By Chris Latoz
Figure 1: Vaccine/Therapeutic Modalities
Beyond the daily COVID-19 metrics, there has been a media focus on the hope new vaccines bring to the global fight against the pandemic. To date, three such vaccines have been granted Emergency Use Authorization (EUA) in the United States by the Food and Drug Administration (FDA). Have you ever wondered about the overall stability strategy and analytical techniques used to assess these vaccines? How is the development accelerated? How is the vaccine expiry determined? Why do different vaccines require different storage conditions? Also, what role does stability knowledge play in supply chain logistics, with “just in time” delivery of vaccines?
These and many more related topics were discussed during the recent two-day AAPS Virtual Workshop Vaccine Stability Considerations to Enable Rapid Development and Deployment, held on March 24-25, 2021, and sponsored by Catalent and Eurofins. Many distinguished speakers and panelists from across industry, academia, regulatory, as well as healthcare leaders, discussed their insights and solutions to meet the need of rapid development and deployment to respond to the Sars-Cov-2 pandemic (Table 1).
Table 1: List of distinguished speakers and panelists
Several common themes to enable rapid development and deployment of the Covid-19 vaccine surfaced during the workshop presentations and panel discussions. The most critical points included: leveraging prior knowledge from existing vaccines, using kinetic modeling with accelerated stability data, and understanding the supply chain to ensure the success of the vaccine program.
Leveraging Prior Knowledge
Leveraging prior knowledge (production processes, analytical techniques, formulation materials, stability data, and toxicological concerns) was critical to enable rapid development of the COVID-19 vaccines. Typically, it takes an average of ten years for the development of a vaccine. The COVID-19 vaccine process took less than 10-months to be granted Emergency Use Authorization (EUA). This would have not been possible utilizing a start-from-scratch approach.
The workshop started with Dr. Lisa Kueltzo discussing how the NIH formulation strategy consisted of identifying the optimal pH and ionic strength for the protein antigen by leveraging historical data on viral fusion proteins (Figure 1). This allowed for a minimal stress assessment, such as freeze-thaw, Tg’, Tc (Lyo), agitation, etc. to be conducted. The amount of real-time and accelerated stability data generated was driven by material availability and timeline. Kueltzo also stressed the importance of mimicking the container closure system, using Good Manufacturing Practices (GMP) analogous assays in drug product stability, and the need for frequent communication with the regulatory agencies when leveraging development stability data for Investigational New Drug (IND) applications.
Dr. Arjen Scholten discussed Janssen’s Covid shelf-life model and how they developed different aspects of the shelf-life model in parallel to meet the EUA timelines. Many shelf-life challenges were overcome, including extremely tight timelines, limited COVID-19-specific data on shelf-life elements, limited real-time stability data, and simultaneous development of the shelf-life model and supply routes (Figure 2).
Figure 2: Challenges for Janssen’s COVID-19 Shelf-life Model
Janssen’s success was accomplished primarily by leveraging existing platform experience from Janssen’s AdVac® and Per.C6® technologies. From a manufacturability perspective, Janssen was also able to leverage their learnings from the Ebola vaccine to scale up production for the COVID-19 vaccine. A condition of the EUA process is confirming and updating of the shelf-life modeled from the leveraged platforms with actual COVID-19 vaccine-specific data from the ongoing studies.
Kartik Narayan, Ph.D. followed on with a discussion of leveraging existing platforms to characterize a spiked protein construct. Narayan used an example where using existing platforms, the lab was able to confirm that the recombinant spike produced was pure and properly folded. They were also able to confirm the interaction between spike protein and ACE2 receptor can be disrupted by use of nAbs and that pH had a major effect on binding to ACE2 receptor.
Using Kinetic Modeling and Accelerated Stability Data
The principles outlined in the WHO guidelines on evaluation of vaccine stability were presented by Tim Schofield. The presentation included the importance of setting scientifically or clinically justified specification limits and the “grand delusion” of thinking that release and end-of-shelf-life limits should be the same (thus allowing no room for the inherent instability of many vaccines). It also covered the formulation of release and stability models using accelerated stability conditions and an understanding of the kinetics of degradation. Such statistical modeling was contrasted against less rigorous approaches or routine monitoring. Statistical approaches for post-licensure stability modeling were also presented, as a form of “process monitoring” or “continued stability verification” (csv). Stability models can be very important tools to evaluate the impact of any life cycle management (LCM) changes. This is accomplished by using the Arrhenius relationship as a stability “fingerprint” or using a statistically defined “stability space” and equivalence testing. Schofield showed post-licensure examples where using accelerated stability data was a means of discovering peculiar stability properties (when compared to historical behavior), leading to more formal evaluations.
Dr. Andrew Lennard discussed strategies for taking stability off the critical path during vaccine product development. Risk-based predictive stability (RBPS) approaches balance stability data at submission with product medical need. It does this using principles of quality by design (QbD) to justify moving elements of conventional stability to post-approval by including the use of appropriate models to predict product stability profile and shelf-life establishment. Two approaches described in detail were prior knowledge stability models (using reference data sets for “like molecules”) and advanced kinetic analysis (using multiple accelerated temperatures to develop kinetic models to predict shelf-life beyond the product specific real-time data).Examples were presented for both approaches.
Criticality of Supply Chain and “The Last Mile”
The final session of the workshop focused on the challenges of the supply chain and deployment of the vaccine (the “last mile”). The recent disruption in global supply chains has not only impacted the ability to purchase consumer items, it has also hampered the manufacture and delivery of the COVID-19 vaccine to distribution centers, globally. Nada Sanders, Ph.D., described supply chain vulnerabilities, such as complexity, capacity, poor visibility, and heavy reliance on ust-in-time manufacturing.
For the mRNA vaccines, cold-chain storage and “the last mile” were especially problematic because many vaccination centers were not equipped with the -80°C freezers required to store Pfizer’s vaccine. Tony Reed, Ph.D. addressed this very topic in his presentation and discussed the many challenges when deploying the COVID-19 vaccine including purchasing at least a -80°C freezer to allow for receipt of the Pfizer vaccine, developing a tracking list to allocate vaccines to the right groups, training staff, and establishing logistics to ensure all the vaccines are used. Reed indicated that there is no room for errors and they must remain flexible to ensure the vaccines get to the proper group of people, as assigned and prioritized.
Temple University Health System addressed the last mile by starting with considering patient entry (Figure 3). It is critical to ensure the right patients were selected; thus, priority was assigned to the sickest patients first for the initial 30,000 doses. As the vaccine entry involved the supply chain from beginning to end, the distribution sites needed to have the ability to accommodate specific storage of either the Pfizer or Moderna vaccines.Consequently, the process had to eliminate the potential mix-up of the two vaccines. Lastly, the vaccinator entry involved establishment of proper training at the vaccination centers to administer the vaccine without losing a single dose, and preventing issues with back-ups at the center.
Figure 3: Entries that Impact The Last Mile
The two-day workshop successfully discussed different aspects of vaccine development and deployment. It consisted of eight presentations, four panel discussions, and 12 speakers, gathering almost 70 participants from many different pharmaceutical and healthcare companies. To register to view the recording of the entire workshop or individual sessions on Vaccine Stability Considerations to Enable Rapid Development and Deployment, go to https://www.aaps.org/education-and-research/workshops/stability.
Acknowledgments
The author would like to acknowledge the workshop organizing committee consisting of Mark Alasandro, Ph.D.; Jianmei Kochling, Ph.D.; Kim Huynh-Ba, M.S., FAAPS; Nanda Subbarao, Ph.D.; Lori McCaig, Ph.D.; Laure Larkin, M.S.; Yan Wu, M.S.; Dilip Choudhury, Ph.D.; and Yajie Zhang, Ph.D., for their planning and execution of the workshop. Special acknowledgement also for their technical review and contribution to this briefing.