top of page
  • Writer's pictureAntleron NV

Factories of the future - a far dream or close reality?

Based on the lecture of Jan Schrooten at the ATMP Supply Chain Conference, Feb 8th, 2024

It is no secret that manufacturing advanced therapies remains a bottleneck and heavily impacts the typical pharma business model. Economies of scale can go only so far with (predominantly) personalized products using complex starting materials. Armies of operators will not be an answer and will certainly not provide the scalability so direly needed by the cell and gene therapy (CGT) field. Even the FDA in a recent guidance suggested that all major manufacturing changes should be resolved prior to the clinical phase. This point is further driven home by Srivastava and Foster who argue that waiting until later clinical phases with developing the manufacturing process and understanding cost-of-goods (CoG) will put the commercial success at risk.

The AstraZeneca investment in manufacturing facilities and the shocking acquisition of Catalent by Novo Holdings earlier this year, sent ripples throughout the field, impacting the availability of manufacturing capacity for an already strained cell and gene therapy field. Indeed, the remarkable clinical results of this class of therapeutics, and ability to address debilitating diseases for which no other treatments exist, ensure continuously growing interest from industry and patients alike. For example, between 2019 and 2023, CGT grew from 1 to 10% of FDA approvals. The number of clinical trials has already further grown (e.g., 11% growth in Phase I products) since the beginning of 2024 and spending on CGT increased by almost 40% in 2023.

Unfortunately, 7 out of 25 CGT and tissue-engineered products authorized in Europe were withdrawn for commercial reasons, such as unscalable manufacturing, prohibitive cost and resulting difficulties with reimbursement, leading to low adoption. This only highlights the importance of considering the bioprocess and its economics as early as possible. Indeed, CoGs and scalability are key process parameters to bear in mind during the development of a bioprocess.

Considering the substantial CGT clinical pipeline and expected around 10 yearly FDA approvals, it is not hard to imagine that the value of such products, compared to their costs, will come (and already has been) under intense scrutiny. Health systems under pressure will drive down on reimbursement, meaning that economics of cell and gene therapies will become even more critical for their societal success and sustainability, favoring players able to find efficiencies and drive down costs, while solving medical needs.

All of the above indicate that a change in paradigm is needed. As a study by McKinsey & Company indicated, no one production system fits all assets, meaning that companies must recognize that standardization may not always be possible in the world of cell and gene therapy. Efficiently deploying custom bioprocesses is not evident in the field. It is not impossible, however, as will be shown further.

Firstly, let us consider where the field is now. It is noteworthy that production systems thought of as innovative today (e.g., Lonza’s Cocoon or Miltenyi’s Prodigy) have been around for two decades. The concept of modular, closed, and automated systems is not new, it just takes time to implement in the strictly regulated pharmaceutical (bio)manufacturing context. Additionally, some key enablers, such as digital twins and reliable sensors allowing for accurate bioprocess control are only starting to make their way into widespread use, accelerating the transition.

Simultaneously, new players are emerging, offering novel solutions and business models, ranging from Integrated Development and Manufacturing, supported by proprietary pods or containers (Cellares and Orgenesis) to highly intensified, closed and integrated up- and downstream production systems (Univercells Technologies). Many innovative auxiliary devices for different unit operations are also being launched by start-ups to help cell processing in more robust, scalable, and sustainable ways (e.g., Trince, Limula,). Following this philosophy and to respond to the growing demand for bespoke devices and bioprocessing technologies, we engaged with Comate and ICsense in the Biodevice-CRO partnership, thus providing biotech and life science actors with customizable, modular bioprocess solutions.

Regardless of the groundbreaking character of these production technologies, their implementation still needs to happen within a context of a broader strategy, especially if each therapeutic asset will require a semi-custom solution. The processes and materials in CGT remain variable and sensitive to changes in process parameters. Therefore, changing a bioprocess at a later clinical stage is considered almost impossible because of all the quality validations that need to be performed. With already lengthy development timelines, there is a tendency to continue with a bioprocess defined early on, while disregarding the potential detrimental impact on the economics of the final product. As highlighted at the beginning, this is a recipe for problems later.

Thus, a common misconception forms that innovation must be risky and very costly. Yet, with some of the cases that we have worked on at Antleron, we have shown that it works if done in a smart way. Enabled by combining digital solutions with custom and scalable bioreactors, we can limit costs while implementing closed, scalable systems, and still hitting bioprocess targets. In other words, fast-tracking development at an acceptable cost, while maintaining scalability and quality, is very much within the realm of possibilities.

The first key enabler is artificial intelligence, or more specifically machine learning algorithms. As we have recently demonstrated in the peer-reviewed paper Bayesian cell therapy process optimization, these powerful statistical tools cut experimentation time significantly (up to 70% cost savings) while doubling bioprocess efficiencies, easily outperforming traditional approaches. Joining Bayesian optimization are hybrid models, CFD simulations and economics modeling tools, completing the bioprocess digital twin.

Secondly, the intelligent, proprietary Antleron platform of scalable, smart, and customizable bioreactors, linked to the bioprocess digital twin allows us to generate real-life bioprocess data at small scale and directly make predictions for larger scales. The increased number of conditions and data supplemented with digitally conducted experiments enables greater bioprocess understanding, in turn allowing more robust optimization and control strategies, all at a fraction of the cost and time of typical bioprocess development runs. Additionally, an investment made into such a scalable platform needs to be only made once, while remaining applicable across a product portfolio. A flexible platform, scalable and easily customizable can be a powerful motor for innovative, sustainable bioprocesses.

Despite all the progress and modern technologies entering the market, the main question remains – will CGT manufacturing mature to a true industry or remain a craft? In Belgium we have many pieces of the puzzle to enable that transition: world class academic institutions, favorable clinical trial climate, a progressive regulator, and a mature biotech industry. Simultaneously, in our ecosystem we find examples of more traditional bioprocessing approaches (Legend Biotech) and companies pursuing a decentralized approach driven by platforms such as the Cocoon (Galapagos). Even more biopharma companies are entering the field (UCB), standing in front of strategic choices on how to approach their CGT manufacturing. Inevitably, this mix of expertise and approaches will result in innovation diffusing into daily practices, particularly if supported by robust data, and clinical and market results.

Bringing all these competences together in end-to-end cases would allow to usher in a next generation of bioprocesses, and therapies, which are scalable and sustainable. Some initiatives, such as the GEne Therapy INnovation Training Network (GET-IN), are currently underway. GET-IN unites academic and industrial partners across Europe to collaboratively address prevalent knowledge and expertise gaps in the gene therapy value chain. Particularly reaching higher vector titers, minimizing empty capsids, and keeping costs under control are important goals. The critical mass of scientific and industrial expertise within GET-IN should enable solving these pain points and subsequently the sustainable translation of gene therapy innovations along the value chain.

To conclude, while we may have factories of the future, or we are quite close to having them, we still need next-generation bioprocesses to run in them. This, however, requires novel bioprocess development strategies and a certain openness to risk. Distributing the risk through co-development partnerships that bring all necessary expertise and technology to the table should provide an incentive to take the leap. Working in a smarter way will still be a smaller risk than sticking to the old ways. Only by fully embracing the future can we make cell and gene therapies a scalable and sustainable success story for the sake of patients.


bottom of page