The clinical trial is a critical stage of drug development workflow, with an estimated average success rate of about 11% for drug candidates moving from Phase 1 towards approval. Even if the drug candidate is safe and efficacious, clinical trials might fail due to the lack of financing, insufficient enrollment or poor study design. [Fogel DB. 2018].
Artificial Intelligence (AI) is increasingly perceived as a source of opportunities to improve operational efficiency of clinical trials, and minimize clinical development costs. Typically, AI vendors offer their services and expertise in the three main areas. AI start-ups in the first area help to unlock information from disparate data sources, such as scientific papers, medical records, disease registries, and even medical claims by applying Natural Language Processing (NLP). This can support patient recruitment and stratification, site selection, improve clinical study design and understanding of diseases mechanisms. As an example, about 18 % of clinical studies fail due to insufficient recruitment, as 2015 study reported.
Another aspect of success in clinical trials is improved patient stratification. Since trial patients are expensive - the average cost of enrolling one patient was $15,700-26,000 in 2017 -- it is important to be able to predict which patient will have greater benefit or risk from treatment. AI-driven companies operate with multiple data types, such as Electronic Health Records (EHR), omics and imaging data to reduce population heterogeneity and increase clinical study power. Vendors could use speech biomarkers to identify neurological disease progression, imaging analyses to track treatment progression, or genetic biomarkers to identify patients with more severe symptoms.
AI is also streamlining the operational processes of clinical trials. AI vendors help to track patient health from their homes, monitor treatment response, and patient adherence to the trial procedures. By doing that AI companies decrease the risk of patient dropouts, which accounted for 30% on average. Usually, the Phase 3 clinical study stage requires 1000-3000 participants, with a part of them taking placebo. That’s why the development of synthetic control arms - AI models that could replace the placebo-control groups of individuals thus reducing the number of individuals required for clinical trials - might become a novel trend.
Below we summarize a list of notable AI-vendors providing advanced tools for clinical development.
Informational and analytical engines
The company provides real-world evidence (RWE) services for precision oncology. It has established the broadest clinical network through the partnership and licensing with community oncology networks, thus getting access to Electronic Medical Records, Results of NGS diagnostics, and patient-reported outcomes. Concerto then analyses these data and generates evidence for new therapeutic approaches.
Saama claims it is the number one in clinical trial analytics and offers to provide clients with valuable clinical and operations insights and risk-based monitoring. Early this year, it partnered with Pfizer to deploy Saama’s Life Science Analytics Cloud platform, under the agreement Pfizer is providing clinical data to train Saama’s model.
PathAI is a supplier of AI imaging analyzing tools for pathology. It is mainly focused on cancer research with the aim to deliver precision medicine to every patient and help them to benefit from novel therapies. PathAI partners with leading life science companies, including collaboration with Bristol Myers Squibb, where PathAI worked on the evaluation of PD-L1 expression.
The company uses federated learning to train and develop its machine learning models specifically to increase clinical trial efficiency. They have built a catalog of 30 life models, enabling them to identify new biomarkers from imaging, genomics, and clinical data. As an example, Owkin is working on identifying patients with the most severe disease progression that might respond to the treatment.
GNS is a company from Cambridge that was founded in 2000. The company’s technology operates a wide variety of complex data to create in silico patients - accurate computer models of disease that enable modulation of individual drug response. GNS focuses on oncology, immunology, CNS, and cardio-metabolic disease. Their technology supports better patient stratification and can identify patients who should receive therapy in the first and second lines. GNS Healthcare validated its in silico patients through over 50 scientific publications.
It developed a mobile app that helps to monitor clinical trial adherence by analyzing videos made by the participants. The algorithms can recognize the participant's face, make sure that he is taking the right pill, swallowing it, and it isn’t hidden under the patient's tongue or cheek.
Unlearn.AI is a start-up from San-Francisco founded in 2017 by a former principal scientist at Pfizer. Unlearn is working on the concept of “digital twins” profiles, that integrate multiple data from real patients. Unlearn is creating these profiles through their DiGenesis platform with the aim to replace real patients in placebo control groups. This solution removes the need to engage real patients in placebo control, thus reducing the number of participants required for clinical studies. Currently, the company is focused on Alzheimer’s Disease and Multiple Sclerosis.
AI vendors are believed to provide tangible impact on the improvement of the clinical study process. Today, there is evidence that AI may accelerate patient enrollment, one study reported reduced patient screening time by 34 % and improved patient enrollment by 11.1 %. Additionally, IQVIA, an American multinational company, has reported a 20% increase in enrollment. On the other hand, AICure reported that using their platform increased the rate of taking prescribed medication from 72 to 90%. We believe that AI tools will continue to emerge, opening new opportunities for clinical trials to improve.
However, we are in the early days of AI adoption in this area and the efficiency of this new technology still has to be validated by more statistics and reported use cases. Moreover, we believe AI will not be able to completely resolve issues in clinical research: patients and doctors will still be needed as decision-makers in all major contexts.