| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Basic Science Investigations |
1 Positron Emission Tomography Department, Clinical Center, National Institutes of Health, Bethesda, Maryland
2 Istituto di Scienze Radiologiche, Università degli Studi di Milano, Milano, Italy
3 Laboratory of Cerebral Metabolism, National Institute of Mental Health, Bethesda, Maryland
4 Hospital General Universitario Gregorio Marañón, Madrid, Spain
5 Department of Nuclear Medicine, Clinical Center, National Institutes of Health, Bethesda, Maryland
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
48 MBq (1.3 mCi) 18F-FDG were performed in rats anesthetized with isoflurane. Arterial blood sampling was performed throughout the scanning period. At 60 min the rat was killed, and the brain was rapidly removed and dissected into 5 structures (thalamus [TH], cortex [CX], brain stem [BS], cerebellum [CB], and half brain). Activity in the tissue samples was compared with the mean activity of the last 5 min of calibrated PET data. Results: Plasma activity peaked at 0.2 min and then cleared rapidly. Brain activity initially rose rapidly; the rate of increase then progressively slowed until activity was approximately constant between 30 and 60 min. Recovery coefficients (MBq/mL in PET images)/(MBq/mL in tissue samples) were 0.99 ± 0.04, 0.90 ± 0.19, 1.01 ± 0.24, 0.84 ± 0.05, and 1.01 ± 0.17, respectively, in TH, CX, BS, CB, and half brain (mean ± SD, n = 6–9). Cerebral glucose utilization determined by Patlak analyses of PET data measured 30–60 min after injection of 18F-FDG was 31.7 ± 5.2, 23.9 ± 4.8, 29.9 ± 5.0, 39.3 ± 7.3, and 28.1 ± 4.6 µmol/100 g/min (mean ± SD, n = 9) in TH, CX, BS, CB, and whole brain, respectively. These results are consistent with a previous 14C-deoxyglucose study of the isoflurane-anesthetized rat. Conclusion: Expected values for glucose metabolic rates and recovery coefficients near unity suggest that quantitatively accurate dynamic 18F-FDG brain imaging can be performed in the rat with arterial blood sampling and the ATLAS small animal PET scanner.
Key Words: small animal imaging • 18F-FDG • arterial blood sampling • dynamic scan • glucose utilization • brain
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
Fully quantitative PET methods require not only accurate measurement of tissue activity but also a measure of the amount of radiotracer delivered to the tissue during the study interval, the so-called "input function." This represents a second challenge in small animal PET imaging. Sampling of arterial blood provides the most direct and appropriate measurement of the input function, although other less direct measures—such as sampling blood from other mice at specified time points and applying that to the experimental group (2), direct sampling of retroorbital blood from a mouse (3), blood sampling from the left ventricle plus PET scanning of the heart and liver of a mouse (4), and an automated microvolumetric blood sampler to draw the blood from the jugular vein in a mouse (5)—have been used. External measurements of blood activity—for example, from the left ventricle—cannot be used when radiolabeled species other than the parent compound are present in plasma, such as is the case with many receptor ligands, or when the radiotracer does not rapidly equilibrate between red blood cells and plasma to a known equilibrium distribution ratio. An alternative approach used in large animal PET scanning is to infer the input function through measurements in a reference tissue. Accurate determination of the activity in the reference tissue may be more difficult in small animals due to the relatively greater partial-volume effects in small animal imaging. Arterial blood sampling, when feasible, remains the most direct and accurate means of measuring the input function.
In response to the challenge of developing a high-sensitivity, high-resolution small animal PET scanner, the Imaging Physics Laboratory of the Nuclear Medicine Department, Clinical Center, National Institutes of Health, constructed the Advanced Technology Laboratory Animal Scanner (ATLAS) (6). In the present study, we explored the feasibility of using the ATLAS scanner for fully quantitative PET studies in anesthetized rats with rapid dynamic scanning and arterial blood sampling. In a feasibility study, we chose the well-known radiotracer 18F-FDG, which is used for determination of regional rates of glucose utilization in all areas of the brain. Measurements of tissue activities determined with the ATLAS scanner were validated by comparison with those obtained by direct counting of tissue samples dissected immediately after completion of PET scanning. To demonstrate the feasibility of using time-resolved tissue data, regional cerebral glucose utilization (rCMRglu) was determined from the arterial plasma and brain tissue time-activity curves by the multiple-time graphical analysis method of Patlak et al. (7,8). Additionally, we have performed a postmortem static scan of an anesthetized animal and an awake animal, in both of which 18F-FDG had been administered 45 min previously, to illustrate the effect of isoflurane anesthesia on tracer uptake because it is well known that volatile anesthetics affect cerebral glucose utilization (9).
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
System Characteristics of ATLAS Small Animal PET Scanner
All PET images shown in this article were obtained with the ATLAS scanner. ATLAS has an 11.8-cm ring diameter, an 8-cm aperture, a 6-cm effective transverse field of view, and a 2-cm axial field of view. The scanner consists of 18 "phoswich," or depth-of-interaction, detector modules surrounding the imaging volume. The spatial resolution of this system is 1.8-mm full width at half maximum (FWHM) in the central field of view with filtered backprojection reconstruction and better than 1.5-mm FWHM when 3-dimensional ordered-subset expectation maximization (3D OSEM) reconstruction is used with resolution recovery and exact positioning of the lines of response (1). Sensitivity is 1.8% using a window of 250–650 keV and 2.7% with a window of 100–650 keV (6).
For scanner activity calibration, a cylinder 3.1 cm in diameter by 7.5-cm long was used to simulate the rat head and upper torso. This cylinder was filled with 18F in water, centered in the scanner aperture, and imaged. The radioactivity in the cylinder was determined by a dose calibrator, and this value was divided by the volume of the cylinder to give concentration. Given these data, a calibration factor (C) was determined that related image counts to tracer concentration in the animal:
 |
Animal Preparation
Normal adult male Sprague-Dawley rats (220–268 g) were purchased from Charles River Laboratories and maintained in a climate-controlled room on a 12-h light:dark cycle with food and water available ad libitum. They were deprived of food but allowed free access to water for 16 h before surgical preparation for the experiments. All animal procedures were in strict accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and approved by the Clinical Center Animal Care and Use Committee. Animals were anesthetized with halothane (5% induction and 1%–1.5% for maintenance in 30% O2/70% N2O). Polyethylene catheters (PE 50; Clay Adams) were inserted into left and right femoral arteries and the left femoral vein. One arterial catheter was used for continuous monitoring of mean arterial blood pressure (MABP) and the other was used for sampling of arterial blood. The length of the catheter used for blood sampling was fixed precisely at 16 cm to minimize sampling errors due to catheter delay and dispersion. The venous catheter was used for injection of 18F-FDG. The catheters were buried under the skin in a tunnel that was passed from the femoral area to the area behind the neck, and the tips of the catheters were located under the skin in the area behind the neck. This placement of the catheters allows the animal to move freely without damaging the catheters. After the catheterizations, surgical wounds were treated with 5% lidocaine ointment and sutured. Duration of anesthesia during the surgical preparation was usually about 30 min. Throughout the surgical preparations and the experimental periods, body temperature was monitored by a rectal probe and maintained at 37°C by a thermostatically controlled heating lamp (model 73 A; Yellow Springs Instrument Co., Inc.). At least 3 h were allowed for recovery from the surgery and anesthesia before the PET study. The surgical procedures were successful in >90% of the studies.
Physiologic Variables
In 4 animals, MABP was monitored continuously with a Micro-Med Blood Pressure Analyzer (model 300; Micro-Med) that had been calibrated with an air-damped mercury manometer. In all animals, arterial blood partial pressure of CO2 (PCO2), partial pressure of O2 (PO2), and pH were measured with a blood-gas analyzer (model 288 Blood Gas System; Ciba-Corning Diagnostics Corp.). Hematocrit was determined in arterial blood samples centrifuged in a Microfuge B (Beckman Instruments). Arterial plasma glucose concentration was determined with a Beckman Glucose Analyzer 2 (Beckman Instruments). Physiologic variables other than MABP were measured immediately before injection of the radiotracer and again approximately 35 min later.
PET Study
PET scans were performed in 9 rats. After at least 3-h recovery from catheterization surgery, the rat was reanesthetized with isoflurane (5% for induction and 2% for maintenance in 100% O2) using a nose cone. The tunneling of the catheters was undone so that the catheters were readily accessible. After 15-min preparation, the rat was placed in a prone position on the platform of the scanner. With the help of a laser beam attached to the scanner, the rat was positioned so that the center of the field of view was 12 mm caudal to the line between the lateral edges of the eyes. 18F-FDG (49.9 ± 10.5 MBq [1.348 ± 0.284 mCi]; range, 37–62.9 MBq [1.0–1.7 mCi]; 300–700 µL in volume) was administered via the venous catheter over a period of about 15 s, and the catheter was immediately flushed with 100 µL of saline from a syringe connected with a custom-made Y-connector. Scanning was initiated at the onset of 18F-FDG administration. A sequence of one hundred twenty 30-s images was acquired; total scanning time was 60 min. After the last scan, the rat was killed with a bolus venous injection of euthanasia solution.
To obtain a high-quality image and to visualize the effects of 2% isoflurane anesthesia on the uptake of 18F-FDG, we conducted an ex vivo static scan of the head for 60 min on one rat who was awake and one rat who was anesthetized during the 45-min uptake period of 63 MBq (1.7mCi) 18F-FDG.
Arterial Blood Sampling
To determine the plasma time-activity curve, the blood sampling schedule of the quantitative autoradiographic 2-deoxy-D-14C-glucose (2-14C-DG) method (9) was followed—that is, 6 continuous samples over 15 s and 13 additional blood samples taken at progressively increasing intervals (0.5, 0.75, 1, 2, 3, 5, 7.5, 10, 15, 25, 35, 45, 60 min). Blood samples, approximately 100 µL in volume, were withdrawn into polyethylene microcentrifuge tubes coated with heparin-lithium-fluoride (Beckman Instruments, Inc.), plasma was separated from red blood cells by centrifugation, and 40 µL of plasma were counted in a 
-counter (Wizard model 1480; PerkinElmer Life Sciences Inc.). The remaining plasma was used for the measurement of glucose concentration. To minimize blood loss, all dead-space volumes of blood collected in the sampling process were reinjected intravenously into the animal.
Direct Tissue Counting
Immediately after the rat was euthanized, the brain was taken out and separated into 2 parts, each including one hemisphere and half of the CB and BS. One part was weighed and counted; its concentration of radioactivity was used to calculate the whole brain radioactivity. From the other part, 4 structures were dissected (thalamus [TH], cortex [CX], cerebellum [CB], and brain stem [BS]), weighed, and counted in a -counter. All activities were expressed in MBq/mL, assuming the specific gravity of tissue to be 1.
Analysis of PET Data
Data from the scanner were formatted into 27 time frames (8 x 30 s, 6 x 60 s, 5 x 120 s, and 8 x 300 s). Images were reconstructed by 3D OSEM (11). No correction was made for attenuation or scatter. Counts detected by the scanner were converted into MBq/mL by use of a cross-calibration factor (325.63 scanner cps/mL per MBq/mL [0.01205 scanner cps/mL per nCi/mL]) previously determined in a phantom study. Additionally, one image containing the sum of all counts collected over the full 60-min interval was constructed for placement of regions of interest (ROIs). Due to its relatively low spatial resolution, ROIs were placed on the summed image, not by use of landmarks in that image, but rather by identifying the 3D coordinates of each structure in a rat brain atlas (12) and locating the approximate corresponding position in the PET image. TH, CX, and whole brain ROIs were drawn on coronal sections; the first coronal slice was 4 mm caudal from the tip of the frontal lobe. Each slice was 1.125-mm thick. CB and BS ROIs were placed on sagittal sections. ROIs were then transferred to the image for each time frame of data, and weighted average tissue activity was computed. A background ROI was also drawn outside the brain region. In all cases, the apparent activity in this region was very small during the course of a study and reflected the small contributions from random coincidence events and from scatter. Tissue time-activity curves were constructed by subtracting background activity from activity measured in each structure. Background-corrected tissue activity in the last frame of data (55–60 min after injection) was used for comparison with the directly counted tissue samples.
To verify the reproducibility of the ROI placement, 2 different readers separately drew the ROIs on the same regions of each animal and the results were compared. The interreader variability was 4.6% at its maximum with a SD of 3.4%. In addition, both readers redrew the ROIs after 1 mo or more from the first reading and again the results were compared. Reader 1 intravariability was 2.1% at its maximum value with a SD of 4.3% and reader 2 intravariability was 0.23% ± 0.73%.
Determination of rCMRglu
The multiple-time graphical analysis technique of Patlak et al. (7,8) was used to calculate rCMRglu. This technique requires that the normalized plasma activity (
(T) = 
0T Cp*(t)dt/Cp*(T), where Cp*(t) is the arterial plasma concentration of 18F-FDG) be plotted on the abscissa and the normalized tissue activity (Ci*(T)/Cp*(T), where Ci*(T) is the tissue activity) be plotted on the ordinate. After the 18F-FDG in the tissue pool has equilibrated with that in the plasma, the graph becomes a straight line with slope K equal to the rate constant for net uptake and metabolism of 18F-FDG in the brain. In this study, the 6 frames of data acquired in the interval 30–60 min were used to estimate K. rCMRglu was then calculated from the following equation:
 |
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
-counting of ex vivo tissue samples and used the dynamic scans to determine tissue time-activity curves for determination of rCMRglu.
Physiologic Variables
Rats in the present study had a MABP slightly lower than the normal range in unanesthetized animals, but within the range found in isoflurane-anesthetized rats (13). The decrease in the hematocrit was 7% from the 19 blood samples taken (total volume, 1.9 mL); although this was statistically significant, it probably did not have a great impact on the physiologic status of the animal because the animals were at rest and, therefore, did not require increased oxygen. In studies with radiotracers that can use whole blood for measuring the input function—that is, those that do not require separation of plasma—it may be possible to further reduce the blood sample volumes. Increases in plasma glucose concentration were moderate and may be due to the stress of blood loss. The plasma glucose concentration, however, remained within the normoglycemic range. pH was lower, PCO2 was slightly higher, and PO2 was higher than the normal limits in awake animals, consistent with effects usually seen under gas anesthetics.
PET Images
PET scanning in animals requires that the animal be immobilized; as there are currently no accepted alternatives for immobilization, all animals undergoing scanning must be anesthetized. In the present study we used isoflurane, which had profound effects on glucose utilization. To illustrate the differences in rCMRglu between the awake and isoflurane-anesthetized animal, we performed a postmortem, single scan of long duration on 2 additional animals: one awake during the period of uptake of 18F-FDG and the other under isoflurane anesthesia during the uptake period. The effect of restraint on the biodistribution of 18F-FDG in awake animals is unknown. The coronal images from these scans are shown in Figure 1. In the awake rat, there is the expected heterogeneity between the gray matter and the white matter. In the cortical area of the awake rat, the auditory CX has the highest uptake of 18F-FDG (third and fourth slices in row 2). In contrast, the image of the anesthetized rat shows decreased activity and decreased heterogeneity in almost all regions. There is, however, an area that shows an increased uptake of 18F-FDG compared with the awake rat (second slice, bottom row). This is in the area of the vestibular nucleus, but precise identification of the region needs to be determined with a method that has higher spatial resolution—for example, the quantitative autoradiographic 2-14C-DG method. The effect of isoflurane anesthesia on rCMRglu in the vestibular nucleus is not presently known.
The loss of heterogeneity in the cortical area under isoflurane anesthesia is also evident in Figure 2, which shows activity in one animal integrated over the entire 60-min uptake interval of the dynamic PET acquisition. The integrated image also shows an area of very high uptake in the area of the vestibular nucleus. The less heterogeneous pattern of tracer uptake under isoflurane anesthesia is similar to that found with thiopental anesthesia (9) and reflects basal glucose metabolism. Even though ATLAS has better spatial resolution than previous-generation scanners (14), it is still difficult to determine detailed structures from the image. Therefore, in the present study, we outlined ROIs based on 3D coordinates from an atlas of the rat brain (12).
Time-Activity Curves
In the present study, we demonstrated the feasibility of performing arterial blood sampling during dynamic PET scans in rats by using the blood sampling techniques originally designed for use in the quantitative autoradiographic 14C-deoxyglucose method. As illustrated in Figure 3, plasma radioactivity increased rapidly for 0.2 min and then cleared rapidly, similar to the pattern that is observed after injection of 2-14C-DG. The present study required 1.9 mL of blood to fully characterize the input function, and it is well understood that both the volume and the frequency of the samples must be reduced for measurements in mice. 14C-Deoxyglucose studies in freely moving mice have demonstrated that glucose and 14C-deoxyglucose concentrations can be measured in micro samples to reduce the total blood sampling volume to as little as 270 µL (15). The experience developed from these studies will be helpful for adapting PET methods for use in mice.
The time-activity curves in each of the tissues examined (whole brain, CX, TH, CB, and BS) were similar in shape. They initially rose rapidly, and then the rate of increase progressively slowed until activity was approximately constant between 30 and 60 min after injection of 18F-FDG. The loss of heterogeneity is recognized in the time-activity curves from the fairly narrow range of tissue activities in the last frame of data: Activities ranged only approximately 1.5-fold. Although partial-volume effects due to the limited spatial resolution of the scanner contribute to the narrow range of activities, anesthesia plays a substantial role. In a previous quantitative 14C-deoxyglucose study, the ratio of average gray matter to average white matter rCMRglu in isoflurane-anesthetized rats was 1.95:1, whereas the conscious animals exhibited a much greater range of activities with a ratio of 2.8:1 (13).
Comparison Between Direct Tissue Counting Versus PET Image Analysis
The whole brain, TH, and BS showed the best agreement between activities measured by PET and those measured by direct counting of the tissue samples. The whole brain was expected to give the best agreement due to lesser partial-volume effects. The TH is located in the center part of the brain, and, therefore, it had a higher degree of certainty in placing the ROI. The lower recovery rate in the CB may be due to difficulty in obtaining a clean dissection separating this structure from the adjacent region that has a very high uptake of 18F-FDG, as shown in the PET images (Figs. 1 and 2). The decreased heterogeneity in the anesthesized rat brain may have reduced the partial-volume effects (13). Nevertheless, our data provide strong evidence that ATLAS can perform a dynamic 18F-FDG scan because its data match well with the direct tissue counting.
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
|
FOOTNOTES |
|---|
For correspondence contact: William C. Eckelman, PhD, Positron Emission Tomography Department, Building 10, Room 1C495, 10 Center Dr., MSC 1180, Bethesda, MD 20892.
E-mail: Eckelman{at}nih.gov
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  B. Bolon, D. C. Anthony, M. Butt, D. Dorman, M. V. Green, P. B. Little, W. M. Valentine, D. Weinstock, J. Yan, and R. C. Sills "Current Pathology Techniques" Symposium Review: Advances and Issues in Neuropathology Toxicol Pathol, October 1, 2008; 36(6): 871 - 889. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. L. Franc, P. D. Acton, C. Mari, and B. H. Hasegawa Small-Animal SPECT and SPECT/CT: Important Tools for Preclinical Investigation J. Nucl. Med., October 1, 2008; 49(10): 1651 - 1663. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. C. Nawroth, C. A. Greer, W. R. Chen, S. B. Laughlin, and G. M. Shepherd An Energy Budget for the Olfactory Glomerulus J. Neurosci., September 5, 2007; 27(36): 9790 - 9800. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. K. Schiffer, M. M. Mirrione, and S. L. Dewey Optimizing Experimental Protocols for Quantitative Behavioral Imaging with 18F-FDG in Rodents J. Nucl. Med., February 1, 2007; 48(2): 277 - 287. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. A. Foss, R. C. Mease, H. Fan, Y. Wang, H. T. Ravert, R. F. Dannals, R. T. Olszewski, W. D. Heston, A. P. Kozikowski, and M. G. Pomper Radiolabeled Small-Molecule Ligands for Prostate-Specific Membrane Antigen: In vivo Imaging in Experimental Models of Prostate Cancer Clin. Cancer Res., June 1, 2005; 11(11): 4022 - 4028. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| JOURNAL OF NUCLEAR MEDICINE TECHNOLOGY | THE JOURNAL OF NUCLEAR MEDICINE |
