SHORT COMMUNICATION 
Annals of Nuclear Medicine Vol. 8, No. 1, 75-83, 1994 

Autoradiographic analysis of [14C]deoxy-D-glucose in thyroid cancer xenografts: A comparative study with pathologic correlation 

Masahiro KUBOTA,* + Takatoshi TSUDA,* + + Takashi MINASE,** Kunihiro NAKADA*** and Masayori FURUDATE*** 

*Department of Radiology, School of Medicine, Sapporo Medical University 
**Clinical Laboratory of NTT Sapporo Hospital 
***Department of Nuclear Medicine, Faculty of Medicine, Hokkaido University 
+ Department of Radiology, Teine Keijinkai Hospital 
+ + Department of Radiology, Kushiro Chuo Hospital 

An experimental model of thyroid cancer was prepared for evaluating the accumulation of [14C]deoxy-D-glucose ([14C]DG) in thyroid cancer xenografts (AC2). A continuous cell line established from a biopsy specimen of a metastatic thyroid carcinoma possessed the ability to synthesize the cellular protein without increase in cell division after adding bovine TSH in vitro. The histological sections of the xenografts resected from the 131I treated nude mice mainly consisted of structures showing follicular and trabecular growth. Immunohistochemically the cytoplasm of the tumor cells was positive for human thyroglobulin(hTg). These observations provide strong evidence that the AC2 cell originates in the thyroid follicular epithelium. By comparing autoradiographic accumulation patterns of [14C]DG and histopathological examinations, it was found that the uptake of [14C]DG was higher in the granulation tissues surrounding necrosis than in viable tumor cells of trabeculary growing and follicle forming tissues. 
It is suggested that the degree of [14C]DG content reflects not only tumor cell viability and proliferation but also the inflammatory and degenerative reaction accompanying tumor cell growth. 

Key words: thyroid carcinoma xenograft, [14C]deoxy-D-glucose ([14C]DG), autoradiography 

INTRODUCTION 
INCREASED GLYCOLYSIS is one of the most important characteristics of cancer cells. Studies on the glucose metabolism of experimental tumors have been reported by many authors. 1-3 
Clinically, it is reported that glucose consumption measured by [18F]fluoro-2-deoxy-D-glucose (FDG) with PET was found to correlate well with the viability of cancer cells and the prognosis of cancer patients. 4-8 However, these previous studies focused on evaluating the viability and growth activity of cancer cells, which are nonspecific characteristics independent of the cancer type. Cancer cells occasionally express the phenotypes of their 

Received October 21 , 1992, revision accepted November 24, 1993. 
For reprint contact: Masahiro Kubota, Department of Radiology, Teine Keijinkai Hospital, Maeda 1-12, Teine-ku, Sapporo 006, JAPAN. 

original tissues, and the degree of expression may represent the degree of differentiation. A clinical study on the metabolic heterogeneity of thyroid carcinoma carried out with FDG suggests that metastases that accumulate FDG, but not 131I may behave more aggressively than metastases that accumulate 131I, but not FDG. 9 
We investigated the accumulation of [14C] DG in athymic nude mice with thyroid cancer xenografts to analyze the correlation with the degree of differentiation and glucose consumption. In this present study, we report a distinctive pattern of uptake of [14C]DG by the secondary inflammatory reaction cells in a thyroid cancer model in mice. 

MATERIALS AND METHODS 
Thyroid carcinoma 
A 58-year-old Japanese male patient was referred to the Sapporo Medical College Hospital in 1984. Metastasis was found in the right supraclavicular lymph nodes. Two courses of radioiodine therapy with 7.2 GBq of 131I were carried out after thyroidectomy. The primary lesion was pathologically diagnosed as papillary carcinoma. Wide-spread metastases occurred three months after second 131I therapy. At this time, a metastatic tumor of the anterior chest wall was biopsied. Histological examination confirmed metastatic thyroid carcinoma with anaplastic transformation. Thyroid carcinoma was obtained from the biopsied specimen (Fig, 1). 

Establishment of AC2 cell line 
The biopsied tissue was cut as finely as possible, and cells obtained from the tumor pieces were suspended in 20 ml of Eagle's medium (Nissui Pharma, Tokyo, Japan) supplemented with 10% fetal bovine serum (FBS) (Hyclone, Logan, Utah). The cells were cultured in 25 cm 2 flasks (Falcon 3013, Falcon Plastics, Oxnard, CA) at 37deg.C in a humidified atmosphere of 5% CO2 in air. During the subsequent 4 weeks of primary culture, colonies of small epitheloid cells attached to the flasks were observed and grew slowly. The tumor cell colonies were surrounded by a considerable number of fibroblastic cells. During a further 5 weeks of primary culture, it became apparent that a homogeneous population of small epitheloid cells had gradually propagated and reached confluency. The cell line was passaged with 0.25% trypsin and 0.02% EDTA in PBS. 

Radiolabeling of proteins in thyroid cell culture 
[35S] methionine (41.8 TBq/mmol ; E I Dupont de Nemours Inc, N. Billerica, MA) was added to the medium ( 10 micro Ci / ml; Modified Eagle's MEM) 2 hours before the end of the culture with bovine thyroid stimulating hormone (bTSH) ( 100 micro IU/ml) or without bTSH. Cultured cells were extensively but quickly washed with PBS, then isolated by centrifugation and stored at -80deg.C. The cells were homogenized ( 10 x 10 6 cells/0.6 ml) in PBS and mixed with hot trichloroacetic acid (TCA). Then the mixture was centrifuged at 27000 r.p.m. for 5 minutes. The precipitated protein fractions were trapped on a 0.5 micro m glass filter and washed with cold 10% TCA solution. Their radioactivity was measured in a scintillation counter. 

Animal model 
Ten 4-week-old nude mice (BALB/cAJcl-nu) were subcutaneously inoculated with 1 x 10 7 to 1 x 10 8 AC2 cells. Four weeks before the cell inoculation, five mice were given 3.7 MBq of 131I for their thyroid ablation. Four to eight weeks after the inoculation, tumors in the treated nude mice were resected for light microscopical and autoradiographic examinations. Five other mice without 131I treatment were used for histopathological examination only. 
Histopathological examination 
The extirpated xenograft was fixed with 10% formalin in PBS, embedded in paraffin, and stained with hematoxylin and eosin (H-E) for light microscopic examination. For immunohistochemical analysis of human thyroglobulin (hTg), tissue sections were deparaffinized in xylene, washed with PBS, and rehydrated. They were then applied for the detection of hTg. Immunostaining was achieved by the avidin-biotin peroxidase complex (ABC) method. Endogenous peroxidase was removed by incubation in a mixture of 100 ml absolute methanol and 1 ml 30% hydrogen peroxide for 30 minutes. The slides were also incubated with normal horse serum to avoid non-specific binding. Subsequently, anti-human 27S, 19S, and 12S thyroglobulin (hTg) mouse monoclonal antibody (Sanbio Bv-Biologlcal Products, Uden, The Netherlands) was applied to the sections. Goat biotinylated anti-mouse antisera and ABC were then successively placed on the tissue section for 30 minutes, according to the manufacturer's instructions (Vector Laboratories, Burlingame, CA). Peroxidase activity was examined by staining with 0.1 % diaminobenzidine tetrahydrochloride (Sigma Chemical Co, St Louis, MO) in Tris-HCI buffer (50 mM, pH 7.6) containing 0.02% hydrogen peroxide. Control sections were Incubated with PBS in place of the monoclonal antibodies. 

Estimation of bromodeoxyuridine (BrdU) labeling index In a BrdU distribution study, 50% ethanol fixed tissue sections from the 131I treated mice were deparaffinized in xylene, washed in PBS, and rehydrated. Anti-BrdU mouse monoclonal antibody (MBL CO,Tokyo) was used. Staining was also performed by the ABC method described above. The positive nuclear staining rate per 1000 tumor cells was then found. The labeling index (L.I.) was estimated as an indicator of the tumor cell proliferative activity under the various light-microscopic fields.

Deoxy-D-glucose distribution 
Tumor bearing athymic nude mice treated with 131I were given an intraperitoneal injection of 10 micro Ci of [14C]deoxy-D-glucose ([14C]DG) (E I Du Pont de Nemours Inc, N. Billerica, Ma) and sacrificed 1 hour after the injection. For macro-autoradiographic studies, the xenografts and the selected tissue sections were applied to Fuji RX X-ray film and exposed for 8 weeks. Micro-autoradiography was performed according to the previously reported method. 10 Slides bearing HE stained sections were dipped twice in a 1 % solution of celloidin. After drying overnight at room temperature, the celloidin coated slides were smeared with a small amount of egg albumine : glycerine solution ( 1 : l), except over the section itself. In a completely light-proof dark room the slides were dipped for 1-2 seconds in the melted NTB2 nuclear emulsion (Eastman Kodak, USA) and then allowed to dry in a vertical position. The slides were exposed in black plastic boxes for 3 weeks. 

RESULTS 
Effect of TSH in vitro 
A kinetic study on the growth of the cultured cells is shown in Figure 2. From the curve, the population doubling time of TSH treated and untreated cells was estimated to be 1.7 days and 1.8 days, respectively. TSH had no effect on cell division or cellular morphology (data not shown). 
The [35S] activity of cells cultured for 2 hours was significantly increased by TSH, and this is shown in Table 1 . The mean uptake of cells treated with bTSH was about twice as great as that of the cells without bTSH ( p < 0.001 ). 

Histology of xenografts 
Subcutaneous transplantation of AC2 cells into athymic nude BALB/c mice led to the production of tumors that were clearly visible 3 to 4 weeks later. The solid tumors obtained from the 131I treated and untreated nude mice consisted of tissues showing signs of trabecular growth and small follicle formation. In some areas, the morphology of a typical differentiated adenocarcinoma with follicles was observed. In other parts of the tumors, sheets of either very large cells or small cells were observed. Most tumor cells were viable in the tissues from both the treated and untreated mice. The tumors from treated mice consisted of tissues with more follicle formation than those from untreated mice. 
Microscopically, the tissues with follicle formation showed a tendency to grow near the granulation and/or necrotic tissues. Histochemically, more intensely immunoreactive hTg was observed in the cytoplasms of the viable tumor cells from the treated mice than in those from untreated mice (Fig. 3-A, B). 

BrdU distribution study 
In a comparison of BrdU distribution in the tumor cells and histologic features of the tissue, BrdU L.I. were 19.6 +- 6.0% in the trabecularly growing tissues and 21.4 +- 9.5% in the follicle forming tissues, respectively (Fig. 4-A, B). 

Autoradiographic study 
The macro-autoradiogram of the resected specimens from the 131I treated mice is shown in Figure 5-A. Heterogeneous distribution patterns of silver grains within the tumor were seen. No silver grains could be seen within necrotic tissues in the tumor. A combination with light microscopy revealed that there were more dense grains (Fig. 5-B; arrow, arrowhead) in the follicle forming tumor tissues accompanied with granulation (Fig. 5-C) and/or necrotic tissue (Fig. 5-D), than in viable tumor cells of trabecularly growing tissues (Fig. 5-B; open circle, Fig. 5-E). 
Macro-autoradiograms of [14C]DG in formalin and in ethanol fixed sections from four other treated mice showed a similar pattern with a heterogeneous grain density in comparison with portions of the tumor mass (Fig. 6-A-D). In comparison with the H-E stained tissue sections, micro-autoradiograms showed that more dense grains were seen on the cells of the granulation tissue than on the tumor cells (Fig. 7-A). The grains were localized diffusely on viable tumor cells Independent of the histologic features of the tumor tissues (Fig. 7-B). 

DISCUSSION 
The tumors which originated in thyroid follicular epithelium represent various stages of tumor progression. The benign follicular adenomas arc considered to progress at low frequency to the malignant change in follicular carcinomas. In addition, thyroid differentiated cancer sometimes shows anaplastic transformation during its clinical course. Thus, thyroid neoplasms show a wide range of lesions varying from slowly progressive well-differentiated tumors to anaplastic highly malignant neoplasms. 11 
For practical reasons, [18F]FDG has found widespread use in PET and allowed us it to use as the standard radio-pharmaceutical for metabolic studies In the brain and various malignant tumors, 4 such as glioma, 5 hepatoma, 6 lymphoma 7 and lung cancer. 8 The accumulation mechanism of both FDG and DG as an analogue of glucose into cells is based on the intracellular trapping of DG-6-phosphate, 12 which is the product of the hexokinase reaction of deoxyglucose. Further metabolism of DG-6-phosphate is restricted to tissues devoid of glucose-6-phosphatase due to steric inhibition of enzyme activities. We carried out a correlative functional-histological study by means of autoradiographic imaging of thyroid cancer xenografts with [14C]DG. 
TSH significantly increased the total protein synthesis of the AC2 cells in the monolayer culture but was not accompanied with cell proliferation. These observation indicated that the established cell line, AC2, has the property of responding to TSH driving total protein synthesis in vitro. Recently, we reported that endogenous murine TSH appeared to have an effect on the cell differentiation. 13 In the present study, the tumors from treated mice consisted of tissues with more follicle formation than those from untreated mice. Histochemically, more intensely immunoreactive 27S, 19S, and 12S hTg were observed in the cytoplasms of the viable tumor cells from the treated mice than in those from untreated mice. These findings suggest that endogenous murine TSH may have an effect on cell differentiation l4,15 in AC2 cells in vivo. 
Watanabe and associates 3 studied rat brain tumors and reported a significant correlation between malignancy as measured by the bromodeoxyuridine labeling index and the local cerebral glucose utilization. They suggested that an increase in glucose utilization in these tumors might be due to the requirement for nucleic acid synthesis. Minn and associates 16 also observed a high correlation of FDG uptake with the amount of S-phase-cells. These reports support the hypothesis that the degree of FDG uptake may reflect proliferative activity and implicit aggressiveness of a tumor. 
However, we could not find a correlation between the amount of S-phase cells and the histologic features of the tumor. 
Kubota and associates 1 7 re ported that in an intratumoral [18F]FDG distribution study of a C3H/He mouse FM3A tumor model, [18F]FDG accumulation was relatively higher in macrophages and young granulation tissue than in the tumor cells. Our macroautoradiographic examinations also revealed that the areas with granulation tissues and/or the margin of necrotic tissues showed higher accumulation of [14C]DG than tumor cells of trabecularly and follicularly growing tissues of thyroid cancer AC2 xenografts. No accumulation of [14C]DG was observed in microscopically necrotic tissue. 
Our findings therefore indicate that the degree of deoxyglucose uptake may reflect not only tumor cell viability and proliferation but also inflammatory reaction and degenerative reaction to tumors. 
Further studies are under way to evaluate the correlation between the uptake of other metabolic substrates and malignant thyroid tumor grade, using our animal model which will be helpful in understanding the role of PET in thyroid cancer research. 

ACKNOWLEDGMENTS 
A part of this research was supported by a Grant-in-Aid from the Cancer Research Fund provided by the Ministry of Education (Group leader; Yukio Tateno) of Japan. 

REFERENCES 
l. Weber G, Morris HP. Comparative biochemistry of hepatomas. III. Carbohydrate enzymes in liver tumors of different growth rate. Cancer Res 23: 987-993, 1963. 
2. Burk D, Wood M, Hunter J. On the significance of glycolysis for cancer growth with special reference to Morris rat hepatomas. J Natl Cancer Inst 38: 839-863, 1967. 
3. Watanabe A, Tanaka R, Takeda N, Washiyama K. DNA synthesis, blood flow, and glucose utilization in experimental rat brain tumors. J Neurosurg 70: 86-91, 1989. 
4. Strauss LG. Conti PS. The applications of PET in clinical oncology. J Nucl Med 32: 623-648, 1991. 
5. Di Chiro G, DeLaPaz RL, Brooks RA, et al. Glucose utilization of cerebral gliomas measured by [18F]-fluorodeoxyglucose and positron emission tomograpy. Neu-rology 32: 1323-1329, 1982. 
6. Paul R, Ahonen A, Nordman E. Imaging of hepatoma with [18F]fluorodeoxyglucose. Lancet 1 : 50-51 , 1985. 
7. Paul R. Comparison of fluorine-18-2-fluorodeoxyglucose and gallium-67 citrate imaging for detection of lymphoma. J Nucl Med 28: 288-292, 1987. 
8. Nolop KB, Rhodes CG, Brudin LH, et al. Glucose utilization in vivo by human pulmonary neoplasms. Cancer 60: 2682-2689, 1987. 
9. Joensuu H, Ahonen A. Imaging of metastases of thyroid carcinoma with fluorine-18 fluorodeoxyglucose. J Nucl Med 28: 910-914, 1987. 
10. Kopriwa BM, Leblond CP. Improvements in the coating technique of radioautography. J Histochem Cytochem 10: 269-284, 1962. 
11 . Carcangiu ML, Steeper T, Zampi G, et al. Anaplastic carcinoma-A study of 70 cases. Am J Clin Pathol 83: 135-158, 1985. 
12. Reivich M, Kuhl D, Wolf A, et al. The 18F-2-fluolo-2-deoxy-D-glucose method for the measurement of local cerebral glucose utilization in man. Circ Res 44: 127-137, 1979. 
13. Tsuda T. In vitro and in vivo characterizations of established human follicular carcinoma cell line derived from thyroid cancer: A novel model for well-differentiated thyroid malignant tumor. Ann Nucl Med 6: 159-168, 1992. 
14. Valenta L, Kyncl F, Niederle B, Jirousek L. Soluble proteins in thyroid neoplasia. J Clin Endocrinol Metab 28: 442-450, 1968. 
15. Leche AR, Staub J, Kohler-Faden R, Muller-Brand J, et al. Thyroglobulin production by malignant thyroid tumors: An immunocytochemical and radioimmunoassay study. Cancer 57: 1145-1153, 1986. 
16. Minn H, Joensuu H, Ahonen A, Klemi P. Fluorodeoxyglucose imaging: a method to assess the proliferative activity of human cancer in vivo. Comparison with DNA flow cytometry in head and neck tumors. Cancer61 : 1776-1781 , 1988. 
17. Kubota R, Yamada S, Kubota K, et al. Intratumoral distribution of fluorine-18-fluorodeoxyglucose in vivo: High accumulation in macrophages and granulation tissues studied by microautoradiography. J Nucl Med 33: 1972-1980, 1992.