AAPS PharmSci 360: Enhancing Bioanalysis with AI and Other Technologies

Bioanalysis, which is utilized in drug development to determine pharmacology, bioavailability, and bioequivalence, includes pharmacokinetic, toxicokinetic, or biomarker concentration measurements (1) in animal safety and clinical studies. Bioanalytical methods are required by regulators to be validated to ensure their data’s reliability (1). Artificial intelligence (AI), machine learning (ML), and other complex tools are helping drug developers evaluate the complex data involved in these types of studies.

Pharmaceutical Technology® spoke with Dr. Mark Arnold, owner and principal, Bioanalytical Solution Integration to find out how bioanalysis enhances bio/pharmaceutical drug development. Arnold will be presenting the keynote session, “Reflections on Transformations in Drug Development, Bioanalysis, and Regulations”, at AAPS PharmSci 360 on Monday, Nov. 10, 2025 at 1:30 PM.

PharmTech: How can AI and ML be used to enhance bioanalysis in bio/pharmaceutical development?

Arnold (Bioanalytical Solution Integration): AI, large language models [LLMs], natural language processing, and ML have been discussed for a few years in bioanalysis but are now making actual inroads with deployed validated applications and others in development.

The earliest implementations are for reports and quality control (QC) of assay validations and study sample analysis. As the capabilities of the AIs have advanced, so have the applications in bioanalysis. New applications are looking at taking GLP [good laboratory practice] or GCP [good clinical practice] protocols and bioanalytical contracts to write bioanalytical plans. Other applications perform quality control checks that integrate ELN [electronic lab notebook], instrument, and LIMS [laboratory information management system] data on a daily basis to catch errors early and identify trends that lead to failure before they happen, and data analysis as part of investigations into assay failures.

Building on design of experiment concepts, AI will utilize databases of knowledge to improve method development for immunoassays and LC–MS [liquid chromatography—mass spectrometry] methods. I’ve heard that a mass spectrometer manufacturer is looking to apply AI to chromatogram integration to go past algorithm rules to more intelligent peak detection.Ultimately, biotech and pharma companies will be using AI to write the bioanalytical sections of filings. Health authorities are also looking at AI as a tool to analyze bioanalytical data in submissions and during on-site or remote inspections of bioanalytical labs. The resulting analyses will highlight both compliance with regulations and detect problems and errors that need further review.

PharmTech: How can large language models be applied to bioanalytical workflows?

Arnold (Bioanalytical Solution Integration): As noted above, AI, natural language processing, and ML are being implemented in bioanalysis for report writing and bioanalytical plans. Many LLM AIs are designed as predictive tools and operate as black boxes where it cannot be determined how they do what they do. A challenge in regulated bioanalysis in the use of LLM AIs is in their training and predictive nature; they need to find data (content) in different sources and formats and place it appropriately in the report.

For regulated bioanalysis, controls must be in place for the AI or the risk of hallucination (creation of data or content not provided) can occur as has been seen in some AIs. Not knowing if the report content is correct is not something bioanalytical labs can risk and would result in utilizing the same or greater QC and QA [quality assurance] resources to make sure of the report’s accuracy. Which is contrary to what AI is supposed to do: reduce human involvement and improve accuracy.

To address this, at least one AI-based vendor for bioanalytical labs has constraints in place to prevent hallucinated data creation and is including an audit trail in its report generation steps to demonstrate the data reached the right location. Pharma companies have started using LLMs for writing some portions of eCTDs [electronic common technical documents] and the bioanalytical sections are well-structured, making them good candidates for early implementation of AI writing.

PharmTech: What are some of the new instrumentations that are being used in bioanalysis? What makes these unique?

Arnold (Bioanalytical Solution Integration): Most of the technologies in use for pharmacokinetic, immunogenicity, and pharmacodynamic measurements have been around but are being refined for improved sensitivity, accuracy, and precision. These include the various forms of immunoassays, hyphenated chromatographic assays (e.g., LC–MS/MS [liquid chromatography—tandem mass spectrometry], LC–HRMS [liquid chromatography—high-resolution mass spectrometry]), PCR [polymerase chain reaction], flow cytometry, and cell-based technologies. The improved flow cytometry instruments, with some instruments having 30 and 40 color detection ability, are being used not only for cellular characterization, but the improved sensitivities allow them to be applied to circulating exosome biomarkers that can detect and characterize many diseases. Mass cytometry is being used in research, and once suitable applications are identified, likely in measuring biomarkers, it will move into regulated bioanalysis.

PharmTech: How do changing regulations impact advancements in bioanalysis?

Arnold (Bioanalytical Solution Integration): ICH [International Council for Harmonisation] M10 standardized regulations for pharmacokinetic bioanalysis using immuno- and chromatographic assays. Having a standard accepted in many major markets eliminated the need for bioanalysts to have to keep up with changes in each of the individual countries and adjust their practices, especially when there were conflicting expectations. WHO [World Health Organization] recently published M10 which may expand its acceptance beyond the ICH-adherent countries. The expansion of gene and cell therapies require technologies (e.g., PCR, FLOW cytometry) not covered in ICH M10, and so far, the scientific community has been promoting science-based approaches to these technologies. These technologies have been used in clinical laboratories and adapted for use in pharmacokinetic assays.

For immunogenicity, a few countries have developed regulations that are reasonably similar so that achieving a consistent approach is not difficult. However, with the extensive experience for monoclonal antibody drugs, bioanalysts are proposing alternative approaches to simplify the traditional tiered-immunogenicity assay designs that generate as informative data with less effort and cost. Regulators have not formally responded to these proposals.

Another area without a global regulatory perspective is the analysis of biomarkers. In the United States, measuring biomarkers for a medical decision (i.e., those used by doctors and in hospitals) must follow the Clinical Laboratory Improvement Amendments, while those used to support the safety and efficacy of a new drug are to follow a 2025 FDA guidance which is focused on immuno- and chromatographic (e.g., LC–MS/MS) assays.The European Union and Japan have totally separate and different sets of regulations. These differences create not only challenges for the bioanalyst but also add complexities when trying to run clinical trials in multiple global regions.

To be successful in delivering data accepted globally, bioanalysts must be aware of a variety of regulations and how they apply to the purpose of the analysis, type of samples they are analyzing, and the technologies. When no regulations exist, scientists must continue to collaborate on science-based best practices. Since the science is always advancing faster than the regulations, regulators must learn about the science of the technologies and evaluate practices that generate quality data, and when they have enough understanding of the science generate regulations; hopefully, based on or in collaboration with the recommendations of the scientific community.

Reference

1. FDA. Bioanalytical Method Validation, Guidance for Industry (CDER, May 2018).

About Mark Arnold

Mark E. Arnold, PhD is Owner and Principal of Bioanalytical Solution Integration; consulting with biotech and pharma companies to advance their portfolios to bring new therapies to patients in need. In that role, he collaboratively develops the bioanalytical scientific and regulatory strategy to meet current and future client needs. Mark was previously Executive Director of the Bioanalytical Sciences Department at Bristol-Myers Squibb Co. and Director of Science for the Scientific Affairs Department at Labcorp supporting bioanalytical and clinical laboratories. He received a B.S. in biology from Indiana University of Pennsylvania and a PhD in pharmacology from the University of Pittsburgh. For 40 years, Dr. Arnold has been involved in the application of bioanalysis to developing new therapies, including the review and interpretation of regulations and guidance as they apply to the evolving science in bioanalytical and clinical laboratories. He was co-chair of both the AAPS-FDA Crystal City V and VI Workshops and has been a member of AAPS since 1988 holding a variety of committee positions including 2023-2024 Scientific Advisory Chair, 2022 PharmSci360 Scientific Program Committee Chair, 2018 PharmSci360 Bioanalytical Track Chair and is an AAPS Fellow (2014). Dr. Arnold has over 115 peer-reviewed publications and numerous invited podium presentations.