Starting with the End in Mind

Since 2017, the FDA has approved 22 cell and gene therapies, six of which are autologous CAR-T cells. With about 2,000 active CGT clinical trials, the sector is grappling with scale-out solutions as the patient access challenge continues to grow exponentially. I can safely say that the supply chain was not ready for the CGT revolution. The commercial introduction of CAR-T cell therapies led pharma companies to acquire much of the global supply of human AB serum, base medium and cell stacks, crucial raw materials, and tools for developing and scaling cell therapies, which had a knock-on effect on therapies in clinical development.

In the years since, supply of materials for GMP use has increased, but there are still supply chain headaches. For one, human AB serum safety regulations differ depending on geography, which has concentrated supply from US-based sources. Serum-free media are now commercially available and have the potential to decrease reliance on human AB serum, but none have been widely adopted (yet), in part because changing to media with defined, recombinant cytokines or alternatives is costly, and requires time-consuming comparability protocols. 

As more suppliers join the market, the supply chain can be expected to stabilize, but that still won’t solve the problem of scale. To date, only autologous CAR-T cell therapies have reached the market, but various new approaches are in development with the potential for scale-up, such as donor-derived allogeneic cell therapies or in vivo gene therapies. The biopharmaceutical industry is excellent at developing efficient ways to produce large batches (as seen in antibody manufacturing), but even if allogeneic approaches are successfully commercialized, it is unlikely they will ever completely replace the need for personalized autologous therapies.

In other words, the industry must still grapple with its original challenge. If you can’t make many doses per batch, you must make many small batches cost-effectively. Approved therapies, so far, are new ground for biomanufacturing – and the approach is something akin to a car factory; scaling out instead of up, by building large facilities that can manufacture hundreds of individual bespoke doses in parallel. 

As a result of these experiences, the industry now has manufacturing partners with the expertise needed to scale autologous cell therapies. 

Lessons learned
 

Developers that have brought CGT products to market have learned from their successes, as well as the obstacles they have been unable to surmount. Often, biotech companies take the sensible approach of looking at one inflection point at a time; complete one trial, release the next tranche of funding, repeat. This can blur long-term vision but the approach is often inevitable in biotech. Many biotechs start with passionate scientists, who do not necessarily have the required business perspective. Even for those companies with the right skill sets in place, venture capital funding tends to be stage-gated, which can disincentivize early investment in development for late-stage manufacturing processes.

In several cases in the young history of CGTs, this has meant the focus on reaching approval has come at the cost of sacrificing a sustainable commercial launch. Too many advanced therapies, especially in the EU, have been approved and then withdrawn from markets when they failed to secure payer commitments to make commercialization feasible.

Though this is disappointing from a patient aspect, it is to some degree understandable given where commercial manufacturing for biopharma differs from research and development. The industry is notoriously siloed at the early stages, but manufacturing requires a team approach that links process development experts with the inventors of cutting edge biotechnologies. Shared learnings – drawn from multiple disciplines – are necessary to make such complex products available at scale to meet the eligible patient population.

The CGT field is still immature and has a great deal to explore in terms of what it can realistically learn from mainstream biopharma experiences in scaling biologics. It can also be beneficial to look at completely different industries that have similar fundamental challenges to today’s CGT world.

Let’s return to the car factory model. Like autologous cell therapies, technology incorporated into modern vehicles is very different from previous generations, and every car manufactured today is slightly different. To efficiently meet the demands for bespoke, complex products, car manufacturers use Industry 4.0 hallmarks – AI-powered robotics, Internet of Things, and big data analytics. Advances like these have been slow to reach biopharmaceuticals – and especially CGT manufacturing, where some process development often still starts on paper.

The analogy has limitations, of course. Cars don’t require clean rooms; the raw materials aren’t as expensive; and, most importantly, a batch failure does not put a patient’s life at risk. However, integrating Industry 4.0 into CGT manufacturing will still have advantages when it comes to scaling out. Many developers have already seen the value of integrating automation into their platforms or processes – an excellent first step – but they have found it expensive, especially for more advanced programs and commercialized drugs. Incorporating automation at a late stage means a process redesign, and so it only makes sense when capacity is justified.

Some early-stage programs attempt to address the cost of implementing Industry 4.0 technologies by beginning with manual processes, with a vision of where automation should be integrated for commercial manufacture. This approach, however, is always going to be a challenge. Beyond the technical challenges of automating an established manual process, it requires quality management forethought and regulatory flexibility. It’s incumbent on developers to explain to regulators that a change will need to happen, and how to make it possible. Reaching alignment would mean understanding how to review a commercial process that may differ from an early-phase process, and being comfortable with the changes before early trials are even run. Because of the challenges, some companies turn to manufacturing partners, who can advise on coordinating with regulators and planning approvals with a forward-looking perspective that allows room for changes.  A good manufacturing partner will ask the right questions to draw a clear picture of the goal line.

If you’re going to do it in house, I can offer some tips. First of all, where possible, developers are best off exploring automation in the process development laboratory, and then integrating it hand-in-hand with manufacturing development throughout every step. It generally means moving away from the much simpler conveyor belt approach to a completely localized process, grouping all the units of operation in one place, with added reliance on single-use manufacturing techniques.

Second, keeping the transfer of liquids to an enclosed system can reduce the levels of error and improve reproducibility and sterility assurance. These systems can also be designed to manufacture multiple doses in parallel, making better use of limited cleanroom space. Since the closed environment is well controlled, it decreases the resource requirements of clean rooms as well.

Finally, beyond the automation of processes, consider a sophisticated IT infrastructure to automate the flow of data, which can help increase throughput. This will be necessary to remove much of the manual work from batch record assembly, review, and certification, and boost analytical capabilities to improve in-process and release testing.

Preparing for the future
 

Like many of us, I chose to work in this industry because of the incredible potential to impact patients, but there is little point in putting a tremendous amount of effort and resources into developing cures for cancer that nobody can afford. The key to developing the next generation of CGTs and making them accessible and cost effective is to start with the end in mind, and then work backwards.

The CGT manufacturing industry is on a journey. Some of the larger companies behind the first commercial CAR-T cell therapies had deep enough pockets to launch products that were not cost-effective, which have given the entire field the ability to develop capabilities and learn. Given the slow uptake of these first-generation cell therapies, they may not be able to recoup their development costs.

The next generation, led by smaller innovators, won’t have that flexibility, meaning there are risks for the whole field if we fail to capitalize on this moment. Since scale-up won’t be an option for autologous therapies, we need to focus on the right places to incorporate automation to make scale-out viable. Mapping the journey will take collaboration across multiple disciplines, requiring partnership between scientific, quality, regulatory, engineering, manufacturing, control, automation, and commercial teams.