Review of BEBPA’s HCP Conference 2021Key topics & take-home messages

June 4 2021, by Rikke Raaen Lund and Thomas Kofoed

Review of BEBPA HCP conference 2022

BEBPA’s 9th annual Host Cell Protein Conference took place, again as a virtual conference, on May 17-19, 2021.

The BEBPA organizing committee once again assembled an exciting program for the approximately 120 participants. They represent more than 20 countries and various organizations, including biopharma, service- and technology providers, and regulatory agencies.

Like previous years, we created a conference summary to share the key topics and our take‐home messages from the meeting.

On a general note, many of the talks were again concentrating on mass spectrometry (LC-MS). They highlighted how the more detailed information obtained from LC-MS is changing the way we perform HCP analysis.

Risk assessment and focus on specific HCPs, even at very low levels, is a hot topic. Thus, several of the talks showed examples of how to deal with this.

Main Topics

  • Surveys: Summary of different surveys conducted during the conference
  • Regulatory input
  • Specific HCPs of concern
  • Gene Therapy Products
  • Bringing HCP LC-MS analysis into a GMP environment


As in previous years, each session started with some questions to the audience. Here are the results of a few of the most interesting polls:

  • Does your company include MS data on HCPs to regulatory agencies? Yes: 43%, No: 57%
  • How are your normalizing your MS data? Intact proteins: 56%, Peptides: 29%, Other: 15%
  • Are you using or planning to use validated MS analysis for HCP analysis? Yes: 63%, No: 37%

Compared to previous years, the usage of MS is increasing. We look forward to seeing the historical data on these survey questions on the BEBPA SoMe-platforms.


Regulatory Input

This year, only one presentation came directly from the Regulatory Authorities. From the Paul-Ehrlich-Institut and EMA, Erika Friedl focused on the use of mass spectrometry for HCP analysis.

Although still not a requirement, it is an excellent idea to include MS data in applications to the regulatory authorities. As she said, “when 7 of 10 applications include MS data, you need a good excuse for not doing it”. The inclusion of MS data must be followed by a detailed risk assessment, e.g., the use of in-silico immunogenicity evaluation. The risk assessment is not an easy task, but you must perform as well as possible.


Specific HCPs of concern

Mass spectrometry provides the identity and quantity of each HCPs used for risk assessment. Several talks focused, directly or indirectly, on this subject.

Suli Liu, from Biogen, gave a great talk about identifying two residual HCPs produced from a CHO cell line as ‘difficult to remove’ during a downstream purification process. They performed a risk assessment to understand the impact of the HCPs and to guide process development. To understand HCP removal capability at each step and establish a robust biopharmaceutical manufacturing process, they utilized an end-to-end approach from upstream to downstream. Thus, the outcome was an optimized process that reduced the amount of the two HCPs to an acceptable level.

Shawn Li, Merck, showed that sometimes it is meaningful to go to sub-ppm levels when looking at HCPs causing problems. Shawn Li evaluated and optimized the activity-based protein profiling (ABPP) approach very elegantly to identify active enzymes responsible for polysorbate degradation at a very low level.

The study did not find any active Phospholipase B-like 2 (PLBL2) in process-intermediates from the two mAbs projects. Instead, ABPP identified that phospholipase A2 group VII (PLA2G7) and sialic acid acetylesterase (SIAE) were possible root causes of polysorbate-80 degradation. Activity-based profiling promise to enables better polysorbate degradation investigations for biotherapeutic development.

Karen Price, Bristol-Myers Squibb Company, presented a very nice case study from back in 1997 on how observing adverse events in a Phase 2 trial halted a clinical-stage project. Using the available tools at that time, they identified an impurity found as a peak by RP-HPLC analysis, induced in vitro histamine release in whole blood from one subject that experienced a histamine-like reaction.

They also found that the impurity had significant sequence homology with monocyte chemoattractant protein-1 (MCP-1), a potent inducer of histamine release. So, this was one of the first examples of how a single HCP can be a risk for an otherwise promising project.

Martha Stapels, Sanofi showed how to use mass spec data for identifying HCPs and purify them away. Whereas hitchhiker proteins sticking to the DS were harder to remove. Further, this presentation included examples of how a detailed risk assessment and a targeted MS approach can lead to a sensitive identification of proteins that might impact product quality.


Gene Therapy Products

Gene therapy products represent a group of new biologics with additional complexity in HCP control challenges. Many commercial ELISAs provide low coverage and are useless for detailed monitoring of HCPs.

Since many of these projects are intended for severe indications, they often need fast development. This conflicts with the timeline required to develop product-specific ELISA. However, mass spectrometry provides a good alternative. It can accommodate the high complexity of the samples, and development for individual projects takes only a few months.

Rikke Lund, from Alphalyse, gave a rapid-fire talk (only 5 minutes), focusing on the comparison of HCPs found in 11 different Gene Therapy projects using mass spectrometry. She saw that several HCP classes were typical for gene therapy products.

These, although in general recognized by ELISA antibodies found in commercial kits, had a high tendency to show non-specific binding. This tendency makes the use of any ELISA, commercial or process-specific, difficult. Feedback from Alexey Khrenov, FDA, included a comment that this was new and interesting knowledge.

Olaf Stamm, Charles River, found four HCPs accounted for the majority (67%) of the HCP content in an AAV-based gene therapy product. At the same time, they observed insufficient coverage of commercial ELISA (<28%), indicating that the ELISA did not contain antibodies for these four host cell proteins.

Identifying immunogenic peptides and then synthesizing them chemically enabled the optimization of the ELISA antibodies to detect these four HCPs.


Bringing LC-MS analysis of HCPs into a GMP environment

Since introducing mass spectrometry as a tool for monitoring HCPs, it has been a hot topic if it is possible and of interest to bring LC-MS into a GMP environment. A requirement to use the analysis as a release assay.

LC-MS analysis is very complex and has many variables. However, Thomas Kofoed, Alphalyse, showed how using internal standard proteins and automating the sample preparation procedure can control most of these variables, leading to a reproducible and robust assay.

Alphalyse investigated the behavior of the standard proteins included in samples from 10 projects before and after normalization. It was possible to see that normalization reduced most of the variation from sample prep and LC-MS analysis. Reducing this variability will be a crucial feature to bring the analysis into a GMP environment.

The second part of his talk focused on a specific project and the assay qualification for this project. The parameters investigated were compliant with ICH guidelines. The data was quite convincing and used for the IND submission.

Overall, the virtual BEBPA 2021 HCP conference was very well-organized. We look very much forward to, hopefully, be able to meet in person next year in Dubrovnik to celebrate the 10th BEBPA Host Cell Protein Conference in May 2022.


Talk to us

Whatever protein-related challenge or question you may have, we would love to help. Our experts can help you decide on the best analytical approach for your project by email or online meeting - providing advice without obligation.

I would like an
This field is for validation purposes and should be left unchanged.