Noninvasive Quantification of Regional Myocardial Metabolic Rate of Oxygen by 15O2 Inhalation and Positron Emission Tomography
Background The purpose of this study was to validate a novel method for noninvasive quantification of regional myocardial oxygen consumption (MMRo2, mL·min−1·100 g−1) and oxygen extraction fraction (OEF) by use of positron emission tomography (PET) and inhalation of 15O-labeled molecular oxygen gas (15O2).
Methods and Results Twenty-four measurements were performed in eight closed-chest anesthetized greyhounds at baseline and during infusions of adenosine (100 to 200 μg·kg−1·min−1), isoproterenol (1 to 10 μg/min), and propranolol (5 mg bolus+0.2 to 1 mg/min) with morphine (5 mg slow infusion+0.2 to 0.5 mg/min) to obtain a wide range of oxidative metabolism. The PET imaging protocol consisted of 15O2 emission (OEF and MMRo2), transmission, [15O]CO emission (blood pool), and [15O]CO2 emission (myocardial blood flow: MBFpet, mL·min−1·g−1) scans. OEF was calculated from the PET data (OEFpet) by three different analytical techniques: steady-state, 5-minute, and 8-minute autoradiographic analyses. Reference measurements of MBF (MBFref) and OEF (OEFref) were obtained during 15O2 inhalation with radiolabeled microspheres and paired arterial and coronary sinus blood sampling, respectively. MMRo2 was calculated from the PET (MMRo2pet) and the reference (MMRo2ref) data as follows: MMRo2=OEF×MBF×(O2 content of arterial blood). OEF measured by the steady-state PET method was well correlated with the reference data over the range 0.16 to 0.73 (OEFpet=1.03 OEFref −0.01, r=.97), as was MMRo2 over the range 2.4 to 27.5 mL·min−1·100 g−1 (MMRo2pet=0.98 MMRo2ref +0.91, r=.94). OEFpet calculated by use of the 5-minute and 8-minute autoradiographic analyses were equally well correlated with the reference measurements (r=.95 and r=.97, respectively). There were no significant differences between values of MMRo2pet calculated by use of the steady-state, 5-minute, and 8-minute autoradiographic analyses (P=NS by ANOVA). Regional values of MBFpet, OEFpet, and MMRo2pet were homogeneously distributed and similar to the whole-heart values both at baseline and during the various pharmacological interventions.
Conclusions Accurate quantification of OEF and MMRo2 is feasible with 15O2 inhalation and PET imaging using both the steady-state and autoradiographic analytical approaches. These studies suggest the applicability of this method for quantitative assessments of regional cardiac oxidative metabolism in clinical studies.
The energy required for cardiac contraction is derived from the hydrolysis of high-energy phosphates, which are generated by aerobic metabolic pathways culminating in the phosphorylation of ADP to ATP.1 Measurements of MMRo2 and OEF would therefore provide direct assessments of the status of myocardial oxidative metabolism. The ability to perform these measurements in humans may further our understandingof the pathophysiology and metabolic consequences of cardiac diseases as well as providing a means for further exploring the relationship between regional myocardial metabolism and contractile function in humans in normal and disease states.
PET is a radionuclide imaging technique that allows noninvasive assessment of regional tissue metabolism,2 and a number of tracers have been used for investigating cardiac oxidative metabolism by this imaging modality. 11C-labeled palmitic acid has been used previously as a marker of oxidative metabolism of free fatty acids.3 More recently, 11C-labeled acetate has been advocated as a tracer of tricarboxylic acid cycle activity4 and has been used for assessing oxidative metabolism by PET in both experimental animals5 6 7 8 9 and humans.10 11 12 A number of studies have shown that the rate constant describing the clearance of [1-11C]acetate from the myocardium correlates well with AV differences of blood oxygen content.5 6 7 8 9 However, absolute quantification of OEF and MMRo2 with this tracer has not yet been achieved.
A new technique for the quantification of MMRo2 and OEF by use of PET and inhalation of 15O2 has been developed.13 14 The theoretical background and methodological details of this technique are described in-depth in another report.14 The purpose of the present study was to validate the measurement of OEF and MMRo2 with the 15O2 inhalation–PET method by comparing it with the data obtained by direct AV sampling and microsphere myocardial blood flow measurements in closed-chest anesthetized greyhounds.
Eight greyhounds (body weight, 30±19 kg) were initially sedated with 4 mg acetapromazine IM. Anesthesia was induced with 25 mg/kg pentobarbital sodium IV. The animals were intubated and mechanically ventilated with a mixture of oxygen (Fio2, 20% to 40%), air, and nitrous oxide. Anesthesia was maintained throughout the study by inhalation of halothane (0.5% to 1.0%).
Experiments were performed in closed-chest preparations. Catheters were placed in the right and left femoral vessels and in the right external jugular vein. The catheters in the femoral veins were used for infusions of drugs and intravenous fluids. The catheter in the right femoral artery was advanced into the abdominal aorta for withdrawal of reference arterial blood samples during microsphere infusion, for measurement of arterial oxygen content, and for arterial blood pressure monitoring. An 8F pigtail catheter was positioned in the left ventricle via the left femoral artery for injection of microspheres. The catheter in the right external jugular vein was advanced into the coronary sinus under fluoroscopic guidance. Adequate positioning of the catheter in the coronary sinus was confirmed by injection of contrast medium and measurement of the oxygen contents of coronary sinus and peripheral venous blood samples. Arterial blood pressure and an ECG were monitored continuously.
Positron Emission Tomography
All PET scans were performed with an ECAT 931/08-12 scanner (CTI/Siemens Inc), which consists of eight rings of bismuth germanate crystal detectors. This scanner allows 15 planes of data acquisition in a field of view of 10.5 cm; thus, the whole heart may be imaged simultaneously. All emission and transmission sinograms were reconstructed with a Hanning filter with a cutoff frequency of 0.5 of maximum. This resulted in a spatial resolution of 8.4±0.7 mm full width at half maximum for the emission data and 7.7±0.7 mm full width at half maximum for the transmission data15 at the center of the field of view. Further details on the physical performance of this scanner have been reported elsewhere.16
The animals were positioned in the left lateral decubitus position on the scanner table. A 5-minute rectilinear scan was performed by exposure of an external 68Ge ring source to determine the optimal position for imaging the heart. Briefly, the animals inhaled 15O2 (t½=2.05 minutes) at 2.5 MBq/mL at 500 mL/min for 13 minutes. Data acquisition commenced 30 seconds before the start of gas administration, to obtain a measurement of background radioactivity, and lasted for a total of 16.5 minutes. The scan was divided into 18 time frames, as follows: 1×30 seconds, 4×30 seconds, 6×60 seconds, 1×300 seconds (steady state), and 6×30 seconds. After a 15-minute period to allow for decay of 15O radioactivity to background levels, a 10-minute transmission scan was performed by exposure of the external 68Ge ring source. The transmission data were used to correct the emission data for tissue attenuation of the 511-keV annihilation gamma photons. At the end of transmission, the blood pool was imaged by inhalation of [15O]CO, which labels erythrocytes by the formation of carboxyhemoglobin. [15O]CO was administered for 4 minutes at a radioactive concentration of 3 MBq/mL and at a flow rate of 500 mL/min. A 6-minute single-frame emission acquisition was initiated 1 minute after the end of [15O]CO inhalation. Arterial blood samples were taken every minute during the scan, and the [15O]CO concentration in whole blood was measured with a sodium iodide well counter cross-calibrated with the scanner. After a 10-minute period to allow for decay of 15O radioactivity to background levels, MBF was measured by [15O]CO2 inhalation (MBFpet) as a means of administering [15O]H2O. This method has been described previously.17 Briefly, [15O]CO2 gas, which is converted to [15O]H2O in the lung, was inhaled for a period of 3.5 minutes (3 MBq/mL×500 mL/min). A 25-frame dynamic scan was started 28 seconds before the start of [15O]CO2 delivery and lasted for a total of 7 minutes. The effective dose equivalents estimated for humans undergoing the procedures were 2.3, 3.4, and 4.8 mSv for the [15O]CO, [15O]CO2, and 15O2 scans, respectively.
The following intravenous drug infusion protocols were performed to achieve a wide range of MMRo2 and OEF: adenosine (100 to 200 μg·kg−1·min−1), isoproterenol (1 to 10 μg/min), and propranolol (5 mg bolus+0.2 to 1 mg/min) with morphine (5 mg slow infusion by hand+0.2 to 0.5 mg/min). In each animal, 15O2, transmission, [15O]CO, and [15O]CO2 emission scans were performed at rest and during the continuous infusion of two different drugs. Therefore, a total of 24 measurements were made. Stability of hemodynamics and oxidative metabolism throughout each condition was confirmed by monitoring of heart rate, blood pressure, and oxygen saturation and content in arterial and coronary sinus blood.
Paired arterial and coronary sinus blood samples were anaerobically withdrawn at 3, 5, 7, 8.5, and 12.5 minutes after the start of 15O2 inhalation and were analyzed for blood gases (Po2, Pco2, blood oxygen saturation) and blood hemoglobin concentration with a CO oximeter and an automated blood gas analyzer. Arterial blood withdrawn at 8.5, 12.5, and 14.5 minutes after the start of 15O2 inhalation and at 4 and 5 minutes after the start of the [15O]CO2 scan was immediately centrifuged, and the 15O activities in whole blood and plasma were measured in a sodium iodide well counter. These data were used for calculating the 15O2, [15O]H2O, and total 15O concentrations in the arterial blood samples.14
MBFref measurements were made during steady-state 15O2 inhalation by the radiolabeled microsphere method. Ten minutes after the start of 15O2 inhalation, one set of 15-μm-diameter microspheres (1.0×106 to 1.5×106) labeled with either 113Sn, 57Co, or 46Sc (NEN DuPont, UK) were infused into the left ventricular chamber over 30 seconds. A different set of labeled microspheres was injected during each of the three experimental conditions imposed in each animal. Arterial blood was withdrawn from a catheter in the descending aorta with a calibrated Harvard pump at a flow rate of 5.12 mL/min. Withdrawal was initiated 30 seconds before microsphere infusion and lasted for a total of 2.5 minutes. At the end of the study, the heart was removed and the left ventricular myocardium was dissected regionally. The tissue samples were weighed, and the radioactive signals from each set of microspheres were determined simultaneously with a multichannel gamma counter programmed to correct for the cross talk between the counting windows for each isotope.
MBFref values were calculated by the standard reference technique,18 as follows: MBFref=Cm×Qr/Cr, where Cm represents the nuclide activity per gram of myocardial tissue, Qr is the withdrawal rate of the pump (mL/min), and Cr is the total nuclide activity in the reference arterial blood sample. Whole-heart OEFref values were determined from the oxygen contents of arterial and coronary sinus blood by the Fick principle, as follows: OEFref=(Cao2−Cvo2)/Cao2, where Cao2 and Cvo2 represent the oxygen contents of arterial and coronary sinus blood, respectively. These reference values of OEF and MBF were used to generate reference values of MMRo2 (MMRo2ref) as follows: MMRo2=[O2]a·OEF·MBF, where [O2]a represents the total oxygen content of arterial blood.14
All sinograms were corrected for tissue attenuation and reconstructed on a MicroVax II computer (Digital Equipment Corp) by dedicated array processors using standard reconstruction algorithms. Images were transferred to SUN 3/60 workstations for further analysis. Image manipulations and data handling were performed with the Analyze (Mayo Foundation) and the Pro-Matlab (The MathWorks Inc) software packages and special dedicated software. For the calculation of MMRo2 and OEF from the PET data (MMRo2pet, OEFpet), a considerable degree of image manipulation is required for the generation of images of blood volume, lung gas volume (from the transmission data), extravascular tissue density, myocardial [15O]H2O, and buildup and steady-state 15O2 images. These processes are described in detail in the companion article.14
Seven sets of ROIs were drawn. Myocardial tissue ROIs were positioned in the anterior, lateral, inferior, and septal walls in five to seven image planes. In addition, a set of whole-heart regions was also drawn. These regions of interest were projected onto the 15O2, [15O]CO2, lung gas volume, and blood volume data sets as described previously.13 Arterial ROIs were positioned in the left ventricular and left atrial chambers for the generation of arterial time-activity curves during 15O2 and [15O]CO2 inhalation, respectively.14 These data were used to calculate global and regional values of MBFpet, OEFpet, and MMRo2pet. OEFpet was determined by the three different analytical approaches described in the companion article, ie, steady-state, 5-minute autoradiographic, and 8-minute autoradiographic techniques.14 Values of MBF were calculated from the PET data acquired during [15O]CO2 inhalation by use of a previously validated single-tissue-compartment model that included corrections for partial volume effect and spillover from the left ventricular chamber.17
Measurements of OEF and MMRo2 by PET were compared with the reference measurements by linear regression analysis. Regional variability in OEF and MMRo2 by PET was assessed by calculation of the coefficient of variation of the regional data and ANOVA. Comparison of multiple data sets was performed by ANOVA, and specific differences were identified by a Student's t test corrected for multiple comparisons with the Bonferroni inequality adjustment.19 All data are expressed as mean±SD.
Hemodynamic data during the steady state of O2 inhalation for the different pharmacological interventions are summarized in Table 1⇓. Both adenosine and isoproterenol infusion caused a significant increase in heart rate without changing systemic blood pressure. Infusion of propranolol with morphine decreased blood pressure without significantly changing the heart rate. The RPP, an indirect index of MMRo2, was calculated as the product of systolic blood pressure and heart rate. The linear relationship between RPP and MMRo2, assessed by both PET and reference methods, is illustrated in Fig 1⇓. At the higher flow values induced by administration of adenosine, the relationship between blood flow and oxidative metabolism deteriorates significantly, consistent with the existence of luxury perfusion induced by adenosine.
The pharmacological interventions used in this study allowed comparison of the PET techniques with invasive AV sampling over a wide range of OEF and MMRo2. Measurement of OEFpet by the steady-state method compared very favorably with that measured by AV sampling (OEFpet=1.03 OEFref−0.01, r=.97) (Fig 2⇓). Although measurement of MMRo2 by PET also compared well with the reference data (MMRo2pet=0.98 MMRo2ref+0.91, r=.94) (Fig 3⇓), the comparison was not as good as for OEF. This was predominantly because of the slight difference in MBF as assessed by radiolabeled microspheres and PET. The effects of the various pharmacological interventions on the stability of the experimental preparation and OEF measured by the reference and PET techniques are summarized in Tables 2⇓ and 3⇓, respectively.
An alternative method for determining OEF (and therefore MMRo2) from the PET data can be achieved by autoradiographic analysis of the dynamic 15O2 PET data.13 This analysis was performed over both the first 5 and the first 8 minutes of 15O2 inhalation. The correlations between OEF measured by these two analyses and the reference measurements are summarized in Figs 4⇓ and 5⇓, respectively. The data indicate that both analyses provide accurate determinations of OEF, although the analysis performed over the first 8 minutes of 15O2 inhalation provided data slightly superior to the 5-minute analysis, as evidenced by the higher correlation coefficient and the slope of the regression line being closer to the line of identity. The values of OEF, and consequently of MMRo2, derived from the autoradiographic analyses were similar to those obtained from the steady-state analysis (P=NS by ANOVA) (Table 3⇑). The accuracy of the steady-state, 5-minute, and 8-minute autoradiographic analyses relative to the reference determinations as a function of the prevailing blood flow is summarized in Fig 6⇓. It is noteworthy that in all but two cases (under high-flow conditions), OEF measured by PET was within 20% of that measured from analysis of AV differences.
Regional Values of OEF and MMRo2 by PET
One of the major benefits of PET imaging is the ability to record data from all parts of the heart simultaneously, thus allowing regional measurements to be made. Fig 7⇓ shows the homogeneous distribution of the regional values of MMRo2 in the different anatomic segments of the left ventricle calculated by the steady-state technique. The values of OEF were also homogeneously distributed and similar to the whole-heart measurements. Similar data were obtained from both the 5-minute and 8-minute autoradiographic analyses. It can be seen from Fig 7⇓ that the changes in MMRo2 during the different pharmacological interventions occur homogeneously throughout the different regions of the left ventricle. A similar pattern of change was observed for the OEF measurements. In addition, the values of OEF and MMRo2 in each individual region were similar to the values derived from the whole-heart ROIs.
Intrasubject regional variabilities, expressed as the coefficient of variation of the individual regional values, are shown in Table 4⇓ for all the analytical approaches. These data demonstrate that for a given condition, the regional variability in OEF by PET is similar for the different analytical approaches. The steady-state approach gave the lowest values of intrasubject regional variability, and there was little difference between the values for the 5-minute and 8-minute autoradiographic approaches. Furthermore, it appears that regional variability of OEF increases under conditions of increased MBF. Values of regional coefficient of variation for the MMRo2 by PET measurements were higher than for the measurements of OEF by PET. Although only values of MMRo2pet calculated by the steady-state analysis are shown (Table 4⇓), similar data were observed when both the 5-minute and 8-minute autoradiographic analyses were used.
This article describes the experiments performed to validate the accuracy of OEF and MMRo2 measurements by 15O2 inhalation and PET imaging. The kinetic model used in this study was first described by Iida et al.13 14 Initial studies in normal human volunteers provided values of OEF and MMRo2 by PET that were consistent with the literature values obtained by invasive AV sampling techniques.13 14 20 21 However, no rigorous validation of this technique has been performed previously. The data presented in this study demonstrate the accuracy of noninvasive quantitative assessments of cardiac oxidative metabolism with 15O2 inhalation and PET.
Comparison of PET With Reference Techniques
The pharmacological interventions used in this study allowed validation of the PET techniques over a wide range of OEF and MMRo2 values. Isoproterenol is a specific β-adrenergic receptor agonist that increases cardiac work, and hence MMRo2, by increasing heart rate and myocardial contractility. MBF increases after isoproterenol administration because of metabolic and direct vasodilatation.22 Adenosine is a vasodilator that acts by causing smooth muscle relaxation and hence vasodilatation at the level of the coronary resistance vessels.23 The increase in MBF after administration of adenosine is accompanied by a rise in MMRo2. This increase in MMRo2 is more than offset by the increase in MBF, and thus, the net effect of adenosine infusion is to decrease OEF. A combination of propranolol with morphine decreased MMRo2 by virtue of their negative inotropic effects. The reduction in cardiac contractility and hence metabolic demand results in a secondary decrease in MBF due to the physiological coupling between MBF and myocardial metabolism. Although MMRo2 decreased after the administration of propranolol with morphine, OEF did not increase above control values. This is consistent with a physiological coupling between MBF, cardiac mechanical work, and metabolic demand. The cardiodepressant effect of propranolol with morphine resulted in a decrease in MBF as well as a decrease in MMRo2. The increase in OEF that may occur as a result of a reduction in MBF is offset by the decrease in myocardial metabolic demand. Therefore, this may explain why the net effect of administration of propranolol with morphine was not to change OEF.
Oxygen extraction fraction for whole-heart ROIs measured by PET correlated well with those measured by AV sampling over a wide range of fractional extractions (0.16 to 0.73) for all the analytical procedures tested, ie, steady-state (Fig 2⇑), 5-minute (Fig 4⇑), and 8-minute (Fig 5⇑) autoradiographic analyses. These data confirm the accuracy of OEF measurements by PET.
The measurement of OEF is potentially of great use in its own right. Under normal circumstances, blood flow is coupled to cardiac metabolism. Therefore, in most cases, measurements of blood flow alone will accurately reflect the metabolic state of the myocardium. However, under certain pathophysiological circumstances, such as ischemia and reperfusion, this relationship may be altered. Interestingly, recent studies in dogs have shown that partial coronary occlusion produced an increase in arterial oxygen extraction.24 In addition, studies using [11C]acetate and PET have indicated that the sustenance of cardiac oxidative metabolism is a good prognostic indicator for contractile recovery after myocardial infarction.25 The methods validated in the present article for quantifying MMRo2 noninvasively by PET scanning should be suitable for performing similar studies. They may not only provide an additional approach for assessing myocardial viability but also give further insight into the mechanisms by which the myocardium adapts metabolically to a reduced level of perfusion, by virtue of the ability to measure OEF independently.
Good correlations were also obtained between MMRo2 by PET measurements and reference techniques for all three analytical approaches. The correlation was not quite as strong as for the measurement of OEF, mainly because of the added variability of the MBF measurement by PET in the calculation of MMRo2pet. Although the two measurements of blood flow were well correlated, MBF measured by PET differed slightly from the corresponding microsphere determination. In the experimental protocol, radiolabeled microspheres were injected during steady-state 15O2 inhalation and not simultaneously with the [15O]CO2 MBF measurement, which was performed about 40 minutes later. During this time delay, it is conceivable that MBF changed to some extent, despite the most stringent efforts to maintain stable experimental conditions (Table 2⇑). This problem, however, could be overcome in the future by quantifying MBF from the washout of [15O]H2O from the heart at the end of 15O2 inhalation, thus obviating the need for a separate MBF measurement.
Comparison of Steady-State and Autoradiographic Analyses
This study demonstrated that the steady-state and autoradiographic analyses provided accurate measurements of both OEF and MMRo2 compared with the invasive reference measurements. The steady-state method marginally gave the best correlation with invasive AV sampling, although there were no significant differences between the values obtained by the steady-state and both of the autoradiographic analytical techniques. Of the autoradiographic methods, the analysis over the first 8 minutes of 15O2 inhalation gave a slightly better correlation than the 5-minute analysis. Intrasubject regional variabilities, as assessed by coefficient of variation, for the different procedures were found to be similar.
Determination of OEF and MMRo2 from the PET data requires that three emission scans be performed in addition to the transmission, ie, [15O]CO, [15O]CO2, and 15O2 scans. Although the steady-state analysis is relatively easy to implement and provided the best correlation with the invasive methods, a number of practical disadvantages are associated with this technique. In particular, the steady-state approach requires the subject to inhale 15O2 for ≈20 minutes. This increased scan duration results in a larger radiation dose to the recipient and increases the chance of subject movement during the scan. Alternatively, the use of the autoradiographic analyses provides accurate estimates of OEF and MMRo2 that are not significantly different from either the steady-state or the reference measurements, with the advantage that they require a shorter period of 15O2 inhalation, hence reducing the radiation dose to the subject without significantly altering the accuracy of the PET measurements. The mathematical formulation required to calculate OEF and MMRo2 is more complex for the autoradiographic approaches than for the steady-state analysis. In addition, the autoradiographic approach requires the continuous measurement of the arterial input function either by continuous measurement of the 15O concentration in blood withdrawn from a radial arterial catheter or from the PET data by ROIs positioned in the left ventricular chamber, as performed in this study.
For clinical studies in humans, it is suggested that an autoradiographic approach should be used because of the shorter duration of 15O2 inhalation and the lower radiation dose to the patient compared with the steady-state protocol, without significant loss in the accuracy of the OEF and MMRo2 measurements. On the evidence presented in this study, there seems to be no significant difference between the 5-minute and 8-minute autoradiographic analyses. Computer simulation studies have suggested, however, that the statistical fluctuations in the calculation of OEF and MMRo2 are greater in the 5-minute than in the 8-minute analysis.14 Therefore, it is suggested that the 8-minute autoradiographic approach be adopted for clinical studies, because it should provide accurate and stable values of OEF and MMRo2.
Regional Measurements of OEF and MMRo2
Regional values of resting OEF and MMRo2 have been measured previously in hearts of anesthetized dogs.26 27 28 These studies used the technique of microspectrophotometry to measure oxygen saturation of hemoglobin in arterioles and venules in frozen sections of myocardium. These experiments indicated that AV differences in O2 saturation were similar in the lateral and septal walls of the left ventricle and in the right ventricle.26 It was further demonstrated that resting MMRo2 values in septal and free wall regions of the left ventricle were similar. The average value of resting MMRo2 was 11.0±0.4 mL·min−1·100 g−1,27 which was very similar to that measured by PET in the present study.
The major advantage of the microspectrophotometric assay is that the resolution of the technique is sufficient to discern transmural gradients in AV differences in oxygen saturation. Indeed, it was shown that in the free wall of the left ventricle, OEF was ≈18% higher in the subendocardium than in the subepicardium.26 No such difference was identified either in the interventricular septum or in the right ventricular myocardium.26 The data presented in this study using PET represent average transmural values of OEF and MMRo2 for different macroscopic regions of the heart, because the spatial resolution of PET cameras is not sufficient to measure either OEF in the right ventricular wall or transmural differences in OEF. Indeed, one of the assumptions in the kinetic model used in the present study is that oxygen is extracted from arterial blood homogeneously throughout the designated ROI. This assumption of homogeneity is common to all the compartmental models currently used in cardiac PET studies. However, the average transmural values of OEF and MMRo2 obtained with PET in the different regions of the left ventricle are very similar to those obtained previously by microspectrophotometry.26 27
The main advantages of the PET technique for measurement of OEF and MMRo2 over previous methods are that it is noninvasive, regional, quantitative, and accurate and that it allows repeat studies to be performed in vivo in the same subject within the same scanning session. Microspectrophotometry can be performed only ex vivo after only a single experimental intervention and is applicable only to experimental animal studies. The data in the present study indicate that the changes in OEF and MMRo2 are regionally homogeneous as a result of the various pharmacological interventions, with the values for each individual region being similar to the values for the whole-heart ROIs. These data suggest that the techniques presented in the present study should be useful in the assessment of the effects of various pharmacological interventions on regional myocardial oxidative metabolism in humans.
Comparison With Other PET Tracers
Several tracers have been used for the assessment of regional cardiac oxidative metabolism, for example, [11C]palmitate.3 However, quantification of free fatty acid utilization by use of this tracer was hindered by the dependence of the myocardial extraction on substrate availability and MBF and by the overcomplexity of the tracer model required to accurately describe its tissue kinetics.2 A number of other free fatty acid analogues have been used in PET studies,2 but currently few such investigations are being performed.
More recently, [11C]acetate has been used for assessment of regional MMRo2 by PET. Studies in experimental animals have demonstrated that the rate constant describing the clearance kinetics of [11C]acetate from the myocardium is linearly correlated with MMRo2 measured from AV differences in the rat heart4 and isolated perfused rabbit heart5 and in vivo in dogs6 7 8 9 and humans.10 11 12 Preliminary studies have recently shown that MMRo2 calculated by the 15O2 inhalation technique correlated with the clearance of [11C]acetate in humans both at rest and after intravenous dobutamine infusion.29 Although studies have demonstrated a linear relationship between the clearance rate constant for [11C]acetate and MMRo2,5 6 7 8 9 absolute quantification of MMRo2 has not been possible. Furthermore, the [11C]acetate technique does not at present allow direct determinations of OEF, which may be an important parameter in its own right. When the quantitative relationship between the [11C]acetate turnover rate constant and MMRo2 has been elucidated more precisely, it may be possible to derive estimates of OEF from the [11C]acetate data with separate MBF measurements. Thus, the 15O2 inhalation technique validated in the present study is the only one to date that allows accurate noninvasive quantification of OEF and MMRo2 in absolute terms by PET.
Although the 15O2 inhalation technique provides absolute quantification of MMRo2 and OEF, its implementation requires the resolution of a number of methodological problems14 and is practically more complex to perform than the [11C]acetate method, in that the scan duration is longer and multiple scans need to be performed. These issues have been discussed in more detail elsewhere.14 However, application of this technique has been shown to be feasible in humans,14 29 and future refinements of this technique should significantly improve many of the difficulties that exist at present.
In summary, we report the successful experimental validation of a novel approach for noninvasive quantification of OEF and MMRo2 by PET and 15O2 inhalation. Three different analytical approaches were tested, namely, the steady-state and the 5-minute and 8-minute autoradiographic analyses. The data show that all three techniques provide accurate measurements of OEF and MMRo2 compared with invasive reference techniques, although the 8-minute autoradiographic approach is recommended for clinical studies. This technique will allow quantitative assessments of regional cardiac oxidative metabolism in the clinical setting.
Selected Abbreviations and Acronyms
|MBF||=||myocardial blood flow|
|MMRo2||=||myocardial metabolic rate of oxygen consumption|
|OEF||=||oxygen extraction fraction|
|PET||=||positron emission tomography|
|ROI||=||region of interest|
This study was supported by the Medical Research Council. Dr Yamamoto was a visiting research fellow from the Research Institute of Angiocardiology and Cardiovascular Clinic, Kyushu University, Fukuoka, Japan. Dr de Silva was a postgraduate student in the Division of Biomedical Sciences, King's College, University of London, supported by a Medical Research Council postgraduate research studentship, and is currently supported by a Foulkes Foundation Fellowship. Dr Iida was a visiting scientist from the Research Institute of Brain and Blood Vessels, Akita, Japan. Drs Yamamoto and Iida were supported by the Japanese Heart Foundation and Bayer. We are greatly indebted to the staff of the MRC Cyclotron Unit. In particular, we would like to thank Dr Jon Heather, Peter Bloomfield, and John Ashburner for their assistance with the computer analysis; Claire Taylor, Andreanna Williams, and Graham Lewington for their technical support; and the radiochemistry department for the provision of tracers.
Presented in part at the 64th Scientific Sessions of the American Heart Association, Anaheim, Calif, November 11-14, 1991, and published in abstract form (Circulation. 1991;84[suppl II]:II-47).
- Received October 16, 1995.
- Revision received April 30, 1996.
- Accepted May 6, 1996.
- Copyright © 1996 by American Heart Association
Phelps ME, Mazziotta JC, Schelbert HR. Positron Emission Tomography and Autoradiography: Principles and Applications for the Brain and the Heart. New York, NY: Raven Press; 1986.
Schon HR, Schelbert HR, Najafi A, Robinson G, Huang SC, Banio J, Phelps ME. C-11 labeled palmitic acid for the noninvasive evaluation of regional myocardial fatty acid metabolism with positron emission tomography, 1: kinetics of C-11 palmitic acid in normal myocardium. Am Heart J.. 1982;103:532-547.
Buxton DB, Schwaiger M, Nguyen A, Phelps ME, Schelbert HR. Radiolabelled acetate as a tracer of myocardial tricarboxylic acid cycle flux. Circ Res. 1988;63:628-634.
Brown M, Marshall DR, Sobel BE, Begmann SR. Delineation of myocardial oxygen utilization with carbon-11-labeled acetate. Circulation. 1987;76:687-696.
Buxton DB, Nienaber CA, Luxen A, Ratib O, Hansen H, Phelps ME, Schelbert HR. Noninvasive quantitation of regional myocardial oxygen consumption in vivo with [1-11C]-acetate and dynamic positron emission tomography. Circulation. 1989;79:134-142.
Armbrecht JJ, Buxton DB, Schelbert HR. Validation of [1-11C] acetate as a tracer for noninvasive assessment of oxidative metabolism with positron emission tomography in normal, ischemic, postischemic, and hyperemic canine myocardium. Circulation. 1990;81:1594-1605.
Buck A, Wolpers G, Hutchins GD, Savas V, Mangner TJ, Nguyen N, Schwaiger M. Effect of carbon-11 acetate recirculation on estimates of myocardial oxygen consumption by PET. J Nucl Med. 1991;32:1950-1957.
Henes CG, Bergmann SR, Walsh MN, Sobel BE, Geltman EM. Assessment of myocardial oxidative metabolic reserve with positron emission tomography and carbon-11 acetate. J Nucl Med. 1989;30:1489-1499.
Walsh MN, Geltman EM, Brown MA, Henes CG, Weinheimer CJ, Sobel BE, Bergmann SR. Noninvasive estimation of regional myocardial oxygen consumption by positron emission tomography with carbon-11 acetate in patients with myocardial infarction. J Nucl Med. 1989;30:1798-1808.
Armbrecht JJ, Buxton DB, Brunken RC, Phelps ME, Schelbert HR. Regional myocardial oxygen consumption determined noninvasively in humans with [1-11C] acetate and dynamic positron emission tomography. Circulation. 1989;80:863-872.
Iida H, Rhodes CG, Yamamoto Y, Jones T, De Silva R, Araujo LI. Quantitative measurements of myocardial metabolic rate of oxygen (MMRO2) in man using 15O2 inhalation and positron emission tomography. Circulation. 1990;82(suppl III):III-614. Abstract.
Iida H, Rhodes CG, Araujo LI, Yamamoto Y, De Silva R, Maseri A, Jones T. Noninvasive quantification of regional myocardial metabolic rate for oxygen by use of 15O2 inhalation and positron emission tomography: theory, error analysis, and application in humans. Circulation. 1996;94:792-807.
Araujo LI, Lammertsma AA, Rhodes CG, McFalls EO, Iida H, Rechavia E, Galassi AR, De Silva R, Jones T, Maseri A. Noninvasive quantification of regional myocardial blood flow in coronary artery disease with oxygen-15 labeled carbon dioxide inhalation and positron emission tomography. Circulation. 1991;83:875-885.
Wallenstein S, Zucker CL, Fleiss JL. Some statistical methods useful in circulation research. Circ Res. 1980;47:639-652.
Messer JV, Wagman RJ, Levine HJ, Neill WA, Krasnow N, Gorlin R. Patterns of human myocardial oxygen extraction during rest and exercise. J Clin Invest. 1962;41:725-735.
Leight L, DeFazio V, Talmers FN, Regan TJ, Hellmens HK. Coronary blood flow, myocardial oxygen consumption and myocardial metabolism in normal and hyperthyroid human subjects. Circulation. 1956;14:90-99.
Klocke FJ, Kaiser GA, Ross J Jr, Braunwald E. An intrinsic adrenergic vasodilator mechanism in the coronary vascular bed of the dog. Circ Res. 1965;16:376-382.
Crystal GJ, Downey HF, Bashour HF. Small vessel and total coronary blood volume during intracoronary adenosine. Am J Physiol.. 1981;241:H194-H201.
Parsons WJ, Rembert JC, Bauman RP, Duhaylongsod FG, Greenfield JC, Piantadosi CA. Myocardial oxygenation in dogs during partial and complete coronary artery occlusion. Circ Res. 1993;73:458-464.
Weiss HR, Sinha AK. Regional oxygen saturation of small arteries and veins in the canine myocardium. Circ Res. 1978;42:119-126.
Weiss HR, Neubauer JA, Lipp JA, Sinha AK. Quantitative determination of regional oxygen consumption in the dog heart. Circ Res. 1978;42:394-401.
Parker JA, Beller GA, Hoop B, Holman BL, Smith TW. Assessment of regional myocardial blood flow and regional fractional oxygen extraction in dogs, using 15O-water and 15O-hemoglobin. Circ Res. 1978;42:511-518.
Bol A, Iida H, Essarmi B, Vanbutsele R, Labar D, Grandin C, Wijns W, Melin J. Assessment of myocardial oxidative reserve with PET: comparison of C-11 acetate kinetics with quantitation of metabolic rate of oxygen (MRO2) using O-15 O2. J Nucl Med. 1991;32(suppl):988. Abstract.