In current pharmaceutical sciences, when we talk about innovative drug discovery, we should think of target-based drug discovery, which is to design synthetic drugs based on a target. Then, have we thought about how efficient this drug development model is in the real world? Especially compared to another well-known phenotypic drug development model. To answer this question, Iranian scientist Arash Sadri recently published a review on JMC, which conducted in-depth analysis with logical reasoning.
First, the author found that the overall efficiency of drug development is unsurprisingly low. The probability of a drug from entering clinical trials to being approved for marketing is about 13%, and for more complex CNS and cancer, the probability is 0.4%. The average cost is 1.3 billion US dollars, and some even reach 6.4 billion US dollars. So where is the problem? Of course, there are many reasons, such as the “low hanging fruit” that is already picked, but one of the main reasons is that drug development is simplified to target-based drug discovery. This model simplifies the therapeutic effect from interfering with the complex network composed of cellular and extracellular components and their complex feedback arcs to regulating one or several proteins.
The author also found that the early days drug discovery started from the discovery of the phenotype caused by a substance. But with the development of molecular biology techniques and the huge success of some targeted drugs such as imatinib, the target-based drug discovery has attract more attention. However, with the development of systems pharmacology, network pharmacology, and multi-pharmacology, as well as the fact that most of the original drugs discovered in the 1999-2008 period were discovered using phenotypic methods, the voices of anti-simplification are growing louder. So what is the truth? After reading the following questions and answers, I think you will have a preliminary conclusion.
Question 1: How many of the drugs on the market are discovered by target-based drug discovery?
Answer: The author analyzed and found that of the 1,144 approved small molecule drugs, 123 were discovered by target-based drug discovery, and 1,021 were discovered using phenotypic methods. And in the number of drugs approved each year, the number of drugs discovered based on targets has never exceeded the number of drugs discovered based on phenotypes.
Question 2: Can the therapeutic effects of all drugs discovered based on targets be attributed to binding to a single protein?
Answer: Obviously not entirely. The author analyzed and found that many drugs discovered based on targets have many off-targets, such as donepezil, an acetylcholinesterase inhibitor, but later found that there are 40 mechanisms that are not related to this enzyme. Another example is crizotinib, which can bind to 146 kinases in addition to inhibiting MEK and ALK. These hints suggest that many of the 123 drugs that are attributed to target-based discovery are not effective because of their intended action on a single target.
Question 3: What is the theoretical basis of target-based drug discovery?
Answer: The author analyzed and found that this model is to treat the human body as a machine. Because a machine is made up of parts, if a part of the machine breaks down, then you fix the broken part. So this kind of thinking is naturally applied to drug development. But the function of the human body is not like this. It is the result of natural selection. Natural selection does not impose a function on a certain part of the body, but selects based on the final phenotype. So treating it as a machine obviously simplifies the human body.
In conclusion, the author believes that target-based drug discovery is undoubtedly valuable, such as the success of monoclonal antibodies and drugs developed based on some disease-driving proteins. However, we should also be aware of the problems with target-based drug discovery. In the beginning of this drug development, it is not based on the improvement of phenotype to obtain molecules, but only uses the improvement of phenotype for evaluation at the end. This single-directional end-blinding mode is undoubtedly risky. In contrast, phenotypic drug development focuses on phenotype improvement from the beginning.
For more information, please see the original text (DOI: 10.1021/acs.jmedchem.2c01737).