Radiol Oncol 1996; .30: 104-7.
Insulin dependency of F-18-fluorodeoxyglucose accumulation in breast carcinoma cells compared to Tl-201 uptake
Heike Wolf, Winfried Brenner, Volker Stein, Stephan Tinnemeyer, Karl Heinz Bohuslavizki, Ulrich Teichert, Malte Clausen, Eberhard Henze
Clinic of Nuclear Medicine, Christian-Albrechts-University of Kiel, Kiel, Germany
The effect of euglycemic hyperinsulinism on '"F-fluorodeoxyglucose (FDG) uptake in cultures of breast cancer was determined, in comparison to 2I"TI. Measurements of bolh tracers were performed on 168 cell culture tubes, with incubation inteivals ranging from 1 to 240 min. Linear accumulation of FDG over time was observed bolh with aiul without insulin. A significant increase from 3.52 ± O.74 % to 5.10 ± 0.32 % over 240 min was attained after adding insulin. In contrast, 2"'TI revealed only a slightly significant increase after insulin. By extrapolating these results to FDG PET tumor imaging, a markedly improved tumor targeting might be obtained by providing a stale of euglycemic hyperinsulinism, i.e. replacing the commonly used single FDG injection by a continuous FDG/glucose/insulin infusion. An optimum imaging period of 150 min after starting the infusion can be derived froin our dala, considering the decay of ' "F
Key words: breast neoplasms; tumor cells, cultured; euglycemic hyperinsulinism; deoxyglucose, lsF-fluo-rodeoxyghicose; thallium radiosiotopes, 201T1
Introduction
Increased glycolysis is an important characteristic of cancer cells.1- 2 Positron emission tomography (PET) with [l!iF]-2-fluoro-2-deoxy-D-glucose (FDG) is used as a suitable indicator of the glycolytic activity of tumors. FDG is rapidly transported into the tumor cells and phosphorylated by hexokinase to FDG-6-phosphate, but is not further metabolized, • meanwhile the physiologic substrate glucose enters the glycolytic pathway. Increased FDG uptake imaged by PET has been reported in many types of human tumors, e.g., head and neck cancer,5 lung,'' colon,7 ' liver9- 10 and breast cancer."• '2201 Tl SPECT imaging might work as well as FDG PET in the detection of viable tumor tissue based on the relatively enhanced tumor blood supply. In neoplastic cell cultures the 20lTl uptake increases in
Correspondence to: HeikeWolf, Ph. D. Christian-Albrechts-University of Kiel, Clinic of Nuclear Medicine, Arnold-Heller-Str. 9, D-24105 Kiel, Germany
UDC: 618. l 9-006.6-092.4:616.l53.455.01
conjunction with the cell's metabolic activity, thereby confirming that it might also reflect tumor growth rather than just tumor perfusion.11
In the in-vivo and in-vitro studies so far available5- 12 '4-''J controversial effects of plasma insulin and/or glucose levels on FDG tumor targeting have been reported. Agreement consists in most studies that elevated cold glucose levels seem to compete with FDG, thus hampering FDG accumulation.
The purpose of this study was to determine the effect of euglycemic hyperinsulinism on the cellular uptake of FDG in cultures of breast cancer cells, and to compare it with the uptake of 20IT| under identical conditions, since insulin effects on 20IT| tumor accumulation are also hardly known.
Materials and methods
Radiopharmaceuticals
FDG (BTZ Beschleunigei- und Tomographiezentrum Hamburg, Germany) was obtained with a radiochemical purity higher than 98 %. Impurities like ethanol, acetonitrile and acetone were less than
Insulin dependency of F-lS-fl.uoivdeoxytflucose accumulation
10,5
0.06 mg/ml and the glucose concentration was less than 0.6 mg/ml.
A commercially available 201TI thallous chloride was used (Mallinckrodt, Hennef, Germany) as a reference tracer.
Cell cultivation
The human breast carcinoma celi line (MCF 7, dkfz -Tumorbank, Heidelberg, Germany) chosen for examination was maintained in a medium containing 5 mmol/1 glucose, supplemented with I 0% fetal calf serum (Boehringer Mannheim, Mannheim, Germany), 0.5 % L-glutamine 20 mmol/1 (Biochrom, Berlin, Germany) and I ml gentamycin 0.1 mg/ml (Merck, Darmstadt, Germany). The celi line grew well in vitro as a onolayer and had a doubling time of approximately 40 h. A Fuchs-Rosenthal counting chamber was used for celi counting and the trypan blue exclusion method for viability determination was conducted using an inverted Leitz
icroscope.
Tiie total number of cells ranged from 7.9 to 9.2 million cells per tube with a median of 8.5 million. Celi viability was higher than 90 %.
Uptake measurements
.30 MBq FDG and 3 MBq 201 TI were added to 250 ml medium, thus maintaining a continuous supply during the whole incubation period. For each tube, 3 ml medium was used, and the exponentially growing cells were incubated with TOG and 10'Tl at 37 °C. Influx was stopped at different incubation intervals ranging from I minute to 4 hours, by removing the medium. .Subsequently, cells were washed rapidly 3 times with 4 °C saline solution for a total of 15 seconds and harvested with irypsin-EDTA (Sigma Chemie, Deisenhofen. Germany).
Tiie radioactivity was measured with a gamma well counter in a definite geometry. Total activity per culutre tube was attained by easuring the tube before removing the medium. The cellular uptake was determined alter the washing phase. TOG measurements were performed first and samples of medium and cells were stored for two days to determine the 201 TI uptake, after the complete decay of lsF. Ali results were corrected for physical decay.
To assure adequate measurement statistics, six tubes per exposure period were used, totaling 84 culture tubes in ali. The experiment was repeated exactly under the same conditions, but with an additional administration of insulin, in a quantily of 0.5 IU to the medium, i.e. 2 mIU/ml, providing a complete receptor saturation.
ptake results are presented as the percentage of the activity accumulated within the cells, related to the activity added in each tube normalized to 1 million cells, and expressed as mean ± 1 s.d. For statistical comparison, Student's t-test for unpaired data was used.
Results
As shown in Figure 1, linear accumulation of FDG over time was observed in nearly ali the exposures, up to a maximum of 240 min. With the longest exposure period of 240 min an uptake of 3.52 ± 0.74 % was measured under baseline conditions. A further significant increase of up to 5.10 ± 0.32 % was attained after adding insulin, with p 0.02 already beginning after an incubation lime of 20 in.
incubation time [min]
Figure 1. Cellular tiplake of "FDG expressed as % of medium activily per I million breasl carcinoma cells with and withoul insulin (0.5 mlU/ml medium) at various incubation lime intervals. Resulls are given as mean of 6 culture lubes ± 1 SD. Additional insulin: filled squares; without insulin: open circles.
In contrast, 201 Tl accumulation occured at a high rate during the first I 0 to 20 min and then reached a plateau. Adding insulin did not induce such marked effects, as illustrated in Figure 2.
By lumping ali the tubes, with and without insulin, together, the groups differ with 0.33 ± 0.06 % and 0.25 ± 0.05 %, respectively. Significant differences with p < 0.02 might be obtained in the initial phase, such as at I 0, 30, 45 or 90 min, while after a 150 min exposure time, both curves converge on each other.
106
WolfH etat.
, 0,5
Sä
p 0,4
0,3
0,2 • I
11 with insulin O withoutinsulin
0 100 200
incubation time [min]
Figure 2. Cellular uptake of ""Tl expressed as % of medium activity per 1 million breast carcinoma cells with and without insulin (0,5 mIU/ml medium) at various incubation time intervals. Results are given as mean of 6 culture tubes ± 1 SD. Additional insulin: filled squares; without insulin: open circles.
Discussion
Tumor targeting and quantitative evaluation of tumor tissue viability with FDG and PET have shown encouraging results, with high tumor/background ratios. Attempts have been made to further increase FDG accumulation in tumor tissue for even better scintigraphic tumor delineation, mainly by using additional glucose administration on induce a state of hyperglycemic hyperinsulinism.14-'7- 2" In agreement with in-vitro studies, examing the same issue, FDG accumulation mainly decreased with elevated plasma or media glucose concentrations12 l5' 21 except in brain tumor imaging.17 This was interpreted as a competing effect between cold and labeled glucose leading to the conclusion that FDG tumor imaging ought to be performed under conditions of food abstinence. '4- b-18
The purpose and the results of this study have to be seen in that context. Because of the obviously hampering FDG accumulation at increased plasma glucose levels, the approach was different to all previously performed studies. The experimental design simulated a stale of euglycemic hyperinsu-linism in conjunction with a relatively constant supply of FDG, maintained lor up to 240 min. Providing this environment to breast cancer cells a signili-cant increase of FDG uptake in tumor tissue was observecl as seen in Figure 1. This result seems to support the above cited hypothesis of competition between cold and labeled glucose.
The effect of the same study design on 201T1 cell accumulation was considerably less, and completely concomitant to recent results.19 Particularly during the initial exposure period of up to 150 min, however, insulin increased the 20'Tl uptake by 50 %. Providecl this also holds true for in vivo scintigraphy, an increase of the tumor/background ratio by a factor of 1.5 would stili represent a great progress in scintigraphic tumor targeting.
Based on this data it seems thus justified to propose a modified application protocol for both 20lTl and FDG tumor imaging, suggesting euglycemic hyperinsulinism and maintaining constant tracer supply for a longer period of time. In the case of 20'TI, an approximately 60 min infusion might be suggested, which causes no additional problem, since insulin clamping requires a continuous infusion anyway. In the case of FDG, the short tracer half life has to be considered As seen in Figure 1 constant tracer supply results in linear accumulation, if correctecl for physical decay. Considering 1SF decay an accumulation type of curve is obtained, as shown in Figure 3, with maximum tracer concentration and, thus, a postulated optimum lor imaging, at approximately 150 min after starting the FDG/insulin infusion. The dose of insulin should be moderate, i.e. 20-30 IU/h, so as not to induce significant hypoglycemia ancl/or clinical symptoms, but to raise the plasma insulin level.
Clinical trials are needed to validate this posiu-lated approach. The outcome of such trials is difficult to predict, since the effect of euglycemic hyper-
• with insulin ° without insulin
100 time [min] 200
Figure 3. Cellular uptake of "FDG with (filled squares) and without (open circles) insulin considering physical decay of "F, as calculated from the regression curves of Figure l.
(
I
0
Imidin depe//de//ry o/' F-/(S'-/l//o™deoA"3J;'///ro.ve «cc»/»»/«//o»
107
insulinism on bolh the tumor and the surrounding tissue in vivo is uncertain because of the various distribution of the, at least five different, glucose transporting molecules, and the unknown magnitude as to whether these glucose transporters are involved in the facilitated transmembraneous glycolytic flux.12, U~IS
Conclusion
The accumulation of FDG and, to a lesser extent, that of 2"'T1, in breast carcinoma cells can be significantly increased by insulin and euglycemia. The potential improvemeni of PET and SPECT tumor targeting ought to justify further clinical testing.
Acknowledgements
The generous supporl in delivering of FDG by Dr. B. Nebeling (Cyclotron, University of Hamburg) is highly appreciated.
References
1.	Warburg O. On the origin of cancer cells. .SV/rnre 19.56; 12.3: 309-14.
2.	Weber G. Enzymology of cancer cells. N ling! J /Wed 1977; 296: 486-93. "
3.	Gallagher BM, Fowler JS. Gutterson NI, MacGregor RR, Wan CN, Wolf AP. Metabolic trapping as a principle of radiopharmaceutical design: some factors responsible for the biodistribution of ['"F] 2-deoxy-2-lluoro-D-gluco.se..! Nucl Med 1978; 19: 1154-6l.
4.	Som P. Atkins HL. Bandoypadhyay D,etal. A fluorinated glucose analog, 2-['sFj-lluoro-2-deoxy-D-gluco-se: nontoxic tracer for rapid tumor detection. ./ N//W /Wed 1980; 21: 670-5.
5.	Haberkorn U, Strauss LG. Reisser Ch, el al. Glucose uptake. perfusion, and celi proliferation in head and iieck tumors: relation of positron emission tomography to flow cytometry. ./ Nud Med 1991; .32: 1548.55.
6.	Nolop KB. Rhodes CG, Brudiii LH. et al. Glucose utilization in vivo by human pulmonary neoplasm. C«//cer 1987; 60: 2682-9.
7.	Strauss LG, Clorius JH. Schlag P. et al. Recurrence of colorectal tumors: PET evaluation. RV/d/ology 1989; 170: 329-32.
8.	Haberkorn U, Strauss LG, Dimitrakopoulou A. et al.
PET studies of fluorodeoxyglucose metabolism in pa-
tients with recurrent colorectal tumors receiving radiotherapy. ./ N//W Med 1991; 32: 1485-90.
9. Yonekura Y, Benua RS, Brili AB, et al Increased accumulation of 2-deoxy-2-"F-fiuoro-D-glucose in liver metastases from colon carcinoma. ./ Nurl Med 1982; 32: 1133-7.
10.	Oya N, Nagata Y. lshigaki T, et al. Evaluation of experiment liver tumors using fluorine-l8-2-fluoro-2-de-oxy-D-glucose PET../ Nwl Med 1993; 34: 2124-9.
11.	Kubota K, Matsuzawa T. Amemiya A. et al. lmaging of breasl cancer with ['sF]fluorodeoxyglucose and positron emission tomography. ./. Com/;»/ J\.v.v/.vi Tomogr 1989; 13: 1097-8.
12.	Wahl RL. Cody RL, Hutchins GD, Mudgett EE. Primary and metastatic breast carcinoma: initial clinical evaluation with PET with the radiolabeled glucose analogue 2-l'*FJ-fluoro-2-deoxy-D-glucose. R«d/o/ogy 1991; 179: 765-70.
13.	Maublant JC, Zhang Z. Rapp M. Ollier M. Michelot J. Veyre A. In vitro uptake of Technetium-99m-Tchoro-xime in carcinoma cell lines and normal cells: comparison with Technetium-99rn-Sestamibi andThallium-201. ./ Nur/ /Wed 1993; 34: 1949-52.
14.	Langen K-J, Braun U. Rota Kops E. et al.The inlluence of plasma glucose levels on fluorine-l 8-11uorodeoxy-glucose uptake in bronchial carcinomas. J Nur/ Med 1993; 34: 355-9.
15.	Lindholm P Minn II. Leskinen-Kallio S. Bergman J, Ruotsalainen U. Joensuu 11. Influence of the blood glucose concenlration on FDG uptake in cancer a PET study. J Nad /Wed 1993; 34: 1-6.
16.	Minn II. Leskinen-Kallio S. Lindholm P. et al. ['"FJ Fluorodeoxyglucose uptake in tumors: kinetic vs. steady-state methods with reference to plasma insulin. ./ Comp/// Av.v/.w 7ö»/og/. 1993: 17: 115-23.
17.	lshizu K. Nishizawa S, Yonekura Y, et al. Effects of hyperglycemia on FDG-PET in patients with brain tumors. 7 Nurl Med 1994; 35: 43P.
18.	Wahl RL. Henry ChA, Ethier StP Serum glucose: effects on tumor and normal tissue accumulation of 2-['"F)-nuoro-2-deoxy-D-glucose in rodents with mammary carcinoma. /i«d/ology 1992; 183: 643-7.
19.	Slosrnan DO. Pugin J. Lick of correlation between tritialed deoxyglucose, tliallium-201 and teclrnetium-99m-MIBI cell incorporation under various celi stresses. J N»W Med 1994; 35: 120-6.
20.	Okazurni S, lsono K. Enomoto K. et al. Evaluation of liver tumors using lluorine-18-Fluorodeoxyglucose PET: characterization of tumor and assessment of effect of treatment. ./ Nurl Med 1992: .3.3: 333-9.
21.	Amos II. Muslimer TA Asdourian II. Regulation of glucose carriers in chick fibroblasts. ./ .Supra/Holer«/«/. S/mr/ure 1977; 7: 499-513.