Uncoupling of 2-fluoro-2-deoxyglucose transport and ... - Nature

9 downloads 0 Views 483KB Size Report
1998 Stockton Press All rights reserved 0969-7128/98 $12.00 http://www.stockton-press.co.uk/gt. Uncoupling of 2-fluoro-2-deoxyglucose transport and.
Gene Therapy (1998) 5, 880–887  1998 Stockton Press All rights reserved 0969-7128/98 $12.00 http://www.stockton-press.co.uk/gt

Uncoupling of 2-fluoro-2-deoxyglucose transport and phosphorylation in rat hepatoma during gene therapy with HSV thymidine kinase U Haberkorn, ME Bellemann, L Gerlach, I Morr, H Trojan, G Brix, A Altmann, J Doll and G van Kaick Department of Oncological Diagnostics and Therapy, German Cancer Research Center, Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany

This animal study investigates the application of positron emission tomography (PET) with tracers of tumour metabolism for monitoring suicide gene therapy with herpes simplex virus thymidine kinase (HSVtk). After transplantation of HSVtk-expressing Morris hepatoma cells into ACI rats, dynamic PET measurements of 18F-labeled 2-fluoro-2deoxyglucose (FDG) uptake were performed in animals 2 days (n = 7) and 4 days (n = 5) after the onset of therapy with 100 mg ganciclovir (GCV)/kg body weight as well as after administration of sodium chloride (n = 8). The arterial FDG plasma concentration was measured dynamically in an extracorporeal loop and the rate constants for FDG transport (K1, k2) and FDG phosphorylation (k3) were calculated using a three-compartment model modified for heterogeneous tissues. Also, quantification using the metabolic rate of FDG turnover and the standardized uptake value (SUV) was done. Furthermore, the thymidine incor-

poration into the tumour DNA was determined after i.v. administration of 3H-thymidine. An uncoupling of FDG transport and phosphorylation was found with enhanced K1 and k2 values and a normal k3 after 2 days of GCV treatment. The increase in FDG transport normalized after 4 days whereas the phosphorylation rate k3 increased. Quantification using the metabolic rate or the SUV showed congruent but less sensitive results compared with the modeling approach. The thymidine incorporation into the DNA of the tumours declined to 10.5% of the controls after 4 days of GCV treatment. The data indicate that PET with 18FDG and 11C-thymidine may be applied for monitoring of gene therapy with the HSVtk/GCV suicide system. Increased transport rates are evidence of stress reactions early after therapy. The measurement of thymidine incorporation into the tumour DNA can be used as an indicator of therapy efficacy.

Keywords: PET; FDG; gene therapy; hepatoma; HSV thymidine kinase

Introduction Gene therapy of malignant tumours using suicide genes is performed in two steps. First, the tumour is infected with recombinant viruses to introduce the suicide gene into the cells. To obtain a sufficient level of enzyme activity in the tumour, multiple infections may be necessary. For the planning and the individualization of gene therapy, the enzyme activity induced in the tumour has to be estimated in order to express levels of enzyme sufficient for therapy before the application of the prodrug. This may be done with radiolabeled specific substrates of the suicide enzyme which has been transferred into the tumour.1–4 In the second step, the nontoxic prodrug is administered systemically. The measurement of therapy effects on the tumour metabolism after the onset of systemic prodrug application may be useful for the prediction of therapeutic outcome at an early stage of the treatment. Positron emission tomography (PET) using tracers of tumour metabolism has been applied for the evalu-

Correspondence: U Haberkorn Received 3 October 1997; revised 28 January 1998; accepted 10 February 1998

ation of treatment response during chemotherapy and radiotherapy in a variety of tumours,5–9 indicating that these tracers deliver useful parameters for the early assessment of therapeutic efficacy. However, PET has not yet been applied to monitoring tumours during suicide gene therapy. In a previous in vitro study with genetically modified Morris hepatoma cells bearing the herpes simplex virus thymidine kinase (HSVtk) gene, we were able to show that 2-fluoro-2-deoxyglucose (FDG) uptake is enhanced shortly after therapy with ganciclovir (GCV). This increase in FDG uptake was mainly due to an increase in the transport capacity of the tumour cells.10 The present study was undertaken to evaluate in vivo the effects of gene therapy in the same tumour model.

Results Growth inhibition To determine the therapeutic efficacy in vivo, animals bearing a control tumour (LXSN6) or the HSVtk-expressing tumour (LXSNtk8) were treated with GCV or with sodium chloride. The control tumours showed no growth inhibition during GCV treatment, whereas the HSVtk-

Monitoring gene therapy with PET U Haberkorn et al

expressing tumours revealed a decline in tumour volume which led to disappearance of the tumours in six cases and growth inhibition in two animals (Figure 1). To exclude a systemic influence on tumour growth in HSVtk-expressing tumours, a follow-up was done in these tumours after sodium chloride administration showing normal growth behaviour (Figure 1).

PET analysis of FDG uptake Figure 2 shows the normalized (SUV) PET-FDG images of animals after sodium chloride administration (Figure 2a), and 2 days (Figure 2b) and 4 days (Figure 2c) after the onset of GCV therapy. After 2 days GCV treatment animals showed an increased FDG uptake as compared with the controls, whereas in four of five animals a hypometabolic area was seen in the center of the tumour after 4 days GCV therapy. Using a three-compartment pharmacokinetic model modified for heterogeneous tissues, the rate constants for FDG transport and phosphorylation were calculated from the dynamic PET data of control animals (n = 8) which received sodium chloride, animals after 2 days treatment with GCV (n = 7) and animals after 4 days of GCV application (n = 5). The variation coefficients (mean ± s.d.) of parameter estimation were 25 ± 18%, 30 ± 18% and 11 ± 5% for rate constants K1, k2 and k3, respectively. We found a significant increase of K1 after 2 days of treatment (P ⬍ 0.001), whereas after 4 days of GCV therapy K1 was not different (P = 0.622) from the value in the control animals (Figure 3a and Table 1). Also, k2 was significantly enhanced after 2 days of therapy (P = 0.014), while no change was observed after 4 days of GCV treatment (P = 0.171; Figure 3b). In contrast, k3 was unchanged after 2 days therapy (P = 0.232) and enhanced after 4 days GCV administration (P = 0.006; Figure 3c). An HPLC analysis of tumour lysates was carried out for three animals of each group, revealing 91.9 ± 3.8% (control tumours), 91.8 ± 6.3% (tumours after 2 days GCV

881

a

b

c

Figure 2 FDG uptake (80–90 min p.i.) in animals with HSVtk-expressing tumours after sodium chloride administration (a) as well as 2 days (b) and 4 days (c) after the onset of treatment with 100 mg GCV/kg bw. The images are standardized to the injected dose and the body weight of the animals. R, right; L, left.

Figure 1 Tumour volume (normalized to the pretreatment value) during treatment with 100 mg GCV/kg bw in control hepatomas (LXSN6 + GCV, n = 6) and HSVtk-expressing hepatomas (LXSNtk8 + GCV, n = 8) as well as in HSVtk-expressing tumours during sodium chloride administration (LXSNtk8 + NaCl, n = 6). Mean values and s.d. are shown.

therapy) and 97.0 ± 4.2% (tumours after 4 days GCV therapy) of FDG-6-phosphate compared with the total radioactivity (FDG and FDG-6-phosphate). The difference between the groups was not statistically significant. The metabolic rate was also enhanced after 2 days, however, with no statistical significance (P = 0.054) and no change was seen after 4 days (P = 0.222; Figure 3d). PET quantification using the SUV (SUVPET) showed an enhancement of FDG uptake after 2 days (P = 0.009) and

Monitoring gene therapy with PET U Haberkorn et al

882

Figure 3 FDG uptake in untreated animals (n = 8) as well as after 2 days (n = 7) and 4 days (n = 5) of treatment with 100 mg GCV/kg bw. Changes in K1 (a), k2 (b), k3 (c), MRFDG (d), SUV measured by PET (e), and SUV measured by well counter (f).

Monitoring gene therapy with PET U Haberkorn et al

883 Table 1 Change of tumour parameters (kinetic rate constants (K1, k2, k3), metabolic rate (MRFDG), SUV (measured by PET and well counter), and thymidine incorporation (TdR)) after 2 (2 day GCV) and 4 (4 day GCV) days of therapy with GCV as compared with the controls Parameter K1 k2 k3 MRFDG SUVPET SUVWC TdR

2 Day GCV

4 Day GCV

↑ ↑ − (↑) ↑ (↑) −

− − ↑ − − − ↓

−, no change; ↓, decrease; ↑, increase; (↑), increase not statistically significant.

no change after 4 days of GCV treatment (P = 0.622; Figure 3e). The ex vivo data (SUVWC) revealed no statistically significant differences (P = 0.072 for 2 days and P = 0.524 for 4 days of treatment with GCV) as compared with the control group (Figure 3f). The SUVWC (measured by well counter 90 min post-injection (p.i.)) and the SUVPET (measured by PET 80–90 min p.i.) were correlated with r = 0.853 (significant at the 0.1% level; Figure 4).

Thymidine incorporation into DNA In order to assess the effects of gene therapy with the HSVtk/GCV system on DNA synthesis, we determined the thymidine incorporation into the DNA of untreated and treated tumours (Figure 5). After 2 days GCV therapy, no significant change of thymidine incorporation into the tumour DNA was observed (P = 0.189). However, we found a significant decrease of thymidine incorporation (P = 0.002) to 10.5% of the controls after 4 days of treatment with GCV.

Figure 4 Correlation of FDG uptake as measured by PET (SUVPET; 80– 90 min p.i.) and as measured by well counter (SUVWC; 90 min p.i.) in untreated animals (n = 8) as well as in rats after 2 (n = 7) and 4 days (n = 5) of treatment with 100 mg GCV/kg bw. r = regression coefficient.

Figure 5 Thymidine (TdR) incorporation into DNA in untreated Morris hepatomas (control; n = 8) as well as in tumours after 2 (n = 7) and 4 days (n = 5) of therapy with 100 mg GCV/kg bw.

Discussion In our study, we observed a growth inhibition in HSVtkexpressing Morris hepatoma, where a significant decrease in tumour volume occurred after 5 days of GCV treatment and no change in the LXSN6 control tumour (Figure 1). However, immune reactions to antigens of the foreign enzyme, which may be presented after intracellular processing on the tumour cell surface, may influence the tumour growth behaviour. To exclude these systemic effects on the growth behaviour of the HSVtk-expressing tumours, we also measured the tumour volume in the HSVtk-expressing hepatomas during daily administration of sodium chloride and found no evidence for growth delay. Monitoring of gene therapy using imaging procedures for the assessment of morphological changes has been performed with magnetic resonance imaging techniques in rats bearing C6 rat glioblastomas and also in patients with glioblastoma.11–13 In these studies, either marked tumour necrosis after interleukin-2 gene transfer or regression after induction of HSVtk expression and GCV application were observed. Maron et al13 found an initial response to GCV treatment in 90% of the animals and a complete regression in two-thirds of the treated rats. Also, tumour recurrence could be observed. However, in these studies the therapeutic efficacy was evaluated using changes in tumour volume with examination intervals of 2 months between the end of the treatment and the first follow-up examination.12 The measurement of metabolic changes after therapeutic intervention has proven to be superior to morphological procedures for the assessment of early therapy effects. In this respect, the FDG uptake has been demonstrated to be a useful parameter for the evaluation of glucose metabolism.5–9 Since the HSVtk/GCV system induces DNA chain termination,14 we also expect changes in thymidine incorporation into tumour cell DNA. 11C-thymidine has been applied to determine DNA synthesis in vivo.15–18 Therefore, a double tracer

Monitoring gene therapy with PET U Haberkorn et al

884

study was performed using 18FDG and 3H-thymidine. We found an increase of the FDG transport after 2 days of GCV treatment which normalized after 4 days and an increase of the phosphorylation rate after 4 days therapy (Figures 2 and 3). Quantification by metabolic rate and a procedure which normalizes the measured activity to the injected dose and the body weight (SUVPET) also revealed an increase of these parameters after 2 days of GCV administration. However, the difference failed to be statistically significantly different for the metabolic rate MRFDG (P = 0.054) and for the SUVWC measured ex vivo (P = 0.072). The SUV data are influenced by patient body composition, length of the uptake period, plasma glucose level and partial volume effects (underestimation of radioactivity concentration in small structures).19 To exclude partial volume effects on the quantification of the PET measurements, we also determined the radioactivity ex vivo in three tumour tissue samples per animal using a well counter (SUVWC). The SUVWC and the SUVPET were highly correlated (Figure 4). A deviation from the line of identity was seen with an underestimation of higher uptake values by PET. This underestimation was determined by a constant multiplicative factor. Furthermore, the ROI size was not significantly different in the control group and the treated animals. Therefore, partial volume effects can be neglected as cause for systemic bias in our study. Another confounding factor may be loss of body weight during GCV therapy. We observed a weight reduction of 5–10% after 4 days and up to 30% after 8 days of GCV therapy. This shows that GCV is also a substrate for the host thymidine kinase leading to side-effects when higher doses of the drug are used. However, we performed our PET studies 2 and 4 days after the onset of GCV treatment where the small changes in weight loss are not likely to influence the SUV quantification of the PET data. Enhancement of FDG uptake in tumour cells early after therapeutic intervention was found in vitro as well as in vivo7,10,20–24 and has also been observed in nonmalignant tissues as an early reaction to cellular stress with a redistribution of glucose transport proteins to the plasma membrane as the underlying molecular mechanism.25–27 Our findings of an increased inward transport 2 days after therapy fit well into previous in vitro findings of enhanced 3-O-methylglucose uptake in the same tumour cells after therapy with GCV.10 Also, the changes in the transport constants and the metabolic rate or the SUV were congruent. The normalization of the transport rates after 4 days may be due to the therapeutic effects of DNA synthesis inhibition and the consequent death of tumour cells. This explanation is evidenced by the fact that a significant decrease of thymidine incorporation into the tumour DNA occurred only after 4 days of treatment. Furthermore, a central necrosis was seen after this period in four of five tumours (Figure 2). In the in vitro study,10 we found an increased 3-Omethylglucose uptake together with a decreased FDG phosphorylation capacity after 2 days of treatment with GCV.10 This is at variance with our in vivo data showing an increased inward transport with a normal phosphorylation rate after 2 days and an increase of phosphorylation together with normal inward transport after 4 days. This may be caused by a different drug concentration obtained in vivo compared with the ideal in vitro situation and by

the fact that in vivo tumours also contain nontumour cells such as endothelial cells, macrophages and lymphocytes,28,29 which lack the suicide gene transfected to the Morris hepatoma cells. These nontumour cells contribute to the signal obtained in metabolic studies as shown by MR spectroscopy.30 Another explanation for the different behaviour of the phosphorylation rate may be the small size (n = 5) of the group after 4 days of treatment which may lead to statistical problems. However, even after elimination of a tumour showing an extremely high phosphorylation rate, the difference between this group and the control group is still significant (P = 0.016). Finally, the changes in FDG phosphorylation may be caused by redistribution phenomena in a similar way as to the glucose transporter. The hexokinases I and II are found in a soluble form and a form which is bound to the outer membrane of the mitochondria.31 Since the mitochondrial-bound hexokinase is much less sensitive to product inhibition by glucose-6-phosphate than the soluble form and has privileged access to the mitochondrial-produced ATP, this form confers a metabolic advantage, which is needed in tissues with high dependence on glucose for the cellular metabolism as in the brain as well as in tumours.31,32 This form of hexokinase favours high rates of glucose-6-phosphate synthesis. Since glucose-6-phosphate is a precursor of glycolysis as well as for the biosynthesis of nucleic acids and phospholipids, enhanced synthesis of these compounds may be essential for rapid cell division, membrane biosynthesis, and repair of cell damage. In our study, the thymidine incorporation into tumour DNA decreased after 4 days of GCV treatment (Figure 5), demonstrating the efficacy of the therapy. However, in a previous in vitro study, which used 3H-thymidine, a decrease of thymidine uptake in the nucleic acid fraction was found together with an increase in the cytoplasmic fraction, indicating that quantification of 11C-thymidine PET studies based on the assessment of total intracellular radioactivity may fail to detect early therapy effects on the DNA synthesis.4,22 The interpretation of 11C-thymidine measurements requires information about the size of the different metabolite fractions, which may be delivered by HPLC analyses of the arterial blood plasma.17,18 These may be used in a modelling analysis for the determination of 11C-thymidine incorporation into the DNA. However, other nucleoside analogues with a more favourable metabolism (less metabolites) may be better candidates for the measurement of tumour proliferation.33 In the present study, we used a stable transfected cell line with 100% of the tumour cells expressing the HSVtk gene. Such a high transduction efficiency is not achieved in vivo. However, in our previous in vitro study with the same cell line, we found a dependence of the changes in FDG transport and phosphorylation on the amount of HSVtk-expressing cells.10 Furthermore, in vivo transduction of tumours with retroviral vectors results in suicide gene transfer not only to tumour cells but also to the endothelial cells of the tumour vasculature. This may result in changes of tumour perfusion and substrate delivery. Therefore, the use of a stable HSVtk-expressing cell line allowed the delineation of therapy-induced changes in FDG transport and phosphorylation from treatment-induced ischemia as a consequence of impaired tumour perfusion.

Monitoring gene therapy with PET U Haberkorn et al

In conclusion, an uncoupling of glucose transport and phosphorylation during GCV treatment of HSVtkexpressing Morris hepatoma was found. We observed an enhanced FDG transport with normal phosphorylation after 2 days of treatment as an early event after GCV administration which is interpreted as nonspecific stress reaction to the cell damaging agent. This increase in transport normalized after 4 days whereas the phosphorylation increased. The changes in the phosphorylation rate in vivo need further investigation. Quantification using the metabolic rate or the SUV showed congruent but less sensitive results compared with the modelling approach. The measurement of thymidine incorporation into the tumour DNA can be used as an indicator of therapy efficacy.

Materials and methods The experiments were performed in compliance with German laws relating to the conduct of animal experimentation. All data were statistically analyzed using the Mann–Whitney rank sum test (SigmaStat, version 2.0; Jandel Scientific, Erkrath, Germany).

Growth inhibition Two cell lines, LXSN6 (a control line with the empty vector; in vitro doubling time, 17.8 ± 1.0 h) and LXSNtk8 (a GCV-sensitive line bearing the HSVtk gene; in vitro doubling time, 16.8 ± 0.7 h), derived from the Morris hepatoma cell line MH3924A were used.4 Tumour cells (4 × 106) were transplanted s.c. into the right thigh of young male adult ACI rats. A therapy study was performed 3–4 weeks later. Animals with the LXSNtk8 tumour were treated by daily i.p. injections of 100 mg GCV (Cymeven; Syntex, Aachen, Germany)/kg body weight (bw) (n = 8) or sodium chloride (n = 6), whereas animals with the LXSN6 tumour were treated with 100 mg GCV/kg bw (n = 6). The tumour size was measured every 2 days in two orthogonal dimensions and the tumour volume was calculated assuming an ellipsoid volume. Positron emission tomography Since no effect of GCV therapy was seen on the LXSN6 tumour and also no change in FDG uptake could be observed in a previous in vitro study,10 we concentrated on the LXSNtk8 tumour in the PET experiments. The measurements were performed 21–28 days after inoculation when the tumours had reached a diameter of more than 20 mm. Two (n = 7) or four days (n = 5) before the PET study, the animals (weighing 260–393 g) were treated by daily i.p. administration of 100 mg GCV/kg bw. A control group (n = 8) received sodium chloride for 2 days. After premedication with 50 mg ketamine (Ketanest; Parke Davis, Berlin, Germany)/kg bw and 0.5 mg N-(3′-dimethylamino-propyl)-3-propionyl-phenothiazine-phosphate (Combelen; Bayer, Leverkusen, Germany)/kg bw, the animals were kept in an inhalation narcosis with enflurane (volume-%, 0.4), nitrous oxide (flow, 1000 ml/min), and oxygen (flow, 500 ml/min) during the PET examination. The production of 18FDG was done according to the method described by Oberdorfer et al.34 The radiochemical purity was determined by high performance liquid chromatography (HPLC) with values above 98%. The PET studies were performed with a PC2048–7WB whole-

body scanner (Scanditronix AB, Uppsala, Sweden). After administration of 23.7–48.9 MBq 18FDG in a permanent catheter positioned in the jugular vein, the FDG uptake was measured dynamically over 90 min (acquisition frames, 12 ×5 s, 6 × 10 s, 6 × 30 s, 5 × 60 s, 6 × 300 s, and 5 × 600 s). The 18FDG concentration in the arterial blood plasma of each animal was determined with a closed extracorporeal loop placed between the carotid artery and the jugular vein. The total 18F activity concentration in the extracorporeal loop was measured continuously over 90 min using a bismuth germanate fluid monitor (Scanditronix AB) with a temporal resolution of 500 ms. At 5, 10, 30, 60 and 90 min or after 18FDG injection, blood samples (volume approximately 100 ␮l) were taken from the extracorporeal loop and the 18F radioactivity concentrations in the whole blood and plasma were determined with an automated NaJ(Tl) well counter (Cobra II; Canberra Packard, Meriden, CT, USA) referenced to a 511keV 68Ga/68Ge calibration standard (Amersham-Buchler, Braunschweig, Germany). The measured arterial 18F time-activity profiles were calibrated using the whole blood samples, corrected for the FDG exchange between the plasma and the red blood cells, and standardized to the injected radioactivity and the body weight of the animals. Before each PET study, the plasma glucose level (Cp) of the animals was determined using a blood glucose sensor electrode (MediSense, Waltham, MA, USA), revealing plasma glucose levels of 105 ± 21 mg% for the control group, 130 ± 57mg% for the animals after 2 days GCV therapy, and 151 ± 54 mg% for the animals after 4 days GCV treatment. The differences between the groups were not statistically significant. Simultaneously with the 18FDG administration, 3.7 MBq (methyl-3H)thymidine (Amersham-Buchler, specific activity, 185 GBq/mmol; radioactivity concentration, 37 MBq/ml; radiochemical purity, 97.5%) were injected. Before the PET emission scanning, a 30-min blank scan and a 15-min transmission scan were performed using a 68Ge rod source. The emission sinograms were corrected for scattered radiation35 and attenuation36 as described elsewhere. PET images with slice thickness of 8 mm (cross plane) and 11 mm (direct planes) were generated by use of a high-resolution maximum-likelihood reconstruction algorithm37 on a DEC 3000/400 AXP workstation (Digital Equipment Corporation, Maynard, MA, USA). To minimize partial volume effects, the inplane resolution of the images was improved by the implementation of the line-spread function of the PET scanner into the process of iterative image reconstruction which resulted in an almost constant spatial resolution of 3.9 mm over the examined field-of-view.37 The image matrix was 256 × 256 elements with a pixel size of 1 × 1 mm. For quantitative evaluation, regions-of-interest (ROIs) were defined in the hottest area of the tumours with region sizes exceeding 114 pixels in all cases. There was no statistically significant difference of the ROI size in the different treatment groups. The FDG uptake was expressed as standardized uptake value (SUV) according to SUVPET =

C , D/m

with C the tissue activity concentration (in Becquerels per

885

Monitoring gene therapy with PET U Haberkorn et al

886

gram), D the injected dose (in Becquerels), and m the body weight (in grams). Immediately after the PET examination, the animals were killed and the tumours were removed and weighed. Three tumour samples per animal were taken for the measurement of the 18FDG uptake in an automated NaJ(Tl) well counter (Canberra Packard) and expressed as SUV. The three tumour uptake values were averaged to yield a well-counter standardized uptake value (SUVWC) at 90 min after tracer administration. The same tissue samples (weighing 109 ± 6 mg) were used for DNA extraction. Furthermore, an HPLC analysis of tumour lysates was done after perchloric acid extraction for three animals of each group as described elsewhere.10

Pharmacokinetic analysis A three-compartment pharmacokinetic model, originally developed for the brain38 and modified for heterogeneous tissues,39,40 was employed for the assessment of FDG transport and phosphorylation in the tumours. The rate constants that describe the compartmental fluxes include K1 (in millilitres per minute per gramme) for forward transport of FDG, k2 (in reciprocal minutes) for backward transport of FDG, and k3 (in reciprocal minutes) for phosphorylation of FDG in the tumour tissue. The model configuration included three additional fit parameters: an input function time delay parameter Td (in seconds) to correct for the time delay of 18FDG activity arrival in the extracorporeal detector loop as well as the fit parameters ␣ (dimensionless) and ␤ (in reciprocal minutes) which define the time dependence of k2 and k3 according to the pharmacokinetic model configuration for FDG uptake in heterogeneous tissues.39,40 The model configuration was realized within the ADAPT II program environment41 on a DEC 3000/400 AXP workstation (Digital Equipment Corporation) and included nonlinear least-squares fitting for parameter estimation. Since the lumped constant to account for the differences between glucose and FDG for transport and phosphorylation38 is unknown for Morris hepatoma, we restricted our analysis to the FDG metabolism. Hence, the metabolic rate of FDG utilization (MRFDG) (in micromoles per minute per gramme) was computed as described38 MRFDG =

CP K1 · k 3 · , LC k2 + k3

with Cp the plasma glucose concentration (in micromoles per millilitre) and LC the lumped constant with LC = 1.

DNA extraction Tumour pieces were minced and digested overnight at 55°C in 700 ␮l lysis buffer (100 mm NaCl, 10 mm Tris at pH 8, 25 mm EDTA at pH 8, 0.5% sodium dodecyl sulfate, and 0.1 mg proteinase K/ml). After addition of 3 ml phenol, the lysates were washed for 30 min and rotated at 1500 g in a Heraeus Minifuge for 10 min. The supernatant was washed with 3 ml of a phenol–chloroform-isoamylalcohol (CIA; ratio, 24:1) mixture (ratio phenol/CIA, 1:1) for 15 min. After centrifugation (10 min at 1500 g), the supernatant was washed with CIA for 15 min and rotated (10 min at 1500 g) to remove the remaining phenol. To the supernatant, 0.1 vol 3 m sodium acetate and 1 vol isopropanol were added to precipitate the DNA. After centrifugation, the pellet was washed with 70% ethanol and air dried. The pellet was resuspended in water and the

DNA concentration was determined photometrically at 260 nm. Thereafter, the incorporated radioactivity was measured using Pico-Fluor-15 (Canberra Packard) in a scintillation counter (LSC TRICARB 2500TR, Canberra Packard) and normalized to the DNA content (in Becquerels per milligramme of DNA).

Acknowledgements We thank Ellen Bender, Werner Konowalczyk, Heike Marx, Armin Runz, Anne Theobald, Ulrike Wagner, Klaus Weber and Wolfgang Weber for their help in performing this study. We also thank Anthony Shields (Karmanos Cancer Institute, Detroit, MI, USA) and Ronald Blasberg (Memorial Sloan-Kettering Cancer Center, New York, NY, USA) for fruitful discussions.

References 1 Tjuvajev JG et al. Imaging the expression of transfected genes in vivo. Cancer Res 1995; 55: 6126–6132. 2 Tjuvajev JG et al. Noninvasive imaging of herpes simplex virus thymidine kinase gene transfer and expression: a potential method for monitoring clinical gene therapy. Cancer Res 1996; 56: 4087–4095. 3 Haberkorn U et al. Monitoring of gene therapy with cytosine deaminase: in vitro studies using 3H-5-fluorocytosine. J Nucl Med 1996; 37: 87–94. 4 Haberkorn U et al. Monitoring gene therapy with herpes simplex virus thymidine kinase in hepatoma cells: uptake of specific substrates. J Nucl Med 1997; 38: 287–294. 5 Haberkorn U et al. Fluorodeoxyglucose imaging of advanced head and neck cancer after chemotherapy. J Nucl Med 1993; 34: 12–17. 6 Haberkorn U et al. PET studies of fluorodeoxyglucose metabolism in patients with recurrent colorectal tumours receiving radiotherapy. J Nucl Med 1991; 32: 1485–1490. 7 Rozenthal JM, Levine RL, Nickles RJ, Dobkin JA. Glucose uptake by gliomas after treatment. Arch Neurol 1989; 46: 1302– 1307. 8 Minn H, Paul R, Ahonen A. Evaluation of treatment response to radiotherapy in head and neck cancer with fluorine-18 fluorodeoxyglucose. J Nucl Med 1988; 29: 1521–1525. 9 Wahl RL et al. Metabolic monitoring of breast cancer chemohormonotherapy using positron emission tomography: initial evaluation. J Clin Oncol 1993; 11: 2101–2111. 10 Haberkorn U et al. Multitracer studies during gene therapy of hepatoma cells with herpes simplex virus thymidine kinase and ganciclovir. J Nucl Med 1997; 38: 1048–1054. 11 Sobol RE et al. Interleukin-2 gene therapy in a patient with glioblastoma. Gene Therapy 1995; 2: 164–167. 12 Izquierdo M et al. Long-term rat survival after malignant brain tumour regression by retroviral gene therapy. Gene Therapy 1995; 2: 66–69. 13 Maron A et al. Gene therapy of rat C6 glioma using adenovirusmediated transfer of the herpes simplex virus thymidine kinase gene: long-term follow up by magnetic resonance imaging. Gene Therapy 1996; 3: 315–322. 14 Keller PM et al. Enzymatic phosphorylation of acyclic nucleoside analogs and correlations with antiherpetic activities. Biochem Pharmacol 1981; 30: 3071–3077. 15 Christman D, Crawford EJ, Friedkin M, Wolf AP. Detection of DNA synthesis in intact organisms with positron-emitting (methyl-11C)thymidine. Proc Natl Acad Sci USA 1972; 69: 988– 992. 16 Poupeye EM et al. Evaluation of (11C)thymidine for measurement of cell proliferation in fast dividing tissues. Nucl Med Biol 1993; 20: 359–362. 17 Shields AF, Coonrod DV, Quackenbush RC, Crowley JJ. Cellular sources of thymidine nucleotides: studies for PET. J Nucl Med 1987; 28: 1435–1440.

Monitoring gene therapy with PET U Haberkorn et al

18 Shields AF et al. Utilization of labeled thymidine in DNA synthesis: studies for PET. J Nucl Med 1990; 31: 337–342. 19 Keyes JW. SUV: standardized uptake or silly useless value? J Nucl Med 1995; 36: 1836–1839. 20 Haberkorn U et al. Fluorodeoxyglucose uptake in vitro: aspects of method and effects of treatment with gemcitabine. J Nucl Med 1994; 35: 1842–1850. 21 Minn H et al. Uptake of 2-fluoro-2-deoxy-D-(U-14C)-glucose during chemotherapy in murine Lewis lung tumour. Nucl Med Biol 1992; 19: 55–63. 22 Haberkorn U et al. PET 2-fluoro-2-deoxyglucose uptake in rat prostate adenocarcinoma during chemotherapy with gemcitabine. J Nucl Med 1997; 38: 1215–1221. 23 Slosman DO, Pugin J. Lack of correlation between tritiated deoxyglucose, thallium-201 and technetium-99m-MIBI cell incorporation under various cell stresses. Eur J Nucl Med 1994; 35: 120–126. 24 Haberkorn U et al. Metabolic and transcriptional changes in osteosarcoma cells treated with chemotherapeutic drugs. Nucl Med Biol 1994; 21: 835–845. 25 Widnell CC et al. Cellular stress induces a redistribution of the glucose transporter. FASEB J 1990; 4: 1634–1637. 26 Clancy BM, Czech MP. Hexose transport stimulation and membrane redistribution of glucose transporter isoforms in response to cholera toxin, dibutyryl cyclic AMP, and insulin in 3T3 adipocytes. J Biol Chem 1990; 265: 12434–12443. 27 Wertheimer E, Sasson S, Cerasi E, Ben-Neriah Y. The ubiquitous glucose transporter GLUT-1 belongs to the glucose-regulated protein family of stress-inducible proteins. Proc Natl Acad Sci USA 1991; 88: 2525–2529. 28 Steele RJC, Brown M, Eremin O. Characterization of macrophages infiltrating human mammary carcinomas. Br J Cancer 1985; 51: 135–138. 29 Whitford P, Mallon ES, George WD, Campbell AM. Flow cytometric analysis of tumour infiltrating lymphocytes in breast cancer. Br J Cancer 1990; 62: 971–975. 30 Smith TAD et al. The effect of intra-tumour heterogeneity on the

31 32

33

34

35

36 37

38

39

40

41

distribution of phosphorus-containing metabolites within human breast tumours: an in vitro study using 31P NMR spectroscopy. NMR Biomed 1991; 4: 262–267. Pedersen PL. Tumour mitochondria and the bioenergetics of cancer cells. Prog Exp Tumor Res 1978; 22: 190–274. Arora KK, Pedersen PL. Functional significance of mitochondrial bound hexokinase in tumour cell metabolism. J Biol Chem 1988; 263: 17422–17428. Shields AF, Grierson JR, Kozawa SM, Zheng M. Development of labeled thymidine analogs for imaging tumour proliferation. Nucl Med Biol 1996; 23: 17–22. Oberdorfer F, Hull WE, Traving BC, Maier-Borst W. Synthesis and purification of 2-deoxy-2-18fluoro-d-glucose and 2-deoxy-218 fluoro-d-mannose: characterization of products by 1Hand 19F-NMR spectroscopy. Int J Rad Appl Instrum A 1986; 37: 695–701. Hoverath H et al. Scatter correction in the transaxial slices of a whole-body positron emission tomograph. Phys Med Biol 1993; 38: 717–728. ¨ Ostertag H, Kubler WK, Doll J, Lorenz WJ. Measured attenuation correction methods. Eur J Nucl Med 1989; 15: 722–726. Brix G et al. Use of scanner characteristics in iterative image reconstruction for high-resolution positron emission tomography studies of small animals. Eur J Nucl Med 1997; 24: 779–786. Huang SC et al. Noninvasive determination of local cerebral metabolic rate of glucose in man. Am J Physiol 1980; 238: E69E82. Schmidt K, Mies G, Sokoloff L. Model of kinetic behavior of deoxyglucose in heterogeneous tissues in brain: a reinterpretation of the significance of parameters fitted to homogeneous tissue models. J Cereb Blood Flow Metab 1991; 11: 10–24. Lucignani G et al. Measurement of regional cerebral glucose utilization with fluorine-18-FDG and PET in heterogeneous tissues: theoretical considerations and practical procedure. J Nucl Med 1993; 34: 360–369. D’Argenio DZ, Schumitzky A. ADAPT II User’s Guide. Biomedical Simulation Resource, University of Southern California: Los Angeles, 1992.

887