Q&A: LC-MS to reinvent impurity analysis in gene-based drugs

by ,

Kerstin Pohl, sr. global manager, Gene Therapies, at SCIEX, and Thomas Kofoed, co-founder of Alphalyse, discussed the harmful effects of HCPs and strategies to mitigate them.

Key Insights:

  • Host cell proteins (HCPs) may disrupt drug functionality and pose the risk of severe levels of undesired immunogenicity. They must be thoroughly assessed and eliminated during manufacturing.
  • Immunoassays frequently only estimate the overall HCP quantity and cannot provide information about individual HCPs, and they tend to neglect weak immunogens.
  • Liquid chromatography-mass spectrometry (LC-MS) offers robust, sensitive and unbiased profiling of HCPs and can be leveraged to minimise process-related impurities.

Genomic medicine and gene-based vaccines show immense promise for treating cancer and rare diseases, correcting for gene deficiencies and accelerating vaccine production. However, purification of these products can be quite intricate, and incomplete purification can lead to complications. Most recently, one of the FDA-approved adenoviral COVID-19 vaccines was associated with vaccine-induced thrombosis with thrombocytopenia syndrome (VITT) with 1:125,000 frequency, inducing blood clotting in vital organs. Although the factors leading to VITT are still being studied, investigation of the vaccine components revealed adenoviral proteins that may have stimulated antibody production for attacking platelet factor 4 (PF4), triggering coagulation.

This raised concerns about process-related impurities in gene-based vaccines and safety implications. During viral vector development, proteins expressed by the packaging cells (host cells) can co-purify with the delivery vehicle (the viral vectors). These host cell proteins (HCPs) are impurities that can alter the mechanism of action in many ways.

Kerstin Pohl (KP): How exactly do HCPs affect vaccine quality and safety? How does this impact the vaccine production timeline?

Thomas Kofoed (TK): Certain HCPs, such as lipases and proteases, could impair stability and integrity by degrading the vaccine vector, stabilisers or adjuvants. Others can activate an immune response and trigger adverse effects, such as excessive inflammatory cytokine release or thrombosis.

Once the harmful effects from HCPs show up in later stages of the product life cycle and clinical settings, optimising the product becomes an insurmountable challenge. Therefore, pharmaceutical companies must analyse and eliminate these impurities as early as possible.

KP: Manufacturers have commonly used ELISAs to analyse HCPs. What are the advantages and disadvantages of this approach?

TK: One of the critical quality attributes in pharmaceutical production is the HCP concentration or amount in the final product, typically established as 100 ng per mg of therapeutic (100 ppm). Manufacturers mainly use enzyme-linked immunosorbent assays (ELISAs) to estimate overall HCP content. ELISAs have proven valuable since they can rapidly detect impurities at levels as low as parts per million.

However, it is challenging to develop a set of antibodies to recognise extensive HCP profiles, especially small HCPs with relatively weak immunogenicity to raise antibodies for ELISAs. There is then the problem of HCP composition, which ELISAs cannot provide since they only reveal the overall HCP quantity. Generally speaking, the insight gained from ELISA does not suffice to distinguish and quantify individual HCPs, so an orthogonal approach is required.

KP: Liquid chromatography-mass spectrometry (LC-MS) is often called an orthogonal approach. Can you elaborate on that? What makes LC-MS feasible for HCP analysis?

TK: LC-MS can be a useful complement to ELISA because of its ability to identify and quantify individual HCPs from a mixture. We have gained extensive experience on this topic over the past seven years, owing to our partnership with SCIEX and the use of high-resolution mass spectrometry (HRMS). While ELISAs are blind towards changes of in composition, HRMS can reveal changes and identify individual HCPs, allowing for informed decision-making. You can detect thousands of proteins, and then identify and quantify them, in a single LC-MS injection. HRMS allows for the quantification of HCPs present in trace amounts as well.

This initial analysis can then be used to identify problematic HCPs, which can be monitored in a targeted manner with high throughput using triple quadrupole LC-MS through multiple reaction monitoring (MRM) at later stages of development or even quality control (QC).

KP: Is LC-MS for HCP analysis used only at later stages of development, or is it flexible enough to be implemented for process optimisation?

TK: One of the advantages of LC-MS is its flexibility. It can be used to analyse both the final product and in-process samples at various stages during downstream processing. In fact, we have several clients using the provided LC-MS results to optimise their downstream purification processes, remove challenging HCPs more efficiently and thus improve yield. Furthermore, LC-MS can be used for analysing the final product - we recently showed that LC-MS for HCP analysis can be used under GMP as a release assay. This is the world’s first case of using LC-MS in that context, and it significantly shortened the development time of our client’s virus-like particle-based vaccine booster.

KP: We sometimes hear about concerns related to the bias of the methods used for HCP analysis. Is there bias in LC-MS-based HCP analysis, and if so, how do you overcome it?

TK: While both ELISA and LC-MS can be affected by their own biases for different reasons, LC-MS is a less biased technique overall. However, when using data-dependent acquisition (DDA), the LC-MS results will be skewed towards more abundant proteins. This can be explained by the stochastic acquisition process of DDA, during which the most abundant n analytes are fragmented at any given time point. Less abundant analytes might not be chosen for fragmentation or maybe only in an unreproducible manner.

This can be overcome by using data-independent acquisition (DIA), which enables unbiased and reproducible analysis. SWATH DIA, for instance, ensures that the entire detectable HCP spectrum is analysed, without prior knowledge of the analytes. This approach minimises the risk of overemphasising high-abundance proteins and missing the ones in trace amounts. It also reduces the variability seen with DDA methods as it relies on a stochastic acquisition process. Thus, this unbiased HCP detection method eliminates variation between replicates and drives reproducibility.

KP: How do you see advancements in LC-MS technology contributing to HCP analysis in the future?

TK: Commercial ELISA kits cannot account for the variations in process-related impurities, especially when it comes to novel gene-based therapies and vaccines, which are often based on many different cell lines and production methods. Developing ELISAs tailored to specific manufacturing processes would involve lengthy antibody production, which is very time-intensive. In contrast, LC-MS workflows can be adopted within days and meet the rising demand for efficient analysis of gene-based products.

The HCP content of a biotherapeutic or vaccine is subject to change depending on process variations and process developments, such as changes in culture conditions and scale-up. Therefore, HCP analysis must begin from the early drug development phase. This will inform process developers about the possible trigger points and their downstream effects. This knowledge can be used to optimise the production workflow to minimise contamination at later stages, which would otherwise be time-consuming and costly to resolve.

The technological advancements in HCP identification and quantification pave the way for better interpretation and elimination of impurities. Drug innovators must implement benchmarks to reveal the impact of individual HCPs and determine the potentially harmful ones. Once there is sufficient information about the impact of various HCPs, research must also focus on eliminating these impurities with high efficiency and ideally high specificity. This can reshape the HCP analysis framework and the entire biopharmaceutical manufacturing process, and LC-MS plays a crucial role here.