2013 Volume 230 Issue 3 Pages 155-169
Positron emission tomography (PET) with 18F-fluorodeoxyglucose (18F-FDG), a glucose analog, is widely used throughout the world as an indispensable imaging modality for the management of cancer treatment. This article reviews the pioneering achievements of PET in oncology with a focus on the development of PET that occurred from 1980 through the early-1990s. 18F-FDG was first applied for imaging of animal tumors in 1980 and for brain tumor imaging clinically in 1982. 18F-FDG enabled to visualize liver metastasis as clear positive image that could not be obtained by conventional nuclear imaging. Subsequently, 18F-FDG was used for imaging various cancers, such as lung, pancreas, colorectal and hepatoma. 11C-L-methionine (11C-MET) that reflects amino acid transport of cancers has an advantage that its uptake is lower in the brain and inflammatory tissue compared to 18F-FDG, and was first applied for imaging lung cancer and brain tumor. 18F-FDG and 11C-MET were proved to be sensitive tracers that can be used to objectively evaluate the effectiveness of cancer treatment. The diagnostic accuracy of PET, which is critical in clinical practice, was evaluated for the differential diagnosis of malignant and benign lung nodules using 18F-FDG or 11C-MET. In addition to 18F-FDG and 11C-MET, many radiopharmaceuticals were developed, such as 18F-labled thymidine analogs for evaluating proliferative activity, 18F-fluoromisonidazole for imaging of hypoxia, and 18F-fluorodeoxygalactose for evaluating liver-specific galactose metabolism and for imaging of hepatoma that retains galactose metabolic activity. These early efforts and achievements have greatly contributed to the development and clinical application of 18F-FDG PET in oncology.
Positron emission tomography (PET) using radiopharmaceuticals labeled with short-lived positron emitter can provide biochemical information of living human though quantitative tomographic images. 18F-fluorodeoxyglucose (18F-FDG), an analog of glucose and 2-deoxyglucose, was first developed by Ido et al. (1978). 18F-FDG is transported into the cells by glucose transporter, where it is phosphorylated to 18F-FDG-6-phophate by hexokinase and then remains trapped. Thus, accumulation of 18F-FDG in the tissue represents glucose consumption, or energy metabolism, in the tissue. 18F-FDG was developed firstly for the visualization and quantification of glucose utilization of the human brain using PET based on a rat model using 14C-2-deoxyglucose (Sokoloff et al. 1977). Later, Phelps et al. (1979) developed and validated a method for the tomographic measurement of glucose utilization in the human brain with 18F-FDG using PET. The method was applied for the evaluation of regional glucose consumption in diseased brain (Benson et al. 1981) and normal brain under stimulated conditions (Phelps et al. 1981). 18F-FDG has been used also for the imaging and evaluation of metabolism in the heart (Phelps et al. 1978).
An increased rate of glycolysis is one of the characteristics of cancer cells (Warburg 1956), and therefore, 18F-FDG accumulates and remains trapped in cancer cells. PET using 18F-FDG enables to visualize and quantify amount of glucose metabolism in cancers, which is closely correlated to viability of the tumors. In addition to glucose metabolism, Itoh et al. (1982) examined oxygen consumption of human brain tumors using 15O2 with PET and confirmed that brain tumor does not demand much oxygen even in relatively hypoxic condition. This is the first report that elucidated Warburg effect in human cancers. Currently, the most frequent clinical application of 18F-FDG PET is in oncology. In particular, 18F-FDG PET/CT, a PET system combined with X-ray computed tomography, is widely used as an indispensable diagnostic tool for the cancer management in the clinical setting (Gambhir et al. 2002). For example, 18F-FDG PET has been used to assess the staging of lung cancer patients (Fischer et al. 2001; Schrevens et al. 2002). Determining the stage of cancers are essential in not only determining optimal therapeutic strategy, such as surgery, radiotherapy or chemotherapy but also for predicting patient outcome. 18F-FDG PET is also used to monitor various malignancies (Poeppel et al. 2009). The uptake of 18F-FDG in cancer cells is correlated to tumor viability; therefore, 18F-FDG PET is useful to evaluate the effectiveness of therapies and for monitor recurrence of tumors. Recently, 18F-FDG PET was adopted as a method to objectively evaluate the effectiveness of tumor therapies (Wahl et al. 2009).
In addition to glucose metabolism, increased amino acid transport and protein synthesis are also characteristics of cancer cells. 11C-methionine (11C-MET) is currently the most commonly used for PET tumor imaging at present time because of its high tumor uptake and ease of radio-synthesis, although many labeled amino acids were developed.
Using labeled thymidine or -uridine analogs, PET enables to visualize and quantify non-invasively proliferative activity of human tumor, which is important to manage cancer treatment. Although many radiopharmaceuticals have been developed, 18F-fluorothymidine (Shields et al. 1998) is most commonly used.
Cancer cells occasionally express the phenotypes of their original tissues and the level of expression may represent the degree of differentiation. For example, iodine is a source for the synthesis of thyroid hormone, and radioiodine accumulates in well-differentiated thyroid cancers. Thus, in addition to evaluation of growth activity of tumors, another direction of PET tumor imaging is the visualization of distinct cancer cell phenotypes using phenotype specific radiopharmaceuticals. 18F-fluorodeoxygalactose (18F-FDGal) was developed to evaluate liver specific galactose metabolism (Fukuda et al. 1986). Another example is imaging of hypoxic cells in the tumor, which are resistant to radiotherapy or chemotherapy. Imaging of tumor receptor expression using labeled receptor ligands is the typical examples of this concept, however, no paper was published during the selected period in this article (1980-to the early 1990s).
As described above, PET tumor imaging has become an indispensable imaging modality for the evaluation and management of cancer. This article reviews the original and pioneering achievements of PET in oncology with a central focus on the development of PET that occurred from 1980 through early 1990s. Many Japanese research groups, especially from Tohoku University, significantly contributed to the development of PET in oncology. Based on these achievements, Matsuzawa et al. (1985) organized the first international symposium on “Current and Future Aspects of Cancer Diagnosis with Positron Emission Tomography: Biological and Clinical aspects.” This symposium was held two years prior to the second international symposium on “Positron Emission Tomography and Magnetic Resonance in Oncology” organized by Strauss et al. The fundamental achievements discussed in this review article are valuable for the further development and innovation of PET in oncology.
Som et al. (1980) was a pioneer in using 18F-FDG in oncological field. She published a paper describing the high and rapid accumulation of 18F-FDG in leukemia cells transplanted into mice, although she did not present any image. 18F-FDG was first utilized clinically for the visualization and evaluation of human brain tumors by Di Chiro et al. (1982). They showed that 18F-FDG uptakes of high-grade gliomas were higher than low-grade gliomas, although high uptake of 18F-FDG in the normal brain is disadvantage for brain tumor imaging. 18F-FDG PET has also been used for the differential diagnosis between recurrence of brain tumor and necrosis of the normal brain after treatment (Patronas et al. 1982). Subsequently, 18F-FDG has been used for imaging of various types of tumors.
Intrahepatic tumors (liver metastasis)Fukuda et al. (1982a, b) showed an increased uptake of 18F-FDG in a rat tumor model and visualized intrahepatic tumor clearly (Fig. 1). Yonekura et al. (1983) and Fukuda et al. (1983a, b, 1984) were the first to image clinically liver metastases from colon cancer using 18F-FDG-PET. Immediately after 18F-FDG injection, a dynamic scan was performed every 5 min for a total of 50 min. They observed that tumor radioactivity increased with time, whereas liver radioactivity decreased with time, which resulted in the clear visualization of the intrahepatic tumor (Fig. 2). This was difficult to achieve using gallium-67, which is mainly used in nuclear oncology performed in hospitals where PET is not available. The ability to obtain positive and clear images of intrahepatic tumor is attributed to the relatively low uptake of 18F-FDG in the normal liver. This is because the liver has a high activity of glucose-6-phosphatase, which de-phosphorylates 18F-FDG-6-phosphate to 18F-FDG so that it can easily re-enter into the blood stream.
An autoradiogram of a AH109A tumor-bearing rat one hour after injection of 18F-FDG.
The intrahepatic tumor, which was directly inoculated into the liver, was clearly visualized as a highly radioactive area (arrow). Fukuda, H., et al. (1982) Eur. J. Nucl. Med., 7, 294-297 (reproduced with permission of Springer-Verlag GmbH).
PET imaging of liver metastases with 18F-FDG.
A dynamic scan every 5 min after injection of 18F-FDG (from top left to bottom right). Increased uptake of 18F-FDG with time (A: arrow) was observed in liver metastases from colon cancer on CT (B: arrow), while the uptake in the normal liver and spleen decreased with time. The PET image was obtained by ECAT-II (EG & G ORTEC), a single-slice PET with spatial resolution of 15 mm.
Fukuda et al. (1983a) were the first to report successful imaging of hepatocellular carcinomas (HCC) with 18F-FDG. The uptakes of 18F-FDG in 3 tumors was semi-quantified by differential absorption ratio (DAR), which normalizes tissue tracer uptake to subject body weight and amount of injected radioactivity and were determined to 2.10, 3.87 and 4.10 at 50 min after the injection of 18F-FDG were. Later, Paul et al. (1985) reported 18F-FDG imaging of a HCC using a gamma camera and not PET. Relatively low uptake of 18F-FDG in HCC might be attributed to high glucose-6-phosphatase activity in well-differentiated HCCs. Okazumi et al. (1992) confirmed this hypothesis by comparing the uptake of 18F-FDG and HCC of various grade. A 2-compartment, 4-parameter (k1-k4) analysis of 18F-FDG uptake in HCC revealed that a higher grade of HCC had larger k4 values, which is a back flow constant from the second to first compartment and may correspond to glucose-6- phosphatase activity.
Pancreatic cancerIncreased 18F-FDG uptake in pancreatic cancers was first reported by Fukuda et al. (1983a, 1984, 1987a). Fig. 3 shows the increased uptake of 18F-FDG that was observed in a pancreatic head tumor by X-CT. 18F-FDG was injected in 3 separate cases of pancreatic cancer, including one with liver metastasis. After 50 min, DAR values were determined to be. 2.83, 1.94 and 3.55. Later, Klever et al. (1992) and Bares et al. (1993) reported on several cases of pancreatic cancer.
PET imaging of pancreatic cancer with 18F-FDG.
A: A PET image with 18F-FDG in a pancreatic cancer patient. The accumulation of 18F-FDG was observed in the pancreas head tumor (arrow), which was shown on the CT image (B). K: kidney
B: CT image. Pancreas head tumor (arrow).
Fukuda, H. et al. (1984) CYRIC Annual Report 1983, 244-249.
Fujiwara et al. (1984) were the first to report increased 18F-FDG uptakes in the malignant lung tumors of 9 patients. They performed a dynamic scan every 5 min and showed that 18F-FDG uptake in the lung tumor increased with time in all cases. The results were confirmed by DAR values that increased from initial value of 3.0 to 9.4 after 50 min. By fitting an 18F-FDG-accumulation curve, they calculated the increase rate of DAR (γDAR) and showed that squamous cell carcinomas had larger γ DAR than adenocarcinomas. Later, Nolop et al. (1987) expanded the initial experience of them (Fujiwara et al. 1984).
Orbital tumorKiyosawa et al. (1983) examined 18F-FDG uptakes in a VX2 tumor transplanted into the orbit of rabbits. 18F-FDG uptake in the VX2 tumor was high, whereas uptake in chemically induced inflammation was relatively low. Kiyosawa et al. (1985) extended this study to clinical trials and performed 18F-FDG PET imaging of orbital tumors. They observed high uptake of 18F-FDG uptake in malignant lymphomas were high; however, they observed high uptake in inflammatory lymphoid tumors. These were the initial applications of FDG PET in ophthalmology.
Breast cancerKubota et al. (1989a) reported a focally increased uptake of 18F-FDG in a breast cancer patient. Subsequently, Wahl et al. (1991) showed that 12 breast cancer patients with primary tumors larger than 3 cm had increased glucose metabolism. Several authors have confirmed these findings in small study groups and Adler et al. (1993) evaluated the diagnostic accuracy of FDG-PET in 28 patients.
Malignant lymphomas and the correlation with patient outcomePaul (1987) the first to image lymphoma with 18F-FDG, but Okada et al. (1991) evaluated the correlation between the 18F-FDG uptake of 21 patients with malignant lymphoma and the initial outcome. 18F-FDG uptake was evaluated by the tumor-to-normal soft tissue contrast ratio (TCR) and the glucose utilization rate (GUR), which was calculated by a graphical plot method by Patlak and Blasberg (1985). Higher TCRs and GURs were observed in patients with poor prognoses. Okada et al. (1992) compared the uptake of 18F-FDG in malignant lymphomas to the tumor proliferative activity of the tumor using Ki-67 labeling index. The 18F-FDG uptake as quantified by TCR or DAR correlated well with Ki-67 labeling index.
Colorectal cancer recurrenceStrauss et al. (1989) first reported the significance of PET imaging in 29 patients with recurrent colorectal cancers. Ito et al. (1992) further investigated the value of 18F-FDG PET and MR imaging in differentiating recurrent rectal cancer from scar tissue in 15 patients with suspected recurrence. All 11 masses with confirmed cancer recurrence showed significantly higher uptakes as expressed by DAR with values of 4.73 ± 2.28, whereas 4 scar tissues that were confirmed by biopsy had decreased 18F-FDG with low DAR values of 0.97 ± 0.15. Combined with MR imaging, most recurrent tumors could be differentiated from scar tissues except one case. They concluded that PET and MR were complementary modalities for the differential diagnosis of tumor recurrence and scar tissue. Ito et al. (1996) also examined the correlation between the 18F-FDG uptakes in rectal cancer and the tumor histology and observed that the amount of 18F-FDG uptake correlated with the cellularity of the tumor tissue in histology.
18F-FDM is an isomer of 18F-FDG and was obtained as a byproduct of an electrophilic substitution reaction of 18F-FDG synthesis. Fukuda et al. (1982b) showed that the uptake of 18F-FDM and 18F-FDG in a poorly differentiated rat hepatoma was almost identical. In addition, 18F-FDM uptake was found to be 30% less than 18F-FDG uptake in the brain and 18F-FDM was cleared more rapidly from the blood. PET with 18F-FDM successfully visualized a VX2 tumor that was transplanted into a rabbit. These results suggested that 18F-FDM could be a more suitable radiopharmaceutical for PET cancer imaging than18F-FDG. However, a clinical PET study was not performed at the time because of difficulty in synthesizing 18F-FDM with a high yield and purity. Since then, substantial improvements and developments in radio-synthesis have been achieved in recent years. Recently, Furumoto et al. (2009) developed a new method of 18F-FDM synthesis using a nucleophilic substitution reaction to with a high yield and purity, and bio-distribution and PET imaging studies of tumor bearing rats confirmed the potential of this tracer in clinical application (Furumoto et al. 2013, in press).
In addition to increased glucose metabolism, increased amino acid transport and metabolism is also characteristics of cancer cells. Methionine is a physiological amino acid that is transported into cells by neutral amino acid transporter and metabolized. 11C-MET was developed by Comar et al. (1976) and the first clinical application of 11C-MET PET for pancreatic cancer imaging was reported by Syrota et al. (1982). However, pancreatic cancer could not be distinguished from chronic pancreatitis because both lesions were visualized as a decreased uptake of 11C-MET MET. However, Kubota et al. (1983) successfully imaged a lung cancer using PET with 11C-MET (Fig. 4). Later, Kubota et al. (1985) analyzed 10 cases of lung tumors and reported that 8 lung cancers showed higher uptake than 2 benign lung tumors. Fujiwara et al. (1989) evaluated the correlation between the histological type of 16 lung cancers and 11C-MET uptake and reported that large cell carcinomas had higher 11C-MET uptake compared to squamous cell carcinomas.
Although many labeled amino acids have been developed, such as 11C-luecin (Barrio et al. 1983), 11C-tyrosine (Bolster et al. 1986) and the non-natural amino acid-11C-aminocyclo-pentanecarboxylic acid (ACPC) (Conti et al. 1986), 11C-MET is commonly used at present time because of its higher tumor uptake and ease of radio-synthesis. Kubota et al. (1984) compared the uptakes of 10 different 11C-labeled amino acids in rat tumors and found that 11C-MET demonstrated the highest tumor uptake followed by 11C-ACPC.
Because of their low uptake in the normal brain tissue, 11C-MET and other amino acids are superior to 18F-FDG in brain tumor imaging. Lilja et al. (1985) and Ericson et al. (1985) were the first to report successful brain tumor imaging with PET using 11C-MET. Another advantage of 11C-MET compared to18F-FDG is relatively low uptake in the inflammatory tissues (Kubota et al. 1995).
Imaging of lung cancer with 11C-MET
Twenty minutes after injection of 11C-MET, the tumor in right upper lobe was clearly visualized. Kubota, K., et al. (1983) The Lancet, 2, 1192-1193 (reproduced with permission of Elsevier Limited).
Boron neutron capture therapy (BNCT) is a radiotherapy for cancer utilizing boron-10 (10B), which efficiently absorbs thermal neutrons and releases high LET α-particles (230 keV/µm) via a 10B(n, α)7Li reaction. The range of high LET α particle is only approximately 10 µm, which is comparable to the diameter of a cell, therefore, thermal neutron irradiation causes substantial damage limited to cells that take up the tumor-seeking 10B-compound. Mishima and Fukuda et al. initiated a clinical trial of BNCT for malignant melanoma with successful results (Mishima et al. 1989a, b; Fukuda et al. 1989, 1994, 1999, 2003). However, 10B-concentrations in both the tumor and normal tissues could not be directly measured during neutron irradiation, which cause an inaccurate estimation of radiation dose. Therefore, use of PET preferred to visualize and quantify 10B-content in the tumor by the fluorine-18 fluorination of BPA. Ishiwata et al. (1991) developed the 18F-fluorination of 10B-BPA and Imahori et al. (1996, 1998a, b) performed clinical trials and proposed the 18F-FBPA PET system for BNCT. Currently, 18F-FBPA PET is an indispensable method for BNCT for determining indication of BNCT, predicting absorbed dose by BNCT and evaluating effectiveness of treatment after BNCT. A tumor-to-normal tissue uptake ratio of 2.5 or more with 18F-FBPA is considered to be indicative of BNCT (Fig. 5), and the ratio is subsequently used for the estimation of radiation dose to be delivered to the tumors and normal tissues by BNCT.
PET image using 18F-10B-parabononophenylalanine (BPA) in a patient with a right parotid cancer.
18F-10B-BPA PET clearly visualized the submandibular tumor (arrow) with a tumor-to-contralateral normal tissue ratio of 5.0. A: PET image, B: contrast-enhanced CT (Courtesy of Prof. Jun-ichi Hiratsuka).
Proliferative activity can be evaluated using labeled thymidine analogs such as 14C or 3H-labeled thymidine and 125I-lableled iodo-deoxyuridine in animals or in vitro models. Fluorinated pyrimidine analogs such as 5-furuorouracil (5-FU), 5-fluoro-uridine (5-FUR) and 5-fluorodeoxyuridine (5-FdUR) have been used as antitumor agents, and their effects have been well acknowledged. Accordingly, 18F labeling of these compounds enables visualization of tumor proliferative activity with PET. Abe et al. (1983a) and Ishiwata et al. (1984) compared the uptakes of 18F-fluorinated pyrimidines, 18F-5-FU, 18F-5-FUR, and 18F-5-FdUR in an animal tumor model. They performed bio-distribution studies in tumor-bearing rats and obtained PET image of rabbit tumors before concluding that 18F-5-FdUR was superior in terms of tumor uptake and tumor-to-normal tissue ratio. A preliminary PET study with 18F-5-FdUR in lung cancer patients revealed a clear visualization of the tumors (Abe et al. 1985). Kiyosawa et al. (1986) performed PET imaging of rabbit orbital tumors with 18F-5-FdUR and clearly visualized the tumor with a high tumor-to-normal orbit ratio. Tsurumi et al. (1990) tested the feasibility of 18F-5-FdUR for PET imaging of nucleic metabolism using a rat brain tumor model. A double tracer autoradiogram with 18F-5-FdUR and 14C-thymidine revealed similar brain-tumor images. Furthermore, a metabolite study revealed that the radioactivity of the nucleotide and acid-insoluble fractions increased with time, whereas the radioactivity of entire tumor remained unchanged. These results indicated that 18F-5-FdUR uptake in brain tumors correlates with nucleic acid metabolism and that 18F-5-FdUR might be a useful tracer for PET tumor imaging.
However, 18F-5-FdUR had not been used in further studies, because of its low tumor-to-background ratio. 11C-thymidine was developed by Sundro-Wu et al. (1984), however, the tracer yielded many labeled metabolites in the blood, which made it difficult to evaluate rate of proliferative activity. Shields et al. (1998) developed 18F-fluoromethyl-thymidine, which is an effective substrate for thymidine kinase, found to be trapped in proliferating cells, and it has been now commonly used.
The ability to eliminate galactose is an important function of the liver and was used as liver function test. Tada et al. (1984, 1987) developed 18F-FDGal to image galactose metabolism in the liver using PET. Fukuda et al. (1986) demonstrated high 18F-FDGal uptake in the normal rat liver (4.47% injected dose/g of tissue), however, uptake decreased by 13% (3.89% ID/g) in livers damaged by CCl4 treatment. Ishiwata et al. (1988) confirmed that 18F-FDGal is phosphorylated by galactokinase in the liver and remains trapped after the second metabolic step of the galactose pathway. Ishiwata et al. (1989) confirmed that this two-step metabolic trap also occurs in hepatomas. Because pronounced activity of these enzymes is found specifically in the liver, 18F-FDGal can be used to evaluate cirrhotic liver function and to image hepatocellular carcinoma. After confirming the acute toxicity of the compound and determining the radiation dose (Fukuda et al. 1987c), Fukuda et al. (1987d) initiated a clinical PET study. They observed a high accumulation in the liver, relatively high in the kidney and small intestine, and low uptake in other tissues. In a preliminary PET study, Fukuda et al. (1987b, 1988) found that 18F-FDGal accumulates in hepatocellular carcinomas (HCCs) that originate from hepatocytes, but not in metastatic liver tumors of other origins. To investigate the mechanisms underlying the high uptake in HCCs, Fukuda et al. (1993) examined 18F-FDGal uptake in the following animal tumor model: well-differentiated spontaneous hepatoma (Spt HCC) in C3H mice, well-differentiated Morris hepatoma (5123D) in Buffalo rats and poorly differentiated hepatoma (AH109A) in Donryu rats. The uptake of 18F-FDGal by well-differentiated spHCC and 5123D tumors was 92% and 33% of the normal liver uptake, respectively. However, the uptake by the poorly differentiated AH109A tumors was only 16% of the normal liver uptake (Fig. 6). Later, Fukuda et al. (1994) reported visualization of bone metastases from HCC with very low background levels using 18F-FDGal PET. Fig. 7 demonstrates an orbital metastasis from HCC. Currently, a whole body PET scan with 18F-FDGal would be useful for a systemic survey of HCC metastases.
Autoradiograms of tumor-bearing rats and mice with 18F-FDGal.
A: A Donryu rat bearing a poorly differentiated hepatoma, AH109A.
B: A Buffalo rat bearing a well-differentiated Morris hepatoma, 5123D.
C: A highly differentiated spontaneous hepatoma of a C3H mouse. Arrow heads and arrows indicate tumor and normal liver, respectively. Fukuda, H. et al. (1993) J. Nucl. Med., 34,780-786 (reproduced with permission of Society of Nuclear Medicine).
PET images of an orbital metastasis from hepatocellular carcinoma with 18F-FDGal and 18F-FDG.
The mass in the right orbit on CT (B) was confirmed by a biopsy to be a metastasis from hepato-cellular carcinoma. The 18F-FDGal uptake by the tumor was high with a low background activity (A). Dashed line; contour of the head.
Glucosamine and its derivatives are essential structural components of many important biological macromolecules, including membrane glycoproteins and mucopolysaccharides. Hyaluronic acid is the most abundant mucopolysaccharide present in the cell membrane. It was reported that an invasive VX2 carcinoma grown in rabbit contained a greater quantity of hyaluronic acid than the same tumor grown in nude mice, where it was less aggressive. The evaluation of hyaluronic acid synthesis in vivo may be useful for the evaluation of the invasiveness of cancer cells. Therefore, Tada et al. (1989) synthesized a structural analog of N-acetyl-glucosamine, N-[18F]fluoroacetly-D-glucosamine (18F-FAG). Fujiwara et al. (1990) demonstrated that the uptake of 18F-FAG in well-differentiated, spontaneous hepatoma of C3H mice (Spt HCC) was substantially high, suggesting that 18F-FAG might be a potential tumor-imaging agent.
18F-fluoromisonidazole: imaging of hypoxic cells in the tumorMisonidazole and its derivatives are metabolically trapped in hypoxic cells, which are resistant to radiotherapy and chemotherapy. Therefore, visualization and quantification of hypoxic fraction within a tumor are useful tools in determining effective treatment strategy. 18F-fluoromisonidazole (18F-FMISO) was developed by Jerabeck et al. (1986) and Koh et al. (1992) reported initial success with PET imaging of 8 human malignant tumor. Currently, 18F-FMISO is used to determine treatment methods and predict patient’s outcome after radiotherapy, which based on the kinetic behavior of FMISO in tumor tissue (Eschmann et al. 2005).
Abe et al. (1983b, 1986) first assessed the effect of radiotherapy on experimental mouse and rat tumor models using 18F-FDG. They used two types of mouse mammary carcinomas models, FM3A and MM48 that were radiosensitive and radio-resistant, respectively. The tumor volume of the radiosensitive FM3A began to decrease two days after 20-Gy irradiation and the uptake of 18F-FDG rapidly decreased 1 day after irradiation. However, the tumor volume of the radio-resistant MM48 tumor continued to increase even after irradiation, with no decrease in 18F-FDG uptakes. Using the rat hepatoma line AH109A, they evaluated the radiation dose-response of tumor growth and the 18F-FDG uptake kinetics (Fig. 8). After 10-Gy irradiation, the tumor volume decreased one day after irradiation and increased again at 6 days. The decrease and increase of tumor 18F-FDG uptake after irradiation well correlated with tumor volume changes. Abe et al. (1990) first evaluated the 18F-FDG uptake of the human lung cancer before and after irradiation and showed that 18F-FDG can be used for the assessment of radio-therapeutic effectiveness in the clinical setting.
Kubota et al. (1989b) examined the effects of radiation on 11C-MET uptake by tumors. The 11C-MET uptake by tumors decreased 50% 12 hours after 20-Gy irradiation. However, the tumor volume continued to increase until one day after irradiation, and then decreased. He measured the necrotic fraction within the tumor after irradiation and found that the decreased rate of 11C-MET was larger than the decreased viable fraction of the tumor as measured by histological section. He concluded that 11C-MET could be a more sensitive marker for the radiation response of the tumor than 18F-FDG. Next, Kubota et al. (1991) compared the different uptake patterns after irradiation of 5 tracers (18F-FDG, 18F-Flurodeoxyuridine, 11C-MET, 3H-Tdr and 67Ga citrate) and showed that 11C-MET and 3H-Tdr were the most rapid and sensitive tracers for monitoring the effects of radiotherapy. This paper was also the first to demonstrate that the decrease in 18F-FDG uptake by tumors after irradiation was almost parallel to the decrease in the viable cell fraction in the tumor. Later, Higashi et al. (1993) demonstrated that the 18F-FDG uptakes by tumors correspond to the fraction of viable tumor cells but not to proliferating cells. Kubota et al. (1992a) showed that after 10 Gy irradiation, the 11C-MET uptake decreased earlier than the tumor volume reduction and later, it significantly increased earlier than the recurrent tumor growth and that 11C-MET is a sensitive indicator for monitoring radiotherapeutic effect and recurrence.
Kubota et al. (1993) extended the 11C-MET study to the clinical setting and examined 11C-MET uptake in 21 patients with lung cancer before and during the two weeks after radiotherapy. The group with local control (no recurrence) showed a larger decrease in 11C-MET uptake. The early-recurrence group showed the smallest decrease in 11C-MET uptake. The late-recurrence group (after 11 months or more) showed a similar decrease to the no-recurrence group in the amount of 11C-MET uptake. However, no recurrence- and late-recurrence groups could be distinguished by the amount of tumor volume decrease observed by CT. The late-recurrence group showed a smaller volume decrease. This report was the first to use 11C-MET for the evaluation of the radio-therapeutic effect on cancers in the clinical setting except for a case report by Kubota et al. (1992b), in which they demonstrated that 11C-MET could detect lung cancer recurrence, while the residual mass volume remained almost constant during the follow-up period after radiotherapy.
18F-FDG uptakes of rat tumor after different radiation doses.
18F-FDG uptake continuously decreased after irradiation except for the 10-Gy-irradiation group, where uptake again increased at 7 days. The secondary increase in uptake corresponded with tumor re-growth after 4 days (left panel). Left: tumor volume, Right: 18F-FDG uptakes. Control (●), 10 Gy (○), 20 Gy (▲), 40 Gy (△). Abe, Y., et al. (1986) Eur J Nucl Med., 12, 325-328 (reproduced with permission of Springer-Verlag GmbH).
The evaluation of the effectiveness of anti-cancer chemotherapy is commonly performed based on changes in tumor size. As described above, 18F-FDG and 11C-MET are more sensitive markers for the evaluation of radio-therapeutic effects on cancers than tumor size change. Takahashi et al. (1985) examined the 18F-FDG uptake changes of rat AH272 tumors after chemotherapy with mitomycin C (MMC) or Adriamycin (ADR). The former is an effective chemotherapy agent against AH272 tumors, and the latter is not. A substantial decrease (−63%) of 18F-FDG uptake was observed after the administration of 1.25 mg/kg of MMC with a 29% decrease in tumor volume. In contrast, the administration of 4.0 mg/kg of ADR yielded no tumor volume reduction and only a small reduction (−8%) in 18F-FDG uptake. Takahashi et al. (1986) examined the changes in 18F-FDG uptake of the tumor before and after chemotherapy in 4 advanced-cancer patients. The response estimated by the DAR ratio (DARbefore / DARafter) of 18F-FDG correlated well with conventional response criteria. He concluded that 18F-FDG appeared to be promising for the evaluation of the efficacy of cancer chemotherapy.
Kubota et al. (1990) first reported the diagnostic accuracy of 18F-FDG PET with a prospective study for the differential diagnosis of 22 lung nodules. The sensitivity, specificity and accuracy for the differential diagnosis of benign and malignant lung nodules were 83%, 90% and 86%, respectively. These values were confirmed in many later papers (Gambhir et al. 2002; Schrevens et al. 2002). This may be a key paper that promoted and realized clinical PET using 18F-FDG in oncology because high diagnostic accuracy is very important in clinical setting. In this paper, he also first reported the diagnostic accuracy of 11C-MET PET for the differential diagnosis of 24 lung nodules in a prospective study with a sensitivity, specificity and accuracy of 93%, 60% and 79%, respectively.
Abe et al. (1988) compared the intra-tumoral regional distribution of 14C-MET and a blood flow tracer, 18F-fluoro-antipyrine (FAP), using a double-labeled autoradiogram. The density of the film was digitized, and their correlation was tested by a (x, y) scatter plot of the two tracers. The results revealed that the regional distribution of 14C-MET was almost linearly correlated with FAP. Metabolite analysis revealed that 50% of the 14C-MET was found in the acid-insoluble fraction 10 min after administration. The results indicated a coupling of 14C-MET metabolism and blood flow within an experimental tumor. He also examined correlation between regional uptake of 18F-FDG and FAP and found no correlation between two.
Microscopic auto-radiographic studyKubota et al. (1992) developed micro-autoradiogram techniques using the short-lived F-18. Using this technique, they examined the intra-tumoral distribution of 18F-FDG at the tissue or cellular level and reported a substantially high uptake of 18F-FDG in macrophages and the granulation tissue around the tumor and in tumor cells. This paper had a striking impact on 18F-FDG PET in oncology. We should consider that the 18F-FDG uptake of tumors includes the substantially high fraction of uptake by infiltrating immune cells and granulation tissues. In particular, we should consider that the 18F-FDG uptake of tumors after anti-cancer treatment includes not only the residual uptake by tumors but also uptake by inflammatory tissues induced by the therapy. Kubota et al. (1994) examined the 18F-FDG uptake dynamics in cancer cells, macrophages, granulation tissue and pre-necrotic tissues and showed that the uptake per unit volume in macrophages was the highest and most rapid among the tissues and continued to increase during the 60-min study period. However, uptakes in the FM3A and MH134 mouse mammary carcinomas and in the granulation tissue were relatively lower than in macrophages and reached a plateau approximately 30 min after injection of 18F-FDG. From these results, she concluded that viable neoplastic tissue could be differentiated from non-neoplastic tissue by means of a dynamic analysis of the tracer uptake.
Kubota et al. (1995) compared the intra-tumoral distribution of 11C-MET and 18F-FDG at the cellular level using micro-autoradiography. A substantial uptake of 11C-MET was found mainly in viable tumor cells with a relatively low uptake in granulation cells and macrophages. This is in contrast to the 18F-FDG uptake, which was found in granulation cells and macrophages as high as in the tumor cell. From these results, she concluded that 11C-MET might be suitable for treatment evaluation.
Fukuda et al. (1983a) first proposed the differential absorption ratio (DAR) as a semi-quantitative parameter of the 18F-FDG uptakes in tumors. DAR represents tissue uptake (%dose /g tissue) normalized by injected radioactivity and body weight of the animal, and has been used as a measure of the relative tissue uptake of radiotracers in animal studies. Kubota et al. (1985) used the differential uptake ratio (DUR) instead of the DAR in his paper. Strauss et al. (1989) referred to DAR and used it as a parameter for 18F-FDG uptakes in his paper describing PET imaging of recurrent colorectal cancers. After Fukuda introduced DAR, the standardized uptake value (SUV) (Hübner et al. 1996; Freedman et al. 1996) was proposed and has been commonly used instead of DAR, although the definition of SUV was completely identical to that of DAR or DUR.
When Fukuda et al. (1983a) first introduced DAR as a semi-quantitative parameter for 18F-FDG uptake of tumors, most of the scientists in PET research claimed that the absolute glucose utilization CMRglc (mg/ml-tissue/min) should be used, even for tumors. However, there are at least two reasons for not calculating the tumor CMRglc, which is based on the normal brain model. First, lumped constant (LC), which is a total efficiency of glucose consumption in a given glucose analog compared to that by glucose, may be higher in cancers than that in normal brains, and LC may vary depending on tumor type or aggressiveness. There are classical papers in biochemistry describing the low substrate specificity of tumor hexokinase for glucose and the higher uptake of deoxyhexoses (Bessel et al. 1972). This is comparable to higher LC values for cancers. Therefore, the LC values for the normal brain, 0.42 (Phelps et al. 1979), or 0.52 (Reivich et al. 1985), should not be applied to cancers. Later, a paper was published describing higher (LC = 1.4) and variable LC values in human brain tumors (Spence et al. 1998) compared to the normal brain. Another reason for not using the CMRglc is that the brain and the tumor show different competition patterns of 18F-FDG uptakes with blood glucose levels (transport and phosphorylation levels). The 18F-FDG uptakes by the brain negatively correlated with blood glucose levels for a certain range of blood glucose; however, the tumor 18F-FDG uptakes are relatively insensitive to blood glucose levels (Yamada et al. 1985; Ishizu et al. 1994). Currently, the SUV (DAR) is widely accepted as a parameter for relative tumor 18F-FDG uptakes.
As introduced here, Japanese scientists, mainly members at Tohoku University, published many original papers and substantially contributed to the development of PET in oncology, in particular at the initial developmental stage from 1980 to early 1990s. These distinguished achievements were summarized in Table 1. Kubota et al. (2001) has already summarized the clinical significance of PET and their biological bases obtained from our studies in a previous review article. A review article on “oncological application of FDG-PET” by Czrernin (2004) was helpful for preparing this manuscript and determining who performed it first in PET oncology.
Chronological table for the development of PET in oncology.
BNL, Brookhaven National Laboratory; TU, Tohoku University; Heidelberg, Heidelberg Cancer Center; TUA, Technical University of Aachen.
The authors thank Drs. Iwata, R., Ishiwata, K., Takahashi, T., Tada, M., Furumoto, S., Ido, T., Watanuki, S. (Cyclotron and Radioisotope Center, Tohoku Univ.), and Yamada, K., Itoh, M., Abe, Y., Yoshioka, S., Hatazawa, J., Kiyosawa, M., Yamada, S., Ito, K., Fujiwara, T., Kubota, R., Yamaguchi, K., Sato, T., Takahashi, J. (IDAC, Tohoku Univ.) for their cooperation and contributions in studies described in this article.
We declare no conflict of interest.
