 5, 
; P<0.001 compared with
). The similarity between the data of Jennings et
al7 concerning K+
concentrations and our data concerning 39K MRI is
consistent with our conclusion that 39K
MR image intensity primarily reflects regional K+
concentrations.
In previous studies of 23Na MRI by our group,23 24 we found in the same rabbit model (40 minutes of occlusion, 60 minutes of reperfusion) that 23Na image intensity in areas of infarction were 142% of that of remote regions and that the Na+ concentration of these infarcted tissues were 163% of that of remote regions. Similarly, others have found that irreversible injury is characterized by an increase in Na+ and a decrease in K+, and the time course of these changes has been studied.6 7 These results suggest that, for infarcted myocardium, the information reflected by 23Na MRI may be comparable with that reflected by 39K MRI.
During ischemia, however, the information reflected by 23Na compared with 39K MR image intensities may be different. Hill and Gettes1 report that K+ concentrations change within the first 15 to 30 seconds of ischemia, whereas Pike et al39 report that changes in Na+ occur over the time scale of minutes to tens of minutes. Although K+ and Na+ are cotransported by the Na+-K+ ATPase, each ion also has unique transport channels and, in certain situations, 23Na and 39K MRI may reflect different pathophysiological information.
Relationship of Image Intensity to Absolute
[K+]
Although image intensity appears to primarily reflect regional
intracellular K+ concentrations, direct
comparisons of 39K image intensity with absolute
tissue concentration may be more complex. One can compare our results
with estimates of tissue-level K+ concentrations.
Assuming the intracellular space is 75% of the water space, that the
water space is 80% of the total, that intracellular
K+ concentration in normal myocardium
is 140 mmol/L, and that extracellular K+
concentration is 4 mmol/L, tissue K+ would
be 85 mmol/L [=0.8x(0.75x140+0.25x4)]. Experimentally,
however, we found that tissue K+ concentrations
in normal myocardium were only 24.3±1.4 mmol/L by MRS
(Figure 5). Isolated hearts are often subject to mild injury,
which may decrease intracellular K+
concentrations below in vivo values.40 Additionally,
however, this apparent discrepancy might be partially explained by the
results of other investigators, who have suggested that a portion of
the intracellular K+ pool in the heart may be
"invisible" to NMR.41 This "invisibility" may be
due to different magnetic environments within myocardial tissue.
To investigate whether the tissue environment may have influenced our
MRS results, we performed the following experiment. First, we measured
39K spin density by MRS as previously described
and found, as reported in Figure 5, that
K+ concentration in normal myocardium
was 24 mmol/L. Next, the tissue sample was removed from the test
tube, placed in a clean metal bowl, and freeze-thawed several times
with liquid nitrogen. The frozen sample was also crushed with a mortar
while bathed in the liquid nitrogen. Then, the sample was transferred
back to the test tube and dissolved by adding 1 mL of either 4 mol/L
NaOH or 3N HCl. Seven days later, we repeated the identical MRS
experiment on each tissue sample. Interestingly, we found that measured
K+ concentrations increased from 24 to 37
mmol/L (n=4, P<0.001). This result provides strong evidence
that the magnetic environment experienced by K+
ions is in some way affected by the structure of the tissue itself,
resulting in a net reduction in the detected 39K
MR signal. Although this effect appears to measurably affect the
magnitude of the myocardial 39K signal, regions
containing lower K+ concentrations will still be
reflected as regions of reduced image intensity compared with normal
areas.
Study Limitations
The 39K MR images of this study were
acquired at high field (4.7 T) and were of isolated, nonbeating hearts.
To date, the ability to acquire similar images in living humans has not
been demonstrated. The data from this study, however, strongly suggest
that useful physiological information is reflected
by regional 39K MR image intensities.
Summary
In irreversibly injured regions, 39K MR
image intensities were reduced, the spatial extent and locations of
regions of reduced image intensity correlated with infarct size
histologically, and regions of reduced
39K image intensity were characterized by
electrolyte imbalances associated with a loss of intracellular
K+. We conclude that, in the pathophysiologies
investigated, 39K MR image intensity primarily
reflects regional intracellular K+
concentrations.
|
Acknowledgments |
|---|
This work was supported by a biomedical engineering research grant from the Whitaker Foundation (to R.M.J.), NIH Grant HL53411 (to R.M.J.), and a biomedical engineering graduate fellowship from the Whitaker Foundation (to D.S.F.).
Received July 15, 1998; accepted February 5, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1. Hill JL, Gettes LS. Effect of acute coronary artery occlusion on local myocardial extracellular K+ activity in swine. Circulation. 1980;61:768–778.
2. Shine KI, Douglas AM, Ricchiuti N. Ischemia in isolated interventricular septa: mechanical events. Am J Physiol. 1976;231:1225–1232.
3. Hirche HJ, Franz CHR, Bos L, Bissig R, Lang R, Schramm M. Myocardial extracellular K+ and H+ increase and noradrenaline release as possible cause of early arrhythmias following acute coronary artery occlusion in pigs. J Mol Cell Cardiol. 1980;12:579–593.[Medline] [Order article via Infotrieve]
4.
Kleber AG. Resting membrane potential, extracellular
potassium activity, and intracellular sodium activity during acute
global ischemia in isolated perfused guinea pig hearts.
Circ Res. 1983;52:442–450.
5.
Janse MJ, Wit AL.
Electrophysiological mechanisms of
ventricular arrhythmias resulting from myocardial
ischemia and infarction. Physiol Rev. 1989;69:1049–1168.
6. Whalen DA, Hamilton MB, Ganote CE, Jennings RB. Effect of transient period of ischemia on myocardial cells: effects on cell volume regulation. Am J Pathol. 1974;74:381–398.[Medline] [Order article via Infotrieve]
7. Jennings RB, Sommers HM, Kaltenbach JP, West JJ. Electrolyte alterations in acute myocardial ischemic injury. Circ Res. 1963;14:260–269.
8. Lowry OH, Gilligan DR, Hastings AB. Histochemical changes in myocardium of dogs following experimental temporary arterial occlusion. Am J Physiol. 1942;136:474–485.
9. Conrad GL, Rau EE, Shine KI. Creatine kinase release, potassium-42 content and mechanical performance in anoxic rabbit myocardium. J Clin Invest. 1979;64:155–161.
10. Johnson RN, Sammel NL, Norris RM. Depletion of myocardial creatine kinase, lactate dehydrogenase, myoglobin, and K+ after coronary artery ligation in dogs. Cardiovasc Res. 1981;15:529–537.[Medline] [Order article via Infotrieve]
11. Beller GA. Clinical Nuclear Cardiology. Philadelphia, Pa: WB Saunders Co Inc; 1995:37–38.
12.
Schelbert HR, Ashburn WL, Chauncey DM, Halpern SE.
Comparative myocardial uptake of intravenously administered
radionuclides. J Nucl Med. 1998;15:1092–1100.
13.
Bonow RO. Identification of viable
myocardium. Circulation. 1996;94:2674–2680.
14.
Weich HF, Strauss HW, Pitt B. The extraction of
thallium-201 by the myocardium. Circulation. 1977;56:188–191.
15.
Strauss HW, Harrison K, Langan JK, Lebowitz E, Pitt B.
Thallium-201 for myocardial imaging: relation of thallium-201 to
regional myocardial perfusion. Circulation. 1975;51:641–645.
16.
Dahlberg ST, Leppo JA. Myocardial kinetics of
radiolabeled perfusion agents: basis for perfusion imaging.
J Nucl Med. 1987;28:878–885.
17.
Beller GA, Watson DD, Ackell P, Pohost GM. Time course
of thallium-201 redistribution after transient myocardial
ischemia. Circulation. 1980;61:791–799.
18. Gerry JL, Becker LC, Flaherty JT, Weisfeldt ML. Evidence for a flow-independent contribution to the phenomenon of thallium redistribution. Am J Cardiol. 1980;45:58–61.[Medline] [Order article via Infotrieve]
19. Beller GA. Clinical Nuclear Cardiology. Philadelphia, Pa: WB Saunders Co Inc; 1995.
20.
Burstein D, Fossel ET. Nuclear MR studies of
intracellular ions in perfused frog heart. Am J
Physiol. 1987;252:H1138–H1146.
21. Cope FW, Damadian R. Cell potassium by 39K spin echo nuclear MR. Nature. 1970;228:76–77.[Medline] [Order article via Infotrieve]
22. Parrish TB, Fieno DS, Fitzgerald SW, Judd RM. Theoretical basis for sodium and potassium MRI of the human heart at 1.5 T. Magn Reson Med. 1998;38:653–661.
23.
Kim RJ, Lima JAC, Chen E-L, Reeder SB, Klocke FJ,
Zerhouni EA, Judd RM. Fast 23Na MR imaging of
acute reperfused myocardial infarction: potential to assess myocardial
viability. Circulation. 1997;95:1877–1885.
24. Kim RJ, Judd RM, Chen E, Fieno DS, Parrish TB, Lima JAC. Relationship of elevated 23Na MR image intensity to infarct size and tissue sodium redistribution after acute reperfused myocardial infarction. Circulation. In press.
25. Hardy CJ, Bottomley PA, Rohling KW, Roemer PB. An NMR phased array for human cardiac 31P spectroscopy. Magn Reson Med. 1992;1:54–64.
26. Bottomley PA, Lugo Olivieri CH, Giaquinto R. What is the optimum phased array coil design for cardiac and torso MR? Magn Reson Med. 1997;37:591–599.[Medline] [Order article via Infotrieve]
27. Hoult DI, Lauterbur PC. The sensitivity of the zeugmatographic experiment involving human samples. J Magn Reson. 1979;34:425–433.
28.
Resar JR, Livingston JZ, Yin FCP. In-plane myocardial
wall stress is not the primary determinant of coronary
systolic flow impediment. Circ Res. 1992;70:583–592.
29.
Resar JR, Judd RM, Halperin HR, Chacko VP, Weiss RG,
Yin FCP. Direct evidence that coronary perfusion affects
diastolic myocardial mechanical properties.
Cardiovasc Res. 1993;27:403–410.
30. Judd RM, Atalay MK, Rottman JR, Zerhouni EA. Effects of myocardial water exchange on T1 enhancement during bolus administration of MR contrast agents. Magn Reson Med. 1995;33:215–223.[Medline] [Order article via Infotrieve]
31. Rehwald WG, Reeder SB, McVeigh ER, Judd RM. Techniques for high-speed cardiac MR imaging in rats and rabbits. Magn Reson Med. 1997;37:124–130.[Medline] [Order article via Infotrieve]
32. Fishbein MC, Meerbaum S, Rit J, Lando U, Kanmatsuse K, Mercier JC, Corday E, Ganz W. Early phase acute myocardial infarct size quantification: validation of the triphenyl tetrazolium chloride tissue enzyme staining technique. Am Heart J. 1981;101:593–600.[Medline] [Order article via Infotrieve]
33. Henkelman RM. Measurement of signal intensities in the presence of noise in MR images. Med Phys. 1984;12:232–233.
34. Hagler HK, Burton KP, Greico CA, Lopez LE, Buja LM. Techniques for cryosectioning and x-ray microanalysis in the study of normal and injured myocardium. Scanning Electron Microsc. 1980 (pt 2):493–498, 510.
35. Hagler HK, Lopez LE, Flores JS, Lundswick RJ, Buja LM. Standards for quantitative energy dispersive x-ray microanalysis of biological cryosections: validation and application to studies of myocardium. J Microsc. 1983;131:221–234.[Medline] [Order article via Infotrieve]
36.
Buja LM, Burton KP, Hagler HK, Willerson JT.
Quantitative x-ray microanalysis of the elemental composition
of individual myocytes in hypoxic rabbit myocardium.
Circulation. 1983;68:872–882.
37. Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet. 1986;1:307–310.[Medline] [Order article via Infotrieve]
38. Burstein D, Litt HI, Fossel ET. NMR characteristics of "visible" intracellular myocardial potassium in perfused rat hearts. Magn Reson Med. 1989;9:66–78.[Medline] [Order article via Infotrieve]
39.
Pike MM, Kitakaze M, Marban E.
23Na-NMR measurements of intracellular sodium in
intact perfused ferret hearts during ischemia and reperfusion.
Am J Physiol. 1990;259:H1767–H1773.
40.
Barclay JA, Hamley EJ, Houghton H. Electrolyte content
of rat heart atria and ventricles. Circ Res. 1960;8:1264–1267.
41. Fossel ET, Hoefeler H. Observation of intracellular potassium and sodium in the heart by NMR: a major fraction of potassium is "invisible." Magn Reson Med. 1986;3:534–540.[Medline] [Order article via Infotrieve]
This article has been cited by other articles:
 |
|
 |
|
  D. J. Pennell, U. P. Sechtem, C. B. Higgins, W. J. Manning, G. M. Pohost, F. E. Rademakers, A. C. van Rossum, L. J. Shaw, and E. K. Yucel Clinical indications for cardiovascular magnetic resonance (CMR): Consensus Panel report Eur. Heart J., November 1, 2004; 25(21): 1940 - 1965. [Full Text] [PDF]
|
|
|
|
 |
|
 W. G. Rehwald, D. S. Fieno, E.-L. Chen, R. J. Kim, and R. M. Judd Myocardial Magnetic Resonance Imaging Contrast Agent Concentrations After Reversible and Irreversible Ischemic Injury Circulation, January 15, 2002; 105(2): 224 - 229. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. B. Reeder, Y. P. Du, J. A. C. Lima, and D. A. Bluemke Advanced Cardiac MR Imaging of Ischemic Heart Disease RadioGraphics, July 1, 2001; 21(4): 1047 - 1074. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||