HF-1 cells and induced minimal apoptosis (�5%). In HeLa cells, ON01910 was found to be more cytotoxic than CP incomparable doses (survival fraction of cells incubated with 1.0 mM CP: 0.3 vs. 0.1 for 1.0 mM ONC treated cells).Cytotoxicity was further enhanced by the addition of CP and irradiation. The survival fraction of various treatmentregimens are summarized below.

Conclusions: At the clinically relevant 2 Gy dose, ON01910 significantly enhances radiosensitivity and augments the combinedcytotoxicity of cisplatin and RT. This is the first report of a small molecule inhibitor of Plk1 that exhibits selective cytotoxicityand radiosensitivity towards tumor cells and could translate towards a beneficial therapeutic ratio for chemoradiation therapyof cervical carcinomas.

Surviving Fraction of Different Treatment Regimens

1046 Schedule-Dependent Drug Effects of Oral 5-iodo-2-pyrimidinone-2�-Deoxyribose as an In VivoRadiosensitizer in U251 Human Glioblastoma Xenografts

Y. Seo,1 T. Yan,1 J.E. Schupp,1 T. Radivoyevitch,2 T.J. Kinsella1

1Radiation Oncology, Case Western Reserve University/University Hospitals of Cleveland, Cleveland, OH, 2Epidemiologyand biostatistics, Case Western Reserve University/University Hospitals of Cleveland, Cleveland, OH

Purpose/Objective: 5-iodo-2-pyrimidinone-2�-deoxyribose (IPdR) is an oral prodrug of 5-iodo-2�-deoxyuridine (IUdR), aneffective in vitro/in vivo radiosensitizer. We have previously shown that a once daily oral administration of IPdR results insignificant radiosensitization and improves the therapeutic index compared to a continuous infusion of IUdR using a mousemodel with multiple different human tumor xenografts. IPdR can be rapidly converted to IUdR by a hepatic aldehyde oxidase.However, we found that the enzymatic conversion of IPdR to IUdR can be transiently saturated and this may affect the IPdRdosing schedule used for optimal tumor radiosensitization. The purpose of this study is to assess pharmacological profiles, tumorradiosensitization, and systemic toxicity using alternative IPdR dosing schedules.

Materials/Methods: Three different IPdR treatment schedules (three times a day, TID; every other day, QOD; every third day,Q3D), compared to a once daily (QD) schedule, were analyzed using athymic nude mice with human glioblastoma (U251)subcutaneous xenografts. Plasma concentration of IPdR and IUdR, hepatic aldehyde oxidase activities, and % IUdR-DNAincorporation in tumor and normal proliferating tissues were measured as pharmacological parameters. The extent of tumorradiosensitization was assessed by a tumor regrowth delay following fractionated radiotherapy. Systemic toxicity was assessedby weight loss.

Results: The TID schedule with the same total daily doses as the QD schedule (250, 500, 1000 mg/kg/day) showed lessreduction of aldehyde oxidase activity and improved the efficiency of IPdR conversion to IUdR. As a result, the %IUdR-DNA incorporation was higher using the TID schedule in the tumor xenografts as well as in normal small intestineand bone marrow. However, the enhanced IUdR-DNA incorporation was greater in the normal proliferating tissues thanin tumor. Using a fixed dose (500 mg/kg) per administration, the QOD and Q3D schedules showed greater IPdRconversion than the QD schedule, related to a greater recovery of hepatic aldehyde oxidase activity prior to the next drugdosing. The QOD schedule provided a comparable level of IUdR-DNA incorporation in the tumor while the incorporationin the normal proliferating tissues was significantly reduced, leading to an improved therapeutic index. In the tumorregrowth assay using 2 Gy/d 4 days, all the IPdR schedules showed significant increases of regrowth delays comparedto the control without IPdR (QOD 29.4 days; QD 29.7 days; TID 34.7 days; Radiotherapy alone 15.7 days). The extentof radiosensitization of each schedule was quantitated as a radiosensitizer enhancement ratio; QOD 2.17, QD 2.20, TID2.56. There was no statistically significant difference between the QOD and the QD schedules. The TID schedule showedsignificantly enhanced radiosensitization compared to the QD schedule (p �.05). The weight loss also differed among theIPdR treatment schedules (TID � QD � QOD � Q3D).

Conclusions: The alternative dosing schedules improved the efficiency of enzymatic activation of IPdR to IUdR and changedthe drug effect as measured by % IUdR-DNA incorporation and tumor regrowth delay. These alternative schedules will beconsidered for future clinical testing if the pharmacokinetic analyses of our initial QD 28 day dosing schedule in humansshow significant hepatic aldehyde oxidase saturation.

1047 mTOR Inhibitors Are Safe and Effective Radiosensitizers in Glioblastoma Multiforme Pre-ClinicalModels

E. Shinohara, C. Cao, K. Niermann, Y. Mu, F. Zeng, D. Hallahan, B. Lu

Vanderbilt University, Nashville, TN

Purpose/Objective: mTOR inhibitors, such as Rapamycin have been used safely in humans as an antibiotic and immunosup-pressant. Because of mTORs role in cell survival in response to mitogenic stimuli, and the safety of mTOR inhibitors, mTORprovides an attractive target for cancer treatment. This study investigates the role of mTOR inhibition in radiosensitization,specifically in GBM, and by what mechanism this sensitization may be occurring.

S172 I. J. Radiation Oncology ● Biology ● Physics Volume 63, Number 2, Supplement, 2005

Materials/Methods: Two mTOR inhibitors, rapamycin and Rad001 were used in this study. Western blot analysis wasused to determine the phosphorylation status of mTOR in both HUVEC (human endothelial) and GL261 (Glioma) celllines. Caspase 3 levels and clonogenic assays were used to determine levels of apoptosis. Matrigel assays were used todetermine mTOR and radiations effects on vasculature in vitro. GL261 tumor-bearing mice were treated with 3 Gy twicea week for two weeks. Rad001 was given orally (2.5 mg/kg), one hour prior to radiotherapy. These mice were also usedin Doppler studies and in histological studies. Doppler studies used ultrasound to determine blood flow and tumorvascularity. Histological sections were stained for vWF and were used to quantify vasculature. Tumor growth delaywas determined in mice treated biweekly with either vehicle, Rapamycin (2 mg/kg), 3 Gy, or Rapamycin 1 hour prior to3 Gy.

Results: Clonogenic assays showed that both rapamycin (100 nM) and Rad001 (5 nM) were effective radiosensitizers inHUVEC cells but had little effect on irradiated GL261 cells. HUVEC showed an increase in phosphorylated mTOR(p-mTOR) levels 15 minutes after treatment with 3 Gy as shown by western blot analysis. A similar increase in p-mTORwas not seen in irradiated GL261 cells. S6 is known to be phosphorylated by mTOR in response to mitogenic stimuli.HUVEC cells treated with 3 Gy showed increased phosphorylation of S6 in response to 3 Gy. However, this effect couldbe inhibited by pretreating HUVEC with Rad001 or LY294002, a PI3K inhibitor. In vitro matrigel assays showed thatthere was a slight decreased formation of microtubules in HUVEC treated with 100 nM of rapamycin or 3 Gy but therewas a significant decrease in HUVEC treated with rapamycin and 3 Gy. Tumor vasculature was studied in vivo usingDoppler imaging of mice implanted with GL261. Mice treated with Rad001 or radiation alone showed a slight decreasein blood supply compared with controls, however, mice receiving combination therapy were the only treatment group toexpress significant decreases in blood supply, with a 32% reduction in vascularity (P�0.05) and 36% reduction in flow(P�0.03). Tumors were then removed and sectioned. Vessels were stained for using antibody for vWF and these werequantified microscopically, with tumors treated with combined therapy having the greatest reduction in blood vessels.While radiation and Rad001 alone showed a decrease in the number of vessels, the greatest decrease was in the combinedtreatment group. Tumor volume studies showed that rapamycin and radiation given alone could induce a short growthdelay but combined therapy induced an increase in growth delay of 18 days and 15 days, respectively, when comparedwith rapamycin and radiation alone.

Conclusions: This study suggests that irradiation activates mTOR cell survival signaling in glioma vascular endothelium viaa PI3K/Akt dependent pathway. mTOR inhibition resulted in marked increase in radiosensitization of HUVEC with littlesensitization of GL261 cells. Rapamycin’s effects on vascular endothelium were further confirmed by both in vitro and in vivostudies. mTOR inhibitors are an effective radiosensitizer in the GL261 glioma model.

1048 WITHDRAWN

1049 Systemic Administration of GMCSF Combined with Localized Tumor Hyperthermia Induces SystemicAntitumor Immunity

A. Mukhopadhyay, J. Mendecki, A. Alfieri, L. Liu, S. Kalnicki, C. Guha

Radiation Oncology, Montefiore Medical Center - Albert Einstein College of Medicine, Bronx, NY

Purpose/Objective: Localized tumor hyperthermia (LTH) treatment of prostate adenocarcinoma can potentially serve as asource of tumor antigen, where dying apoptotic/necrotic cells release tumor peptides slowly over time. In addition, heat shockproteins released from LTH treated cells can chaperone antigenic peptides to antigen-presenting cells (APCs), such as dendriticcells (DCs). We attempted to discern whether sequential application of LTH and systemic administration of the DC proliferatingagent granulocyte macrophage colony stimulating factor (GMCSF) could activate antitumor immune response in a syngeneicmurine model of prostate cancer.

Materials/Methods: RM1 (105) cells were injected in the dorsum of the foot of C57Bl/6 male mice. Palpable tumors (approx.100mg) were subjected to LTH (43.7°C for 1 hr) 2, separated by 5 days. One day after first LTH treatment animals receivedsystemic murine GM-CSF by recombinant adenovirus (8109 particles). Control animals received either AdenoLacZ orAdenoGFP. RM1 tumors were followed by volume measurement. Antitumor immune responses were measured by cytokinerelease assays, ELISPOT and LDH release to measure T helper and CTL response of splenocytes. In addition, DC and T cellinfiltration of RM1 tumors were analyzed by FACS scan with antimouse fluoroprobes to CD4, CD8 and CD11c.

Results: LTH alone (n�10) was responsible for significant tumor growth delays when compared to untreated controls.Systemic injection of adenoviruses (LacZ and GFP) did not alter tumor control of HT-treated animals (p�0.67). Whereas,AdGM-CSF (n�13) significantly retarded tumor growth (p�0.001) in HT-treated RM-1 tumors. Systemic administration ofAdGMCSF induced T-helper cell (IFN� ELISPOTS) and CTL activity (LDH release 19% vs 5%). There was increasedinfiltration of CD4� and CD8� T cells and CD11c� DCs in HT�GMCSF-treated tumors, compared to controls.

Conclusions: The combination of LTH and AdGMCSF inoculation induced a systemic tumor-specific immune response andenhanced local control of HT-treated RM-1 tumors. Immunomodulation of LTH by DC-stimulating cytokines could be a usefuladjuvant in the treatment of local and systemic recurrence of prostate cancer.

1050 Low Levels of Quinazoline-Based EGFR Inhibitors Enhance Growth of Glioma Cells

C. Chang, M. Yuan, B. Li, H.G. Shu

Radiation Oncology, University of Pennsylvania, Philadelphia, PA

Purpose/Objective: Glioblastoma multiformes are highly aggressive primary brain tumors. Epidermal growth factor receptor(EGFR) is commonly altered in these tumors and this is believed to be important in its pathogenesis. Thus, inhibition of EGFR

S173Proceedings of the 47th Annual ASTRO Meeting