Optimizing Complex mAb Purification

Purification of complex mAb modalities: Ask the specialists Antibody therapeutics White paper | Bioprocess resins Jett Appel Field Application Scientist, Thermo Fisher Scientific Jett Appel has been a field application scientist for purification at Thermo Fisher since 2021, supporting purification development and scale-up for monoclonal antibodies, viral vectors, nucleic acids, and recombinant proteins. Prior to joining Thermo Fisher, Jett worked at Avid Bioservices, a contract development and manufacturing organization (CDMO) based in Orange County, CA. During his 5 years of working at Avid, Jett first worked in process development and eventually transitioned to a senior engineer role in the Manufacturing Sciences & Technology (MSAT) team, where he supported purification development and technology transfers for monoclonal IgGs, IgMs, bispecific antibodies, scFvs, and recombinant enzymes up to the 2,000 L bioreactor scale. Jett received a bachelor’s degree in Chemical and Biomolecular Engineering from UCLA in 2016. While at UCLA, he was involved in epigenetics research in the lab of Steve Jacobsen, which included studying the pathways involved in RNA-directed DNA methylation in Arabidopsis thaliana. Nicholas Bardol Manager, Manufacturing Services, Thermo Fisher Scientific Nicholas Bardol has been working in the pharmaceutical industry for the past 13 years as a process development scientist focusing on biologics purification and small molecule conjugation. His work at CDMOs has supported over a dozen novel molecules through CGMP manufacturing and regulatory agency filings. He joined the St. Louis, MO, Thermo Fisher site supporting the Pharma Services location in 2018 as a member of the downstream process development group. As the complexity of novel monoclonal antibody (mAb) modalities increases, traditional purification and polishing processes may no longer be effective. Outdated resins and archaic polishing methods can decrease the efficiency and effectiveness of mAb purification, while new mAb structures can also bring new impurities that are difficult to remove. We asked three in-house specialists at Thermo Fisher Scientific for their advice on how to overcome the challenges of purifying complex mAbs, and how to optimize your polishing strategies for maximum efficiency. Q: What are the main challenges faced in the purification of complex mAbs compared to traditional monoclonal antibodies? Jett Appel (JA): It’s hard to pinpoint the main challenges because it depends a lot on antibody structure, and antibody structures are diverse, creating diverse challenges. Often, when we talk about complex mAbs, we are talking about bispecific or multispecific antibodies, antibodies where the Fc domain—or even antibody fragments, such as Fabs or scFvs—are modified. So, for example, you may not be able to utilize traditional affinity steps for purification, or it could result in product impurities that are harder to separate. Good yield or purity tends to be obtained differently from these different complex mAbs. Nicholas Bardol (NB): Increasing numbers of drug candidates are bispecific, which are very difficult to separate, because bispecific expression systems produce a lot of parent mAb fragments and paratopes as impurities. Most fragments not only share a lot of the primary characteristics we would usually exploit in traditional purification strategies, but they can also be more biologically active. So, if they outcompete the drug candidate for the epitope they’re targeting, then the efficacy of the drug candidate really diminishes. Kelly Flook (KF): Because structures are growing more and more complex, heterogeneity and aggregation become increasingly challenging. However, it’s not just complex mAbs that are suffering from increased aggregation levels. As the industry moves towards more cost-efficient mAb production, higher cell densities and processes to increase titer can impose additional challenges downstream with respect to process- and productrelated impurities. Q: How do the structural characteristics of complex mAbs impact the purification process and pose unique challenges? JA: It depends on the exact structure of the complex mAb. Bispecific or multispecific antibodies often have different heavy chains or light chains and this can result in mispairing. Consequently, these antibodies are very similar in physicochemical properties and can be hard to separate or may get copurified with traditional protein A affinity steps. Other formats of complex mAbs—for example, antibodies that have been multimerized—are much larger than traditional mAbs. This can result in steric hindrance or lower diffusivities. That’s a unique challenge because it may lead to lower resin capacities and therefore larger columns, which could ultimately reduce productivity as scale increases. NB: In the case of antibody–drug conjugates, they’ve been modified specifically to take on a drug linker, and that can pose many challenges—we see issues with free thiol or unbound drug linker. In the case of bispecifics, it’s hard to use charge-based separation because the sequence of impurities will overlap a lot with the sequence of the target antibody. Just as importantly, we may not be able to use our typical polishing strategies of anion and cation exchange. Instead, we have to rely more on things like multimodal or hydrophobic interaction chromatography (HIC) as a different tool to target another characteristic of those impurities and see if we can remove them with a different mechanism than one the industry typically relies on. Kelly Flook Senior Manager, Product Management, Purification Products, Thermo Fisher Scientific Kelly Flook is a dynamic product manager specializing in downstream purification of biotherapeutics and new product development. With over 15 years of experience in product development, Kelly has an excellent track record of bringing the right product to the market. She started her career as a scientist specializing in the development of polymerbased analytical columns for HPLC before moving into purification R&D and then product management. Kelly went to school in the UK, receiving a bachelor’s degree in Analytical Chemistry from the University of Northumbria, Newcastle, and a PhD in Polymer Chemistry from the University of Durham. 2 thermofisher.com/antibody-derived-therapeutics Q: What are some of the common impurities or contaminants encountered during the purification of complex mAbs? And how are they typically addressed? JA: Some common impurities overlap with traditional mAbs— e.g., host cell proteins, residual DNA, and aggregates. Often in these complex mAbs, we tend to see higher levels of aggregates because, as these structures are engineered, the impurities are either produced more in the cell culture itself or can be generated under traditional mAb purification parameters, such as exposure to low pH conditions. The complex mAbs can be less stable and form impurities inherently within the process. Other common impurities can be mispaired homodimers, for example. They’re typically addressed through a traditional ion exchange to separate the target product from these impurities. But, as Nick was alluding to earlier, this can often be a challenge. So, you may need to implement HIC or multimodal chromatography as well. NB: I already alluded to impurities that we see with unbound drug linker or bispecific isoforms, but the biggest one that we deal with is host cell protein (HCP). In some of the newer mAbs, there is a lot of change in the isoelectric point (pI) of the target antibody, and it can overlap with the HCP pI. Traditionally, antibodies have high pI relative to the HCPs, making charge-based separation very straightforward: you bind up your antibody, and flow through all the HCP. But suppose the target is sitting right in the middle. Then you either need multiple steps to capture and resolve everything on either side of the pI or you have to look at another option like multimodal chromatography or HIC again, just to get rid of those impurities that have historically been relatively simple to remove. KF: Many manufacturers have developed platform purification workflows for their traditional mAbs, but these don’t always work for their more complex molecules. This is when you need to turn to HIC or multimodal chromatography. For affinity purification, although protein A continues to be the workhorse, it’s not always the best option, depending on the binding domains of the specific antibody. This is where our Thermo Scientific™ CaptureSelect™ affinity resin product line comes in. Q: Could you discuss any specific strategies or technologies employed to overcome the challenges associated with complex mAb purification? JA: We’ve seen customers implementing unique affinity technologies. Typically, we’ve utilized traditional protein A affinity chromatography, but we can see that, depending on how these complex mAbs are engineered, other affinity ligands that target different domains of the antibody could be advantageous. For bispecifics, we’ve seen engineered antibodies where one arm doesn’t have a CH1 domain. So, for example, you could utilize an affinity resin that targets CH1. That homodimer won’t bind to the resin and will just flow through instead of getting copurified with protein A. Another interesting strategy that customers use is to leverage the avidity of the target molecule or the impurity to the resin and utilize pH gradients. This is used, depending on the antibody format, when there is more than one specific domain, or if the target molecule or the impurity has fewer domains than the other. Where traditional operating conditions for polishing steps work for standard mAbs, they may not work for complex mAbs. These may then need a unique set of conditions; for example, operating cation exchange under a more alkaline condition may help improve resolution. As Nick mentioned earlier, other technologies—such as multimodal chromatography or HIC—can be a useful tool if the standard methods do not work. KF: Our ion exchange resins have high capacity and robust salt tolerance to be able to handle higher titers in bind-and-elute mode. Using a flow-through process, your capacity becomes capacity for the impurity, which is typically <5% at this point, meaning you can significantly improve productivity and reduce resin consumption. NB: I want to highlight a couple of things both upstream and downstream of purification. The industry has been chasing high titers in their clone selection for as long as biologics have been around, but choosing the clone with the highest titer may not be the best strategy. Instead, choosing the cell line with the lowest levels of a complex impurity might be the best mechanism. So be very forward when thinking about clone selection. Another thing is analytical strategies. A lot of these new impurities come with bispecifics and isoforms that share the same charge; using all the traditional analytical techniques, you might not get total resolution. They may co-elute, and you won’t know the contaminant is present until the end stages, or you may chase something that isn’t there. So, looking at orthogonal assays for size or isoform pattern and charge heterogeneity can be hugely beneficial. Some less commonly used techniques can be powerful here; using two-dimensional liquid chromatography (2D-LC) and native mass spec to see those hidden isoforms can save a lot of time when you’re trying to develop your purification strategy. 3 thermofisher.com/antibody-derived-therapeutics Q: Are there any particular purification techniques or methods that have shown promising results in effectively purifying complex monoclonal antibodies? JA: Utilizing innovative affinity technologies that target different domains can be useful to leverage the unique structures of these complex mAbs. That’s one technique that has shown promise. Innovations looking at different optimization conditions for polishing steps and unique polishing resins can be a useful strategy as well. A lot of these complex mAbs may not be stable and can form higher-level aggregates; therefore, the ability to operate a wider set of conditions and separate species using resins that might be more selective for different product species can be useful. NB: Affinity capture is the big space where we’re seeing major changes and alternative mechanisms for purifying complex mAbs. For bispecifics, they have both a lambda and kappa light chain, which is something that you can target—you can capture one side, then capture the other side. That’s a perfect mechanism to remove all the fragments of the parent antibodies that would otherwise need convoluted polishing strategies to remove. KF: We recently launched the Thermo Scientific™ POROS™ Caprylate Mixed-Mode Cation Exchange Chromatography Resin, our first mixed-mode cation exchange resin. As many may know, caprylic acid is used to pull down high levels of aggregates upstream. Attaching this to the bead allows aggregation to be effectively removed downstream, reducing aggregation from 20% to less than 2% without the need to then remove the caprylic acid. An added benefit is that this resin was designed specifically to be used in flow-through mode to allow high recovery over a wider range of loads compared to other resins positioned for aggregate removal in flow-through mode. Q: How do the scalability and throughput of complex mAb purification processes compare to traditional mAbs? And what considerations need to be taken into account? JA: It depends on the antibody and the specific challenges present. Some of these complex mAbs can have lower capacities, so you may need to utilize resins at lower loading densities or, in order to achieve appropriate resolution, operate at lower flow rates. Ultimately, this can lead to larger columns as you scale up, increased buffer consumption, and longer processing times, which can reduce efficiency. Therefore, it’s important to optimize early on and utilize different strategies—e.g., using conditions that allow you to operate at faster processing times and higher capacities to reduce that burden as you scale up more complex mAbs. NB: We’ve seen a need to prioritize resolution, and we were seeing lower loading on columns through the entire downstream process as we tried to get higher resolution. This can mean more unit operations in manufacturing and taller polishing columns that are more difficult to work with, as well as big increases of buffer volumes and longer residence times. So, the variations may become a little more complicated to transfer and scale up. To alleviate these issues, many programs are using very specific pH ranges or nontraditional buffers (e.g., Good’s buffers such as MES or bicine) that we haven’t seen used on a manufacturing scale historically. Pressure-tolerant resins can also be very beneficial—you can get the tall bed heights and the high resolution that you need. New inline dilution strategies to make the buffer volumes more manageable are also hugely beneficial when you’re scaling up. Q: What role does process optimization play in improving the efficiency and yield of complex mAb purification? And what are some key factors to consider during optimization? JA: Process optimization plays a critical role because your goal is to maximize purity and recovery and still maintain the quality of the product. But you also want to ensure that it’s scalable, robust, and overall low-cost once you have a fully scaled-up process. Monitoring the process is important to better understand how it operates, ensuring that you’re not operating at the edge of failure and identifying the parameters that could impact your process. NB: Complex mAbs tend to have lower bioreactor titers overall because they are more difficult to grow, so even a couple of percent gain can be meaningful. Relying more heavily on design of experiments (DOEs) and quality-by-design frameworks can help labs to get the most information possible with the least material, so that those optimization strategies can be executed with just one batch. 4 thermofisher.com/antibody-derived-therapeutics For research and development use and further manufacturing in support of FDA-regulated end uses. Not for diagnostic use or direct administration into humans or animals. © 2025 Thermo Fisher Scientific Inc. All rights reserved. All trademarks are the property of Thermo Fisher Scientific and its subsidiaries unless otherwise specified. WTP-10889550 0625 Learn more at thermofisher.com/antibody-derived-therapeutics Q: What are the potential future advancements or innovations in complex mAb purification that could address existing challenges and further improve the process? JA: Firstly, there are continuing advancements of different affinity technologies. Different affinity ligands target different domains, so designing new ligands that are specific for different domains or have different characteristics that can be leveraged for separating difficult-to-resolve impurities in these complex mAbs is important. Another interesting advancement is the use of chromatography modeling software. Mathematical models can reduce the number of DOEs required and help us have a better understanding of the process. We’re even starting to see customers utilize AI or machine learning techniques for their large datasets of different antibody sequences and structures. AI can look at these datasets and use pattern identification to predict the behavior of different chromatographic resins. NB: The other thing that we’re seeing customers move towards is process analytical technologies. Not all these mAbs can be tightly controlled just by looking at a UV signal. Having inline or online detection or mass-based detection—instead of just looking for chromophores that may not exist or may completely overlap with your antibody—has real potential to solve challenging separation issues that we may not be able to resolve at scale. KF: Achieving high purity is all about having the right resin for your molecule. Chromatographic efficiency can be achieved by reducing the particle size, but that comes with the expense of flow rate or bead height due to the pressure limitations of hardware. Modulating or designing the right chemistry or ligand on the surface to be selective between target and impurities is a more effective way to increase purity and yield. But as molecules become more complex, this means more tools are needed across different variations from the traditional mAb. mAb developers can engineer in things like a C-tag to enable selective removal or a specific sequence that interacts with specific ion exchange or HIC groups.