Analytical Strategies Enabling rapid and Nimble mRNA Product Development
David Cirelli, Senior Director, Analytical Research & Development, Pfizer
At the TIDES USA conference in San Diego, David Cirelli, Senior Director, Analytical Research & Development, at Pfizer, began his presentation by focusing on a recent publication describing the strategy that has enabled Pfizer and Biontech to successfully deliver billions of doses of vaccines to more than 170 countries worldwide. While the network of GMP manufacturing and testing sites is gradually expanding, there is a 50% reduction in manufacturing and release testing time, leading to an acceleration of vaccine development.
Cirelli said that experience shows that the lipid nanoparticle (LNP) manufacturing process is effective, scalable, and agile. He highlighted their commitment to respond to COVID-19 variants within 100 days.
He discussed the mRNA-LNP vaccine manufacturing process, which starts with a high-quality DNA template that is used in an in vitro transcription process to produce mRNA transcripts. After purification and formulation, the drug is combined with lipids. RNA and lipid come together to form lipid nanoparticles, which are further purified and formulated and filled into the final drug container, he said.
mRNA Traits
Cirelli briefly touched upon the key quality attributes of mRNA transcripts including 5’ cap, 5’ untranslated region (UTR), open reading frame (ORF), 3’ UTR, and poly(A) tail, which are affecting translation efficiency and RNA stability in some way.
He said that 5’ cap, ORF and poly(A) tail are routinely controlled at QC level, and there are a variety of methods to measure RNA integrity for batch testing. Full-length RNA is measured using CR (fragment analyzer) and product impurities and truncated products lacking portions, especially poly(A) tail, which can be detected. To test this, they used control samples lacking 5’ cap or poly(A) tail and showed that 5’ cap and poly(A) tail elements are required for antigen expression.
For further analysis, they isolated fragmented species using chromatography and showed that these impurities do not give rise to any off-target antigen expression or truncated protein.
Following mRNA attributes, he spoke about composition on LNP, including ionizable cationic lipid, PEG lipid, structural lipids, and RNA inside. Several parameters such as ratio of lipids, LNP size, lipid content, and RNA encapsulation are routinely controlled. They also ensure that LNPs can be taken up, and RNA can be released and translated into protein of interest. They highlighted the use of NMR to assess surface properties of the LNP and to give an idea about the product stability.
Regarding potency assessment, Cirelli mentioned in vitro expression assay, which is a useful tool for screening formulations and determining the best performers. In this assay, cells transfected with LNPs are incubated for antigen production. Then the cells are harvested and stained for viability. Using an antibody cocktail specific for intended antigen, cells can be analyzed using flow cytometry. Regardless of the cellular localization, IVE assay enables us to compare benchmarks to the other constructs in terms of IC50, % positive staining cells and mean fluorescent intensity.
Platform Approach
In the final part of his presentation, Cirelli emphasized the importance of a platform approach. Looking at quality attributes such as DNA starting material, mRNA drug substance, and mRNA/LNP drug substance, there is a wide range of methods that are used to control the quality and properties of the drug or vaccine. He said if we think about a platform approach, many of these methods become a platform that does not need to be applied when we make changes to the product. For example, in the case of a COVID variant that does not require significant changes in operating conditions, platform method approaches work without having to repeat extensive ICH validations.
Depending on which features change in the product, additional validation or a new methodology can be performed for the new attribute. This framework has been used for new drug applications for COVID-19 vaccines. Cirelli explained this concept using two scenarios as examples. First, there is only a change in the mRNA sequence. Therefore, we should evaluate what is unchanged or very similar. Although there is only a change in the DNA sequence encoding the mRNA, only a small number of methods such as DNA identity, RNA identity, RNA identity/strain ratio, and potency are required for additional validation. Second, there is a change from a monovalent to a multivalent drug. The only thing that changes is the RNA ratio. The other parameters are either unchanged or remain very similar. One needs to focus only on key methods that are at the product level, such as RNA identity/strain ratio and potency.