Cardiac Maythem Takayuki

Saeed, Masui,

DVM, MD

PhD #{149}Michael F. Wendland, Charles B. Higgins, MD

PhD

Yasuo

Takehara,

#{149}

Radiology

MD

#{149}

Reperfusion and Irreversible Myocardial Injury: Identification with a Nonionic MR Imaging Contrast Medium’

I

The potential of a new nonionic gadolinium complex-gadodiamide injection-to (a) allow distinction between reperfused and occlusive infarction and (b) enable differentiation between reperfused reversible and irreversible myocardial injury

bidity

was

tant

investigated.

rats were reperfused

irreversibly with

occlusive

ministration there was signal

groups

infarction.

Before

After

normal

administration

in and

on Ti-weighted

imof gadodi-

amide injection (0.2 mmol/kg), reversibly injured myocardium indistinguishable

ad-

material, difference

between

regions

of

reversibly injury, 10 with injury, and 10

of contrast no significant

intensity

injured ages.

Three

used: 10 with myocardial reperfused

from

the was

normal

myocardium, while the reperfused irreversibly injured zone showed prominent and homogeneous enhancement. Occlusive infarcts showed three zones of differential enhancement consisting of normal, periinfarction, and infarction regions. Gadodiamide injection provides differential enhancement in reversibly reperfused, irreversibly reperfused, and occlusive infarcts. Thus, it may be useful as a marker of reperfusion and extent of infarction after thrombolytic therapy. Index

terms:

nance

(MR),

Gadolinium contrast

dium, infarction, 511.1214

511.771

Radiology

182:675-683

1992;

Magnetic

#{149}

enhancement Myocardium,

#{149}

resoMyocar-

#{149}

MR,

with thrombolytic or angioplasty to limit

NTERVENTION

therapy myocardial

injury

acute

coronary

serving

left

and

in patients

occlusion ventricular

reducing

both

(3-5). and

at pre-

function

(1,2)

mortality the

the

extent

and

success

mor-

has

been

mentally fication (8,9).

used

clinically

and

The

recognition

edema;

this

infarc-

ensues

over several hours after the occlusion of a coronary artery (10-13). MR imaging is not capable of depicting ischemic but not yet infarcted myocardium without the use of contrast agents. Therefore, trast media have

MR been

Medium

of a nonionic

gadolinium

experi-

edema

A New Contrast

composed

tion with MR imaging is dependent on the increase in signal intensity on T2-weighted images as a result of myocardial

Injection: Imaging

of reper-

and quantiinfarction

of acute

Gadodiamide Nomonic MR

METHODS

Gadodiamide injection (Omniscan; Nycomed AS, Sanofi Winthrop, New York) is

of myocardial

for the detection of acute myocardial

AND

impor-

injury after reperfusion. Depiction and characterization of myocardial injury can potentially provide important clinical data for guiding therapeutic intervention (6,7). The new methods of reperfusion (thrombolytic therapy or early bypass surgery) dictate the need for an effective noninvasive method for detecting the extent of myocardial salvage after reperfusion. Magnetic resonance (MR) imaging

imaging conused in the de-

tection of myocardial ischemia and infarction (13-22). The purposes of the current study were to determine the capability of gadodiamide injection for (a) allowing differentiation between reperfused reversible and irreversible myocardial injury (infarction) distinction between occlusive myocardial

From the Department of Radiology, University of California, San Francisco, 505 Parnassus Aye, San Francisco, CA 94143. Received August 20, 1991; revision requested September 23; revision received October 18; accepted November 4. Address reprint requests to C.B.H.

after

aims

It is increasingly

to determine

fusion

MATERIALS

and (b) allowing reperfused and infarction.

chelate

complex,

diethylenetriaminepentaacetic

acid bismethylamide

(Gd-DTPA

BMA,

gadodiamide), (CaNa-DTPA contrast agent

and caldiamide sodium BMA). This new MR imaging reduces both Ti and T2 re-

laxation

of water

rates

other ity

gadolinium in water

at

protons,

similar

complexes. 10 MHz

The

is 4.6

to

relaxiv-

(mmol/L’

sec’ for Ti and 5.i (mmol/L)’ sec’ for T2 (23), which are similar to those of gadopentetate dimeglumine (24). In comparison with gadopentetate dimeglumine, this

new agent

has the following

tics: gadolinium 17%, osmolality kg,

and

characteris-

content of 27% versus of 789 versus 1,940 mosm/

viscosity

at 20#{176}C of 2.0

versus

4.9

(25). Gadopentetate dimeglumine has a net charge of -2, while gadodiamide injection is a nonionic complex with a zero net charge.

have been

Reduced

cardiovascular

found

with

effects

gadodiamide

injec-

tion compared with gadopentetate dimeglumine (26). The dose of gadodiamide injection used in the current study (0.2 mmol/kg) has a safety index of 170, compared with 60 for gadopentetate dimeglumine used at its standard dose of 0.1 mmol/kg (27). The pharmacokmnetics, biodistribution, and excretion of gadodiamide injection have been studied in experimental animals; like gadopentetate dimeglummne, it demonstrates extracellular

distribution

and

rapid

renal

excretion

(25). The high stability and low toxicity of this new agent are mainly attributed to three factors: (a) the ligand has a strong affinity for Gd3, (b) minimal transmetallation in vivo of Gd3 is expected, and (c) the nonionic nature makes it less likely to cause cardiovascular effects. Moreover, the high

median

lethal

dose

of this

contrast

I

,

RSNA,

1992

Abbreviations: amine-N,N’-diacetic triphenyltetrazolium

BMA

= bismethylamide, DPDP = N,N acid, DTPA = diethylenetriaminepentaacetic chloride.

‘-bis-(pyridoxal-5-phosphate)ethylenediacid,

SI

=

signal

intensity,

TTC

=

675

medium formulation may be due in part to the addition of 5% caldiamide sodium to minimize the possibility of in vivo Gd3 transmetallation (25). Toxicity studies have shown

that

tolerated. mmol/kg ministered symptoms mmol/kg weeks

gadodiamide

injection

is well

The median lethal dose was 34.4 when the compound was adintravenously in mice. No toxic were seen with doses of up to 2 given three times per week for 3 (25).

view,

of Myocardial

Injuries

Sprague-Dawley

rats

Female

were

ostomy pirator Mass). cotomy, cluded

intubated

(Charles

a cervical

appendage.

were

made left

Reperfused

by placing

time for each image was approxi2 minutes. To locate the heart,

scout

coronal

coronary

flow after hours (n

artery

=

transaxial

images

were

20-30

seconds,

and

five

images

were

of 500

mmol/L.

MR

Image

The

Signal

Statistical All

were

loop

values

measured

from

in regions

MR

images

of interest

that

were defined by a rectangular cursor (six to nine pixels, with each pixel 0.23 x 0.23 mm). The size and position of the regions of interest were determined first on the

around

images injection

blood re10) or 2

(n = to produce

kept On

obtained at 15 or 30 minutes of contrast material and

constant each

for the

image,

rest

the

of the

regions

after

images.

of interest

cardial injuries, respectively. Reperfusion of the previously ischemic zone continued for 1 to 2 hours before acquisition of the first MR image. A third set of animals (n = iO) were subjected to 3-4 hours of permanent

to produce

this

model,

high

reversible

or irreversible

coronary

occlusive

occlusion

infarction.

In this

animal

occlusion of the left coronary duces ischemia of the anterior wall

of the

wall.

left

Regional

development tractility

ventricle

but

not

was

confirmed

within

and

i minute

clusion. A tail vein used as the injection

the

septal by

lack

of con-

of coronary

oc-

was cannulated and site for the contrast

agent and supplemental of 57 rats were used ever, 27 died during

MR

artery proand lateral

ischemia

of cyanosis

reperfusion

myo-

anesthetic. in the study; the occlusion

stages

A total howor

of the experiment.

zone

was

signal

identified

as a region

intensity

[SI]

Imaging

Copper forelimb

needles were inserted into a and the abdominal muscles of

Staining

each

and

At the end of each imaging protocol, the extent and location of the injured region were defined by staining the myocardium

rat

connected

to a patient

elec-

trocardiographic monitor (Accusyn-6L; Advanced Medical Research, Milford, Conn) to provide an R-wave signal for triggering

the

animals

were

MR

pulse

placed

sequence.

supine

with

The

inside

a 5.5-

cm-diameter birdcage imaging coil. A cylindric oil phantom was positioned adjacent to the animal to enable standardization of signal intensity for evaluating small changes

course

in repetition

of each

time

during

the

at 85.6 Medical

MHz

with

tion

Fremont, Calif). Spin-echo imaging parameters were as follows: echo time, 20 msec; repetition time, one R-R interval (about 250

msec);

section

676

#{149} Radiology

thickness,

3 mm;

with

TTC.

chemical

Systems,

field

phthalocyanine

Each

dye

of the

remained blue dye infarcts of

Injuries

blue

dye

or

triphenyltetrazolium chloride (TTC). To confirm the presence or absence of irreversible myocardial injuries, hearts subjected to reperfused myocardial injuries (reversible and irreversible) were stained heart

to five sections and stain for is minutes

study.

Images were obtained a CSI 2.0-T system (GE

either

ment

was

produced

viable

sliced

immersed

into

cells,

while

four

TTC histocobra-

in 2%

at 37#{176}C. This brick-red

necrotic

cells

unstained. The phthabocyanine was not used in the reperfused because

it prevents

of the histochemical

the

staining

of contrast

of zones

in the same

discern-

(TTC).

were

expressed

as

of the mean. (before and

material)

heart

and

were

Time after

SI ratios

evaluated

by using a repeated-measure analysis of variance test followed by the Tukey test for multiple comparison. Comparison of contrast values (SI of the injured zone/SI of the normal myocardium) for reperfused injuries (reversible and irreversible) and infarcts

was

ing an unpaired value differences atP < .05.

determined

by

Student t test. were deemed

us-

Mean significant

RESULTS Reperfused

Myocardial

Reversible the

core of the infarction), and (d) the oil phantom (as an SI standard to compensate for small changes in heart rate during the course of the protocol). Changes in the SI values of each region after injection of the contrast medium were measured. The percentage of enhancement produced by contrast material in each region was calculated by using the formula (postcontrast SI/precontrast SI) x 100. The contrast ratio between the injured and the normal myocardium was computed by dividing the SI of the injured zone by the SI of the normal myocardium.

of Myocardial

injection

of

surrounding

data

± standard error SI measurements

occlusive

were

were located in the following sites: (a) the anterolateral wall of the left ventricle (the center of the infarction), (b) the interventricular septum (the normal myocardium), (c) the perimnfarction zone (the rim of the infarcted zone in occlusive infarcts only;

reperfused

Analysis

numerical

means course

Analysis

intensity

To confirm persistent and complete occlusion of the left coronary artery, occlusive infarcts were defined by injecting phthalocyanine blue dye (0.2 mL of a 2% solution) intravenously via the tail vein and allowing it to circulate for 2-3 minutes before killing the animal. This dye imparts a blue color to normally perfused myocardium, while the territory with the occluded artery remains unstained. This stain does not necessarily indicate infarcted myocardium but rather myocardium without blood supply. Stained hearts were photographed and compared with MR images.

stock

% calcium/diso-

trache-

to allow

either iS minutes 10) of occlusion

and

acquired. Baseline transaxial images were obtained at the level of the midventricle before administration of the contrast agent. Gadodiamide injection (0.2 mmol/ kg) was then injected into the tail vein

solution contains 5 mol dium-DTPA BMA.

infarcts

a snare

block acqui-

sition mately

concentration

and ventilated with a rodent res(Harvard Apparatus, South Natick, After performance of a left thorathe left coronary artery was ocnear its origin beneath the left

atrial the

via

phase-encoded per step; and points. Total

acquired at 3, 15, 30, 45, and 60 minutes. Gadodiamide injection was supplied as a sterile, aqueous, and colorless solution at a

River Laboratories, Wilmington, Mass) weighing 240-280 g were anesthetized with an intraperitoneal injection of pentobarbital sodium (Nembutal; Anthony Products, Arcadia, Calif) (50 mg/kg). Animals

128

four excitations 512 complex data

over

Models

x 60 mm;

60

steps; size,

Injuries

myocardial

injury.-In

10

rats, the change in SI in reversible ischemic injury was monitored for 60 minutes. Figure 1 shows a series of transaxial images of a rat subjected to 15 minutes of coronary occlusion folbowed by 1’/2 hours of reperfusion,

obtained utes

before

after

and

3, 30, and

administration

mmol/kg of gadodiamide Before administration agent,

ence

there

was

no

in SI between

versibly

injured

60 mm-

of 0.2

injection. of the contrast

significant

normal

differ-

and

myocardium

re(Fig

1,

Table). After administration of gadodiamide injection (n = 10), homogeneous increases in the SI of both normal myocardium (from 0.33 ± 0.03 to 0.68 ± 0.05) and previously ischemic regions (from 0.35 ± 0.03 to 0.70

± 0.05)

were

in Figure

2, similar

hancement

were

observed.

As shown

percentages found

of en-

in both

nor-

mal and reversibly injured myocardium. Both regions had the same pattern of washout of the contrast agent during the 60-minute observation period (Table, Fig 2). Differential contrast

gion/SI tween

(SI of reversibly

of normal the

reversibly

injured

re-

myocardium) injured

beand

nor-

mal regions was not observed after administration of gadodiamide injection (Fig 3). TTC staining performed March

1992

terventricular

wall

septum

and

of the left ventricle,

normal

myocardium,

terolateral indicating Animals

posterior

indicative whereas

wall remained the presence demonstrated

of

the

an-

unstained, of infarction. transmural

infarcts in at least three tions (2-3-mm thickness).

to four

Occlusive

Infarct

Myocardial

sec-

Figure 6 shows transaxial images of a heart subjected to 3’/2 hours of permanent coronary occlusion before and after administration of 0.2 mmol/kg of gadodiamide injection. SI values (0.29 (0.30

of normal ± 0.01) and ± 0.02) were

different images

myocardium the infarcted region not significantly

on Ti-weighted spin-echo without use of contrast me-

dium (Fig 6a-6d, Table). istration of gadodiamide

After admininjection,

three distinct enhancement farcted, and

regions of differential of normal, penininfarcted myocardium

were

(Fig

noted

6e).

The

core

of the

infarct appeared significantly darker than normal myocardium during the first

few

minutes

after

the

injection

(Fig 6e). At the earliest time point ter administration of gadodiamide injection,

d.

C.

Figure 1. Transaxial MR images of a rat subjected to 15 minutes of left coronary occlusion followed by 1 Y2 hours of reperfusion, obtained before (a) and 3 (b), 30 (c), and 60 (d) minutes after administration of 0.2 mmol/kg of gadodiamide injection. There is no difference in SI between the normal and previously ischemic zones.

the

an animal

subjected

to 2 hours

coronary occlusion hours of reperfusion.

weighted

images

followed by Spin-echo

showed

of left 1’/2 Ti-

no signifi-

cant difference in SI between normal (0.28 ± 0.03) and irreversibly injured (0.34 ± 0.04) myocardium before injection of the contrast medium (n = 10). The ratio of precontrast SI of irreversibly injured myocardium to that of normal myocardium was 1.18 ± 0.07. After administration of gadodiamide injection, the SI of normal myocardium increased from 0.28 ± 0.03 to 0.61 ± 0.02 at 3 minutes

and

then

gradually

decreased

to

0.41 ± 0.02 at 60 minutes. Enhancement of the irreversibly injured region was significantly greater (from 0.34 ± 0.04 to 1.27 ± 0.05 at 3 minutes; P < .05) than that of normal myocarVolume

182

#{149} Number

3

dium

(Figs

4, 5; Table).

Maximum

en-

hancement of the reperfused infarcted region was achieved 3 minutes after injection (393% ± 23% of the control value) and persisted for at least 60 minutes (294% ± 22% of the control value) (Fig 5, Table). The distribution of gadodiamide injection in the reperfused region was uniform and homogeneous. Figure 3 demonstrates the contrast (SI of infarcted region/SI of normal myocardium) between normal and irreversibly injured myocardium throughout the 60-minute observation period. In reversibly and irreversibly injured myocardium, similar baseline SI ratios for injured to nor-

mal myocardium were found, but there were different SI ratios after injection of the contrast agent (Fig 3). The sites of greater myocardial enhancement emia and

the

presence

chemical produced

on MR reperfusion

images after correlated

of infarction

ischwith

at histo-

staining (TTC). TTC staining a brick-red color in the in-

zone

was

de-

lineated by a rim of peripheral enhancement surrounding a central zone that did not show significant enhancement. During the course of 60 minutes, enhanced, hancement delayed

to the approximately 3 hours after initial occlusion revealed no evidence of infarction in this group of animals. Irreversible myocardial injury.Figure 4 shows typical MR images in

infarcted

af-

the

central region slowly and the pattern of the ensuggested that there was delivery of the contrast agent

central

a moderate

zone

mal

myocardium

0.64

± 0.06;

bowed

(Fig 7). There

P

in SI of the nor(from 0.29 ± 0.Oi to < .05) at 3 minutes, fob-

by gradual

diminution

over

course of 60 minutes, reaching of 0.46 ± 0.03 (Table). A bright zone in the form

doughnut normal

was

increase

the

a level of a

was recognized between and infarcted myocardium;

this was the periinfarction zone (Fig 6e). The SI of the periinfarction zone reached a peak value (0.9i ± 0.08) at 30 minutes and did not change significantly during the remaining 30 mmutes (0.84 ± 0.06) (Table). Enhance-

ment

values

for the central

and periinfarction cantly different,

zones with the

infarction were significentral in-

farction zone becoming clearly delineated over the course of 60 minutes (Fig 8). A crescentic area of high SI adjacent to infarcted ing. This ing blood tricle along

the endocardiab wall region was a consistent

likely

represents

in the cavity the infarcted

of the find-

slowly of the wall Radiology

flow-

left yen(Fig 6e). #{149} 677

--, -

-r Gadodiamide

Injection

After Administration

..

in Rats Subjected of Contrast

to Three

Different

Agent 45 minutes

60 minutes

0.48

±

0.04k

0.45

±

0.03*

0.50 0.43

±

0.04k 0.02*

0.49

±

0.04k

±

0.4i

±

0.02*

1.03 0.47 0.85 0.51

±

0.04

0.98

±

±

0.02* O.O7t 0.05*

0.46 0.84 033

±

0.04t 0.03* 0.06*t 0.05*

± ±

± ±

,. ..

450

Figure 9 demonstrates the markedly different profile of contrast enhancement (SI of infarcted region/SI of normal myocardium) after administration of gadodiamide injection in both occlusive (central zone) and reperfused myocardial infarcts. There were clear differences in contrast enhancement between the two types of infarction. Figure 10 shows the influence of gadodiamide injection on contrast enhancement (SI of infarcted region/SI

of

normal

myocardium)

SI ratio (ischemic/normai

375

-..-

Reperfused

-0-

inlured Normal

Si)

reversibly zone myocardium

300

SI (%)

225

1

50

75

-..-

-L

control

3

15

30

45

60

control

tIme (mln)

3

15

irreversibly

zone

Reperiused reversibly iniured soy 30 45 60

time (mm)

2.

in

Repertused inlured

05

3.

Figures

2, 3. (2) Influence of gadodiamide injection on SI in normally and reversibly injured myocardium. This figure shows the time course of changes in SI after injection of the contrast agent. There was no significant difference in enhancement between normal myocardium and the reversibly injured region. All values are expressed as a percentage of the control value, which was considered to be 100%. * = J) < .05 for postversus precontrast SI. (3) Effect of gadodiamide injection on SI ratio (SI of injured region/SI of normal myocardium) as a function of time for reperfused reversibly and irreversibly injured myocardium. Note that there is a significant difference in the contrast enhancement profile between the two injuries for at least 60 minutes after injection. * = P < .05 for reversibly versus irreversibly injured (infarcted) reperfused myocardium.

reperfused, irreversibly injured myocardium and the periinfarction zone of occlusive infarcts as a function of time. The enhancement pattern in the two regions was similar OVC time, but a significant difference in the magnitude of the enhancement was observed.

The

site of the infarcted wall on MR correlated well with the region delineated by the phthabocyanine blue dye. Occlusion of the coronary artery was confirmed in all imaged animals. images

DISCUSSION The

current

study

major findings: infarction (both sive

infarcts)

produced

(a) Acute reperfused was

clearly

three

60 minutes,

thereby

delineated

administration of at

permitting

window for image acquisition. (Ii) Reversibly injured reperfused myocardium was not distinguishable from normal myocardiurn. Thus, irreversible myocardial injury is selectively detected after administration of this new contrast agent. (c) Gadodiarnide injection at a dose of 0.2

infarcts, 678

allows and suggesting

#{149} Radiology

distinction occlusive that

of gadodiamide

myocardium

wide

mmol/kg reperfused

tion

myocardial and occlu-

after administration of 0.2 mmol/kg gadodiamide injection. Demarcation of the infarcted zones persisted for

beast

with gadodiamide injection has the potential to document reperfusion. A recent study (28) indicated that myocardial contrast enhancement and its persistence are dose dependent. In that study, there was a close relationship between the concentra-

between myocardial MR

imaging

a

injection

and

signal

of different

0.5 mmol/kg).

Masui

in the

intensity

after

doses

(0.1-

et al (28) also

found that administration of 0.3 mmol/kg of gadodiamide injection produced optimum enhancement in the myocardium. The main advantage of this new nonionic contrast agent, gadodiamide injection, is that it has a high safety index of 170 at a dose of 0.2 mmol/kg. Compared with other contrast agents for Ti-weighted imag-

ing, mine

such (safety

0.1 mmol/kg)

as gadopentetate index

=

dimeglu60 at a dose

(27) or manganese

N, N ‘-bis-(pyridoxal-5-phosphate)ethylenediamine-

N,N’-diacetic

acid

of

(DPDP) (safety index = i2 at a dose of 0.4 mmol/kg) (i4,i9), gadodiamide injection has a safety index that is 5.6 and 7.i times higher, respectively, on a molar basis. Moreover, gadodiamide injection causes no hemodynamic alterations after administration of 0.i, 0.3, and 0.5 mmol/kg into the jugular vein as a bobus. However, gadopentetate dimeglumine produces dosedependent negative inotropic and vasodilatory effects even at a dose of 0.1 mmol/kg (26). Previous reports have indicated that detection and characterization of acute myocardiab infarction with MR imaging is based on alterations in tissue relaxation rates (Ti and T2), with resultant changes in regional SI on MR images (29,30). At MR imaging, reperfused myocardial infarcts in dogs subjected to 3 hours of coronary occlusion followed by 30 minutes of reflow showed a significant increase in SI and T2 (3i-34). Dogs with occluMarch

1992

450

375

300

SI

(%)225

ISO mnlured zone Normal myocardlum

75 -0-

control

3

15

30

45

60

time (mm)

Figure

5.

Influence

odiamide

of 0.2 mmol/kg SI in normal

injection

of gadmyocar-

on

dium and irreversible myocardial injury (n = iO). There is a significant difference

(t

P