Shared article

Drug Discovery and Development

1. Introduction

Drug development is a lengthy process, involving many difficulties and costs. Successful drugs take at least ten years from conception to clinic, costing upwards of $2.6 billion. Clinical trials and drug approval alone can cost $1 billion. There is a high attrition rate which causes around 90% of the drugs to fail after the first clinical trials. Drug safety tests have a great impact on drug development leading to drug withdrawals and restricted use policies. Therefore, to avoid highly expensive late-stages failures, toxicology assays have been developed, which aim to identify drugs that are highly probable to fail at the early stages (“fail-early” strategy). Another approach was later developed to help identify potential drugs with no safety liabilities and is used in conjugation with the “fail-early” strategy. However, using in vitro, ex vivo and in vivo tests it is possible to avoid testing everything during the early phase.

Therefore, the high attrition rate caused by the approval of only a few medicines out of the thousands and sometimes millions of compounds that have been tested during the early Research and Development (R&D) stages explains the high cost of a newly developed medicine. 


2.1. Drug discovery and development

Drug development strategies can be divided into several stages: basic research, drug discovery, drug development, FDA review and Clinical Use (Figure 1). This article will review each of the elements and will focus on drug discovery and development processes. 

Figure 1: The pipeline of drug development

2.2. Basic research- target identification

The initial stages of drug development involve identifying potential pathways or protein involved in disease. Identification of these targets is a challenge, as cellular pathways are complex with many regulatory mechanisms and may be involved in key cellular functions. As a result, modification of certain pathways may have extreme toxic effects and be unsuitable for further drug development. A variety of biological entities can be used as a drug target: genes, RNA or proteins. However, to become a good target, the molecule should be “druggable”, for instance, be biologics or a small molecule and express a biological effect upon its binding. Moreover, it should be safe, efficacious and pass clinical and commercial requirements.

The number of promising targets to be identified has significantly increased due to the development of data mining. For instance, employing bioinformatics not only helps to identify potential targets, but also prioritises these for a particular condition. Alternatively, looking into genetic association with diseases can bring some hints into the link between the condition. For example, a specific mutation, like in Alzheimer's Disease (AD) where alterations in the presenilin genes lead to the development of this neurodegenerative disease. Finally, the phenotypic screening  can be used, which works by identifying antigens (through the monoclonal antibodies) that are specific to tumour cells.

After the target is selected, further validation is done which brings to the lead discovery stage, where a promising ”druggable” molecule is selected which if successful, might progress into clinical trials and the market.


2.3. Drug discovery

Following on from target identification, compounds can be screened during "drug discovery". The process of drug discovery includes several steps (Figure 2) and can last in excess of ten years.

Figure 2: The process of drug discovery.

2.3.1.  Target validation

Promising targets are evaluated using a variety of validation techniques. Many approaches are used at this stage, to increase the confidence in the target and justify future costs of drug development.  

Example techniques may include genetic manipulation of target genes. This may involve knocking down the gene (shRNA, siRNA), knocking out the gene (CRISPR) or knocking in the gene (viral transfection of mutant genes). All these techniques can be used in vitro at lower costs. Following these experiments, the results may be validated using in vivo models, looking at the global effects of knock outs or mutant genes. Using tissue specific promoters, it can be possible to restrict knock outs to certain organs. For example, the MMTV promoter can be used to target mammary gland.

Monoclonal antibodies can also be used to validate the target. These antibodies selectively bind to their target with high affinity and block further interactions. However, as the monoclonal antibodies cannot pass through cell membranes, these are restricted for the use on secreted and surface proteins.

Finally, chemical genomics which determines the effect of chemical compounds on the genomic responses brings great promises to the future of target identification and validation as it aims to provide chemical approaches against every genome encoding proteins.


2.3.2. Hit discovery

During the hit identification and discovery phase the screen assays are developed. A “hit” compound is the one, which has the required activity in the screen and this is validated upon repeat runs. A variety of screening strategies are employed by the pharmaceutical companies to identify the target molecules. For instance, in the high-throughput screen (HTS), a compound library (large number of compounds) are screened against the drug target. NMR screen can be also employed to determine the structure of the molecules. In a series of activities, different chemistry programmes are used in parallel to improve the safety, efficacy and physicochemical properties of the target compound. Finally, the data still continues to be acquired which confirms the theory of the efficacy at the disease stage upon the drug intervention at the target site.


2.3.3. Assay development

A great variety of assays can be used to aid the compound screening. The choice depends on several factors, such as the biology of the drug target protein, the mechanism of action (activator/ inhibitor) of a potential drug and the available equipment. Cell-based assays normally provide the results for the functional activity of the biologics, such as ion channels or nuclear receptors. In contrast, biochemical assays are focused on determining affinity between the test compound and target protein and can be applied for receptor and enzyme targets.

Nevertheless, regardless the assay format, the following principles should be met:

  1. Pharmacological relevance of the assay; e.g.: assay performed on the known ligand with the molecule under investigation and to demonstrate the required potency and mechanism of action
  2. Reproducibility: the assay is reproducible across the plates and duration (e.g.: the duration of the entire discovery program)
  3. Minimized cost, e.g.: by employing 384 instead of 96-well screen.
  4. Quality. The Z’ factor measures a signal window and the variance around maximal and minimal signals in the assay. The Z’ ranges from 0 to 1 and a “good” Z’ value is above 0.6. 

Secondly, the controls should be present in the assays and their limits define the range into which test compounds will fall.

  1. Effect of compounds in the assay. Assays cannot be sensitive towards concentrations of solvents (e.g.: DMSO) in which compounds are dissolved.


 2.3.4. Defining a hit series

Drug-like molecules must obey certain chemical parameters to be considered. Typically, the Lipinski Rule of  Five is followed:

  1. molecular weight ≤400 and 
  2. logP (lipophilicity measurement) ≤4
  3. No more than 5 hydrogen bond donors
  4. No more than 10 hydrogen bond acceptors

After the HTS has generated the compounds’ hits, the next step is to define the most “promising” molecules. Firstly, compounds which are considered as common hitters in HTS are removed. Also, compounds should represent a broad variety of chemical families. Then, for each hit in the primary assay the dose-response curves are generated which define the potency of the candidate compounds through a half maximal inhibitory concentration. The next step is to establish the surviving hits in the secondary assay for the target. For instance, cell-based assays of the target's functional response. The advantageous compounds would be those that can undergo SAR (structure-activity relationship), i.e.: compounds that share a common chemical motif and upon addition of different functional groups change their potencies. Several groups of compounds now can be examined for identification of the vital elements in their structure and its relationship to activity. Alongside this, in vitro assays should be performed which provide the information on the pharmacokinetic and physicochemical measurements by generating data on the absorption, distribution, metabolism and excretion (ADME) of the compound.


2.3.5.   Hit-to-lead phase

During this stage the hit series are improved to produce more efficacious compounds and potency can be observed using in vivo models. Different assays are performed which measure the activity of the potential drug at the predicted sites and a deeper  attention is put into physiochemical and in vitro ADME properties of the key selected compounds. The future drugs should also get a “green light” in the solubility and permeability assessments. For instance, Hepg2 hepatox assay can identify potential drugs that can cause human liver toxicity or MDR1 permeability assay is performed to test drug’s intestinal and brain permeability. Then the compounds, which meet physicochemical and ADME requirements as well as have the target potency and selectivity properties, undergo pharmacokinetic assessments in rats (Hughes et al).


2.3.6.  Lead optimisation

This is the final stage of the drug discovery process which aims to improve the drawbacks in the lead structure, while maintaining its favourable biochemical properties. At this stage, the compounds meet the requirements of the lead optimisation phase and soon can be forwarded to the pre-clinical trials. However, meanwhile the research work is still ongoing to discover the back-up molecules in case the drug undergoing pre-clinical phase fails. Dependent on the company, a variety of tests are performed. For instance, the Ames test which detects the compounds that have the potential to cause DNA mutations. All this information, together with the data obtained from the toxicological and chemical screens form the basis for the regulatory submission which allows clinical trials to begin.

The further down along the developmental pipeline the drug goes, the harder it is to terminate the project. This is because of the high cost (at the more advanced stage it is more expensive to halt the research than at the earlier phase) and also by the fact that at the late stage the drug discovery becomes a public knowledge and therefore, its termination can affect the public confidence.

2.3.7. Toxicology screening

A vital part of the discovery process are the safety and toxicology screens and the risk assesment. These aim to reduce the attrition rate of pipeline drugs when they reach the clinical phase. The discovery toxicology screen can be subdivided into three categories: drugs that cause seious adverse effects (AE) and as a result, are commonly withdrawn; toxicities that immediately affect further drugs’ development and toxicities present in animals (Figure 3). These toxicities and examples are discussed below in greater details.  



Figure 3: Types of drug toxicities

A.)  Serious AEs that alter drug development or lead to withdrawal.

The key systems toxicity tests look at are liver, heart ad nervous system. For the cardiovascular system, subjects are assessed for cardiac conduction (QT interval and electrocardiogram (ECG)), hypotension, hERG and/or blood pressure. Compounds are classified into low, intermediate or high risk according to the obtained hERG IC50 data. While hERG is usually screened during hit-to-lead stage, the effect of compounds on the ion channels (cardiac sodium channel or cardiac L-type calcium channel) is normally determined during lead normalisation. In the later stage of lead optimisation, ex vivo work (e.g.: cardiac electrophysiology, coronary blood flow) is performed on Langendorff hearts that are isolated from the guinea pigs. The Langendorff model is the preferred method (e.g.: over dog telemetry) as it can detect even a small increase in the QT interval and is directly related to the QT prolongation in humans caused by hREG inhibition. Follow up experiments are dose-dependent and measure the effect of the drug on the heart rate and blood pressure in anaesthetized rats.  Then, the mechanistic studies (e.g.: to measure cardiac contractility) are conducted in concious animals (e.g.: dogs).

Drug hepatoxicity is usually detected early during drug development, however some effects may only be indetified in late clinical development or post- market. This may be a result of population variance (age, drug metabolism or lifestyle), which is not normally taken into account during the earlier stages of drug development. To reduce potentially hepatoxic drugs making their way too far, several high throughput assays have been developed. These assays focus on mitochondrial toxicity and oxidative stress by assessing mitochondrial membrane potential, nuclei counts or nuclear area. These high-throughput assays produce quantitative data to determine the IC50 and the lowest toxic dose. The potential candidates are further tested in the in vivo toxicity screen, where liver weight or clinical pathology markers in serum are measured.


B.)  Drug toxicities that immediately affect further development

Drugs are also assayed for additional effects related to toxicity. Drugs are assessed early in development for genotoxic effects to prevent any serious complications. Many genotoxic chemical structures are already known and can be excluded in silico prior to the screen. Before the drug proceeds into the clinical trials, it is tested in the Ames test, which identifies mutagenic alterations in the coding sequence  caused by the interaction between the DNA and the chemicals. Further in vitro genotoxic studies can be also performed on rat liver fractions.

Immune suppression and immune enhancement (e.g.: hypersensitivity or autoimmunity) can cause immunotoxicity.  Immunotoxicity screens are usually evaluated in vivo by assessing changes in leukocyte counts and phenotypes. Also, the functional properties of the specific immune cells and whether these have been been altered should be examined using in vitro (e.g.: lymphocyte proliferation) or in vivo assays (e.g.: T-cell dependent antibody response assay).


C.)  In vivo toxicity

The main advantage of in vivo screening over in vitro is its ability to see the drug’s effect on the entire organism. Animal studies are a reliable way to detect toxic compounds before entering human trials. For most essential target organs, animals and humans share similar responses, with a correlation rate of more than 50%, with humas sharing a toxicity rate of 43% with rodents and 63% with non-rodents. Usually several factors are assessed, such as: mortality, clinical signs, body and organ weights and clinical pathology. Other endpoints can be also investigated and these are based on the target safety assessment (e.g.: biomarkers or toxicogenomics).   


3. Drug development. Clinical trials

Pre-clinical research should exclude most potentially toxic compounds, however it is important to determine any unexpected side effects with the human body. Prior to the clinical study, the aims of the research for each of the different phases are set up (Table1) and the investigational new drug process (IND) begins. The IND application includes the data from in vivo and toxicity studies, manufacturing information and the study plans.

Table 1: Phases of clinical trials, highlighting: the number of participants involved, the aim and t

4. Regulations and Approval: FDA Review

The Food and Drug Administration (FDA) regulatory review aims to evaluate the complete data sets and propose the labeling and manufacturing plans. Upon the positive confirmation of safety and efficacy results from the clinical trials, a new drug application (NDA) is submitted to get approval for drug marketing. In the application, the results from pre-clinical and clinical development programs as well as proposals for manufacturing and labeling are considered.

The review can be accelerated, e.g.: through a “fast track” if it is suitable for people, suffering from serious conditions or there is an unmet medical need. The drug can also be classified as a “breakthrough therapy” when it demonstrates significant improvements over the marketed drugs. When the drug aims to treat a serious condition or targets an unmet disease, it’s approval can be fast tracked based on a surrogate or an unmet clinical endpoint. Finally, the drug can undergo a “priority review” when it is predicted to bring significant improvements in the efficacy and safety of the diagnosis and/or treatment of a serious condition.

Drug approval is a very long and thorough process and occasionally, additional screens and tests are requested to prove the novel drug has an advantage over the marketed drug.


4.1.  Manufacturing: high quality production of new medicines on a large scale

The approved drugs can be administered to a small population of patients or to millions of people. In either case, during each stage of the manufacturing procedure, quality and safety checks are conducted to ensure that companies meet good manufacturing practice (GMP) requirements approved by FDA. During manufacturing, the cutting edge technologies and advances are implemented (i.e.: nanotechnology) which are closely coordinated with the automation and software.


5. Post-approval research and monitoring

Upon completion of R&D work the drug research continues and monitors its efficacy and new long-term side-effects. Moreover, it can be repositioned to a new patient population. However, additional beneficial properties of a drug can be revealed leading to expansion of its use. For instance, the drug’s therapeutic value is maximised when for a given indication it has a greater therapeutic value than seen during clinical trials, it can be used to target a different disease, has a new way of delivery or brings an advantageous effect when used as a combination therapy.



6. Conclusion

To conclude, drug development is a very long and complex process, sometimes taking up to 10-15 years to develop a new medicine. Increased understanding in human biology and disease biochemistry opens the doors for new advances in drug discovery and possibilities to discover a new treatment option.

As the complexity of science increases, so do the challenges facing drug development, with more thorough checks on pharmaceutical efficacy and safety properties. This can be explained by a broad range of toxicological effects drugs can cause (Section: 2.3.7). To overcome these, a variety of in vivo screens (e.g.: Mitotox) have been developed which aim to predict the possible adverse effects drugs can cause. By improving out understanding of the relationship between drug structure and activity, a pipeline drug can move to a further stage, where it is tested in vivo and if successful, in humans. All the obtained data and analysis with the proposed manufacturing procedure and packaging are later submitted to FDA and even upon it’s approval, it is further monitored for its efficacy, possible adverse reactions or possibilities to be repositioned.

Considering the advances in research and medical innovations, there is a hope that  the success of drug development will improve in the near future. 



Anon. Immunotoxicity Studies for Human Pharmaceuticals. Note for Guidance on Immunotoxicity for Human Pharmaceuticals (S8). The European Agency for the Evaluation of Medicinal Products, (2006).

Bailey, W. J. et al. 'A Performance Evaluation Of Three Drug-Induced Liver Injury Biomarkers In The Rat: Alpha-Glutathione S-Transferase, Arginase 1, And 4-Hydroxyphenyl-Pyruvate Dioxygenase'.Toxicological Sciences 130.2 (2012): 229-244. Web.

Bass, Alan S. et al. 'Exploratory Drug Safety: A Discovery Strategy To Reduce Attrition In Development'. Journal of Pharmacological and Toxicological Methods 60.1 (2009): 69-78. Web.

Bell, Robert M., Mihaela M. Mocanu, and Derek M. Yellon. 'Retrograde Heart Perfusion: The Langendorff Technique Of Isolated Heart Perfusion'. Journal of Molecular and Cellular Cardiology 50.6 (2011): 940-950. Web.

Bertram, Lars, and Rudolph E. Tanzi. 'Thirty Years Of Alzheimer's Disease Genetics: The Implications Of Systematic Meta-Analyses'. Nature Reviews Neuroscience 9.10 (2008): 768-778. Web.

Budnitz, Daniel S. et al. 'National Surveillance Of Emergency Department Visits For Outpatient Adverse Drug Events'. JAMA 296.15 (2006): 1858. Web.

Chessell, Iain P. et al. 'Disruption Of The P2X7 Purinoceptor Gene Abolishes Chronic Inflammatory And Neuropathic Pain'. Pain 114.3 (2005): 386-396. Web.,. 'Clinical Research'. N.p., 2015. Web. 12 Nov. 2015.

Guth, Brian D et al. 'Developing A Strategy For The Nonclinical Assessment Of Proarrhythmic Risk Of Pharmaceuticals Due To Prolonged Ventricular Repolarization'. Journal of Pharmacological and Toxicological Methods 49.3 (2004): 159-169. Web.

Hornberg, Jorrit J. et al (I). 'Exploratory Toxicology As An Integrated Part Of Drug Discovery. Part I: Why And How'. Drug Discovery Today 19.8 (2014): 1131-1136.

Hornberg, Jorrit J. et al (II). 'Exploratory Toxicology As An Integrated Part Of Drug Discovery. Part II: Screening Strategies'. Drug Discovery Today 19.8 (2014): 1137-1144.

Henning, S., and G. Beste. 'Loss-Of-Function Strategies In Drug Target Validation'. Current Drug Discovery 5 (2002): 17-21. Print.

Hughes, JP et al. 'Principles Of Early Drug Discovery'. British Journal of Pharmacology 162.6 (2011): 1239-1249. Web.

Kola, Ismail, and John Landis. 'Opinion: Can The Pharmaceutical Industry Reduce Attrition Rates?'. Nature Reviews Drug Discovery 3.8 (2004): 711-716. Web.

Kurosawa, G. et al. 'Comprehensive Screening For Antigens Overexpressed On Carcinomas Via Isolation Of Human Mabs That May Be Therapeutic'. Proceedings of the National Academy of Sciences 105.20 (2008): 7287-7292. Web.

Lipinski, Christopher A. et al. 'Experimental And Computational Approaches To Estimate Solubility And Permeability In Drug Discovery And Development Settings'. Advanced Drug Delivery Reviews 64 (2012): 4-17. Web.

Michelini, Elisa et al. 'Cell-Based Assays: Fuelling Drug Discovery'. Anal Bioanal Chem 398.1 (2010): 227-238. Web.

Moore, K., and S. Rees. 'Cell-Based Versus Isolated Target Screening: How Lucky Do You Feel?'.Journal of Biomolecular Screening 6.2 (2001): 69-74. Web.

Olson, Harry et al. 'Concordance Of The Toxicity Of Pharmaceuticals In Humans And In Animals'. Regulatory Toxicology and Pharmacology 32.1 (2000): 56-67. Web.

Paolini, Moreno, and Giorgio Cantelli-Forti. 'On The Metabolizing Systems For Short-Term Genotoxicity Assays: A Review'. Mutation Research/Reviews in Mutation Research 387.1 (1997): 17-34. Web.

Peet, Norton P. 'What Constitutes Target Validation?'. TARGETS 2.4 (2003): 125-127. Web.,. 'Phrma | Pharmaceutical Research And Manufacturers Of America'. N.p., 2015. Web. 5 Oct. 2015.

Tamargo, Juan et al. 'Investigational Positive Inotropic Agents For Acute Heart Failure'. CHDDT9.3 (2009): 193-205. Web.

Van Deuren, Bruno et al. 'The Fentanyl/Etomidate-Anaesthetised Beagle (FEAB) Dog: A Versatile In Vivo Model In Cardiovascular Safety Research'. Journal of Pharmacological and Toxicological Methods 60.1 (2009): 11-23. Web.

Yang, Yongliang, S. James Adelstein, and Amin I. Kassis. 'Target Discovery From Data Mining Approaches'. Drug Discovery Today 14.3-4 (2009): 147-154. Web.

Zanders, Edward D, David S Bailey, and Philip M Dean. 'Probes For Chemical Genomics By Design'. Drug Discovery Today 7.13 (2002): 711-718. Web.



Fastbleep © 2019.