Over 10 years experience of Traceability Solutions
By Pharmatrax Author
Category: News, Press Release
No Comments
External expertise is beneficial in helping companies select the right tools at the right stages of development to ensure success.
IPOPBA – STOCK.ADOBE.COM
According to research, post-pandemic R&D growth is healthy in the pharma industry with the number of drugs in the R&D pipeline now exceeding 20,000 in 2022—which represents an expansion rate of 8.22% from 2021 (1). This heightened spending into researching novel chemical modalities and substances can be absorbed by some pharma companies as an internal cost; however, development of a suitable and ultimately successful formulation strategy to advance a molecule to commercialization can benefit from external expertise. As such, it has also been projected that the expanding R&D pipeline will be a driver for the formulation development outsourcing market (2).
“An outsourcing partner should have a good understanding of CMC [chemistry, manufacturing, and controls] perspectives with the developer’s interests in mind, which enables the developer to have a sharp vision of potential scenarios to develop for first-in-human (FIH) trials, but also considers factors that will influence filings and the product’s commercial success,” says Thierry-Thien K Nguyen, director of Product Development, Catalent Biologics. “In addition, an outsourcing partner should collaborate with customers to develop PrOACT (identify problems, objectives, alternatives, consequences, and trade-offs) and enable customers to make the most informed decisions to reduce risk and increase the program’s chances of success.”
To support optimal formulation strategies, not only must outsourcing partners invest in building a resource pool with the right skill set, but they must also provide continuous training development programs and bring in the right infrastructure, adds Anil Kane, PhD, senior director, global technical scientific affairs, pharma services, Thermo Fisher Scientific. “Employing the right tools (modeling, simulation, predictive) at the right stage of pre-clinical, clinical or late-stage development is important and has a huge potential for application, interpretation, and science-based decision making in adopting smart formulation strategies,” he says.
A tool that is aiding in early development is computer-based predictive modeling, reveals Kane. “This predictive modeling avoids trial-and-error-based search of the right solution to solve the complex technology and excipient selection process, thus helping to reduce the time and investment in early clinical development,” he specifies. “An algorithm-based in-house developed software can predict technologies that will offer a higher probability of success among several options such as micronization, amorphous solid dispersions (spray drying or hot melt extrusion), lipid-based delivery, complexation, and others.”
“During formulation and process development, at an early preclinical/clinical stage, an in-depth understanding of material properties and their impact on formulation and process is critical,” Kane continues. It is possible to simulate the behavior of bulk materials in oral solid dosage forms using discrete element modeling (DEM)/compaction simulation tools, he explains. “The use of these tools mitigates formulation/process challenges early in development and helps speed up process development and troubleshooting,” Kane says.
Additionally, Kane emphasizes that modeling has found further use in aiding the design of formulation strategies for immediate-release and controlled/extended-release oral solid dosage forms through physiologically based pharmacokinetics modeling. “Predicting animal/FIH doses, in-vitro–in-vivo correlation (IVIVC), food effects, bioequivalence, pharmacokinetic/pharmacodynamic (PK/PD), and so on, using software has helped in the assessment of formulation impact on PK, compartmental PK analysis of animal data, evaluation of animal bioavailability, and defining an FIH clinical dosing strategy,” Kane states.
Focusing on large molecules, Nguyen points out that there have been several innovations that have significantly impacted formulation strategies during recent years. In addition to predictive models using artificial intelligence (AI) and machine learning (ML), Nguyen highlights high-purity, parenteral-grade excipients in cryopreservation media for autologous and allogeneic cell therapies; novel excipients for use in vaccines, viral therapies, and high-concentration biologics formulations; lipid nanoparticle components and associated manufacturing technology for encapsulation; and continuous sterile freeze-drying technology, as key innovations.
However, each of these innovations faces several challenges, such as toxicology/pharmacology and pharmacovigilance data requirements for novel excipients before they can be widely adopted, Nguyen stresses. “Some innovations are more acceptable, such as lipid nanoparticles (LNP) in messenger RNA/LNP formulations,” he says. “To overcome these challenges and enable technology adoption, those aspects identified using a quality-by-design (QbD) approach need to be addressed in order that patient-centric requirements are satisfied, and a clear understanding of product quality attributes and associated process parameters is demonstrated.”
There has been much advancement in the field of information technology recently, confirms Nguyen, such as the implementation of AI- and ML-based techniques in a variety of industries. “However, the pharmaceutical industry has been slower than others to adopt AI- and ML-based techniques, perhaps because of the lack of clear regulations, a paucity of funding for research into these fields, and a scarcity of reputable quality training programs,” he notes.
Yet, Nguyen continues, when digitalization is implemented both thoroughly and properly into formulation development strategies, it is possible to improve cost and time efficiencies in drug product formulation and process development. If applied correctly, it is possible to enable past information on formulations, APIs, and processing conditions to be amassed; allow the creation of a large common library to teach/train new ML algorithms; support the creation of algorithms that predict how similar molecules/entities would behave
in given formulation environments; and will provide the potential for quicker access to data and greater processing power, which would make the formulation candidate selection more efficient, he specifies.
“Finally, with a predictive AI model, minimal data sets could more accurately anticipate a molecule’s stability,” Nguyen adds. “For formulation in a traditional manufacturing site, digitalization benefits could be realized in the next few years, which have already been explored in both academia and industry. For personalized medicine, bedside and in situ formulation, and in combination with wearable devices, there is still some way to go, but the future is exciting.”
Advances in analytics and digitalization are being leveraged by many Big Pharma companies to improve drug development productivity, asserts Kane. Approaches to increase productivity or success rates in drug design that are adopted by researchers, medicinal chemists, and pharmaceutical experts include big data, AI, ML, neural networks, quantitative structure-activity relationship modeling, and others, he states.
“The traditional approach to drug formulation development relies on iterative trial and error, requiring a large number of resource-intensive and time-consuming in-vitro and in-vivo experiments,” Kane says. “ML-directed drug formulation development offers unparalleled opportunities to fast-track development efforts, uncover new materials and innovative formulations, and generate new knowledge in drug formulation science (3).”
At the dosage form level, advancements are allowing for more comprehensive data collection. For example, digital pills that feature ingestible sensors in an oral drug delivery dosage form will revolutionize diagnostics, clinical monitoring, data collection, and data analytics in the healthcare industry, Kane emphasizes. “Smart medicines involve a target applied directly to the medicine via a coating or ink, and a sensing device (a cell-phone application) is used to read the medicine and convert the signal into a digital format in a matter of seconds. With smart medicines, every dose becomes a basis for patient communication, engagement, and data generation,” he says. “The transition to smart medicines offers an enormous source of value creation to the pharmaceutical industry.”
It is important, however, to ensure the quality of the data being collected, warned Nguyen. When collecting data that will be used to build a library and then to train machines, for example, the data should conform to ALCOA+ [attributable, legible, contemporaneous, original, accurate, and additionally that it is complete, consistent, enduring, and available] guidance and should be unbiased, he explains.
“Good input and good process shall yield good output, otherwise, with less representative data the predictive model would misrepresent, and unintentionally skew toward an unwanted design space,” Nguyen says.
“Continuous manufacturing (CM) continues to be of interest for specific products by the emerging, mid-size and Big Pharma companies for one or more of the well-documented and proven reasons—reduced API usage from early to mid-stage development to commercialization, robust simplified formulation, scale-up efficiencies, speed to clinic and to market, supply chain flexibility, reduced cycle time, and potential reduced overall cost of supply while having a greater assurance of quality,” Kane asserts. “A CM process certainly changes the formulation strategies, as there will be far more controls and checks at critical processes as compared to a batch process that utilizes multi-unit operations. Selection of excipients and a process needs to be aligned with the continuous manufacturing equipment design and scale.”
Process analytical technology (PAT) tools, such as near infrared and Raman spectroscopy, are integral for continuous manufacturing programs, continues Kane. Additionally, to be able to expand the continuous solid dose program and bring customer programs to commercialization, it has been beneficial at Thermo Fisher Scientific to use real-time release—for finished product release testing—as well as predictive models and in-suite tools to support the required testing, he adds.
“Advanced continuous manufacturing approaches are often considered for mAbs [monoclonal antibodies], therapeutic proteins, or other biologics that are compatible with such a process,” states Nguyen. “[Advanced CM] is more challenging to adapt for cell therapies and less-standard modalities made through complex, smaller-scale processes.”
Across the board, however, demonstrating homogeneity in bulk drug substance and assurance of dose uniformity are key aspects of formulation and fill/finish, Nguyen points out. “From this perspective, CM and process intensification, while making much advancement in terms of drug substance upstream (cell culture) and downstream processing (purification), still faces such challenges that it is often decoupled from fill/finish operations. As a result, at the end of the process, bulk drug substance containers need to be stored and handled within the correct conditions and shipped to fill/finish sites to process as batches upon demand,” he says.
“For a product that requires only a single formulation and single strength, formulation considerations could be built in during the ultrafiltration/diafiltration (UF/DF) step and the finishing step to create formulated bulk material,” Nguyen continues. “Upon receiving at the fill/finish site, it would require pooling, blending, filtration, aseptic filling, and packing.”
When considering the quality requirements, as is the same with batch processing, the framework of CM should start with a patient-centric design, the target product profile, followed by CMC target profile, Nguyen states. “At each unit operation, risk needs to be evaluated, and the learnings carried over to the next steps to understand what can be further removed, polished, controlled, mitigated, or monitored,” he says. “Understanding the endpoint and requirements would help developers anticipate what shall be considered upfront to avoid costly correction or redevelopment later in a product’s lifecycle. Moving from early phase to late phase, as product and process knowledge are accumulated, the presumptive attributes and process parameters would be further refined as critical versus non-critical, or key versus non-key for a product and its process.”
“Digitization of operations, use of augmented and virtual reality, AI and ML, increased use of robotics, and using connected data systems to drive better quality and business decisions need to be actively pursued in many parts of the business,” asserts Kane.
Through digitization, it is possible to reduce the number of conditions that need to be performed in the design of experiment within formulation, Nguyen adds. “If a reliable predictive model can be built using a high-quality, diverse, and unbiased database, then that will drastically help improve formulation strategies in terms of time, cost, and likelihood of success, and would contribute to bringing drugs to patients more quickly.”
“Therefore, expertise and experience in utilization of the modeling tools is critical to bring speed and efficiency to clinical programs,” Kane summarizes. “In late-stage clinical studies, it is critical to adopt sound, science-based risk-assessment and risk-mitigation strategies to optimize formulations and process of manufacture.”
Source: https://www.pharmtech.com/view/adopting-smart-development-strategies