Search

3101 Background: RO4929097 is a oral inhibitor of gamma-secretase, which results in Notch signaling inhibition. Prior work has demonstrated that Notch-signaling inhibition enhances chemosensitivity of colon cancer cells. This study sought to combine RO4929097 with capecitabine (cape) and determine maximum tolerated dose (MTD), toxicities and efficacy. Preclinical and prior phase I work demonstrated possible autoinduction of RO4929097, so pharmacokinetic (PK) evaluation of RO4929097 and cape was planned. Methods: Adult patients with refractory solid tumors were eligible and received RO4929097 and cape at a fixed dose of 1000 mg/m2 BID with escalating doses of RO4929097 on a 21-day cycle according to a 3+3 design. Cape was administered for 14 days and the RO49029097 once daily, 3 days per week. RO4929097 plasma concentrations were evaluated by LC/MS/MS on days 3 and 10 of cycle 1, and PK parameters analyzed with WinNonLin version 5.2. Results: 4 dose levels have been completed (20, 30, 45 and 68 mg); 18 of 19 patients are evaluable for toxicity and PK data is available for 11 patients. One DLT has been observed (intolerable grade 2 fatigue) at 68 mg. There have been 2 confirmed partial responses: fluoropyrimide-refractory colon cancer (for 12 cycles) and cervical cancer (for 6 cycles). The half-life, Cmax and AUC of RO4929097 all significantly decreased on day 10 compared to day 3, with an increase in clearance for all doses. The half-life was significantly shorter in dose level 3 (p=0.0012) and dose level 4 (p=0.0126) compared to dose level 1. Conclusions: RO4929097 plus cape was well tolerated. Activity was seen in cervical and colon cancer. Autoinduction of RO4929097 was seen both with increasing cycle number and increasing dose. However, all plasma concentrations of RO4929097 were above those needed for Notch inhibition (1.9 ng/mL based on prior studies). Dose level 1 (cape 1000 mg/m2 BID and RO4929097 20 mg daily) is the recommended phase II dose. [Table: see text]

Notch-1 is a transmembrane receptor whose activation upon ligand binding or calcium depletion leads to translocation of the Notch-1 intracellular domain (NICD) to the nucleus where it activates transcription of Notch target genes. Targeting Notch signaling, such as with inhibitors of γ-secretase that is required for Notch-1 activation, is currently in early clinical development. It is critical to characterize the expression of Notch-1, an important member of Notch family of receptors in human cancers with a validated assay suitable to monitor pharmacodynamic responses to Notch inhibitors and to carry out clinical validation investigation of Notch-1 as predictive/prognostic biomarker. This could improve our understanding of pathobiology of Notch signaling and in turn increase the odds of success of targeting Notch signaling. The aims of this study were to develop an immunohistochemistry assay for Notch-1/NICD (activated Notch-1), and to evaluate expression levels and patterns of Notch-1/NICD in paraffin-embedded tumor samples. Human breast cancer cell lines MCF-7, MDA-MB-231 and MDA-MB-468, and colorectal cancer HCT-116 cells were treated with and without a calcium chelator EDTA for 10 min and embedded in paraffin after formalin fixation. Expression levels and patterns of Notch-1 were examined by Western blot, and immunocytochemistry as well as immunohistochemistry and were quantitated using an Automated Cellular Imaging System. There was an increase in Notch-1/NICD expression in cells treated with EDTA versus without EDTA in MCF-7 (staining index: 28 vs. 10) and MDA-MB-231 (81 vs. 62) cells by immunocytochemistry and confirmed by Western blot. Interestingly, membranous/cytoplasmic Notch-1 was in part translocated into the nucleus in MCF-7 and MDA-MB-468 cells following EDTA treatment, suggesting Notch-1 activation and specific detection of Notch-1 in distinct cellular compartments. In addition, a predominant cytoplasmic/membranous expression of Notch-1 was observed in MDA-MB-231 and HCT-116 cells with and without EDTA treatment. Low to intermediate levels of Notch-1 with a predominantly cytoplasmic localization, relative to moderate to high levels in cell lines, were detected in 13% (6/45), 4% (2/48), 16% (7/45) of tumor cells in carcinomas of the breast, lung, and ovary. Nuclear/cytoplasmic staining of Notch-1 was observed in one case of lung cancer. In addition, Notch-1 expression was found in a significant fraction of endothelial cells of human solid tumors. We have established a fit-for-use immunohistochemistry assay with an antibody that specifically detects Notch-1 in formalin-fixed and paraffin-embedded human tumors. Its application could facilitate the evaluation of pharmacodynamic responses to various Notch targeting agents undergoing clinical development and subsequent clinical validation studies. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr 746. doi:1538-7445.AM2012-746

219 Background: Notch signaling is a key determinant of cell fate. RO492 is a selective γ-secretase inhibitor blocking cleavage of Notch. We hypothesize that anti-androgen therapy will lead to an enriched population of basal prostate cancer stem cells. Targeting Notch following anti-androgen treatment will lead to delay in re-growth of mature luminal PC cells. We are using a novel double blind, placebo controlled randomized discontinuation design to treat men with rising PSA. A non-castrating anti-androgen (BIC) was used to induce initial tumor regression Methods: Pts with rising PSA after primary therapy for PC, and no radiographic evidence of disease were enrolled in this ongoing study. All pts received BIC 50mg/day for 16 weeks (induction phase). Pts achieving a >50% (modified from 80% decrease after the first 8 pts) decrease in PSA were randomized to placebo or RO492 20mg orally 3 days on/ 4 days off per week (randomization phase). Primary endpoint is PSA progression (increase in PSA by 50% over nadir with minimum PSA rise of 2ng/mL) during randomization phase. Pts meeting primary endpoint could receive 12 months of RO492 and BIC (combination phase). Serum is collected for evaluation of soluble markers of gamma secretase inhibition. 29 pts/arm in randomization phase gives 90% power to detect a decrease in1 yr PFS from 80% to 45% with 5% alpha. Results: We report on the first 16 patients that have enrolled. Median pretreatment PSA was 6.3ng/ml (0.35 to 34). 9 of 16 pts on the induction phase came off study due to PSA progression, although 5 of 9 pts would have been randomized using the 50% PSA response criteria. 6 pts have been randomized to RO492 or placebo (1 pt not yet randomized) and 2 pts have proceeded to combination phase. Median PSA decrease during induction phase in 6 pts who went on to randomization was 86% (69–97). Overall decline in PSA in induction phase was 67% (7.3-99%). Placebo/RO492 has been well tolerated with 1 pt experiencing g1 fatigue, and 1 pt experiencing g1 nausea in the combination phase. 8/16 pts had g1 breast tenderness with BIC. Conclusions: In this ongoing randomized phase II trial, BIC decreased PSA by >50% in 12/16 pts. Placebo/RO492 is well tolerated. Enrollment continues with analysis of plasma correlates to be compared to clinical results.

Background: Dysregulated Notch signalling has been implicated in a variety of cancers. R is a potent and selective inhibitor of -secretase, a key enzyme in Notch signalling. Phase I data demonstrated R to be tolerable but PK profile revealed auto-induction with repeated dosing at high doses (Tolcher et al. ASCO 2010). The primary aim of this study is to determine the optimal regimen of R using alternative dosing schedules. Its secondary aim is to establish mouse XG from tumor biopsies to enable correlative biomarker research. Methods: Pts with advanced solid tumors were enrolled in one of 6 different schedules and starting doses of R. Serial PK samples were collected on Day 1 and at steady state. Mandatory pre and post treatment tumor biopsies were procured if medically safe and implanted subcutaneously into NOD-SCID mice for propagation in up to 4 serial generations. Tumor DNA from second XG passage was analyzed using the Sequenom OncoCarta Panel v1.0 (SOCP). Results: To date, 24 pts have been enrolled. PK data are available for 21 pts, summarized in Table 1. The maximum tolerable dose for schedules C-F has not been reached. Commonest adverse events (AE) of all grades with at least possible attribution to R were fatigue (n=14), nausea (13), hypophosphatemia (11), anorexia (9) and diarrhoea (8). Grade 3/4 AEs consisted of hypophosphatemia (n=2), lymphopenia (2), vomiting (1) and QTc prolongation (1). Data on engraftment are available from biopsies of 18 pts (18 pre and 16 post treatment samples). Primary tumor sites included gastrointestinal (n=6), breast (4), melanoma (2), ovarian (2), pancreatic (1), neuroendocrine (1), parotid (1) and lung cancers (1). In pre treatment biopsies, the rate of initial engraftment was 12/18 (67%); while in post treatment biopsies, the rate of initial engraftment was 14/16 (88%). Further passages (2nd and 3rd) from viable engrafted cancers demonstrated a 100% engraftment rate. The SOCP analyses were carried out in 9 different XG models with mutations detected in 4 (KRAS [n=3] and BRAF mutations), all of which were from metastatic colon cancers. Conclusions: Preliminary PK evaluation demonstrated auto-induction of R in schedules A and B (concordant with prior data), but to a lesser degree in schedules C, D & E. It is unclear whether this observation was due to the lower administered doses of R or the alternative schedules employed. This study demonstrated the ability to generate mouse XG models from different tumor histologies using core biopsies, with a mean engraftment rate of 78% at first implantation. We also confirmed the feasibility of detecting point mutations in cancers using SOCP platform on XG specimens. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2011 Nov 12-16; San Francisco, CA. Philadelphia (PA): AACR; Mol Cancer Ther 2011;10(11 Suppl):Abstract nr B50.

Background: The NOTCH signaling pathway has been implicated in the tumorigenesis of several human cancers. Pre-clinical models have shown that deficiency of components of the NOTCH pathway results in abnormal vascular remodelling. Upregulation of NOTCH signaling has been implicated in resistance to anti-VEGF therapy, providing a rationale for the combination of RO, a gamma secretase inhibitor that disrupts NOTCH signaling, and Cd, a VEGFR-tyrosine kinase inhibitor. The primary objective of this study is to determine safety, tolerability and the recommended phase II dose (RP2D) of the combination of RO and Cd. Secondary objectives are preliminary anti-tumor efficacy, as per RECIST v1.1 measurements, as well as pharmacokinetic and pharmacodynamic studies. Methods: Adults with histologically confirmed solid malignancies for whom no effective conventional treatment is available are eligible. A standard 3+3 dose-escalation design is used. Cycle 1 is 42 days (D), where RO is given orally once daily D1–3, 8–10, 15–17, 22–24, 29–31, 36–38 and Cd is given orally once daily D22–42. Cycle 2 and onwards are 21 days, with RO given D1–3, 8–10, 15–17 and Cd given continuously D1–21. The dose-limiting toxicity (DLT) evaluation period is 42 days (cycle 1). Results: As of August 15 2011, 11 patients (pts) have been treated at two dose-levels (DL). 7 pts were treated at DL1 (RO = 10 mg; Cd =20 mg,) and 4 were treated at DL2 (RO = 20 mg; Cd = 20 mg). Pts had received a median of 3 prior lines of treatment prior to enrolment and received a median of 3 cycles on study. The most common drug related adverse events were: diarrhea (n = 9), hypertension (n = 6), nausea (n = 5), headache (n = 4), fatigue (n = 4), liver enzyme elevation (n = 3), hypophosphatemia (n = 3), anemia (n =3) and hypothyroidism (n =3). There were 2 DLTs observed: grade 3 hypertension occurred in a pt with metastatic melanoma treated at DL1 and grade 4 aspartate aminotransferase (AST) elevation in a pt with metastatic uterine leiomyosarcoma treated at DL2. Best response was stable disease in 7 pts, progressive disease in 3 and 1 pt was inevaluable. A prolonged response of over 6 cycles was observed in 3 pts (1 pt with cholangiocarcinoma withdrew consent after completing cycle 6, 1 pt with endometrial stromal sarcoma is still on study and stable after 8 cycles, 1 pt with colon cancer progressed after completing 7 cycles). Reasons for discontinuation were progressive disease in 6 pts, toxicity in 1 pt and withdrawal of consent in 1 pt. 3 pts are still on study. Conclusion: The combination of RO and Cd appears feasible; dose escalation is ongoing to establish the RP2D. Pharmacokinetic and pharmacodynamic studies to evaluate soluble markers of angiogenesis will be conducted. The combination of RO and Cd shows encouraging early signs of activity, with 3 of 11 patients achieving stable disease for at least 6 cycles. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2011 Nov 12-16; San Francisco, CA. Philadelphia (PA): AACR; Mol Cancer Ther 2011;10(11 Suppl):Abstract nr C17.

Induction of epithelial-to-mesenchymal transition (EMT) in cancer stem cells (CSCs) can occur as the result of embryonic pathway signaling. Activation of Hedgehog (Hh), Wnt, Notch, or transforming growth factor-β leads to the upregulation of a group of transcriptional factors that drive EMT. This process leads to the transformation of adhesive, non-mobile, epithelial-like tumor cells into cells with a mobile, invasive phenotype. CSCs and the EMT process are currently being investigated for the role they play in driving metastatic tumor formation in breast cancer. Both are very closely associated with embryonic signaling pathways that stimulate self-renewal properties of CSCs and EMT-inducing transcription factors. Understanding these mechanisms and embryonic signaling pathways may lead to new opportunities for developing therapeutic agents to help prevent metastasis in breast cancer. In this review, we examine embryonic signaling pathways, CSCs, and factors affecting EMT.

3065 Background: We evaluated the safety of a combination of bevacizumab (B) and a TKI against VEGFR-cediranib (C). Methods: This was a phase I study with 3+3 design in patients (pts) with advanced...

Abstract 3975 Background: Bortezomib (btz, VELCADE®) is approved in the US for the treatment of patients (pts) with multiple myeloma, and for the treatment of pts with mantle cell lymphoma who have received at least one prior therapy. Btz is oxidatively metabolized by hepatic cytochrome P450 enzymes to pharmacologically inactive deboronated metabolites. Thus, btz pharmacokinetics (PK) may be altered in pts with hepatic impairment (HI). This phase 1 study was undertaken to evaluate the PK and safety of btz in pts with varying grades of HI and to inform dosing recommendations in these subpopulations. Here we present the available interim results. Methods: Pts with advanced malignancies were assigned to four groups based on hepatic function (bilirubin and aspartate aminotransferase [AST] levels relative to the upper limit of normal [ULN]) using the classification developed for organ dysfunction studies by the National Cancer Institute's Cancer Therapy Evaluation Program. Normal function was defined as bilirubin and AST ≤ULN. Pts with mild HI had bilirubin ≤ULN and AST >ULN, or bilirubin >1.0–1.5 × ULN (with any AST). Moderate and severe HI were defined as bilirubin >1.5–3 × ULN and >3 × ULN, respectively, with any AST. Pts received btz on days 1, 4, 8, and 11 of 21-day cycles. Pts with normal hepatic function received btz at the standard dose of 1.3 mg/m2. Pts with severe, moderate, and mild HI received escalating doses of btz from starting doses of 0.5, 0.7, and 1.0 mg/m2, respectively, via a standard 3+3 design to 1.3 mg/m2. Serial blood samples were collected for 24 hours post-dose on days 1 and 8 of cycle 1 for measurement of btz plasma concentrations. Individual values of dose-normalized (DN) area under the concentration–time curve from time zero to the point of last quantifiable concentration (AUC) and maximum concentration (Cmax) were calculated by noncompartmental analysis. PK parameters were summarized by hepatic function and subject to an analysis of variance to estimate the ratio of geometric means and 90% confidence intervals (CIs) for each HI group relative to the control group. Results: A total of 53 pts were treated, 14 with normal hepatic function, and 16, 9, and 14 with mild, moderate, and severe HI, respectively. Median age was 62 years (range 30–85), 53% were female, 94% were white, and 11/58/30% had ECOG performance status 0/1/2. Dose escalation proceeded to 1.3 mg/m2 in pts with mild HI; data collection is ongoing at 1.0 mg/m2 in pts with moderate and severe HI. The most common adverse events (AEs) were fatigue (53%), nausea (42%), and dyspnea (32%). For most AEs, no association was apparent between increased incidence and poorer hepatic function; the incidence of elevated AST was 7, 25, 33, and 36% in the normal group and mild, moderate, and severe HI groups, respectively. PK data were available for 51 pts. Geometric mean DN AUC was 30.2, 23.0, 50.3, and 47.7 (ng.hr/mL)/(mg/m2) on day 1 and 52.2, 51.9, 83.4, and 80.3 (ng.hr/mL)/(mg/m2) on day 8 in pts with normal hepatic function and mild, moderate, and severe HI, respectively. Respective values for geometric mean DN Cmax were 56.7, 38.2, 66.0, and 84.2 (ng/mL)/(mg/m2) on day 1, and 88.9, 79.6, 59.2, and 96.1 (ng/mL)/(mg/m2) on day 8. Geometric least square mean ratios for DN AUC and DN Cmax between groups are shown in the table. Compared to pts with normal hepatic function, mild HI did not alter DN AUC and DN Cmax. However, mean DN AUC was increased by approximately 60% in pts with moderate or severe HI. Conclusions: Pts with mild HI do not require a starting dose adjustment and should be treated per the recommended dose of btz. Pts with moderate or severe HI should be started at a reduced dose of 0.7 mg/m2 during the first cycle, and a subsequent dose escalation to 1.0 mg/m2 or further dose reduction to 0.5 mg/m2 may be considered based on pt tolerance. These data have resulted in an update to the US Prescribing Information for VELCADE®. Comparison vs normal Geometric least square mean ratio (90% CI) Day 1 Day 8 DN AUC Mild HI 0.833 (0.603–1.151) 0.954 (0.664–1.371) Moderate HI 1.662 (1.136–2.430) 1.538 (0.969–2.440) Severe HI 1.578 (1.139–2.186) 1.477 (1.046–2.085) Average of moderate/severe HI 1.619 (1.199–2.186) 1.507 (1.075–2.113) DN Cmax Mild HI 0.697 (0.446–1.090) 0.879 (0.528–1.463) Moderate HI 1.162 (0.688–1.962) 0.612 (0.318–1.178) Severe HI 1.483 (0.947–2.323) 1.023 (0.633–1.653) Average of moderate/severe HI 1.313 (0.868–1.986) 0.791 (0.493–1.271) Disclosures: Venkatakrishnan: Millennium Pharmaceuticals: Employment. Off Label Use: Discussion of Velcade in patients with advanced non-hematologic malignancies is included. Takimoto:Johnson & Johnson: Employment, Equity Ownership. Neuwirth:Millennium Pharmaceuticals, Inc.: Employment. Parasuraman:Millennium Pharmaceuticals: Employment, Equity Ownership. LoRusso:Millennium Pharmaceuticals, Inc.: Membership on an entity's Board of Directors or advisory committees, Research Funding.

Spinal cord compression in Wilms' tumor is a rare event, generally caused by invasion of the canal by paraspinal lesions or metastatically involved vertebral bodies. This case report reviews the clinical presentation, radiologic evaluation, and emergent therapy in two cases of spinal cord compromise involving patients with widely metastatic Wilms' tumor. One of these is the only known report of intradural metastasis in a child with this malignancy. Both cases illustrate the importance of anticipating and rapidly responding to neurologic complications that may arise in patients with aggressively metastatic Wilms' tumor.

Background: Angiogenesis inhibitors may be classified as direct or indirect inhibitors. Direct inhibitors like bevacizumab (B) that target the microvascular endothelial cells and prevent them from responding to different endothelial mitogens. Conversely, indirect inhibitors like cediranib (C), target growth-factor tyrosine kinases and its VEGF receptors on tumor endothelium. We hypothesized that maximizing VEGF inhibition with a combination therapy consisting of a direct angiogenesis inhibitor (B) and an indirect inhibitor (C) with different VEGF blocking mechanisms may result in increased anti-tumor activity and a safe toxicity profile. As a first step, we conducted a phase I study of the combination of bevacizumab and cediranib in advanced cancer patients. Methods: Pts with advanced solid tumors refractory to standard therapy were eligible. The trial is a conventional phase I 3+3 dose-escalation design. Cediranib was administered orally continuously for 21 days. Bevacizumab was given every two weeks (days 1 and 15) at escalating doses. Correlative studies including pharmacokinetic studies of cediranib, DCE-MRI, and circulating tumor (CTCs) and endothelial cells (CECs). Results: A total of 27 pts have been treated. Dose level 3 (C-30 mg/day and B-5mg/kg/2 weeks) was determined to be the MTD with the major dose-limiting toxicity of hypertension. The most common drug related toxicities include hypertension (66%), fatigue (41%), and anorexia (33%). Currently, of the 17 evaluable pts by RECIST, 7 pts had SD for >4 cycles (months). One confirmed PR was noted in a metastatic basal cell carcinoma pt (−33%) and one unconfirmed PR (−31%) in ovarian cancer pt. Several notable minor responses by RECIST have also been observed in the SD pts; 2 pts with RCC (−21%, −23%), an alveolar soft tissue sarcoma pt (−24%), and a prostate cancer pt (−21%). Overall, immature CECs (CD105+) enumerated by CellSearch™ revealed a dose-dependent significant decrease (1,290% to 93%; p=0.0001). A 3-fold induction in apoptosis was observed at 24 hrs compared to baseline in the CD105-positive CECs. Overall, mature CECs (CD31+) assessed for VEGFR2 activity revealed an 83% and 1.9% significant (p=0.019) inhibition in pVEGFR2 expression at low (B;3mg/kg) and high (B;5mg/kg) doses, respectively. In the non-responders, mature CECs revealed a dose-dependent (C;20mg/kg–30mg/kg) significant increase (−6.8% to 63%;p=0.031) in pERK/ERK expression levels. PK and DCE-MRI analysis is pending. Conclusions: The MTD was determined to be cediranib 30mg/day and bevacizumab 5mg/kg/2weeks. Activity has been seen in a diverse number of tumor types. Inhibition of pVEGFR2 and induction of apoptosis in CECs shows target effect by the combination. Grant support: National Cancer Institute grant and 2 UO1 CA062461-15 Citation Information: Mol Cancer Ther 2009;8(12 Suppl):A4.

Abstract 2729 Poster Board II-705 Background: There are limited effective treatment options for patients with diffuse large B cell lymphoma who relapse post autologous stem cell transplant or are transplant-ineligible. The detection of vascular endothelial growth factor (VEGF )A, B and C isoforms and their receptors on many large cell lymphoma samples suggests that the VEGF pathway is critically important and may contribute to disease progression. Sunitinib maleate, is an orally bioavailable inhibitor of VEGF receptor-1 (VEGFR-1), -2, and -3, PDGFR-α and β as well as KIT, FLT3, RET and CSF-1 (Chow, LQ JCO 2007). We tested the efficacy, safety and biomarker activity of sunitinib in patients with relapsed diffuse large B cell lymphoma in a multi-center prospective phase II study. Methods: Eligibility included age 18 or older, histologically confirmed relapsed or refractory diffuse large B-cell (DLBCL), primary mediastinal (PMBCL), or transformed indolent lymphoma, at least one and no more than 2 prior chemotherapy regimens (one anthracycline-containing). The primary endpoint was objective response defined by 1999 Cheson criteria. The secondary endpoints were progression-free and overall survival, toxicity and the effect of sunitinib on peripheral blood circulating endothelial cells (CECs) and their precursors (CEPs). Patients self administered sunitinib 37.5 mg po daily with no breaks in 4 week cycles for up to 1 year. CT imaging was performed every 2 cycles and CEC/CEP assays were done at baseline, day 1 of cycles 2, 3 then q 3 months thereafter. A Simon 2-stage design was used with at least 1 response needed after 15 evaluable patients to complete a planned accrual of 25 patients total. Results: Between Feb 2007 and September 2008 we enrolled 19 patients at 7 Canadian sites - 17 were eligible and are evaluable for toxicity and 15 are evaluable for response. The median age was 65 and patients were a median of 20.3 months(m) from diagnosis (range 5.8–132 m). Fourteen (82%) had a diagnosis of DLBCL, 10 (58%) had responded to their preceding line of chemotherapy - 5 (29%) had relapsed post high dose chemotherapy. The median number of cycles of sunitinib received was 2 (1–5) with only 5 patients remaining on drug for 3 or more cycles. Only 35% of patients received > 90% planned dose intensity with 14 patients missing doses and 5 undergoing dose reductions necessitated by toxicities. Hematological toxicity (grade 3 neutropenia in 5 pts, grades 3–4 thrombocytopenia in 6 pts) and was the most common reason for dose omission. Of the 15 evaluable pts, no objective responses were seen, and 9 achieved stable disease (median duration 3.4 m); and 6 had primary progressive disease. As a result, the study was closed at the end of the first stage. With limited serial sampling, there was no discernable relationship between the change in absolute or apoptotic CEC over time with clinical response as measured by best response or change in bi-dimensional measurements. Conclusions: Sunitinib 37.5 mg po daily showed no evidence of anti-tumour activity in relapsed/refractory large B cell lymphoma and is associated with greater than expected hematological toxicity. Disclosures: Buckstein: Celgene: Honoraria, Research Funding; Novartis: Honoraria, Research Funding. Off Label Use: It is being tested in NHL.

Abstract 3744 Poster Board III-680 HSP-90 is abundantly expressed in a variety of cancer cells, including lymphoma. HSP-90 chaperones several client proteins that are involved in the oncogenic process, and therefore has become an appealing target for cancer therapy. Previous in vitro studies demonstrated that inhibition of HSP-90 by the small molecule 17-AAG (17-allylamino-17-demethoxy-geldanamycin) can induce apoptosis in a dose and time dependent manner by reducing the cellular contents of critical survival proteins, including Akt and cyclin D1 in a variety of lymphoma cell lines. With this background, we conducted a phase II study of 17-AAG in adults with relapsed/refractory mantle cell or Hodgkin lymphoma. Patients were eligible if they had a relapsed disease, and adequate organ function. Patients who had primary refractory disease were required to have a maximum of 4 prior treatment regimens. Twenty-two patients were enrolled (13 classical Hodgkin lymphoma, and 9 mantle cell lymphoma) with a median age of 49.5 years. Median prior treatments were 3 (range 1-4) and 36% (n=8) received prior autologous stem cell transplant. 17AAG was given at 220 mg/m2 by intravenous infusions, initially over 1hour, and subsequently over 6 hours. Doses were given on days 1, 4, 8 and 11 of 21-day cycles. Twenty-two patients received at least one dose of 17AAG and were evaluable for toxicity. The most common non-hematologic toxicity were diarrhea (n=8), fatigue (n=7), nausea (n=6), and fever (n=6). Grade 3 and 4 toxicities occurred in 8 patients; pain and thrombocytopenia being the most common. Three patients who were treated with the 1 hour infusion died after receiving 1-2 doses of therapy. One patient had extensive lung disease, including pleural effusions, developed worsening of respiratory symptoms requiring intubation, a second patient had extensive disease in the mediastinum and large pleural effusion who also complained of worsening shortness of breath that was associated with nausea and vomiting few hours after the first dose of 17-AAG requiring visit to the emergency center, an EKG showed no major abnormalities, however he had a cardiac arrest shortly thereafter, and a third patient with primary refractory extensive disease also had pulmonary and liver disease complained of chest pain and subsequently had a cardiac arrest. After revising the study to prolong the infusion time to 3 to 6 hours, and to exclude patients with pulmonary and cardiac diseases, no additional deaths were observed. Eighteen of the 22 patients enrolled were evaluable for treatment response, of whom, 7 (39%) had reduction in their tumor measurement, and 2 (11%) achieved partial remissions. Paired core needle biopsies from diseased lymph nodes were obtained from selected patients before and within 24 hours of the first infusion of 17-AAG, and tissue sections were evaluated by immunohistochemistry for changes in the expression of selected client proteins. Remarkably, even after one dose of therapy, 17-AAG downregulated cyclin D1 and Akt proteins in tumor cells, decreased the percentage of tumor cells expressing Ki-67, and increased the percentage of cells expressing activated caspase-3. Collectively, these data demonstrate for the first time that targeting HSP-90 can modulate the level of several client proteins in lymphoma cells in vivo, and can achieve clinical responses in patients with relapsed lymphoma, providing a proof-of-principle for further developing HSP-90 targeted therapy in patients with lymphoma. Disclosures: Off Label Use: 17-AAG in lymphoma.

1133 Background: Pazopanib, an oral small molecule inhibitor of VEGFR, PDGFR, and KIT, has demonstrated activity in phase I, with a recommended phase II dose of 800 mg/d (Hurwitz H et al, J Clin Oncol. 2005;23[16 suppl]:3012.1). We evaluated the activity of single agent pazopanib in recurrent or metastatic breast cancer (MBC). Methods: In this 2-stage design, patients with recurrent or MBC received pazopanib 800 mg/d. The primary endpoint was objective response rate (ORR) of 20%. Response in 3 out of 18 patients was required to go to stage 2. Treatment was continued until progression. Results: 21 patients entered stage 1; 67% were ER positive and all were HER-2-negative. Prior lines of chemotherapy were 1 in 76% and 2 in 14%. Of the 19 evaluable patients, 2 patients remain on treatment. 14 (74%) stopped due to progressive disease, 2 (10%) due to adverse events, and 1 (5%) due to patient request. Best response was partial response (PR) in 1 (5%), stable disease (SD) in 11 (58%), and progressive disease in 7 (37%). Clinical benefit rate (CR, PR, or SD for ≥ 6 months) was 26%. Median time to progression (TTP) was 3.7 months (95% C.I. 1.7 months - not reached). 9 out of 18 patients (50%) with measurable target lesions had some decrease in target lesion size. Estimated progression-free survival at 3 months was 55%, and 28% at 6 months. Adverse events were grade 3/4 elevations in AST (14%) and ALT (10%), and grade 3 hypertension and neutropenia (14% each). Other common events were grade 1/2 lymphopenia, neutropenia, diarrhea, fatigue, skin hypopigmentation, hypertension, nausea, vomiting, anorexia, and headache. Conclusions: Pazopanib is well tolerated and demonstrates activity in pretreated breast cancer. While the target ORR of 20% has not been met, rates of SD and TTP are comparable to other active agents in this setting, and therefore pazopanib may be an interesting agent for future studies in breast cancer. [Table: see text]

2508 Background: 3-Aminopyridine-2-carboxaldehyde thiosemicarbazone (3AP) is an inhibitor of ribonucletoide reductatse with activity in hematological malignancies as well as melanoma and prostate cancer. This study describes the population PK of 3AP and the relationship between 3AP disposition and patient covariates. Methods: 40 pts with advanced cancer from two phase 1 studies were included in the PK model building. Pts received 3AP 25–105 mg/m2 IV on day 1. 3AP plasma and erythrocyte levels were sampled at 10 timepoints over a 24-hour period and measured by a validated HPLC method. Data were analyzed by a nonlinear mixed- effects modeling approach with NONMEM system Version V. Xpose10 and S-PLUS were used for goodness-of-fit assessment and model evaluation to obtain clearance (CL), volume of distribution (V) for compartment (C) 1, plasma (V1); V for C2, erythrocytes (V2); and V for C3 (V3). 15 covariates were evaluated, including: weight, BSA, gender, PS, age, ABCB1[C1236T, G2667T, C3436T] genotypes, toxicity, concurrent chemotherapy [irinotecan, doxorubicin], response and number of cycles administered. Normal renal and hepatic function were required for study entry and were not included as covariates. Response was assessed after 2 cycles of therapy. Results: 3AP PK were described as a 3-compartment model with first-order elimination. Estimated paramaters were: V1 =5.68 L/m2 (95%CI 4.3–7.1); V2=18.9 L/m2 (95%CI 14.6–23.2); V3= 40.0 L/m2 (95%CI 23.8–55.4); CL 25.4 L/hr/m2 (95%CI 22.4–28.4). Gender was associated with V; women had a lower V2 (p No significant financial relationships to disclose.

e13509 Background: E7389 (E) is a synthetic analogue of halichondrin B, an investigational tubulin-based antimitotic drug. Gemcitabine (G) is a nucleoside analogue clinically active in several human tumours. These two drugs exhibit synergistic cytotoxic effects against the H522 non-small cell lung cancer (NSCLC) xenografts. Methods: A phase I study of these two drugs in combination was initiated in patients with advanced solid tumours. Two prior chemotherapy regimens for metastatic disease were allowed. Results: Patient characteristics: male 11/female 10; median age 59 (range 28–84); performance status 0 /1/2: n=1/13/7; prior chemotherapy 21, prior radiotherapy 7, prior immunotherapy 1; tumour types: ovarian cancer 3, endometrial cancer 3, NSCLC 3, gastric/esophageal adenocarcinoma 3, miscellaneous 9. Total number of cycles given 52, median cycles/patient (range) 2 (1, 8). Cohort (CT) 1: E/G given days 1,8,15 q28 days. Due to DLT, regimen changed in CT 2 with E/G given days 1, 8 q21 days. Cycles (C) given: median 2 range 1–8, total 52. Response: partial response 1 (ovarian cancer), stable 8 [minor response 4 (NSCLC 2, endometrial cancer 1, head and neck cancer 1)], progression 8, inevaluable 4. Conclusions: Further phase II investigation of this regimen should be considered at a dose of E 1.0 g/m2/G 1000 mg/m2 q 3 weeks. [Table: see text] No significant financial relationships to disclose.

3525 Background: Rare circulating tumor cells (CTCs) and endothelial cells (CECs) offer a feasible approach for studying the pharmacodynamic effects of investigational agents. We investigated the effects of bevacizumab (B) and cediranib (C) on inhibition of the VEGFR pathway and correlated these changes with dose and clinical response. Methods: Peripheral blood was obtained at baseline, 24hrs and at C2D26–30 post-treatment from patients (n=14) undergoing dose escalation of intravenous B and oral C. CTCs and CECs (CD31+ or CD105+) were isolated and immunofluorescently stained. Laser scanning cytometry (LSC) was used to quantify phosphorylated and total-VEGFR2 (pVEGFR2/VEGFR2), pERK/ERK, and apoptosis in each phenotype. Changes in each biomarker were correlated with partial response (PR) or stable disease and progression > 2 months, evaluated using RECIST. Results: Overall, immature CECs (CD105+) enumerated by CellSearch™ revealed a dose-dependent significant decrease (p=0.0001). A 3-fold induction in apoptosis was observed at 24 hrs compared to baseline in the CD105+ CECs. Mature CD31+ cells assessed for VEGFR2 activity revealed an 83% and 1.9% significant (p=0.019) inhibition in pVEGFR2 expression at low (B;3mg/kg) and high (B;5mg/kg) doses, respectively. In the non-responders, mature CECs revealed a dose-dependent significant increase (-6.8% to 63%;p=0.031) in pERK/ERK expression levels. No significant changes were observed in CTC enumeration by CellSearch™. LSC-mediated CTC enumeration revealed a 4.77 % and 2.33% increase in CTCs following treatment in the non- responders and responder (p=0.809), respectively. Analysis of pVEGFR2 in CTCs revealed a 58% inhibition in the responder versus a 163% increase in expression in the non-responders (p=0.63). Conclusions: Inhibition of pVEGFR2 and induction of apoptosis in CECs confirmed the target therapy. An increase in CD105+ CECs is consistent with the hypothesis that anti-angiogenic efficacy induces endothelial cell shedding. Assessment of CECs indicates that B and C are more biologically active at lower doses, and resistance may be attributed to ERK activity. Support: UO1 CA062461 (RK) No significant financial relationships to disclose.

Targeted agents are promising new therapeutics that have entered clinical trials for glioblastoma. AZD2171 (cediranib) is a potent, oral pan-VEGF receptor tyrosine kinase inhibitor with a half-life of 20 hours compatible with once-daily dosing. A primary target of AZD2171, VEGFR2, is expressed on glioblastoma endothelium. We have demonstrated normalization of tumor vessels in recurrent glioblastoma patients administered AZD2171 on a daily basis. Normalization has rapid onset, is reversible, and is associated with alleviation of brain edema (Batchelor TT, et al. Cancer Cell 2007;11:83-95). In this Phase II study of 31 recurrent glioblastoma subjects we now report clinical outcomes, including the proportion of patients alive and progression-free at 6 months (APF6) as the primary endpoint. Secondary endpoints include radiographic response proportion; progression-free survival; and overall survival and toxicity. At this time we are presenting radiographic response data, APF6, PFS, and OS on 31 patients. Thirty patients have experienced disease progression and one patient remains in follow-up without progression. The primary and secondary endpoints are tabulated below: Radiographic Responses Partial Response (more than 50% decrease in volume) in 16/30 (56%) APF6 [95% CI] 25.8% [14.7%, 46.9%] PFS [95% CI] 117 days [88, 145] OS [95% CI] 221 days [172, 285] Only one of the first 16 patients was removed from the study due to toxicity (fatigue). Dose limiting toxicities were observed in 9/16 patients with hypertension; fatigue and diarrhea were encountered most often. There have been no intracerebral hemorrhages. AZD2171 alleviated brain edema, a major cause of morbidity in glioblastoma patients, and had a steroid-sparing effect in many of the patients enrolled. Blood biomarkers were serially assessed and it was observed that plasma VEGF, PlGF, and SDF1α were increased after treatment. Plasma PlGF and VEGF decreased upon cediranib discontinuation. Plasma bFGF and SDF1α and viable circulating endothelial cells increased when tumors escaped treatment with AZD2171. AZD2171 is a promising therapy for glioblastoma and an attractive therapeutic partner for chemotherapy or radiation.

3566 Background: 17-DMAG binds to heat-shock protein 90, leading to degradation of oncogene client proteins, inhibiting tumor growth. Preclinically, 17-DMAG has potent antitumor activity. Methods: 17-DMAG was given by 1–2 hour infusion twice weekly for 4 weeks/28-day cycle. Starting dose: 1mg/m2/dose. Eligibility: ECOG = 2, adequate organ function. Exclusions: prolonged QTc/uncontrolled illness. An accelerated titration escalation design was used; one patient (pt)/dose level was entered until a single pt experienced dose-limiting toxicity (DLT) or 2 pts have grade (gr) = 2 toxicity during the first cycle. Ending the accelerated phase, it converts to a standard 3–6 pt/cohort design. Objectives: To determine the toxicity profile, pharmacokinetics (PK), pharmacodynamics (PD) and MTD of twice weekly 17-DMAG. Plasma samples collected before and up to 48 hours after 17-DMAG infusion were analyzed by LC-MS. Results: 23 pts were accrued and treated; Diagnosis: renal cell-1, pancreatic-2, medullary thyroid-1, NSCLC-3, colorectal -5, peritoneal mesotelioma-1 adrenocortical carcinoma-1, pheochromocytoma-2, malignant thymoma-1, melanoma-2, sarcoma-1, hepatocarcinoma-1 and head & neck -1 and ovary-1. 2 DLTs (gr 3 peripheral neuropathy and renal failure) were observed at 27 mg/m2/dose. Toxicities at the MTD dose (21 mg/m2/dose) have been acceptable: Myalgias/arthralgias being the most common. Other gr 1–2 toxicities included: diarrhea, weight loss, fatigue, elevated transaminases, hair loss, electrolyte abnormalities (hyponatremia and hypo/hyper-magnesemia) and leukocytopenia. PK is linear, with AUC and Cmax increasing as dose was escalated. Drug clearance is independent of dose, but highly variable (10.75 ± 5.56 L/hr/m2). Half-life: 10.0 to 31.7 h, with a median of 17.2 h. At the MTD: Cmax was 523.6 ± 265.3 ng/mL; Very high interindividual variability (CV = 70.3%) and a mean AUCinf of 2467.5 ± 1733.8 hr*ng/mL. Disease stabilization was achieved in a pt with mesothelioma (19 mos +) and another with head and neck cancer (8 mos +). Conclusions: We recommend 21mg/m2/dose twice weekly as the Phase II dose. The study continues to accrue at the MTD to collect paired biopsies for proteomic evaluation on Hsp90 client proteins analysis. No significant financial relationships to disclose.

3531 Background: The combination of B with A was evaluated to recommend a phase II dose for using A in combination with B, since raf-kinase and VEGF inhibition by distinct pathways could result in synergy and prevent resistance to B. Patients and Methods: Eligibility included histologically confirmed advanced malignancy, at least 4 weeks since prior systemic cancer therapy and/or radiation, adequate organ function and performance status of 0–2. B was administered orally at fixed dose of 400 mg twice daily starting 2 weeks prior to A, which was added in escalating dose cohorts starting at 300mg/m2 intravenously over 3 hours on days 1, 8 and 15 in a 28 day cycle. Toxicity was assessed weekly and clinical response evaluated every 2 cycles. Results: 19 patients have been registered (6 clear cell renal, 1 papillary renal, 4 colorectal, 1 each with cervix, adrenal, oral, thyroid, pancreas cancer and 3 with melanoma). 16 have been treated for at least the first two cycles at 4 dose levels on this ongoing study. Dose limiting toxicity has not yet been reached at 400mg/m2 of A. Major toxicities were grade 3 fatigue in 1 patient and grade 2 nausea and diarrhea in 3 and 2 patients respectively. Grade 3 hand foot syndrome related to B was noted in 3 patients. Of 14 response evaluable patients, 1 demonstrated partial response (thyroid cancer) and 6 had stable disease;3 clear cell renal,1 each with papillary renal and cervix cancers, and 1 melanoma. Latter patient had stable disease for 10 cycles and was resected to render disease free. PK analysis reveals that steady-state was achieved after 1 week of B; addition of A did not alter PK of B. PK levels of A will be performed. Studies of Hsp90a in the plasma have been performed on 10 patients. Of the 10 patients evaluated for plasma Hsp90a (baseline values pre-A and at 4 and 18 days after A), 5 of the 6 clinically stable patients had 1.8 to 6.8- fold increase; 3 patients with clinical progression and 1 clinically inevaluable patient had induction of 0.8 to 1.7-fold. Conclusion: The combination of A and B is well tolerated and clinical benefit may correlate with induction of plasma Hsp90a. Supported by NIH# 5UO1CA062487- 14, 2MO1RR016500–006 and NCI CTEP. No significant financial relationships to disclose.

2561 Background: Becatecarin (rebeccamycin analogue-RA) is an anti-tumor antibiotic with inhibitory activity against both topoisomerase II and I as well as DNA intercalating properties. We performed a phase I trial to a) determine the maximum tolerated dose (MTD) of RA in combination with oxaliplatin; b) determine the dose limiting toxicities (DLT) (c) obtain data on pharmacokinetics and (d) observe for any antitumor activity. Methods: Eligibility criteria included patients with advanced solid tumors refractory to standard therapy; performance status 0–2; adequate hematologic, renal and liver function. Patients were treated with RA as a 1 hour infusion daily x 5 and oxaliplatin on day 5 only, after RA infusion. Treatment was repeated q 21 days. The following dose levels were evaluated: Dose level 1: RA 80 mg/m2/d and oxaliplatin 90 mg/m2; Dose level 2: RA 80 mg/m2/d and oxaliplatin 130 mg/m2; Dose level 3: RA 110 mg/m2/d and oxaliplatin 130 mg/m2. Results: A total of 15 evaluable patients were enrolled. Median age was 56 (8 male, 7 female). A variety of tumor types were enrolled. A total of 56 cycles were administered. DLT occurred at a dose of RA at 110 mg/m2/d x 5 days and oxaliplatin at 130 mg/m2 and consisted of grade 3 hypophosphatemia and grade 4 atrial fibrillation. At this dose level 2 of 3 enrolled patients also developed grade 3 neutropenia. The MTD and recommended phase II dose was RA at 80 mg/m2/daily x 5 along with oxaliplatin 130 mg/m2 day 5 q 21 days. Three confirmed partial responses were observed in patients with hepatocellular, gallbladder and esophageal cancers. Six patients experienced stable disease. Conclusions: At the MTD combination RA and oxaliplatin is well tolerated and given the response rate and stable diseases observed, phase II studies are recommended. Supported by Grants U01 CA62502, MO1-RR-00080, K23 CA109348–01 from the National Institutes of Health. No significant financial relationships to disclose.

2001 Background: AZD2171 (cediranib) is a potent, oral pan-VEGF receptor tyrosine kinase inhibitor with a half-life of 20 hours compatible with once-daily dosing. A primary target of AZD2171, VEGFR2, is expressed on glioblastoma endothelium. We have demonstrated normalization of tumor vessels in recurrent glioblastoma patients treated with daily doses of AZD2171. Normalization has rapid onset, is reversible and is associated with alleviation of brain edema [Cancer Cell 2007; 11: 83]. Methods: In this phase II study of 30 recurrent glioblastoma subjects the primary endpoint was the proportion of patients alive and progression-free at 6 months (APF6). Secondary endpoints include radiographic response proportion; progression-free survival; overall survival and toxicity. At this time we are presenting radiographic response data and toxicity on the first 16 consecutive patients and APF6, PFS and OS on all 30 patients. Complete information will be available on all 30 patients at the time of presentation. Results: Twenty-eight patients have experienced disease progression and two patients remain in follow-up without progression. The primary and secondary endpoints are tabulated below: Only one of the first 16 patients was removed from the study due to toxicity (fatigue). Dose limiting toxicities of hypertension, fatigue and diarrhea were observed in 9/16 patients. There were no intracerebral hemorrhages. AZD2171 alleviated brain edema, a major cause of morbidity in glioblastoma patients, and had a steroid-sparing effect in the first 16 patients enrolled. Blood biomarkers were serially assessed and elevated levels of bFGF, SDF1a and viable circulating endothelial cells correlated with disease progression. Conclusions: AZD2171 has activity in patients with recurrent glioblastoma. Combination studies of AZD2171 with radiation and chemotherapy are planned. [Table: see text] [Table: see text]

15553 Background: 17-Allylamino-17-Demethoxygeldanamycin (17-AAG) is a benzoquinone ansamycin antibiotic with anti- proliferative activity in several mouse xenograft models including prostate cancer models. Serum IL-6, IL-8, and maspin are potentially important markers of prostate cancer biology. Methods: Patients (pts) with metastatic, hormone-refractory prostate cancer progressing on at least one prior systemic therapy with rising PSA were eligible. All pts received 17-AAG at a dose of 300 mg/m2 IV weekly for three out of four weeks. Primary objective was to assess the PSA response (50% decrease in PSA). Secondary objectives included Adverse Events (AEs) and correlative serum markers including IL-6, IL-8 and maspin levels. A Simon two-stage design required a total of 25 pts with early termination if < 2 responses occurred among the 1st 16 eligible patients. Results: Seventeen pts were enrolled of which 15 were deemed eligible. Median age was 68 and median PSA was 252 ng/mL. Pts received 17-AAG for a median number of 2 cycles. No pt had a PSA response. No grade 4/5 AEs occurred. Grade 3 AEs included fatigue (4 pts), lymphopenia (2 pts) and back pain (2 pts). The median PSA progression free survival was 1.8 months (95% CI: 1.3–3.4 months). The six-month overall survival was 61% (95% CI: 37%-100%). Due to the lack of PSA response, accrual was stopped per study design. At day 15, the median IL-6 and IL-8 increase from baseline was 0.4 pg/ml (p=0.57) and 3 pg/ml (p=0.73), respectively. Maspin levels had day 15 increase of 6-fold (p=0.44). At treatment failure (TF), the median IL-6 increase from baseline was 4.47 pg/ml (p=0.03) and IL-8 decrease was 1.8 pg/ml (p=0.31). Maspin levels had a 29-fold increase at TF (p=0.09). Conclusions: 17-AAG did not show any activity with regards to PSA response. Serum IL-6 was significantly increased at the time of TF. Further evaluation of 17-AAG at a dose of 300 mg/m2 IV weekly in this patient population is not warranted. No significant financial relationships to disclose.

5549 Background: Although cisplatin (cis) is an effective radiosensitizer, its activity in patients (pts) with advanced and recurrent CC is disappointing. Oxaliplatin (O) might exhibit greater activity and be suitable for combination with paclitaxel (P). Methods: Pts with advanced and recurrent CC who had not received prior chemotherapy except radiosensitizing cis were treated with P 175 mg/m2 IV and O 130 mg/m2 IV every 21 d. Response (R), as determined by RECIST criteria and confirmed at 9 weeks (wk), and toxicity were primary outcomes. If R in = 8 patients (pts) of 18, accrual to 46 could ensue. Results: Of 17 pts enrolled, 16 were treated. Histology: 7 adeno/10 squamous.The median age was 57 (range 33–78) and 13 had had prior radiation (10 with cis). Median cycles: 3 (0–8). One CR and 4 PRs were achieved for an overall response rate of 29% (95% confidence interval: 10.3%, 56.0%). Additionally, 4 had stable disease, but this was not confirmed at 15wk. Time to progression in those responding was 21 wk (range 11–51). 7 pts are alive. 6 pts experienced grade (gr) 3/4 hematologic toxicity. Neuropathy occurred following cycle 3 in 2 pts (gr3). Additonal gr 3/4 toxicities were gastrointestinal (GI)/metabolic in 6, thrombotic in 1 and hypersensitivity in 1. There were no treatment-related deaths. Conclusions: In the stage I of our Phase II trial, O and P demonstrate a 29% response rate when used to treat advanced and recurrent CC. Toxicities of this regimen are hematologic, gastrointestinal, and neurologic. Supported by N01-CM-62204. No significant financial relationships to disclose.

The heat shock protein Hsp90 functions to stabilize important cell survival- and proliferation-related kinases such as AKT, IKK, c-Src, Raf-1, and cdk9. Cancer cells have activated Hsp90 as compared to normal cells, making this a relevant therapeutic target to eliminate these kinases. The semi-synthetic geldanamycin derivative 17-allylamino-17-demethoxygeldanamycin (17-AAG) binds to and inhibits the activity of Hsp90, and previous work demonstrated the ability of 17-AAG to deplete AKT in several cancer types in vitro. A newer geldanamycin derivative, 17-(dimethylaminoethylamino)-17-demethoxygeldanamycin (DMAG) has improved pharmacological properties including solubility and oral bioavailability, and was shown to be more effective than 17-AAG in melanoma and pancreatic carcinoma mouse xenograft mouse models. We therefore examined the effects of 17-AAG and DMAG against CD19-positive tumor cells from CLL patients. Cell viability was examined by the MTT assay, and AKT and IKK protein expression was examined by immunoblot analysis. In samples from seven CLL patients, 1.0uM DMAG resulted in 31.5% viability (95% CI: 13.1–50.0%), compared to 61.5% viability (95% CI: 45.0–78.0%) using the same concentration of 17-AAG. In four CLL patient samples treated with 1.0uM DMAG for 24 hours, AKT was decreased an average of 72.5% (95% CI: 57.7–87.3%) relative to the untreated controls. This is in comparison to 1.0uM 17-AAG, which caused a 52.7% decrease in AKT (95%CI: 39.7–65.6%). IKK protein was also decreased at similar levels in all patient samples examined. This data indicates that in CLL cells, DMAG has superior activity both in cytotoxicity and in reduction of relevant Hsp90 client proteins. 17-AAG is currently undergoing Phase I clinical testing in CLL, and DMAG is completing phase I clinical development in solid tumor malignancies. Overall, our data and that of others support clinical development of DMAG in CLL, based on the improved pharmacologic properties, enhanced efficacy relative to 17-AAG, and expected down-regulation of target proteins. In addition, our in vitro observations support using measurement of protein down-regulation of AKT and IKK as pharmacodynamic biomarkers of activity in patients undergoing therapy with these agents.

12007 Background: B inhibits proteasome function and may be synergistic in causing apoptotic death with platinum agents. We were interested in combining B with OX but concerned with dose limiting (DL) NT based on our prior B phase I study [Hamilton et al., JCO 2005]: therefore this Phase I dose-escalation study (alternating increases of B and OX) focusing on NT was planned. Methods: Patients (pts) with metastatic solid tumors, PS 0–2, platinum or taxane naive, no peripheral neuropathy and adequate organ function, received B (D1, 4, 15, 18) and OX (D1, 15) every 28 days in a dose escalation design (see table ). Baseline and monthly assessments were performed by an independent neurologist. Results: 27 (18 gastrointestinal, 3 melanoma, 3 ovarian, 3 others) were accrued; pt characteristics: 14 male/13 female; median age 55 years (range 35–75); 2 median cycles (range 1–10). NT was not DL because it did not occur within the first cycle. Late and limiting NT was observed in levels 2–5 after 2–9 cycles, but serial neurologic evaluations showed reversible NT. With an amended new dose level to lower B to 1.0 mg/m2 (level 6) to avoid late NT, NT was not observed. Of 22 evaluable pts, there were 3 partial responses (ampullary, GE junction, biliary), 6 stable disease, and 13 disease progression by RECIST criteria. Conclusions: biweekly BOX is tolerable at B 1.0 mg/m2 and OX 85 mg/m2 with no DL NT. Additional observations on late NT are ongoing. This suggests B is a major contributor to NT observed in dose levels 2–5 and may potentiate the effects of OX. [Table: see text] [Table: see text]

2032 Background: Bortezomib is a highly selective and reversible inhibitor of the proteasome with activity in multiple myeloma and other malignancies. Patients (pts) with renal impairment have been treated in previous trials, but there has not been a systematic investigation into the effects of renal dysfunction on dosing. Study objectives were to characterize the pharmacokinetic (PK) and pharmacodynamic (PD) profile of bortezomib, and to determine the maximum tolerated dose (MTD) in adults with advanced malignancy and renal insufficiency ranging in severity from mild to dialysis dependence. Methods: Fifty-one pts have received intravenous bortezomib at 0.7 mg/m2 up to the approved dose of 1.3 mg/m2 on days 1,4, 8, and 11 every 3 weeks. Pts were stratified by 24-hour creatinine clearance (CrCL) normalized to a body surface area of 1.73 m2 into five cohorts per the table . Doses were escalated in cohorts of three pts in groups B-E. Blood samples were assayed for bortezomib concentration, as well as the PD endpoint of 20S inhibition. Results: Escalation of bortezomib doses to 1.3 mg/m2 was well tolerated in all groups with CrCL ≥ 20 mL/min/1.73 m2. There has been only one instance of dose limiting toxicity (group C at 1.3 mg/m2) which did not prevent successful completion of this cohort. No patients discontinued therapy due to renal deterioration. Doses of 0.7 mg/m2 were tolerable in Group D patients (CrCL< 20 mL/min/1.73 m2 ). Five dialysis pts have been treated; 3 at 0.7 mg/m2, and 2 at 1.0 mg/m2. All tolerated therapy well, and accrual to Groups D and E continues. PK and PD assays are underway and the analysis will be reported in full. Conclusions: This study is the first comprehensive evaluation of bortezomib in pts with various degrees of renal insufficiency including dialysis dependence. Bortezemib at the approved dose of 1.3 mg/m2 on this schedule is well tolerated by pts with CrCL ≥ 20 mL/min/1.73 m2. Results of this trial will allow for dosing recommendations for bortezomib use in pts with renal insufficiency. [Table: see text] [Table: see text]

2004 Background: Ixabepilone (Ix) is a semisynthetic Epothilone B derivative with antitumor activity in breast cancer pts previously treated with taxanes and in chemotherapy naïve prostate cancer pts. Because Ix is hepatically metabolized, the following study was performed to define dosing recommendations for pts with varying degrees of hepatic impairment. Methods: Pts were classified into hepatic dysfunction cohorts defined by a modified NCI Organ Dysfunction Working Group (NCI) schema. Starting doses were escalated in new pts independently in each cohort using a standard phase I design. Results: Overall, 71 pts were registered and 66 pts are evaluable for cycle 1 dose limiting toxicities (DLTs). Ix was administered at 10–40 mg/m2 as 10 minute infusions q3wks. Dose levels reached 40, 40, 30, and 20 mg/m2 for pts in Groups A, B, C and D, respectively. In group B, DLTs were observed in 2/12 pts treated at 30 mg/m2 (febrile neutropenia, grade (gr)3 mucositis, and gr3 diarrhea) and 3/8 pts at 40 mg/m2 (febrile neutropenia, gr3 nausea/vomiting, gr3 hyponatremia). In group C, DLTs were observed in 2/10 pts at 20 mg/m2 (gr3 dehydration, gr3 muscle weakness) and 2/3 pts at 30 mg/m2 (febrile neutropenia, gr 4 neutropenia). In group D, DLTs were observed in 2/9 pts at 10 mg/m2 (gr3 infection and gr3 renal failure) and 1/5 pts treated at 20 mg/m2 (gr3 infection). Otherwise, Ix was generally well tolerated. Pharmacokinetic parameters are currently being analyzed. No objective responses have been documented. Conclusions: Ix dose reduction is required in pts with moderate to severe liver dysfunction. The recommended Ix dose for group C patients is 30 mg/m2. To define the impact of mild liver impairment, Group B has been stratified further into B1 (Bili ≤ ULN and AST > ULN) and B2 (ULN < Bili ≤ 1.5 × ULN, AST any). Accrual continues to groups B1 and B2 at 40 mg/m2 and group D at 20 mg/m2. [Table: see text] [Table: see text]

9678 Background: Proteasome inhibition has profound effects on cell cycle control and on the activation of transcription factors and anti-apoptotic signals such as NF-kB. It may potentiate DNA dama...

2076 Background: We previously demonstrated that CPDG2, which rapidly hydrolyzes MTX to inactive metabolites, when combined with LV and Thd, effectively rescues patients (pts) with HDMTX-induced nephrotoxicity. This study assessed the role of Thd administration in these pts. Methods: Pts (n=100) with HDMTX-induced renal dysfunction and significantly delayed MTX clearance received 1–3 doses of CPDG2, 50 U/kg IV, and standard LV rescue. In the initial cohort pts (n=35) also received 8 g/m2/d of Thd by continuous IV infusion. Subsequently, Thd was restricted to pts with prolonged (> 96 hours) exposure to MTX or with severe MTX toxicity at study entry (n=9). Plasma MTX concentrations (available in 75 pts), MTX associated toxicities, and outcome were compared in pts who did (Thd+) vs. did not (Thd-) receive Thd. Results: CPDG2 was administered at a median of 96 and 66 hours after start of the MTX infusion in Thd+ (n=44) or Thd- (n=54) pts, respectively. Median plasma MTX concentrations measured by HPLC pre CPD...

3050 Background: 17AAG is a benzoquinone antibiotic known to down-regulate oncoproteins by binding to heat shock protein 90. Previous studies have evaluated weekly, daily x 3 or daily x 5 schedules...

3030 Background: Therapy directed at the molecular chaperone HSP90 causes degradation of multiple client proteins critical in cancer cell proliferation and survival. 17AAG, a HSP90 directed agent, has demonstrated in vitro and in vivo antitumor effects in a variety of cancers. Methods: We performed a phase I trial to determine the maximum tolerated dose (MTD) and the dose limiting toxicities (DLT) of 17AAG infused on days 1, 4, 8 and 11 of a 21 day cycle, to characterize the pharmacokinetics of 17AAG and its effect on chaperone and client proteins in peripheral blood mononuclear cells (PBMCs). An accelerated titration design was utilized. Surrogate markers were measured in PBMCs at baseline, day 1 and 4. Pharmacokinetic analysis was performed on day 1 only. Results: Twelve patients received 35 courses (median, 2) at 6 dose levels. Both patients treated at 308 mg/m2 experienced one or more of the non-hematologic DLTs: hyperglycemia, dehydration, diarrhea (with maximal supportive treatment), ALT, AST, and a...

3031 Background: 17AAG is a benzoquinone antibiotic known to down regulate oncoproteins, by binding to heat shock protein 90 (HSP90). The study objective was to establish the dose limiting toxicity...

Background Inhibition of angiogenesis is a novel strategy for the treatment of cancer. We evaluated the safety and efficacy of cediranib, a potent small molecule inhibitor of the vascular endothelial growth factor receptor, in patients with refractory or recurrent small cell lung cancer (SCLC). Methods Patients with SCLC with progression after prior platinum-based chemotherapy only; performance status (PS) of 0 to 2; and adequate bone marrow, renal, and hepatic function were included. The dose of cediranib was 45 mg PO once a day for the first 12 patients and was reduced to 30 mg PO once a day for the subsequent patients because of intolerance of the higher dose. Treatment was given on a daily continuous schedule. The primary end point was determination of the response rate. Results Twenty-five patients were enrolled. Patient characteristics were as follows: 13 men; median age 61 years; PS 0 (12 pts), PS 1 (12 pts). A median of two cycles were administered. Salient grade 3/4 toxicities were fatigue, diarrhea, hypertension, proteinuria, and elevated liver enzymes. Tolerability was better with the 30 mg dose once a day. Nine patients had stable disease, but none had a confirmed partial response. The median progression-free survival and overall survival were 2 and 6 months, respectively. Response criteria to proceed to full accrual were not met. Increase in circulating endothelial cell count was noted at the time of progression in several patients. Conclusions Cediranib failed to demonstrate objective responses in recurrent or refractory SCLC at the dose and schedule evaluated. The 45 mg dose was intolerable in a majority of SCLC patients.

Relapsed small cell lung cancer (SCLC) carries a poor prognosis. Topoisomerase I and II inhibitors and DNA-damaging agents are considered among the most active agents against SCLC. Rebeccamycin analog (RA, Becatecarin) is an antitumor antibiotic with inhibitory activity against both topoisomerase I and II, and DNA-intercalating properties. We performed a phase-II trial of RA in relapsed, sensitive SCLC with the primary end point of response rate. Patients with previously treated SCLC who relapsed more than 60 days after the completion of first-line chemotherapy were treated with RA-administered intravenously at a dose of 140 mg/m2 on days 1 to 5 of 21-day cycles for a maximum of six cycles. Eligibility included Eastern Cooperative Oncology Group performance status 0 to 2 and adequate organ function. A two-stage design was employed. Twenty evaluable patients were enrolled. Median age was 61 years. Two patients (10%) had a partial response and six had stable disease. The clinical benefit rate was 40% (95% confidence interval [CI], 23–64%). The median progression-free survival was 2 months (95% CI, 1.2–5.2 months). The median survival was 6.7 months (95% CI, 3.3–8.0 months). No treatment-related deaths occurred. Grade-4 neutropenia and thrombocytopenia occurred in 23% and 14% of the patients, respectively. In conclusion, RA has single-agent activity in relapsed, sensitive SCLC with manageable toxicities but is unlikely to provide any superiority compared to existing agents for this disease.

10004 Background: Aberrant Hedgehog and Notch signaling is seen in sarcoma and anti-tumor activity is enhanced by inhibiting both pathways. This first in man Phase Ib study evaluated safety and efficacy of the combination of GDC-0449 (G), a smoothened inhibitor with RO4929097 (R), a gamma-secretase/notch inhibitor. Methods: The study evaluated fixed dose G (150 mg qd) for 21 days followed by G + R at either 10 mg (level 1) or 15 mg (level 2) QD. At MTD pts were evaluated with G + R (without “lead in” G) or single agent R. Pre and post treatment tumor biopsies were obtained to assess notch and hedgehog inhibition. PK was assessed for each drug. Key eligibility: advanced sarcoma, ECOG ≤ 2, age ≥ 18 and prior therapies ≤ 4. RECIST response was assessed Q6 wks. Results: 34 pts had a median age of 53 yrs (23-78), median priors 3 and various histologies [liposarcoma (7), chondrosarcoma (7), leiomyosarcoma (4), osteosarcoma (2), GIST (2) and other (12)]. No DLT was seen. Common (≥10 %) grade ≤3 toxicity was hypophosphatemia (18%). Systemic exposure (AUC0-24 and Cmax) of G was similar to established studies (not shown). AUC0-24 and Cmax of R was significantly lower (~70%) when administered with G (Table). However, measurement of unbound, free drug (Cfree) of R showed comparable levels between single agent R and G + R, but not when G was a “lead in” (Table). Target inhibition of the notch pathway (cleaved notch) and pAkt was observed in matched tumor biopsies with both arms. Durable SD was seen in both arms: myxoid chondrosarcoma (56 wks), liposarcoma (24+ wks), clear cell (21+ wks), desmoid (17+ wks), spindle cell (15+ wks) and chondrosarcoma (13+ wks). Conclusions: The combination of G+R is well tolerated and the RP2D is G 150 mg qd and R 15 mg qd without a “lead in” G. Despite reduction in total levels of R in the combination, free drug was similar and inhibits target. Disease stability was seen with R and R + G. Further clinical development of notch and hedgehog inhibitors in sarcomas is warranted. [Table: see text]

Normal ASB and the mutant enzyme in Maroteaux-Lamy syndrome (MLS) have been shown to be immunologically indistinguishable when compared using rabbit anti-ASB antiserum. In the present study a monoclonal antibody library was obtained by fusion of spleen cells from mice immunized with purified ASB and X63NS1/1-Ag-4 (NS1) mice myeloma cells. Seventy-six hybridoma lines with ASB antibody specificity were obtained, of which 64 were sub-cloned and further studied. Of these, 56 reacted with both control and mutant ASB. Five reacted only with the active form of ASB and not with either the mutant form or with partially denatured (enzymatically inactive) control ASB. Three reacted only with heat denatured or urea treated control ASB and with the mutant enzyme but not with the active form. The specificity of these antibodies toward ASB was documented using Western blotting. Several of these hybridoma lines were injected into Pris-tane sensitized mice, and the specific monoclonal antibodies were purified from the ascites fluid. A solid phase radioimmunoassay was developed using IV-16-8 antibody (reacts with control and mutant ASB and has high binding affinity to polyvinyl chloride), 125I-labeled V-4-7 (reacts with both mutant and control ASB), and 125I-labeled V-18-8 (reacts specifically with active ASB). Using this assay the ratio between active and inactive ASB was determined in liver, fibroblast and leukocyte homogenates from controls, homozygotes and heterozygotes for MLS.

Background Notch1 transmembrane receptor is activated through ligand-binding- triggered proteolytic cleavages and, upon release, the intracellular domain (N1-ICD) translocates into the nucleus and modulates target gene transcriptions. Notch activation has been implicated in tumorigenesis in an increasing number of human malignancies including non-small cell lung cancer (NSCLC). However, Notch1 in distinct expression patterns and activation status with tumor progression remains to be defined in NSCLC. Methods Notch1 and activated Notch1, N1-ICD, were examined by immunohistochemistry in 58 cases of stage I to IV NSCLC tumors. Association between Notch1 or N1-ICD expression and clinicopathological factors was assessed via correlation coefficient r statistics. P-values are two-sided. Results Detectable tumor Notch1, predominantly localized to the membrane and cytoplasm, was observed in 29 cases (50%, 95% Blyth-Still-Casella confidence interval 37 – 63%). It was negatively associated with stage (r = - 0.43, P