An MIT study on clinical trial success rates found that the overall probability of success for all drugs and vaccines is 13.8 percent. Two-thirds of trial failures occur due to inadequate efficiency, flawed design, inappropriate endpoints, or under-enrolment.
While many professionals are aware of the use of Artificial Intelligence (AI) to support clinical data analytics, the full range of its potential is less well known. AI has the potential to transform many of the key steps in clinical trials, from protocol design to study execution, reduce the cost of conducting clinical trials and increase the probability of studies’ success. AI can also help enterprise and mid-market sponsors, as well as contract research organizations (CROs), analyse and develop strategies from vast amounts of historical data, which can also be a good database for future trials but an impossible manual task.
Improving clinical operations in trial design
When we think of AI in clinical trials, we usually imagine using data analytics after data is already collected. However, AI can be equally if not more important in the pre-trial phase. Trials leveraging AI technology and big data in the planning phases can optimize trial design. Through measuring and benchmarking site burden, real world evidence can help researchers identify target patient populations, allowing for better matching of trials to sites with required site characteristics. In a recent survey of clinical trial sites, respondents identified patient recruitment and enrolment as their top two concerns. Patient recruitment has also been called the “biggest challenge” in clinical trials.
Imagine if you had started a clinical trial for a rare disease treatment but realize midway through the patient recruitment process that incidence rates are lower than expected in the country. We saw this last April, when several COVID-19 vaccine clinical trials in China had to be cancelled due to shortage of patients. This was due to a combination of factors, including that not many of the existing patients in China met study requirements (age, gender, and prior health conditions) and a lack of total patient population for large-scale advanced drug studies. Restarting and relocating clinical trials requires time, costs, staffing, and more. Identifying optimal countries and sites with the help of AI and big data can help in balancing speed, cost, and quality. It can also predict enrolment at site, country, and trial levels.
However, it is not industry standard for researchers to consider implementing AI at the pre-trial stage. Especially among enterprise mid-market sponsors, traditional operating processes and lack of resources may limit access to clinical trial databases and intelligent trial solutions.
Single arm trials as a practical solution to patient recruitment
Technologies such as single arm trials can leverage available standardized data from historical trials to increase sample size and offer a practical solution to studies where randomization is not feasible. This can be especially beneficial for researchers targeting patients with life-threatening diseases who are dissuaded from participating in a trial due to the possibility of landing in a control arm, such as placebo or ineffective standard-of-care treatment. Additionally, research studies targeting small or hard-to-reach patient populations such as rare diseases may find it even more challenging to recruit a substantial pool of participants. Compounded with the inevitability of some patient incompletions, data collection from these patient populations are further limited.
While randomized controlled trials are the gold standard for evaluating the safety and efficacy of new medical treatments, maintaining a concurrent control arm is sometimes not feasible and can lead to increased patient burden and threaten the completion of a trial. Instead of collecting data from patients recruited for a trial who have been assigned to the control or standard-of-care arm, synthetic control arms use real-world data that has previously been collected from sources such as health data generated during routine care, including electronic health records, claims data, patient-generated data, and historical clinical trials data.
With AI technology simulating trial data, more patients can be treated with the experimental therapy. A traditional, randomized controlled trial needing 1,000 patients to demonstrate the effectiveness of a new therapy — 500 for the active arm, 500 for the control arm — need now only recruit 500 participants when a synthetic control arm is employed. By reducing or eliminating the need to enroll control participants, a synthetic control arm can increase efficiency, reduce delays, lower trial costs, and speed lifesaving therapies to market.
According to a Deloitte survey, top biopharma companies ranked synthetic control arms as one of the most impactful future applications of real-world evidence. Despite this recognition, 70 percent of companies in the survey noted that the lack of research-grade data is hindering efforts to leverage real world evidence in R&D. Greater partnerships with health systems and other stakeholders is critical in curating purpose-built data sets, for SCA and related technology to ultimately better inform clinical development.
In addition, scepticism from regulators and lack of methodology guidelines and standards on such innovative technologies trickle down to reluctance from researchers. Real world evidence and synthetic arms could also play a role in evaluating existing and repurposed drugs that could be effective against Covid-19 and similar cases in the future. Fortunately, attitudes are changing along with the pandemic, as shown through its approvals to be used in Covid-19 vaccine trials.
Even when you work tirelessly to get the best data for your clinical development program, you can still end up with data gaps and unconnected data that can jeopardize the success of clinical trials. Real world evidence and AI technology can help deliver insights to increase efficiencies and probabilities for success. Yet, apart from a small segment of large companies, most biopharma companies lack the internal knowledge, expertise, and resources to effectively leverage real world evidence and AI technologies to improve business and research outcomes. Nonetheless, the notion that AI and real-world evidence are trends for the near future is outdated; AI in clinical trials is what we need now – AI enabled technologies can be a necessary tool for clinical trials during and beyond this pandemic. Transforming the industry requires partnerships and collaborations across the ecosystem to ensure sustainable AI adoption and application. [APBN]
References:
- Wong, C. H., Siah, K. W., & Lo, A. W. (2018). Estimation of clinical trial success rates and related parameters. Biostatistics, 20(2), 273–286. https://doi.org/10.1093/biostatistics/kxx0692.
- Shim, H. (2021, April 21). Medidata’s Acorn AI solution raises clinical trials’ success rates. Korean Biomedical Review. http://www.koreabiomed.com/news/articleView.html?idxno=109833. Oliver, P. (2021, May 14).
- Clinical Trial Sites Share Frontline Insights into COVID-19 Impact. Medidata Solutions. https://www.medidata.com/en/life-science-resources/medidata-blog/clinical-trial-sites-share-frontline-insights-into-covid-19-impact
- Rutter, K. (2018, October 9). Why patient recruitment is the biggest challenge in clinical trials – INDUSTRY VOICES. Informa Connect. https://informaconnect.com/patient-recruitment-challenge-clinical-trials-industry-voices/
- Zhuang, P. (2020, April 17). China cancels coronavirus clinical trials due to shortage of patients. South China Morning Post. https://www.scmp.com/news/china/society/article/3080453/china-cancels-coronavirus-clinical-trials-due-shortage-patients
- Huang, J. (2020, August 2). Chinese drugmakers turn to COVID-19 drug trials overseas amid infection surge. S&P Global Market Intelligence. https://www.spglobal.com/marketintelligence/en/news-insights/latest-news-headlines/chinese-drugmakers-turn-to-covid-19-drug-trials-overseas-amid-infection-surge-595543247.
- https://www.medidata.com/wp-content/uploads/2020/04/Acorn-AI-Intelligent-Trials-Fact-Sheet.pdf
- https://www.medidata.com/wp-content/uploads/2018/10/ASCO-Annual-Meeting_Presentation_20170531_Medidata-2.pdf
- Groth S.W. Honorarium or coercion: use of incentives for participants in clinical research. J N Y State Nurses Assoc. 2010;41(1):11
- Goldsack, J. (2019, February 5). Synthetic control arms can save time and money in clinical trials. STAT. https://www.statnews.com/2019/02/05/synthetic-control-arms-clinical-trials/
- Shah, S. (2020, May 7). Race to Develop COVID-19 Vaccines/Treatments Could Transform the R&D Process in Biopharma. Deloitte United States. https://www2.deloitte.com/us/en/blog/health-care-blog/2020/race-to-develop-covid-19-vaccines-treatments-could-transform-research.html
- Shah, S. (2020, May 7). Race to Develop COVID-19 Vaccines/Treatments Could Transform the R&D Process in Biopharma. Deloitte United States. https://www2.deloitte.com/us/en/blog/health-care-blog/2020/race-to-develop-covid-19-vaccines-treatments-could-transform-research.html
- Pfizer & BioNTech. (2021, March 11). Real-World Evidence Confirms High Effectiveness of Pfizer-BioNTech COVID-19 Vaccine and Profound Public Health Impact of Vaccination One Year After Pandemic Declared [Press release]. https://www.pfizer.com/news/press-release/press-release-detail/real-world-evidence-confirms-high-effectiveness-pfizer
- Medidata Solutions. (2020). The State of Real-World Evidence in Biopharma. https://www.medidata.com/wp-content/uploads/2020/05/State-Real-World_White-Paper_20200527.pdf
About the Author
Edwin Ng, Senior Vice President, General Manager, APAC, Medidata Solutions Edwin joined Medidata in 2014 and heads the Life Science industry of Dassault Systemes across the Asia Pacific region. Under Edwin’s leadership, the company’s business in the region has recorded significant year on year growth and achieved multiple leadership positions and awards in different countries. With the push for transformation in the healthcare sector through digitalization, Medidata is committed to serve our sponsors and partners by continued investments in the APAC sales, marketing, solution consultant, professional services and customer support teams. The Medidata, BIOVIA and other Dassault Systemes combined applications across research, development, quality, manufacturing, and clinical trials create a broader range of end-to-end solutions for our customers, and we are excited to extend the reach of our platform beyond clinical trials to research and development, commercialization, and manufacturing. Edwin has spent the last 26 years of his career managing enterprise solutions businesses, including at companies such as Dell, APJ Hewlett-Packard, StorageTek and Sun Microsystems.Edwin is a graduate of Nanyang Technological University in Singapore with a degree in Bachelor of Applied Science (Computer Engineering).