By Thomas Kofoed, Rikke Lund and Ejvind Mørtz, Alphalyse
On May 16-19, we attended BEBPA’s annual HCP conference. The organizing committee had prepared an exciting conference program. It consisted of high-level scientific talks, audience poll questions, and panel discussions on topics related to Host Cell Protein analysis by ELISA, LC-MS, and related technologies. Each session was chaired by experienced persons in the field, facilitating high-level discussions and interaction with the virtual audience.
A large audience of skilled HCP experts from biopharma, service providers, and regulatory agencies attended the virtual event. By the way, a format that the organizers handled exceptionally well.
If you are curious about what the attendees discussed, we have summarized the key topics and take‐home message below.
The main topics included
- Regulatory perspectives on HCP-related risk management from FDA and EMA
- HCP analysis in new modalities; IVIGs, AAV, Adenovirus, and Lentivirus gene therapy products
- ELISA reagent characterization
In addition, there were Large Panel Discussions on these hot topics:
- MS standards for LC-MS HCP analysis
- Moving LC-MS/MS into GMP
- Risk assessment
- Rare reagents
HCP Risk Assessment
The HCP Conference saw several presentations about HCP risk assessment and potentially problematic HCPs.
Alexey Khrenov (FDA-CBER) presented the updated quality risk management (QRM) framework. He highlighted the high level of complexity and uncertainty in the risk control using ELISAs. For instance, the differences in HCP profiles between ELISA standard and drug substance does not allow accurate HCP quantification, and immunoassays may not detect all HCPs.
Robust knowledge management and risk-based decision-making in the QRM activity are critical. Hence, it is relevant to look at specific HCPs of possible concern in your drug and make a risk assessment of these as well.
Fengqiang Wang (MSD) and Georgeen Gaza-Bulseco (AbbVie) updated us on the BioPhorum online database of HCPs and other residual impurities. They highlighted specific examples of HCPs of possible concern divided into
A) HCPs with direct biological function,
B) HCPs known to elicit an immune response or be potentially immunogenic in patients, and
C) HCPs that affect drug stability and quality.
The database contains HCPs from CHO-produced biologics. You can find it on https://www.biophorum.com/host-cell-proteins/ along with other common impurities. BioPhorum also continues to update it with frequently detected and “high-risk” CHO HCPs seen in biologics.
Furthermore, the speakers recommended mass spectrometry analysis of final product samples to ensure product purity. In particular, when ELISA detects elevated levels of total HCPs or when observing dilutional nonlinearity.
A clinical risk assessment sub-team within the BioPhorum plans to establish “safe” levels of HCPs based on toxicity and immunogenicity. Thus, they look into immunogenicity analysis tools to define the criticality of “high-risk” proteins.
Finally, Christina de Zafra (Seagen) gave an overview of the risk assessment framework published by her (de Zafra et al., Biotech. Bioeng. 112: 2284-2291, 2015) and widely used in the industry. Christina gave three excellent case examples on HCP risk assessment, including the lebrikizumab and recombinant Factor IX stories, to demonstrate how you use the tool.
Process- and product-related impurities in Covid-19 vaccines
Stefan Kochanek (Ulm University) told a very relevant and exciting story with analytical results of adenovirus-based SARS-CoV-2 vaccines obtained and published during the Covid-19 pandemic. The vaccines included concerning and higher-than-expected HCP levels and a significant variation in the HCP content between some vaccine batches.
He also relayed his learnings from communicating analytical results to the public, regulatory reviewers, and vaccine manufacturers during the Covid-19 pandemic.
Ejvind Mørtz (Alphalyse) showed how to meet the HCP analysis demands for pandemic vaccines in the future. The presentation also included case examples of LC-MS analysis of multiple Covid-19 vaccine batches, including two adenovirus-based and one VLP-based vaccine.
The LC-MS method in the examples used intact protein standards as internal calibrants. It demonstrated high reproducibility and the ability to identify and quantify different viral vector proteins and HCPs for batch consistency comparisons. The method was recently validated, and now a Covid-19 vaccine developer is using it as a release test and for fast-tracking their vaccine.
Finally, Frances Terry (EpiVax) presented a bioinformatics immunogenicity prediction tool based on multiple parameters. These included potential HLA epitope binding, activation of CD4+T cells, the humanness of epitopes, and HCP abundance. Frances Terry also showed a case study of the HCPs identified in Covid-19 vaccines.
Regulatory Perspectives on HCP-related risk management from FDA and EMA
Like the previous years, the BEBPA conference included presentations by two regulatory bodies; Alexey Khrenov from US FDA-CBER and Erika Friedl from the Paul-Ehrlich-Institut. Furthermore, Niomi Peckham presented from the US Pharmacopeia.
As mentioned earlier, Alexey Khrenov discussed the updated Quality Risk Management framework (QRM), focusing on HCP-related risks. It was surprising to hear that many projects struggle with the project-life-time supply of ELISA reagents for HCP analysis. Furthermore, the changes in reagents or ELISA method lead to significant delays.
Alexey estimated that half of the projects involving a change of reagents ran into issues and problems bridging from the old to the new assay. Therefore, it is critical to have a life-cycle approach to QRM for HCP, a supply forecast plan, and a backup plan with alternative methods.
HCP characterization and life-cycle management were also the main points of the talk by Erika Friedl. Erica presented a case where a product had been on the market for several years: the reagents ran out, and the manufacturer thus needed to switch to a new assay.
The new assay showed much lower sensitivity, problems with dilutional linearity, and a change in MW distribution of covered HCPs compared to the old assay. Therefore, it gave challenges with a lack of comparability and insufficient characterization of the new assay.
Erika Friedl mentioned an increasing regulatory focus on problematic HCPs and that you need to detect, monitor, and remove these specific HCPs. Again, it emphasizes that you need to plan for ELISA reagent characterization and supply to ensure that you can correctly monitor the HCPs throughout the whole lifetime of your product.
Niomi Peckham (USP) presented the current status of the ‘Host Cell Protein Expert Panel’ on the proposed new USP chapter with the working title: Identification and Quantification of HCP Impurities in Biological Products using Mass Spectrometry. The draft is expected to be released for review in early 2023 and will contain general guidance and best practices.
HCP analysis in new modalities; IVIGs, AAV, Adenovirus, and Lentivirus gene therapy products
Kevin Van Cott (Prolytix) focused on Immune Globulin Intravenous (IGIV) products and analytical challenges for HCP analysis. The IGIV is polyclonal and purified from thousands of individual human serum donors. Thus, the active ingredient IGIVs are highly heterogeneous, as are the human “HCP” impurities. Kevin showed some exciting examples of analytical improvements and also case studies.
Yiling Bi (Sangamo) presented on Host Cell Protein analysis for gene therapy products. She summarized the manufacturing platforms and process steps for adeno-associated virus (AAV) vectors for gene therapy products and the impurity analysis strategies for different expression systems.
The HCP characterization is challenging for many reasons; the complexity of manufacturing, variation in upstream and downstream processes, interference from other impurities, lack of representative standards for assay development, and diversity in the methods for AAV concentration measurements. Yiling also gave case examples for comparing results obtained with different HCP Coverage methods. These included 2D gels, the ELISA-MS method, and quantitative LC-MS data for multiple production lots.
Albert Molina (Orchard) had analyzed the protein profiles of lentiviral vector products for gene therapy by LC-MS. He gave an excellent overview of the LVV biology, production processes, and the multiple sources of residual protein impurities in CGT products.
The presentation showed several case examples, including different purification processes, GMP batches, and PPQ runs. The cases emphasized HCP clearance, comparison to selected ELISA results, and batch-to-batch comparisons on the complex protein profiles.
The complexity arose from proteins originating from various sources, including the lentiviral vector, human cell line, fetal bovine serum (FBS), and other proteins used throughout the manufacture.
ELISA reagent characterization
Several conference presentations discussed the development of ELISA antibodies against HCPs and the characterization of critical reagents in the ELISA method.
Martha Stables (Sanofi) gave a fascinating insight into using MS as an orthogonal method for developing and characterizing HCP ELISAs. She showed how to characterize various platform-specific ELISAs, using affinity enrichment with ELISA antibodies combined with 2D-PAGE and MS.
She further highlighted the benefits of using MS analysis on different mock samples before immunization and generating HCP ELISA antibodies. The analysis ensures that the composition of HCPs in the mock mimics the production process.
She then gave an example of how a mock sample containing a highly abundant HCP resulted in more than 50% of the antibodies recognizing this specific HCP. Thus, the HCP ELISA would measure the abundant HCP instead of the total HCP amount in some samples analyzed.
Jan Pippel (Biogenes) brought an interesting case study of accelerated implementation of a hybrid HCP ELISA for a client. Initially, the client developed their clinical candidate for a different project, but it now showed promising results for treating another disease.
The project had no time for creating a process-specific HCP ELISA, so they looked into the possibilities of repurposing reagents from related processes instead. After assessing various reagents, they set up an HCP ELISA method and qualified it in a much shorter time frame.
Summing it up
Overall, the virtual BEBPA 2022 HCP symposium was well-organized and gave us much more knowledge about HCP impurity control, risk assessment, and inspiration for new projects and improvements.
We will undoubtedly participate again at the BEBPA HCP Conference in May 2023.