MAGMA MagneticResonanceMaterialsin

Physi~BiologyandMedicine

ELSEVIER

Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191-207

Abstracts

31p-MR-spectroscopy before and after aortic valve replacement

Meinrad Beer, Mardn Viehrig, Tobias Seyfarth, J6rn Sandstede, Thomas Pabst, Werner Kenn, Susanne Bengsch, Gerald Bertsch, Michael Horn, Wilfried Landschiitz, Martin Meininger, Markus yon Kienlin, Stefan Neubauer, Dietbert Hahn Wiirzburg, Germany Introduction: Cardiac high-energy phosphate metabolism, measured as the phosphocreatine-to-adenosintriphosphate (PCr/ATP) ratio using 31p-MR-spectroscopy, is deranged in patients with aortic valve disease [1,2]. Moreover, reductions of PCr/ATP ratios correlate with left ventricular end-diastolic pressures and with end-diastolic wall stress. Purpose of the present work was therefore, to examine whether changes in cardiac metabolism are reversible 3 months after surgical valve replacement. Materials and methods: Using a double-oblique 3D-CSI technique all spectra were obtained on a 1.5 T Magnetom VISION scanner (Siemens Medizintechnik, Erlangen, Germany). A voxel of 25 ccm was positioned in the septal myocardium. Spectra were corrected for saturation, nuclear Overhauser enhancement (nOe) and blood contamination. The postprocessing was done using the 'AMARES'-software [3]. Left ventricular (LV) function was analyzed by short axis cine MRI breath hold sequences (slice thickness 8 mm). First, 15 healthy agematched volunteers were examined. Until now, 5 patients with aortic valve disease (2 stenoses, 3 regurgitations) were examined before and 3 months after SVR. Results: The PCr-linewidth of all spectra

(before as well as after valve replacement) was below 10 Hz. The PCr/ATP ratio of healthy volunteers was 1.72+0.31 (SD). The PCr/ATP ration of spectra, acquired from the septal myocardium of all patients before valve replacement, was 1.14 + 0.21. Fig. 1 show a patient with aortic stenosis before and 3 months after valve replacement. 3 months after valve replacement left myocardial mass decreased from 250.5 ± 39.7 to 187.4 _ 41.4 g (p < 0.05), showing regression of LV hypertrophy. However, the PCr/ATP ratio remained depressed (1.36_ 0.18; p = 0.22). Conclusion: 31p-MR-spectroscopy demonstrates abnormal energy metabolism in aortic valve disease. In spite of regression of hypertrophy, energy metabolism remained depressed 3 months after surgery, Further follow-up studies I2 months after valve replacement are currently being performed. References [1] Neubauer et al. (1997) Cardiac high-energy phosphate metabolism in patients with aortic valve disease assessed by 31P magnetic resonance spectroscopy. J. Invest. Med. 45: 453-462. [2] Conway et al. (1991) Detection of ionophosphocreatine to ATP ratio in failing hypertrophied human myocardium by 31P magnetic resonance spectroscopy. The Lancet 338: 973-976. [3] Vanhamme L et al. (1997) Improved method for accurate and efficient quantification of MRS data with use of prior knowledge. J. Magn. Reson.

Fig. 1. Cardiac 31p spectra before (a) and 3 months after (b) SVR. 1352-8661/98/$ - see front matter © 1998 Elsevier Science B.V. All rights reserved. PII S 1352-8661 (98)00065-9

192

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191 207

Growth hormone-induced left ventrieular hypertrophy is not but pre~ sure overload hypertrophy is associated with increased susceptibility to ischemia

Mark de Groat, Michael Horn, Andrea Leupold, Charlotte Dienesch, Stephanie Hiigel, Antonio Cittadini, James P Morgan, Stefan Neubauer, Hinrik Str6mer Department of Medicine, Wiirzburg University, Germany

Introduction: Growth hormone (GH) is under clinical investigation for heart failure treatment. It is unclear whether susceptibility to ischemia and reperfusion in GH-induced left ventricular hypertrophy (LVH) is increased as it is in pressure overload LVH. We tested the recovery of developed pressure (AP), high energy phosphate metabolism and Ca 2 + handling in isolated isovolumic rat hearts after 20 min of no-flow ischemia and 30 min of reperfusion in 2 groups of LVH: After 4 weeks of GH treatment (LVH-GH) or 8 weeks after banding of the ascending aorta (LVI-I-AB) compared to age-matched controls. Methods: Isolated hearts from male Wistar rats (355 _+28 g) were perfused isovolumetrically (LV balloon) with Krebs-Henseleit buffer (1 1 mM glucose, phosphate-free, oxygenated, 37°(2) at a coronary flow of 12 ml rain 1 g 1 heart weight- 1. Left ventricular balloon volume was set to 50% of the volume at which maximal developed pressure was attained. Hearts were allowed 15 min of stabilization before 20 min of total global ischemia followed by 30 min of reperfusion. Intracellular [Ca2+ ]i and high-energy phosphate metabolism were measured by aequorin normalised by fractional luminescence ( n = 3 × 20) or 31 P-MRS ( n = 3 × 9) respectively. NMR spectra were recorded on a 7.05 T Bruker AM system every 5 min with 4 K data points by signal-averaging 152 FIDs, TR = 1.93 s,

Contribution of NO to ischemia/reperfusion injury in the saline-perfused heart--a study in endothelial NO-synthase knockout mice

Ulrich Flrgel, Ulrich K.M. Decking, Axel Godecke, Jiirgen Schrader Institut fiir Herz-und Kreislaufphysiologie, Heinrich-Heine-UniversitS.t D/isseldorf, Postfach 101007, D-40001 Diisseldorf; e-mail: [email protected] de

Introduction: Under physiological conditions, nitric oxide (NO) exerts a variety of functions in the heart: NO is involved in the regulation of vascular tone, has been reported to modulate myocardial contractility, and is known to reduce leucocyte adhesion at the endothelial surface. The role of NO in ischemia and reperfusion, however, is not fully understood [1]. Due to the rise in cytosolic Ca 2 + during ischemia, a burst of NO formation from endothelial NO synthase (eNOS) can be expected upon reperfusion. This most likely contributes to post-ischemic hyperemia and may temporarily decrease cardiac contractility, thereby increasing oxygen supply and reducing ischemia/reperfusion injury. On the other hand, an enhanced NO production may cause increased peroxynitrate formation, thereby exacerbating myocardial damage [1]. While some studies reported a cardioprotective effect of NOS inhibitors-consistent with less ischemia-reperfusion injury due to reduced oxidant stress from both hydroxide radical and peroxynitrite-contradictory results were obtained in others [2-4]. This may be put down in part to species- and system-related differences but also to side-effects of the NOS inhibitors employed, possibly masking the NO-induced beneficial or detrimental effects. In order to investigate the contribution of endogenous NO to ischemia/reperfusion damage without inhibitor interference, isolated perfused hearts of endothelial NO synthase knockout (eNOS - / - ) mice [5] were compared to that of wild-type animals. In an additional series of hearts the NOS inhibitor L-N-monomethylarginine (L-NMMA) was applied. Hearts were subjected to 16 rain of global ischemia and 60 min reperfusion. In addition to the

at a flip angle of 45 ° and were corrected for partial saturation.

Results: Both LVH groups showed an increase in LV-weight/LV-volume ratio which is a measure of concentric growth. Post-ischernic recovery of AP was impaired in LVH-AB but not in LVH-GH. The maximum [Ca z+ ]i-overload occurring in the first minute of reperfusion was exacerbated in LVH-AB only. No differences were found for high energy phosphates and pH i.

Control

LVH-GH

LVH-AB

LV-weight/LV8.8+2.6 Volume [g/ml] AP baseline 87+2.6 [mmHg] AP recovery [% of 72+4 baseline] [Ca2 + ]i-overload 1.3+0.1 [~M] PCr recovery [%] 78+6

11.1+0.3"

11.9+3.4"

113+2.3"

117+3.4"

73+6

52+4* #

1.5+0.1

2.3+0.2* #

72+6

72+8

*p < 0.05 versus control, # p < 0.05 LVH-AB versus LVH-STH. Conclusion: Despite similar concentric growth in GH-induced and pressure overload hypertrophy, only the latter shows increased susceptibility to ischemia and reperfusion. This effect can be explained by an increased Ca 2 + overload upon reperfusion but not by differences in high-energy phosphate metabolism or pHi.

evaluation of cardiac contractile function, 31p NMR spectroscopy was used to monitor the energy state and the intracellular pH (pHi) of the hearts. Materials and methods: Mice weighing 25 to 30 g were injected with 250 U heparin IP and anaesthetized with urethane (1.5 g/kg). The hearts were rapidly excised, the aorta was cannulated, and hearts were retrogradely perfused by a nonrecirculating Langendorff technique at a constant pressure of 50 mmHg. The heart was placed inside a 10-ram NMR tube and transferred into a temperated (37°C) probe inside a Bruker AMX 400 NMR spectrometer. Coronary flow, left ventricular developed pressure (LVDP), heart rate, dP/dt, and coronary perfusion pressure (CPP) were continuously recorded. After hearts have stabilized inside the magnet, cardiac pacing (600 min :) was initiated and continued throughout. Fifteen rain later, the coronary perfusion rate was fixed to the steady state flow achieved during pacing and was maintained constant. Hearts were subjected to 16 rain total global ischemia, followed by 1 h reperfusion at the initial flow rate. Partially saturated 31p NMR spectra were recorded from a 10 mm IH/31p dual probe. 240 transients at a flip angle of 70° were accumulated with a repetition time of 1 s resulting in a temporal resolution of 4 min. Spectra were continuously acquired: at baseline, during 16 min of global ischemia, and during the 60 min of recovery. Results and discussion: Basal cardiac function, PCr/ATP ratio and pHi were identical in e N O S - / and wild-type hearts (each n = 8, Fig. 1). Ischemia induced a rapid decline of both LVDP and phosphocreatine (PCr) in each group. During 16 min of ischemia, ischemic contracture developed resulting in a left ventricular enddiastolic pressure (LVEDP) of approximately 70 mmHg in all hearts, while pHi fell to approximately 6.1. During ischemia, there was neither a significant difference in the time course of LVDP and PCr (Fig. 1), nor in LVEDP, dP/dt, CPP, or adenosinetriphosphate (ATP), pH i, and inorganic phosphate between the two groups. In reperfusion, however, a significant improvement of the post-ischemic functional and metabolic recovery became apparent in the eNOS - / - hearts.

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191-207 While in the wild-type group, LVDP only recovered to about 55% of its basal value, in the e N O S - / group LVDP attained approximately 85% of basal levels at the end of 60 min reperfusion (Fig. 1). Similarly, the recovery of PCr was significantly enhanced in the transgenic hearts as compared to the wild-type controls (Fig. 2). eNOS-/hearts also showed a better restoration of dP/dt (92 + 8% vs. 72 + 10% of the basal value in wild-type hearts) and a significant lower LVEDP (25 ___8 mmHg vs 40 ___10 mmHg in wildtype hearts), the latter indicating improved recovery from ischemic contracture after one hour of reperfusion. Moreover, a slightly better recovery of myocardial ATP was observed in the hearts of transgenic mice. Perfusion of wild-type hearts with 100 ~tM L-NMMA increased CPP from 50 to 58 +_2 mmHg but induced no alterations in functional or energetic parameters under control conditions. When compared to non-treated controls, L-NMMA had no effect on the decline in cardiac function and energy status during 16 min of ischemia but tended to improve the recovery of both LVDP and PCr in reperfusion. However, the restoration of functional and metabolic parameters (LVDP 43-t-6 mmHg, LVEDP 32 + 8 mmHg, PCr 76 + 10%, n = 8) was less pronounced than that of the eNOS - / - hearts after

70

one hour of reperfusion. The data obtained provide clear evidence that endogenously formed NO significantly contributes to ischemia/ reperfusion injury in the saline-perfused heart, possibly by peroxynitrite fornaation from NO [7,8]. The lesser effect of pharmacological inhibition of NO synthase by L-NMMA may be due to incomplete blockade of NO formation which may partially explain conflicting results previously reported in saline-perfused hearts of other species [1,6]. It remains to be seen to what extent inhibition of NOS improves cardiac function in ischemia/reperfusion also in the heart in situ. References [1] Nonami Y 1997 Jpn Circ 61: 119-132. [2] Schulz R et al. 1995 Cardiovasc Res 30: 423-439. [3] Depr6 C et al. 1995 Circulation 92: 1911-1918. [4] Yang BC et al. 1997 Loife Sci 61: 229-236. [5] G6decke et al. 1998 Circ Res 82: 186-194. [6] Naseem SA et al. 1995 J Mol Cell Cardiol 27:419 426. [7] Wang P e t al. 1996 J Biol Chem 271: 29223-29320. [8] Yasmin W e t al. 1997 Cardiovas Res 33: 422-432.

LVDP

ischemia

193

ischemia

PCr

120

60 , ~ 50 "IE 40

&. 30 ~

20 10 L

'~,........... t

0 -~

0 , , ,

0

t6

32

48

64

time [mini

¢,,

. . . . .

, , , , , , ,

0

0 , , o , , ,

16

.

.

.

.

32

.

.

.

.

.

48

.

,,

.

.

.

.

.

.

.

.

64

time [rain]

Fig. 1. Temporal changes of mechanical function (LVDP), at the left) and the high energy phosphate PCr (at the right) during 16 min of global ischemia and subsequent reperfusion for 1 h. O, wild-type hearts; O, e N O S - / hearts; each n = 8; *p < 0.05; **p < 0.01.

Mathematical models and 13C-NMR spectroscopy for evaluating the TCA cycle flux in rat hearts perfused with various substrates M. Herve I, J.A. Hoerter 2, P. Mateo 2, S. Tran-Dinh 1 I S.B.P.M./D.B.C.M., CEN Saclay, 91191 Gif-sur-Yvette, France. 2 U 446 INSERM, Universit6 Paris Sud, 92296-Chatenay-Malabry, France.

Introduction: Many mathematical models have been proposed since 1951 for evaiuating the flux in the TCA cycle but most of them are very complicated and only consider the special case where the TCA cycle is not associated with another. This assumption is not always true. For example, the TCA cycle is in connection with the malate-aspartate shuttle in hearts or with the glyoxylate cycle in yeast [1, 2]. In the present communication, two mathematical models were constructed for evaluating the TCA cycle flux in rat hearts for the following cases: (I) the TCA cycle is not associated with another pathway; (2) the TCA cycle is coupled to the malate-aspartate shuttle. Materials and methods: (1) Wistar male rats were anesthetized with ethyl carbonate and the hearts were perfused by the Langendorff technique at a constant flow with 100% enriched [2-a3C]acetate 10 mM [3] or [U-13f]glueose 5 mM in the absence of insulin. A left

ventricular balloon allowed to impose working conditions which was estimated by the rate pressure product (RPP). The cells extracts were prepared by adding cold perchloric acid (1.5 M) and then adjusting to pH = 7.0. (2) ~3C-NMR spectra were recorded at 125.7 MHz in 5 mm tubes on a Bruker AMX-500. Mathematical models: In general, 13C-NMR spectra of glutamate result in the superposition of the spectra of 32 isotopomers. Each isotopomer will split after one turn in the TCA cycle into four or 16 different others with unequal population. Our matrix method was used for studying the isotopic transformation of the glutamate carbons. The global transformation through the TCA cycle is given by C¢~+ 1 = MC~ where G~ and C¢~+ 1 represent the glu-isotopomers in the state (n) and (n + 1), respectively; M is the matrix corresponding to the isotopic transformation of the glu-isotopomers. (1) Model with constant Glu-pool size: When the TCA cycle is not associated with another pathway and the glutamate concentration remains constant during incubation, the time dependence of all glutamate isotopomers (Fig. 1) can be expressed by a simple general equation: dG/dt = a [ M - I]G [1] where a is the flux constant. (2) Model where the TCA cycle is associated with the malate-aspartate shuttle: dGi/dt - bEmijAj -- aG i [2]; dAi/dt = bEnijGj - bAh- [3] where Gi and A i are the isotopomers of glutamate

194

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191-207

and aspartate, respectively; a e t b represent the fraction of glutamate and aspartate in isotopic exchange with the TCA cycle, respectively. Results: The first model has been used for evaluating the TCA cycle flux in rat hearts perfused with [2-zaC]acetate and the second one for rat hearts perfused with [U-13C]glucose. Surprisingly, in the absence of insulin, aspartate instead of glutamate was found to be the most abundant metabolise of [U-13C]glucose, indicating that aspartate transaminase (which catalyses the reversible reaction: [Glutamate + Oxaloacetateo2Oxoglutarate+Aspartate]) is highly active. The amount of aspartate was about two times larger than glutamate (Fig. 2). Excellent agreement between the computed and experimental data was obtained, showing that acetate represents the main precursor of acetyl-CoA. By contrast, with glucose in the absence of insulin, only

1

i

i

1

i

i

i

i

i

i

~'

PZ)

"~

\/

,"

g .~ 0.s \ g. .04 :<

S

,

I/,

i I,,

"

:" \

\

O-~r:' p,. ~

l l

"-.,

r-el

/ t

....•_

, ".

G1~-¢4

,"

0

2

(

1.

\Ul

.

~

]~z2~,

~

_

~

1

rain

4 ,':

[

-I

/

I~a o

Ch-C3

"J

,,,

\ P .

1o:

I Asp I I I Glu I = 2

...'"

w

o~ 0.7

References: [1] Tran-Dinh, S., Beganton, F., Nguyen, T.T., Bouet, F. and Herve, M. (1996) Eur. J. Biochem. 242, 220-227. [2] Tran-Dinh, S., Bouet, F., Huynh, T. and Herve, M. (1996) Eur. J. Biochern. 242, 770-778. [3] Tran-Dinh, S., Hoerter, J.A., Mateo, P., Bouet, F. and Herve, M. (1997) Eur. J. Biochem. 245, 497-504.

]

For hearts perfused with [U-13C]glucose

0.9 o8

41% of acetyl-CoA is formed from glucose while the rest is derived from endogenous substrates. The exchange between aspartate and oxaloacetate or between glutamate and 2-oxoglutarate is fast in comparison with the biological transformation of intermediate compounds by the citric acid cycle.

6

l 10 12 Ttme (rain)

14

16

18

20

45

40

30

Fig. 2. N M R spectra of perchloric acid extracts obtained at various times from rat hearts perfused with [UJ3C]Glc in the absence of insulin.

Fig. 1. Changes in the population of glutamate isotopomers computed for a = 0.35 rain 1 All the glutamate carbons are unlabeled at t = 0 (i.e. P12 = I).

Combined high speed NMR imaging methods for determination of perfusion and vasodynamics in the isolated rat heart Karl-Heinz Hiller~, Wolfgang R. Bauerb, Sabine Volla, Axel Haase a a Physikalisches Institut, Universit~it Wtirzburg, 97074 Wiirzburg, Germany b II. Medizinische Universit~itsklinik Mannheim, 68135 Mannheim, Germany

Introduction: The aim of this study was to combine a non-contrastagent NMR method of measuring perfusion and a NMR-microscopy method which reflects the state of vasodynamics of coronary microvessels to obtain a powerful tool to investigate the effects of vasoactive substances. Both methods were developed by our group [1, 2]. The principle of the perfusion measurement is that spins of a selected slice in the short axis view of the heart are inverted and a perfusion sensitive TI relaxation of these spins is observed. According to model calculations variations of perfusion P are determined from: A(1/T1) = AP/~. (~ = tissue/perfusion partition coefficient of water). Perfusion changes were also determined by simultaneous injection of colored microspheres This microsphere method was modified for simultaneous use in N M R microscopy and previously reported [3].

Methods: Six isolated rat hearts were studied (perfusion with Krebs Henseleit buffer in the Langendorff mode, on-line registration of coronary flow, left ventricular pressure). Measurements of perfusion changes and vessel diameters were performed before, 10 rain after infusion of the vasodilator nitroglycerine (0.5 mg/ml) and 10 min after switch off the infusion. NMR-imaging was performed on an 11.75 Tesla magnet (AMX 500, Bruker). Perfusion measurement: Spins of a slice (short axis view) 4-6 mm below the valvular plane were inverted (slice thickness = 3 mm) and T1 maps were gained in this slice by 16 x Snapshot FLASH images (spatial resolution 140 pm in plane, slice thickness = 1.5 mm, TR = 3.6 ms). Corresponding microsphere measurements were always performed simultaneously (20.000 MS/color). Measurement of vasodynamics: Coronary microvessels were imaged by middiastolic triggered flow-weighted 2-D gradient echo multislice pulse sequence, with an TE = 1.6 ms and TR = 1 heart cycle ~ 200 ms. Eight slices were gained simultaneously in the short axis view (slice thickness = 200 pm, slice distance = 500 pm, spatial resolution 70 ~tm in plane). Only vessels which intersectinned perpendicularly the imaging plane were considered. Corresponding vessel cross sections were compared and the relative variations during and after nitroglycerine infusion were evaluated.

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191-207

195

Results: The changes of global perfusion determined by T1 vs. colored microsphere measurements are in the same range during nitroglycerine infusion (3.23-t-0.44 and 3.26_+0.78 ml/min/g + SEM) and afterwards ( - 5.09 + 1.19 and - 5.57 + 1.16 ml/min/g + SEM). Calculations of global perfusion by ultrasonic measured coronary flow were also in agreement with these data (3.09 + 0.75 and - 4 . 8 1 +0.92). The corresponding relative changes in vessel diameters showed an increase during nitroglycerine infusion (11.09% + 2.66) followed by a decrease after switch off the infusion (15.23% + 4.03). Conclusion: It could be shown that it is possible to obtain accurate values of perfusion changes with the slice selective

spin inversion NMR technique. The combination of the two NMR imaging techniques will be a helpful tool kit by any physiological or pharmacological studies in organs where alterations of perfusion and vasodynamics are supposed. So this may be the first step to get an 'one-stop-shop' for microcirculatory rat heart studies.

Improved air temperature control system for isolated organ NMR experiments RJF Houston, R J Legtenberg, A Heerschap, B Oeseburg Faculty of Medical Sciences, University of Nijmegen, Post Box 9101, NL-6500 HB Nijmegen, The Netherlands

Improved control of experimental conditions reduces variability of results, increases the power of experiments, and reduces necessary group sizes, and thus in this type of work, the number of experimental animals required. This new system is simple, cheap and efficient and allows perfused organs to be maintained in air at an accurately controlled temperature.

Purpose: Isolated organ models are widely used in NMR experiments, e.g. the isolated heart to test cardioprotective techniques [1], which must be rigorously validated before introduction into the clinic [2]. A perfused heart suspended in a moving stream of heated room air and subjected to ischaemia rapidly cools to room temperature due to wind chill. If submerged in a thermostatically controlled bath, temperature is maintained, but a bath of electrolyte can reduce the NMR signal strength or affect shimming. Methods: Working rat hearts perfused with a phosphate-free erythrocyte suspension were subjected to ischaemia while suspended in air. Tissue phosphates were monitored using 31p NMR spectroscopy. A warm air system and a new humidified system (Fig. 2) were compared, air temperature being maintained at normothermia, while heart and air temperatures were measured with Luxtron ® fibre optic probes. Results: In heated room air at 37°C, the hearts cooled almost to room temperature after 10 min ischaemia (Fig. 1). After 20 min, phosphocreatine content dropped to less than 25% of control, and recovery of left ventricular output was 95%. With air at 100% relative humidity, heart temperature remained at 38°C (true normothermia in rat), and the same degree of damage occurred after only 10 rain ischaemia. Discussion: The effect of changing temperature on different metabolic systems is different and poor temperature control can cause misleading results.

References [1] K.-H. Hiller et al., ESMRMB (abstract) 1997, p. 91 [2] W.R. Bauer et al., Circulation 92, 1995, 968 977 [3] K.-H. Hiller et al., JMCC 28, 1996, 571-577

References [1] Houston RJF, Heerschap A, Skotnicki SH, Verheugt FWA, Oeseburg B (1997). ATP content measured by 31p NMR correlates with post ischaemic recovery of external work output in an isolated rat heart perfused with erythrocyte suspension. J Mol Cell Cardiol 29:1763 1766. [2] Buckberg G D (1994). Normothermic blood cardioplegia. Alternative or adjunct? J Thorac Cardiovasc Surg 107:860-867

Heart in air

1o rainischaemia l

0 "o

20

/ 5

,

I 10

,

......................... ,

Tune~ln) Fig. 1. Effect of ischaemia on heart temperature.

38"c ® temp.sensor

>38°C ventilator humidifier

NMR probe

heater ]element RH -

compressed air

temp ,,

1OO% temp 38"C

,

20

15

I heater t element Fig. 2. Air Temperature control system.

1 "°

/ 20 25

196

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191 207

Beneficial effects of the p-receptor-blocker bisoprolol and the angiotensin-converting enzyme inhibitor captopril on cardiac energy metabolism in rats post-myocardial infarction S. Hiigel, M. Horn, M. de Groot, K. Hu*, H. Remkes, C. Dienesch, G. Ertl*, S. Neubauer Medizinische Universit/itsklinik Wiirzburg, Germany *II. Medizinische Klinik, Mannheim, Germany

of residual intact myocardium was analyzed in terms of total and isoenzyme creatine kinase activity, steady state levels (ATP, phosphocreatine) and turnover rates (creatine kinase reaction velocity) of high-energy phosphates (31p-NMR) and total creatine content (HPLC). Both bisoprolol and captopril prevented post-MI hypertrophy and partially prevented left ventricular contractile dysfunction (maximum left ventricular developed pressure 185 + 11, 133 + 9", 223 -I- 14:~, 160 _+ 10t" in sham, MI, MI + Bisoprolol and MI + Captopril hearts) (Fig. 1). Residual intact failing myocardium showed a 25% decrease of total, a 26% decrease of MM- and a 37% decrease of mito-CK activity. Total creatine was reduced by 15%, phosphocreatine by 21% and CK reaction velocity by 41%. Fig. 2 shows representative 31p-NMR spectra of all six groups. Parallel to improved function treatment with bisoprolol or captopril largely prevented all of these changes in infarcted hearts. Conclusions: Thus, the favourable functional effects of 13-receptor blockers and ACE inhibitors post-myocardial infarction are accompanied by substantial beneficial effects on cardiac energy metabolism. Improvement of cardiac energetics may be one mechanism of action of chronic treatment with these compounds in heart failure.

Background: Chronic treatment with B-receptor blockers or angiotensin-converting enzyme (ACE) inhibitors in heart failure can reduce mortality and improve left ventricular function, but the mechanisms involved in the beneficial action of these compounds remain to be fully defined. Our hypothesis was that [3-receptor blockers and ACE inhibitors prevent the derangement of cardiac energy metabolism in parallel with their favourable functional effects. Methods and results: Rats were subjected to left coronary artery (MI) or sham-operation. After surgery, animals were treated with bisoprolol (60 mg/kg/day), captopril (2 g/L drinking water) or remained untreated. Two months later, cardiac function was measured by a left ventricular balloon (pressure-volume curves), and energy metabolism 250 200

----0---

E E

150

.~

MI

•- . - o . - -

100 > 50

!

I

I

0.20 0.25

I

0.30 0.35

I

0.40

I

I

sham

sham+Capto

:

MI+Capto

,~

sham+Biso

,t

MI+Biso

I

0.45 0.50 0.60 Left ventricular v o l u m e (ml)

Fig. 1. Pressure-volume curves of the six experimental groups.

Sham

PCr

MI

ATP

Untreated

Captopril

Bisoprolol

10

5

0

-5

-10

-15

-20ppm

10

5

0

-5

-10

Fig. 2. Representative 31p-NMR spectra of the six experimental groups.

-15

-20ppm

Abstracts~Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191-207

Joubert, F., Mateo, P. Gillet, B. Beloeil, J.C., Hoerter J.A. U-446 INSERM, Univ. Paris-Sud, 92296-Chatenay Malabry, France & ICSN, CNRS, 91 Gif S/Yvette, France Despite considerable efforts devoted in the last fifteen years to the study of the creatine kinase (CK) flux, there is still some controversy about the equivalence of the various magnetization transfer techniques and about the physiological interpretation of the N M R measured CK fluxes. Renewed interest in the determination of CK flux has come from the development of transgenic mice with specific CK isoforms knockout. If saturation of ATP gives consistently straightforward measurements of the forward CK flux PCr---, ATP (Ff), a large range of values of ATP ~ PCr (reverse flux Fr) are found in the literature. Discrepancies observed between Ff and Fr measured by saturation transfer have been suggested to be due to the existence of other reactions involving ATP: the 2 sites exchange model being an oversimplification of the system [1,2], or to the existence of ATP compartmentation [3]. The existence of an ATP NMR-visible compartment which is not in exchange with PCr has been suggested [4] while the postulate of an exchange between PCr and a non-visible non-saturated ATP compartment has led to the modeling of mitochondrial CK flux [5] Inversion transfer experiments has also been widely used. A simplified analysis of the initial slope allows to determine one flux, however the complete time analysis of the evolution of the magnetization of the inverted and non inverted species allows the determination of all parameters of interest T1ATp, Tlpcr, kf and kr in the same experiment. Furthermore this analysis of the inversion transfer data can be performed without imposing CK equilibrium of the system while this simplification of the Bloch equation is usually applied in the analysis of saturation transfer. We thus compared the Ff and Fr flux obtained by inversion of ATP (inv-ATP) with those found by inversion of PCr(inv-PCr) in the isovolumic heart perfused with acetate. The analysis was performed without imposing CK equilibrium and compared with the kinetics of F f measured by time dependent saturation transfer(TDST) of ATP. Table shows CKflux in isovolumic perfused heart (Flux are expressed in MA.s 1; MA = magnetization area of signal). Type of n experiment

kf (s-Z)

fluxF

Saturation ATP

0.63

1.13

+ 0.06

+ 0.09

7- 4

3.

•

kr (s ])

inversion forward Flux

Fluxr

0.55

1.14

1.87

2.14

0.58 + 0.07

Inversion 6

+0.I1 1.01

-I-0.17 1.78

+0.26 1.27

_+0.24 1.16

1.54 + 0.23

3'-ATP

+ 0.13

___0.27

-+ 0.07

-+ 0.06

Inversion 7 PCr

Can inversion transfer experiments evidence a fraction of ATP which does not participate in the myocardial ereatine kinase flux?

Such an analysis performed on CK in vitro revealed no difference between the flux measured by inversion transfer and TDST; the forward and reverse flux are equal. Since the heart is in steady state and NMR measures an average of CK flux over time and space (the whole cell and organ), Ff should also be similar to Fr in viv0. This was clearly not observed: Ff/Fr were markedly different from 1 in inv-ATP and inv-PCr protocols. We examine the hypothesis that the underestimation of Ff/Fr in the inv-PCr and its overestimation in the inv-ATP protocols results from the existence of a fraction ATP visible in N M R but not involved in the CK reaction (as earlier suggested 6). Classically in an inversion transfer the solutions of McConnell equations are given by: Mvcr(t) = Mp°cr+ C] eMt+ C2exit, with C ] , C2, C3, C4 = f ( k i , Tli, M ~ , Mi°): M A T p ( t ) = M ~ T v + C31 e x : + C3exzt, with Xl,2 =f(ki, Tli). If a fraction of ATP, ATP2 = x.ATP does not participate in the exchange with PCr, errors are made on the determination of Fr (due to error in MArp), but also on the determination of Ff if this fraction is magnetically perturbated. The error can be predicted by a modification of the McConnell equation taking into account x.ATP. The magnetization of PCr is not affected by this fraction but the magnetization of ATP is now described by M~Tp(t) = MATp(t) + x.M~T P in the inv-Per protocol and M~rl,(t ) = MAxp(t) + Cse - t / x) in the inv-PCr protocol This TIATP2 where Cs =f(MAxi, , z, MATp2, o model suggest that a fraction of about 40% of the total ATP which does not participate in the CK reaction could explain our experimental data in the perfused heart (Fig. 1). In conclusion the inversion of PCr protocol allows a correct measurement of Ff and a direct evaluation of ATP fraction by comparing Ff and Fr. We will further check this approach in various perfusion conditions designed to change the size of the ATP compartment. References [1] Ugurbil, J. Mag. Res., 64: 207,1985. [2] Brindle, Prog. NMR Spear., 20,257,1988. [3] Nunnally & Hollis, Biochemistry, 18,3642,1979. [4] Koretsky & al., Magn. Res. Med, 2:586,1985. [5] Zahler & al. Biophys. J. 51,883,1987. [6] Meyer & al., Am. J. Physiol., 242, C1,1982.

Ff/Fr

inversion

ATP

2

parameters:

PCr

kf=0.6s -I

reverse Flux -t

'

•

2

• O

• in M~s

•

•

•

•

1

•

II

•

i 20

i 30

i 40

i 50

0

i 0

i 10

i 20

i 30

! 40

! 50

O

O

O

"

O

2

! 10

! 20

3'0

' 40

50'

0

= 10

i 20

! 30

i 40

i 50

•

•

•

Flux 0

197

T1pcr=3S T1ATp=T1Aann=0.6s la--- adjusted to obtained a constant

!

i

0

lO

1

Fr=Ff=-l.11MA.s -1 hypothesis: all ATP is visible all ATP is inversed

FflFr 0 0

0

% of ATP not involved in CK reaction

Fig. 1. Parameters: k f = 0.6 s - 1 ; T l e c r = 3 s; T1ATP = T1ATP2 = 0.6 S; kr adjusyed to obtain a constant; Fr = F f = 1.I 1 M A . s - ]. Hypothesis: all ATP is visible; alll ATP is inversed.

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191-207

198

Cardiac energetics measured by 31p magnetic resonance spectToscopy is abnormal in Friedreich's ataxia.

correction to the ATP signal for blood contamination (as determined from the 2,3-DPG resonance). Left ventricular dimensions were measured from echocardiograms using a HP5500 system and fractional shortening (FS) was calculated. Results: The figure shows a spectrum from a patient and a control. In FA patients the mean PCr/ATP was 1.16 (0.37) (range = 0.64-1.81) compared to controls in whom it was 1.81 (0.08) (range = 1.67-1.93). This difference was highly significant (p = 0.0001). Ten (77%) of the patients had PCr/ATP ratio < 1.4. There was no significant correlation between LV hypertrophy or LV dimension and the PCr/ATP ratios although there was a trend for FS to decrease as PCr/ATP ratio fell [table, mean(sd)] (Fig. 1).

R. Lodi, A.M. Blamire, C.H. Davies, J.M. Cooper, P. Styles, A.H.V. Schapira, B. Rajagopalan. MRC-MRS Unit, Dept. Cardiovascular Medicine, John Radcliffe Hospital, Oxford, UK Dept. Clinical Neuroscience, Royal Free Hospital, London, U K

Purpose; The genetic abnormality in Friedreich's ataxia (FA) is known to lead to a severe reduction or absence of a mitochondrial protein, Frataxin. Both hypertrophy and dilatation have been described in the hearts of patients with FA. It has been shown in a yeast model of the disease that the deficit of Frataxin leads to a profound deficit in mitochondrial respiration. We hypothesised that this mitochondrial abnormality may be an important factor in the development of cardiac abnormalities. We used in vivo 31p spectroscopy to investigate cardiac bioenergetics in these patients. Methods: Thirteen patients with a clinical diagnosis of FA (mean age 33) and 9 controls (mean age 31) were studied in a 2 T Oxford magnet interfaced to a Bruker Avarice spectrometer. Patients lay prone in the magnet and standard spin-echo MRI was used to position the heart in the centre of the magnet. Cardiac 31p spectra were acquired using a 7 cm circular surface coil placed below the chest. Data were acquired following localised shimming using a slice selective, 1D spectroscopic imaging sequence. A single oblique outer volume suppression slab was position using the MR images and was used to presaturate 31p signal from the lateral chest wall to improve the localisation. Spatial resolution was set to 1 cm and encoding was performed in the anterior-posterior direction. Spectroscopic imaging rows corresponding to the heart were identified from the MR images and extracted from the data set. Data were analysed using a purpose written interactive frequency domain fitting program. Spectral fitting ineluded adjustments for mixed Gaussian-Lorentzian lineshape and missing data points from the beginning of the FID (due to the phase encode duration). PCr to ATP ratios were calculated including a

<1

1 1.4

> 1.4

FS-% LV mass-g

32.3 (9.8) 232 (52)

37.9 (13 9) 212 (71)

40.8 (6.6) 259 (57)

Discussion: In previous patients with mitral valve disease FS was correlated with PCr/ATP. In other cardiac diseases this correlation has been less strong but in general low PCr/ATP were associated with poor mechanical function. It was interesting that in patients with FA quite marked decreases in PCr/ATP ratio were not associated with gross functional changes or hypertrophy on the echocardiogram. This raises the possibility that the metabolic abnormality precedes the dilatation and hypertrophy that have been seen in other studies. It is possible that these studies included patients in whom the disease had progressed much further. It is also possible that the metabolic changes are related to the size of the GAA triplet expansion on chromosome 9 which is thought to the cause the phenotype in this disease.

PCr/ATP ratio in <1 FA FS- % 32.3 (9.8)

232 (52)

LV mass - g

PCr/ATP ratio in FA

I - 1.4

> 1.4

37.9 (13.9)

40.8 (6.6)

212 (71)

259 (57)

Control

FA patient

PCr/ATP = 1.93

PCr/ATP = 0.64

2-3 DPG

7

. . . .

I

. . . .

10

I

5

. . . .

I

0

. . . .

I

-5

. . . .

I

. . . .

-10

I

. . . .

-15

I

.

.

.

.

.

ppm

.

.

.

I

. . . .

10

Fig. 1. Spectra.

. . . .

!

0

. . . .

i

-5

. . . .

I

. . . .

-10

I ' "

-15

I

. . . .

ppm

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191-207 Cardiac multi-phase k-space segmented echo planar imaging with black blood preparation for fast velocity mapping of myocardial motion M. Markl, J. Hennig Department of Diagnostic Radiology, University of Freiburg, D 72106 Freiburn Germany Introduction: Cine or multi-phase cardiac imaging [1] offers a variety of applications such as assessment of the global cardiac function, measurement of the heart wall motion with phase contrast [2,3] or tagging techniques [4,5] as well as determination of flow inside the ventricular system. Here, we present a method based on interleaved Echo Planar Imaging (EPI) with multiple RF excitations per phase of the heart movement. Methods: All experiments were performed on a 1.5 T scanner (Magnetom Vision, Siemens, Germany) with a standard gradient system (25 mT/m and 600 ~ts risetime) using a 4-segmented phased array coil. Images (slice thickness = 8 ram, 250 mm FOV) were acquired in short axis view. The pulse sequence consisted of a blipped multi-shot EPI scan with linear ramping of the readout gradients and an echo train length of 8 echoes per excitation (flip angle ~ = 25 °, effective TE = 10 ms, TR = 23 ms, bandwidth = 780 Hz/pixel). Phase encoding gradients were rewinded for each excitation. An even/odd phase correction was performed using a calibration scan acquired under reversed switching gradients. In order to achieve short overall scan times data was read out with 3 shots per phase. These parameters allow multi-phase images to be acquired in 6 heart beats during breath-hold. The measurement during the first ECG cycle serves as a dummy scan including acquisition of 2 lines of data per time frame for even-odd phase correction. In order to reduce image artifacts resulting from

199

blood-flow inside the ventricles we have used a spatially selective preparation pulse which nulls all signal from blood directly outside the observed slice. This black blood pulse is applied before every RF pulse in the sequence timing table. Fat saturation using 4 rectangular pulses was performed for every phase (Fig. 1). Results: Fig. 2 displays the result of a multi-phase measurement of a healthy volunteer's heart in short axis view. The temporal resolution of the venticular motion was 77.5 ms. Acquisition time for 5 slices including positioning is about 10-15 minutes. Discussion: We demonstrated a high speed imaging method to acquire cine multi-shot EPI images using a, standard MR scanner. For measurements of myocardial motion with the phase contrast technique a bipolar gradient can be added after each RF-pulse to the otherwise identical sequences in any of the three dimensions. It is thus possible to obtain full in plane velocity information of the moving heart with only one breath-hold measurement over 16 heart beats (dummy scan and acquisition of phase correction data need to be carried out only once). References [1] Atkinson D.J., Edelman R.R., Radiology 178,357:360, 1991 [2] Pele L.R., Sayre J., Yun K., Castro L.J., Herfkens R.J., Miller D.C., Pelc N.J., lnvst. Radiol. 29,1038, 1994 [3] Hennig J., Markl L., Peschl S., Schmialek A., Schneider B., Krause T., Laubenberger J., Proc. Vth Meeting ISMRM, Vancouver, 390, 1997 [4] Buchalter M.B., Weiss J.L., Rogers W.J., Zerhouni E.A., Weisfeldt M.L., Beyar R., Shapiro E., Circulation, 81, 1236, 1990 [5] Young A.A., Imai H., Chang C., Axel L., Circulation, 89, 740, 1994

a) ~. ECG :.,'.-,i,,.-.-,I . . . . . .

~

I,,..-,i~,,....,t . . . . . .

/ 5g

IIFI

RF II

RF I11

n[-~ u

AAAAAAA

n

G$

u

~AA~A

AA

n Gp

b) !1

k-space, 8 segments:

m3(5) ,~

I

1

Iz

t

I

i5

!, !7 is

I

J

J

1,4(1) 114(1) IH4(I) I4(2)

I~ ~', •

lX4a)

~i

1114(2)

•,

I4(s)

~,

114(5) 1114(5)

I,, b~

L~l)

4

kx Fig. 1. Sequence design and measurement timing, a) ECG gated multi-phase data acquisition and pulse sequencefor each phase of the heart motion. B)Ordering of recieved lines of data in k-space.

200

Abstracts / Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191-207

Fig. 2. Black blood segmented EPI images of 8 phases of ECG cycle of a healthy volunteers heart. The cardiac cycle starts at the top left and ends at the bottom right.

The use of in vivo 31p MRS for evaluation of cardiac energy status in the transgenic mice overexpressing growth hormone E. Omerovic, E. Bollano, B. Madhu, M. Bohlooly, J. T6rnell, F. Waagstein, J. Isgaard and B. Soussi Wallenberg and Lundberg laboratories, Sahlgrenska University Hospital, Dept. of Physiology, Research Center for Endocrinology and Metabolism, University of G6teborg, Sweden Introduction: Recent advances in transgenic mice technology make the mouse a particularly interesting small-animal species in cardiovascular research. A mice-model for in vivo noninvasive 31p MRS would be useful for studies of cardiac energy metabolism in different transgenic models of altered cardiac physiology. Disturbed myocardial energy metabolism have been proposed to be contributory factor responsible for compromised cardiac function in the settings of different diseases of the myocardium. The aims of the study were: to develop mice-model for in vivo non-invasive assessment of myocardial bioenergetics with volume-selective 31p MRS; to examine effects of chronically elevated plasma levels of growth hormone on cardiac energy metabolism. Subjects & methods: Normal mice (BW = 20-40 g, n = 6) and transgenic mice overexpressing bovine growth hormone (bGH) (BW = 50 56 g; n = 3) were used. Cardiac gated MR imaging and spectroscopy were performed on a 2.35 T horizontal magnet with a 20 cm bore (Bruker Biospec 24/30). A double tuned (1H & 3]p) Helmholtz coils of 2 and 3 cm were used for transmission and reception of RF. Image Selected In vivo Spectroscopy (ISIS) was used as localization method. The volume of interest was 8 x 4 x 4 mm (128 gL) and included whole left ventricle. Acquisition parameters were 4096

normal wansgmic

scans with RT = 2,5 s giving total scan time ca 3 h. The animals were maintained anaesthetized by continued gas anesthesia with isoflurane. Body temperature was kept constant at 37 __+0,5°C using specially adapted homeothermic blanket system (Harvard apparatus). The heart rate was maintained within the range 450-570 bpm The mice were investigated with transthoracal echocardiography for the assessment of cardiac function. Results: All animals tolerated well the examination procedure. The PCr/ATP values corrected for partial saturation and echocardiography data are summarized in the table. Markedly lower PCr/ATP ratio and left ventricular fractional shortening were observed in the transgenic mice. An interesting finding was the high phosphodiester (PDE) signal in the transgenic mice spectra (Fig. 1).

Normal Transgenic

PCr/ATP

FS

2,0 + 0,3 1,3 _+ 0,1

36,9 _+6,0 25,4 _+ 3,0

In the above table FS = left venbicular fractional shortening. Conclusions: It is feasible to estimate phosphorous metabolises levels of the mouse heart in vivo noninvasively with 3~p ISIS under maintained physiological conditions. Chronic exposure of the myocardium to increased plasma growth hormone may have adverse effect on cardiac energy metabolism and left ventricular function. This model could be useful for repeated evaluation of cardiac energy status as well as effects of pharmacological treatments during the progress of myocardial disease in different mice models. PCr/ATP

FS

2,0 + 0,3

36,9 ± 6,0 2 5 , 4 . + 3,0

1,3 ± 0,1

FS = Ic~ vcumcular ~a~onal shortcning PCr

PCr 2,3 DPG + Pi

ATP 2,3 DPG ~ i

n p spectrum from the normal mouse

31p spectrum ~om the mouse ovez~,xpx~miugbGH Fig. 1. Spectra.

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191 207

Acquisition weighted chemical shift imaging Rolf Pohmann and Markus yon Kienlin Department of Biophysics, University of W/irzburg, Germany Introduction: Chemical shift imaging (CSI) already plays an important part in observing in vivo the biochemical processes in studies on animals and humans. The precision of the quantification, however, is strongly reduced by the shape of the pointspread function (PSF) of such an experiment, which mainly affects measurements with a small spatial resolution on less sensitive nuclei like phosphorus. A considerable improvement of the PSF can be achieved by different averaging of the phase encode steps. By choosing an optimum number of averages for each phase encode step, the shape of the PSF can be improved without loss in signal-to-noise ratio or spatial resolution [1]. We implemented acquisition weighted CSI on a whole-body scanner to improve the quality of phosphorus spectroscopy on humans. Theory: The shape of the PSF in normal CSI is due to the limited region in k-space that is sampled: The PSF is the Fourier-transform of the sampling window and for a rectangular window has the shape of a sinc-Funktion (Fig. 1): {PSF(x)= [sin nN(x--xo)/FOV]/[n(xx0)], N being the number of phase encode steps and FOV the field of view of the experiment. The wiggles at the side lobes of this function lead to strong contamination from adjacent voxels. They can be reduced by applying a stronger weighting in the center of k-space than at the corners. The optimum averaging scheme has been shown [1] to be the Hanning-function: {w(k) - 0.5(1 + cos ((2nFOV)/(N)). Itis common practice in imaging to apply a Harming filter on the acquired data to reduce ringing. In CSI-experiments this would cause an unacceptable broadening of the PSF and thus a strong loss in spatial resolution. The weighting can, however, already be applied during the acquisition by adding a different number of averages for

201

different positions in k-space: For phase encode steps at the corners of k-space, a smaller number of averages are performed than in the center. The increased linewidth can be prevented by scanning a larger part of k-space. Then neither signal-to-noise ratio nor resolution are affected, but the signal contamination between voxels is considerably reduced. Results: Acquisition-weighted CSI sequences were implemented on a Bruker 2 T whole body scanner and a AMX 11.7 T microscopy system. For the comparison of the results of this technique to those of normal CSI, for both cases a two dimensional experiment was performed with a total of 256 repetitions each. For the CSI measurement, a matrix size of 8 × 8 phase encode steps was averaged four times. An equal width of the PSF was obtained by an acquisition weighted experiment 12 x 12 phase encode steps. The number of averages was detm~nined by the Harming function and varied from 6 at the center of k-space to 0 at the corners. In Fig. 1, the spatial dependence of the signal of a point source is plotted. The improvement of the PSF by acquisition weighting is clearly visible. Fig. 2 shows the results for a simple phantom. While the fat signal from the voxel on the right is considerably contaminated by the water peak in the CSI spectrum, it is almost totally suppressed in the corresponding acquisition-weighted spectrum, while the signal-tonoise ratio is similar in both experiments. Conclusion: For experiments where a low signal-to-noise ratio demands a low spatial resolution and averaging, acquisition weighted CSI is able to considerably reduce contamination without affecting the sensitivity or the spatial resolution of the experiment. Implementation of acquisitionweighted schemes on whole body instruments will strongly improve the quality of phosphorus in vivo spectroscopic imaging. Reference [1] Mareci, T.H., Brooker, H.R., J. Magn. Resort. 92, 229, 1991

Fig. 1. Measured PSF withuot (left) and With acquisition weighting, demonstrating substantially reduced contamination between adjacent voxels.

fat

B; ~ ~

Fig. 2. Spectra from a water/fat phantom, acquired with normal (A) and acquisition weighted (B) CSI.

202

Abstracts /Magnetie Resonance Materials in Physics, Biology and Medicine 6 (1998) 191-207

In Vivo imaging of coronary arteries of the mouse J. RufJ~ F. Wiesmann, K.-H. Hiller, E. Rommel, S. Neubauerl, A. Haase Physikalisches Inst., Universit/it Wtirzburg, 1Med. Universit~itsklinik, Wfirzburg, Germany Purpose: In the recent years, the mouse has been established as model for cardiovascular pathophysiological research. N M R investigations include anatomical and functional cardiac studies. The purpose of this work was to optimize three-dimensional (3D) N M R angiography of the coronary arteries of the mouse heart in vivo. The challenges of this study were, the fast motion of the heart (up to 600 beats per minute), the microscopic dimensions of the coronary vessels (diameter about 1501am) and in addition the low flow within these vessels. Methods: 3D images were acquired with an echocardiogram (ECG)-triggered segmented F L A S H sequence (8 phase-steps per R-R-interval). Images were acquired with a repetition time of 4.3 ms and an echo time of 1.6 ms. With zerofilling in the third dimension an isotropic resolution of 100 ~tm was achieved. Four averages leaded to a total acquisition time of half an hour. N M R angiography was obtained due to the inflow contrast in fast FLASH sequences. Five 3D slabs were acquired to image the complete heart. Male C57bl/6 mice with body mass of 20 g were explored on a 7-T-Bruker BIOSPEC spectrometer with a microscopy gradient system (@ 70 mm; maximum field strength 870

roT/m) and a birdcage RF-coii. During the experiment, the mouse was anesthetized with 1.5% Isoflurane/1 L 02. The maximum of signal intensity within coronary arteries in relation to heart phase was examined with a 2D ECG-triggered cine gradient echo sequence. Results: For the first time coronary vessels of the in vivo mouse heart could be visualized. 2D cine studies showed, that the maximum signal intensity within the coronary arteries was achieved from end diastole to mid systole (maximum of flow velocity).This result leaded us to trigger the segmented 3D-FLASH sequence to late diastole. Fig. 1 shows transversal planes of the 3D-data. The coronary arteries were clearly visible from the base of the heart up to midventricular slices. With MIP processing, the reconstructed vessel tree can be visualized from different projection angles. A surface reconstruction of the coronary arteries after hysteresis thresholding and segmentation showed many branches of the vessel tree. Discussion: In spite of the microscopic dimensions of coronary arteries, 3D gradient-echo N M R imaging revealed the feasibility o f N M R angiography for in vivo mouse studies. The resolution of one or two pixels per vessel-diameter is sufficient to visualize coronary arteries. To do vaso-dynamic studies, the resolution has to be further increased. With optimized surface coils, either an enhancement of spatial resolution, or a minimization of acquisition time can be realized and N M R angiography will become a standard procedure to the study of gene knock-out mouse models.

Fig. 1. Transversal slice a) base of the heart with left coronary artery (LCA) and aortic root b) midventricular with smalles coronary arteries. Magnetic resonance imaging reliably detects the area of left ventricular outflow tract in patients with hypertrophic obstructive cardiomyopathy as a parameter to characterize clinical outcome

J. Schulz-Menger, M.G. Friedrich, O. Strohm, R. Dietz Franz-Volhard-Klinik Charite Berlin,, Campus Buch Max Delbriick Centrum fiir Molekulare Medizin, Germany Purpose:The degree of left ventricular outflow tract (LVOT) obstruction is essential in the evaluation of hypertrophic obstructive cardiomyopathy (HOCM) and in the follow up investigation after interventional ablation of a septal artery. Interventional ablation of a septal artery is a new therapeutic approach in patients with hypertrophic obstructive cardiomyopathy. Catheter treatment is performed for reduction of septal myocardium in order to reduce left ventricular outflow tract (LVOT) obstruction. But the pressure gradient is susceptible to hemodynamic variations and the correlation to the clinical outcome is poor. Furthermore, the echocardiographic followup of diameter and LVOT pressure gradient strongly depends on the 'ultrasound window' and on the investigator's experience. The area of LVOT as the key parameter could only be estimated. Magnetic resonance imaging (MR1) enables the accurate visualization of the myocardium and of the blood flow with free choice of slice orientation. Method: We investigated 15 patients with a conventional MRI scanner

(Siemens Expert/l.0 T; Siemens AG, Erlangen, Germany). Using gradient-echo sequences we directly measured minimal orifice area of LVOT during naaximal blood flow in systole. Additionally, we quantified myocardial muscle mass, ventricular volumes and endsystolic wall stress. Furthermore we performed MRI before embolisation and at day 3,7,14 and 28 after the intervention in 4 patients. In these patients we could detect after application of contrast medium (0.1 mg/kg GadoliniumDTPA) the myocardial infarction and its relation to the septum by Tl-weighted spinecho-sequences. There were no problems during investigation. Mean acquisition time was 30 minutes. Results: We found a very close relationship ( R = 0.8, p <0.0001) between LVOT orifice area and the NYHA class. Follow-up measurements after embolization of septal arteries reinforces the observation of a close relationship existing between change in symptoms and change in the LVOT area. Discussion: In conclusion, systolic LVOT orifice area seems to be a reliable parameter in quantifying the LVOT obstruction in HOCM. Magnetic resonance imaging reliably detects the area of left ventricular outflow tract in patients with hypertrophic obstructive cardiomyopathy as an parameter to characterize clinical outcome. MRI is able to reliably detect the functional and morphological changes in the follow up of patients after catheter treatment of hypertrophic obstructive cardiomyopathy.

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medic.me 6 (1998) i 9 I - 2 0 7 Decreased creatine kinase reaction velocity but unchanged high energy phosphate levels in chronic diabetic myocardium Matthias Spindler, Kurt W. Saupe, Rong Tian, *Saadia Ahmed, *M. Abdul Matlib, Joanne S. Ingwall Brigham & Women's Hospital, Harvard Medical School, Boston, MA, USA *Department of Pharmacology & Cell Biophysics, University of Cincinnati, OH, USA Several studies have demonstrated that creatine kinase (CK) activity is decreased in diabetic myocardium. To define whether decreased energy reserve via CK is a major factor leading to decreased contractile performance in diabetic hearts, we studied CK isoenzyme activity in vitro and CK reaction velocity in vivo in 36 isolated perfused rat hearts using 31p NMR spectroscopy. After 4 weeks of diabetes, total CK activity decreased by 22%. The decreases in CK-MI3 ( 54%) and CK-mito (--35%) were greater than for CK-MM ( - 1 1 % ) . There were no further changes after 6

Noninvasive quantification of the aortic valve area in patients with calcified aortic valve stenosis by magnetic resonance imaging O. Strohm, J. Schulz-Menger, F. Uhlich, M.G. Friedrich, R. Dietz. Charite, Medizinische Fakultat der Humboldt-Universitat, FranzVolhard-Klinik am MaxDelbriick-Centrum, Berlin., Germany

Purpose: The quantification of the residual aortic valve area is of great importance in evaluation of the hemodynamic relevance of aortic stenosis. Routinely used techniques include transthoracic (TTE), multiplanar transesophageal echocardiography (TEE), and cardiac catheterization (CATH). Both, TTE and CATH compute the value using measured parameters like flow or stroke volume and empiric multiplication factors. These parameters suffer from a high intraindividual variability. Furthermore, accompanied insufficiency of the aortic or mitral valve impairs their applicability. TEE is an invasive method, which is highly dependent on the experience of the investigator. Magnetic resonance imaging (MRI) is capable of visualizing blood flow nonivasively and reproducibly, thus making MRI a valuable tool in diagnosis and follow-up of cardiac patients. We sought to determine the diagnostic value of MR1 in measuring the aortic valvular orifice area in patients with anglographically proven calcified aortic valve stenosis. Method~: 1. Flow model Using a conventional MRI scanner (Siemens Expert/1.0 T; Siemens AG, Erlangen, Germany) we developed an easy-to-perform subtraction technique in gradient-echo-techniques to measure the residual area of the blood flow crossing the assessed plane. In a custom-designed, constant-flow model of physiologic-size aortic stenosis with different degrees of stenosis, ranging from 0.5 cm2 to 2.0 cm2 we performed gradient-echo measurements (TE = 6.1 msec, TR - 300 msec, FOV = 250 x 250 ram) at different levels below and above the stenotic area. The true opening area of the model was measured by scanning the model and measuring the translucent area; using the CATH-formula, the opening area was calculated

Estimation of 02 consumption in frozen heart tissue using 1~C NMR H.G.J van Mil and J.H.G.M. van Beck Laboratory of Physiology, Institute of Cardiovascular Research, Free University, Van der Boechorststraat 7, 1081 BT Amsterdam, The Netherlands.

Purpose: Heterogeneity of the myocardial perfusion may cause aerobic metabolic heterogeneity under normal conditions [1]. Such heterogeneity may increase under pathological conditions [2]. To assess in more detail the transmural 02 consmnption (VO2) profile, a method

203

weeks of diabetes. Cardiac performance of diabetic hearts, estimated as ratepressure product (left ventricular developed pressure times heart rate) under identical perfusion conditions, was only 50% that of control. ATP, phosphocreatine (PCr) and total creatine concentration were the same in control and 4 or 6 weeks diabetic hearts. However, after 4 weeks of diabetes, in vivo CK reaction velocity (measured by 3~p magnetization transfer) decreased 22%, in proportion to the decline of the total CK activity, The further decline of the CK reaction velocity after 6 weeks of diabetes ( 45%) was greater than predicted from the CK rate equation. This unexpected result suggests that CK enzyme regulation occurs in addition to substrate control. During increased workload, induced by increasing perfusate calcium concentration, cardiac performance more than doubled in control and diabetic hearts and in vivo CK reaction velocity increased by 16%, thereby maintaining a constant ATP concentration in both groups. Thus, we conclude that in chronic diabetic myocardium the CK reaction velocity decreases but does not limit the availability of high energy phosphates to support contractile performance over the range of workloads studied.

using the pressure gradient over the model and the 'blood'-flow in the model. 2. Clinical study. CATH and TTE were done in the standard approach. We studied 22 consecutive patients with calcified aortic valve stenosis and no or only small valve insufficiency diagnosed by TTE. Mean examination time was 20 minutes. The values for the orifice area were then correlated to the results of the CATH procedure, which was performed within 5 days. Results: The values of the orifice area measured by MRI in the model correlated excellent to the true opening area (ra= 0.99, correlation equation: areatrue=0.1 +0.95 x areaMR) and to the calculated opening area (r2= 1.0, correlation equation: a r e a ~ l ¢ = - 0 05+0.89 x areaMR) All patient tolerated the MRl-examination well and without complications. Image quality was sufficient in all subjects to allow evaluation. The mean difference to the area computed by CATH was 0.103 cm2 with higher values obtained by the MRI technique. The correlation to CATH was 0.892 (areaMR = 0.27+ 1.54x areaOATH; R2=0.80; p<0.0001; 95% confidence interval! 0.7540.955). Discussion: Flow-sensitive sequences in MRI allow to evaluate not only flow in the investigated plane, but also the exact determination of a stenotic area without the need for geometrical or physical assumptions about the relationship between hemodynamic values and pressure gradients. We found a very close correlation between the MRl-opening area and the calculated and true opening area. In the patient study there was a trend towards higher values measured by MR1 in orifice areas >_ 1.0 cm2. Whether this is due to underestimation of the orifice area by CATH remains to be assessed. However, in the range of severe stenoses the difference was very small. So, the therapeutic decision would not have been different. We conclude, that MR1 has the possibility to measure the true opening area in patients with aortic stenosis and could be a valuable tool to follow-up patients non-invasively. MR[ seems to be a feasible method to noninvasively determine the valve area in patients with calcified aortic stenosis.

has been devised and tested using 13C NMR [3,4,5,6]. This ~3C methodology, pioneered by Chance et al [7], analyzes glutamate 13C NMR multiplets with a kinetic model to estimate TCA cycle parameters. Our method however, differs from [7,8,9] in that it only analyzes o~e point in time. Here we investigate the sensitivity and properties of this single-time point ~C NMR method using computer simulation methods. Methods: We used our ~3C distribution model to simulate previous experiments. In these experiments, a rabbit heart was exposed to a ~3C acetate double labeling protocol, 4 rain. [2-13C] acetate and 1.5 min 1,2-~3C] acetate [4]. The tissue was quickly frozen and freeze

204

Abstracts~Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191 207

dried. 13C Glutamate fine multiplet profiles were obtained from small samples of the left ventricular (LV) free wall, using high resolution ~3C NMR. A kinetic model of 132 coupled differential equations was then utilized to estimate critical model parameters values. VO 2 was calculated from the flux through the TCA cycle (JTcA) and the enrichment of Acetyl CoA (Fc) as VO2 = (2-Fc) JT(2A" Due to the relative short labeling time of 5.5 min, it was necessary to account for the delay (ttransport) of label incorporation into the acetyl CoA pool after injection into the blood vessel of the ~3C acetate, which is in the order of 20 to 40 sec. We investigated the accuracy of the methodology when using the multiplet fine structure of a single time point of a pre-steady-state double labeling experiment. Therefore, eight multiplets are calculated using the isotopomers obtained from a double labeling experiment simulation with our model. To these multiplets we added realistic NMR noise, comparable to our own experiments, to obtain 20 groups of 8 multiplets. Results: To investigate the effect of ttransport, we used the data of [8] to compare the parameter estimates obtained by our model. Comparable parameter estimates, as found by Yu et al. [8], are obtained for a ttransport of 0 seconds. However, if the weight of the first three time points was increased, realistic values for ttransport must be utilized to obtain comparable results. We estimated the parameters using the 20 set of multiplets and compared these with the original parameter values of the model. Good estimations, that are not significantly different from the values used in the simulation, are obtained for JTCA (SD/JTcA = 0.10) and Fc (SD/F c = 0.06), whereas the size of glutamate pool and ttransport obtained larger SD/parameter ( > 0.2). Less fair estimations, significantly different from the simulation value with SD/parameter > 0.5, are found for transaminase flux between ~-ketoglutarate and glutamate (F-exchange). To investigate the sensitivity of the estimation of JTCA and Fc for F-exchange we estimated the parameters from the 20 sets of multiplets for different values of F-exchange. When F-exchange is larger than 4 gmol/g.dm/min, standard errors were not effected nor the mean value's of JTCA and Fc. The same was done for ttransport. For a good estimation of JTCA and Fc, ttransport must be in a time window of 10 100 see. Discussion: We used the pre-steady state-one point in time-muliplet fine structure of 13C labeled glutamate rather then the time evolution of the incorporation of the label into a specific carbon position. We show that, using this method, an estimation of JTCA and F c for the calculation of VO 2 is possible, although the estimation of the Glutamate pool

BOLD-MRI under pharmacological stress with dipyridamole: T~-measurements in 10 patients with coronary artery disease C M Wacker ~'2, A W Hartlep 1, M Bock 1, S Pfleger 2, G Ertl 2, G van Kaick 1, L R Schad 1, WR Bauer 2 Research program 'Radiological Diagnostics and Therapy', German Cancer Research Center (DKFZ) Heidelberg, Germany 2 1t. Med. Univ. Klinik Mannheim/Heidelberg, Germany

Purpose: It is well known that changes in the apparent transversal relaxation time T* are correlated with the oxygenation state of hemoglobin (BOLD effect). In a pilot study, global and regional changes in myocardial oxygenation were evaluated under pharmacological stress with dipyridamole (DIP) using T*-measurements. Methods: A segmented gradient echo pulse sequence was implemented on a 1.5 T whole body scanner (SIEMENS Vision) to assess myocardial oxygenation. T*-measurements were done at rest and under stress conditions with DIP (0.56-0.84 mg/kg). 20 healthy volunteers and 10 patients with coronary artery disease (CAD, no myocardial infarction) were examined. Patients underwent X-ray angiography, stressechocardiography and MR-exam within 5 days. In one patient, T~-measurements were repeated 10 weeks after performance of coronary angioplasty (PTCA, see Fig. 1). Results: In volunteers, admin-istration of DIP induced a 17% increase in T~ from 36 to 42 ms.

size and F-exchange are poor. It is shown that the estimation procedure is not sensitive to F-exchange over a wide range. The necessity of adjustment with ttransport is illustrated in the sensitivity analysis. Literature: [1] Beek JHGM van. Is local metabolism the basis of the fractal vascular structure in the heart? Int.J.Microcirc.:Clin.Exp. 17:337-345, 1997. [2] Bussemaker J, Groeneveld ABJ, Teerlink T, Hennekes M, Westerhof N, and Beck JHGM van,. Low and high blood flow regions in the normal pig heart are equally vulnerable to ischaemia during partial coronary stenosis. Eur. J. Physiol. (Mugers Arch.) 434:785794, 1997. [3] Beek JHGM van, Bussemaker J, and Westerhof N. Measurement of local oxygen consumption in small frozen tissue samples with a new method shows a higher metabolic rate in subendocardium than in subepicardium in isolated rabbit heart. J. Physiol. 491: 158P, 1996. [4] Beek JHGM van, Bussemaker J, Barends JPF, and Westerhof N. A 13C-NMR technique to determine absolute oxygen consumption in quickly frozen small myocardial samples. FASEB J. 10:A325, 1996. [5] Beek JHGM van, Csont T, Bussemaker J, and Barends JPF. Measuring local myocardial 02 consumption in many samples. Ann.Biomed.Eng. 24:S-19, 1996. (invited) [6] Beek JHGM van, Csont T, Kanter FJJ de, and Bussemaker J. Simple model analysis of t3C NMR spectra to measure oxygen consumption using frozen tissue samples. Adv. Exp. Med. Biol., in press [7] Chance, EM, Seeholzer, SH, Kobayashi K, Williamson JR. Mathematical analysis of isotope labeling in the citric acid cycle with applications to 13C NMR studies in perfused rat hearts. YBiol Chem. 258: 13785-13794, 1983. [8] Yu X, White LT, Doumen C, Damico LA, LaNoue KF, Alpert NM, Lewandowski ED. Kinetic analysis of dynamic 13C NMR spectra: metabolic flux, regulation, and compartmentation in hearts. Biophys J. 69: 2090-2102, 1995. [9] Weiss RG, Gloth ST, Kalil-Filho R, Chacko VP, Stern MD, Gerstenblith G. Indexing tricarboxylic acid cycle flux in intact hearts by carbon-13 nuclear magnetic resonance. Circ Res. 70: 392-408, 1992.

T*-values increased 4 min after onset of DIP-infusion for about 13 rain. In 7 of 10 patients, areas with lower T~-values or no significant increase of T~ after DIP could be clearly identified and corresponded well with findings of X-ray angiography and stress-echocardiography. In 3 patients, suspected ischemic myocardium could not be identified due to severe susceptibility artifacts. Our preliminary results show that alterations in the relationship between myocardial oxygen supply and demand are detectable using T*-exams. Condition of myocardial oxygenation can be described by T*-maps and the time-dependent behavior of the measured T*-values by a series of rapid measurements within 30 40 minutes. In the present study, for the first time DIP-induced regional changes of myocardial oxygenation in patients with CAD could be assessed with T*-exams. Our data show that T~'-exams are an alternative to CA studies and may be used for noninvasive diagnostic and therapy evaluation in patients with CAD. References [1] [2] [3] [4]

Ogawa S [1990] P N A S USA 87: 9868-9872. Li D [1996] Magn. Reson. Med. 36: 16-20. Wacker CM [1998] 6th I S M R M Sydney, 897. Wacker CM [1998] Magn. Reson. Med. submitted.

Abstracts/Magnetic Resonance Materials in Physics', Biology and Medicine 6 (1998) 191 207

205

Fig. 1. Examples of patients ( # 1-3) with 90% stenosis af LAD ( # 1 3) and hypokinesia in anteroseptal region ( # 1 2). T*-maps were obtained by outlining the endo- and epicardial orders of myocardium in the amplitude images (midventricular short-axis view). Expected ischemic areas could be well distinguished in the computed T*-maps. Colorbar from grey to black: T~-values from 15 to 20 ms.

Perfusion-corrected mapping of cardiac regional blood volume versus first-pass-imaging: A comparative study C. Waller, E. Kahler*, K.-H. Hiller*, S. Voll*, A. Haase*, G. Ertl, W.R. Bauer Medizinische Universit/itsklinik, 68135 Mannheim, Germany *Physikalisches Institut, Universit/it Wiirzburg, 97074 Wfirzburg, Germany Purpose: Measurement of intracapillary blood volume (called RBV) in the myocardium in vivo is of paramount importance for the assessment of pathophysiological processes in the heart. Currently, there are two magnetic resonance (MR) methods available for localized and non-invasive determination of RBV, first pass imaging [1,2] and fast T~ imaging [3]. Here, we present a new method that directly provides perfusion-corrected RBV data [4]. This technique is based on the acquisition of slice selective T~-images prior to and after contrast agent application [5]. In addition, first pass studies were performed and the resulting RBV values were compared with the RBV results of our fast T~ imaging method. Methods': We measured 10 male Wistar rats (body weight 200 300 g, anesthesia: sodium pentobarbital 60 mg/kg) on a Bruker Biospec 7 Tes[a spectrometer using a homogeneous transmitter coil and a circular polarized surface receiver coil. For fast T~ imaging we used an inversion recovery snapshot FLASH sequence with an adiabatic sechdnversion pulse and 24 consecutive FLASH images (MR-parameters: T R - 2.25 ms, TE = 1 ms, flip angle ~ 3 °, matrix 64 × 64, FOV = 5 x 5 cm 2, slice thickness - 3 mm). MR-first

Breath-hold contrast enhanced 3D coronary MR angiography using motion matched k-space sampling Yi Wang, Pris'citla A. Winchester, Howard M. Lee Department of Radiology, Cornell University Medical College, New York, NY 10021, USA Purpose: The major challenges of coronary M R angiography are respiratory motion and cardiac motion and limited signal-to-noise ratio. This study investigates contrast enhanced coronary M R A techniques. Contrast enhancement is used to increase signal-to-noise ratio or shorten scan time. Respiratory motion effects are minimized using breath-holding. Cardiac motion effects are minimized by matching the order of k-space sampling to the k-space distribution of M R signal. Contrast enhanced 3D coronary arteries are imaged in a single breath-hod acquisition. Metho&': A breath-hold contrast enhanced 3D coronary MRA technique was developed using motion matched k-space sampling. During the data acquisition period of the cardiac cycle, the center of k-space was acquired in mid-diastole when

pass imaging was performed using a T~-weighted snapshot FLASH sequence (MR-parameters: flip angle-~ 30 °, slice thickness = 3 mm, matrix = 64 x 64). For our T~ imaging method we used Gd-DTPA-albumin as inkavascular contrast agent (in collaboration with Physiologische Chemie, Wiirzburg), while the first pass studies were performed using Gd-DTPA-polylysine, since Gd-DTPA-albmnin was too viscous for theses kind of experiments. Results: With the slice selective method (5 animals) a mean RBV value of 12.8 +_ 0.7 Vol% was obtained over all animals in the left myocardium. Using the first pass method (5 animals) we determined an average RBV value of 13.6 _+ 3.2 Vol% in the whole left myocardium. Discussion: Our results indicate, that we obtained quantitative perfusion-corrected RBV values in good agreement with the first pass RBV studies in rat myocardium. However, in contrast to the first pass imaging technique, the perfusion-corrected method allows more accurate and faster determination of RBV in vivo. Furthermore, it provides quantitative RBV data under steady state conditions and is therefore suitable to visualize RBV alterations over time. References: [1] [2] [3] [4] [5]

Rosen BR et al., [1990], Magn. Res. Med. 14:249-265 Wilke N e t al., [1995],1MRI,5:227-237 Schwarzbauer C et al., [1993], Magn. Res. Med. 29:709-712 Bauer W R et al., [1996], Magn. Res. Med., 35:43-55 Kahler E et al., [1998], in press

motion of the coronary arteries was minimal. The edge of k-space was acquired in an edge-canter manner immediately before the center k-space acquisition, and in a canter-edge manner immediately after the center k-space data acquisition. The whole 3D data acquisition was completed in a breath-hold of 32-heartbeat duration. The pattern for sampling the center k-sPace was edge-canter-edge over the breathhold. Spin preparation was used to suppress background signal. This technique was validated in 5 health subjects. Twenty to twenty-five mL conventional Gd-DTPA agent was injected at 1 mL/sec. Data acquisition began when the contrast bolus arrived at the ascending aorta. The right coronary artery and the left circumflex artery were imaged in one double oblique volume, and the left main artery and the left anterior descending artery were imaged in another double oblique volume. 3D acquisition using motion matched k-space sampling was compared to 3D acquisition using conventional centric view ordering. Results: Motion matched k-space sampling substantially reduced cardiac blurring existing in conventional centric view order. Contrast enhancement provided high vessel-to-background contrast.

206

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 19I-207

Coronary arteries were well depicted and were viewed conveniently via maximum intensity projection (Fig. 1) and multi-planar reformatting (Fig. 2). Discussion : Our preliminary results indicate that motionmatched k-space sampling can reduce cardiac motion effects, and contrast enhanced 3D coronary M R A can be performed within a breathhold. Image contrast of coronary arteries is based on contrast

enhancements, independent of flow effects that may cause image artifacts. Background tissues such as surrounding epicardial fat and myocardium are well suppressed. The coronary arteries can be viewed conveniently via maximal intensity projection and multi-planar reformatting of the 3D data set. Coronary arteries are well depicted by this breath-hold contrast enhanced 3D M R A technique,

Fig. I.

Fig. 2.

Acquisition weighted 13C-chemical shift imaging in the isolated rat heart after infarction Claudia Weidensteiner, Titus Lanz, Axel Haase, Michael Horn*, Markus yon Kienlin Physikalisches Institut and *Medizinische Klinik, Universit/it Wtirzburg, Germany

nominal resolution of 4 mm x 4 mm × 6 mm (voxel size 96 gl). Male Wistar rats were anesthetized and subjected to coronary artery ligation (MI). After 4 weeks the hearts were excised and perfused in the Langendorff-mode. One lifer of Krebs-Henseleit-buffer containing 5 m M 99% enriched [2-~3C]-sodium acetate as substrate was recirculated. The label is incorporated into glutamate via the tricarbonicacid-cycle. Results: Fig. 1 shows the distribution of the glutamate-CA-resonance over a slice in the raiddle of a rat heart. The spectrum in Fig. 1 is taken from the indicated voxel in the myocardium. No signal is coming from the infarcted area. The nmltiplet structure of the glutamate resonances due to the J3C-~3C-coupling can be resolved. Fig, 2 corresponds to a slice without an infarcted area near the base of the same heart, Glutamate is distributed over the whole myocardium. Discussion: We showed the first 13C-CSI of an isolated rat heart after infarction. With acquisition weighted CSI we reached a signal-to-noise-ratio o f about 5 in a 96~tl voxel. This was sufficient to identify healthy and scarred myocardium due to the grade of incorporation of the 13C-label into glutamate.

Introduction: The limiting factor for localized 13C-spectroscopy is the low NMR-sensitivity of this nucleus which requires a long experimental time and large voxels. In order to measure the regional metabolise content in healthy and infarcted isolated rat hearts we performed acquisition weighted Chemical Shift Imaging (CSI) experiments [1,2]. Spectra over the whole object volume can be acquired with CSI. In comparison to conventional CSI the signal contamination between adjacent voxels is dramatically reduced in acquisition weighted CSI without a loss in sensitivity. Methods: We performed the experiments on a Bruker AMX-500 microscopy system. A homebuilt probehead with a double-tuned ~3C-~H-four-ring birdcage resonator [3,4] was optimized to obtain the essential high ~C-sensitivity. The 1H-channel was used for shimming, FLASH-imaging of the heart, WALTZ-16-decoup/ing and saturation of the IH-spins for NOE. The weighting function of the number of averages of each phase encoding step was a Hanning-window. The 3D-CSI dataset was acquired in 55 min with 13 × 13 x 13 phaseencoding steps (8 averages in the k-space-center, repetition time 1.5 s). With a field of view of 32 mm in x- and y-direction and 48 mm in z-direction, this yields a

References: [1] [2] [3] [4]

Maudsley AA et al. [1983] J.Mag.Res. 51:147-152 Mareci TH, Brooker H R [1991] J.Magn.Res. 92:229 246 Lanz T et al., abstract ESMRMB, Bruxelles 1997 Lanz T et al. [1997] M A G M A 5:243-246

Abstracts/Magnetic Resonance Materials in Physics, Biology and Medicine 6 (1998) 191-207

207

a)

Fig. 1. (a) Chemical shift image of the glutamate-C4-resonance in a transversal slice with scarred myocardium of an isolated, infarcted rat heart. The contours of the NMR-tube, the heart and the pressure balloon in the left ventricle (LV) are taken from an FLASH-image. (b) aH-decoupled t3C-spectrum showing the glutamate resonances in the indicated voxel.

Fig. 2. Chemical shift image of the same heart of the glutamate-C4-resonance in a slice without scarred myocardium.