EglenMed Founder Du Tao: AI in Clinical Trials — The Next Investment Hotspot in AI-Driven Drug Development

Editor's Note: This article comes from PharmaCube, and VCBeat is authorized to reprint it.

In recent years, the application of AI technology in the innovative pharmaceuticals field has been booming. Domestic AI drug discovery startups have emerged like mushrooms after rain, with hundreds of companies joining this wave of AI transformation; additionally, established pharmaceutical enterprises and internet giants have also entered the AI drug discovery track, showcasing an unprecedented momentum for AI + innovative drugs.

Since 2022, China's biopharmaceutical industry has experienced a capital "winter." Although last year ChatGPT brought the AI pharmaceuticals sector back into the spotlight, the industrial and investment communities, after gradually returning to rationality, have begun to reassess the future development path of AI-driven drug research and development, especially as it moves forward.Clinical Regulatory PhaseAfterward, how to leverage AI technology to navigate the entire drug development process remains a focal point of attention for all parties.

Compared with the bustling AI drug and target discovery track, at which stage has the global AI clinical track currently developed? Why are there relatively fewer companies in China's industrial chain focusing on "AI + clinical"? For the "most expensive" clinical trial环节, what empowerment can AI technology currently provide? What breakthroughs are expected to be achieved in the future?

Recently, at the closed-door conference titled "New Revolution, New Momentum, New Opportunities – AI Technology Leading the Future of Innovative Drug Industry," hosted by Zhangjiang Group, Zhangjiang AI New Drug R&D Alliance, and PharmaDJ, we...We were fortunate to interview Dr. Tao Du, founder and chairman of Evergreen Therapeutics, for an in-depth discussion on the logic, breakthroughs, and future potential behind AI + clinical applications.

Evergreen Therapeutics is one of the earliest companies in China to enter the AI clinical field. Relying on its self-developed AI algorithm platform, it has obtained FDA clinical approvals for five candidate drugs in less than five years since its establishment, with the fastest project already entering Phase III clinical trials.

"Globally, the AI clinical track has already taken shape. Currently, there are dozens of companies worldwide engaged in AI clinical services, and many multinational pharmaceutical enterprises have directly invested in AI clinical design companies. According to the data, since 2021, the number of clinical trials designed using AI technology in the United States has exceeded triple digits annually. Therefore, I believe,"The entry of AI into the costly field of clinical practice is inevitable, and this track will certainly become a hot spot for investment and an important position in the industry in the future."Dr. Du Tao said."

Q: How did Evergreen Therapeutics develop into an AI pharmaceutical company?

Dr. Tao Du:Evergreen Therapeutics is an innovative pharmaceutical company co-founded by several former FDA reviewers, including many pharmaceutical industry experts from both China and the U.S., forming the company's management system. The company was established in September 2019, and we encountered information related to AI and pharmaceuticals in the summer of 2020, which was more than half a year after the company’s founding. By chance, we discovered that Roche, Pfizer, and Merck all had AI teams consisting of one or two hundred people. This prompted us to conduct in-depth research, allowing us to understand the significance of AI in the entire pharmaceutical process, from molecular discovery to clinical trials. Therefore, we decided to begin incorporating AI into our own drug development process.

The benefits of using AI, in addition to reducing costs and shortening cycles, most importantly, lie in increasing the success rate of pharmaceutical R&D. Initially, we planned to have companies providing AI technical services assist us with development. To that end, we reached out to several AI pharmaceutical enterprises in Europe and the U.S. However, we later discovered that these companies could not offer the services we needed. Therefore, we decided to build our own AI team and use our self-developed AI platform for the development of our proprietary pipeline.

It took nearly two years to develop the first candidate drug using AI before obtaining FDA approval to directly enter Phase II clinical trials. After that,The development of the second drug took only half a year.，In this process, we fully realized that AI is indeed a very effective tool.

Our AI team initially included AI scientists, chemists, and professionals from other fields. However, after more than three years of hard work, we...Decide to focus AI applications on the clinical stage.

Q: What prompted Evergreen Therapeutics to choose the AI clinical direction?

Dr. Tao Du:Currently, the main focus of AI-driven pharmaceutical companies in China is mostly on molecules and targets, with very few companies engaged in AI clinical applications. After the pandemic ended in 2023, at China's first DIA conference, the heads of two departments from the FDA disclosed data on AI clinical applications within the U.S. FDA.Since 2021, the number of clinical trials designed by AI in the United States has actually exceeded three digits, and by 2023, it is estimated to be close to 200 clinical trials., which makes us more convinced that our initial choice of pursuing AI in clinical applications was correct. The reason we chose this path at the time was because we had offices abroad, which provided us with a wealth of overseas-related information. We learned that there are at least dozens of companies abroad utilizing AI technology for clinical purposes.

Currently, there are companies in Europe, the United States, South Korea, and Japan that use AI to provide clinical services. Moreover, in recent years, many multinational pharmaceutical enterprises have directly invested in AI clinical design companies. This shows that the use of AI in clinical settings is highly valued by large pharmaceutical companies.

Q: What are the application scenarios of AI in clinical settings?

Dr. Tao Du:Currently, Evergreen Therapeutics' AI clinical applications mainly focus on utilizing disease phenotypes and genetic information.Selection of clinical indications, optimization of clinical endpoints, and determination of qualified subjectsIn fact, AI technology can play a very significant role in clinical settings, ranging from selecting new indications for an old drug or choosing a combination of two drugs—AI can provide crucial assistance in these areas. Additionally, in terms of clinical execution, AI can help pharmaceutical companies identify hospitals with more patients that meet enrollment criteria and assist in clinical monitoring. These applications support decision-making by analyzing and summarizing large amounts of data. Relying solely on human effort for such tasks would hardly yield comprehensive results.

In addition to the several clinical application scenarios that Evergreen Therapeutics focuses on, there are actually other AI clinical application scenarios overseas. For instance, digital twin technology can simulate clinical records and comprehensively model clinical outcomes based on standard care, thereby achieving higher success rates with fewer patients (such as Unlearn AI in the United States). Moreover, it also includes leveraging real-world data (RWD) analysis in clinical development to assess trial feasibility and optimize trial protocols, thus accelerating the clinical process (such as Concert AI in the United States).

AI clinical is also a highly regarded investment hotspot abroad.Currently, there are dozens of companies worldwide engaged in AI clinical services. Among them, Tempus, located in Chicago, is an unlisted company, but its valuation has reached more than 8 billion US dollars. Therefore, we can see that AI entering the costly field of clinical practice has become inevitable and is becoming increasingly important.

I once thought that multinational pharmaceutical companies, with decades of data accumulation, could rely entirely on their own data for clinical trials. However, in reality, many large pharmaceutical enterprises use external AI clinical services. More notably, some multinational pharmaceutical companies (such as Eli Lilly, Johnson & Johnson, and Merck) not only sign service contracts with AI clinical companies but also make direct investments. Therefore, overall,The AI clinical track is actually in a state of rapid development abroad.I believe that in the future, there will be more companies in China participating in the field of AI clinical applications.

Q: What are the essential conditions for entering the AI clinical field? Can you share some experiences?

Dr. Tao Du:Before conducting clinical trials, it is mandatory to have a regulatory-approved clinical protocol and obtain consent from the regulatory authorities to confirm the accuracy of the trial protocol and begin execution. Therefore, companies abroad engaged in AI clinical applications typically either have founders who are doctors themselves (e.g., Oncocross in South Korea) or possess a reasonably large medical team (e.g., CytoReason in Israel, where clinical and biological experts construct computational models of the human body, enabling simulations of human diseases at both tissue and cellular levels).

The rapid development of Evergreen Therapeutics in the AI clinical field is largely due to our multidisciplinary R&D team, which integrates clinical medicine, AI science, and regulatory affairs professionals. To succeed in the AI clinical field, these three elements are indispensable.In the future, the application of AI in the clinical stage will become more standardized.Only through an understanding of regulations, regulatory processes, and effective control of key stages can the effectiveness of AI as a research and development tool be maximized, thereby increasing the success rate of drug development and approval.