Challenges in Drug Discovery, Why 90% Of Drugs Fail Before Approval

The pharmaceutical industry's most expensive and comprehensive part is drug development, also called drug discovery. The intricate process of designing a new drug covers the latest drug compound detection or screening for introduction to the pharmaceutical market. The further stage involves preliminary preclinical evaluation and clinical studies to ensure the safety and efficacy of the compounds for human use. All these stages, along with quality control assurance amid regulatory approvals, are essential to calibrate and validate if the medicine is safe for human use. This process is vital to meeting existing medical needs, advancing medicine and improving patient outgrowth. 

Amid all the stages involved in drug development, a few challenges are involved during target identification, preclinical studies, regulatory approval and many more. In this article, we will explore various challenges occurring in drug development and the implementation of AI in drug discovery, as well as provide an overview of the role of AI in drug development for controlling the hurdles and delays. 

The Stages of Drug Discovery:

1) Research and development - In this initial phase, scientists explore disease mechanisms and identify potential drug candidates through laboratory research.  

2) Target identification - Identifying a biological molecule, such as a protein or gene, plays a key role in disease progression so that the drug can modify the target.

3) Preclinical studies-  A laboratory study conducted on animal models to evaluate a drug candidate's safety, efficacy, and pharmacokinetics before human testing begins.

4) Investigation of New Drug (IND) - An application submitted to regulatory authorities, such as the FDA, seeking approval to begin human clinical trials based on existing preclinical data.

5) Clinical trials -  It consists of a four-phase process involving human participants to evaluate a drug’s safety, optimal dosage, and effectiveness in treating the disease. 

6) Review and Approval -  Regulatory agencies evaluate clinical trial data, manufacturing processes, and QA/QC reports to determine if the drug meets approval criteria.

7) New Drug Application - A formal request submitted to regulatory bodies for commercial approval, including comprehensive data on clinical trials, safety, efficacy, and manufacturing details.

Also Read: Drug Development: From Discovery to Market 

What are the Challenges In Drug Discovery

Target Identification

Drug development starts with identifying a target for a drug to act so that the drug compounds can be identified to potentially hit the target. Understanding the cause of a disease condition helps to know the pathways of the condition and develop a target-oriented drug. Thus, selecting the appropriate biological molecule is essential. Though this process is crucial in drug development, it is often complicated due to biological systems and how targeting specific molecules would impact the effect of disease. Diseases like Alzheimer's, cancer, cardiovascular disorder and diabetes are a few diseases with multiple target approaches that face significant challenges while identifying exact proteins for complex diseases, detecting binding sites of effective target proteins, and handling potential unexpected adverse effects of other proteins. 

Developing Biomarkers

Biomarkers, or biological markers, are measurable traits that indicate the organism's status and are utilized for clinical evaluation of health, disease, and therapeutic efficacies. Biomarkers can be genes, molecules, proteins, tissues or cell types indicative of the body's pathological or normal processes. Biomarkers in drug development are applied to monitor the response of the drug and evaluate its safety profile.

One of the main challenges in biomarker discovery is the high number of false discoveries. This occurs when other laboratories or independent samples cannot reproduce initial scientific findings associated with a new biomarker.  It is difficult to identify the biomarkers and eventually develop targeted drugs if the biological mechanisms of the disease are unknown. Correct clinical validation methods are essential in biomarker development because improper implementation during development can negatively impact patient outcomes. 

Animal models

In the process of drug development, preclinical trials are in-vivo analyses of drug molecules consisting of animal models. These non-human models may occasionally fail to predict the drug's accuracy, efficacy and therapeutic index. This is likely due to the inability of animals to mimic disease comprehensively due to biological differences, such as variation in metabolism, which can result in inaccurate effectiveness. Preclinical experiments in cells, tissues, and animal models are imperfect representations of human disease treatment due to differences in organ structure, immune responses, genetic makeup, and signalling pathways, further contributing to the challenges; hence, the positive results in animal model systems may not replicate in human participants. 

These limitations have led to significant drug failures in clinical trials. For example, specific gene therapy vectors considered safe in animal studies caused severe liver failure and brain swelling in children. Similarly, HIV vaccines, stroke treatments, inflammatory disease agents, and Alzheimer’s therapies showed promise in animal models. Still, they ultimately failed in human trials, highlighting the need for more reliable predictive models in drug development.

Clinical Trials

Clinical trials are the primary need for the development of drugs as it is essential to carry out trials on a minor number of human volunteers initially to analyze efficacy and safety and to perceive the pharmacology of the drug and the occurrence of adverse drug reactions. A limited patient population, recruitment of patients from diverse geographical regions and genders, and lack of patient involvement are significant challenges in clinical trials. Enrolling patients for rare diseases and maintaining diversity in trial volunteers, patient retention, management and identification of ADRs are substantial issues in trials as they can terminate or halt the drug discovery process. Gen Li, the president and founder of Phesi, a clinical data science company that uses AI to analyze clinical trial data, said, “Clinical trials never get started. Once they get started, they never get to end.”

High cost- Drug development Is Expensive!

Drug development is a costly process; finance plays a significant role in the initial stage, from drug research to target identification validations, lead optimization, clinical-preclinical trials, manufacturing, marketing, and post-marketing surveillance. The quality and efficacy of the drug depend on the standard of raw materials used. High drug discovery costs stem from high failure rates, expensive clinical trials, complex regulatory processes, and costly patent protection; also, there's no promising return on investment. The limited funding for orphan drugs and the high cost of R&D are the additional hurdles to financial challenges.

The development expense of neurodegenerative diseases, gene therapies, and personalized medicine is also too high. Potential solutions include public-private partnerships, advanced computational modelling, biomarker development, and improved reimbursement policies. These approaches aim to reduce costs, enhance efficiency, and innovation in drug development for unmet medical needs.

Data from “Estimated Research and Development Investment Needed to Bring a New Medicine to Market, 2009–2018,” JAMA 

Lack of information

As drug development is a novel process, there is a rare or very low possibility of pre-existing data. The treatment duration and dosage regimen are not specified for preclinical trials; hence, the minimum tolerated dose is approached, which may not result in the desired outcomes of the drug. Also, drug development for rare diseases can be significantly challenging due to a lack of knowledge, which can also delay research. A lack of information is a significant challenge in drug discovery, and it includes unclear disease pathways, limited patient data, inadequate animal models, and the absence of reliable biomarkers. 

Unclear disease mechanisms, as seen in Alzheimer’s, make it difficult to identify effective drug targets. Insufficient data on diverse populations affects drug efficacy, particularly for rare diseases. Inaccurate animal models can mislead preclinical results, especially in neurological conditions. Additionally, the absence of biomarkers, such as in early-stage cancer, complicates treatment monitoring and patient selection. 

Regulatory process and approval

The regulatory approval process is the most complex, expensive, and time-consuming procedure, as it is very critical to adhere to the agency's norms. IND applications are precisely reviewed, and any inadequacies that occur can put clinical trials on hold, leading to delay or even rejection. The lengthy process, often exceeding a decade, requires extensive safety and efficacy testing. Stringent requirements can delay approvals, high compliance costs, and drug development expenses surpassing billions. Evolving regulations, especially in gene therapy, make ongoing adjustments a necessity. Additionally, approval variability across regions complicates global launches. These challenges highlight the need for streamlined regulatory pathways to accelerate drug availability.

Rising Antimicrobial Resistance (AMR)

Another major challenge in drug development is the rising antimicrobial resistance (AMR), which causes existing antibiotics to be ineffective against evolving pathogens. Despite the urgent need for new antibiotics, R&D in this field is slow due to high failure rates, regulatory hurdles, and limited financial incentives. Many pharmaceutical companies hesitate to invest in antimicrobial drug discovery because of low profitability and stringent approval processes. This crisis threatens global health, making it crucial to overcome barriers to antibiotic innovation.

Increasing Pressure On Applying Sustainable Practices

With the rising awareness about the environmental impact, people are seriously looking forward for sustainable solutions. Pharmaceutical manufacturing generates substantial waste, consumes vast resources, and contributes to environmental pollution. Regulatory bodies and stakeholders now expect drug developers to adopt greener practices, such as reducing carbon emissions, minimizing chemical waste or dumping chemical waste more ethically, and ensuring honest sourcing of raw materials. However, integrating sustainable practices into drug development without compromising efficiency or escalating costs remains a significant hurdle for the industry.

What is the role of generative AI in drug discovery?

Artificial Intelligence (AI) is transforming drug development by enhancing efficiency across various stages of drug discovery. In 2024, Charles River and Lundbeck used AI-powered Logica® to speed up development in neurological drugs. The latest news is that Lisata and GATC, a clinical-stage pharmaceutical company and a biotech firm, utilize the AI-driven Multiomics Advanced Technology™ (MAT) platform to accelerate and optimize drug research.

AI improves patient selection, predicts molecular effects, and enhances adherence monitoring in clinical trials. AI evaluates vast datasets for target selection and validation using deep learning and graph-convolutional networks to prioritize molecular targets. AI-driven preclinical study models like Deeptox predict toxicity and pharmacokinetic properties. AI also revolutionizes compound screening and lead optimization through virtual screening, speeding up drug candidate identification. These advancements streamline research, reduce costs, and accelerate drug development, making AI a crucial tool in modern pharmaceutical innovation.

Conclusion

The difficulties mentioned above are common but prominent challenges in drug development. The process also includes a lack of disease awareness, lack of resources, clinical holds, negative results, and molecular incompatibility. The drug discovery process is a complex journey that involves multiple stages of drug discovery, from target identification in drug discovery to regulatory approval. 

Despite challenges like high costs, lengthy trials, and uncertain outcomes, advancements in AI in drug discovery are reshaping the landscape. What is drug discovery without failures? With artificial intelligence in drug discovery, we can accelerate research, improve accuracy, and reduce costs. The role of generative AI in drug discovery is particularly promising, offering innovative solutions for molecule design and predictive modelling, bringing us closer to faster, safer, and more effective treatments.