Taking Stock of a Maturing CGT Sector

Published on: 
Pharmaceutical Technology, Pharmaceutical Technology, July 2024, Volume 48, Issue 7
Pages: 10–12

As the CGT market matures, increased focus lays in securing optimal efficacy while ensuring safety.

Advances in development and manufacturing of cell and gene therapies (CGTs) have been focusing on increasing therapeutic efficacy while ensuring product purity and improving patient safety. Efficacy, purity, and safety were major themes at the American Society of Gene and Cell Therapy (ASGCT) Annual Meeting held in May 2024.

On the gene therapy front

Sven Moller-Tank, director, Regeneron Genetic Medicines, Regeneron, points out that part of his company’s approach is to directly address and overcome the commonly identified industry challenges around gene therapy development, such as safety, delivery, packaging constraints, immunogenicity, and durability. In terms of safety, for example, Moller-Tank says that Regeneron is investigating strategies for detargeting adeno-associated viruses (AAVs), which are commonly used viral vectors for gene therapy delivery, away from the liver.

“We have data demonstrating that reducing vector tropism to the liver significantly reduces liver damage and complement activation in non-human primates. In concert with this, we are using antibodies to confer novel tropism to AAVs and enhance their efficacy in desired target tissues. We have used this approach for enhanced muscle and central nervous system targeting and are currently expanding to new tissues,” Moller-Tank states.

A dual AAV approach may be one solution to alleviating packaging constraints encountered within the AAV capsid. “We now know [that an] oversized gene delivery approach can be successful in humans. We are solving for immunogenicity concerns using … [tools] for immunomodulation to combat pre-existing anti-AAV antibodies and enable redosing,” he explains. “Finally, we are applying a targeted liver gene insertion platform we co-developed with Intellia to hopefully enhance the durability of liver gene therapies and have shown in mouse [studies] the potential to sustain expression despite rapid liver growth, a critical feature for treating younger patients.”

AAV vectors have become the most common viral vector used in gene therapy manufacturing, but, until recently, their target specificity has not been quite successful. When focusing on the liver, Moller-Tank explains that some level of specificity can be achieved by targeting the liver through systemic administration. However, specifically targeting extrahepatic tissues has been a challenge, he points out. “Because of this, the development of novel vectors to enhance specificity has been a major focus of the field,” he says.

Moller-Tank emphasizes that using local injection is one method to achieve specific delivery outside of the liver. Meanwhile, the company’s pursuit of improving target specificity has taken it beyond just the liver. Several groups were recently able to uncover the fact that the cochlea is an organ that could benefit from AAV gene therapy via the delivery of a virus locally to the inner ear. “Demonstration of safety and efficacy of AAV delivery to the cochlea in humans may enable the development of several novel treatments for hearing loss,” Moller-Tank says.

Redirecting target specificity

A common approach to increase on-target specificity is utilizing a library-based selection strategy to screen for capsids that have desired properties, according to Moller-Tank. “However, this approach has had very mixed success when selecting for tropism to target organs and can yield capsids whose tropism does not translate across species. Now, the field is moving toward utilizing the same library-based screening approach to select AAV capsids that bind to specific receptors expressed in target tissues of interest,” he adds.

A different approach is to use antibodies to redirect the virus to the receptor of interest rather than selecting for receptor binding by the capsid itself. “In this way, we are able to fully characterize our antibodies independently of the capsid, giving us an extra layer of control over the binding properties of antibody targeted AAVs and allowing the platform to be modular and applicable to several different AAV serotypes,” Moller-Tank states.

Regeneron’s experience in developing antibodies for all types of challenging targets has given the company a formidable understanding of how to push antibodies to their limits. “It remains to be seen how versatile peptide libraries are for binding diverse target receptors,” Moller-Tank says.

AAV clearance

In addition to manipulating AAV vectors themselves to improve target specificity, another major concern is the adequate separation and purification of AAV-vector-based gene therapies in downstream processing. At ASGCT 2024, a myriad of research efforts was shown to be underway to improve the downstream processing stage of gene therapy production. Much of the work is centered around the clearance of empty and partially full capsids (process-related impurities) from full capsids (the product).

For example, one study demonstrated how certain additives or a combination of additives in downstream bioprocessing play a role in improving the separation and clearance of empty and partially full capsids during the manufacturing of viral vectors. Ohnmar Khanal, PhD, Downstream Technology Lead, Spark Therapeutics, and her team focused on copper ions and conducted oversaturated loading experiments using multicolumn anion exchange (AEC) chromatography.

“A current challenge in the downstream purification of recombinant [AAV] is the separation of transgene-containing capsids from empty [capsids] and undesired capsid variants,” Khanal stated in a presentation at ASGCT 2024 (1). “AEX [anion exchange chromatography] resin is often used below its saturated binding capacity to separate empty and full AAV particles on a single column. However, the outcome of the traditional single-column operation can be impacted by variability in load composition, sensitivity to buffer composition, and fractionation strategies,” she explained.

The removal of empty and partial capsids with a high yield has rarely been demonstrated in the scientific literature, Khanal and her team found. Through their study, the team presented three-stage separation strategies for two recombinant AAV (rAAV) constructs in which they demonstrated a way to overcome the trade-off between yield and purity.

The first strategy used resin chemistry (Poros XQ, Poros HS, Poros I) combined with a concoction of kosmotropic agents (i.e., additives), such as ammonium sulfate or sodium sulfate, which promoted capsid stability at low ionic strength. Metal ions (e.g., copper, magnesium, calcium) were then used to promote selectivity of the desired capsid. “Using copper ions improved the resolution between empty and full capsid separation and helped remove other product-related impurities,” Khanal stated in her presentation.


In the second strategy, an appropriate resin/media structure and geometry were used to exclude AAV from smaller resin pores. This approach minimized particle dispersion and improved resolution. “Monolith and membrane chromatography showed better peak resolution than traditional resin based on favorable transport behavior,” Khanal also stated.

The team’s third strategy involved the transition from single-column to multiple-columns-in-series chromatography. This multi-column series chromatography was done by loading 1.5x the resin saturation capacity, which allowed for over 40% reduction of partial capsids and over 95% removal of empty capsids. “Using these strategies, we provide the best modality to achieve maximum empty and partial capsid removal with a yield of over 80%,” Khanal concluded in her presentation.

Meanwhile, in a different presentation given at ASGCT 2024 by Alex Meola, associate director of AAV Downstream Process Development at Oxford Biomedica, it was demonstrated that hydrophobic interaction chromatography (HIC) can be used to separate non-deamidated capsids from deamidated capsids (2). According to Meola, the industry has been focusing on how the deamidation of capsid protein from viral protein 1 (VP1)-specific residues, including the N57 domain, is linked to a loss of in-vivo potency (3).

“Temperature, pH, and storage time have been identified as key factors causing this phenomenon,” Meola said in his presentation (2). “In our quest at Oxford Biomedica—to try and resolve this—we have a static binding capacity residence screening protocol that we run where we look at a variety of different residents. In this case, we decided to look at salt-tolerant residents,” he stated.

Aside from process control to prevent deamidation, Meola’s team hypothesized that extrusion of VP1 and deamidation of N57 are related phenomenon and, furthermore, that this connected phenomenon results in the generation of empty, partial, and full capsids (4). The team also believed that the generation of these mixed capsids significantly compromises the ability of the AEX process to remove empty capsids and deliver functional AAV gene therapies.

Meola noted that charge-driven separation techniques have not provided the much-needed resolution to allow differentiation between deamidated and non-deamidated capsid species. Thus, his team used novel approaches that can be applied to the AEX process to address this complex challenge of clearing both empty capsids and deamidated intact capsids.

In one approach, Meola and his team used HIC to separate distinct peaks with near baseline resolution. After separation, each peak was then isolated and individually reprocessed with AEX. “Surprisingly, we discovered that HIC resolved two different species of AAV capsids with near baseline resolution. Each species was reprocessed on AEX, and all product quality attributes were assessed for the intermediate peaks that were generated. We found that the capsids that were more hydrophobic were also more negatively charged,” he explained in his presentation.

Data from liquid chromatography–mass spectrometry analysis further demonstrated that the more hydrophobic and more negatively charged capsids exhibited significant levels of VP1-specific N57 deamidation., which has been linked to a loss in gene expression, Meola noted. The team’s study thus demonstrated that HIC can be a useful method to separate non-deamidated capsids from deamidated capsids.

On the cell therapy front

Chimeric antigen receptor T-cell (CAR-T) therapies are another sector incurring industry focus as this market matures as well. Similar to AAV vector-based gene therapies, CAR-Ts require an enhanced targeting mechanism. Philip Gregory, senior vice-president, Regeneron Cell Medicines, Regeneron, notes that, while conceptually simple, the redirection of a killer T cell to a tumor target using a CAR is a complex and multi-component problem.

“Thus, success required innovation and iteration in three key areas. Firstly, the methods for the isolation and expansion of the desired T cells from the peripheral blood. Second, the design of the CAR itself to ensure appropriate activation of the relevant signaling pathways in response to binding antigen. Third, the identification and selection of the best target antigens and antibody formats to enable the CAR to recognize the tumor cells,” Gregory states. He emphasizes that advances across all three areas were necessary to see the depth and durability of responses that have driven the recent excitement surrounding the CAR-T modality.

Gregory points out that the initial successes of CAR-T cell therapies have been in liquid tumors, such as lymphoma and myeloma, driven, in part, by the identification of cell-surface-expressed CAR-T cell target antigens that are universally found on both the surface of the tumor cells and on the normal tissue from which the tumor is derived. “For example, B cell maturation antigen (BCMA) is expressed on the surface of plasma cells (the cell type which is cancerous in multiple myeloma). Such universal and high levels of expression are, however, not the norm,” Gregory says.

Furthermore, while B cells and plasma cell aplasia can be compensated for by the infusion of immunoglobulins, most normal tissues are not “dispensable”, Gregory also points out. “Solid tumors, by contrast, lack such broadly expressed target antigens and have the additional constraint that the tumor microenvironment may be immunosuppressive (via the secretion of factors that dampen T-cell activity). Thus, we believe that solid tumors will require multi-layered approaches, such as the ones we have been developing (combining T cells and antibody-based approaches), to achieve the depth and durability of response desired,” he adds.

Gregory highlights the following advances CAR-T cell therapy design:

  • Regulatable CAR-T cell approaches. These approaches include Regeneron’s work on a dimerizing agent–regulated immunoreceptor complex (DARIC) platform. DARIC offers a physician the ability to turn the CAR-T cell off (and back on) in response to a small-molecule drug. This development could have an impact in both normal tissue salvage and in the prevention of rapid exhaustion of the T cells via persistent antigen stimulation.
  • Dual and multi-target approaches. “In these approaches, one can deploy novel architectures such as those used in bbT369 (Regeneron’s investigational CAR-T cell approach for B cell lymphoma) to enable the T cell to kill tumors expressing either, or both, tumor targets. Alternatively, the T cell can be engineered to be a delivery vehicle for a tumor-targeted bispecific antibody. Potentially powerful strategies to prevent tumor escape (the loss of the target antigen),” says Gregory.

What the future holds

Both Moller-Tank and Gregory see a dynamic future ahead for gene therapies and cell therapies. “We are hopeful that we can continue to expand the types of tissues and cell types we can reach with our antibody-targeted AAV delivered gene therapies. We are excited about applying our emerging strategies for addressing immunogenicity as it relates to challenges with redosing and pre-existing antibodies,” Moller-Tank states.

Regeneron is also expecting to make first-of-its-kind clinical progress with its clustered regularly interspaced short palindromic repeats (CRISPR)-mediated targeted gene insertion platform. The company anticipates that its platform has the potential to pave the way for more durable liver-directed gene therapies.

Gregory is confident that CAR-T cell approaches will remain an important modality for treating cancer, both as a standalone modality and in combination with biologics. “We are particularly excited about the potential for the combination of CAR-T with the antibody-based biologics we are developing at Regeneron. To outsmart the tumor will require multi-layered strategies best addressed via these types of innovative combinations. Moreover, we are in a privileged position to be able to explore these in the clinic,” he concludes.


1. Mirasol, F. ASGCT 2024: Additives Play a Role in Removal of Empty and Partial AAV Capsids. Pharm. Technol. online, May 8, 2024.
2. Mirasol, F. Research Shows that HIC Can Effectively Separate Deaminated Capsids from Non-Deaminated Capsids (ASGCT 2024). Pharm. Technol. online, May 11, 2024.
3. Giles, A.; Lock, M.; Chen, S. J.; et al. Significant Differences in Capsid Properties and Potency Between Adeno-Associated Virus Vectors Produced in Sf9 and HEK293 Cells. Hum Gene Ther. 2023, 34 (19–20),1003–1021. DOI: 10.1089/hum.2022.116
4. Meola, A.; Aguiar, A.; Chong, C.; et al. Not Created Equal: Impact of Deamidation on Capsid Heterogeneity and Anion Exchange Chromatography (AEX) Performance. Paper Abstract. annualmeeting.asgct.org (accessed May 8, 2024).

About the author

Feliza Mirasol is the science editor for Pharmaceutical Technology®.

Article details

Pharmaceutical Technology®
Vol. 48, No. 7
July 2024
Pages: 10–12


When referring to this article, please cite it has Mirasol, F. Taking Stock of a Maturing CGT Sector. Pharmaceutical Technology 2024, 48 (7), 10–12.