Phenotypic screening in oncology drug discovery – Hosein Kouros-Mehr

A recent paper* summarized the history of oncology drug discovery and analyzes the contibutions of phenotypic drug discovery vs traditional target-based drug discovery.  The authors analyze the development and mechanisms of all small molecule-based cancer drugs approved by the FDA over the past 15 years and those currently in clinical development.


The authors show that the majority of small molecule inhibitors for oncology originated in target-based discovery.  However, a significant number of inhibitors were identified in phenotypic drug screening approaches.  The authors suggest that the rate-limiting step in bringing compounds identified in phenotypic screenings into clinical development is the lack of mechanistically defined cellular models for the cancer phenotypes in question and also the reliance on traditional drug effects such as cytotoxicity and mitotic arrest, which are only a component of the hallmarks of cancer.  


The authors suggest that mechanistically informed phenotypic models may better enable compounds identified in phenotypic screenings to successfully complete clinical development.  These models would enable confirmation that the targeted agents have a molecular mechanism of action, which would enable PD biomarker development and development of diagnostic hypotheses and patient tailoring hypotheses.

* Moffat, et. al. (2014) Phenotypic screening in cancer drug discovery — past, present and future.  Nat Rev Drug Discovery, doi:10.1038/nrd4366

Hosein Kouros-Mehr

New study links cholesterol, Wnt signaling and cancer – Hosein Kouros-Mehr

A recent paper* has reported a direct biological link between cholesterol and cancer.  The Wnt pathway plays an important role in cancer formation and metastasis, particularly in colorectal cancer in which the Wnt-related gene APC is frequently mutated.  In this report, the authors find that cholesterol is enriched around the Wnt-related Frizzled and LRP5/6 receptors in cancer cells.  This suggested to the authors that perhaps cholesterol itself could activate Wnt signaling.  The authors found that cholesterol directly recruits the Wnt-related Dishevelled (Dvl) scaffold protein to the cell membrane through interaction with its PDZ domain.  Dvl then activates the canonical Wnt signaling pathway.  In this way, cholesterol may directly regulate activation of canonical Wnt signaling and may regulate balance between canonical and non-canonical signaling.  Future studies will be needed to determine effects of anti-cholesterol medicines in silencing Wnt signaling and perhaps inhibiting Wnt-driven cancers.


* ​Sheng, R., et. al. (2014).  Cholesterol selectively activates canonical Wnt signalling over non-canonical Wnt signalling.  Nature 5 (4393)

Hosein Kouros-Mehr

RUNX1 as a biomarker for triple-negative breast cancer – Hosein Kouros-Mehr


With the advent of microarray expression profiling over a decade ago, many new biomarkers have been identified for ER+ breast cancers, such as the transcription factors GATA-3 and XBP-1.  A recent paper* identified a novel biomarker for triple-negative (ER- PR- HER2-) breast cancer.  RUNX1 is a transcription factor that is the most frequently mutated gene in human leukemia, and the RUNX family plays essential roles in haematopoiesis, osteogenesis and neurogenesis.


In this report, the authors utilized a tissue microarray continaing biopsies from 483 patients with invasive ductal breast adenocarcinoma.  The microarray was stained with antibodies for immunihistochemistry.  RUNX1 immunostaining was signficantly asociated with pooere cancer-specific survival in patients with ER-negative and triple-negative breast cancer.  However, RUNX1 was associated with progesterone receptor positive tumors as well, which is a bit confounding.  Interestinlgly, RUNX1 was associated with more CD4+ and CD8+ T-lymphocyte infiltration and CD68+ macrophage infiltration, which have been observed as markers for poor prognosis in breast cancer patients.  


Triple negative breast cancer is an unmet medical need and lacks suitable biomarkers for patient stratification.  RUNX1 may be used as a biomarker and prognostic indicator correlating with poor prognosis specifically in the triple negative subtype of human breast cancer.


Hosein Kouros-Mehr

Retrospective analysis of a pharmaceutical company’s R&D pipeline – Hosein Kouros-Mehr

A major pharmaceutical company has published* a thorough analysis of its R&D pipeline from 2005-2010 and summarized the primary reasons for successes and failures of its small-molecule drug projects.  While R&D investment in the industry has reached record high levels and there has been a wealth of new technologies and new drug targets to explore, the rate of new drug launches has been steady and drug development costs have increased substantially.  The conclusions of this article can be used to boost productivity of drug development efforts in hopes of increasing the number of successful drug launches.  
An interesting point made in the paper is that volume-based metrics to boost portfolio projects and drug candidates have not necessarily yielded better performance.  The authors state, “This volume-based approach damaged not only the quality and sustainability of R&D pipelines but, more importantly, also the health of the R&D organizations and their underlying scientific curiosity. This is because the focus of scientists and clinicians moved away from the more demanding goal of thoroughly understanding disease pathophysiology and the therapeutic opportunities, and instead moved towards meeting volume-based goals and identifying an unprecedented level of back-up and ‘me too’ drug candidates. In such an environment, ‘truth-seeking’ behaviours to understand disease biology may have been over-ridden by ‘progression-driven’ behaviours that rewarded scientists for meeting numerical volume-based goals.” The authors suggest that volume-based metrics should be substituted with a more in depth understanding of drug targets, biologies, and patient selection metrics.  
The company describes a comprehensive review undertaken for 142 drug discovery and development projects from candidate phase to Phase II.   Data were gathered from more than 80% of these 142 projects and for 95% of projects in clinical phases.  Of the projects analysed, 94 closed during the period assessed; 33 closed before clinical testing and a further 61 closed during clinical testing. The review was performed by submitting structured questionnaires to a cross-functional group of scientists and clinicians drawn from the project teams.  The primary cause of failure for projects up to Phase II was unacceptale safety, which accounted for more than half of all project closures.  The majority of these failures occurred before clinical testing (primarily during regulatory GLP toxicology testing), with safety issues being the reason for 82% of preclinical project closures.  The analysis suggested a  crucial need for teams to pay attention to preclinical safety signals.  
Based on the analysis, the paper concludes that there 5 variable that predict success for an R&D portfolio:
(1) Right target – Strong link between target and disease; differentiated efficacy; available and predictive biomarkers
(2) Right tissue – Adequate bioavailability and tissue exposure; definition of PD biomarkers; clear understanding of preclinical and clinical PK/PD; understanding of drug-drug interactions
(3) Right safety – DIfferentiated and clear safety margins, understanding of secondary pharmcology risk, understanding of reactive metabolites, genotoxicity, drug-drug interactions, understanding of target liability
(4) Right patients – Identification of patient population for tailoring of molecules, definition of risk-benefit for a given population
(5) Right commercial potential – Differentiated value proposition versus future standard of care, focus on market access (payer, provider), personalized health care strategy (diagnostic, biomarkers)


Cook D et. al., , (2014). Lessons learned from the fate of AstraZeneca’s drug pipeline: a five-dimensional framework Nature Reviews Drug Discovery 13, 419–431 (2014)


Hosein Kouros-Mehr

Gene amplifications in cancer TCGA datasets – Hosein Kouros-Mehr

The Cancer Genome Atlas (TCGA) projects have advanced our understanding of the driver mutations, genetic backgrounds, and key pathways that drive cancer types. We recently published* our efforts to identify genes that are commonly amplified in cancers and display a cancer driver signature.  Put in another way, we devised a bioinformatics screening strategy to identify putative cancer driver genes amplified across TCGA datasets
We carried out a GISTIC bioinformatics analysis of TCGA datasets spanning 16 cancer subtypes and identified 486 genes that were amplified in two or more datasets.  These cancer types include BLCA – Bladder Urothelial Carcinoma, BRCA – Breast invasive carcinoma, CESC – Cervical squamous cell carcinoma and endocervical adenocarcinoma, CRC – Colorectal Cancer (COAD and READ studies combined together), GBM – Glioblastoma multiforme, HNSC – Head and Neck squamous cell carcinoma, KIRC – Kidney renal clear cell carcinoma, LGG – Brain Lower Grade Glioma, LUAD- Lung adenocarcinoma, LUSC –
Lung squamous cell carcinoma, OV – Ovarian serous cystadenocarcinoma, PAAD – Pancreatic adenocarcinoma, PRAD – Prostate adenocarcinoma, SKCM – Skin Cutaneous Melanoma, STAD – Stomach adenocarcinoma, UCEC – Uterine Corpus Endometrioid Carcinoma.
From the 486 genes, we identified 75 cancer-associated genes with potential “druggable” properties. The majority of the genes were localized to 14 amplicons spread across the genome.  Genes within an amplicon tended to be amplified in the same cancer subtypes.  To further identify potential cancer driver genes, we analyzed gene copy number and mRNA expression data from individual patient samples and identified 42 putative cancer driver genes linked to diverse oncogenic processes. Oncogenic activity was further validated by siRNA/shRNA knockdown and by referencing the Project Achilles datasets.
The amplified cancer driver genes represented a number of gene families, including epigenetic regulators, cell cycle-associated genes, DNA damage response/repair genes, metabolic regulators, and genes linked to the Wnt, Notch, Hedgehog, JAK/STAT, NF-KB and MAPK signaling pathways. Among the 42 putative driver genes were known driver genes, such as EGFR, ERBB2 and PIK3CA. Wild-type KRAS was amplified in several cancer types, and KRAS-amplified cancer cell lines were most sensitive to KRAS shRNA, suggesting that KRAS amplification was an independent oncogenic event. A number of MAP kinase adapters were co-amplified with their receptor tyrosine kinases, such as the FGFR adapter FRS2 and the EGFR family adapters GRB2 and GRB7. The ubiquitin-like ligase DCUN1D1 and the histone methyltransferase NSD3 were also identified as novel putative cancer driver genes.
The data presented can be used for patient tailoring efforts — in other words, to tailor novel therapeutics to the patients whose cancers contain the genetics drivers in question and would benefit most from the targeted therapy.  The data can also be used to identify potential novel opportunities for drug discovery efforts.

Hosein Kouros-Mehr


Chen YMcGee JChen XDoman TNGong XZhang YHamm NMa XHiggs REBhagwat SVBuchanan SPeng SBStaschke KA,Yadav VYue YKouros-Mehr H. (2014)  Identification of Druggable Cancer Driver Genes Amplified across TCGA Datasets.  PLoS One. 2014 May 29;9(5):e98293. doi: 10.1371/journal.pone.0098293. eCollection 2014.