«Prev‘Open for Innovation’ outlines post-Brexit path NextOpen Access Journals in the 2015 Journal Citation Process»

Stream the presentation, "Reaching Beyond the Data to Prevent Target Identification Failure"

Over the last several years, when planning the best approach for the earliest stages of the drug discovery process, the research community has witnessed a back and forth between phenotypic and target-based screening.

Phenotypic drug discovery (PDD) looks at the effects (or phenotypes) that compounds induce in cells, tissues or whole organisms. Target-based drug discovery (TDD) measures the effect of compounds on a purified target protein via in vitro assay (involving in many cases screening large libraries of chemical compounds in automated high-throughput cellular assays that measure the levels of various proteins or effects on characteristics such as cell proliferation).

Beginning in the 1980s and especially in the 1990s, advances in molecular biology and genomics led to phenotypic screens largely being replaced by screens against defined targets implicated in disease. A study by Swinney et al.1 reported that among the 183 small-molecule drugs across all therapeutic areas approved between 1999 and 2008, 58 (32%) were discovered using phenotype-based approaches. Importantly, 28 (56%) of the 50 small-molecule, first-in-class new molecular entities (NMEs) identified in the study resulted from phenotypic screening approaches, whereas 17 (34%) resulted from target-based approaches. A more recent article by Eder et al. reports some updated data: Of the 113 first-in-class drugs approved by the FDA between 1999 and 2013, target-based research approaches accounted for 70% of the new drugs2. Focusing on TDD in the absence of validated molecular mechanisms of action was suggested to be a major technical factor contributing to the high attrition rates of small-molecule, first-in-class medicines observed in the clinic3.

Nevertheless, phenotypic screening is making resurgence in drug discovery, as some researchers consider that reductionist approaches such as target-based screening are useful but may also limit the breadth of new findings. However, returning to phenotypic screening practices may not be the solution for all drug development concerns. It is still a challenge to find the target or targets that are being hit by the disease-modifying candidate drug molecule (target deconvolution) for understanding compound mechanisms of action and a number of drug discovery groups prefer to plan their research focused on target-based screening.

Despite the novelty of phenotypic “hits” and their use as tools for further dissection of a given biological process, the majority of compounds are unlikely to progress as leads or drug candidates for several reasons such as promiscuous modes of action and their frequent presence in screening campaigns. Some of these molecules can even be considered artifacts of high throughput screening4 .

Additionally, target-based approaches are often simpler to execute than phenotypic assays, and provide the knowledge that the result seen involves a specific genetic target. Phenotypic assays generally take more effort and resources, but hopefully the results are more directly related to the disease state.

On the basis of the above, it can easily be stated that there is no one-size-fits-all drug discovery solution. For every unique drug discovery challenge, its own unique solution is often required. As much as it has become necessary to look at integrated systems during drug discovery, knowing the target related to the drug activity is essential. Therefore combining the two approaches gives the greatest opportunity to understand the broad mechanism of drug action and be able to discover targets with new therapeutic potential. Hence, as more integrative technologies become available, the focus of the discussion should shift from prioritizing the different approaches to finding strategies that can combine their complementary strength5. Namely, Moffat et al.3, propose that in practice a considerable proportion of cancer drug discovery falls between pure phenotypic drug discovery and target-based drug discovery, in a category that they identify as “mechanism-informed phenotypic drug discovery” (MIPDD). This category includes inhibitors of known or hypothesized molecular targets that are identified and/or optimized by assessing their effects on a therapeutically relevant phenotype, as well as drug candidates that are identified by their effect on a mechanistically defined phenotype or phenotypic marker and subsequently optimized for a specific target-engagement mechanism of action.

About targets information and opportunities

Regardless of the selected approach to start the research, targets still have a leading role, not only in the process of new drug discovery, but also in exploring other indications for known drugs – the so called “indications discovery.” William T. Loging at Mount Sinai stated in an interview last year6 that opening up other indication areas for known targets (or drugs) could potentially lead to millions of dollars in increased revenue, hence the need for understanding new opportunities in this space.

Returning to the topic of new drug discovery process, Richard Harrison, Chief Scientific Officer of Clarivate Analytics, recently published an analysis7 to understand the reasons for clinical failure, essential for decreasing attrition in clinical development, concluding that efficacy is still the greatest reason for failure (Figure 1), but we are hopeful that this can change as trial populations are increasingly stratified using efficacy biomarkers and understanding of disease biology and target selection improve.

FIGURE 1:  Reasons for clinical trial failures 2013−2015. The pie charts illustrate the reason for failure for phase II and phase III trials for which a reason for failure was reported for all 174 clinical trials (part a), according to therapeutic area (part b), in all therapy areas for phase II only (part c) and in all therapy areas for phase III only (part d). Data are from Clarivate Analytics and Drugs of Today.

 

When it comes to choosing the right target for a new drug discovery project, researchers at pharma companies and research institutions rely not only on the experimental or in silico data gathered in house, often in collaboration with other institutions. Researchers also need their decisions to be strongly supported by the analysis of the key findings published by others that will enable them to balance the scientific validity of a target with factors such as the number of other companies developing drugs for said target, the intellectual property around them, the development stages of said drugs, the availability of animal models for the preclinical testing and the potential for patient stratification. Understanding these factors presents multiple challenges but also presents opportunities for the rapid development of first- and/or best-in-class treatments.

Given the importance of this knowledge generating process, it is crucial to make the right choice and the right use balance of public data and curated data with the right equilibrium of volume versus quality. Both medicinal chemists and biologists seek high quality chemistry and biology data to support their project portfolio. In either case, it should be noted that commercial databases rely on high curation quality of largely manually extracted data, with the assistance of software tools. The public domain resources, however, beyond their submission filtration pipelines, are dependent on the quality of depositing sources. Multiple reviews of public domain data sources indicate that, in the main, data quality issues arise that are independent of the submitter.8

In relation to the competition, it is important to consider how current crowded markets have led to high-risk research. Selecting a new target for research requires researchers to genuinely differentiate from the competition and find a novel niche at the very start of the process. In our recently published article9 , we reviewed how the target landscape has changed in the past 3 years, concluding that competition for novel targets (targets with no drugs under active development that have launched, registered or recommended approval stages) has substantially increased over the past three years and competition for proven targets remains high. Other data from the study includes the number of targets with only one company pursuing them, which has decreased by 40% since 2013 with a similar percentage increase of targets with five or more companies pursuing them, indicating a substantial increase in competition.

In addition to competitive intelligence, another important aspect for the target validation process is to understand the chemical, biological and pharmacological evidence published around the biological entity (gene, protein, RNA) which will provide guidance on the likelihood of that entity to succeed in the process of drug discovery and development. Furthermore researchers need to rapidly evaluate the target landscape around an indication or pathway.

With review of the aspects above, it is easy to infer how crucial it is to spend the right time and effort at the early stages of any drug discovery research planning to avoid drug-development failures later in the pipeline due to a wrong target selection, not only by validating it in the lab but also by considering and understanding the challenges and opportunities around it in a holistic manner in order to reach fast and well-informed decisions.

Target Druggability in action

At Clarivate Analytics, we are developing workflow-based solutions under the Drug Research Advisor solutions strategy. Target Druggability (Figure 2) is the first application in the Drug Research Advisor suite, and has been developed to support precision in the target selection process. Information found in Target Druggability brings together trusted scientific content from highly regarded resources: Integrity and MetaCore.

FIGURE 2:  from Target Druggability: Representation of the target landscape for a given condition, target view page relating the gene/protein with different conditions and a pathway indicating related targets upstream and downstream from the target of interest

An example of Target Druggability in action is illustrated in this case study on Crohn’s disease, with a proposal of a new putative target to explore for developing a new drug for this condition.

There may be different scenarios depending on users’ expertise in the area. Target Druggability can guide users new in the therapeutic area to focus on those targets most likely to succeed first by using graphic visualizations to interact with complex information, or by using the target prioritization algorithm that will rank the targets based on druggability and novelty criteria. Nevertheless, if users have some previous context in the area, they might be aware that several studies have reported the importance of T Helper 17 (Th17) cells in the pathogenesis of Crohn’s disease as mediators of intestinal inflammation. Interleukin-23 (IL-23) is key for the differentiation of Th17 lymphocytes; therefore, exploration of the interleukin-23 signaling pathway may lead to the identification of novel targets for the treatment of this inflammatory disease.

To learn more about themes covered in this article, please watch, “Reaching Beyond the Data to Prevent Target Identification Failure” or contact the author, Montse del Fresno, PhD, Product Manager – Preclinical Solutions. Watch now.

We also invite you to review a presentation given at the Molecular Medicine Tri-Con 2017 (MMTC): “New approaches to Target Identification,” by Dr. Richard Harrison, Chief Scientific Officer, Clarivate Analytics.

To learn more about Drug Research Advisor – Target Druggability, click here.

References:
1 Swinney, D. C. et al. Nature Rev. Drug Discov. , 2011; 10: 507
2Eder, J. et al. Nature Rev. Drug Discov., 2014; 13(8): 577
3Moffat, J.G. et al. Nature Rev. Drug Discov., 2014; 13: 588
4Cox, J.A.G. et al. Sci Rep., 2016; 6: 38986.
5Hassan, A.A., Med. Chem. Commun., 2016; 7: 788
6http://stateofinnovation.thomsonreuters.com/computational-tools-help-con...
7Harrison, R.K., Nature Rev. Drug Discov. , 2016; 15: 817
8Lipinski, C.A. et al. J. Med. Chem., 2015; 58(5): 2068; Williams, A.J. Drug. Discov. Today, 2012;17(13-14): 685
9Lafferty-Whyte, K. et al. Nature Rev. Drug Discov. , 2017; 16: 10
10Eken, A. et al. Inflamm. Bowel Dis., 2014; 20(3): 587