An AAPS PharmSci 360 overview: Making better decisions faster, earlier, with fewer resources
New Orleans hosted this year’s AAPS PharmSci 360 at the Ernest N. Morial Convention Center. For those fortunate enough to attend, the show provided a targeted delivery of innovation and science.
One of the most coveted aspects about AAPS PharmSci 360 is exposure to some of the most relevant topics across differing areas of the industry, side by side. In one presentation session, we hear how teams are improving commercial manufacturability for molecules with solubility challenges. In the next presentation, you hear research groups present clinical data on an emerging therapeutic area.
It’s clear that the maturity of existing methods and further understanding of newer delivery technologies has grown beyond the question of where could a new technology work? Instead, the focus was on: how do we make better decisions faster, earlier, and with fewer resources? We saw this in discussions on oral delivery of biologics and long acting therapeutics, where attention is turning towards in-depth material characterisation, product behaviour, and delivery mechanisms. Additionally, the same conversation occurred around nitrosamines, with characterisation and risk assessments increasingly moving earlier in development.
This goes hand in hand with a theme that continues to be relevant within our industry – we cannot rely on experimental test data alone. We must be able to describe the under-pinning mechanisms behind that cause of that data to ensure safety and quality of our products and services. Controlling the outcome requires knowledge of the attributes and parameters, whether human or material, actually matter for our product performance.
Demonstrating value in early product development strategy
Physiologically based pharmacokinetic (PBPK) modelling was a hot topic, and what interested us was how the growth of interest from industry is being used to support reduced development efforts and experiments. Some discussions were flatly concerned with predictions for the clinic, while others were more focused on how a process change or drug attribute may matter for the patient experience. Typically, those discussions are siloed between different teams. Hearing them come up within the same circles was one of the more interesting signs of how the development landscape is changing. Modelling is no longer a special area of interest, but is increasingly demonstrating value in early product development strategy – bridging formulation, manufacturing, material characteristics, product attributes, and clinical performance.
As expected, there was a heavy focus on artificial intelligence and machine learning topics. It was clear from the listening areas and poster presentations that AI is continuing to demonstrate itself as a valuable tool in supporting decisions across the entire drug product lifecycle. The focus for presentation of these new technologies has evolved and is no longer about novel potential uses. Data was presented on a modified release AI formulation that had recently been dosed in the clinic. Highlighting these efforts is now on practical applications for improving development timelines and decision making.
AI and machine learning were no longer being applied to specific areas of interest or within a traditional departmental perview. Some of the most impactful case studies were where these technologies leveraged experimental data combined with understanding across an organisation's differing teams. Several discussions included machine learning around formulation development and prediction being combined with material characterisation through advanced imaging and computational analysis and clinical outcomes.
Concerns were voiced as to whether regulators will accept the data from these newer technologies or not. We learned the response will not be a direct yes or no answer. Regulatory expectations seem more focused on how the qualifications of a model and the use of that data are relied on for its intended purposes. Particularly insightful was a retrospective analysis of machine learning models over recent years. It allowed us a glance into whether models continued to remain accurate or whether they generically are simply ‘overall informative’.
Process development integrated with drug product development
We also saw that poor solubility remains a major hurdle for oral formulation development. Dissolution is no longer viewed as just a QC test. Biorelevant dissolution is a valuable in vitro characterisation tool to assess formulation performance and maybe more predictive to animal models. The industry has been working to expand enabling technologies, and there’s still plenty for us to learn about ensuring success at increased process scale while continuing to link these back to relevant biopharmaceutic properties. With the use of newer process modelling, the discussion is at a different stage now, allowing faster decision making with less data. It's clear that process development is being integrated with drug product development, instead of being considered separate activities – formulation versus manufacturing.
AAPS PharmSci 360 is still a great event for bringing together academics, research, consultants, and industry. The conference showed shared focus across existing, maturing, and more novel areas of interest. Though the show seems condensed from previous years, several vendors seem to have downsized their presence; indicating a shift in how individuals and organisations use these conferences as they ask themselves not whether to attend or not, but how to best present to their targeted audience while still showcasing differentiation. This may create unique opportunities in future conferences, allowing organisations to have a more concentrated voice on the most relevant, ongoing scientific themes and greater exposure.
Despite the smaller footprint, the quality and the relevance of the scientific discussions were undiminished. Attending AAPS PharmSci 360 continues to show value across organisations operating within academia, manufacturing, clinical, analytical, quality, and regulatory areas.
About the authors
Robert Cornog is senior drug development consultant at Quotient Sciences. He has over 26 years of experience in product development and process design. Across a broad range of dosage forms and technologies, Cornog has focused on advancing pharmaceutical manufacturing through science and technical innovations. Through application of a deep understanding of manufacturing sciences and risk-based methodologies, he has successfully driven the adoption of new technologies and expanding manufacturing capabilities for late stage and commercial programmes.
Dr Vanessa Zann is VP of translational pharmaceutics and clinical pharmacology at Quotient Sciences. Dr Zann has over 25 years’ industry experience providing expert biopharmaceutic support to drug discovery, early drug development, and clinical programme design. Dr Zann holds a PhD in Pharmaceutical Science and completed postdoctoral research in buccal transport. She is currently head of the US scientific consulting team, providing programme design, scientific oversight, and support to business development.
