In preclinical models, the histone deacetylase inhibitor vorinostat sensitizes breast cancer cells to tubulin-polymerizing agents and to anti-vascular endothelial growth factor-directed therapies. We sought to determine the safety and efficacy of vorinostat plus paclitaxel and bevacizumab as first-line therapy in metastatic breast cancer (MBC), and the biological effects of vorinostat in vivo. For this purpose of this study, 54 patients with measurable disease and no prior chemotherapy for MBC received vorinostat (200 or 300 mg PO BID) on days 1–3, 8–10, and 15–17, plus paclitaxel (90 mg/m 2 ) on days 2, 9, 16, and bevacizumab (10 mg/kg) on days 2 and 16 every 28 days. The primary objective of the phase I study was to determine the recommended phase II dose (RPTD) of vorinostat, and for the phase II to detect an improvement of response rate from 40 to 60% (alpha = 0.10, beta = 0.10). No dose limiting toxicities were observed, and the RPTD of vorinostat was 300 mg BID. For the primary efficacy analysis in 44 patients at the RPTD, we observed 24 objective responses (55%, 95% confidence intervals (C.I) 39%, 70%). The adverse event profile was consistent with paclitaxel–bevacizumab, with the exception of increased diarrhea with the addition of vorinostat. Analysis of serial tumor biopsies in seven patients showed increased acetylation of Hsp90 and α-tubulin following vorinostat. Vorinostat induces histone and alpha tubulin acetylation and functional inhibition of Hsp90 in breast cancer in vivo and can be safely combined with paclitaxel and bevacizumab.

Abstract BackgroundVorinostat, also known as suberoylanilide hydroxamic acid or SAHA, is a potent, small molecule, pan-inhibitor of histone deacetylase (HDAC) with excellent oral bioavailability, and is commercially approved for cutaneous T cell lymphoma. Pan HDAC inhibitors induce acetylation of histone tails, relaxing condensed chromatin for transcriptional regulation of specific genes and can also target non histone nuclear proteins. In pre-clinical studies of HER-2 overexpressing breast cancer cells, vorinostat(V) synergistically augmented the effects of trastuzumab(T) by downregulating Her-2, pAKT, pERK 1/2, reduced apoptosis by decreasing Bcl-2, Bcl-XL, XIAP and survivin levels, while inducing pro-apoptotic Bim and Bak. Therefore, we hypothesized that adding vorinostat to trastuzumab would reverse trastuzumab resistance in patients with HER-2 positive metastatic breast cancer.MethodsAn open label dose de-escalation schema for the initial portion of the trial was planned. V 200 mg bid continuous daily dosing with 6 mg/kg T every 21 days was administered. Eligible patients: HER-2 positive by IHC (3+) or FISH, subsequently centrally confirmed; measurable metastatic or chest wall disease; any number of prior therapies including endocrine, chemotherapy, prior treatment with T and/or lapatinib were allowed; evidence of progressive disease on prior T was required. The Phase II primary objective was response rate by RECIST criteria.ResultsMedian age - 54 (range 40-69); PS was 0 or 1; 6 patients - ER+; median number of prior therapies - 3 (range 2-6). Six patients were treated at the starting dose of 200 mg bid daily V combined with 6 mg/kg T every 21 days. There were no dose limiting adverse events at this dosing level, eliminating the need to de-escalate and this was determined to be the optimal Phase II dose. A total of 16 patients were enrolled. The most common AEs (all grades) were 62%-diarrhea, 37%-nausea, fatigue; 31%-anemia, anorexia, vomiting, elevated creatinine, hyperglycemia, hypokalemia; 25%-thrombocytopenia. Grade 3,4 toxicities were dyspnea(n=2) and thrombocytopenia(n=2). Of the 16 total enrolled, 10 were centrally confirmed to have HER-2 positive disease by the ECOG Pathology Coordinating Office, 5 were HER-2 negative and 1 had insufficient tissue. None of 11 patients with centrally confirmed HER-2 positive disease (or unknown) had an objective response. One of 5 patients with centrally confirmed HER-2 negative disease had an objective response.ConclusionsIn this heavily pre-treated population who had either relapsed or progressed during trastuzumab containing therapy (either alone or in combination with chemotherapy), we failed to confirm our hypothesis that adding the histone deacetylase inhibitor, vorinostat, could reverse trastuzumab resistance. Correlative studies including hyperacetylation status of PBMC, CTC enumeration and peripheral as well as tumor methylation are ongoing. Citation Information: Cancer Res 2009;69(24 Suppl):Abstract nr 5084.

Abstract Abstract #404 Background: The oral HDACi vorinostat (V) sensitizes breast cancer cells to tubulin polymerizing agents and anti-VEGF directed therapies in vitro. We sought to determine the safety and efficacy of V plus paclitaxel (P) and bevacizumab (B) as first-line therapy for MBC. Methods: Eligible patients (pts) with measurable disease, ECOG PS 0 or 1, and no prior chemotherapy for MBC, received P (90mg/m2) on days 1, 8, 15 and B (10mg/kg) on days 1,15 every 28 days, plus V 200 mg (N=3) or 300 mg (N=28) BID orally for 3 days given the day before, day of, and day after each P dose. Response was assessed after every 3 cycles (RECIST). The trial was designed to distinguish between a response rate of < 40% vs. > 60% (alpha=0.10, beta=0.10) using Simon's optimal 2-stage design. At least 12 responses were required in the first 28 evaluable pts in the 1st stage and 21 of 41 in the 2nd stage. Results: There were no dose-limiting toxicities among the first 6 pts treated at the 1st (200 mg BID; N=3) and 2nd (300 mg BID; N=3) dose levels. Among the 26 evaluable pts treated at the recommended phase II dose of 300mg bid in the 1st stage, 14 patients had a partial response (PR-54%; 90% confidence interval [C.I.] 36%-71%), 6 had stable disease (SD) at 3 months and remain on treatment, 2 had SD as the best response with progressive disease (PD) at 11.8 and 15 months, and 4 had PD as their best response. Additionally, 2 of 3 pts treated at the first V dose level had a PR, indicating PR in 16 of 29 overall (55%, 90% C.I 38%-71%). Only 11 of 29 evaluable pts (38%) have progressed after a median followup of 10.5 months (R-2.3-14.9). Two pts had tumor biopsies and peripheral blood mononuclear cells (PBMC) collected before and 4 hours after the 3rd V dose (300 mg BID) which showed increased acetylation of K69 lysine residue of the chaperone protein Hsp 90, upregulation of Hsp70, and downregulation of AKT by Western Blot, providing evidence of Hsp90 inhibition in tumor and PBMC. Gr 3-4 toxicities included neutropenia (26%), neuropathy (23%), fatigue (16%), thrombosis (10%), vomiting (6%), hypertension (3%), and diarrhea (3%). Conclusions: Our preliminary findings indicate that the HDACi V inhibits Hsp90 and downregulates AKT in breast cancer in vivo, and enhances the effectiveness of P plus B in MBC. Accrual is nearly complete and the final efficacy analysis will be presented. Citation Information: Cancer Res 2009;69(2 Suppl):Abstract nr 404.

WITHDRAWN. 31 Loss of terminal fucose residue expression associated with Bombay phenotype influences plasma von Willebrand factor and factor VIII antigen levels TA MCKINNON*, MA LAFFAN*, AF RIDDELL, A HANN* and JS O’DONNELL* *Haematology Department, Hammersmith Hospital and Imperial College, London, UK, Katherine Dormandy Haemophilia Centre, Royal Free Hospital, London, UK ABO blood group exerts a major quantitative effect on plasma VWF level. Group O individuals have significantly lower VWF:Ag levels. Furthermore, ABH antigenic determinants are carried on the N-linked glycans of circulating VWF according to the blood group of the individual. The mechanism through which these glycans determine plasma VWF levels remains unknown. To further investigate this mechanism, we have collected plasma samples from 47 individuals with the rare Bombay blood group phenotype. These individuals fail to express alpha-1,2 fucosyltransferase, and thus cannot express A, B or O(H) antigens regardless of their ABO genotype. Using a modified ELISA technique, we confirmed that Bombay plasma VWF in all cases did not express ABO(H) determinants. We found VWF : Ag levels in Bombay patients (median VWF : Ag = 0.69 IU/dl) to be significantly lower than those observed in groups AB, A or B (median VWF : Ag = 1.24, 1.01 and 0.94 respectively; p < 0.05). Moreover, Bombay VWF : Ag levels were also lower than those in group O individuals (median VWF : Ag = 0.78), although this difference failed to achieve statistical significance. VWF : CB was also reduced in the Bombay group (median VWF : CB = 28 Elucidating the role of MLL in adult murine haematopoiesis K MCMAHON*, S HADJUR*, U MENZEL, D KIOUSSIS and HJM BRADY* *Molecular Haematology and Cancer Biology Unit, Institute of Child Health, University College London, London, UK, Molecular Immunology, MRC NIMR, London, UK The gene MLL has a fundamental role in haematopoietic stem cell development. Research so far has concentrated on the role of MLL in embryonic haematopoiesis as MLL homozygous knockout embryos die in utero. Studies in foetal liver, yolk sac cells and chimaeras have shown that lack of MLL leads to a reduction of cells in the myeloid, B and Tcell lineages. However, the role of a given gene in haematopoiesis may differ in the context of the embryo and the adult mouse as shown by studies on other embryonic lethal genes such as SCL. To study the effect of MLL deletion on adult haematopoiesis we have generated conditional knockout mice, carrying an allele of MLL bearing Lox-P recombination sites (‘floxed’) that will be recombined in the presence of the Cre recombinase, truncating the gene. When bred to mice carrying the Cre recombinase, the resulting offspring will lack MLL only in specified haematopoietic lineages, bypassing the embryonic lethality caused by Hox deregulation in other parts of the embryo. We have successfully generated chimaeras bearing the ‘floxed’ alleles which have been bred to create conditional knockouts. These mice are being analysed to determine the impact of the absence of MLL on adult haematopoietic stem cell development and renewal. 0.71 IU/dl) compared to other blood groups, but the ratio of VWF : CB to VWF : Ag remained unchanged. Furthermore, VWF multimer pattern in the Bombay plasmas demonstrated no loss of HMW multimers. ABO blood group also influences plasma FVIII : Ag levels. In Bombay individuals, we found FVIII : Ag levels to be significantly higher (median FVIII : Ag = 0.91 IU/dl) than corresponding VWF : Ag levels (median VWF : Ag = 0.69 IU/dl), resulting in an elevation of the FVIII : VWF ratio compared to other ABO groups (medians 1.32 and 1.0 respectively) However in VWF-FVIII binding ELISA, the FVIII–binding capacity of Bombay VWF was not increased. Our findings support the hypothesis that alterations in terminal carbohydrate moiety expression on plasma VWF-FVIII constitute important determinants of both VWF : Ag and FVIII : Ag levels. 32 A randomised control trial of patient selfmanagement of oral anticoagulation compared with patient self-testing C GARDINER*, KE WILLIAMS*, IJ MACKIE, SJ MACHIN and H COHEN* *Department of Haematology, University College London Hospitals, London, UK, Department of Haematology, UCL, London, UK Several studies suggest that patient self-management (PSM) may improve the quality of oral anticoagulant therapy (OAT) as measured by time in INR target range. Whether this improvement is due to PSM itself, or more frequent testing is unclear. We performed a randomised control study to determine whether the quality of treatment afforded by PSM is superior to that achieved by patient self-testing (PST) alone. 13% of eligible patients from our hospital anticoagulant clinic (receiving long-term OAT for >8 months) agreed to participate. 104 patients aged 22 to 88 years (median = 59.8) were randomised to PSM (n = 55) or patient self-testing (PST) (n = 49). Following satisfactory completion of a nurse-led training course, patients in both groups measured their INR using the CoaguChek S (Roche Diagnostics) every two weeks, or more frequently if required, for a period of six months. The PST patients telephoned their INR result to a nurse specialist for interpretation and adjustment of warfarin dose, whereas those in the PSM group adjusted their own warfarin dose on the basis of their INR using an algorithm provided by the clinic. 77/104 (74%) patients completed the study (PSM = 41, PST = 36). The ‘drop out’ rates for both groups were similar, with difficulty in obtaining an adequate capillary sample the most common reason given. There was no significant difference in median time in target therapeutic range between the two groups; PSM 71.8% (95% CI 39.7%–92.4%) and PST 70.3% (95% CI 41.3%–92.6%). We conclude that, in the majority of suitably trained patients, the quality of OAT achieved through PSM is comparable to that obtained by self-testing patients managed by a specialised hospital anticoagulation clinic. PSM is therefore an effective model for selected patients. 33 A thromboelastograph study on the effects of tissuetype plasminogen activator (t-PA), urokinase-type plasminogen activator (u-PA) and thrombin activatable fibrinolysis inhibitor (TAFI) on whole blood fibrinolysis SL HARRIS*, MJ GALLIMORE*, DW JONES*, K TAPPENDEN* and M WINTER* *Kent Haemophilia Centre, Kent and Canterbury Hospital, Canterbury, Kent, UK, Department of Biosciences, University of Kent at Canterbury, Canterbury, Kent, UK Thrombin activatable fibrinolysis inhibitor (TAFI) is a carboxypeptidase which has been shown to reduce t-PA-induced fibrinolysis in plasma and whole blood in vivo. It has been shown, however, that there is a significantly lower risk of myocardial infarction (MI) in individuals with elevated TAFI levels. Urokinase-type plasminogen activator (u-PA) was thought to be involved in extracellular plasminogen activation and not fibrinolysis. However, it has recently been speculated that u-PA may have a role in plasminogen activation via single chain u-PA (scu-PA) association with platelets (which carry a novel receptor for u-PA). In the present study, we investigated whether fibrinolysis induced in whole blood by u-PA and t-PA was affected by increasing levels of TAFI and whether u-PA-induced fibrinolyis was affected by TAFI in whole blood in a different manner to t-PA. We used the thromboelastography analyser (TEG) to measure clot formation and fibrinolysis in blood containing u-PA and t-PA in the presence of various concentrations of TAFI. Following clot formation, fibrinolysis was monitored. Various concentrations of u-PA and t-PA were added to whole blood samples from five donors and the samples subjected to TEG. Concentrations of u-PA and t-PA were selected to give approximately 50% lysis 60 minutes after clot formation. The addition of increasing levels of TAFI in samples containing t-PA gave reduced fibrinolysis in a dose dependent manner. However, the addition of increasing levels of TAFI in blood samples containing u-PA did not show a dose dependent effect and in some cases actually increased fibrinolysis. In some donors, TAFI with u-PA gave reduced coagulation. Our results suggest that, the interaction of TAFI and u-PA may cause a different response in platelet rich clots compared to platelet poor clots and may be a significant factor in the lower risk of MI in individuals with higher TAFI levels. 34 Bleeding complications of oral anticoagulant therapy in a single centre ES GREEN*, S BOND*, S RHODES*, M TAYLOR* and C EMMAS *Swindon and Marlborough NHS Trust, Swindon, UK, Astrazeneca, UK Bleeding is the most serious complication of the use of oral anticoagulation in the prevention and treatment of thromboembolism. We prospectively audited the frequency and severity of over anticoagulation and bleeding in our Anticoagulant service at a District hospital serving a population of 330 000 patients (3900 anticoagulant patients) providing Anticoagulant Practitioner-led inpatient and outpatient anticoagulant dosing. Data was collected over 23 months including INR at time of event, whether complications necessitated hospital attendance and/or admission, investigations carried out and treatment given and an assessment of the associated costs made. There were 443 events. 146 were asymptomatic high INRs (>8). The majority of these were treated with oral vitamin K as outpatients. There were 297 bleeding episodes, 73 major and 224 minor. The most common sites of bleeding were epistaxis (87/297), gastrointestinal (66/297), haematuria (43/297), haematoma (22/297) and bruising (20/297). 152 patients required hospital admission. 23 patients died, 5 as a result of the bleeding episode. 48/443 patients had received anticoagulant therapy for less than 4 weeks at the time of the event, 145/443 were unstable, 224/443 were stable and 26/443 were GP monitored patients with no INR data available. At the time of the bleeding event 48% of patients had an INR abov