Report
Molecular (functional) imaging for radiotherapy applications: an RTOG symposium

https://doi.org/10.1016/S0360-3016(02)04215-3Get rights and content

Introduction

The Translational Research and Image-Guided Radiotherapy Committees of the Radiation Therapy Oncology Group (RTOG) organized this symposium for the winter 2002 semiannual meeting. Although the term “molecular imaging” is used currently to describe several diverse techniques under investigation for delineating cancer phenotype and genotype, the focus of this symposium was on those that could produce tumor maps for image-guided planning for radiotherapy. Because the combination PET/CT scanner is rapidly becoming a standard treatment-planning platform (despite its cost), the topics were selected to cover several applications of positron-emission tomography (PET) and single-photon emission computed tomography (SPECT) imaging in radiotherapy practice. The high specific activity of fluorodeoxyglucose (FDG) and several other novel nuclear medicine agents facilitates clinical investigation with only trace amounts of marker that are nontoxic. Furthermore, the spatial resolution of nuclear medicine images, although not as good as that of CT or MRI, can delineate tumor voxels whose dimensions approach 0.5 cm on a side. Thus, the purpose of this symposium was to review the current use of FDG-PET imaging in radiotherapy treatment planning and to report on novel markers being developed that could be used for radiation oncology research in the near future.

After a brief welcome from Dr. W. J. Curran (the RTOG chairman, Thomas Jefferson Cancer Center), Dr. Michalski (Washington University) gave an introduction to the symposium. Functional imaging with FDG has already led to improved staging of disease and improved treatment planning in several cancer centers by ensuring better dose delivery to gross tumor volume (GTV) while avoiding critical normal tissues. Additional research will likely delineate other benefits to radiotherapy planning from FDG imaging. As well, the development of novel radiolabeled markers associated with tumor prognosis and prediction of radioresistance and/or outcome could impact significantly on treatment planning. Such information could be used to design more conformal radiation dose distributions, “paint” additional dose to specific tumor microregions, or to rationally prescribe adjunctive therapy targeted to specific tumor phenotype and genotype. The Image-Guided Radiotherapy and Translational Research Committees of the RTOG routinely evaluate novel markers and imaging techniques that might be exploited in their clinical trials to improve treatment outcome. An introduction to the field of molecular (functional) imaging followed.

Dr. Michael Welch (Mallinckrodt Institute of Radiology at Washington University) described the scientific principles of molecular imaging and defined some standard terms. For several years, functional imaging has been a basic research tool for scientific investigation, mainly with animal models. As early as 1945, Tobias et al. (1) proposed that tissue biochemistry might be investigated noninvasively with 11C-labeled substrates, and the first PET scan of 15O distribution in a tumor-bearing mouse was reported in 1958 by Ter-Pogossian and Powers (2). Thus, functional imaging has been under development for many years. The first human PET scanner was built and tested by Ter-Pogossian et al. in 1974 (3). However, through the 1970s and 1980s, the high-resolution images of anatomic structure produced by CT and MRI became the “workhorses” for delineating tumor volumes for both radiologic diagnosis and planning radiation treatments. Although some nuclear medicine procedures became important for visualizing tumor spread, SPECT and PET were essentially secondary imaging modalities for oncology over this time. Early PET research was usually performed with short-lived isotopes like 11C, 13N, and 15O, which had to be produced by cyclotrons in close proximity to the imaging device. In the 1990s, positron-emitting isotopes of longer half-life (such as 18F) have come to play a much greater role in functional imaging research important for cancer diagnosis and staging.

Radiolabeled markers have been used to detect hormone receptor status, hypoxia, proliferation rate, blood vessel development/angiogenesis, and apoptosis, to name some applications of potential use to radiotherapy. Several of these applications were discussed in detail by other speakers. Dr. Welch reviewed research from his laboratory on hormone receptor measurement. 18F-estradiol was developed to investigate the estrogen receptor status of tissues and tumor in vivo 4, 5. In breast cancer patients, the uptake of 18F-estradiol correlated with estrogen receptor concentration, could distinguish between receptor expression levels in the primary tumor vs. node metastases, and could predict for outcome to tamoxifen therapy when uptake before and 10 days after treatment was compared. In fact, the response of breast cancer patients to tamoxifen could be predicted equally well when before-and-after scans of FDG were compared. This makes these molecular imaging procedures useful for following the response of individual tumors to various hormone therapies. Analogous markers of androgen receptors may play a role in predicting response in prostate cancer (6). In prostate cancer patients, 18F-dihydrotestosterone associated selectively with tumor tissue and could indicate androgen receptor blocking by flutamide.

The current difficulty associated with molecular image coregistration with patient anatomy was discussed. Larger imaging research centers and industry have invested considerable effort in developing algorithms that can overlay PET, SPECT, and MRI images with the CT image used for treatment planning. These rely on bony structures and/or fiducial markers, and the techniques vary based on the different imaging platforms. None are yet completely satisfactory. The introduction of combination PET/CT scanners will facilitate this problem significantly, at least for these two imaging modalities. More will be said by other speakers about these techniques.

Dr. Welch concluded his lecture by reviewing the important role of small-animal imaging systems in the development of molecular imaging agents. High-resolution MRI and CT scanners for animals have been developed and have produced useful information that can guide the application of novel patient procedures. As well, optical imaging has developed several probes for tissue genotype and phenotype that are important for research. Optical techniques are limited, because useful information can be acquired only from tissue depths of less than 1 cm. Concorde Microsystems (Knoxville, TN) has marketed high-resolution microPET scanners for small and larger animal studies. The availability of these devices, along with the production of several novel radioisotopes for research, has opened the door to additional possibilities for nuclear medicine imaging research. Some of these imaging procedures will now be described in greater detail.

Dr. Steven Larson (Memorial Sloan-Kettering Cancer Center) described the current use of FDG-PET in cancer diagnosis and treatment planning. He emphasized that its utility had been further improved by the introduction of combination PET/CT scanners. FDG imaging is based on the “tracer principle,” that is, the administration of a radiolabeled precursor for some metabolic pathway whose uptake can inform about a specific biochemistry. FDG is taken up into cells by glucose transporters, cannot be phosphorylated by hexokinases, and is subsequently trapped. Measurements (images) of retained radioactivity within various tissues yield good estimates of glycolysis or metabolic rate. FDG uptake correlates strongly with the presence of cancer, but can also reveal areas of infection and/or abscess. In some regions of the body, its predictive power for identifying cancer is ∼95% (7). Several clinical case examples were reviewed that showed how PET images, when superimposed upon CT images, provided a superior definition of the primary GTV, of nodal involvement, and of distant metastases. PET imaging was also effective for directing the procurement of biopsy material that, in most cases, confirmed the presence of cancer.

FDG imaging has become a reimbursable procedure for the diagnosis and staging of non-small-cell lung cancer, colorectal cancer, head-and-neck cancer, melanomas, lymphomas, and esophageal cancers. This resulted from the clear demonstration of its power to significantly improve tumor staging, to detect nodal and other metastases, and to usefully guide biopsy procurement. It can also play a role in measuring tumor response to specific treatment, but this use is not currently reimbursed. Because CT is now the major imaging platform for radiotherapy treatment planning, the acquisition of PET image on a similar digital image array allows for rapid fusion of metabolic and anatomic information. Increased dose can be targeted to the metabolically active regions of the GTV, and nonviable CT image can be excluded, making these merged images useful for intensity- modulated radiotherapy.

Two examples of imaging proliferation kinetics were reported. The first used 124I-labeled iododeoxyuridine as a precursor nucleoside that is taken up into the DNA of all proliferating cells (in place of thymidine) and gives a measure of tumor proliferation. Incorporated iododeoxyuridine can be imaged over several days, allowing for excretion of the unincorporated precursor. Three days after its administration to a patient with a brain tumor, it was observed in a circular, peripheral subregion of the tumor volume that was also FDG positive. Information about zones of proliferation within GTV could be important for treatment planning and more important for evaluating treatment response. Another animal study employed 18F-labeled FIAU, a “false” nucleoside that is recognized by the thymidine kinase of some viruses, but not by mammalian cells, to monitor gene delivery to tumor cells by a viral vector. The detection of this agent within the tumor indicated successful transfection and could have utility in the future for monitoring the success of gene therapy.

Dr. Larson concluded his talk by demonstrating that the gating of PET image acquisition from lung cancers yielded a tumor image of smaller volume and greater metabolic activity than that acquired throughout the breathing cycle. The role of the medical physicist in both the acquisition of image and the delineation of GTV for planning treatment dose will necessarily expand with the increase in use of combination PET/CT scanners.

Dr. Yee Ung (Toronto-Sunnybrook Regional Cancer Center) then described research into the use of hybrid PET/CT in radiation therapy treatment planning. Coregistration of hybrid PET and CT images allows for the correlation of functional and anatomic information. External fiducial markers were placed, and the patient was CT and PET imaged in the radiation treatment position. Intravenous contrast agent was not used for CT imaging, because of its effects on tissue attenuation and hence radiation dose calculations, but the use of oral contrast helped to identify physiologic uptake in the gastrointestinal tract. Common sites of FDG uptake into normal tissue are the brain, heart, and urinary tract. Variable but significant uptake can be seen in the gastrointestinal tract, thyroid gland, muscle, and sites of infection (8).

A pilot study on 8 patients with carcinoma of the anal canal was described (9). CT imaging alone is often poor in defining the primary tumor in the anal canal (10), and the availability of lymphangiography to detect pelvic lymphadenopathy is limited (11). The anal canal primary was clearly visualized by PET (8/8) with only 4 of the 8 detectable by CT alone. Two patients had biopsy-proven inguinal nodes, and both of these were positive on FDG. One patient had an indeterminate inguinal node by size criteria, and this node was negative on FDG. In this pilot study, significantly improved primary tumor volume definition was achieved with hybrid PET/CT images.

Hybrid PET/CT images were used also to define the GTVs of locally advanced non-small-cell lung carcinomas (NSCLC). The primary aim of this prospective study was to determine the impact of integrating hybrid PET/CT images on treatment planning in the radical treatment of NSCLC as evaluated by coverage of the planning target volume (PTV) and dose-volume histogram analysis of target and critical normal structures. The second aim of the study was to determine whether the integration of hybrid PET/CT images reduced the interobserver variation in the localization of the GTV. Three experienced radiation oncologists independently contoured the GTV in 30 patients with NSCLC, using CT images along with diagnostic CT radiographs, bronchoscopy, and mediastinoscopy information and pathology reports. Then the coregistered hybrid PET/CT images were viewed, and the GTV modified based on the additional information provided by hybrid PET imaging (12). Separate treatment plans were generated for both the CT-based and hybrid PET/CT-based planning target volumes. Treatment strategy for 7/30 (23%) patients was changed from curative to palliative based on the imaging information obtained by PET, which detected extrathoracic disease or disease too extensive to be encompassed by a radical dose of radiation. Hybrid PET/CT imaging was also useful for distinguishing atelectasis from tumor tissue, and this resulted in normal-tissue sparing with a reduction in radiation treatment volumes. PTV coverage based on CT information alone was inadequate for 17%–29% of patients, depending on the individual observer. The primary reason for inadequate coverage was the finding of unsuspected lymph node involvement on hybrid PET that was not seen on CT. There was a large variation in the contoured GTV among the various observers (13). The use of hybrid PET information reduced interobserver variation as seen by a reduction in the mean ratio of the largest to smallest GTV from 2.31 to 1.56 in this study. It was concluded that a more consistent definition of GTV could be obtained when coregistered hybrid PET and CT images are used. Ultimately, the radiation treatment plans for the majority of these patients were altered based on the improvements achieved in defining GTV. It should be noted that some interobserver variation in GTV definition will always exist. This prospective study enrolled primarily NSCLC patients who had tumor boundaries obscured by atelectasis, pleural effusion, and contiguity of the primary with mediastinal lymphadenopathy, all leading to variability in target definition. In addition, physician experience, treatment philosophy, and understanding the patterns of spread for lung cancer ultimately determine the final PTV. Defining the GTV is a fundamental step in radiation treatment planning, and given the difficulties in visualizing the primary and abnormal lymph nodes, as well as in observer variation, the use of combined FDG-PET/CT imaging for radiation treatment planning should become the standard in lung cancer for GTV localization before attempting dose escalation with conformal radiotherapy or intensity modulated radiation therapy.

Dr. Jeffrey Bradley (Washington University) described experience with PET imaging in defining GTV and planning conformal radiotherapy for carcinoma of the lung. In a pilot study, 15 patients were imaged with FDG-PET and CT on two different imagers using the α-cradle in the radiotherapy position. Laser positioning, similar to that used in patient setup for radiation treatment, was employed. The images were fused using fiducial markers (metal beads for CT and FDG-containing beads for PET) contained within adhesive strips and a locally developed algorithm. The incorporation of metabolic images into CT planning produced better definition of the primary tumor and identified previously undetected nodal disease. Specifically, PET was useful for identifying previously occult distant metastases and additional nodal involvement and for delineating between tumor and atelectatic lung. Dose-volume histograms of lung and esophagus were generated to estimate normal-tissue toxicities associated with altered treatment plans. Tumor stage (TNM) changed for 6/15 patients, and treatment plans were significantly altered for 8/15 patients. Significant alterations were defined as a change in the radiation therapy portals to include areas that were not included in the plan using CT alone. Either major or minor variations in tumor volume contours were seen for all patients. Dr. Bradley referred to recent reports by Pieterman et al. (14) and MacManus et al. (15) that clearly demonstrate the important role of PET imaging for lung cancer diagnosis, staging, treatment prescription, and radiation planning.

Discussion on FDG-PET/CT imaging for radiotherapy treatment planning that followed addressed the cost of the combo-imagers (∼ $2.5 M), the time to acquire and fuse the images, the problems associated with organ motion, the need for gated image acquisition, and a concern about the variation in GTV delineation by different observers. It was noted that FDG images of early-stage prostate cancer were usually inferior to those obtained from others tumors but that (11C)acetate or (11C)methionine could provide useful information.

Dr. J. Donald Chapman (Fox Chase Cancer Center) reviewed the current status of radiolabeled markers of tumor hypoxic microenvironment. The RTOG has invested considerable resources to evaluate several techniques for overcoming tumor resistance associated with hypoxia. The identification of bioreducible drugs that become bound to viable cells of low oxygen tension was a “spin-off” from hypoxic radiosensitizer research (16), much of which was performed by this cooperative group.

There are three distinct classes of hypoxic markers being investigated today whose mechanisms of hypoxic deposition and pharmacology are profoundly different. Those that contain the 2-nitroimidazole (azomycin) moiety, which upon reduction can covalently bind to cellular molecules, were named “vanilla” markers (17). The vanilla markers include misonidazole (MISO); fluoromisonidazole (FMISO); the azomycin nucleosides, of which IAZAF, IAZGP, IAZXP have been widely investigated; fluoroerythromisonidazole; and EF-5 (18). Their lipid/water partition coefficients, uptake into severely hypoxic relative to aerobic cells (hypoxia-specific factor [HSF]), and oxygen concentration dependency of binding vary significantly. The optimal hypoxia marker should have a large HSF (high sensitivity), and bind to cells in tumors over the oxygen concentration range that defines the radiobiologic oxygen effect. Their uptake into individual tumors should predict strongly for radioresistance and tumor response. As well, the acquisition of hypoxia-specific radioactive signal from tumors in patients should be at times when the hypoxia/background signal is maximal. These constraints for optimizing this technique make hypoxia imaging with 18F-labeled agents difficult, because hypoxia signal at early times after marker administration is small relative to the signal from nonspecific marker that has not been excreted. Of the various vanilla markers tested to date, β-D-IAZGP labeled with 124I should produce the optimal maps of human tumor hypoxia at the highest resolution (19).

Markers that do not contain azomycin but rely on the reduction of a chelated metal for their selective deposition in hypoxic tissue were named “chocolate” markers. The most widely investigated chocolate markers are (99mTc)HL-91 and (60Cu)ATSM. The HSFs of these agents are significantly lower than those of the vanilla class, and their oxygen dependency of binding varies significantly (20). After i.v. injection, these markers distribute rapidly to different tissues in animals at distinct levels that can differ by one to two orders of magnitude. Their retention in liver and kidney tissue is always greater than in experimental rodent tumors. That tumor uptake is predictive of individual tumor hypoxic fraction, and radioresistance remains to be demonstrated. Nycomed Amersham plc (Amersham, UK) has discontinued the development of (99mTc)HL-91 as a hypoxic marker, but (60Cu)ATSM continues to be investigated by researchers at the Mallinckrodt Institute of Radiology (St. Louis), as well as by others.

A third class of hypoxia marker that contains a metal-chelating ligand complexed with one or more azomycin substituents has been named “chocolate swirl.” These markers were synthesized to take advantage of the well-investigated bioreduction properties of azomycin and the labeling ability of various ligands. Bracco Research USA Inc. (Princeton, NJ) produced the 99mTc-labeled markers, BMS181321 and BMS194796, to monitor brain and heart ischemia. These markers have also been characterized for their ability to measure tumor hypoxia in animal models (21). Dr. Chapman’s laboratory has synthesized several azomycin-cyclam derivatives and has identified some with good hypoxia labeling potential (20). The chocolate swirl markers distribute rapidly in tumor-bearing animals to all tissues at distinct levels, similar to the chocolate markers. In fact, it is likely that the metal component of the marker governs their distribution and uptake into various tissues. That uptake of the chocolate swirl markers into individual tumors correlates strongly with hypoxic fraction, and its associated radioresistance has yet to be demonstrated.

Examples of SPECT and PET images of (123I)IAZAF, (18F)FMISO, and (60Cu)ATSM uptake into human tumors were presented. The validation that marker uptake into specific tumor regions strongly correlates with the presence of viable hypoxic cells is difficult, because no gold standard for measuring this tumor property currently exists. Comparative studies in animal models between Po2 measurements by microelectrodes, 31P magnetic resonance spectra, and cellular radiosensitivity measurements by the paired survival curve and comet assays have been reported. Before clinical evaluation, potential hypoxia markers should undergo testing to determine their power to predict for individual tumor radiosensitivity. Significant correlations between in vivo/in vitro assays of tumor cell radiosensitivity and uptake of MISO, EF-5, and β-D-IAZGP have been reported (19).

Dr. Chapman suggested that one or two agents with optimal hypoxia marking properties should proceed to advanced clinical testing in human tumors where hypoxia has been shown to predict significantly for their treatment outcome. High-resolution maps of hypoxic tumor microenvironment could then be used to guide the application of additional dose to resistance sites and/or allow for the rational prescription of hypoxia-targeted therapy, such as etanidazole, tirapazamine, RSR-13, pretherapy transfusion, etc. (22).

Dr. Paul Okunieff (University of Rochester Medical Center) discussed several hypoxia-measuring techniques that use MRI or magnetic resonance spectroscopy. Direct assays include the spectroscopic assessment of 31P-containing cellular molecules, most of which can inform about cellular metabolism and energetic phosphorous states. The ratio of inorganic phosphate (Pi) to β-ATP or phosphocreatine increases significantly in tissues of lower oxygen tension. In addition, as the oxygen tension of tissue decreases, a pronounced reduction occurs in the resonance peak width of several phosphorylated molecules. These changes in signal from endogenous cellular molecules can report on intracellular oxygen tension. The blood oxygen level dependence technique measures oxygen-dependent changes in the proton spectra of the water/hemoglobin complexes (23). This technique reports on oxygen concentration in the intravascular space, which can inform about tissue and tumor oxygenation. Relative changes in tissue oxygen levels can be inferred by subtracting T1 or T2 images acquired during carbogen and/or oxygen breathing from those obtained during air breathing. Dr. Okunieff suggested that this technique could be adapted for some clinical studies.

Indirect MRI and magnetic resonance spectroscopy methods for measuring tissue oxygen level require the administration of a reporter molecule, such as a fluorinated hydrocarbon or hexafluorobenzene (24). The fluorohydrocarbons are usually injected into the blood compartment and report on intravascular oxygen concentration, whereas hexafluorobenzene must be injected intratumorally to yield information about cancer. Both techniques have been useful for investigating tumor oxygenation in animal models, but their clinical application poses several problems, not the least of which is the mM concentrations of 18F that are required to produce useful signal by currently available scanners.

Dr. Janet Eary (University of Washington) described research into methods for imaging tumor proliferation and growth rate. Direct measures of cell proliferation within solid tumors could be predictive of tumor response and might inform about the efficacy of specific therapy. Cancer biology research, including that of the RTOG, has investigated the role of mitotic index, S-phase fraction, PCNA, Ki-67, p105, and other markers of cell proliferation to evaluate their potential role in predicting for treatment response (25). Several nuclear medicine research groups have investigated the acquisition of related information by a noninvasive procedure. At the University of Washington, (11C)thymidine (TdR) uptake into brain tissue was investigated as a marker of cell proliferation. Images were presented that showed the (11C)TdR distribution within the brain of a glioblastoma patient to be significantly different from that of FDG. (11C)TdR images acquired early after treatment showed a significant decrease in tumor proliferative activity relative to the pretreatment scan. That reduced (11C)TdR uptake into tumor is strongly predictive of response to treatment requires validation studies with patient outcome. Effects of blood flow and the rapid production of radiolabeled catabolic products that are unrelated to cell proliferation complicate this imaging procedure. Techniques have been developed that exploit (11C)CO2 imaging to subtract “anomalous” image from the acquired tumor scans, and data analysis with a compartmental model has improved the image of tumor proliferation. The technique can readily distinguish necrosis from tumor regrowth in posttreatment scans. Based on imaging success with [11C]TdR, which is the native molecule incorporated into cell DNA, research interests have also turned to substrates of thymidine kinase that are not metabolized and are labeled with isotopes of longer half-life. Fluorothymidine, fluorouridine, and fluoroarabinouracil have been investigated 26, 27, 28. These studies may identify a metabolically less complex radiotracer for cell proliferation that can be used for tumor growth rate and tumor response to therapy assessment.

Dr. Roland Haubner (Technische Universität München) reported on radiotracer techniques for the noninvasive determination of tumor-induced angiogenesis. The role of angiogenesis in tumor growth and treatment response has received much attention lately, and the RTOG has formed a working group to monitor the clinical use of antiangiogenic therapies. Most drugs target growth factor receptors, matrix metalloproteinases, or integrins, and some use endogenous inhibitors such as Endostatin. Because most antiangiogenic therapies are mainly cytostatic, novel methods for determining tumor treatment response may be required.

PET techniques already developed should be useful for investigating functional and physiologic changes during antiangiogenic therapy (29). These include measures for blood flow, blood volume, and vascular permeability, as well as measures of tissue hypoxia, oxygen metabolism, cell proliferation, and glucose use. Newer strategies are focused on the development of radiolabeled markers of vascular endothelial growth factor receptors, tyrosine kinases, cryptic angiogenesis inhibitors, extracellular matrix proteins, matrix metalloproteinases, and integrins (30). Although some studies with antibodies specific to fibronectin isoforms (31) and with peptide chains that inhibit specific metalloproteinases (32) have been reported, the majority of effort has concentrated on developing radiolabeled antagonists to ασβ3 integrin. The ασβ3 integrin mediates migration of activated endothelial cells through the basement membrane during formation of new blood vessels (33). It is highly expressed on activated endothelial cells, but not on quiescent endothelial cells of established vessels. ασβ3 binds to the tripeptide sequence Arg-Gly Asp (RGD) of extracellular matrix proteins such as vitronectin, fibronectin, or fibrinogen. Blocking these interactions with low-molecular-weight antagonists results in the detachment of the endothelial cells, leading to apoptosis. A variety of compounds that react with the RGD sequence have been synthesized. Dr. Haubner’s laboratory synthesized the iodinated derivatives [125I]cyclo(-Arg-Gly-Asp-d-Tyr-Val-) and [125I]cyclo(-Arg-Gly-Asp-d-Phe-Tyr-), which showed high ασβ3 affinity and selectivity in vitro and receptor-specific tumor accumulation in vivo (34). These tracers exhibited a predominantly hepatobiliary excretion with high activity found in the liver. More water-soluble tracers were developed by adding sugar substituents. The glycopeptides [125I]Gluco-RGD (35) and [18F]Galacto-RGD (36) showed decreased localization in liver and improved activity accumulation in the ασβ3-positive tumor. A small animal PET scanner was used to validate [18F]Galacto-RGD uptake into receptor-positive tumors in mice. Quantitative autoradiography studies with a transgenic mouse model of pancreatic islet cell carcinogenesis and angiogenesis (37) were used to correlate marker uptake with tumor progression and ασβ3 expression. Other RGD-based markers of tumor angiogenesis labeled with 99mTc, 188Re, and 90Y were described 38, 39, 40. The addition of chelation groups to these polypeptides did not seem to interfere with their specificity. Dr. Haubner concluded that, of the several approaches to labeling tumor vasculature and angiogenesis, [18F]Galacto-RGD allows for good visualization of ασβ3-positive tumors and could play a role in monitoring the response of some tumors to antiangiogenic therapy.

Dr. William Strauss (Memorial Sloan-Kettering Cancer Center) began his presentation by emphasizing that nuclear medicine was close to producing high-resolution maps of several tumor components and functions, including information about vasculature, blood flow, oxygenation and hypoxia, proliferation, and evidence of cell death. With regard to cell death, he reviewed the current understanding of the role of apoptosis in normal tissue and tumor cell death. Apoptosis or programmed cell death is a normal process by which cell death is initiated and accomplished by endogenous molecules in response to external signals or internal molecular damage. It is thought that tumor response to drugs and to some molecular therapy is by this cell death pathway. This mechanism dominates in the radiation killing of cells of lymphoid origin, but seems to play a smaller role in the killing of carcinoma cells, those that constitute the majority of tumors whose optimal treatment includes radiotherapy.

When cells die and their membranes disintegrate, phosphatidylserine, which normally resides on the inner membrane surface, becomes externalized. This event precedes the fragmentation of DNA into nucleosome ladders, one hallmark of cells undergoing apoptosis (41). Annexin V is a naturally occurring human protein that binds avidly to membrane-associated phosphatidylserine and whose intracellular concentration is about 100× higher than its extracellular concentration. Recombinant human annexin V has been produced and complexed with a ligand to which 99mTc can be linked. Radiolabeled annexin localizes in the hearts of experimental animals undergoing rejection after transplants (42), in animals with collagen-induced autoimmune arthritis (43), in the brains of animals exposed to hypoxic-ischemic brain injury, and in humans with neoplasia and acute myocardial infarction (44). [99mTc]Annexin V also localizes at sites of inflammation and in some tissues undergoing remodeling, especially during development. In cancer patients, tumor cell death has been visualized after chemotherapy in non-small-cell lung cancer, small-cell lung cancer, lymphoma, breast cancer, and sarcoma. One pilot study has shown that with 1-year follow-up, posttreatment [99mTc]annexin V uptake is associated with improved time to progression of disease and time of survival (45). This imaging agent adds a novel diagnostic tool to the “arsenal” of oncologists in that it provides predictive information about the role of apoptosis in the response of individual tumors and a method for measuring effectiveness of various therapies. These studies have used SPECT, which can also play a useful role in cancer imaging, even though its current resolution is inferior to that of PET.

Section snippets

Conclusions and recommendations

This symposium has highlighted several imaging techniques that are available and can provide important information for the advanced radiation treatment planning of cancer. FDG-PET has already been used to improve cancer staging and the definition of GTV for radiation treatment planning. It is important to demonstrate that such changes to conventional treatment strategy result in improved cancer outcomes. Whether maps of treatment-resistant subregions of tumors treated with additional dose by

Acknowledgements

The assistance of Pat Bateman with manuscript preparation is appreciated.

First page preview

First page preview
Click to open first page preview

References (45)

  • L. Tarli et al.

    A high-affinity human antibody that targets tumoral blood vessels

    Blood

    (1999)
  • F.G. Blankenberg et al.

    Noninvasive strategies to image cardiovascular apoptosis

    Cardiol Clin

    (2001)
  • L. Hofstra et al.

    Visualisation of cell death in vivo in patients with acute myocardial infarction

    Lancet

    (2000)
  • C.A. Tobias et al.

    The elimination of carbon monoxide from the human body with reference to the possible conversion of CO to CO2

    Am J Physiol

    (1945)
  • M.M. Ter-Pogossian et al.

    The use of radioactive oxygen-15 in the determination of oxygen content in malignant neoplasma

    Proc 1st UNESCO International Conference of Radioisotopes in Sci Res, Paris, 1957

    (1958)
  • M.M. Ter-Pogossian et al.

    A positron-emission transaxial tomograph for nuclear imaging (PETT)

    Radiology

    (1975)
  • K.D. McElvany et al.

    16α-[77Br]bromoestradiolDosimetry and preliminary clinical studies

    J Nucl Med

    (1982)
  • M.A. Mintun et al.

    Breast cancerPET imaging of estrogen receptors

    Radiology

    (1988)
  • M.J. Welch et al.

    Steroid hormone receptors as targets for diagnostic imaging

  • P.D. Shreve et al.

    Pitfalls in oncological diagnosis with FDG PET imagingPhysiologic and benign variants

    Radiographics

    (1999)
  • Y.C. Ung et al.

    18FDG Hybrid PET and CT fusion improves target volume definition in treatment planning for carcinomas of the anal canal

    Int J Radiat Oncol Biol Phys

    (2001)
  • A. Scherrer et al.

    CT of malignant anal canal tumors

    Radiographics

    (1990)
  • Cited by (76)

    • Recent advances in hybrid imaging for radiation therapy planning: The cutting edge

      2011, PET Clinics
      Citation Excerpt :

      who established the concept of biologic imaging and moved forward the role of PET in RT, thus allowing it to enter the clinical arena. Since that time, the technical aspects of PET/CT-guided RT have been described more thoroughly in the scientific literature.3,7,41–45,50–52 The success of these initial studies prompted significant interest from the major medical imaging equipment manufacturers, who all have introduced commercial PET/CT scanners equipped with the required accessories (flat couch insert, positioning system, respiratory gating, and other accessories) and software tools (virtual simulation, visualization and segmentation tools, support of Digital Imaging Communication in Medicine (DICOM) RT object definition, and other tools) for clinical use.46

    • Molecular imaging of tumour hypoxia

      2009, Cancer/Radiotherapie
    • Molecular PET/CT Imaging-Guided Radiation Therapy Treatment Planning

      2009, Academic Radiology
      Citation Excerpt :

      This was further stimulated and put into perspective after the introduction of the concept of biological imaging (136). Thereafter, the technical aspects of PET/CT-based RT were described in many reports and state-of-the-art reviews (18,22,137–140). The fundamental motivation for the use of PET/CT in radiation therapy is that it provides better visualization than CT simulation in the sense that in some cases CT misses some areas that light up on the PET study or that the lesion volume is in reality smaller on the PET/CT study than on the CT alone.

    View all citing articles on Scopus

    This symposium was made possible by an unrestricted educational grant from CTI/P.E.T. Net Pharmaceuticals, Inc., Knoxville, TN.

    View full text