| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Experimental Studies |
1 From the Department of Radiology, Division of Nuclear Medicine, Johns Hopkins University Hospital, 601 N Caroline St, Rm 3223, Baltimore, MD 21287-0817. Received March 27, 2002; revision requested June 10; final revision received October 11; accepted October 23. Address correspondence to R.L.W. (e-mail: rwahl@jhmi.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Fluorine 18 fluorodeoxyglucose (FDG) was mixed with different concentrations of contrast agent with the same syringe (phantom 1), and the phantom was scanned. After image reconstruction with various attenuation maps, radioactivity concentrations were compared. Then, FDG solutions with (phantom 2) or surrounded by (phantom 3) various concentrations of contrast agent were scanned repeatedly, and radioactivity concentration was compared. Finally, PET and CT with and without contrast agent were performed in a dog. PET images were reconstructed by using different attenuation maps, and radioactivity concentrations were compared. The radioactivity concentration on germanium 68 (68Ge)-based corrected images was regarded as standard, and percentage bias, defined as difference divided by measured activity of 68Ge-based corrected images, was assessed. The relationship between the concentration of contrast agent and the percentage bias was assessed with the Pearson coefficient r, and the significance of correlations was evaluated with the Fisher z test.
RESULTS: All phantom studies demonstrated that presence of a contrast agent resulted in overestimation of emission data. CT numbers showed a strong positive correlation with the percentage bias in phantoms 2 (r = 0.999) and 3 (r = 0.987); the maximum percentage bias at 1,360 HU reached approximately 45%. These effects were independent of FDG concentration. In a canine model, presence of a contrast agent also increased emission activity, but the percentage bias was less than 15% in the liver and smaller in all other organs except the kidney (26%).
CONCLUSION: High concentrations of a contrast agent caused considerable overestimation of apparent tracer activity in phantom studies; however, the emission bias was relatively modest in vivo, except in areas with very high contrast agent concentrations.
© RSNA, 2003
Index terms: Animals • Computed tomography (CT), contrast media, **.12111, **.12112, **.121142 • Contrast media, effects • Phantoms • Positron emission tomography (PET), **.12163, **.12166
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
CT scanning in the PET/CT examination has generally been performed without intravenous administration of a contrast agent, because the influence of attenuation factors by contrast agents for PET imaging is unknown. Since intravenous contrast agents have widely been used in clinical CT and are considered desirable for interpretation of images of many areas of the body, CT images without intravenous contrast agents may be suboptimal for diagnosis, even in a PET/CT examination. The use of intravenous contrast agents is expected to affect the CT measurements of attenuation correction in the following ways: (a) the intravenous contrast agent has a high attenuation at x-ray energies, which can cause too high an estimate of tissue attenuation when translated to a 511-keV attenuation map, (b) the attenuation of an intravenous contrast agent changes rapidly after injection, and the sequential transmission and emission scans would be expected to have different distributions. The effects of contrast agents could be avoided by performing contrast material–enhanced scanning for fusion after unenhanced CT for attenuation correction; however, if contrast-enhanced CT can also be used for attenuation correction, the radiation exposure of the patient could be substantially reduced. Moreover, if it is possible to perform contrast-enhanced CT before PET without adversely affecting the PET images, it could shorten total examination time, which would allow more patient throughput.
The purpose of this study was to investigate the effects of intravenous contrast agents on the quantitative emission values obtained with a PET/CT scanner by using several phantoms and a dog.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
Phantom 1 Study
Preparation of phantom.—Phantom 1 consisted of a 60-mL syringe attached to a three-way stopcock that was rigidly attached to a stable plastic mounting platform (Fig 1). This mounting was devised to minimize phantom movement among the scans. The three-way stopcock was used to vary contrast agent concentrations in the phantom to study the effects of varying the CT attenuation correction maps applied to the same emission data sets. This phantom was designed to evaluate different attenuation maps for the identical emission data sets.
|
Data analysis.—For one data set made with the five reconstructed images, 4 x 4 pixels (1.56 x 1.56 cm) of regions of interest (ROIs) were placed (Y.N.) over the identical locations, and the mean values (kBq/mL) were obtained. We also evaluated the corresponding CT number (in Hounsfield units) by placing the same ROIs on the CT images. The exact corresponding location was acquired by using the commercial fusion software on the scanner’s workstation (eNTEGRA; Elgems, Haifa, Israel). To assess the difference, the percentage bias was calculated with the following formula: (XCT - XGe)/XGe x 100, where XCT was the ROI radioactivity concentration value determined from the CT-based corrected image and XGe was the corresponding radioactivity concentration value determined from the 68Ge-based corrected image.
Phantom 2 Study
Preparation of phantom.—Eight 50-mL syringes with varying concentrations of iohexol (solution 0, no contrast agent; solutions 1–7, 1:1,000, 1:500, 1:100, 1:50, 1:20, 1:10, and 1:5 dilution of contrast agent), designed to cover the realistic range of concentration achievable in vivo, were mixed with a standard concentration of FDG (74 kBq/mL). CT scans and 5-minute transmission scans with 68Ge were obtained. Then, 5-minute emission scans were repeated every 30 minutes, until 240 minutes after injection. Image reconstruction was performed with OSEM with segmented attenuation correction by using a 68Ge-based correction map or with OSEM with measured attenuation correction by using a CT-based correction map at each time point. Thus, a total of 18 imaging files were obtained.
Data analysis.—ROIs were placed (Y.N.) in the center of the phantom (3 x 3 pixels, 1.17 x 1.17 cm) for the intermediate three sections. The radioactivity concentrations and CT numbers obtained with this technique were determined.
Phantom 3 Study
Preparation of phantom.—In clinical CT imaging, liver tumors are often detected as areas of low attenuation with or without peripheral enhancement. To investigate the influence of the contrast agent surrounding the radioactivity, another phantom was designed. Fifty milliliters of various concentrations of iohexol (solution 0, no contrast agent; solutions 1–7, 1:1,000, 1:500, 1:100, 1:50, 1:20, 1:10, and 1:5 dilution of contrast agent) were prepared. Then, 10-mL syringes containing 6 mL of FDG solution (74 kBq/mL) were inserted into an empty 50-mL test tube. The inside of the 50-mL test tube was filled with various concentrations of the contrast agent to eliminate any remaining residual air bubbles. CT scans and 5-minute transmission scans with 68Ge were obtained, followed by 5-minute emission scans, which were repeated eight times every 30 minutes. Image reconstruction was performed with OSEM with segmented attenuation correction by using a 68Ge-based correction map or with OSEM with measured attenuation correction by using a CT-based correction map at each time point. Thus, 16 imaging files were obtained.
Data analysis.—ROIs were placed (Y.N.) in the center of the inner syringe in the phantom (3 x 3 pixels, 1.17 x 1.17 cm) for the intermediate three sections and in the outer syringe area that contained various concentrations of contrast agent. The radioactivity concentrations and CT numbers obtained with this technique were determined.
In Vivo Canine Study
Preparation of study.—To validate in vitro findings in vivo, a mongrel dog with a weight of 27 kg (Bruce Rotz Kennel, Shippensburg, Pa) was imaged with general anesthesia. Approximately 1 hour after 236.8 MBq (6.4 mCi) of FDG was injected intravenously, 5-minute emission scans per bed position were obtained at four bed positions (emission 1), and afterward, 3-minute transmission scans were obtained with 68Ge at the same bed positions, which were the same parameters as with a clinical protocol. Then, CT scans of the same region were obtained without the use of contrast material. Subsequently, 54 mL of undiluted iohexol was administrated intravenously at a rate of about 2 mL/sec. CT scanning was then repeated at 40 seconds, 2 minutes, and 8 minutes after the start of injection. Finally, 5-minute emission scanning was repeated at all four levels (emission 2). All CT images were obtained in the suspended end-expiratory phase by pausing the ventilator, as this allowed us to acquire images at the same tidal volume in the lung to obtain similar images of the same section at different time points and because end-expiratory images have been shown to best match the position of the CT- and 68Ge-based transmission images (15).
The study was conducted with the approval of our institutional animal care and use committee and complied with the Guide for the Care and Use of Laboratory Animals (National Institutes of Health, Publication No. 80-23, 1985).
Data analysis.—Two emission data sets (emission 1 and emission 2) were reconstructed with OSEM. Attenuation correction of each data set was performed by using 68Ge and CT without contrast material and at three phases of CT scanning with contrast material. Thus, we generated a total of 10 emission imaging files. We placed 3 x 3 pixel ROIs (1.17 x 1.17 cm) in left ventricle, skeletal muscles adjacent to the spine, renal cortex, renal pelvis, and physiologic but intense bowel uptake. A larger ROI (4 x 4 pixels, 1.56 x 1.56 cm) was applied to the liver, excluding vessels such as hepatic or portal. All ROIs were placed by the same author (Y.N.). The CT number of each organ was measured to assess the level of contrast enhancement. The percentage bias in the activity of CT-based corrected images was calculated with reference to that of the 68Ge-based corrected images, as previously defined.
Statistical Analysis
The relationships between the concentrations of contrast material and the percentage bias were assessed by means of the Pearson coefficient r and were plotted with a linear regression equation by using computerized statistical software (StatView, version 5.0.1; SAS Institute, Cary, NC). The significance of the correlations was assessed with the Fisher z test. P values less than .05 were considered to indicate a significant difference.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
In all phantom studies, the presence of contrast material on the CT image used for attenuation correction resulted in overestimation of PET emission values; however, the percentage bias of overestimation was only 10%–15% in the presence of contrast material at concentrations of up to 200 HU. This overestimation was independent of tracer activity concentration. There are some differences in the results between phantom 2 and phantom 3 studies. These differences may be caused by the different structures of the phantoms or may be due to different CT numbers of the same concentration of contrast material caused by variations in dilutions. However, two phantom studies substantially demonstrated similar results; percentage bias increased with increased concentration of contrast material, and radioactivity did not affect the percentage bias.
Consistent with the phantom study findings were the findings of the canine study, which confirmed that PET emission data had a tendency to be overestimated in regions of contrast enhancement, including the liver and the kidneys. The overestimation of tracer activity in the liver in emission 1 was 14.8% (124 HU) during the early phase and 11.8% (93 HU) during the delayed phase 1. The overestimation in renal cortex in emission 2 was 25.5% (192 HU) during delayed phase 1 and 24.6% (181 HU) during delayed phase 2. These differences in percentage bias are not likely to change most visual interpretations, which is widely accepted in clinical PET evaluation, although a difference of this magnitude could affect quantitative data.
Clinically relevant regions that are affected by contrast material at CT include many lesions, the head and neck, the mediastinum, the abdomen, and the pelvis. The concentration of contrast material in arteries, such as the aorta, is high in the arterial phase of CT scanning (<300 HU) (16); however, for most whole-body CT scans in PET/CT performed to aid in tumor localization, the predominant arterial phase is not preferred, rather, the slightly delayed phase (arteriovenous phase) would be preferred because opacification of vessels, arteries, and veins is most desirable. Tumor lesions in soft tissues and lymph nodes do not typically enhance at the equilibrium phase, and therefore, their FDG activity will be little affected. The muscles and soft tissues showed no significant overestimation of FDG activity with use of contrast material. The major vessels are expected to have the highest CT contrast enhancement, and their identification would aid in tumor localization. Overcorrection of activity in vessels may not affect tumor diagnosis because of their relatively low FDG uptake at 1 hour after injection and moderate arterial CT attenuation values at equilibrium phase imaging. Nevertheless, tumor-to-blood ratios could decline even slightly owing to increased blood tracer levels, which could affect the diagnostic criteria for pulmonary lesions, especially in quantitative analysis. This issue should be addressed in human studies.
For lesions within organs, such as the liver, that enhance with contrast material, the overcorrection of normal liver during the early phase was relatively small (15%). The liver is an organ in which relatively high physiologic FDG accumulation is usually seen at 1 hour after injection, and it is also well enhanced at the arteriovenous phase on CT scans. When contrast material is injected at a rate of 5 mL/sec, peak attenuation of the liver parenchyma reached about 150 HU (17). In our canine study, the CT number in the liver increased to more than 120 HU. Findings of our previous study of 28 patients, in which we compared CT-based and 68Ge-based corrected images demonstrated up to 20%–40% bias in noncontrast-enhanced CT-based corrected PET images in several patients (18), therefore, the percentage bias due to intravenous contrast material was relatively small. In patients with liver tumors, a 15% overestimation bias of the normal hepatic parenchyma could result in a loss of contrast enhancement, and therefore tumor-to-background ratios could be reduced by the use of contrast material in PET imaging. Further investigations are needed to confirm this issue with larger populations of patients with liver tumors.
Regions of very high concentration of contrast material include the kidney, the ureter, and the bladder. For these organs, FDG activity is also extremely high, and overestimation of activity was also high, especially in emission 2. PET evaluation in these regions is always difficult due to physiologic excretion; therefore, an expected large overestimation in these organs is unlikely to make a significant difference in diagnosis. Moreover, overestimated high renal FDG activity, which should be showing a diffuse, not focal, pattern, may not be meaningful in the absence of a renal mass detected at CT.
In the region of the bowel, our results showed only small changes (<5% bias) with use of contrast material. Because oral contrast material is now used for PET/CT examinations in our institute, and coregistered CT helps identify bowel location, this small overestimation is less likely to affect visual diagnosis.
In the canine investigation, we did not evaluate the direct influence of intravenous contrast material on PET emission scans. Emission data may be influenced by contrast material being present at emission scanning due to scatter and attenuation. Two consecutive emission scans were obtained before and after the CT scan for attenuation correction, but since it took a considerable time to perform the repeat CT scanning, we were unable to compare directly between the pre- and postCT emission scan because of the physiologic kinetics of FDG. Since we cannot perform PET emission scanning before and after administration of contrast material at exactly the same time, we cannot determine whether the difference was caused only by the existence of contrast material, although findings of in vitro phantom studies demonstrated that the concentration of radioactivity did not affect the percentage bias. This is a limitation of our study.
In emission 2, the overestimation of tracer activity in the renal cortex and pelvis was remarkable, but the clinical effect of tracer uptake in regions of excreted urine in the renal region is doubtful. In the liver, underestimation rather than overestimation of FDG activity was observed on images corrected at CT without contrast material and at CT with contrast material performed in the delayed phase, and a positive correlation was seen between CT numbers and the measured radioactivity levels. However, the absolute percentage bias was estimated at less than 10%. Since overestimation caused by the presence of contrast material in CT attenuation correction was considered a major problem, mild underestimation caused by the presence of contrast material should be noted but may not be of great clinical relevance.
In this study, we evaluated PET images reconstructed by means of iterative reconstruction (OSEM) with segmented attenuation correction. If we investigated quantitative values on PET images reconstructed by means of the conventional filtered back-projection method, the results might be different, because OSEM with segmented attenuation correction often shows mild underestimation compared with the filtered back projection with measured attenuation correction (19). Nevertheless, we demonstrated only data based on OSEM with segmented attenuation correction, because now it is the most commonly accepted reconstruction method in clinical practice.
In conclusion, the presence of contrast material on CT scans used for attenuation correction may affect the measured radioactivity concentration, especially in well-enhanced regions. However, if CT scanning is performed in the arteriovenous phase, which is considered to be the best timing for fusion of PET and CT images, the effects from contrast material would be modest in most tissues on the basis of our preliminary data. Since the liver is an enhancing area in human body, where physiologic FDG uptake is also seen, overestimation of well-enhanced normal hepatic parenchyma might cause slight reduction in tumor-to-liver ratio. Similarly, tumor-to-blood ratios might decline with contrast enhancement, which could affect the diagnostic criteria for pulmonary lesions. Further evaluations in humans are needed to confirm these issues.
Practical application: The presence of intravenous contrast material may affect the quantitative emission values at PET in PET/CT examinations, but they should be modest when transmission CT scanning is performed in the arteriovenous phase, which can show anatomic information most effectively. Lesion-to-background ratios could be reduced in liver metastases, where normal hepatic parenchyma should be overestimated, while uptake in the metastatic liver tumors should be less influenced by the presence of contrast material. Similarly, tumor-to-blood ratios could be expected to be reduced slightly. The effects will be most noticeable when quantitation is applied. These issues warrant evaluation in clinical studies.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Abbreviations: FDG = fluorodeoxyglucose, OSEM = ordered-subset expectation maximization, ROI = region of interest
Author contributions: Guarantor of integrity of entire study, Y.N.; study concepts, Y.N., B.B.C.; study design, Y.N., B.B.C., R.L.W.; literature research, Y.N.; experimental studies, Y.N., B.B.C., D.L.K., L.P.L., L.T.M.; data acquisition, Y.N., B.B.C., D.L.K.; data analysis/interpretation, Y.N., B.B.C.; manuscript preparation and editing,, Y.N., B.B.C.; manuscript definition of intellectual content, Y.N., B.B.C., R.L.W.; manuscript revision/review, B.B.C., D.L.K., R.L.W.; manuscript final version approval, R.L.W.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODS RESULTS DISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. J. Gollub, R. Hong, D. M. Sarasohn, and T. Akhurst Limitations of CT During PET/CT J. Nucl. Med., October 1, 2007; 48(10): 1583 - 1591. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. A. Weber and R. Figlin Monitoring Cancer Treatment with PET/CT: Does It Make a Difference? J. Nucl. Med., January 1, 2007; 48(1_suppl): 36S - 44S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. Kuehl, P. Veit, S. J. Rosenbaum, A. Bockisch, and G. Antoch Can PET/CT Replace Separate Diagnostic CT for Cancer Imaging? Optimizing CT Protocols for Imaging Cancers of the Chest and Abdomen J. Nucl. Med., January 1, 2007; 48(1_suppl): 45S - 57S. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Rodriguez-Vigil, N. Gomez-Leon, I. Pinilla, D. Hernandez-Maraver, J. Coya, L. Martin-Curto, and R. Madero PET/CT in Lymphoma: Prospective Study of Enhanced Full-Dose PET/CT Versus Unenhanced Low-Dose PET/CT J. Nucl. Med., October 1, 2006; 47(10): 1643 - 1648. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Blake, A. Singh, B. N. Setty, J. Slattery, M. Kalra, M. M. Maher, D. V. Sahani, A. J. Fischman, and P. R. Mueller Pearls and Pitfalls in Interpretation of Abdominal and Pelvic PET-CT RadioGraphics, September 1, 2006; 26(5): 1335 - 1353. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Delbeke, R. E. Coleman, M. J. Guiberteau, M. L. Brown, H. D. Royal, B. A. Siegel, D. W. Townsend, L. L. Berland, J. A. Parker, G. Zubal, et al. Procedure Guideline for SPECT/CT Imaging 1.0 J. Nucl. Med., July 1, 2006; 47(7): 1227 - 1234. [Full Text] [PDF]
|
|
|
|
|
|
D. Delbeke, R. E. Coleman, M. J. Guiberteau, M. L. Brown, H. D. Royal, B. A. Siegel, D. W. Townsend, L. L. Berland, J. A. Parker, K. Hubner, et al. Procedure Guideline for Tumor Imaging with 18F-FDG PET/CT 1.0 J. Nucl. Med., May 1, 2006; 47(5): 885 - 895. [Full Text] [PDF]
|
|
|
|
 |
|
 O. Mawlawi, J. J. Erasmus, R. F. Munden, T. Pan, A. E. Knight, H. A. Macapinlac, D. A. Podoloff, and M. Chasen Quantifying the Effect of IV Contrast Media on Integrated PET/CT: Clinical Evaluation Am. J. Roentgenol., February 1, 2006; 186(2): 308 - 319. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Ames, T. Blodgett, and C. Meltzer 18F-FDG Uptake in an Ovary Containing a Hemorrhagic Corpus Luteal Cyst: False-Positive PET/CT in a Patient with Cervical Carcinoma Am. J. Roentgenol., October 1, 2005; 185(4): 1057 - 1059. [Full Text] [PDF]
|
|
|
|
|
|
R. E. Coleman, D. Delbeke, M. J. Guiberteau, P. S. Conti, H. D. Royal, J. C. Weinreb, B. A. Siegel, M. F. Federle, D. W. Townsend, and L. L. Berland Concurrent PET/CT with an Integrated Imaging System: Intersociety Dialogue from the Joint Working Group of the American College of Radiology, the Society of Nuclear Medicine, and the Society of Computed Body Tomography and Magnetic Resonance J. Nucl. Med., July 1, 2005; 46(7): 1225 - 1239. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. F. Branstetter IV, T. M. Blodgett, L. A. Zimmer, C. H. Snyderman, J. T. Johnson, S. Raman, and C. C. Meltzer Head and Neck Malignancy: Is PET/CT More Accurate than PET or CT Alone? Radiology, May 1, 2005; 235(2): 580 - 586. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-Y. Yau, W.-S. Chan, Y.-M. Tam, P. Vernon, S. Wong, M. Coel, and S. K.-F. Chu Application of Intravenous Contrast in PET/CT: Does It Really Introduce Significant Attenuation Correction Error? J. Nucl. Med., February 1, 2005; 46(2): 283 - 291. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Gutzeit, G. Antoch, H. Kuhl, T. Egelhof, M. Fischer, E. Hauth, S. Goehde, A. Bockisch, J. Debatin, and L. Freudenberg Unknown Primary Tumors: Detection with Dual-Modality PET/CT--Initial Experience Radiology, January 1, 2005; 234(1): 227 - 234. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Antoch, L. S. Freudenberg, T. Beyer, A. Bockisch, and J. F. Debatin To Enhance or Not to Enhance? 18F-FDG and CT Contrast Agents in Dual-Modality 18F-FDG PET/CT J. Nucl. Med., January 1, 2004; 45(90010): 56S - 65. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. L. Wahl Why Nearly All PET of Abdominal and Pelvic Cancers Will Be Performed as PET/CT J. Nucl. Med., January 1, 2004; 45(90010): 82S - 95. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
), muscle (
), liver (
), renal cortex (x), renal pelvis (+), and bowel (
) are demonstrated. In both emission 1 and 2, strong positive correlations between CT numbers and the percentage bias were observed in all organs except muscle; however, the pattern of emission 2 was not equivalent with the emission 1 data.