May 2008

Positron Emission Tomography (PET):
Principles and Applications


Amy LeBlanc, DVM, DACVIM (Oncology)
Assistant Professor, College of Veterinary Medicine; Knoxville, Tennessee




Introduction: Principles of Positron Emission Tomography

Positron Emission Tomography (PET) is a nuclear medicine technique that uses positron-emitting radionuclides and measures distribution of these substances within the body. The applications of PET are widespread and stem from first using 2-[18F]-fluoro-2-deoxyglucose or FDG to study glucose metabolism of the human brain in the 1970s.1 The utility of PET in imaging of malignancy was realized due to increased glucose uptake by tumor cells, and since then PET has revolutionized the diagnosis, staging and management of neoplastic disease in humans.2,3 In 2005, an estimated 1,129,900 clinical PET patient studies were performed at 1,725 sites around the country. PET has numerous applications in other non-neoplastic disorders such as neurologic, infectious, and cardiovascular diseases. Potential applications to clinical veterinary oncology are numerous and include diagnosis and initial staging of malignancy, assessment of response to therapy, and detection of recurrent disease after treatment. PET also has many applications as a research tool in studying spontaneous cancer in animals and aiding in novel radiotracer development.

Basic Physics of Positron Emission


The images produced by PET utilize the unique physical properties of positron-emitting radionuclides.4 Positrons (positively charged electrons, or b+) are emitted from atoms that are unstable because of high proton/neutron ratio. In order to achieve a more stable resting state, the extra proton in the nucleus is converted into a neutron, releasing energy in the form of a pair of particles (positron and neutrino). Positron-emitting radionuclides are created by bombarding a stable target element with protons in a cyclotron. When synthesizing 18F inside a cyclotron, 18O-enriched water is bombarded with protons, resulting in a mixture of H2(18F) and 18O-enriched water. 18F is isolated from this mixture in an automated radiochemical process.

Once the positron is emitted from the nucleus, it loses energy through collision with electrons in the surrounding tissue until it annihilates with an electron, producing two 511 keV photons that are emitted about 180 apart. (Figure 1).5 The mean positron range is smaller in dense structures such as bone but longer for structures such as air-filled lung. These photons are captured in coincidence by a ring-shaped array of detectors within the PET scanner, and are the foundation of PET imaging.





Figure 1: Schematic representation of the decay of a neutron-deficient, positron-emitting radionuclide and detection in coincidence of the annihilation photons within a specific time interval. After injection of the labeled radiopharmaceutical and the detection of a pair of annihilation photons in coincidence by a multi-ring PET camera, the events are collected and reconstructed to produce a whole-body image that maps the uptake of the radionuclide throughout the patient. This image depicts a dog in right lateral recumbancy with a large mast cell tumor involving the left axilla within which significant uptake of radionuclide is visible (thick arrow). The site of FDG injection is visible on the right antebrachium. Note normal uptake of radionuclide within the brain, salivary glands, myocardium and kidneys (thin arrows).

Radiopharmaceuticals


The most common radiopharmaceutical used in modern PET imaging is FDG. Developed in 1976 for the purpose of mapping regional cerebral glucose metabolism, this molecule is an analogue of glucose that is used to quantify the rate at which the hexokinase reaction of glycolysis is occurring in a tissue or organ.6 The development of this compound is based on the intracellular fate of 2-deoxyglucose (2-DG), an analogue of glucose that is phosphorylated in a similar manner by the hexokinase enzyme, the first step of glycolysis.6 However, once phosphorylated, 2-DG-6-P is not a substrate for glucose phosphate isomerase and is therefore trapped within the cell, unable to undergo the ensuing steps of glycolysis or the pentose phosphate shunt.

FDG is synthesized by replacing the hydrogen molecule at the C-2 position of 2-DG with 18F. Phosphorylated FDG, just as 2-DG-6-P, cannot be further metabolized, so all accumulated intracellular radioactivity over time is proportional to the rate of the hexokinase reaction in the observed tissue. At steady state conditions, and in the absence of significant glucose-6 phosphatase activity which will dephosphorylate glucose and FDG, this represents the rate of glycolysis in the tissue.6 FDG is an excellent tracer for malignancy as increased utilization of glucose by tumor cells is well-documented.3 However, FDG is also normally taken up by metabolically active organs such as brain, myocardium, salivary glands, active skeletal muscle, and kidneys. Active areas of inflammation can also account for non-neoplastic FDG uptake.7





Other radionuclides are used in PET such as 11C, 13N, 15O but because of their shorter half-lives, use in clinical patients is limited. 18F-labeled biomarkers, with a half-life of 110 minutes, can be transported from the cyclotron to the patient within a few half-lives.2

SCANNER DESIGN


The PET camera designed to detect the pair of annihilation photons from decay of the positron-emitting isotope is comprised of a ring of block detectors that encircles the patient. Each detector is composed of scintillation crystals coupled to a photomultiplier tube. Because the isotopes used in PET emit photons of much higher energies compared to those used more commonly in nuclear medicine (511 keV for FDG compared to 140 keV for 99mTc), detectors must have higher stopping power than those of a traditional gamma camera. Most PET scanners manufactured today use detector materials composed of bismuth germinate (BGO), cerium-doped lutetium oxyorthosilicate (LSO), or cerium-doped gadolinium oxyorthosilcate (GSO).4

The photons produced by positron-electron annihilations are captured in coincidence by opposing detectors within the ring array of detectors, therefore a 'count' is registered when photons strike the opposing detectors within a specific time window. Detection of photons in coincidence indicates that the annihilation reaction occurred somewhere along a Line of Response (LOR). The location of the radiotracer accumulation can be mapped by creating multiple LORs and reconstructing the data to determine their source(s) within the patient. (Figure 1) The directionality of photon emission also provides natural collimation, making PET much more efficient than other nuclear medicine techniques that employ single-photon emitting radionuclides. The resulting images represent radiopharmaceutical accumulation in specific areas of the body closely related to the underlying biologic process of interest.

PET/CT FUSION


Even with recent advances in scanner design and concomitant increases in spatial resolution of tracer uptake, anatomical localization of functional abnormalities is difficult with PET alone. Accurately-aligned fused images of anatomy and function obtained with PET/CT offer substantial advantage to the study interpreter through accurate localization of tracer accumulation, the distinction of normal uptake from pathology, and the verification that a suspicious finding on one modality can be confirmed by the other modality.8 The recent fusion of PET with computed tomography (CT) has been an important step in maximizing the attributes of both modalities. (Figure 2)5,8,9 The fused scanner design allows anatomy and function to be assessed in a one scan session with single positioning of the patient, minimization of organ movement, and no requirement for labor-intensive image registration algorithms as when the scans are obtained separately.8 The majority of PET scanners currently being manufactured now incorporate CT, and it is predicted that more than 90% of PET will be PET/CT in the near future.





Figure 2: Schematic representation of a typical PET-CT fusion scanner design



Practical Aspects of Pet Imaging

Patient Preparation

Human patients are fasted for approximately 6 hours to maximize tumor-associated FDG uptake and minimize cardiac uptake. Blood glucose is measured prior to the FDG administration, and a level < 150 mg/dl is desired. Good control of blood glucose is essential because the uptake of FDG into cells is competitively inhibited by glucose, due to their shared intracellular transport mechanism (GLUT1). Patients then receive an IV injection of FDG and are instructed to sit quietly without talking or moving for 60-90 minutes to allow FDG uptake, and then are scanned while awake. To perform this type of scan in companion animals, light general anesthesia is required for patient positioning and minimization of movement during the scan. During the time of FDG uptake after injection, the animal should be under general anesthesia or cage confined with sedation to minimize skeletal muscle uptake of FDG.

Image Interpretation and the Standardized Uptake Value (SUV)


Uptake of FDG as a marker of glucose metabolism can be semi-quantified using the standardized uptake value (SUV), which is utilized to determine the relative significance of uptake. The SUV is obtained by quantifying the radioactivity within a region of interest (ROI) placed over the lesion or organ of interest, taking the ratio of the ROI value (in Ci/mL) to the injected dose, divided by the patient's body weight.4



There are ranges of SUV values which are typically observed in areas of postoperative scarring, inflammation, infection, or neoplasia. An SUV or 2-2.5 may be used as a cut-off for benign disease, with malignancies usually having SUVs of 3 and higher. Although the SUV can suggest abnormal uptake consistent with active malignancy, a tissue biopsy is still required for definitive diagnosis.

Physiologic uptake of FDG can be seen in organs known to have high glucose requirements. Among these the brain, being exclusively dependent on glucose metabolism, is most intense. In the myocardium, which normally uses free fatty acids as the primary energy substrate, a recent meal can switch this process to glycosis. Therefore, fasting for 4-6 hours prior to PET using FDG is recommended to reduce physiologic myocardial FDG uptake. FDG is excreted in the urine and is not reabsorbed in the renal tubules, therefore intense FDG activity can be seen in the urinary collecting system. Generally a low level of FDG uptake is seen in liver, spleen, renal cortex and bone marrow. Significant skeletal muscle uptake is seen after exercise, in the respiratory muscles after hyperventilation, in the cervical muscles with tension, and the in laryngeal muscles with vocalization. Homogenous, symmetrical uptake in lymphatic tissues and salivary glands may be seen as a normal variant. Uptake within the GI tract is variable but appears especially intense in the large bowel of normal cats in our study.



Potential Applications in Veterinary Medicine

Infectious Disease

Based on work in humans using FDG-PET to characterize and guide the clinical management of solitary pulmonary nodules (SPN), a study evaluating PET in experimentally induced blastomycosis granulomas in rodents was initiated at UTCVM in 1997. This work led to using PET to characterize a subset of dogs with naturally-occuring blastomycosis, along with a small cohort of dogs with lymphoma. These were the first efforts to image companion dogs with naturally-occuring disease using whole-body PET, and found that the average SUV of canine blastomycosis lesions were higher than canine lymph nodes affected by lymphosarcoma.11 We also recently reported on a case of cerebral blastomycosis in which post-operative brain PET scan was used to document success of craniectomy for removal of a fungal granuloma in a cat. These findings emphasize the lack of specificity of FDG-PET for malignancy, as infectious and inflammatory conditions known to have increased glucose useage can exhibit excessive FDG uptake and be successfully imaged with PET.

Oncology

Based on the human oncologic experience, PET has theoretical application to many common veterinary tumors such as osteosarcoma, lymphoma, mast cell tumor, and genitourinary carcinomas. Currently, the limited availability of this technology has resulted in few veterinary reports of its use.10,11 At the University of Tennessee College of Veterinary Medicine, we have performed pilot studies to evaluate normal distribution of 18FDG in clinically normal dogs and cats. These investigations are important to evaluate FDG-PET in future studies of disease states. Further, we have applied FDG-PET to tumor-bearing dogs, cats and birds with a variety of malignancies.

Tumor Staging


As a non-invasive, whole-body imaging technique, PET and PET/CT are immensely applicable to the staging of many common veterinary malignancies. In humans, PET has become a standard technique in staging of Hodgkin's disease, Non-Hodgkin's lymphoma, and breast, lung, and colorectal tumors.12-15 PET has also proven invaluable in restaging patients that have completed a prescribed course of therapy and present with clinical symptoms that may indicate recurrence of their disease. PET can detect abnormalities within tissue before anatomical derangements occur, theoretically improving upon the detection rate of relapsed disease when scans are performed as a part of routine patient follow-up.1,2

In our investigations of whole-body FDG-PET in dogs with lymphoma and mast cell neoplasia, we have found both tumor types can be imaged successfully with this modality. In several cases, the results of whole body PET did not agree with results of routine tumor staging, specifically in identifying MCT that were not detected on physical examination or in fully characterizing regional lymph nodes suspected of containing metastatic MCT.

Evaluation of Response to Therapy


During a prescribed course of therapy, PET can be used as an early indicator of response. Several human studies indicate that PET can predict outcome by the reduction in SUV seen within known areas of tumor in the early stages of chemotherapy. Significant changes in SUV were seen 72 hours after chemotherapy administration to patients with hepatic metastases from colorectal cancer.16 In non-small cell lung cancer, reduction of metabolic activity assessed by FDG uptake is closely correlated with final outcome of therapy.17 Persistently abnormal FDG uptake after first-line chemotherapy in non-Hodgkin's lymphoma (NHL) is highly predictive for residual or recurrent disease.18 Further, a greater than 90% reduction in intensity of abnormal FDG uptake after 2 courses of reinduction therapy was correlated with a favorable outcome in chemosensitive relapsed NHL.19 In the UTCVM pilot study of whole-body FDG-PET, three dogs with lymphoma had paired scans before and during chemotherapy to evaluate each dog's response to therapy and confirm the clinical diagnosis of remission. Future directions include using radiolabeled tracers such as 2-[11C]-thymidine or 2-[18F]-thymidine, that can be early predictors of response to chemotherapy in other types of cancer through quantitative changes in DNA synthesis.20,21

Research Applications


PET and PET/CT are commonly used in tumor-bearing humans and rodent models for oncologic research. Although this technology is in its infancy in veterinary medicine, the potential for application of PET research to spontaneous cancer in animals cannot be overemphasized. Tumor-bearing companion animals can be incorporated into studies involving novel radiotracer and antineoplastic drug discovery, as well as optimization of existing chemotherapy protocols, surgical and radiation therapy planning.

References
  1. Alavi A, et al. Seminars in Nuclear Medicine 2004, 34(1): 56-69.
  2. Rohren EM, et al. Radiology 2004, 231(2): 305-332.
  3. Warburg O. Science 1956, 123: 306-314.
  4. Zanzonico P. Seminars in Nuclear Medicine 2004, 34(2): 87-111.
  5. Townsend DW. Annals of the Academy of Medicine Singapore 2004, 33: 133-145
  6. Fowler JS, Ido T. Seminars in Nuclear Medicine 2002, 32(1): 6-12.
  7. Cook GJR, et al. Seminars in Nuclear Medicine 2004, 34(2): 122-133.
  8. Townsend DW. Molecular Imaging and Biology 2004, 6(5): 275-290.
  9. Beyer T, et al. Journal of Nuclear Medicine 2000, 41: 1369-79.
  10. Page RL, et al. Journal of Nuclear Medicine 1994; 35: 1506-1513.
  11. Bassett CLM, et al. Molecular Imaging and Biology 2002, 4(3):201-207.
  12. Hustinix R, et al. Seminars in Nuclear Medicine 2002; 32: 35-46.
  13. Dobert N, et al. Cancer Biotherapy and Radiopharmaceuticals 2003, 18(4): 565-571.
  14. Peterson JJ, et al. Clinical Orthopedics 2003; 415(Suppl): S120-8.
  15. Blum RH, et al. Clinical Lymphoma 2003, 4(1): 50-51.
  16. Bender H, et al. Hybridoma 1999, 18: 87-91.
  17. Weber WA, et al. Journal of Clinical Oncology 2003; 21: 2651-2657.
  18. Spaepen K, et al. Journal of Clinical Oncology 2001; 19: 414-419.
  19. Schot B, et al. British Journal of Haematology 2003, 123(2): 282-287.
  20. Wells P, et al . Biochemica et Biophysica Acta 2004, 1705: 91-102.
  21. Wells P, et al. Cancer Research 2002, 62: 5698-5702.



© 2008 - Amy LeBlanc, DVM, DACVIM - All rights reserved