Finding Success in AAV Manufacturing

Published on: 
BioPharm International, BioPharm International-03-01-2021, Volume 34, Issue 3
Pages: 34–37

Homology Medicines discusses how the clinical-stage biopharmaceutical company met its need for supply of adeno-associated virus vectors.

Gene therapy is a young and growing industry, and suppliers have been working to meet the demand for flexible manufacturing capacity to produce the viral vectors needed for clinical trials and commercialization of gene therapies. “Innovations in manufacturing are a key driver of the success of these products,” says Tim Kelly, chief technical officer of Homology Medicines. The clinical-stage biopharmaceutical company is addressing rare genetic diseases using its proprietary technology platform, which is based on a novel set of adeno-associated virus (AAV) vectors derived from human hematopoietic stem cells for either nuclease-free gene editing (inserting a corrected gene into a person’s DNA) or gene therapy (transferring corrected genes to specific cells in a person’s body). In 2018, the company built its own good manufacturing practice (GMP) manufacturing facility for internal production of its AAV vectors, beginning with 500-L bioreactors and scaling up to 2000-L bioreactors in 2019. BioPharm International spoke with Kelly about the challenges of AAV manufacturing and the benefits they have found in having a flexible facility using single-use technology.

AAV manufacturing

BioPharm: What do you see as the top challenges with AAV manufacturing today?

Kelly (Homology): Unlike the manufacturing of small or large molecules, AAV manufacturing presents a different and novel set of challenges. Some of these challenges led us at Homology to our early decision to develop our own in-house manufacturing and process development capabilities. These challenges include the practical availability of capacity; physical and technical expertise; and the breadth of technical difficulties associated with developing novel technologies and the learning curve required. Yes, there are growing contract development and manufacturing organizations (CDMOs) that can assist, but we believe that owning, controlling, and executing all of the development and manufacturing processes allows us to accelerate and better control that steep learning curve. We do continue to work with external partners, but having our own internal capabilities provides us the opportunity to be in full control of our own destiny.

Another challenge facing the industry is the limited experience and expertise with gene therapy manufacturing platforms and techniques. Organizations cannot rely solely on systems and workflows of the past but need to hire a team of experienced technical operations experts who are willing to contribute to building a new set of expertise. At Homology, we intentionally challenged the ‘old way’ of thinking and developed a model that fits our needs across our development and preclinical pipeline of gene therapy and nuclease-free gene editing programs. We are fortunate to have on our team some of those who were responsible for the first FDA-approved single-use bioreactor, and they bring that know-how to our in-house process development and manufacturing operations. This unique advantage enables us to leverage these experiences and insights across our dual platform and continues to inform the development of new treatments.

I am hopeful that as more focus is directed toward efforts on potential one-time treatments, key learnings will continue to drive gene therapy manufacturing advances. In fact, it should be expected that the AAV manufacturing learning curve will be faster than previous learning curves in the biotech industry. 

BioPharm: What is the role of analytical methods in AAV process development and manufacturing? Can you share any best practices you have developed?

Kelly (Homology): Since we focus on rare genetic diseases, there is potential for a quick transition from the early development phase to a registrational trial followed by approval and commercialization. Because of this need for speed, Homology invested early on in robust analytical development capabilities and a platform model that produces reliable data, which allows us to quickly identify opportunities for improvement while gaining vital process characterization knowledge. In fact, our early analytical methods inform development at every step along the way, and they are crucial to driving learning and understanding our AAV vectors.

Start early and utilize platform methodologies are the two best pieces of advice I can give as it relates to analytical development. Anticipate that regulators will continuously increase requirements and expectations, and definitely plan to qualify methods as early as possible. Additionally, being able to apply these methods across all your programs and platforms will help expedite the development timelines for early phase programs.

We have developed 43 assays that we use across products for yield, purity, and more broadly, characterization. Our analytical development team uses these methods to test our constructs and check for any differences. Analytical development capability and implementation is vital to drive quality, consistency, and learning. 

In-house vs. outsourced manufacturing

BioPharm: Can you explain some of the pros/cons you considered in making the decision to do AAV manufacturing in-house?

Kelly (Homology): When considering the decision to build AAV manufacturing in-house, we were fully aware of the expense associated with building internal process development and manufacturing, and the potential challenges of convincing the Board and others that investing early on is the best strategy. When contemplating this decision, we looked at the gene therapy industry as a whole, and the reality was that the industry was booming. While the technology is still quite emerging, there are over 200 clinical trials with AAV-based therapies. What did this mean for us? It became clear that we wanted to ensure control over timelines and drive the mastering of this new technology without having to worry about capacity and competing interests. We searched for a space where we could operate under the same roof as our research and process development labs. We chose to locate our corporate headquarters in Bedford, MA, where there was greater access to space. We built our 25,000-ft2 internal GMP manufacturing facility in 12 months from design to operation alongside our corporate headquarters. This was indeed a large investment, especially when we were still a preclinical company, but it has proven to be a key strategic asset of Homology, especially during the pandemic, which has significantly impacted companies that are reliant on external sources that are experiencing major capacity constraints. At this point, it is difficult to imagine a world where we are not operating our own vector production capabilities and forging our own path forward in our internal facility that we have built to support all of our gene therapy and gene editing programs.


Furthermore, it is clear that a great source of value for our company is the ability to learn, master, and leverage our internal vector capabilities. These capabilities allow us to develop, learn from, and optimize methods and processes supporting an integrated system that works best for our technology. For example, starting in 2019, we executed a 2000-L bioreactor scale with our HEK293 serum-free suspension platform. This project would have taken many more years to accomplish if not done internally.

BioPharm: What were the keys to success in building and starting up your GMP facility?

Kelly (Homology): The most important key to success is building a strong and experienced team who are hungry to rapidly learn and who are in complete alignment with our ‘One Homology Team’ ethos. When building and running our GMP facility, we benefited from incorporating learnings from the industry and our own internal experiences to develop ‘Homology best practices.’ By establishing these, we were able to avoid adopting bad habits in manufacturing genetic medicines. Another key to success includes having our own data readily available, as data-driven decisions are crucial and often time-sensitive. We also saw the immediate value of investing in our team. We understand that manufacturing is just as essential as our research and development efforts, and our technical operations team has grown in line with the importance we place on the function. We built a core team of employees who are innovative thinkers and team players and who have the proven technical and operational experience.

An important thing to consider when beginning to manufacture AAVs is to determine the most appropriate process. After evaluating a variety of systems, we saw several advantages of using a HEK293 cell transfection system, which demonstrates proven performance and is a faster and more simple and consistent system. In a head-to-head study, our data showed that with our vectors, the HEK293 system resulted in superior in-vitro infectivity and in-vivo efficacy compared to a baculovirus process (1). With this system, we have executed a 2000-L bioreactor scale using Homology’s commercial manufacturing platform process and showed clear and linear scalability of our HEK293 platform. Although not aligned with general industry assumptions claiming that this type of manufacturing system may not be as robust or easily scaled, our data-driven analysis proved that this was the most beneficial system for our gene therapy and gene editing programs.

BioPharm: What benefits do you see in having CDMO partners in addition to your in-house manufacturing? What do you see as the key requirements for an outsourcing partner?

Kelly (Homology): As we planned ahead, having a dual source was advantageous. We knew it was important to have strong partnerships with CDMOs to complement our internal efforts and stay adaptable as Homology grows. Selecting a CDMO that fits the company’s evolving needs and can align culturally is paramount to the continued success of the collaboration. Our selection process is a lengthy one of looking at the timely supply to meet the needs of our programs and assessing their strengths with the intention of building a long-term relationship. Over time, the trust and mutual respect between the companies becomes increasingly valuable as we advance our programs, including the world’s first phenylketonuria gene therapy clinical trial, pheNIX, and prepare for potential commercialization.

Flexible facility

BioPharm: What have you seen as the benefits of using single-use systems and a flexible facility design?

Kelly (Homology): I can’t emphasize enough the benefit of our single-use bioreactor system that serves both our gene therapy and gene editing programs, truly allowing us to accelerate timelines in a highly efficient manner. Our single-use containers and systems, which are clean, sterile, and fully closed off from the environment, result in a fast and flexible manufacturing process, as we do not need to have downtime for cleaning or other things that would be required with a different system.

BioPharm: What are the key technologies that have allowed you to create a flexible facility?

Kelly (Homology): We designed a facility that is right-sized for Homology as we are today, with the ability to expand in the future as needed. Our single-use technology allows for rapid turnaround times, and we have semi-automated systems to balance robustness and flexibility. For us, speed and ability to rapidly integrate learnings were at the forefront of our internal capacity expansion strategy. 

Overall, we have established a high quality, reproducible, commercial-scale manufacturing platform. We have developed a comprehensive ‘plug-and-play’ system, whereby we can switch out capsids and transgenes as needed across our gene therapy and gene editing programs. This platform now includes the execution of more than 450 unique constructs and is now approaching 600 unique suspension lots. Not having to reinvent the wheel for each new product allows for greater efficiencies and may benefit patients over time by reducing product development and potential commercialization timelines. 


1. M. Lobikin, S. Ahmed, N. Avila et al., “Development of a Scalable Platform for GMP Production of High Quality, Novel Clade F rAAV Vectors Following Comparison of HEK293 Mammalian and the SF9-Baculovirus Systems,” Presentation at American Society of Gene & Cell Therapy Annual Meeting (Washington DC, 2019).

About the author

Jennifer Markarian is manufacturing editor for BioPharm International.

Article Details

BioPharm International 
Vol. 34, No. 3
March 2021
Pages: 34–37


When citing this article, please refer to it as J. Markarian, "Finding Success in AAV Manufacturing," BioPharm International 34 (3) 2021.