Takamitsu Oku

We report a series of studies that assess the feasibility and sensitivity of imaging of herpes virus type one thymidine kinase (HSV1-tk) gene transfer and expression with [124I]-5-iodo-2'-fluoro-1-beta-D-arabinofuranosyluracil ([124I]-FIAU) and positron emission tomography (PET) and the ability of [124I]-FIAU-PET imaging to discriminate different levels of HSV1-tk gene expression. Studies were performed in rats bearing multiple s.c. tumors derived from W256 rat carcinoma and RG2 rat glioma cells. In the first set, we tested the sensitivity of [124I]-FIAU-PET imaging to detect low levels of HSV1-tk gene expression after retroviral-mediated gene transfer. HSV1-tk gene transduction of one of preestablished wild-type W256 tumor in each animal was accomplished by direct intratumoral injection of retroviral vector-producer cells (W256-->W256TK* tumors). Tumors produced from W256 and W256TK+ cells served as the negative and positive control in each animal. Highly specific images of [124I]-FIAU-derived radioactivity were obtained in W256TK* tumors (that were transduced in vivo) and in W256TK+ tumors but not in nontransduced wild-type W256 tumors. The level of "specific" incorporated radioactivity in transduced portions of both W256TK* and W256TK+ tumors was relatively constant between 4 and 50 h. In the second set, we tested whether [124I]-FIAU and PET imaging can measure and discriminate between different levels of HSV1-tk gene expression. Multiple s.c. tumors were produced from wild-type RG2 cells and stably transduced RG2TK cell lines that express different levels of HSV1-tk. A highly significant relationship between the level of [124I]-FIAU accumulation [% injected dose/g and incorporation constant (Ki)] and an independent measure of HSV1-tk expression (sensitivity of the transduced tumor cells to ganciclovir, IC50) was demonstrated, and the slope of this relationship was defined as a sensitivity index. We have demonstrated for the first time that highly specific noninvasive images of HSV1-tk expression in experimental animal tumors can be obtained using radiolabeled 2'-fluoro-nucleoside [124I]-FIAU and a clinical PET system. The ability to image the location (distribution) of gene expression and the level of expression over time provides new and useful information for monitoring clinical gene therapy protocols in the future.

Noninvasive imaging of herpes simplex virus type 1 thymidine kinase (HSV1-tk) gene expression is possible with a clinical gamma camera and by single-photon emission tomography (SPECT) using 131I-labeled 2'-fluoro-2'-deoxy-1-beta-D-arabinofuranosyl-5-iodo-uracil (FIAU). Studies were performed in rats bearing s.c. tumors. Tumors were produced by injection of wild-type RG2 glioma or W256 mammary carcinoma cells into one flank and RG2TK+ glioma or W256TK+ mammary carcinoma cells (that had been transduced in vitro with the HSV1-tk gene) into the opposite flank. In some animals, HSV1-tk gene transduction of the pre-established wild-type tumors was accomplished in vivo by direct intratumoral injection of retroviral vector-producer cells. Imaging studies were performed 2 weeks after tumor transduction to allow time for production and spread of the retroviruses through the tumor and for sufficient growth and increase in size of the tumors to facilitate imaging. The gamma camera and SPECT images revealed highly specific localization of [131I]FIAU-derived radioactivity to areas of HSV1-tk gene expression at 24, 36, and 48 h after i.v. administration of 1.6-2.8 mCi of [131I]FIAU. Comparative analysis of quantitative autoradiographic images obtained from the same tumors confirmed that the high levels of [131I]FIAU-derived radioactivity (> 1% dose) were localized to areas of HSV1-tk gene expression demonstrated by immunohistochemical staining for HSV1-tk protein. In contrast, significantly lower levels of [131I]FIAU-derived radioactivity (< 0.01%) were observed in the surrounding nontransduced tumor tissue, contralateral wild-type tumors, and other tissues that showed no immunohistochemical staining for the HSV1-tk protein. The magnitude of FIAU accumulation in RG2TK+, W256TK+, and wild-type tumors corresponded to the in vitro ganciclovir sensitivity of the cell lines used to produce these tumors, which indicates that the magnitude of FIAU accumulation reflects the level of HSV1-tk gene expression. We suggest that "clinically relevant" levels of HSV1-tk gene expression in transfected tissue can be imaged with [131I]FIAU and a gamma camera or SPECT, and that a significant improvement in imaging sensitivity and resolution is expected with [124I]FIAU and PET.

The goal of this study was to determine the magnitude of "facilitated" amino acid transport across tumor and brain capillaries and to evaluate whether amino acid transporter expression is "upregulated" in tumor vessels compared to capillaries in contralateral brain tissue. Aminocyclopentane carboxylic acid (ACPC), a non-metabolized [14C]-labeled amino acid, and a reference molecule for passive vascular permeability, [67Ga]-gallium-diethylenetriaminepentaacetic acid (Ga-DTPA), were used in these studies. Two experimental rat gliomas were studied (C6 and RG2). Brain tissue was rapidly processed for double label quantitative autoradiography 10 minutes after intravenous injection of ACPC and Ga-DTPA. Parametric images of blood-to-brain transport (K1ACPC and K1Ga-DTPA, microL/min/g) produced from the autoradiograms and the histology were obtained from the same tissue section. These three images were registered in an image array processor; regions of interest in tumor and contralateral brain were defined on morphologic criteria (histology) and were transferred to the autoradiographic images to obtain mean values. The facilitated component of ACPC transport (deltaK1ACPC) was calculated from the K1ACPC and K1Ga-DTPA data, and paired comparisons between tumor and contralateral brain were performed. ACPC flux, K1ACPC, across normal brain capillaries (22.6 +/- 8.1 microL/g/min) was >200-fold greater than that of Ga-DTPA (0.09 +/- 0.04 microL/g/min), and this difference was largely (approximately 90%) due to facilitated ACPC transport. Substantially higher K1ACPC values compared to corresponding K1DTPA values were also measured in C6 and RG2 gliomas. The deltaK1ACPC values for C6 glioma were more than twice that of contralateral brain cortex. K1ACPC and deltaK1ACPC values for RG2 gliomas was not significantly higher than that of contralateral cortex, although a approximately 2-fold difference in facilitated transport is obtained after normalization for differences in capillary surface area between RG2 tumors and contralateral cortex. K1ACPC, deltaK1ACPC, and K DTPA were directly related to tumor cell density, were higher in regions of "impending" necrosis, and the tumor/contralateral brain ACPC radio-activity ratios (0 to 10 minutes) were very similar to that obtained with 0 to 60 minutes experiments. These results indicate that facilitated transport of ACPC is upregulated across C6 and RG2 glioma capillaries, and that tumors can induce upregulation of amino acid transporter expression in their supporting vasculature. They also suggest that early imaging (e.g., 0 to 20 minutes) with radiolabeled amino acids in a clinical setting may be optimal for defining brain tumors.

Vascular endothelial growth factor (VEGF), also known as vascular permeability factor, has been investigated as a potent mediator of brain tumor angiogenesis and tumor growth. We evaluated the effect of VEGF expression on the pathophysiology of tumor growth in the brain. Human SK-MEL-2 melanoma cells, with minimal VEGF expression, were stably transfected with either sense or antisense mouse VEGF cDNA and used to produce intracerebral xenografts. Vascular permeability, blood volume, blood flow, and tumor fluorodeoxyglucose metabolism were assessed using tissue sampling and quantitative autoradiography. Tumor proliferation was assessed by measuring bromodeoxyuridine labeling indices. Tumor vascular density and morphological status of the blood-brain barrier were evaluated by immunohistochemistry. SK-MEL-2 cells transfected with sense VEGF (V+) expressed large amounts of mouse and human VEGF protein; V+ cells formed well-vascularized, rapidly growing tumors with minimal tumor necrosis. V+ tumors had substantial and significant increases in blood volume, blood flow, vascular permeability, and fluorodeoxyglucose metabolism compared to wild-type and/or V- (antisense VEGF) tumors. VEGF antisense transfected V- expressed no detectable VEGF protein and formed minimally vascularized tumors. V- tumors had a very low initial growth rate with central necrosis; blood volume, blood flow, vascular permeability, and glucose metabolism levels were low compared to wild-type and V+ tumors. A substantial inhibition of intracerebral tumor growth, as well as a decrease in tumor vascularity, blood flow, and vascular permeability may be achieved by down-regulation of endogenous VEGF expression in tumor tissue. VEGF-targeted antiangiogenic gene therapy could be an effective component of a combined strategy to treat VEGF-producing brain tumors.