| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Experimental Studies |
1 From the Depts of Diagnostic and Interventional Radiology (G.A., H.K., J.F.D.), General and Transplantation Surgery (G.M.K., H.Z., S.W., C.E.B.), Radiation Therapy (A.B.M.), Pathology (K.A.M.), and Nuclear Medicine (S.P.M., A.B.); and Institute of Medical Physics (A.B.), Univ Hosp Essen, Hufelandstrasse 55, 45127 Essen, Germany. Received Sep 24, 2002; revision requested Nov 26; final revision received Jul 21, 2003; accepted Aug 6. Address correspondence to G.A. (e-mail: gerald.antoch@uni-essen.de).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Central bile duct resection followed by creation of a biliodigestive anastomosis was performed in nine pigs. Six of these pigs were also treated with intraoperative radiation therapy (IORT) (20 Gy) in the area of the anastomosis. Two, 4, and 8 weeks postoperatively, contrast material–enhanced fluorine 18 fluorodeoxyglucose (FDG) PET/CT of the liver was performed in all of the animals. The radioactive tracer concentration in the region of the anastomosis was quantified, and the values were compared intraindividually with the values at the liver periphery. Histologic evaluation of the liver was performed 8 weeks postoperatively. The PET/CT images were assessed for changes in liver volume and bile duct diameter over time.
RESULTS: In all nine pigs, the region of the anastomosis could be clearly defined on the fused PET/CT images. PET/CT revealed a decreased concentration of FDG in the irradiated field 2 and 4 weeks after IORT. At 8 weeks, however, the distribution of the tracer in the irradiated pigs did not differ from that in the nonirradiated pigs. Homogeneous tracer uptake in all liver regions was observed in the nonirradiated animals. The CT images showed an increase in liver volume in all pigs and bile duct dilatation that increased over time in the irradiated pigs.
CONCLUSION: The morphologic and functional changes due to IORT in liver tissue can be accurately monitored with dual-modality PET/CT. By enabling the integration of functional and morphologic data, PET/CT may have an important role in monitoring radiation treatment.
© RSNA, 2004
Index terms: Bile ducts, surgery • Dual-modality imaging, PET/CT, 761.12162, 761.12163, 765.12162, 765.12163 • Experimental study • Liver, CT, 761.12111, 761.12112, 761.12115 • Therapeutic radiology, intraoperative
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Malignant tumors are often characterized by increased glucose metabolism, which leads to the focal uptake of the radioactively labeled tracer in the tumor. When determining the site of residual tumor or tumor recurrence with FDG PET, however, the lack of morphologic information frequently leads to inaccuracies (4). Furthermore, identification of the irradiated field is known to be difficult when it is based on functional information alone. Software-based fusion of morphologic and functional data sets has been proved to be beneficial for brain imaging (5,6). Coregistration inaccuracies due to respiration and different patient positioning on the examination table, however, have limited the use of these fused data sets in the assessment of the chest and abdomen (7). Hence, the ability to integrate morphologic and functional data by using a single imaging system is highly desirable.
Dual-modality PET/computed tomographic (CT) imaging yields morphologic and functional data sets that are acquired in a single examination. One examination table includes both CT and PET components to ensure accurate image coregistration (8). Furthermore, attenuation correction of the PET emission data can be performed on the basis of CT data, and this obviates a separate PET transmission scan and thereby decreases the PET image acquisition time by approximately 30% (9). Consequently, the number of patients scanned can be greatly increased.
For clinically reliable monitoring of the effects of radiation therapy, PET/CT has to yield accurate morphologic and functional information about the region of radiation treatment. The following data need to be acquired: (a) the accurate size and location of the irradiated field, (b) functional data on the cells’ response to radiation treatment, (c) morphologic and functional information about the possible effects of radiation treatment on the normal tissue adjacent to the irradiated field, (d) morphologic information about changes due to surgery, and (e) accurate image fusion—that is, image fusion that results in a substantial benefit compared with the benefit of using either of the two imaging modalities alone.
The purpose of this study was to assess the value of dual-modality PET/CT for monitoring radiation therapy on the basis of the outlined data requirements. Postoperative and radiation therapy–induced effects were studied with PET/CT in a pig model and correlated with the macroscopic and histologic findings obtained after the animals were sacrificed.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Animal Model
The animal study was approved by the supervising state agency (license number G 641/01) and performed in full accordance with all state and federal guidelines. The study was conducted by using nine young female Pietrain Hamshire pigs (Ellegaard, Skalskor, Denmark) weighing 30–40 kg according to an established animal model protocol (10). Three animal groups consisting of three pigs each were chosen on the basis of the regulatory issues set forth by the supervising state agency, which limits the number of animals used for initial studies.
Three pigs were randomly assigned to each group. In group 1 (control group), hilar bile duct resection was performed and a biliodigestive anastomosis was created. In group 2, in addition to the resection-anastomosis surgery performed in group 1, intraoperative irradiation of the region of the biliodigestive anastomosis was performed by using a 5-cm-diameter radiation applicator. The animals in group 3 were treated by using the surgery and intraoperative irradiation performed in group 2, with the exception that the applicator used to administer intraoperative radiation therapy (IORT) was 7 cm in diameter.
The animals were kept in sheltered runs at a central animal facility and received a standard meal for pigs ad libitum. Before the resection-anastomosis surgery and before all postoperative imaging examinations, the animals did not receive food for a minimum of 8 hours. Their blood glucose levels were found to be in the normal range prior to surgery and all imaging procedures.
Surgical Procedure and Intraoperative Irradiation
Before the abdomen was opened by means of a midline incision from the xiphoid to the pubic bone, the animals were fully anesthetized with midazolam hydrochloride (Dormicum; Hoffmann-La Roche, Grenzach-Wyhlen, Germany), fentanyldihydrogencitrate (Fentanyl Hexal; Hexal, Holzkirchen, Germany), and propofol (Disoprivan; Astra-Zeneca, Zug, Switzerland). After exposing the hepatoduodenal ligament, we resected the gallbladder and approximately 4 cm of the extrahepatic bile duct. In six of the nine pigs, IORT in the area of resection was performed with a linear accelerator (Mevatron, Siemens Radiation Therapy Division, Concord, Mass). This irradiation system delivers electrons with an effective energy output of 8 MeV at a dose rate of 9 Gy/min. A 20-Gy dose of radiation was delivered through a 5-cm (in three pigs, group 2) or 7-cm (in three pigs, group 3) circular metal applicator. Exposure of other organs to the radiation field was avoided.
For the treated volume, we focused on the hepatic hilum, including parts of the right and left hepatic ducts, the portal vein, the hepatic artery, and the central liver parenchyma. After 20 Gy of radiation was delivered, IORT was terminated and the surgery was continued. End-to-side hepaticojejunostomy was performed by using the Roux-en-Y anastomosis procedure. Anesthesia was terminated, and the pigs were extubated and then transported back to their shelters. In all six animals, the surgical procedure combined with IORT was accomplished without immediate complications.
Imaging Procedures
Contrast material–enhanced CT and combined PET/CT were performed 2, 4, and 8 weeks postoperatively. Morphologic CT combined with functional PET was performed by using a dual-modality tomograph (biograph; Siemens Medical Solutions, Hoffman Estates, Ill). The CT component of the tomograph consists of a single-section spiral CT unit (Somatom Emotion; Siemens Medical Systems, Erlangen, Germany) with a minimal gantry rotation time of 800 msec and a maximal scanning time of 100 seconds (125 rotations). The PET component, an ECAT HR+ unit (Siemens Medical Systems), provides an in-plane spatial resolution of 4.6 mm and a transverse field of view of 15.5 cm for one table position. The maximal coaxial imaging range is 145 cm. Additional transmission scanning is not required because the attenuation correction of PET emission data is based on CT transmission information (9).
For all imaging examinations, the animals were sedated with midazolam hydrochloride (Dormicum) injected through an intravenous cannula placed in an ear vein. First, three-phase contrast-enhanced CT was performed with the animal in a supine position on all examination days (ie, 2, 4, and 8 weeks after surgery). The liver was imaged before contrast material administration, as well as during the arterial (15-second delay) and portal venous (50-second delay) phases following the intravenous injection of 70 mL of an iodinated contrast agent (iomeprol, 350 mg of iodine per milliliter, Imeron 350; Byk Gulden, Konstanz, Germany). Five-millimeter contiguous sections were acquired in a craniocaudal direction with 130 mAs, 130 kV, and a pitch of 1.6. With a table speed of 8 mm per gantry rotation, contiguous 5-mm sections were reconstructed at 2.5-mm increments.
After contrast-enhanced CT, blood samples were taken to ensure that the animals had normal blood glucose levels, and then 250 MBq of 18F FDG was injected intravenously. Combined PET/CT examinations of the liver were performed without administration of CT contrast material 15, 30, 45, and 60 minutes after injection of the radioactive tracer. The chosen scanning parameters were the same as those used for contrast-enhanced CT of the liver. At the end of spiral CT, the examination table moved automatically into the PET gantry and emission data were acquired in a caudocranial direction. The time to scan a single table position at PET was set at 5 minutes. Two table positions were scanned to cover the entire liver. PET images were corrected for attenuation on the basis of the CT data, and iterative reconstruction algorithms (two iterations and eight subsets) were performed. PET data were filtered and scatter corrected.
Image Interpretation
To compare the value of accurately coregistered PET/CT images with the values of both CT images alone and PET images alone in identifying the site of irradiation, CT data sets were evaluated separately by a board-certified radiologist (H.K.) and PET data sets were assessed by a board-certified nuclear medicine physician (S.P.M.). Following separate image evaluation, both data sets were evaluated side by side by the same two physicians in consensus before image fusion was performed. Image fusion was then performed at a separate workstation that featured a syngo-based fusion tool (Siemens Medical Solutions). The fused images were evaluated by the same two physicians in consensus.
The PET/CT images were assessed for areas of altered FDG uptake. The two evaluating physicians quantified the tracer uptake by placing regions of interest in irradiated areas of the liver hilum and in three different regions at the nonirradiated liver periphery (one in the left liver lobe, two in the right liver lobe). Regions of interest that encompassed at least 30 pixels (pixel size, 0.9 mm at CT and 1.17 mm at PET) in the irradiated field and at least 50 pixels at the liver periphery were chosen. We compared tracer concentrations intraindividually by correlating mean concentrations in the irradiated area with mean concentrations at the normal liver periphery; the mean values from four PET/CT examinations performed during 1 hour were calculated. To compare the glucose metabolism between the hilum and the periphery in the different pig groups, the following ratio (R) was determined: R = Chil/Cperiph, where Chil is the activity concentration (in kilobecquerels) in the liver hilum and Cperiph is the activity concentration (in kilobecquerels) at the liver periphery.
In cases in which the glucose metabolism was altered in the central part (ie, the hilum) of the liver, the diameter of the area of decreased tracer uptake was determined by the two evaluating physicians in consensus.
The CT images were evaluated for the presence of bile duct dilatation. The evaluating physicians quantified the dilatation by measuring the bile duct diameter centrally (ie, in the liver hilum) and peripherally (ie, at the liver periphery). To avoid errors caused by intraindividual variability in bile duct diameters, three measurements were obtained in the liver hilum and five were obtained at the liver periphery. Mean bile duct diameters and SDs for the central and peripheral parts of the biliary system were calculated. Contrast-enhanced CT images were also assessed qualitatively for regions of inhomogeneous parenchymal enhancement by using the following grades: severe, meaning inhomogeneous contrast enhancement of the entire liver (including right and left lobes); moderate, meaning inhomogeneous enhancement of the right or left liver lobe; and mild, meaning inhomogeneous enhancement of only a part of the right or left liver lobe.
To evaluate the effect of the surgery and IORT on liver size, the liver volume was quantified on the basis of findings on CT images obtained 2, 4, and 8 weeks postoperatively. The evaluating radiologist determined the liver volumes by manually tracing the contours of the liver on a CT data set with 1-cm section spacing.
Histologic Evaluation
Tissue samples from the hilum and periphery of the liver were fixed in formalin and stained with either hematoxylin-eosin, periodic acid–Schiff, or elastica–van Gieson stain. A pathologist (K.A.M.) evaluated the slides for necrosis, cell atrophy, regenerative and inflammatory reactions, fibrosis, and vascular alterations.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
On the basis of the fused PET/CT image findings, the region of the anastomosis could be clearly defined in all nine pigs. Furthermore, accurately coregistered PET/CT images clearly depicted the irradiated field in all six irradiated pigs on all examination days. Finding separate corresponding PET and CT images proved to be difficult when images were evaluated side by side (ie, nonfused), and this difficulty increased with normalization of glucose metabolism in the central liver regions. When separate PET and CT images were evaluated side by side, the irradiated field could be identified in only five of the six pigs 2 weeks after surgery and IORT, in four pigs at 4 weeks, and in three pigs at 8 weeks.
In the group 2 and group 3 animals, in which radiation was delivered by using 5- and 7-cm applicators, respectively, a decrease in hilar glucose metabolism was noted in the irradiated area 2 and 4 weeks postoperatively, as compared with the glucose metabolism seen at the liver periphery (Fig 1). At 8 weeks, however, the ratios of hilar-to-peripheral glucose metabolism returned to normal. All measurements obtained in the three control animals, in which resection-anastomosis surgery but not IORT was performed, were grouped around a ratio of 1.0, demonstrating a similar degree of glucose metabolism in the hilum and periphery on all examination days (Fig 2). The PET/CT-determined sizes of these areas of minor tracer uptake correlated well with the radiation applicator diameters (mean area sizes at PET, 4.87 cm ± 0.52 with the 5-cm applicator and 6.26 cm ± 0.34 with the 7-cm applicator).
|
|
Central and peripheral bile duct dilatation was verified in all pigs in the irradiated groups, and the dilatation increased gradually but substantially over time (Fig 3). When the irradiated and nonirradiated pigs were compared, bile duct diameters were found to be substantially greater in the irradiated animals (Fig 4, Table).
|
|
|
|
|
|
|
|
|
|
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although CT proved to be the most accurate for determining liver volumes, the accurate assessment of irradiated liver tissue was not possible with CT images: They depicted normal contrast enhancement of the liver hilum in all six irradiated pigs. Inhomogeneous contrast enhancement was visible only along the periphery of the liver in three of the six irradiated pigs. Thus, with CT, it was not possible to differentiate the liver cells affected by radiation treatment from the unaffected cells adjacent to the irradiated field.
Even when separate PET and CT images were evaluated side by side, localization of the irradiated field proved to be considerably more difficult than localization on the fused images. Two weeks after surgery and IORT, the irradiated field could be differentiated in only five of the six pigs. Differentiating the irradiated field became even more difficult over time owing to the normalization of glucose metabolism of the irradiated central liver region, rendering differentiation of the irradiated area from the nonirradiated liver periphery difficult. The main problem with side-by-side evaluations is the difficulty in finding separate corresponding CT and PET images. This problem will increase if CT and PET images are acquired by using different imaging systems.
Different patient positions on the examination tables and varying breathing protocols are also known to render side-by-side evaluations of PET and CT images—and even software-based image coregistration—difficult (7). Therefore, the results of this study indicate that the evaluation of accurately coregistered PET/CT images is considerably more reliable for localizing the site of irradiation than is the side-by-side image evaluation approach.
The PET images clearly depicted areas of decreased FDG uptake in accurate coregistration with the irradiated tissue regions. The sizes of the regions of decreased uptake correlated well with both the 5- and 7-cm applicators used for IORT. Similar decreases in FDG uptake in irradiated cells have been described before (15–17). In a rat tumor model, decreased tracer uptake was found to be characteristic of the early response of liver tissue to radiation treatment, which was identifiable as early as 1 day following radiation treatment (16). Other authors have found that low glucose metabolism may persist for as long as 3 years after radiation therapy (17). Because FDG uptake is a known measure of the number of viable cells (18,19), reduced glucose metabolism in the initial stage after irradiation is likely to represent cell necrosis.
The results of several studies have proved that there is a direct correlation between the extent of tumor cell necrosis and the degree of reduced glucose metabolism (15,20). The results of this study show that combined PET/CT can be used to accurately assess the response of affected cells to radiation therapy on the basis of the PET data and to precisely delineate the irradiated field on the basis of the CT data.
Other factors, however, should also be considered when evaluating the cause of decreased FDG uptake in irradiated cells. When CT images were evaluated in the present study, inhomogeneous arterial and venous perfusion was observed in three pigs 2 weeks after surgery and IORT. This perfusion deficit was transient, however; it normalized to mild arterial perfusion inhomogeneity at 4 weeks and to completely homogeneous perfusion at 8 weeks. This temporary deficit in blood supply is attributed mainly to an arterio-occlusive syndrome caused by radiation-induced obliteration of smaller branches of the arterial system (21–23). This explanation is supported by the results of our histologic analysis, which revealed arterial obliteration of small vascular branches in the irradiated field. On the other hand, the venous system was predominantly unaffected.
In the control group, the arterial and venous blood vessels of the liver hilum were found to be normal. The formation of collateral vessels ensures homogeneous reperfusion of the liver periphery over time, as demonstrated at consecutive PET/CT examinations performed in the present study. The unenhanced CT images demonstrated no signs of perfusion deficit; this result stresses the need for a fully diagnostic contrast-enhanced CT component of the combined PET/CT examination.
Histologic evaluation of the irradiated field in the liver hilum revealed fibrotic alteration of the liver parenchyma, with a reduced number of viable liver cells compared with the number of viable cells in the nonirradiated parts of the liver. Scar formation—rather than cell necrosis and vascular stasis, which are known to be transient—may be the cause of a persistent lack of FDG uptake in an irradiated field (24–27). PET/CT, however, was not able to depict persistently decreased glucose metabolism. In the present study, follow-up PET/CT images showed that FDG uptake increased gradually until uptake values were comparable to those at the liver periphery. This increase in glucose metabolism at 4 weeks and the subsequent normalization of FDG uptake at 8 weeks in the irradiated region may have had two causes: (a) liver cell regeneration with an increase in hepatocellular function and (b) an increase in glucose metabolism due to regenerative and inflammatory tissue in the irradiated field (18,28). When the histologic findings are considered, the rapid normalization of glucose metabolism in the irradiated field must be attributed mainly to regenerative and inflammatory processes, which were readily demonstrated 8 weeks following IORT.
Although the liver tissue adjacent to the area of irradiation and in the liver periphery had normal FDG uptake, the CT images showed increasing bile duct dilatation in the irradiated animals. Dilated bile ducts are thought to be caused by radiation-induced bile duct stenoses due to fibrotic alteration of the bile duct walls, a well-known side effect of liver irradiation (25,27). Our histologic results verified this hypothesis. If further clinical follow-up of irradiated animals had been desired, stent placement in the anastomoses would have been required to prevent severe bile duct dilatation in the irradiated groups. Bile duct dilatation is also thought to be the reason for the greater increase in liver volume in the irradiated animals than in the control animals.
Although the sample size was limited, the study results clearly show that dual-modality PET/CT yields accurate data on cell response to radiation treatment. While PET emission data accurately indicate functional changes in hepatocytes, CT imaging enables comprehensive quantitative and qualitative morphologic assessments. However, further studies are necessary to assess the value of combined PET/CT in tumor models, in other organs, and in the clinical setting of radiation therapy.
In conclusion, morphologic and functional changes due to IORT in liver tissue can be accurately monitored with dual-modality PET/CT. By enabling the integration of functional and morphologic information, PET/CT may have an important role in monitoring the effect of radiation therapy.
Practical application: Although surgical alterations and changes in liver volume were seen only on the CT images, the parenchymal changes caused by IORT were best delineated on the PET images. Combined PET/CT imaging enabled the most precise morphologic and functional analysis of the treated region during the critical 2–8-week posttreatment period. Thus, PET/CT helped us accurately determine the location and volume of the irradiated tissue and the functional and morphologic changes caused by the combination of surgical and radiation therapies. On the basis of these animal model data, PET/CT promises to be a sensitive radiation therapy–monitoring tool that has substantial benefits compared with morphologic or functional imaging alone.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, G.A.; study concepts, G.A., G.M.K., H.K., S.W., A.B.; study design, A.B.M., K.A.M., H.Z., C.E.B., J.F.D.; literature research, G.A., S.P.M.; experimental studies, G.A., S.P.M., G.M.K., S.W., H.Z.; data acquisition, G.A., G.M.K., A.B.M., H.K.; data analysis/interpretation, G.A., G.M.K., K.A.M., H.Z., H.K., C.E.B., S.P.M.; statistical analysis, G.A.; manuscript preparation, G.A., H.K., G.M.K., K.A.M.; manuscript definition of intellectual content, G.A., A.B., J.F.D., C.E.B., H.K.; manuscript editing, G.A., G.M.K.; manuscript revision/review and final version approval, all authors
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  H. Christiansen, N. Sheikh, B. Saile, F. Reuter, M. Rave-Frank, R. M. Hermann, J. Dudas, A. Hille, C. F. Hess, and G. Ramadori x-Irradiation in Rat Liver: Consequent Upregulation of Hepcidin and Downregulation of Hemojuvelin and Ferroportin-1 Gene Expression Radiology, December 1, 2006; 242(1): 189 - 197. [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]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
