Artificial Intelligence (AI) in Clinical Trials

Drug development is a very time taking process. Launching a new drug in the market takes an average of 10-15 years and USD 1.5-2.0 billion. Clinical trials take 6-7 years time and a significant amount of financial investment to ensure the safety and efficacy of a drug product for a specific disease condition in humans. Conventional clinical trial methods are time-consuming, expensive, and error-prone, but a new approach enabled by artificial intelligence (AI) can make them faster, more efficient, and more effective. 

This article provides an overview of the various applications of AI in clinical trials and examines the potential benefits and challenges and the future of AI in Clinical Trials.

Application of AI in Clinical Trials

Clinical trial design:

Designing a clinical trial is a challenging task. There are many factors to consider, such as the selection of patients, the choice of treatment, and the determination of appropriate endpoints. One of the main challenges in clinical trial design is patient selection. Clinical trials need to enroll patients who satisfy specific inclusion and exclusion criteria. Another challenge is the choice of treatment and endpoint selection. 

AI can help to identify additional factors that can be used to select patients, such as genetic markers or biomarkers. Machine learning (ML) algorithms can analyze large datasets to identify patient characteristics that are associated with positive treatment outcomes, enabling the selection of patient populations that are more likely to benefit from the treatment being tested. Additionally, it can analyze the data from previous studies and helps a doctor to choose the most suitable treatment option(s). AI analysis gives an obvious idea of whether a particular potential drug or combination of the drugs will be more likely to be effective. AI can also help to optimize dosing schedules, which is critical to achieve from a clinical trial with maintaining maximum efficacy and minimizing toxicity.

Furthermore, AI can assist in selecting endpoints by identifying clinical and biological factors that are associated with positive outcomes. By analyzing large datasets, AI algorithms can identify novel endpoints that may not have been considered previously.

Patient Recruitment:

Patient recruitment is a crucial aspect of clinical trials, and also one of the most challenging tasks. AI- and ML-driven systems can assist with patient recruitment by automating the process of finding matches between patients and recruiting trials. These systems can streamline the recruitment process and increase the efficiency of clinical trials.

AI techniques such as natural language processing, reasoning, and ML spontaneously help  to analyze electronic medical records and clinical trial eligibility databases, find matches between patients and recruiting trials, and recommend these matches to doctors and patients.

These AI-based clinical trial matching systems have been successfully demonstrated and have proven their value in real-life cases. Additionally, AI and ML techniques can proactively identify potential matches between trials and specific patients from publicly available web content. The integration of AI will help not only to improve digital patient enrolment and efficiency of the trial but also to reach the trial data easily to the health care providers and common people.   

Patient Monitoring:

Another challenge in a Clinical Trial is the patient dropping out. Patient adherence to medication and other treatment protocols during clinical trials is highly challenging, resulting in high dropout and non-adherence rates. AI can be used to monitor patients during clinical trials, alerting investigators to potential safety issues and helping to reduce the risk of adverse events.

AI techniques with wearable technology open new possibilities for developing power-efficient, mobile, real-time, and personalized patient monitoring systems. Wearable sensors and video monitoring can collect patient data automatically and continuously. ML models can then be used to analyze such data in real-time to detect and log relevant events. AI and ML methods can also dynamically predict the chances of drop-out for a specific patient and detect the onset of patient behavior that suggests the patient may be having difficulties adhering to the study protocol.

Data analysis:

Data analysis in clinical trials can be challenging due to several factors, including the complexity of the data, the need to ensure data accuracy and integrity, and the need to comply with regulatory requirements. AI algorithms can be trained for identifying patterns and relationships within complex datasets, helping researchers to  understand better the underlying biology of disease and the effects of different treatments. AI can also help to improve the accuracy and integrity of clinical trial data by automating data cleaning and curation processes.

Predictive modeling:

AI can be used to predict the outcomes of clinical trials by analyzing large amounts of data, including demographic data, genetics, and previous trial results. It will help to speed up the trial process and improve the chances of success. Using ML algorithms and natural language processing, AI can analyze electronic health records, patient-generated data, clinical trials, and other sources to identify patterns and relationships between different factors. It can improve trial design, reduce trial costs, accelerate drug discovery, and improve patient outcomes and safety.

Potential Impact of AI on Clinical Trial

AI has very potential impact and enhances patient outcomes in several ways:

• Improving Efficiency: Automation of repetitive tasks reduces the time and resources required to complete trials, and launching new treatments to market more quickly.
• Enhancing Accuracy: AI algorithms can analyze vast amounts of data in real time, with more accurate and timely insights into patient health. 
• Improving Patient Outcomes:  AI is playing an innovative role to develop personalized treatment plans and enhancing patient engagement.  

Conclusion

AI has shown tremendous potential in revolutionizing clinical research by enhancing efficiency, accuracy, and patient outcomes. While AI has the potential to transform clinical trials, there are several limitations to its use. It includes maintaining data privacy and security, bias in the algorithm, validation, regulatory approval, cost, and limited experience and understanding.

In conclusion, the integration of AI into clinical trials is transforming the field of medicine and having a profound impact on the way that trials are conducted, data is analyzed, and patient outcomes are improved. Despite significant challenges, AI in clinical trials is a field  of continuous growth and innovation, as researchers and practitioners explore new and exciting ways to leverage AI and other digital technologies to improve the efficiency, accuracy, and safety of clinical trials, and to enhance patient outcomes.

References 

1. Artificial intelligence in clinical trials: A review” by Charles A. Dinarello, Josef Penninger, and Goran Milojevic was published in the journal Clinical and Translational Medicine in 2018, Vol. 7 (1)
2. “The use of artificial intelligence in clinical trials” by Min Zhang, Qingyun Wu, and Xiaoyan Li was published in the journal Journal of Medical Systems in 2018, Vol. 42 (9)
3. Harrer, S., Shah, P., Antony, B. J., & Hu, J. (2019). Artificial Intelligence for Clinical Trial Design. Trends in Pharmacological Sciences, 40(8), 577–591. https://doi.org/10.1016/j.tips.2019.05.005
4. Angus, D. C. (2020). Randomized Clinical Trials of Artificial Intelligence. JAMA, 323(11), 1043. https://doi.org/10.1001/jama.2020.1039
5. Hariry, R. E., Barenji, R. V., & Paradkar, A. (2022). Towards Pharma 4.0 in clinical trials: A future-orientated perspective. Drug Discovery Today, 27(1), 315–325. https://doi.org/10.1016/j.drudis.2021.09.002

Author : Neha Shrivastava
Consultant Regulatory Affairs and Pharmacovigilance , Clinchoice