Artificial intelligence (AI) is rocking the healthcare industry. With its applicability to drug discovery, medical imaging, disease modeling, and conducting clinical trials, we promise to revolutionize the way we conduct research, treat diseases, and collaborate with patients.
In drug discovery, we’ve seen some of the perceptions behind AI hype and early demonstrations that enable targeting and pipeline development. AI can also support diagnostic decisions in the medical imaging field, read scans with exceptional speed and accuracy, and detect anomalies invisible to the human eye.
AI-enabled disease modeling, on the other hand, provides a deeper understanding of the etiology, transmission, and progression of diseases such as motor neuron disease, cancer, and HIV. However, one of the most promising frontiers in this area is conducting clinical trials and increasing the likelihood of regulatory or technical success.
Several different factors need to be carefully adjusted to increase the chances of a successful clinical trial. Clinical trial sponsors are looking for solutions that minimize the timeline while maximizing results. There are a variety of operational and scientific decisions that must be made during the clinical trial process, from site selection to endpoint selection that reduce the risk of clinical trials and can lead to more successful results. AI is increasingly being used to solve some of the challenges facing research teams, such as operations, science, and ethics.
AI creates practical operational insights
From an operational perspective, study sites can differ in terms of performance, especially with respect to the speed and variety of patient enrollment. Through AI analysis, sponsors and contract research organizations (CROs) leverage historical or real-world data to better understand site performance and to be more informed about time and resource allocation. You can make a decision.
This knowledge and oversight can shorten development schedules and ultimately benefit patients. The use of this AI was especially important in the face of Covid-19. There, AI is an influx of Covid patients. Although still in its infancy, AI is being used to assess patient availability and diversity data, allowing sponsors and CROs to reduce the risk of decision making in highly competitive situations.
Scientific hypotheses can be pressure tested by AI
Recipes for successful trials require a deep understanding of the disease in question, the patient population it affects, and potential treatments. Historically, this has been achieved through a review of the scientific literature and previous clinical studies.
AI is currently being used to enhance the intelligence that underpins testing. Analyzing multiple input sets, including historical trial design, drug biology, sponsor characteristics, and clinical trial results for the entire development program, can clarify the protocol and accurately predict the success of the trial.
In particular, incorporating real-world data along with clinical trial data can provide deeper clinical insights into patient outcomes and improve risk monitoring. It can also support endpoint selection decisions and better equip sponsors and CROs to target the best and most clinically relevant endpoints possible. AI is also used to flag real-time trends that emerge in trials that may not have been apparent until the end of the study in which all data were analyzed.
AI that supports a wider variety of trials
Another challenge that has long plagued clinical trials is the lack of diversity of study participants. From both a scientific and ethical point of view, it is essential to address the underestimation of a particular population within the trial. Studies that do not accommodate different ethnicities, ages, genders, and lifestyles do not provide effective treatments that represent the patient population.
AI can help fill this gap by identifying the best trial sites to serve the underrated community. By simulating a patient model, you can reach specific conclusions and hypotheses about the proportion of patients in a subgroup that respond to a specific treatment. This will give you an idea of what your clinical trial team thinks about recruitment and recruitment diversity. However, those involved in the development and adoption of AI systems need to pay close attention to dismantling rather than recreating prejudices in the collection and use of data. This involves building a model that can be translated into a wide range of epidemiologically representative populations. Regulation continues to play a role in shaping an approach to risk management, data proofing, and mandatory transparency.
Synthetic control arm as a powerful data support tool
The Synthetic Control Arm (SCA), also known as the External Control Arm, is another innovative tool enabled by big data, powerful computing, and advanced analytics. While AI helps mimic real life, SCA uses real-life patient-level data and biostatistical techniques to replicate the control arm, eliminating the need for placebo groups.
Like AI, these advanced statistical methods and analyzes require large amounts of data to accurately emulate real life. It should be noted that while established biostatistical approaches may deviate from the definition of “AI”, traditional methods combined with high quality data show great expectations and success in a regulatory environment. Is important.
Beyond diversity, recruiting patients presents other challenges. In particular, the time pressure to recruit as soon as possible, and the ethical impact of recruiting control arms for trials in conditions where effective treatments such as many may not be available, are rare illnesses. Synthetic control arms can proxy patient-level data for real clinical trials and provide representative datasets that provide valuable information about a disease, indication, or treatment.
In addition, the model can be run repeatedly. This means that you can run dynamic datasets in different analyses and model several different results. The submission of a small number of synthetic control arms has been approved by the FDA. This includes recurrent glioblastoma, a hybrid design of phase III trials in diseases with few treatment options and high unmet needs. SCA is just one of a myriad of advanced analytical tools and statistical methods with great potential in the clinical phase of drug development.
Possibility of untapped AI in clinical research
By harnessing the power of AI, we were able to gain a better understanding of illnesses, patient populations and potential treatments. Technology is transforming the way we conduct clinical trials. It improves the elements of study design, such as target population selection, comparator arms, and clinical endpoints. It also improves patient safety and patient enrollment and provides pharmaceutical companies with important insights and analysis of how their drugs work. But we only bite the surface of what we can really achieve. The possibilities are enormous, and there is no doubt that AI will become an indispensable element in clinical research and drug development in the future.
Photo: Blue Planet Studio, Getty Images
..
