How AI and Machine Learning Can Bring Quality Improvements in Biopharma

Artificial intelligence (AI) and machine learning (ML) are easily confused, but it is crucial to recognize the fundamental difference between them. AI refers to prebuilt products that identify patterns based on human behavior around recognition and decision-making that utilize those patterns to enable platforms to answer questions, provide relevant information, or perform requested tasks. ML, meanwhile, employs mathematical algorithms to predict activities or outcomes based on data. ML is a subset of AI, meaning that although all ML derives from AI, not all AI is based on ML.¹

Abney

The pharmaceutical industry has found applications for both. ML’s ability to swiftly analyze vast amounts of data has greatly contributed to disease identification, leading to accelerated diagnoses and improved, life-saving outcomes. AI, on the other hand, is becoming a real force in drug discovery and manufacturing. It elevates productivity, efficiency, and production speed, enabling pharmaceutical companies to launch products faster and at reduced costs.²

This ongoing sea change will have far-ranging impacts beyond drug development and delivery. The impacts on numerous industries and the market value of the technology itself will only increase. The global market for AI-based software is projected to be $126 billion by 2025, up from $10.1 billion in 2018. The main drivers and challenges in the pharmaceutical industry, besides developing new drugs, include meeting rising patient expectations, addressing incurable diseases, and navigating more complex intellectual property rights.²

Although the rapid development of AI and ML implies escalating costs, forward-thinking and market-savvy executives recognize that the benefits of AI/ML-driven predictive analytics, such as greater precision, accuracy, and insights, outweigh the initial expenditure.

The total cost of researching, testing, perfecting, and validating a single drug concept can reach $2 billion. However, by making drugs more effective earlier in the development process, the chances of success increase while reducing costs associated with clinical trials. This reduction in trial and error leads to faster FDA approval and eventually lower research and development expenses.²

Traditional clinical trial processes face limitations such as fragmented data and systems in multiple formats, extensive manual data transcription, repetition of essentially identical data components, and complexities in integrating data from new sources. Patient-related challenges include recruitment and enrollment, monitoring adherence and retention, as well as diversity among clinical trial participants.³

AI and other innovative technologies that gather, process, and analyze data from multiple sources can drive more precise and targeted treatments. These technologies also have the potential to shift the health care ecosystem from the status quo to one that is more personalized, predictive, preventive, and participatory. These shifts will significantly affect patient outcomes in the next decade, particularly in underserved areas, communities, and patient populations.⁴

The advantages of utilizing AI and ML in clinical trials include the positive impacts of collecting trial data, improving the flow of information digitally, and utilizing high-powered wearable devices to enhance participant recruitment. Remote monitoring increases convenience and boosts retention, whereas algorithms track and predict patient behaviors, enabling more meaningful interactions. Efficiently conducted trials reduce time and money, because smart automation decreases the need for rework in processing clinical trial data.

Additionally, AI and ML technologies enable organizations to reuse existing data, eliminating the need to develop new databases for each trial.³

The speed at which AI processes data far surpasses human capabilities, as demonstrated in Pfizer’s partnership with IBM on Watson for Drug Discovery. Instead of a human researcher reviewing up to 300 articles a year, Watson can examine more than 25 million Medline article abstracts and 1 million medical journal articles. This highly accelerated process allows researchers to identify promising prospects earlier, leading to the introduction of new drugs and treatments to the market at a similarly fast pace.⁵ Academia’s utilization of AI-driven modeling and simulation in drug research and discovery has already begun and is expected to increase significantly by 2030.⁴

The pharmaceutical industry has long embraced innovative tools and technologies to ensure the safe and effective delivery of drugs to patients. The recent COVID-19 pandemic underscored the constant sense of urgency in developing vaccines, and the utilization of digital technologies and tools has empowered the pharmaceutical sector to achieve significant milestones in that area.²

AI and ML have played pivotal roles in modernizing the industry and generating impactful outcomes. Over the past 5 years, these innovative technologies have revolutionized the development of new drugs, disease treatments, and therapies. Their power, speed, agility, and utility will only continue to increase. A recent report stated that 70% of surveyed businesses recognize the importance of AI in their survival and growth,⁶ and this significance is particularly meaningful for pharmaceuticals, an industry in which innovation plays a more profound role than in others. AI/ML has emerged as an invaluable catalyst for the research and development of new pharmaceutical products, and its influence will persist and grow for many years to come.

The utilization of AI and ML in drug development presents significant cost benefits. Despite the rapid development and escalating costs associated with research and testing, the ability of AI/ML-driven predictive analytics to achieve greater precision, accuracy, and insights far outweighs initial costs. The integration of AI/ML technologies can make drugs more effective earlier in the development stage, increasing the chances of success and reducing the costs of clinical trials. Viable solutions can be brought to the market faster, and research and development expenses are reduced.

Moreover, AI and ML have the potential to address limitations in clinical trials and improve patient outcomes, especially in patient groups that have long needed and awaited improvements. Furthermore, the speed and efficiency of AI identifies products with the most potential, expediting the introduction of new drugs and treatments to providers, patients, caretakers, and families. The continued adoption of AI and ML will undoubtedly shape the future of the pharmaceutical industry even more than it already has, leading to even greater efficiencies, better patient care, and cost savings for all stakeholders. 

Author Information
Bryan Abney, MS, is vice president of quality assurance and compliance at Syner-G Biopharma Group.

References
1. Rosenberger S. Growth of artificial intelligence in pharma manufacturing. Genetic Engineering & Biotechnology News. January 12, 2023. Accessed August 21, 2023. https://www.genengnews.com/artificial-intelligence/growth-of-artificial-intelligence-in-pharma-manufacturing/
2. Reddy SS. How AI and ML is driving value for global pharma players. Journal of mHealth. August 9, 2021. Accessed August 21, 2023. https://thejournalofmhealth.com/how-ai-and-ml-is-driving-value-for-global-pharma-players/
3. Lingler N, Karia S. Using AI to accelerate clinical trials. Deloitte. February 24, 2022. Accessed August 21, 2023. https://www2.deloitte.com/us/en/blog/health-care-blog/2022/using-ai-to-accelerate-clinical-trials.html
4. Kudumala A. AI in pharma and life sciences. Deloitte. Accessed August 21,2023. https://www2.deloitte.com/us/en/pages/life-sciences-and-health-care/articles/ai-in-pharma-and-life-sciences.html
5. Will AI play an important role in quality assurance and quality control careers? Academy of Applied Pharmaceutical Services. November 14, 2017. Accessed August 21, 2023. https://www.aaps.ca/blog/will-ai-play-an-important-role-in-quality-assurance-and-quality-control-careers
6. 70% of leaders say AI critical for business in next three years. Moonshot. February 6, 2023. Accessed September 24, 2023. https://moonshot.news/news/technology-news/70-of-leaders-say-ai-critical-for-business-in-next-three-years/