Metabolic Response of Normal Human Myocardium to High-Dose Atropine-Dobutamine Stress Studied by 31P-MRS
Background 31P-MRS during cardiac stress may provide (patho)physiological insights into the high-energy phosphate metabolism of the myocardium. Accordingly, the purpose of the present study was to determine the metabolic response of normal human myocardium to severe atropine-dobutamine (A-D) stress. To corroborate the results from the present in vivo study, a 31P-MRS experiment was performed with a moving phantom to simulate respiratory motion.
Methods and Results The phantom experiment showed no relation (P=.371) between the intensity ratio of two separate phosphate peaks and amplitude of phantom excursions. The phosphocreatine (PCr) and ATP signal strength and the PCr/ATP ratio were determined from the left ventricular wall in 20 healthy subjects (posttest likelihood for coronary artery disease was <2.5%) with 31P-MRS at rest and during high-dose A-D stress (rate-pressure product increased threefold). Stress-induced changes were −21% for PCr (P<.001) and −9% for ATP (P<.05). The average PCr/ATP value at rest was 1.42±0.18 and decreased by 14% to 1.22±0.20 during stress (P<.001).
Conclusions The phantom experiment shows that the in vivo decrease of myocardial PCr/ATP due to high-dose A-D stress we observed is not a motion artifact. Consequently, this indicates that myocardial high-energy phosphate metabolism of the normal human heart is altered at high workloads.
Phosphorus-31 MRS is a unique noninvasive tool to study myocardial HEP metabolism. For example, ATP and PCr, which represent the body’s fundamental energy currency and a reservoir of cellular energy, respectively, can be detected by 31P-MRS.1 Acquisition of 31P-MR spectra during cardiac stress testing may provide new (patho)physiological insights into the HEP metabolism of the myocardium.2-5
Spectral acquisition during exercise-induced cardiac stress is most desirable but is subject to several limitations. The main limitations are possible motion artifacts introduced by physical exercise and a relatively low level of attainable cardiac stress. Moreover, many patients are unable to perform an adequate intensity of physical exercise, for various reasons.6 Pharmacologically induced stress has proved to be an attractive alternative for physical exercise.6-12 For example, A-D–induced stress is commonly used in conjunction with 2D echocardiography.6,7,13 A-D stress eliminates the above-mentioned problems of the physical exercise test. In previous animal and human 31P-MRS studies using surface coils, exercise and pharmacological cardiac stress testing were applied. In previous human 31P-MRS studies, cardiac stress was induced mainly by handgrip or leg exercise,14-16 whereas only one 31P-MRS study applied dobutamine stress.17 In these studies, a relatively low stress level was reached. In animal studies,3-5,18,19 however, higher stress levels were applied. In these animal studies, a decrease in the myocardial PCr/ATP ratio of normal hearts was observed if the heart was stressed severely.
The purpose of the present study was to determine the effects of severe pharmacological positive inotropic and positive chronotropic stimulation of the heart on the HEP metabolism of normal human myocardium with 31P-MRS. To investigate the possible effects of motion artifacts on the outcome of these experiments, the motion sensitivity of 3D-ISIS20 was determined separately with a moving multicompartment phantom to simulate breathing motion. Moreover, in volunteers, the 3D-ISIS volume was shifted deliberately to determine the reproducibility of the myocardial PCr/ATP ratio in the presence of small variations in the position of the ISIS volume. We also propose a semiquantitative approach to assess stress-induced changes in myocardial Pi signal intensity. The present study contributes to a better understanding concerning the metabolic response of normal human myocardium to a significant increase in cardiac workload.
A Philips 1.5-T Gyroscan S15 MR system (Philips Medical Systems International) and a 100-mm-diameter surface coil were used to acquire 31P-MR spectra. Subjects were positioned in the supine position, then volumes of interest were selected by image-guided spectroscopy with 3D-ISIS based on transverse and sagittal spin-echo MR scout images. Typical in vivo size of the volume of interest was 60 (caudocranial)×70×70 mm3. Localized homogeneity adjustment was performed with the body coil and an automatic procedure optimizing the 1H-MRS water signal obtained with a volume-selective 90°-180°-180° pulse sequence.21 The radiofrequency level was adjusted to obtain a 180° pulse of 40 μs for the reference sample (methyl-phosphonate) at the center of the 31P surface coil. Other technical details were similar, as described previously.22 All human myocardial 31P-MR spectra were acquired 200 ms after the R wave of the ECG. After each excitation, a 3-second relaxation delay was inserted, and the first R wave to occur triggered the next acquisition. The mean TR was 3.7±0.06 seconds at rest and 3.4±0.02 seconds during stress. Total acquisition time for a single in vivo spectrum was ≈10 minutes for 192 averaged free-induction decays. Phantom compartments were selected by adjusting the volume of interest to their boundaries based on MR images of the phantom acquired at stationary situation. Total acquisition time for a single in vitro spectrum was 3 minutes, with 64 free induction decays to be averaged at a TR of 3 seconds.
31P-MR spectra were transferred to a remote SUN-SPARC workstation to be quantified automatically23 by model function analysis in the time domain, using a priori spectroscopic knowledge to improve the accuracy of the spectral parameters.24-26 Signal modeling and prior knowledge were performed as in previous studies.22,27
Correction for Partial Saturation and Blood Contamination
The above-mentioned TRs for rest and stress acquisitions result in saturation correction factors of 1.34 for rest spectra and 1.36 for stress spectra.28 For the sake of clinical applicability, we applied a standard saturation correction factor of 1.35 to all 31P-MR spectra acquired, which value was based on a TR of 3.6 seconds as reported previously22 (see “Discussion”). The ATP level in the 31P-MR spectra was corrected for the ATP contribution from blood in the cardiac chambers on the basis of a previous study27 in which the 31P-MR spectrum of whole venous blood was quantified. The ratio of ATP to 2,3-DPG was 0.36 and was used in the present study to calculate the contribution of blood ATP to the observed ATP signal in cardiac 31P-NMR spectra acquired at rest and during stress.
SNR and Pi-e
An estimate of the SNR was obtained from the CRSD calculated for the myocardial PCr/ATP ratio, which is an indicator of the accuracy of the spectral quantification.24-26 The CRSD is based on the statistical theory of maximum-likelihood estimation, leading to lower bounds on the statistical errors in the parameter estimates, and is related to measurement noise. For each 31P-MR spectrum, the CRSD of myocardial PCr/ATP was divided by the myocardial PCr/ATP ratio, yielding an rCRSD, which is inversely related to the SNR.22 The ratio of the two resonance peaks in the spectral region of 2,3-DPG served as an estimate of the Pi signal intensity (see “Results” section).
To simulate the heart and surrounding tissues, a phantom was constructed, consisting of three concentric cubic compartments (Fig 1A⇓). The compartments were filled with 1% agarose gel containing various phosphorus compounds with resonance frequencies and T1 values close to the myocardial in vivo situation. T1 values were measured by the inversion recovery method.29 Compartment I (70×70×70 mm3) contained 50 mmol/L Pi at pH 5.0 (T1=9 seconds) and 50 mmol/L tripolyphosphate (T1=4.5 seconds). Compartment I was centered in compartment II (100×100×100 mm3), leaving a gap of 15 mm on all sides; this space was filled with 50 mmol/L phenylphosphonic acid at pH 7.0 (T1=8 seconds). Compartment II was centered in compartment III (200×200×200 mm3), leaving a gap between II and the bottom of III of 10 mm; this space was filled with 25 mmol/L Pi at pH 11.0 (T1=9 seconds).
To simulate respiratory motion, the phantom and surface coil were placed on top of a plateau on rails (Fig 1B⇑). Motion was introduced by connecting the plateau to an eccentric pin on a rotating disk. Phantom excursions (respiratory motion) were 0, 5, 15, 25, 35, and 45 mm, with variable speed (breathing frequency) of 25, 40, and 50 rounds per minute. Motion direction was in the caudocranial direction (z axis), in analogy to the in vivo situation of respiratory motion of the diaphragm and liver during a 31P-MRS examination because a Velcro band around the chest (to secure the 31P surface coil in place) limits the ventrodorsal movement of the chest wall.22 The 3D-ISIS volume encompassed compartment I before motion was introduced; this volume selection was maintained once the phantom started moving. Acquisition parameters of stationary spectra were identical to those of spectra acquired during phantom motion.
A series of 20 volunteers (average age, 50±17 years; 2 women, 18 men) was studied at rest only to determine the reproducibility of the myocardial PCr/ATP ratio, with small variations in the position of the 3D-ISIS volume. The first spectral acquisition was made with the 3D-ISIS volume encompassing the LV.22 The second 31P-MR spectrum was acquired after the ISIS volume was shifted 5 mm in the cranial direction and 5 mm to the dorsal side. Thereby, the volume of interest was at a suboptimal position but still included the entire LV while still excluding skeletal muscle, diaphragm muscle, or liver tissue, which would otherwise contaminate the myocardial 31P-MR spectrum. The 5-mm shift was based on the 3D-ISIS sensitivity profile, which is 5 mm smaller than the cube displayed on MR images.30 Parameter settings for spectral acquisitions were identical for preshift and postshift studies.
Cardiac Stress Testing
Twenty other volunteers (average age, 48±12 years; 5 women, 15 men) were selected for pharmacological stress testing. They were healthy at clinical examination, showed a normal ECG at rest, were not obese (height, 1.80±0.07 m; weight, 72±11 kg), and were normotensive (Table 1⇓), without a history of cardiovascular disease and without any complaints. The normal subjects >40 years old had no ECG-based evidence of ischemia during bicycle exercise testing (3.0±1.2-fold increase in RPP) and were without anginal complaints.15 Their posttest likelihood for CAD was <2.5%.31
A myocardial 31P-MR spectrum was first acquired at rest. Without the subject being moved, three subsequent doses of 10 μg/kg atropine sulfate were administered intravenously through an infusion line to block the vagal inhibition of heart rate.32 Then, dobutamine infusion was started at a dose of 10 μg · kg−1 · min−1 and was increased every 2 minutes until a steady target heart rate was reached. The highest infusion rate allowed was 40 μg · kg−1 · min−1. The target heart rate (in beats per minute) was based on sex and age (in years) of the subject: men (220−age)×0.85, women (200−age)×0.85.7,13 Blood pressures were recorded every 2 minutes with a Dinamap sphygmomanometer (Criticon) and were not allowed to exceed a systolic/diastolic blood pressure of 220/100 mm Hg, although these levels were never reached (see “Results” section). As soon as the acquisition of a 31P-MR spectrum during A-D stress was finished, infusion of dobutamine was stopped. In 4 volunteers, a third 31P-MR spectrum was acquired after a 15-minute recovery period. Acquisition parameters of 31P-MR rest spectra were identical to those of stress and recovery spectra.
Informed consent was obtained from all 40 subjects before the examination, and the study was approved by our institutional committee on human research.
Paired t tests were used to test the hypothesis that observed differences between, for example, rest and stress spectra are not significantly different from zero for a group of subjects. Linear regression analysis was performed to study relations between signal strength of phosphorus compounds in the phantom and phantom excursion or phantom speed and to study the relation between change in RPP and percentage change in myocardial PCr/ATP based on previously reported values. A value of P<.05 was considered to be significant. All t tests passed tests of normality and equal variance.
Fig 1⇑ shows 31P-MR spectra of the phantom at stationary situation and during motion (Fig 1C⇑) and the derived signal changes due to phantom motion (Fig 1D⇑). Before and during motion of the phantom, virtually no contaminating signal was detected from phenylphosphonic acid (compartment II) when the 3D-ISIS volume encompassed compartment I at stationary situation (Fig 1C⇑, panels 3 and 4). It was not possible to discriminate between the small phenylphosphonic acid resonance peak and noise in the spectral fitting procedure. Pi at pH 11.0 (compartment III) was not detected before or after motion was introduced when the volume of interest encompassed compartment I at baseline (Fig 1C⇑, panels 3 and 4).
If the phantom excursion was kept constant at 45 mm, spectral signal strengths were unaffected by changes in phantom speed (P>.05). Consequently, there are no arguments to presume that a change in breathing frequency could affect the outcome of in vivo spectroscopic measurements. The tripolyphosphate-αγ and Pi at pH 5.0 resonance signals were chosen as examples (Fig 1D⇑); the same results hold true for the tripolyphosphate-β resonance peak inside compartment I (not shown).
Reproducibility of the myocardial PCr/ATP ratio acquired before and after the 3D-ISIS volume was shifted was good; differences between the two acquisitions were 0.11±0.31 (P=.14) on average. Spectra of sufficient quality for automated quantification were obtained at rest, during stress, and after 15 minutes of recovery. The rCRSD was 12±3% on average at rest and changed to 15±4% during stress (P<.001); after 15 minutes of recovery, the rCRSD was 10±2% and similar to rest values (P=.30). Spectral line widths for PCr were unchanged when rest values (0.41±0.07 ppm, or 11±2 Hz at 1.5 T) and stress values (0.44±0.09 ppm, P=.08) were compared or when rest and recovery values (0.38±0.02, P=.98) were compared.
Hemodynamic changes are presented in Table 1⇑ and Fig 2⇓. Heart rate during A-D stress (147±12 bpm) was 102±8% of the calculated target heart rate (143±8 bpm, P=.22). An example of in vivo 31P-MR spectra acquired before, during, and 15 minutes after A-D stress is shown in Fig 3⇓. Quantification of the average metabolic response of normal human myocardium to severe A-D stress and the 15-minute recovery period is shown in Table 1⇑ and Fig 4⇓. With a threefold increase in RPP from rest to A-D stress, observed changes were −21% for PCr (P<.001) and −9% for ATP (P<.05). The average myocardial PCr/ATP ratio at rest was 1.42±0.18 and decreased to 1.22±0.20 (−14%) during stress (P<.001). After a 15-minute recovery period, the myocardial PCr/ATP ratio was 1.31±0.06 and comparable to baseline values (P>.05). The semiquantitative assessment of Pi using Pi-e yielded values of 1.44±0.27 at rest, 1.87±0.37 during stress (+33±26%, P<.001), and 1.75±0.18 after a 15-minute recovery period (+24±17% compared with rest, P<.05).
The main conclusion from the present study is that myocardial HEP metabolism may change in the normal human heart under high-level pharmacological stress conditions. The stress-induced reduction in myocardial PCr/ATP by 14% presently observed in the normal heart has not been reported previously in human studies,14-17 although similar changes have been observed in normal animal hearts using 31P-MRS under high-level stress (Table 2⇓).3-5,18,19 Probably the decrease in myocardial PCr/ATP we observed in the normal heart becomes evident only if the level of induced pharmacological stress is high; a threefold increase in the RPP was not attained in previous human studies (Table 2⇓, Fig 5⇓).14-17 That the relatively high stress level is the most likely explanation for the present findings is corroborated by the present spectroscopic results obtained during the recovery phase after stress. Note that the myocardial PCr/ATP ratio is similar to its resting value when acquired after a 15-minute recovery period; at that time, the remaining level of stress equals about 1.3 times the RPP at rest (Table 1⇑), which is a stress level comparable to those reported in previous human studies (Table 2⇓).14-17
The decrease in PCr and ATP signal due to severe pharmacological stress observed here is in agreement with previous animal studies,3,4 in which cardiac stress was also induced by infusion of dobutamine (Table 2⇑).
To exclude the possibility that the metabolic changes observed here are caused by motion artifacts related to increased respiration during stress, a phantom experiment was designed to simulate the effect of motion on the acquisition of spectroscopic data. At stationary situation and during phantom motion, contamination of the 3D-ISIS volume by signals originating from compartments II and III was virtually absent, most likely because of the conservative 3D-ISIS sensitivity profile.30 Even so, excursions of up to 45 mm should have introduced some signal from compartments II and III in the 3D-ISIS volume positioned around compartment I (Fig 1C⇑, panel 4). This signal was probably averaged out because the add-subtract scheme of 3D-ISIS was out of phase with phantom motion. Therefore, in the in vivo situation, with careful selection of myocardium within the volume of interest, in combination with the conservative 3D-ISIS sensitivity profile, it is unlikely that the myocardial spectrum is contaminated by skeletal and diaphragm muscle or liver tissue resulting from increased respiratory motion during the stress test.9 Furthermore, the phantom experiment shows that signal intensities of, for example, Pi at pH 5.0 and tripolyphosphate-αγ, are inversely related to the amplitude of phantom excursion, whereas their ratio is not affected by increased motion (Fig 1D⇑). Consequently, the in vivo decrease of the myocardial PCr/ATP ratio of the normal heart during severe stress observed here is not a motion artifact.
The phantom results are in accordance with what can be expected on the basis of simple calculations. When compartment I of the phantom is entirely encompassed by the 3D-ISIS volume, signal intensity is maximal. Moving the phantom outside this volume causes a partial signal loss. At an excursion of 45 mm, the maximal excursion in one direction (caudal or cranial) is 45/2=22.5 mm; consequently, at least 70−22.5=47.5 mm of compartment I is inside the 3D-ISIS volume, resulting in detection of 47.5/70×100=68% of the original signal strength. When the phantom is moving at an excursion of 45 mm, the predicted signal strength is (100+68)/2=84% of the available signal in compartment I, which is in accordance with the present results (Fig 1D⇑).
Several in vivo results from the present study also exclude the possibility that motion-related artifacts determine the decrease in the PCr/ATP ratio in normal human myocardium during severe stress. First, rCRSDs determined from in vivo acquisitions were low and spectral line widths were small, implying high-quality 31P-MR spectra before, during, and after A-D stress. The high spectral quality during all acquisitions indicates that the 3D-ISIS volume encompassed the LV at all times. The average rCRSD determined for the myocardial PCr/ATP ratio increased during stress, possibly because of signal loss associated with the observed decrease in myocardial PCr/ATP.
Second, the reproducibility of the myocardial PCr/ATP ratio was studied when the 3D-ISIS volume was shifted deliberately. The average myocardial PCr/ATP ratio acquired before and after the volume of interest was shifted was not statistically significantly different. Therefore, small changes in position of the currently applied 3D-ISIS volume do not significantly affect the myocardial PCr/ATP ratio as long as the volume of interest is selecting the entire LV anterior wall and blood only. In analogy to the present volume shift experiment, a previous 3D-ISIS study22 showed that serial spectroscopic examinations in the same subject with a 1-week interval yielded similar myocardial PCr/ATP ratios at both occasions (small interexamination variability). Consequently, the currently applied 3D-ISIS volume selection technique has proved to be highly reproducible.22
Furthermore, there are several in vivo 31P spectroscopic results that suggest that severe stress of the heart induces a genuine metabolic change in the myocardium. First, the commonly observed change in myocardial Pi in animal studies3-5,19,33 due to stress of the normal heart indicates a real change in the in vivo myocardial 31P-MR spectrum. Second, the myocardial PCr/ATP ratio fully recovers within 15 minutes after infusion of dobutamine ceases. Therefore, it is unlikely that observed changes in signal intensities between rest and stress are due to movement of the subject. Third, it appears that changes in, for example, the myocardial PCr/ATP ratio are inversely proportional to the change in RPP from rest to stress (Table 2⇑, Fig 5⇑). By itself, this could be caused by an increase in respiratory motion, but in most animal studies, which showed similar stress-induced changes in myocardial PCr/ATP (Table 2⇑), the phosphorus surface coil was sutured to the myocardium, excluding a direct effect of respiratory motion.
Possible Physiological Explanations
The fall in the myocardial PCr/ATP ratio during severe stress may suggest inadequate adaptation of oxidative phosphorylation of the normal human heart to high work states. This may be secondary to limited coronary reserve under high-work-state conditions, referred to as demand ischemia.4,5 In addition, an increase in Pi and decrease in PCr is not always associated with imminent heart failure.34 The decrease in ATP signal strength observed here may be partly due to an overall ATP utilization that is incompletely matched by ATP synthesis, as reported earlier for the normal canine heart.4
Conversely, a shift in substrate utilization by myocardium from, for example, free fatty acids to glucose may also lead to different steady-state levels of HEP metabolites.35 Also, regulatory mechanisms may change under different physiological conditions, such as increased mechanical loading.36 The latter two issues may also be related to the type of stress.2 Pharmacologically induced or exercise-induced increases in cardiac workload may have differential effects on steady-state levels of myocardial HEP metabolites. Furthermore, the changes in myocardial PCr/ATP we observed could be a transient phenomenon in the presence of adaptation to high stress levels.
Another explanation is that a higher exchange rate of phosphates between PCr and ATP during cardiac stress may lead to changes in T1 relaxation times of the metabolites studied. In a rat heart study,37 it was found that after the phosphoryl exchange between PCr and ATP was inhibited, the apparent T1 values were changed. A consequence of this phenomenon could be that the currently applied partial saturation correction factor for myocardial PCr/ATP determined at rest is not applicable to PCr/ATP ratios acquired during A-D stress. This theory is not supported by saturation transfer experiments in canine hearts,36,38 the results of which indicated that infusion of norepinephrine did not significantly affect the flux of PCr to ATP. On the other hand, even if the higher exchange rate of phosphates between PCr and ATP were to be the only cause of the observed changes in myocardial PCr/ATP due to an increase in cardiac workload, it would still be a stress-induced biochemical change that could be of diagnostic value.
Limitations of the Present Study
To completely rule out the presence of CAD in the healthy control subjects, a cardiac catheterization would be the preferred method, but this is obviously impossible for ethical reasons. Therefore, subjects were submitted to ECG bicycle exercise testing, a noninvasive technique for possible diagnosis of CAD. The pretest probability for CAD in the age group studied here is <10%, and it decreases to a posttest likelihood of <2.5% after ECG bicycle exercise testing with negative results.31 In addition, the eight youngest healthy subjects presently studied (35±7 years old), who had a pretest likelihood for CAD <2.0%, also showed a statistically significant decrease in myocardial PCr/ATP (−11±13%, P<.05) when rest and stress acquisitions were compared. Moreover, all healthy control subjects were without anginal complaints during severe bicycle or A-D stress testing, whereas the increase in RPP was comparable for the two stress tests (threefold). Therefore, the likelihood that the decrease in myocardial PCr/ATP observed here in the normal human heart during severe cardiac stress testing can be explained by unknown CAD is practically zero.
The difference in heart rate between rest and stress affects the TR and therefore the saturation correction factor for the myocardial PCr/ATP ratio. To determine this effect, we recalculated our myocardial PCr/ATP ratios with different saturation correction factors for rest and stress acquisitions based on the actual TRs (see “Methods”). Myocardial PCr/ATP ratios changed from 1.42±0.18 at rest and 1.22±0.20 during stress (P<.001) to 1.41±0.18 at rest and 1.23±0.20 during stress (P<.001) when different saturation correction factors were applied for rest and stress spectra (see “Methods”). Clearly, the small variations in TR between rest and stress cannot explain the observed changes in myocardial PCr/ATP between rest and stress.
In human cardiac 31P-MRS at 1.5 T, it is difficult to discriminate between 2,3-DPG and Pi, even when proton decoupling is applied.27 Instead of a qualitative description of the stress-induced changes in Pi observed here, a semiquantitative estimate for changes in Pi is proposed (Pi-e). At rest, this estimate may be dominated by the 2,3-DPG signal, but in stress spectra, Pi-e may serve as a good indicator of Pi, although changes in LV blood volume due to stress testing may affect the values of Pi-e. This limitation of Pi-e should be taken into consideration when the data are interpreted physiologically. The change in Pi-e found here in normal human myocardium induced by severe myocardial stress seems to be in agreement with previous animal studies,3-5,19 which reported cardiac stress–related changes in Pi content in normal myocardium (Table 2⇑). In contrast to PCr and ATP signal intensities, myocardial Pi-e did not fully recover to its baseline level in the first 15 minutes. Because of the limitations of Pi-e mentioned above, it is not possible to draw firm physiological conclusions regarding this lack of Pi-e recovery after severe cardiac stress. The limitations of Pi-e do not weaken the conclusion that myocardial PCr/ATP is decreased reversibly in the normal human heart during severe cardiac stress.
Because it appears that high work states of the heart change the level of HEP metabolites (Table 2⇑, Fig 5⇑), it would be interesting to determine the threshold value for a change in RPP at which the PCr/ATP ratio in normal human myocardium is significantly different from baseline values. This may lead to criteria for abnormal HEP metabolism in patients at rest and may provide a sensitive diagnostic tool to detect differences between patients and healthy control subjects concerning myocardial HEP metabolism.
The 95% tolerance interval for normal values39 of myocardial PCr/ATP at rest obtained in the present study ranges from 1.03 to 1.81 (n=20). This means that when a PCr/ATP ratio determined in a single patient is <1.03, there is a statistically significant difference with the normal value. In previous studies, differences of this order of magnitude have been reported in, for example, hypertrophic cardiomyopathy.1 For myocardial PCr/ATP determined during severe stress in the present study, this tolerance interval is from 0.79 to 1.65, which implies a significant difference from the normal value if the PCr/ATP ratio during cardiac stress is <0.79. The tolerance interval for the percentage change in PCr/ATP due to A-D stress ranges from −42% to +14%. Consequently, the PCr/ATP ratio of a subject should drop more than 42% during A-D stress to indicate an abnormal response to cardiac stress. Previous studies did not report such low PCr/ATP values during cardiac stress.
The phantom experiment strongly supports the hypothesis that the in vivo decrease in PCr and ATP signal strengths as observed in the present study during cardiac stress is partly caused by increased respiratory motion. Because all phosphate signals decreased to the same extent with increasing excursion of the phantom, the observed decrease of the myocardial PCr/ATP ratio in vivo is not a motion artifact but rather a metabolic adaptation of the normal human heart to severe cardiac stress. This indicates that myocardial HEP metabolism of the normal heart is altered at high workloads.
Selected Abbreviations and Acronyms
|CAD||=||coronary artery disease|
|CRSD||=||Cramér-Rao standard deviation|
|3D-ISIS||=||three-dimensional image-selected in vivo spectroscopy|
|Pi-e||=||estimate of inorganic phosphate|
- Received June 11, 1997.
- Revision received July 29, 1997.
- Accepted August 13, 1997.
- Copyright © 1997 by American Heart Association
Osbakken M, Mitchell MD, Zhang D, Mayevsky A, Chance B. In vivo correlation of myocardial metabolism, perfusion, and mechanical function during increased cardiac work. Cardiovasc Res. 1991;25:749-756.
Bache RJ, Zhang J, Path G, Merkle H, Hendrich K, From AHL, Ugurbil K. High-energy phosphate responses to tachycardia and inotropic stimulation in left ventricular hypertrophy. Am J Physiol. 1994;266:H1959-H1970.
Zhang J, Duncker DJ, Xu Y, Zhang Y, Path G, Merkle H, Hendrich K, From AHL, Bache RJ, Ugurbil K. Transmural bioenergetic responses of normal myocardium to high workstates. Am J Physiol. 1995;268:H1891-H1905.
Massie BM, Schwarz GG, Garcia J, Wisneski JA, Weiner MW, Owens T. Myocardial metabolism during increased work states in the porcine left ventricle in vivo. Circ Res. 1994;74:64-73.
Poldermans D, Fioretti PM, Boersma E, Thomson IR, Cornel JH, ten Cate FJ, Arnese M, van Urk H, Roelandt JR. Dobutamine-atropine stress echocardiography in elderly patients unable to perform an exercise test: hemodynamic characteristics, safety, and prognostic value. Arch Intern Med. 1994;154:2681-2686.
Lamb HJ, Doornbos J, den Hollander JA, Beyerbacht HP, de Roos A. Strategies for cardiac 31P-MR spectroscopy at rest and during dobutamine stress: a study of reproducibility. Proc Int Soc Magn Reson Med. 1995;96.
Lamb HJ, Beyerbacht HP, Doornbos J, den Hollander JA, Pluim BM, van der Wall EE, de Roos A. Metabolic response of human myocardium to atropine-dobutamine stress studied by 31P-MRS. Proc Int Soc Magn Reson Med. 1996;428.
van Rugge FP, van der Wall EE, Spanjersberg SJ, de Roos A, Matheijssen NA, Zwinderman AH, van Dijkman PRM, Reiber JHC, Bruschke AVG. Magnetic resonance imaging during dobutamine stress for detection and localization of coronary artery disease: quantitative wall motion analysis using a modification of the centerline method. Circulation. 1994;90:127-138.
Yabe T, Mitsunami K, Okada M, Morikawa S, Inubushi T, Kinoshita M. Detection of myocardial ischemia by 31P magnetic resonance spectroscopy during handgrip exercise. Circulation. 1994;89:1709-1716.
Conway MA, Bristow JD, Blackledge MJ, Rajagopalan B, Radda GK. Cardiac metabolism during exercise in healthy volunteers measured by 31P magnetic resonance spectroscopy. Br Heart J. 1991;65:25-30.
Massie BM, Schaefer S, Garcia J, McKirnan MD, Schwartz GG, Wisneski JA, Weiner MW, White FC. Myocardial high-energy phosphate and substrate metabolism in swine with moderate left ventricular hypertrophy. Circulation. 1995;91:1814-1823.
Headrick JP, Dobson GP, Williams JP, McKirdy JC, Jordan L, Willis RJ. Bioenergetics and control of oxygen consumption in the in situ rat heart. Am J Physiol. 1994;267:H1074-H1084.
Ordidge RJ, Connelly A, Lohman JAB. Image-selected in vivo spectroscopy (ISIS): a new technique for spatially selective NMR spectroscopy. J Magn Reson. 1986;66:283-294.
den Hollander JA, Mariën AJH, Luyten PR, Oosterwaal B, de Beer R, van Ormondt D. Automated time-domain quantification of in vivo 31P NMR spectra. Proc Int Soc Magn Reson Med. 1991;761.
van Ormondt D, de Beer R, Mariën AJH, den Hollander JA, Luyten PR, Vermeulen JWAH. 2-D approach to quantitation of inversion recovery data. J Magn Reson. 1990;88:652-659.
de Beer R, van Ormondt D. Analysis of NMR data using time domain fitting procedures. In: Diehl P, Fluck E, Gunther H, Kosfeld D, Seelig J, eds. NMR Basic Principles and Progress. Berlin, Germany: Springer-Verlag; 1992;26:201-248.
Bottomley PA, Hardy CJ, Weiss RG. Correcting human heart 31P NMR spectra for partial saturation: evidence that saturation factors for PCr/ATP are homogeneous in normal and disease states. J Magn Reson. 1991;95:341-355.
den Hollander JA, Evanochko WT, Dell’Italia L, Pohost GM. 31P NMR T1 inversion recovery measurements of the human heart. Proc Int Soc Magn Reson Med. 1993;1098.
Jose AD, Taylor RR. Autonomic blockade by propranolol and atropine to study intrinsic myocardial function in man. J Clin Invest. 1969;48:2019-2031.
Buser PT, Auffermann W, Wu ST, Jasmin G, Parmley WW, Wikman-Coffelt J. Dobutamine potentiates amrinone’s beneficial effects in moderate but not in advanced heart failure: 31P-MRS in isolated hamster hearts. Circ Res. 1990;66:747-753.
Balaban RS, Heineman FW. Nuclear magnetic resonance studies of myocardial metabolic responses to alterations in workload. In: Schaefer S, Balaban RS, ed. Cardiovascular Magnetic Resonance Spectroscopy. New York, NY: Kluwer; 1992:93-110.
Osbakken M, Blum H, Wang DJ, Doliba N, Ivanics T, Zhang D, Mayevsky A. In vivo mechanisms of myocardial function stability during physiological interventions. Cardiology. 1991;79:1-13.
Lentner C. Introduction to statistics. In: Lentner C, ed. Geigy Scientific Tables. West Caldwell, NJ: Medical Education Division, Ciba-Geigy Corp; 1982;2:205.