Drug Development: Can AI Change the Game?
Drug development has historically been a long, expensive, and difficult process. How will the use of artificial intelligence impact the future of this industry? Written by Charlie Almoney, Morgan Kofsky, Jake Pearson, and Jenny Zwerling
In 2019, the total estimated cost to research and develop a novel prescription medicine was $2.6 billion per drug. This figure has increased from $800 million in 2003. Additionally, the rate of success of drugs in clinical trials is 12%, a decrease by half over the same time period1. While costs have risen, the success rate of drug development has been driven down, displaying a significant need for the implementation of more cost-effective and efficient development methods. In recent years, Artificial Intelligence has been proposed as one solution to this problem.
Artificial intelligence can increase the efficiency of drug discovery in many ways. Through the use of AI, pharmaceutical firms can better understand whether a specific drug will yield a positive outcome. AI can generate data and models that can impact the drug discovery and development process through all levels. Establishing biomarkers, validating drug candidates, repurposing old treatments and building new treatments, designing preclinical experiments, and optimizing clinical trials are just a few of the ways that AI has already impacted the industry, and we will explore some of these developments below.
The biggest factor that allows AI to have such a large impact on medicine is in the growth of big data in Healthcare. Big data in healthcare is projected to grow faster than in any other field through 2025, with a CAGR of 36%2. What to do with all of this data is where AI and Machine Learning comes in. With the capability to ingest, analyze, organize, and output complex data sets into a form easy to use by humans, companies can leverage these tools to gain insights never possible before the growth of data in healthcare. For example, many start-up companies have developed systems capable of ingesting sets of already approved drugs and matching their capabilities to known biomarkers of diseases without a cure. In this way, researchers can build pipelines of pre-approved drug candidates that can streamline their development process. Additionally, similar software has been built that recruits for and optimizes the management of clinical trials to decrease the cost burden of large and inefficient studies. AI can build patient databases that allow researchers to quickly recruit the best possible participants to increase the likelihood of trial success, as well as manage doctor-patient interactions to improve trial accuracy. Using these tools, researchers can speed up their testing and spend less time and money in recruitment and management logistics, which will in turn lower costs and improve trial success.
AI has seen growth in almost any technology-centered industry and has been applied all across the professional world. What’s special about the drug development space is the increased difficulty of finding, building, and getting medicines approved despite significant technological jumps in research and healthcare. Where technology has lowered costs and dramatically increased access to many industries, drug development costs have only risen. AI can be the tool that solves this problem and provides pharmaceutical companies and research institutions the cost-effectiveness needed to succeed in drug development.
