Decoding diseases: The AI renaissance in clinical trials

AI clinical trials concept

Deepika Khedekar, Associate Centralized Clinical Lead at IQVIA, explains how artificial intelligence could make clinical trials more accurate, accessible and standardised.

Over 19 million1 hearts succumb to cardiovascular disease yearly, while around 10 million2 lives are claimed by cancer every year. To put this scale into perspective, these numbers surpass the combined population of the Netherlands, Hong Kong, and Singapore. While traditional approaches to clinical trials have helped us manage some of these conditions to a certain extent, they alone may not provide us the velocity we need to accelerate the research and conquer these diseases, raising the possibility that technologies such as artificial intelligence (AI) may hold the answer.

Tapping into the capabilities of AI

One reason AI may accelerate clinical trials is its ability to rapidly process and analyse vast amounts of data. By swiftly sifting through complex datasets, AI algorithms can identify patterns, extract valuable insights, and uncover subtle correlations that might elude human researchers. This efficiency expedites the identification of potential therapeutic targets, screening of drug candidates, and selecting eligible trial participants, all of which are crucial aspects of clinical trial design and execution.

AI’s role in drug discovery 

One notable example of AI’s role in clinical trials is a significant breakthrough made by researchers at the University of Toronto3 in partnership with Insilico Medicine in discovering a potential drug for liver cancer. They used an AI tool called AlphaFold to understand the structure of a protein called CDK204, which is associated with liver cancer. Then, with another AI system, Chemistry42, they generated molecules to interact with CDK20. Among these, the molecule ISM042-2-048 showed potential as a new liver cancer drug due to its effective binding with CDK20. This study was published in the journal Chemical Science5 and led by the University of Toronto Acceleration Consortium6 Director Alán Aspuru-Guzik, Nobel laureate Michael Levitt, and a team of researchers from Insilico Medicine. 

Harnessing AI, researchers have significantly expedited the drug discovery process, achieving what traditionally takes years in just 30 days. This method not only promises more efficient drug development but could also revolutionise clinical research and accelerate trials.

Participant diversity & enrolment 

Another example where AI can help solve challenges in clinical trials is by matching participants with relevant trials across the globe. Many trials fail because they cannot enrol a statistically significant number of participants. This was the challenge embraced by researchers at John Hopkins University.

Researchers at the John Hopkins University have partnered with Viz.ai, an AI-driven disease detection startup, to enhance patient enrolment for the National Institutes of Health (NIH) funded Biomarker and Edema Attenuation in IntraCerebral Hemorrhage (BEACH)7. Using Viz.ai’s Viz RECRUIT software, they can quickly identify suitable patients by analysing their brain scans and medical histories in real time. This method not only ensures continuous recruitment but also broadens the participant pool, emphasising diversity. Such AI applications hold promise not just in expediting enrolment but also in potentially improving trial outcomes.

AI-enabled tumour profiling 

Furthermore, scientists from Harvard Medical School8 have pioneered an AI tool named CHARM (Cryosection Histopathology Assessment and Review Machine), which deciphers the DNA of brain tumours in real-time during surgery. Unlike traditional methods that require days or weeks for molecular profiling, CHARM offers immediate insights into a tumour’s molecular identity, aiding surgical decisions about tissue removal and the direct administration of tumour-targeting drugs. Traditional methods, which involve freezing brain tissue for microscopic examination, can distort tissue appearance and compromise accuracy. In contrast, CHARM extracts biomedical signals from frozen pathology slides, enhancing real-time precision oncology. This advancement can refine clinical trials by improving participant selection based on molecular profiles, facilitating personalised treatments, and offering real-time tumour monitoring.

Challenges on the horizon

Although AI can help us overcome many challenges that impede clinical trials, certain risks cannot be addressed solely by using AI in these trials. For instance, adequate funding is paramount for the seamless operation of a trial. This funding ensures staff salaries, the procurement of necessary software and hardware, the maintenance of modern laboratory infrastructure, and, in some instances, compensation for trial participants. Without sufficient financial support, even the most promising trials may falter.

Global clinical regulations vary significantly, leading to disparities in medical treatment approvals. A treatment greenlit in one country might face delays or even non-approval in another. Additionally, accessibility issues arise as most clinical trials are concentrated in major cities, leveraging their advanced infrastructure to ensure participant safety and optimise resource use. However, this urban focus can inadvertently exclude potential participants from remote areas, presenting its own set of challenges, such as a lack of participant diversity in clinical trials. 

Collaboration and standardisation

We can explore one centralised solution to address clinical trial challenges: collaboration and standardisation.

Addressing regulatory discrepancies across countries requires a harmonised approach. Regulatory bodies can form a global consortium to standardise drug evaluation and approval processes. By prioritising treatments for diseases with high mortality rates a standardised approval process can be established. This ensures that a treatment approved in one country becomes rapidly accessible in another. Over time, this framework can be extended to other diseases, streamlining global treatment approvals.

Challenges related to lack of funding, accessibility, diversity, and resource constraints can be addressed by designing and implementing Synergistic and Centralized Group-based Collaborative Models For Clinical Trials (SYNCRO-CT). This novel method calls for a change in how trials are designed and implemented. The goal here is not to view each trial as an isolated entity addressing specific diseases. Instead, we can categorise these trials into distinct groups. In this case, each group denotes a set of trials targeting a specific disease or health condition. This approach allows us to ensure a balanced distribution of participants and resources across each trial group, promoting efficiency and fairness in our research efforts. 

Furthermore, to mitigate the issue of trial accessibility in remote areas, a hybrid model of in-person and remote trial participation can be introduced. 

People and technology are at the heart of each trial, and hence, establishing synergies, enabling global harmonisation of regulatory standards, and technology adoption holds the potential to address the challenges that obstruct the progress of modern clinical trials.

DDW Volume 24 – Issue 4, Fall 2023

References

  1. 2022 Heart Disease & Stroke Statistical Update Fact Sheet, American Heart Association.
  2. Our World In Data, Cancer by Max Roser and Hannah Ritchie
  3. Steinberg B. AI develops cancer treatment in 30 days, predicts survival rate. NY Post 20 March 2023.
  4. Rosso C. AI Finds Drug Candidate for Liver Cancer in 30 Days. Psychology Today 25 January 2023.
  5. Ren F, Ding X, Zheng M, et al. AlphaFold accelerates artificial intelligence powered drug discovery: efficient discovery of a novel CDK20 small molecule inhibitor. Chem Sci 2023;14:1443-145.
  6. Warner E. Researchers use AI-powered database to design potential cancer drug in 30 days. University of Toronto, 19 January 2023.
  7. Viz.ai Collaborates with Johns Hopkins to Expedite Patient Enrollment in Brain Injury Trial. Businesswire, 14 February 2023.
  8. Pesheva E. AI Tool Decodes Brain Cancer’s Genome During Surgery. Harvard Medical School, 7 July 2023.

About the author:

Deepika KhedekarDeepika Khedekar is a Clinical Trial Lead at IQVIA. In her 12-plus years in the pharmaceutical industry, she has led multiple clinical trial programmes for leading US and Australia-based pharmaceutical organisations, such as Gilead Sciences, Macleods Pharma, Arrowhead Pharmaceuticals and Impact Pharma. She started her journey in the field of pharmaceutical research at Pfizer and holds a Master’s degree in Pharmacy from the University of Mumbai. 

Related Articles

Join FREE today and become a member
of Drug Discovery World

Membership includes:

  • Full access to the website including free and gated premium content in news, articles, business, regulatory, cancer research, intelligence and more.
  • Unlimited App access: current and archived digital issues of DDW magazine with search functionality, special in App only content and links to the latest industry news and information.
  • Weekly e-newsletter, a round-up of the most interesting and pertinent industry news and developments.
  • Whitepapers, eBooks and information from trusted third parties.
Join For Free