L1 + L2: Drug discovery and development pipeline Flashcards
the stages of the pipeline and the drug discovery questions at each stage?
- therapeutic concept: what’s the working hypothesis?
- target selection: what’s the molecular target?
- target validation: is a drug acting on that target likely to be effective?
- lead finding: can a (chemical) starting point be identified?
lead optimisation: can the drug-like properties be optimised?
CANDIDATE DRUG
end of discivery phase: development phase begins
preclinical development: will it be appropriate for testing in humans?
clinical development: is it safe and therapeuticaly useful in humans?
regulatory approval: does this drug have value to society?
product: is the drug commercially attractive?
drug discovery and development pipeline description?
Drug discovery and development pipeline
It doesn’t act much like a pipeline! Pipeline- good at transporting oil from x to y. Drug discovery pipeline: 0.2% success rate.
It’s very “leaky” - highly inefficient:
Discovery success rate ~ 0.2%;
Development success rate = 11.8%
It’s very costly: Overall average cost per approved compound = $1.4 billion
It’s very slow: Overall time from concept to product = 10-15 years
It’s 2-directional: Projects can reach a point where they have to turn back/start again
Further along pipeline drug gets, more expensive it will be to fail.
therapeutic concept?
Therapeutic concept - the working hypothesis
Usually combines a research , development and commercial view
Research: different pharmaceutical companies have different areas of specialisation. E.g: cardiovascular, neuroscience etc. pharmaceutical companies look if it is in their area of expertise / if its something they can do.
Research; can it be done?
Development; can it be tested? Have they got an assay, pre clinical, clinical that will give confidence that drug is working as it should.
Animal models- are they predictive of human conditions? Is there a good animal model for the disease? With clinical trials - can it give us the confidence that drug is working (?)
Commercial; can it be sold? Is there a market out there for the drug
We know if treat patient with antibacterial drug, eventually bcteria will become resistant. Only thing to do is to not sell it. So not a good commercial model for pharmaceutical companies. That is why antibacterial development halts.? Governments - if pharmaceutical companies develop ab we will guarantee income/ annual fee even if not selling it. Trying to get around this problem in the pipeline where best way to prevent resistance is to not sell antibiotic.
Sometimes the concept turns out to be correct
Trastuzamab (Herceptin) selectively binds to the Human Epidermal Growth Factor Receptor (Her2) and targets aggressive, Her2+ tumour cells in the treatment of breast cancer
Sometimes the concept has to be changed
Sildenafil (Viagra) selectively inhibits PDE5 for the symptomatic treatment of angina went to clinical trials and turned out to not be good for angina. Repurposed for the treatment of male erectile dysfunction when the purpose was not set out for that.
Sometimes the concept is wrong
Xamoterol (Corwen) acts as a partial β1 agonist to increase cardiac output and prolong survival in congestive heart failure. Complete disaster at clinical trials.
target selection approaches?
Target selection strategy is usually based on some early target validation
Mechanism of existing drugs - perhaps a drug is prod a side effect that can be useful in treating another condition. E.g: isanisan?? Anti tb drug. Side effect- patients who took it and had depression had their depression relieved.
Analysis of disease pathophysiology- essential to understand disease pathology
Information gene expression- genes may be mutated/overexpressed etc. once identify gene you can look at what protein it produces and target it.
target selection?
Understanding Molecular Targets in Drug Development
Human diseases are complex and influenced by multiple factors. Because of this complexity, there are often many potential targets for treatment, rather than a single critical protein responsible for a disease.
Universities play a crucial role in advancing our understanding of diseases by studying protein interactions and signaling pathways. This research helps identify molecular targets that could be useful for drug development. To select and validate effective drug targets, it is essential to understand how different biological systems interact.
Due to these complex interactions, all drugs cause side effects. This is why selecting the right molecular targets is critical for drug safety and effectiveness.
Most Common Molecular Target for Drugs
G-protein-coupled receptors (GPCRs) are the most common molecular targets for drugs.
Studies analyzing human drug targets have revealed:
a) The distribution of drug targets across different gene families.
b) The proportion of small-molecule drugs that target each family.
c) The percentage of chemical compounds, both approved (orange bars) and non-approved (grey bars), that act on various protein targets.
These data are sourced from ChEMBL, a database containing information on 2.5 million bioactive molecules with drug-like properties.
Advancements in Target Selection
Increasing knowledge of protein structure is transforming how drugs are designed. Understanding the molecular structure of targets, such as protein folding and receptor binding sites, allows for more precise drug targeting.
While computational (in silico) drug design is advancing, we are still not at the stage where drugs can be fully developed inside a computer. They must still be tested in biological systems to confirm their effectiveness and safety.
which came first, the target or the drug?
Target:Drug combinations identified by analysis of pathophysiology i.e. where understanding of the disease led to identification of a molecular target and the development of the drug.
Sometimes disease comes first. Find drug that acts on disease and only then do we find a molecular target. Benzodiazaphines - used to treat anxiety or sleep disorders. No idea how these drugs work. But found to cause hyptonic effect in some animal models. So develop drugs to help people sleep and treat anxiety. Now know molecular target: benzodiazaphine receptor part of gaba? Now know molecular target but not when drug developed.
Anti psychotic drugs- schizophrenia. Did not know how drug worked but saw benefit but now know it acts in human dopamine-2 receptor.
Not always the case that we have molecular target and drug. Increasingly this is the case. But we can see drug effective in disease in animal model or prod side effect then after figure out molecular target.
target validation strategies?
Real’ target validation occurs in Phase II clinical studies (first time given to patient with disease)
But preclinical validation gives confidence
Preclinical Target Validation approaches
Pharmacological
Non-selective compounds from literature
Early stage project compounds
Different classes of compounds with the same effect (e.g. antibodies)
Genetic
Antisense oligonucleotides
RNA interference (RNAi)
Transgenic animals (deletion/overexpression)
Good example of this is voltage dependent na channel involved in cnducting action potential of sensory nerves and so involved in pain.
How Preclinical Approaches Validate a Drug Target
Target validation means proving that a specific molecule (e.g., a protein or gene) is directly involved in a disease and can be targeted by a drug to treat that disease. Preclinical approaches help confirm this before testing in humans.
- Pharmacological Approaches (Using chemicals/drugs to test the target)
These methods use compounds to see if blocking or activating a target produces the expected effect.
Non-selective compounds from literature → If existing drugs that affect the target change the disease symptoms, this suggests the target is relevant.
Early-stage project compounds → If new drugs designed for the target show an effect in lab models, this supports its role in disease.
Different compound classes with the same effect (e.g., antibodies) → If multiple drug types targeting the same molecule give the same result, it strengthens confidence in the target’s role.
🔹 Example: If different drugs that block a protein all reduce tumor growth in cancer models, it suggests that protein is important for cancer progression.
- Genetic Approaches (Modifying genes to test their role in disease)
These methods change gene expression to see if it affects disease symptoms.
Antisense oligonucleotides (ASOs) → Block the target’s gene from making protein. If symptoms improve, it confirms the target is important.
RNA interference (RNAi) → Silences the gene. If this reduces disease effects, the target is validated.
Transgenic animals (gene knockout/overexpression) → Mice or other animals are engineered to lack the target or overproduce it. If the disease changes, the target is likely involved.
🔹 Example: If mice missing a certain gene don’t develop Alzheimer’s symptoms, it suggests that gene is a valid drug target.
Example: Nav1.7 and Pain Perception
Scientists used genetic validation to prove Nav1.7 was crucial for pain perception:
Two families (Pakistan & Italy) had mutations in Nav1.7, making them unable to feel pain.
This showed blocking Nav1.7 could be an effective painkiller strategy.
However, drug development failed because there were too many similar sodium channels (Nav1.1–Nav1.9), and selective inhibition was too difficult.
examples: raditine, ticagrelor, adallimumab, imatinib
Example 1 -ratidine
A once a day treatment for gastric ulcers.
Therapeutic concept: Once a day H2 receptor antagonist for peptic ulcer treatment
Pfizer’s strategy
Research: strong chemistry and Drug Metabolism and Pharmacokinetics (DMPK) expertise
Development: clinical trial end points & regulatory strategy established
Commercial: once a day H2 antagonist therapy would improve compliance
Target selection and validation based on:
Mechanism of action of existing drug – cimetidine (taken twice a day)
Analysis of disease pathophysiology - role of histamine acting on H2 receptor in gastric acid hypersecretion
Simetidine peptiulcer treatment? Already knew how to test it. Knew commercial success - once a day drug. Simetidine was 3x a day. So knocked simetidine off the market.
Example 2 - Ticagralor
Therapeutic concept; Oral P2T (P2Y12) receptor antagonist as an anti-platelet agent
Astra/AstraZeneca’s strategy:
Research: Strong chemistry and purinoceptor expertise
Development: Clinical trial end points & regulatory strategy established
Commercial: Product placement focused on differentiation from Cangrelor (administered intravenously ) and in turn clopidogrel (which is a pro-drug requiring activation by hepatic metabolism)
Target selection and validation based on:
Analysis of disease pathophysiology - known role of platelet in CV disease
Mechanism of action of endogenous mediator – ADP
Proof of concept with Cangrelor
Strong chemistry and expertise in these receptors. Development easy. Drug already available (already mentioned in text) Knew how to test. Big advantage- orally active drug. Cangrelor- administred intravenously (problem), clopidegrol - not as effective in patients with liver problems as requires liver for activaton.
Example 3 - Adalimumab
Concept: Humanised anti-TNFα antibody for the treatment of rheumatoid arthritis
Cambridge Antibody Technology’s strategy
Research: Unique phage display based antibody generation technology
Development: Clinical trial end points & regulatory strategy established
Commercial: Company placement based on value of anti-TNF market
Target selection and validation based on:
Analysis of disease pathophysiology - known role of TNFα in RA
Mechanism of action of existing drug – etanercept (fusion protein that mimics TNFα receptor and binds soluble TNF-α)
Example 4 - Imatinib
Concept: Abl-kinase inhibitor for the treatment of chronic myeloid leukaemia (CML)
Ciba-Geigy/Novartis’s strategy
Research: Strong chemistry, kinase inhibitor background
Development: No “prior-art”. Novel, accelerated clinical trials programme established
Commercial: Commercial potential limited but no direct competition; kinase inhibitors toxic & non-specific
Target selection and validation based on
Information from gene expression studies - oncogene product Bcr-Abl constitutively active
Analysis of disease pathophysiology - mutated tyrosine kinase led to malignancy
Clear association of oncogene, Abl kinase and CML
how are lead compounds identified?
Lead compounds identified by analysis of pathophysiology
Molecular target is known at project inception
Project is driven by finding & refining novel chemicals with affinity for the molecular target
As project matures, improved compounds can be used to validate the molecular target
Lead Finding Strategies (Ways to find promising drug candidates)
Virtual Screening
Using computer models to predict how molecules will bind to the target protein.
High Throughput Screening (HTS)
Pharmaceutical companies have built huge chemical libraries over decades.
These contain thousands or millions of chemical compounds.
They test many compounds quickly to see if any bind to the target.
Natural Product Screening
Extracts from plants, animals, and microorganisms are screened for active compounds that might work as drugs.
Fragment Screening
Instead of testing whole molecules, small fragments of molecules are tested.
If a fragment binds well, researchers modify it to create a more effective drug.
Privileged Structure Screening
Some chemical structures are naturally good at binding to proteins.
These structures are prioritized in drug discovery.
Family-Based Compound Libraries
Some drug families (e.g., kinase inhibitors for cancer, GPCR-targeting drugs) are known to work well.
Researchers focus on similar compounds for new drugs.
Natural Ligands
Some molecules naturally found in the body (like catecholamines, purines, tryptamines, histamine) already bind to biological targets.
Drugs are designed to mimic or block these natural ligands.
Repurposing Known Drugs
Existing drugs with overlapping activity might work for new diseases.
leads identified by analysis of drug effects in a bioassay?
Leads identified by analysis of drug effects in a bioassay
Project is driven by refining the desirable drug effect
Often, target is not known at project inception
As project matures, improved compounds can be used to identify molecular target
Example; levetiracetam
Developed invivo in animal models. Approved without knowing how it worked.
Anti-epileptic discovered & developed using complex models of epilepsy in vivo
Approved in the USA for myoclonic seizures in 2000
Putative molecular target (SV2A; synaptic vesicle protein) published in 2004
Did not know at the time how it worked
Note: Regulatory Authorities do not require that the molecular target of a drug is identified prior to authorisation, although some intelligent guesswork, or a working hypothesis is useful.
