Stassen Mario, Leitão Catarina S, Manzano Toni, Valero Francisco, Stevens Benjamin, Schmucki Matt, Hubmayr David, Rubinat Ferran Mirabent, Dessoy Sandrine, Moreira Antonio
Department of Life Sciences Innovation, Stassen Pharmaconsult BV, Aerdenhout, The Netherlands;
Data Science Department, ValGenesis Portugal, R. Castilho 50 4th Floor, 1250-071 Lisboa, Portugal.
PDA J Pharm Sci Technol. 2025 Mar 3;79(1):68-87. doi: 10.5731/pdajpst.2024.012950.
This review paper explores the transformative impact of Artificial Intelligence (AI) on Continued Process Verification (CPV) in the biopharmaceutical industry. Originating from the CPV of the Future project, the study investigates the challenges and opportunities associated with integrating AI into CPV, focusing on real-time data analysis and proactive process adjustments. The paper highlights the importance of aligning AI solutions with regulatory standards and offers a set of comprehensive recommendations to bridge the gap between AI's potential and its practical, compliant, and safe application in pharmaceutical manufacturing. Emphasizing transparency, interpretability, and risk management, the research contributes to establishing best practices for AI implementation, ensuring the highest quality pharmaceutical products while meeting regulatory expectations. The conclusions drawn provide valuable insights for navigating the evolving landscape of AI in pharmaceutical manufacturing. This paper serves as a guideline for implementing AI, Machine Learning and Deep Learning models to the pharma industry. Nevertheless, the specific algorithms used in the CPV of the Future are not relevant for our paper (Good Practices), since we have to generalize the process independent of the algorithm.
这篇综述文章探讨了人工智能(AI)对生物制药行业持续过程验证(CPV)的变革性影响。该研究源于“未来的CPV”项目,调查了将AI集成到CPV中所面临的挑战和机遇,重点关注实时数据分析和主动过程调整。文章强调了使AI解决方案符合监管标准的重要性,并提供了一套全面的建议,以弥合AI在制药制造中的潜力与其实际、合规和安全应用之间的差距。该研究强调透明度、可解释性和风险管理,有助于建立AI实施的最佳实践,确保在满足监管期望的同时生产出最高质量的药品。所得出的结论为应对制药制造中不断发展的AI格局提供了宝贵的见解。本文可作为在制药行业实施AI、机器学习和深度学习模型的指南。然而,“未来的CPV”中使用的特定算法与我们的论文(良好实践)无关,因为我们必须对独立于算法的过程进行概括。