S62 I. J. Radiation Oncology d Biology d Physics Volume 69, Number 3, Supplement, 2007

increased tumor oxygenation (pO2). This study investigated the impact of IFP, pO2, and CT drug sequencing on breast cancerresponse to neoadjuvant CT and on long-term local control.

Materials/Methods: From January 2001 to June 2004, we conducted a phase II randomized study that enrolled 62 women withprimary invasive breast cancer, clinically $3 cm, without distant metastasis. Patients were randomized to one of two neoadjuvantsequential CT regimens: either four cycles of doxorubicin (60 mg/m2) every 2 weeks followed by nine cycles of weekly paclitaxel(80 mg/m2) or, paclitaxel first followed by doxorubicin using the same doses and schedule. Tumor IFP (wick-in-needle technique),pO2 (Eppendorf), clinical tumor volume (product of the two greatest dimensions palpated on exam), and MRI tumor volume weremeasured at three time points: at baseline, after completing the first drug, and after completing the second drug. Patients underwentsurgical resection and received radiation and additional chemotherapy when clinically indicated. The patients were followed clin-ically and radiographically.

Results: Lower IFP after the first course of CT was associated with improved clinical tumor response (p = 0.004) to the first CTagent in a linear regression model (n = 49) that included baseline tumor size (p = 0.058) and estrogen receptor status (p = 0.0166) aspredictive variables. Receiving doxorubicin as the first agent and paclitaxel as the second agent was associated with improved clin-ical tumor response to neoadjuvant CT (p = 0.027) in a linear regression model (n = 43) that included baseline tumor size (p =0.0001) as a predictive variable. A similar benefit was observed using MRI measurements (p = 0.013). Tumor oxygenation didnot impact clinical tumor response to neoadjuvant CT. With a median follow up of 4.0 years, eight of the subjects experiencedlocal relapse. In a Cox regression model, shorter time to local relapse occurred in patients with larger tumors (.5 cm) at presen-tation (HR 6.43, p = 0.049) and those with measurements that indicated improved tumor oxygenation after the first CT agent (HR1.09, p = 0.036). The measured IFP did not impact on local control for patients who subsequently underwent surgical resectionwhich was followed, in most cases, by adjuvant radiation. A survival benefit was associated with receiving doxorubicin as firstdrug, followed by paclitaxel (p = 0.0135).

Conclusions: In this prospective study, improved clinical tumor response to the first neoadjuvant chemotherapeutic agent was as-sociated with decreased tumor IFP. Clinical tumor response to neoadjuvant CT was superior when doxorubicin was given beforepaclitaxel. Decreasing elevated IFP and optimizing CT sequencing may improve the delivery of antineoplastic agents and subse-quent tumor response to CT.

Author Disclosure: K.E. Hoffman, None; R. Abi-Raad, None; M. Ancukiewicz, None; E. Yeh, None; P. Ryan, None; L. Schapira,None; J. Younger, None; B.L. Smith, None; I. Kuter, None; A.G. Taghian, None.

110 In Vivo Imaging of Apoptosis by Annexin V Scintigraphy: Predictive Value for Treatment Outcome

M. Verheij, M. Kartachova, R. Haas, N. van Zandwijk, S. Burgers, H. van Tinteren, F. Hoebers, R. Valdes Olmos

The Netherlands Cancer Institute, Amsterdam CX, The Netherlands

Purpose/Objective(s): Apoptosis has been recognized as an attractive target for anti-cancer therapy. Indeed, many therapeuticstrategies have been designed to enhance apoptosis and increase the tumor response to radiation and/or chemotherapy. 99 mTc-Annexin V scintigtaphy (TAVS) is a non-invasive imaging technique that allows the in vivo visualization and quantification ofapoptosis. Purpose of this study was to correlate early treatment-induced apoptosis as measured by TAVS, with outcome afterradiotherapy, chemotherapy or the combination.

Materials/Methods: Sixty-one patients (NHL n = 27; HNSCC n = 16; NSCLC n = 16; SCLC n = 1; sarcoma n = 1) underwenta TAVS before and within 24–48 hr after the start of treatment. Therapy consisted of low dose (2 � 2 Gy) radiotherapy (n = 27),cisplatin-based concurrent chemoradiotherapy (n = 16) or cisplatin-based chemotherapy (n = 18). The difference between the TAVtumor uptake before and after start of treatment (DU), calculated as maximum count per pixel and expressed as percentage to base-line value, was correlated to response according to RECIST criteria.

Results: A significant correlation (linear regression analysis; p \ 0.001) was found between DU and treatment outcome. All pa-tients with notably increased TAV tumor uptake showed complete or partial response. Less prominently increased or decreaseduptake correlated with stable or progressive disease.

Conclusions: A significant correlation was established between tumor TAVS uptake and treatment outcome in a variety of tumortypes. The predictive value of this test might help to design novel (apoptosis-modulating) strategies and evaluate treatment effectsat an early stage.

Author Disclosure: M. Verheij, None; M. Kartachova, None; R. Haas, None; N. van Zandwijk, None; S. Burgers, None; H. vanTinteren, None; F. Hoebers, None; R. Valdes Olmos, None.

111 Potential use of Early Cytokine Changes as Surrogate Markers in Adults Versus Children Following

a Terrorist Event

J. P. Williams, C. Johnston, E. Hernady, C. Reed, J. N. Finkelstein

University of Rochester Medical Center, Rochester, NY

Purpose/Objective(s): Following a radiological terrorism event or accident, there is a need for accurate biodosimetry for triagepurposes. As part of our investigation into the consequences of such an event, we are looking at cytokine responses to low dosesof radiation for use as biomarkers following both localized (whole lung) and total body irradiation, correlating changes in expres-sion with both early and late pulmonary responses. In addition, due to our special interest in the pediatric population, we have begunto examine the respective roles of the same cytokines in pups at different stages of postnatal development. We contend that earlyinduction of specific pulmonary inflammatory markers detected in the adult post-radiation will not appear during postnatal lungdevelopment and hypothesize, therefore, that adult cytokine profiles will not be useful as biomarkers of exposure in children.

Materials/Methods: Groups of 10 C57Bl/6 mice, 6–8 weeks of age, received either whole lung or total body external irradiation atdoses of 0–10 Gy. Animals were sacrificed at time points between 1 hour and 15 months. Alterations in cytokine mRNA and pro-tein expression were assessed in tissue (lung and liver), serum and lavage (BAL). In addition, C57Bl/6 mice, 4–56 days of age,

Proceedings of the 49th Annual ASTRO Meeting S63

received total body irradiation of 5 Gy. These animals were examined 1 and 6 hours post-radiation. Analysis included cytokine andchemokine mRNA abundance, TUNEL staining and histological analysis.

Results: In the lung tissue, robust, dose response changes were seen in IL1ß, IL1RII and CXCR2 mRNA expression at 1 and 6hours post-irradiation, returning to baseline by 24 hours. These were concurrent with increases in protein levels of KC, a neutrophilchemotactic factor and a ligand for CXCR2, and interleukin (IL)-6 measured in the serum. Total cell counts indicated that there isa rapid dose-responsive decline in total cell numbers. In contrast, in the tissue, there was an abrupt increase in the numbers of in-filtrating neutrophils over the 1–12 hour period, followed by an equally rapid decline. In contrast, changes in IL1ß and IL1RII werenot detected until 14 days of age in the neonates, which coincided with the end of postnatal lung growth. However TUNEL stainingwas detected at all ages examined with peak staining occurring 6 hours suggesting that although inflammatory mediators did notrespond during development after radiation exposure, pulmonary injury did occur.

Conclusions: KC (the murine equivalent to IL8/GROa) and IL6 may be candidates for biodosimeters. Interestingly, log transfor-mation of the values for these two cytokines demonstrated a clear threshold for doses at or above 2 Gy. Their role in the progressionof tissue response is unclear at present, nonetheless, their early change in expression may act as a surrogate for adult lung response.However, a lack of cytokine induction in early life could subject tissues to a greater injury resulting in long-term alterations instructure and function. Different identification end points and drug treatments therefore may be required to treat children followinga radiological terrorist event. Investigations are currently underway to further examine cytokine expression at later time points andcorrelate these early changes with late tissue effects.

Supported by funding from NIAID/NCI (U19 AI067733), P01 HL-071659, P30 ES-01247 and EPA Star PM Center R-827354.

Author Disclosure: J.P. Williams, None; C. Johnston, None; E. Hernady, None; C. Reed, None; J.N. Finkelstein, None.

112 Cells Lacking the PolQ Polymerase are more Sensitive to Ionizing Irradiation In Vitro and In Vivo

J. P. Goff, M. Epperly, D. Shields, T. Smith, M. Seki, J. Wittschieben, R. Wood, J. S. Greenberger

University of Pittsburgh Cancer Institute, Pittsburgh, PA

Purpose/Objective(s): PolQ polymerase plays an important role in base excision repair. It contains a helicase domain and a poly-merase domain. To determine the role of PolQ in repair of DNA damage following irradiation, we isolated bone marrow stromalcells from PolQ +/+ or PolQ �/� mice by establishing long term bone marrow cultures (LTBMCs).

Materials/Methods: The bone marrow was flushed out of the tibias and femurs into T-25 flasks. The cultures were examined ona weekly basis with the number of cobblestone islands counted, nonadherent cell production and Day 7 or Day 14 colony formingcells counted. After the LTBMCs no longer produced cobblestone islands or nonadherent cells, bone marrow stromal cell lineswere isolated. To determine the irradiation sensitivity of the cell lines, cells from PolQ +/+, PolQ �/� clone 1, PolQ �/� clone3, or PolQ �/� clone 5 irradiation survival curves were performed and the data analyzed by linear quadratic or single-hit, multi-target models. To determine if PolQ �/� mice were more sensitive to irradiation, PolQ +/+, ± or �/� mice were irradiated to0.7 Gy and blood collected and the percent of erythrocyte micronuclei determined forty hours later.

Results: Cobblestone islands, areas of stem cell activity, were increased beginning at week 10 to week 23 in the LTBMCs from thePolQ�/� mice. However, there was no increase in number of nonadherent cells or colony forming cell production in the culturesbetween the PolQ +/+ or�/� cultures. Increase in cobblestone islands usually results in increased nonadherent cell production andcolony production due to increased stem cell activity. The role of PolQ in stem cell activity is being pursued. The PolQ�/� bonemarrow stromal cells were cloned and demonstrated to have a lower saturation density of 2.5 ± 0.1, 1.7 ± 0.1 or 1.9 ± 0. 1 � 106

cells for clones 1, 3 or 5 compared to PolQ +/+ cells 5.3 + 0.2 � 106 cells per plate (p = 0.003,\0.0001,\0.0001, respectively).The PolQ +/+ cell line had a doubling time of 18 hours compared to 30 hours for PolQ�/� clones 1 and 3 or 24 hours for clone 5.The plating efficiencies were similar with the PolQ +/+ having a plating efficiency of 18.35 ± 0.63% compared to 11.37 ± 0.59,21.92 ± 2.70 or 10.45 ± 1.10% for PolQ �/� clones 1, 3 or 5, respectively. PolQ �/� cells were more sensitive to irradiation asshown by decreased Do from 1.38 ± 0.06 Gy for PolQ +/+ cells compared to 1.27 ± 0.16, 0.98 ± 0.10 (p = 0.0316), or 1.08 ± 0.01(p = 0.0103) Gy for PolQ �/� clones 1, 3 or 5, respectively. Before irradiation the PolQ �/� mice had an increased number ofperipheral red blood cell micronuclei of 0.91 ± 0.05% of cells with micronuclei compared to 0.41 ± 0.02 or 0.32 ± 0.03% for thePolQ ± or +/+ mice (p\0.0001). Forty hours after irradiation, the percent of micronuclei in the PolQ�/�mice increased to 8.53 ±1.48% compared to 2.35 ± 0.12 or 2.55 ± 0.17% for the PolQ +/+ or PolQ ± mice (p = 0.0032 or 0.0039, respectively).

Conclusions: The data demonstrate that mice and derived cell lines lacking the PolQ polymerase are radiosensitive and implicatePolQ polymerase and base excision repair in DNA repair following ionizing irradiation.

Author Disclosure: J.P. Goff, None; M. Epperly, None; D. Shields, None; T. Smith, None; M. Seki, None; J. Wittschieben, None;R. Wood, None; J.S. Greenberger, None.

113 Association between Rad-51 Levels and Survival in Patients With Glioblastoma Multiforme Treated With

Radiation Therapy

R. Kumar1, J. Welsh2, R. B. Nagle1, S. Green3, B. Stea2

1Department of Pathology, University of Arizona Health Sciences Ctr., Tucson, AZ, 2Department of Radiation Oncology,University of Arizona Health Sciences Ctr., Tucson, AZ, 3Division of Biometry, Arizona Cancer Ctr., Tucson, AZ

Purpose/Objective(s): The Rad-51 protein plays an important role in DNA double stranded break (DSB) repair induced by radi-ation. The Rad-51 is regulated by p53, and in cancers (such as gliomas) where p53 is mutated or absent, the activity of Rad-51 isincreased. Recent findings suggest that cell lines overexpressing Rad-51 are resistant to both chemotherapy and radiation inducedDSBs. We measured Rad-51 levels in glioblastoma multiforme (GBM) tumor specimens obtained at diagnosis and at the time ofrecurrence; these levels were correlated to patient survival.

Materials/Methods: Matched initial and recurrent tumors from 10 patients were used to construct tissue microarrays. All patientshad been treated with radiation (median dose 60 Gy). Immunohistochemistry was performed using Rad-51 antibody (Santa Cruz