James Cai

Preclinical Profile of a Potent γ-Secretase Inhibitor Targeting Notch Signaling with In vivo Efficacy and Pharmacodynamic Properties

Notch signaling is an area of great interest in oncology. RO4929097 is a potent and selective inhibitor of gamma-secretase, producing inhibitory activity of Notch signaling in tumor cells. The RO4929097 IC50 in cell-free and cellular assays is in the low nanomolar range with >100-fold selectivity with respect to 75 other proteins of various types (receptors, ion channels, and enzymes). RO4929097 inhibits Notch processing in tumor cells as measured by the reduction of intracellular Notch expression by Western blot. This leads to reduced expression of the Notch transcriptional target gene Hes1. RO4929097 does not block tumor cell proliferation or induce apoptosis but instead produces a less transformed, flattened, slower-growing phenotype. RO4929097 is active following oral dosing. Antitumor activity was shown in 7 of 8 xenografts tested on an intermittent or daily schedule in the absence of body weight loss or Notch-related toxicities. Importantly, efficacy is maintained after dosing is terminated. Angiogenesis reverse transcription-PCR array data show reduced expression of several key angiogenic genes. In addition, comparative microarray analysis suggests tumor cell differentiation as an additional mode of action. These preclinical results support evaluation of RO4929097 in clinical studies using an intermittent dosing schedule. A multicenter phase I dose escalation study in oncology is under way.

Discovering drug–drug interactions

Motivation: Identifying drug–drug interactions (DDIs) is a critical process in drug administration and drug development. Clinical support tools often provide comprehensive lists of DDIs, but they usually lack the supporting scientific evidences and different tools can return inconsistent results. In this article, we propose a novel approach that integrates text mining and automated reasoning to derive DDIs. Through the extraction of various facts of drug metabolism, not only the DDIs that are explicitly mentioned in text can be extracted but also the potential interactions that can be inferred by reasoning. Results: Our approach was able to find several potential DDIs that are not present in DrugBank. We manually evaluated these interactions based on their supporting evidences, and our analysis revealed that 81.3% of these interactions are determined to be correct. This suggests that our approach can uncover potential DDIs with scientific evidences explaining the mechanism of the interactions. Contact: luis.tari@roche.com

High tumor levels of IL6 and IL8 abrogate preclinical efficacy of the γ-secretase inhibitor, RO4929097

Interest continues to build around the early application of patient selection markers to prospectively identify patients likely to show clinical benefit from cancer therapies. Hypothesis generation and clinical strategies often begin at the preclinical stage where responder and nonresponder tumor cell lines are first identified and characterized. In the present study, we investigate the drivers of in vivo resistance to the γ‐secretase inhibitor RO4929097. Beginning at the tissue culture level, we identified apparent IL6 and IL8 expression differences that characterized tumor cell line response to RO4929097. We validated this molecular signature at the preclinical efficacy level identifying additional xenograft models resistant to the in vivo effects of RO4929097. Our data suggest that for IL6 and IL8 overexpressing tumors, RO4929097 no longer impacts angiogenesis or the infiltration of tumor associated fibroblasts. These preclinical data provide a rationale for preselecting patients possessing low levels of IL6 and IL8 prior to RO4929097 dosing. Extending this hypothesis into the clinic, we monitored patient IL6 and IL8 serum levels prior to dosing with RO4929097 during Phase I. Interestingly, the small group of patients deriving some type of clinical benefit from RO4929097 presented with low baseline levels of IL6 and IL8. Our data support the continued investigation of this patient selection marker for RO4929097 and other types of Notch inhibitors undergoing early clinical evaluation.

Discovery of siRNA Lipid Nanoparticles to Transfect Suspension Leukemia Cells and Provide In Vivo Delivery Capability

As a powerful research tool, siRNAs therapeutic and target validation utility with leukemia cells and long-term gene knockdown is severely restricted by the lack of omnipotent, safe, stable, and convenient delivery. Here, we detail our discovery of siRNA-containing lipid nanoparticles (LNPs) able to effectively transfect several leukemia and difficult-to-transfect adherent cell lines also providing in vivo delivery to mouse spleen and bone marrow tissues through tail-vein administration. We disclose a series of novel structurally related lipids accounting for the superior transfection ability, and reveal a correlation between expression of Caveolins and successful transfection. These LNPs, bearing low toxicity and long stability of >6 months, are ideal for continuous long-term dosing. Our discovery represents the first effective siRNA-containing LNPs for leukemia cells, which not only enables high-throughput siRNA screening with leukemia cells and difficult-to-transfect adherent cells but also paves the way for the development of therapeutic siRNA for leukemia treatment.

Identifying Novel Drug Indications through Automated Reasoning

Background With the large amount of pharmacological and biological knowledge available in literature, finding novel drug indications for existing drugs using in silico approaches has become increasingly feasible. Typical literature-based approaches generate new hypotheses in the form of protein-protein interactions networks by means of linking concepts based on their cooccurrences within abstracts. However, this kind of approaches tends to generate too many hypotheses, and identifying new drug indications from large networks can be a time-consuming process. Methodology In this work, we developed a method that acquires the necessary facts from literature and knowledge bases, and identifies new drug indications through automated reasoning. This is achieved by encoding the molecular effects caused by drug-target interactions and links to various diseases and drug mechanism as domain knowledge in AnsProlog, a declarative language that is useful for automated reasoning, including reasoning with incomplete information. Unlike other literature-based approaches, our approach is more fine-grained, especially in identifying indirect relationships for drug indications. Conclusion/Significance To evaluate the capability of our approach in inferring novel drug indications, we applied our method to 943 drugs from DrugBank and asked if any of these drugs have potential anti-cancer activities based on information on their targets and molecular interaction types alone. A total of 507 drugs were found to have the potential to be used for cancer treatments. Among the potential anti-cancer drugs, 67 out of 81 drugs (a recall of 82.7%) are indeed known cancer drugs. In addition, 144 out of 289 drugs (a recall of 49.8%) are non-cancer drugs that are currently tested in clinical trials for cancer treatments. These results suggest that our method is able to infer drug indications (original or alternative) based on their molecular targets and interactions alone and has the potential to discover novel drug indications for existing drugs.

MDM2 antagonist clinical response association with a gene expression signature in acute myeloid leukaemia.

Acute myeloid leukaemia (AML) is uniquely sensitive to p53 activation 1, 2 as ≈90% of patients carry wild-type TP53 and frequent MDM2 overexpression.3 MDM2 blocks p53 transactivation and targets p53 for ubiquitin-dependent degradation.4, 5 Nutlins have been characterized as potent and selective small-molecule MDM2 antagonists.1, 6–8 RG7112 was the first such MDM2 antagonist to undergo clinical assessment in solid tumors and leukaemia trials.1, 2, 9 As not all patients with functional p53 will respond to MDM2 antagonists, diagnostic tools may identify patients likely to respond.

QUADS: question answering for decision support

As the scale of available on-line data grows ever larger, individuals and businesses must cope with increasing complexity in decision-making processes which utilize large volumes of unstructured, semi-structured and/or structured data to satisfy multiple, interrelated information needs which contribute to an overall decision. Traditional decision support systems (DSSs) have been developed to address this need, but such systems are typically expensive to build, and are purpose-built for a particular decision-making scenario, making them difficult to extend or adapt to new decision scenarios. In this paper, we propose a novel decision representation which allows decision makers to formulate and organize natural language questions or assertions into an analytic hierarchy, which can be evaluated as part of an ad hoc decision process or as a documented, repeatable analytic process. We then introduce a new decision support framework, QUADS, which takes advantage of automatic question answering (QA) technologies to automatically understand and process a decision representation, producing a final decision by gathering and weighting answers to individual questions using a Bayesian learning and inference process. An open source framework implementation is presented and applied to two real world applications: target validation, a fundamental decision-making task for the pharmaceutical industry, and product recommendation from review texts, an everyday decision-making situation faced by on-line consumers. In both applications, we implemented and compared a number of decision synthesis algorithms, and present experimental results which demonstrate the performance of the QUADS approach versus other baseline approaches.

Mining gene-centric relationships from literature: the roles of gene mutation and gene expression in supporting drug discovery

Identifying drug target candidates is an important task for early development throughout the drug discovery process. This process is supported by the development of new high-throughput technologies that enable better understanding of disease mechanism. It becomes critical to facilitate effective analysis of the large amount of biological data. However, with much of the biological knowledge represented in the literature in the form of natural text, analysis and interpretation of high-throughput data has not reached its potential effectiveness. In this paper, we describe our solution in employing text mining as a technique in finding scientific information for target and biomarker discovery from the biomedical literature. Our approach utilises natural language processing techniques to capture linguistic patterns for the extraction of biological knowledge from text. Additionally, we discuss how the extracted knowledge is used for the analysis of biological data such as next-generation sequencing and gene expression data.

Abstract 2835: MDM2 antagonist-based therapeutic response is discriminated by a 4-gene signature in acute myeloid leukemia patients

The activity of p53, a key tumor suppressor is tightly controlled by MDM2-mediated ubiquination and degradation. Nutlins, a class of small-molecule MDM2 antagonists, have been characterized as drivers of p53 re-activation. Acute myeloid leukemia (AML) is uniquely sensitive to p53 re-activation as ∼90% of cases have wild-type TP53 and frequent MDM2 overexpression to overcome mechanisms of oncogene addiction. Personalized theranostic strategies may distinguish patients likely to clinically benefit from MDM2-antagonist therapy. Association between MDM2 antagonist (RG7112) growth inhibition (IC50s) in 287 human cancer cell lines (Cell Lines for Oncology/Chugai Accumulative Tumor Encyclopedia), and pretreatment RNAseq profiling established a classifier comprising MDM2, XPC, BBC3, and CDKN2A. This signature significantly associated with cell-line efficacy to MDM2 antagonist (odds ratio = 2.53; P RG7112 treatment was assessed in a phase 1 dose escalation trial in relapsed/refractory AML patients (NO21279). Signature scores of AML patient blood specimens at baseline significantly associated with clinical response (PD In summary, we demonstrate that a biological classifier discriminates response broadly to MDM2-antagonist therapy. The level of evidence attained by cell line efficacy modeling and response assessments in trial NO21279 (with MDM2 antagonist RG7112) and now in trial NP28679 (with MDM2 antagonist RG7388) adds substantial weight to the validity of this panel. Citation Format: Hua Zhong, Gong Chen, Lori Jukofsky, David Geho, Sung Won Han, Fabian Birzele, Sabine Bader, Lucia Himmelein, James Cai, Zayed Albertyn, Mark Rothe, Laurent Essioux, Helmut Burtscher, Steven A. Middleton, Lin-Chi Chen, Markus Dangl, William E. Pierceall, Gwen Nichols. MDM2 antagonist-based therapeutic response is discriminated by a 4-gene signature in acute myeloid leukemia patients. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 2835. doi:10.1158/1538-7445.AM2015-2835

Abstract A41: Pre-clinical investigation of predictive biomarkers for drug candidates using a molecularly profiled large cancer cell line panel.

To deliver investigational drugs to appropriate patients from the early clinical development stage, it would be ideal to identify biomarker candidates at the preclinical stage. To investigate prediction biomarkers in preclinical models, a large cell panel consisting of over 500 cell lines from various tumor types was molecularly profiled. To find their gene expression profiles, gene copy number variations, single nucleotide variations and gene fusions, we conducted a comprehensive analysis with GeneChip, CGH array, exon sequencing and RNA sequencing by collaborative efforts between Chugai Pharmaceutical and F. Hoffmann-La Roche in the CELLO/CACTEL program (CELl Line profiling in Oncology/Chugai ACcumulative Tumor EncycLopedia). We further integrated the multidimensional profiling data of in-house cell lines together with those of tumor tissues in the public domain into our CELLO/CACTEL database by vocabulary controlling among databases. Using 300 cell lines in a molecularly profiled cell panel, we investigated the drug sensitivity profiles of representative investigational drug candidates, and identified potentially novel predictive features for sensitivity to these drugs. In summary, the CELLO/CACTEL database provides an effective platform for exploring biomarkers at the preclinical stage. Citation Information: Mol Cancer Ther 2013;12(11 Suppl):A41. Citation Format: Satoshi Aida, Hideaki Mizuno, Hiroshi Sakamoto, Yoshito Nakanishi, Hironori Mutoh, James Cai, Helmut Burtscher, Masahiro Aoki, Yuko Aoki, Nobuya Ishii. Pre-clinical investigation of predictive biomarkers for drug candidates using a molecularly profiled large cancer cell line panel. [abstract]. In: Proceedings of the AACR-NCI-EORTC International Conference: Molecular Targets and Cancer Therapeutics; 2013 Oct 19-23; Boston, MA. Philadelphia (PA): AACR; Mol Cancer Ther 2013;12(11 Suppl):Abstract nr A41.