Abstract BACKGROUND Glioblastoma is treated with surgery to remove as much as the tumour as possible then radiation therapy (RT) and temozolomide (TMZ) to cause extensive DNA damage in any remaining tumour cells, triggering cell death by apoptosis. However, glioblastoma cells respond by upregulating the DNA damage response (DDR) triggering DNA repair that contributes to treatment resistance. Progression then generally occurs rapidly leaving patients with a poor prognosis and survival. We previously showed that gartisertib, an inhibitor of ATM- and Rad3-Related protein (ATR), a serine/threonine kinase that triggers DNA repair, sensitises glioblastoma cells to TMZ and RT. We extend this investigation to brain-penetrant ATR inhibitors elimusertib (BAY1895344) and ceralasertib (AZD6738). MATERIAL AND METHODS Patient-derived glioblastoma cell lines were treated with a dose response of ATRi (elimusertib and ceralasertib) combined with a clinically relevant dose of TMZ (35uM) and/or RT (2Gy). A human astrocyte cell line was also treated with a dose response of single agent elimusertib and ceralasertib. Cells were incubated for 7 days and assessed for cell viability through an MTT assay or cell-cell Incucyte analysis. RESULTS Both elimusertib and ceralasertib as single agents reduced glioblastoma cell growth, with elimusertib being more potent compared to ceralasertib. Elimusertib had greater potency in causing cell death of the glioblastoma cell lines than a human astrocyte cell line. However, there was less difference between glioblastoma and astrocyte cell death when treated with ceralasetib. When combined with RT and/or TMZ, elimusertib once again was more potent in reducing glioblastoma cell growth compared to ceralasertib. CONCLUSION We have expanded our in vitro data showing that various ATR inhibitors (elimusertib, ceralasertib and previously gartisertib) are able to reduce the growth of glioblastoma cells in combination with RT and TMZ treatment. This provides further support for initiating a clinical trial of ATR inhibition in combination with standard treatment for patients with glioblastoma.

Dianthin-30 or gelonin versus monomethyl auristatin E, each configured with an 1 anti-calcitonin receptor antibody, are differentially potent in vitro in high grade 2 glioma cell lines derived from glioblastoma 3 4 Roger Gilabert-Oriol, Sebastian G.B. Furness, Brett W. Stringer, Alexander Weng, 5 Hendrik Fuchs, Bryan W. Day, Angela Kourakis, Andrew W. Boyd, David L. Hare, 6 Mayank Thakur, Terrance G. Johns, Peter J. Wookey. 7

Aim/Background: Consequences of higher DNA damage and higher constitutive activity of DNA damage response (DDR) proteins, especially Checkpoint kinase 1 (CHK1), in advanced malignancies such as Glioblastomas (GBM) are unknown. In this study we aim to assess the role of DNA damage modulator, CHK1, and its effectors in the pathogenesis of GBM. Methods: 1323 patient samples were assessed to identify CHK1 effectors and establish the clinical relevance of the key proteins in GBM. Small molecule inhibitors and siRNA approaches were used to dissect the mechanism. Cell viability, proliferation, monolayer, 3D culture and Tetracycline-inducible knockdown system were used to assess the functional impact on perturbing the newly identified pathway in U251MG and primary GBM cells. Finally, heterotopic and intracranial xenograft mouse models for human GBM were used validate the in vitro findings. Results: In contrast to normal brain clinical samples, higher expression and a strong positive correlation for CHK1 and CIP2A expression was observed in samples from glioma patients. Further, high CHK1 and CIP2A mRNA expression predicted for more aggressive form of the disease and reduced overall survival. Accordingly, GBM patients with higher CIP2A copy number had a lower probability of survival. Chemical and genetic inhibition of CHK1 resulted in decreased CIP2A expression. Functionally, using small molecules, siRNAs and inducible-knockdown systems we demonstrate that CHK1 and CIP2A promote viability, clonogenicity and the tumoursphere-forming capacity of GBM cells. Further, depletion of CIP2A induced senescence. Mechanistically, we demonstrate that CHK1 promotes pSer727-STAT3 which in turn regulates CIP2A transcription. Interestingly, CIP2A binds STAT3 and promotes pSer727-STAT3 levels both in vitro and in vivo thereby establishing a positive feedback between the two proteins. Finally, CIP2A depletion in the heterotopic and orthotopic murine models of human GBM results in decreased tumour growth and increased survival of mice. Conclusions: Altogether, the results highlight the prognostic and therapeutic significance of CHK1 and CIP2A in human gliomas and provide novel mechanistic insights into GBM pathogenesis. Disclosure: All authors have declared no conflicts of interest.

Review on EPHA1 (EPH receptor A1), with data on DNA, on the protein encoded, and where the gene is implicated.

Abstract Abstract 3724 Poster Board III-660 Lymphomas account for about 5% of all cases of cancer in the US. In 2009, it is estimated that 74490 Americans will be diagnosed with lymphoma, which includes approximately 8510 new cases of Hodgkin's lymphoma and 65980 new cases of non-Hodgkin's lymphoma (NHL). An estimated 20790 men and women will die from lymphoma in 2009, the majority (93.8%) of which from NHL. Standard treatment includes chemotherapy and radiation to destroy the malignant lymphoma cells; however, newer therapies are currently being explored. Bortezomib is approved for multiple myeloma and relapsed mantle cell lymphoma, and is under clinical investigation in other types of NHL and other B-cell malignancies. Similar to bortezomib, MLN9708 is a modified dipeptidyl boronic acid that is a potent, reversible and specific inhibitor of the proteasome. MLN9708 preferentially binds to and inhibits b5 enzymatic activity of the 20S catalytic core. It is currently in Phase I trials for the treatment of solid and hematologic malignancies. MLN9708 immediately hydrolyzes to MLN2238, the biologically active form, on exposure to aqueous solutions or plasma. MLN2238 was used for all preclinical studies described below. We evaluated the ability of MLN2238 to inhibit tumor growth in two preclinical models of diffuse large B-cell lymphoma (DLBCL), an aggressive lymphoma that accounts for ∼40% of lymphomas in adults. OCI-LY10 and PHTX-22L are two examples of ABC-subtype of DLBCL. The antitumor activity of MLN2238 was evaluated in CB-17 Scid female mice bearing OCI-LY10 lymphoma xenografts. Tumor inhibition was determined by calculating the treatment over control (T/C) ratio of the mean tumor volume and the percentage of tumor growth inhibition (TGI) on the last day of the study. MLN2238 administered intravenous (IV) at 18 mg/kg once weekly (QW) resulted in tumor regression in 6 of 7 animals (T/C = 0.12; TGI = 88%) and a partial response in the remaining animal (43% reduction in tumor volume). MLN2238 administered at 8 mg/kg or 4 mg/kg IV QW resulted in a T/C of 0.42 and 0.36. Proteasome activity in the tumor was inhibited 49 and 44% 1 hour after a single IV dose of MLN2238 at 14 or 7 mg/kg, respectively. The other tumor model, PHTX-22L, is a primary tumor xenograft that was derived from a tumor surgically removed from the right lymph node of a 71-year-old Caucasian male. SCID NOD female mice implanted with PHTX-22L tumor fragments were treated twice weekly (BIW) with IV doses of vehicle, MLN2238 at 4 mg/kg SC once daily (QD), MLN2238 at 14 mg/kg IV BIW or bortezomib at 0.8 mg/kg IV BIW. The strongest antitumor activity was seen in PHTX-22L tumor bearing mice treated with IV MLN2238 14 mg/kg BIW (T/C = 0.14; TGI = 86%, p < 0.001) or SC MLN2238 4 mg/kg QD (T/C = 0.15; TGI = 85%, p < 0.01). These dosing regimens significantly inhibited tumor growth compared with vehicle treatment. In the same study, bortezomib treatment IV at 0.8 mg/kg on a BIW schedule had no antitumor effect (T/C = 0.88; TGI = 12%, p = 0.86) and was not statistically different from the vehicle treated group. Proteasome activity was significantly inhibited in both blood and tumor following a single dose of MLN2238 administered IV at 14 mg/kg or SC at 4 mg/kg. Proteasome activity was inhibited greater than 50% for up to 8 hrs following the IV dose in both blood and tumor, whereas proteasome activity recovered more quickly following the SC dose. In a separate preclinical study, the antitumor activity of MLN2238 administered at lower doses and on alternate schedules was examined. SCID NOD mice bearing PHTX-22L xenografts were administered MLN2238 IV at 11 mg/kg BIW, 7 mg/kg BIW, 11 mg/kg QW or 5 mg/kg QDx2/week resulted in significant antitumor activity (T/C = 0.04, 0.08, 0.15, 0.16; TGI = 96, 92, 85, 84%, respectively). More importantly, drug treatment caused tumor regression (tumor volume at the end of the study was less than 50% of the starting tumor volume) in 8 of 10 mice in the 11 mg/kg BIW, 5/10 in the 7 mg/kg BIW, 2/8 in the 11 mg/kg QW and 1/10 in the 5 mg/kg QDx2/week treatment groups. The efficacy of MLN2238 dosed IV either at 7 mg/kg QW or 3.5 mg/kg BIW in this model was less efficacious (T/C = 0.61 and 0.64). In addition to PHTX-22L, we have access to several other primary lymphoma tumor xenograft models that we plan to evaluate for sensitivity to MLN2238 treatment. These studies demonstrate that MLN2238 is active in preclinical models of lymphoma, and that MLN2238 has antitumor activity in a model of lymphoma that is refractory to bortezomib treatment. Disclosures: Donelan: Milllennium: Employment. Bannerman:Milllennium: Employment. Bano:Milllennium: Employment. Babcock:Milllennium: Employment. Hales:Millennium Pharmaceuticals: Employment. Stringer:Milllennium: Employment. Burke:Milllennium: Employment. Danaee:Milllennium: Employment, Equity Ownership. Faron-Yowe:Milllennium: Employment. Koenig:Milllennium: Employment. Lichter:Milllennium: Employment. Kupperman:Milllennium: Employment.

An early and reliable blood test is one deficiency in diagnosis of malignant pleural mesothelioma (MPM). Megakaryocyte potentiating factor (MPF) and mesothelin variants (MSLN), members of the mesothelin gene family, have been studied as candidate serum markers for MPM. We developed a novel enzyme-linked immunosorbent assay (ELISA) system to compare the diagnostic efficacy of MPF and MSLN in MPM and control groups.MPF and MSLN were assayed with ELISA in 27 consecutive MPM patients and 129 controls including patients with lung cancer and asymptomatic asbestos-exposed subjects.Statistically significant elevation of serum MPF and MSLN levels was noted in MPM patients in comparison with every control group. The area under the receiver operating characteristic curve (AUC) was calculated for differentiation of MPM and lung cancer, healthy asbestos-exposed subjects, and healthy adults. While the AUC for serum MPF was 0.879, cut-off = 19.1 ng/ml (sensitivity = 74.1%, specificity = 90.4%), the AUC for serum MSLN was 0.713, cut-off = 93.5 ng/ml (sensitivity = 59.3%, specificity = 86.2%). Comparison between AUC for MPF and MSLN values shows that MPF is significantly superior to MSLN (p = 0.025). Finally, there was a significant correlation between MPF and MSLN values for MPM (Pearson's correlation coefficient = 0.77; p < 0.001).These findings suggest that diagnostic value of MPF for MPM was better than that of MSLN although both markers showed almost equal specificity for MPM.

13059 Background: The mitotic kinase Aurora A is implicated in the development of multiple tumor types. MLN8054 is an oral, potent and selective small-molecule inhibitor of Aurora A with broad efficacy in preclinical models of cancer. Inhibition of Aurora A by MLN8054 induces accumulation of mitotic cells, followed by apoptosis. This study explores relationships between Aurora A inhibition, mitotic index, and tumor growth inhibition for xenograft models with different sensitivity to MLN8054. The marker response in mouse skin was also studied. Methods: Mice bearing subcutaneous xenografts were dosed orally qd or bid with MLN8054 for 21 days. Pharmacodynamic markers were studied after 1–2 doses. Formalin-fixed xenograft tissues were stained with the mitotic markers pHisH3 and MPM2, or with an antibody to the T288 autophosphorylation site on Aurora A. Tumor growth inhibition (TGI) was calculated using the formula 100 - [ΔT/ΔC * 100], where ΔT is the volume change for treated tumors, and ΔC is the volume change for control tumors. Results: HCT116 human colon xenografts were sensitive to MLN8054 on a qd or bid schedule (84% and 96% TGI respectively for 30mg/kg dose). The T288 autophosphorylation site was used to directly demonstrate inhibition of Aurora A, which resulted in dose-dependent duration of the elevation in mitotic index. Efficacy was similar for qd vs bid dosing of 30mg/kg MLN8054, and accordingly we found that a single dose was sufficient to elevate the mitotic index for about 20–24h in this model. SW480 human colon xenografts have MLN8054 sensitivity similar to that of HCT116, but more modest effects on mitotic index were observed. The mitotic index profile of SW480 is similar to that of MDA-MB-231 xenografts, the most insensitive model studied. Elevated mitotic index was also observed in mouse skin. Conclusions: We found that mitotic index measurements coupled with the T288 autophosphorylation site as a direct marker of Aurora A activity are useful for monitoring inhibition of Aurora A by MLN8054 in tumor and/or skin biopsies. In a sensitive model, greater duration of mitotic index elevation results in greater efficacy. Our continuing work aims to better understand the differences in marker and efficacy responses between xenograft lines, incorporating the pT288 antibody as a direct marker of Aurora A inhibition. [Table: see text]