Accelerating AI Adoption in Regulatory Drug Development

Accelerating AI Adoption in Regulatory Drug Development

Patterns and Principles to Guide Pharmaceutical Sponsors

The purpose of the conversation is to introduce and review current methodologies, opportunities, challenges, and best practices to address the rapidly changing landscape of artificial intelligence (AI), augmented intelligence (IA), and machine learning (ML) in the setting of regulatory drug development (specifically regulatory CMC) and regulatory compliance.

CMC in drug development stands for Chemistry, Manufacturing, and Controls, which refers to the part of the drug development process that deals with the design, development, and production of a pharmaceutical product. This includes all aspects of the drug product’s development and production, including the synthesis of the active ingredient, formulation, quality control, and manufacturing processes.

AI, on the other hand, stands for Artificial Intelligence, which is a branch of computer science that deals with the development of algorithms and computer systems that can perform tasks that typically require human intelligence, such as visual perception, speech recognition, decision-making, and language translation. AI has the potential to revolutionize many industries, including healthcare, finance, and manufacturing.

Objectives/Topics to touch on:

AI has the potential to significantly impact regulatory affairs strategy building and submission authoring by improving efficiency, accuracy, and consistency.

• Efficiency: AI can automate many routine tasks, such as data collection and analysis, freeing up regulatory affairs professionals to focus on more strategic and complex tasks. AI can also help speed up the review and approval process, reducing the time to market new products.
• Accuracy: AI can help identify potential regulatory issues early on and provide accurate and up-to-date information on regulatory requirements, reducing the risk of errors and increasing the chances of approval.
• Consistency: AI can help ensure consistency in the format, style, and content of regulatory submissions, reducing the risk of inconsistencies that could lead to delay or rejection.

Overall, AI has the potential to revolutionize the way regulatory affairs professionals approach strategy building and submission authoring, making the process faster, more accurate, and more efficient (and dare I say less complex and costly). As the use of AI continues to grow in the regulatory affairs space, it is likely that companies that embrace this technology will be better positioned to succeed in a rapidly changing landscape.

The drug development process is rife with risk and complexity, yet digital solutions are paving the way forward. As AI and ML technologies are adopted and integrated into drug research and development pipelines, pharmaceutical companies can unlock a treasure trove of opportunities.

From finding patterns otherwise lost by conventional data analysis to streamlining essential processes (de-risking strategies and submission authoring), the power of AI can lead to greater chances of approval, more efficient use of resources, and a reduced cost for crucial stages of drug development.

The Use of AI and ML in Regulated Drug Development: Paving the Way for Streamlined Strategy, Improved Accurate Submissions, and Reduced Costs

As traditional challenges persist in the drug development process, AI and machine learning technologies are beginning to bridge the gap. From deep learning for image recognition to natural language processing (NLP) for extracting valuable data from GMP source data, these solutions are being used to streamline investigational (IND) and marketing applications (NDA/BLAs), improve the accuracy of data, and ideally increase speed to market.

With every advancement in these transformative technologies, sponsor companies can rely on more reliable insights while strengthening their chances of success — paving the way toward a brighter future in drug development.

AI/ML applications are used in drug development to identify patterns, find relationships and make predictions from data. The quantities of data generated can be difficult for sponsor companies to manage and interpret due to data silos and disparate systems. However, predictive AI models are allowing for greater efficiency in regulatory operations.

By automating many parts of the regulatory process through data mining, analysis, and interpretation, AI-enabled applications are helping to reduce time and resource costs while significantly speeding up the process. With ever-growing capabilities in this field, AI is rapidly becoming a critical asset for sponsors striving for improved accuracy and speed.

AI-based applications and conversational AI technology are increasingly being used to improve operations. Additionally, chatbots can be used specifically to help enhance engagement and reduce subjectivity while supplementing strategy and management of data

The adoption of AI-based solutions has been on the rise over the past few years and is likely to continue. However, there are several challenges that must be navigated in order to make this process successful. Data silos pose a formidable challenge, as they limit data access and opportunities for secondary uses.

The rise of AI and ML in drug development

Recent technologies such as data fabrics, data meshes, and knowledge graphs offer new ways to connect heterogeneous datasets with AI/ML models. Additionally, federated learning provides companies with an opportunity to maintain privacy while working across datasets. Increased shared data trends through publication requirements can further open pathways for a better understanding of operations and improved outcomes.

The rise of AI and ML in drug development, including the areas of manufacturing, supply chain, and logistics, has resulted in fears that it will lead to job loss for humans. However, recent research suggests this may not be the case. In many instances, AI is being used to augment human intelligence rather than replace it.

For example, narrow AI learning algorithms are being used to operationalize data for a specific task such as image recognition. While human involvement is still necessary for implementing and operating the AI platform, automation of more straightforward tasks frees up human expertise to tackle more complex problems.

Moreover, studies suggest that automation-related job loss will be offset by jobs created through new AI technologies — indicating potential growth in specific sectors due to its increasing use.

The Role of Explainable AI and Benchmarking in Drug Development

AI/ML has become increasingly important in drug development, with organizations assessing their initiatives based on several key performance indicators (KPIs), such as the ability to support existing projects, lower costs, achieve faster outcomes, and reduce waste. However, within closely regulated industries like pharma, AI/ML methods must be proven to be repeatable and transparent before they are trusted and embraced by professionals and regulators.

Explainable AI (XAI), also known as Interpretable AI, is gaining importance as a way of providing insight into how and why an AI-driven platform or system arrives at its results. XAI techniques provide transparency to ensure the legitimacy of an AI-powered solution, thus leading to increased trust and collaboration from leadership.

They also help regulatory scientists improve their models by debugging questionable results or behaviors that occurred during modeling. Ultimately, the use of XAI can give organizations assurance that their decision-making is responsible, ethical, and unbiased.

Measuring Return on Investment for AI in Drug Development

Evaluating the success of Artificial Intelligence (AI) implementation in regulatory drug development can be complex, making it difficult to measure one’s return on investment (ROI). While factors like time and cost are essential to consider, they do not provide the whole picture when assessing AI success.

It is also essential to take into account the value of predictions and insights provided by AI/ML as these may have an impact on time and cost, but may take some time before their effect is realized. Additionally, successful AI/ML projects can build internal capabilities that can be applied to similar projects and set up future successes.

Ultimately, measuring ROI allows organizations to gauge the profitability and efficiency of their investments and make informed decisions about future spending.

Areas to Focus on When Measuring ROI for AI Benchmarking

When benchmarking Artificial Intelligence (AI) initiatives, organizations should focus on measuring their return on investment (ROI) in key areas.

These include:

1. costs for licenses, software and computer platforms, data, team members, 3rd party partners, and security
2. savings from cost-reducing activities throughout the R&D process
3. soft profits such as improved productivity and product quality
4. strategic alignment with stated goals and objectives
5. key Performance Indicators (KPIs); future profits from new revenue streams created downstream as a result of upstream AI investment.

By assessing ROI in these areas, organizations can evaluate the profitability and efficiency of their investments and make informed decisions about future spending.

Roadmap to AI Success in Regulatory Drug Development

In order to meaningfully integrate Artificial Intelligence (AI) into regulatory drug development, organizations need to invest in data, technology, and talent strategically. To increase their chances of success they should develop an outcome-based AI roadmap that identifies foundational project requirements, priorities, and goals.

Crafting such a roadmap includes effectively defining and communicating the vision, setting realistic expectations that align with business objectives, building a team with the appropriate skillsets, and arming them with appropriate platforms. Following such steps can help ensure value is delivered throughout the drug development journey.

Key Areas to Consider for an AI Roadmap

When developing an AI roadmap for regulatory drug development, there are several key areas to consider. These include:

• identifying strategic business goals and Key Performance Indicators (KPIs) that are hoping to be achieved
• aligning AI use cases with business objectives
• determining project requirements such as the type of AI approach and available data, in-house development capabilities, frameworks, and compute capabilities
• identifying or engaging third-party or AI-native partners; addressing ethical, legal, or regulatory issues surrounding the use case
• measuring success objectively by developing business metrics/KPIs and Return on Investment (ROI)

By taking these steps, organizations can set themselves up for success when integrating AI into their regulated drug development process.

Conclusion

AI applications offer a unique opportunity to help solve some of drug development’s most challenging problems, such as reducing regulatory development costs, driving process efficiencies, and improving various aspects of drug R&D.

In order for us to fully realize the game-changing potential of these applications, however, the classical regulatory drug development process needs to evolve. This includes tracking and benchmarking success, return on investment, and setting more realistic performance measures.

With more open collaboration and clearer articulation of AI’s challenges and benefits, an AI-driven paradigm shift in drug discovery and development is already underway. Organizations must continue investing in the necessary techniques and technologies today if they wish to remain competitive tomorrow.