In conversation: A smarter way to eliminate host cell protein contamination from biologic drugs
Dr Thomas Kofoed co-founder of clinical research organisation (CRO) Alphalyse speaks to Mani Krishnan, VP/GM of Global Biopharma and Capillary Electrophoresis (CE) at SCIEX, who is working on new solutions in mass spectrometry and CE, on how new bioanalytical tools can help biologics developers better manage their risk.
As a CRO specialising in bioanalysis for clients developing biologics, what would you say the main ‘burning issue’ is for those companies?
Biologics development is still a lot younger as an industry than traditional pharma, so there are still many challenges. One of the main issues that clients come to us about is host cell proteins (HCPs). HCPs are incredibly important because they can be biologically active and interfere with how the drug works or maybe worse, act on the patient directly. Even if the HCP isn’t bioactive, it can act as an adjuvant to raise an immune response in the patient and make them resistant to drug therapy. Given it can impact efficacy as well as safety, regulatory authorities designate HCPs as a critical quality attribute (CQA). That means any contamination by HCPs needs to be detected and monitored in the development of the bioproduction process, and carefully controlled throughout the scaled-up biomanufacturing process. This ensures any HCPs present in the final drug product are within the acceptable limits pre-specified by regulators. HCPs are a make-or-break factor for biologics.
What’s the main challenge with HCPs?
Biopharma companies struggle to find a good, reliable way of detecting and removing HCP contamination. They often ask us about how to get a good analytical method for HCPs. Especially for many new modalities, from gene therapies to next-generation antibodies, which have more complex biomanufacture considerations than normal biologics. They use various expression systems for the drug production, such as Chinese hamster ovary, E. coli, and Pichia cells. These host cells alone contain over a thousand HCPs. In addition, there are other sources of protein contamination, such as from cell culture medium supplements, for example, bovine serum albumin. So, removing all this protein contamination is a major challenge. Of course, to do that, they first need to be able to robustly detect the HCPs. That way, they can also detect when the HCPs have been successfully removed. But developing a really good single assay that can analyse the different HCP profiles hasn’t been possible using the approach favoured by traditional pharma; that is ligand binding assays, especially ELISAs.
ELISAs are fast, sensitive, and easy to use, but for biologics, we’ve found issues with reproducibility and coverage. For example, clients come to us saying that from batch to batch, they get different results from their ELISAs, so the assay and resulting data is not convincing or reliable. Of course, this can create problems when you have to explain why to the FDA. Also, if the biologic has been produced using a human host cell line, these HCPs are not very immunogenic, so it’s difficult to raise antibodies to them in animals to create a good ELISA. So, commercial ELISAs can miss a substantial proportion of the HCPs in a drug product. This poor coverage can be a real issue, especially if there’s a missed HCP that can affect the patient or the drug function. Indeed, it’s not something you want to find out at the stage of a Phase 3 clinical trial, when you use for the first time an orthogonal method to ELISA, in response to the FDA asking why the ELISA number might be a “bit high”.
What do biopharma companies do then, if ELISAs are so tricky and HCPs are so critical?
We’ve discussed this with a lot of biopharma companies, and they tell us that they have very few options. One is to develop a process-specific ELISA, which is designed only for one drug product and production process. It’s a lot of investment, and it usually takes 1.5–2 years – time that many developers don’t have, especially if they’re developing their drug on a fast-track schedule.
We provide a service based on a new assay using mass spectrometry (MS), which has a lot of advantages over the limitations of ELISAs. It’s much more reliable and has incredibly good coverage, sensitivity and precision, and it works for any biologic product. It’s unique for gene therapy developers, especially those who have even more complications to deal with than developers of normal biologics. MS offers some unique benefits. First, the mass spec can detect HCPs from different organisms simultaneously, so it doesn’t matter if the proteins in the drug sample are from the host cells, viruses, or any cell culture supplements. Second, we use an application called SWATH Acquisition, a data-independent acquisition (DIA) technique, which allows you to pretty much identify and quantify everything, even without knowing exactly what HCPs we’re looking for, and no matter how high or low the amount [1–5]. Anyone can immediately apply our method to any product, without a 1.5- or 2-year development time. It’s probably the only method or certainly the best you can use for gene therapy products. Now, we’re working on getting Good Manufacturing Practice (GMP) certification for our MS assay, as many gene therapy clients are asking for that.
How often do you find new HCPs?
I would say in each product, we find new proteins that the ELISA didn’t see. Although some proteins are not of concern to the FDA, developers will still want to remove the high-abundance ones. MS characterises HCPs, so with information like an HCP’s molecular weight, process developers can easily adjust the columns and buffers to remove the HCP. With ELISA, you just get a number that is rather arbitrary. So, MS is a tool for process development, for removing and checking clearance of HCPs.
Do you have to optimise your assay at all? How long does that take?
Typically, we say about 8 weeks from when we start the project until the analysis is verified.
And you can validate it. You can use it all the time, in different labs, and still get the same data. I think one of the reasons MS hasn’t been used before for HCPs in the past is because, to have the sensitivity, you had to run the instrument in DIA using nanoflow LC-MS systems. Whereas one of the key criteria we had for selecting a SCIEX TripleTOF 6600 instrument was that you could go up in flow rate and still have the sensitivity we needed, so we use microflow, which is much more reliable and robust than nanoflow [6].
How much cost and time has been saved when you integrate the MS assay in complement to the ELISA in process development?
We actually have a client who asked about how to best document the ELISA or select the best ELISA for his specific process, so we showed that there was a commercial ELISA kit that was suitable for his project. He said this work saved him about 1 year of development time for a process-specific ELISA and about $1M USD for assay development and evaluation.
Impressive! So, what are you working on now, other than getting GMP certification?
As ELISAs are not very good for HCP analysis for developing production processes for newer biologics, we developed a better method: a new ELISA-HCP-coverage method, where we use ELISA as an immuno-capture step, then MS to find out which proteins the ELISA detects [7]. This MS method empowers you to select the best commercial ELISA assay or document your process-specific ELISA assay. We are also working on ways to raise the throughput of our HCP methods, by making them faster. In general, we continue talking to our clients, identifying their analytical pain points and working with then to develop better solutions.
Figure and caption
References
1. Lund RR, Pilely K, Mørtz E. Using LC-MS for Efficient HCP Clearance. BioPharm International. June 30, 2019. http://www.biopharminternational.com/using-lc-ms-efficient-hcp-clearance (accessed April 2020).
2. Lund RR, Nielsen SB, Crawford J, et al. Identification and Absolute Quantification of Individual Host Cell Proteins by SWATH® LC-MS. https://www.alphalyse.com/wp-content/uploads/2018/11/poster_hcp_id_and_quantification_by_swath.pdf (accessed April 2020).
3. SCIEX. SWATH Acquisition, Ensuring Nothing is Missed. https://sciex.com/technology/swath-acquisition (accessed April 2020).
4. Carapito C. Dual Data-Independent Acquisition Method using SWATH® Acquisition-MS for Host Cell Proteins (HCP) Profiling and Absolute Quantification of Key Impurities during Bioprocess Development. On-demand Webinar. SCIEX. https://sciex.com/events/hcp-profiling-quantification-impurities (accessed April 2020).
5. Ivosev G, Cox DM, Bloomfield N, et al. Scanning SWATH® Acquisition Method for Improved Compound Screening. https://sciex.com/Documents/posters/asms2018/Technology/354_Monday_Ivosev.pdf (accessed April 2020).
6. Hunter C, Morrice N. Microflow SWATH® Acquisition for Industrialized Quantitative Proteomics. https://sciex.com/Documents/tech%20notes/Microflow_SWATH_industrialized_quantitative_proteomics.pdf (accessed April 2020)
7. Pilely K et al. A novel approach to evaluate ELISA antibody coverage of host cell proteins—combining ELISA-based immunocapture and mass spectrometry. Biotechnol Prog. 2020 Feb 22: e2983 (doi: 10.1002/btpr.2983). https://doi.org/10.1002/btpr.2983 (accessed July 2020).