In the fall of 2010, I began my doctoral studies at the University of California, San Diego (UCSD). I quickly began to use a model that was novel to me and the world: human pluripotent stem cells (hPSCs) derived cardiomyocytes. When I first thawed them, I was surprised by the wide range of cell distribution, viability, and propagation rates in several hPSC vials. I experimented with various methods to control cell plating and cycling, which are critical to the differentiation protocol. Nonetheless, whenever I opened the incubator the next day, I hoped for identical replicates. Every day, I found myself on the edge between losing my mind and finding God.
There’s no way my method would be able to consistently provide the quantity and quality of cells needed to do my experiments, let alone support clinical trials or a globally effective personalized medicine. There is a significant difference between developing a concept in a lab using know-how and lab notebooks and producing it on a tightly regulated, highly standardized commercial scale. Making the most recent iteration of genome-edited autologous cell therapy in a dish may have the potential to treat autoimmune diseases, but it’s useless if it can’t make it over regulatory hurdles and exceed financial expectations.
New biologic, same problems
The challenges for antibodies—high cost of goods, difficult to manufacture, analytical assays that aren’t that informative, concerns about high costs, and a couple of other attributes—are the same as what the field of cell and gene therapy faces today. Jason Bock, PhD, CEO and founder of the Cell Therapy Manufacturing Center (CTMC), said, “Looking back in history, the yields were terrible, and we were analyzing antibodies on SDS-PAGE coomassie stained gels, saying, ‘Look! It’s a single band! I think? I don’t see any other bands there; it must be pure. The first antibodies weren’t great therapeutics—they were chimeric with pieces of mouse sequence and were fairly immunogenic. But they worked okay and met some unmet medical needs, so there was a niche market but not a huge spotlight on the infrastructure to produce these at low cost because there wasn’t much demand for them.”
Flash forward 20 years, and antibodies are doing pretty well in cancer and even common diseases like asthma and chronic migraine because manufacturing costs have been driven down, making antibodies affordable and accessible. Bock thinks the same thing will happen with cell and gene therapy and, in some ways, it could be even faster because, unlike in the days of antibodies, people weren’t thinking about infrastructure and manufacturing early on.
“The science has been evolving for 30 years, but no one had been working on the infrastructure—everything was done in academic labs,” said Bock. “We have to act with urgency and patience because this infrastructure will come, and we will figure all of these things out. We will drive the costs lower and figure out different approaches to manufacturing these. The opportunity we have is that we can find solutions to accelerate this transformation. But if we don’t do anything, it could take 30 years. If you look back at antibodies, all those 17 innovations happened in silos, and slowly people connected the dots and put things together. But it took a long, long time.”
Patient-to-patient, paycheck-to-paycheck
Two weeks before the 2024 Cell and Gene Meeting on the Mesa, Lonza announced it had won the bid to manufacture Vertex’s Casgevy, the cell and gene therapy field’s crown jewel. According to Thomas Fellner, PhD, global head of operations for the Cell and Gene Technologies Business Unit at Lonza Pharma & Biotech, Lonza’s expertise with cell products gives them an edge as one of the first CDMOs to provide cell and gene therapy services.
But even for Lonza, which has an established global footprint as a CDMO, the cost of cell and gene therapy is something that they will need to de-escalate quickly. This becomes particularly difficult when considering bespoke treatments such as autologous cell therapies like Casgevy. “I’ve got to get the drug product to the patient before getting paid,” said Fellner. “For autologous therapies, you have slots like a hotel room and must ensure a patient fills the hotel room. If a patient drops out, the hotel room stays empty and the cost needs to be absorbed—the more of these missed slots you have or the more failures you have, the higher your manufacturing cost becomes for your drug. In the end, this does not leave anything left from a profitability perspective.”
Fellner’s colleague Senthil Ramaswamy, PhD, VP of R&D and Innovation for the Cell & Gene Division at Lonza, thinks that the solution to making cell and gene therapy cost-effective and profitable lies in automation, digitization, and tight management of clinics. “You have to be able to make the process more autonomous, making sure that everything is digitally available, like batch records, and that your QC and process data are connected so that the release process is seamless,” said Ramaswamy. “Can you get to a point where you’ve got enough data and are confident that if it meets certain parameters and release criteria, you don’t need to review the batch?”
Ramaswamy also thinks that innovation will bring down the cost of materials, which will be essential to making cell therapies like CAR T and Casgevy available earlier in the treatment process and expanding them to indications like Alzheimer’s disease, diabetes, and autoimmune diseases.
Shoot for the moon
A sobering fact about why the cell therapy commercial manufacturing world is so profitable is that thousands of patients are left out to dry on waitlists for transplants or even CAR Ts each year. “Patients are dying on the waitlist, even though there’s FDA-approved cell therapies on the market that would most likely cure them—that’s tragic,” said Fabian Gerlinghaus, who co-founded Cellares to address the cell therapy manufacturing shortage. “8,300 patients were treated with FDA-approved CAR T cell therapies. We’re building out capacity for some of between 150,000 and 300,000. That’s really what we need in this industry because a single autoimmune disease drug can have a patient population with millions of patients.”
At Cellares, Gerlinghaus has been building a “smart factory” model called an Integrated Development and Manufacturing Organization (IDMO) that automates end-to-end cell therapy manufacturing and GMP QC with two core technologies, the Cell Shuttle and Cell Q. The entire process for a customer looking to automate manufacturing of cell therapy has three key stages: (1) tech transfer and automation; (2) CGMP bridging work, including complete analytical transfer and small comparability studies; and (3) cGMP manufacturing in support of either clinical or commercial scale.
While Cellares and Gerlinghaus have dazzled investors and announced several partnerships with big pharma, whether Cellares will quickly jump from a futuristic concept to an actual cell therapy-producing staple remains to be seen. The company has yet to treat someone with a Cellares product, which Gerlinghaus said he anticipates will happen in mid-2025. Cellares almost seems too big and too promising to fail. Still, the company and Gerlinghaus may have an Achilles heel by focusing entirely on the manufacturing process and not applying their same approach to the process for orders being placed by providers.
“It’s just a matter of the pharma company coordinating between the treatment side and us and letting us know how many patient doses and starting materials from different patients we’re going to receive on this day and on that day,” said Gerlinghaus. “Depending on the pharma company, that can be more or less automated. In some cases, spreadsheets are involved; in other cases, there are automated software systems.” It is too soon to say whether this seemingly overlooked coordination process will be a growing pain for Cellares to solve or if the initial crack will create propagating fissures throughout the company.
The burning tower
Jason Jones has spent his career developing processes in boxes at Miltenyi Biotec, where he developed the CliniMACS Prodigy for magnetic cell separation. He did not want to make another “box” when he joined Cellular Origins as global business development lead.
“We’ve got to keep things flexible so you can change the process for the betterment of the process of the product if you need to, but you need to be able to go all the way along that clinical development line to get to a commercial product where certain things are going to have to get fixed at some point,” said Jones.
Edwin Stone, CEO of Cellular Origins, thinks the technology already exists to solve the scale problem. “The core thesis behind that is that we have all of these awesome unit operations of which, at the end of it, a wonderful biology emerges that saves lives,” said Stone. “We need to take away the humans that are the bottleneck at the moment to be able to scale those processes and do that with an automation bank. That in itself is not unique to the industrialization of any processes. Historically, that’s been through conveyor belts and more recently through robotics, which we use.”
Cellular Origins’ modular approach to cell and gene therapy manufacturing puts it between BioCentriq, which starts from an empty room and builds around consumer needs, and Cellares, which fits consumer needs inside its box.
Cellular Origins’ team began by considering what tools they needed to cover 80% of the processes with enough confidence to last five to ten years. The answer, Stone said, was that there wasn’t the perfect combination of tools to do so. “That led us to something that can be flexible…and agnostic to ensure that the best biology can win,” said Stone. “It’s critical people can choose the unit operation and that they’re not forced onto a bioreactor that maybe isn’t quite perfect for their process.”
Ultimately, Cellular Origins must find a solution that benefits clients and patients. “I hate the idea that this amazing therapy has been created that can treat hundreds of thousands of people, but because the particular tech that is made on, the particular bioreactor or method of cell modification is not part of a scalable system, does not see the light of day. The scientists working on cell and gene therapies want routes to scale that do not hamper innovation.”
Playing “small-ball”
When he joined BioCentriq two years ago, David Smith, PhD, VP of development, advocated for narrowing the global CDMO’s offerings to cell therapies. Not just any cell therapy, either. “We would be stupid to abandon the CAR T world because it accounts for 90% of the cell therapy market,” said Smith.
However, Smith didn’t point BioCentriq in the direction of manufacturing for CAR T commercialization. Instead, Smith decided there was much opportunity in the stages before commercialization. “How do we engage with the early preclinical phase where we feel that there’s a big gap right now that a lot of the traditional CDMOs move towards the commercial side of the business?” said Smith. “The business model is easier to understand, and the profit margins can be greater at that end.”
Smith thinks the preclinical and clinical stage cell therapy manufacturing industry is lucrative because most drug developers do everything in-house. Smith stated, “80–90% of the market requires phase I trials and must be successful. If you choose a large CDMO, you will not receive the necessary handholding and speed for Phase I.”
To ensure that Phase I trials do not fail due to manufacturing issues that do not translate from the lab into a clinical-grade product, Smith stated that the key is to de-skill—replacing knowledgeable, hands-on scientists with technology and workers to reduce costs and streamline manufacturing. “What we are doing is staffing the clean room with knowledgeable personnel because, for Phase I, it’s not about de-skilling,” said Smith. “Maybe when you go commercial clean room, de-skilling the operators and simplifying it. But for the first-in-human trial, you need highly skilled personnel to make decisions in the clean room while being flexible and fast.”
The new kid on the block
Five years ago, Jason Bock moved from running a 600-person site developing monoclonal antibodies for a pharma company to starting a new group at MD Anderson to help innovative researchers at one of the leading clinical trial hospitals to industrialize their cell therapy ideas. The problem he had to solve was how to cut out everything a researcher would have to do to launch a company to get a cell therapy to market, including raising millions of dollars and running an entire business.
As employee number one looking to build infrastructure manufacturing, Bock identified a local cell therapy development and manufacturing company, Bellicum Pharmaceuticals, with a huge facility, which MD Anderson acquired. In the process of bringing the facility up to speed to help MD Anderson PIs develop their products, Bock said that he was approached by some VCs sitting on cell therapy company boards who were struggling to get from research into the clinic, particularly for CAR T. Repeatedly, he heard the same story from VCs coming from board meetings touting great mouse data and wanting to move into clinical phases but didn’t have the resources, people, or know-how.
These interactions led Bock to develop this new division at MD Anderson into an entity that enables smaller organizations to robustly reach clinical proof of concept and be flexibly positioned for commercialization from process development to integrated regulatory support. In that time, Bock’s team has brought numerous cell therapies to IND and clinical trials, such as Obsidian Therapeutics’ novel TIL therapy, OBX-115. Today, CTMC is a joint venture between National Resilience and the MD Anderson Cancer Center, incorporated in May 2022 as a spin-off from the Therapeutics Discovery and Development Division of MD Anderson.
Bock said, “It’s been tremendously gratifying to build something pretty differentiated at MD Anderson that I think will help provide a different way of developing cell therapies that’s much more speedy and capital efficient so that not every company doesn’t need to raise $100 million to get one product with 15 patients in one building. That’s good for the whole field.”
Bit by bit
My time spent researching and working in a lab came to a close in 2017, around the same time academic neurosurgeon Mark Kotter, MD, PhD, of the University of Cambridge, was getting ready to launch bit.bio. Kotter’s company aims to accomplish what I failed miserably at—the industrial-scale, rapid, and consistent conversion of induced pluripotent stem cells (iPSCs) into specific human cell types. To achieve his objective of producing as many distinct cell types as possible—rather than just a handful—Kotter must first determine the code that produces each type of cell and then figure out how to produce reproducible batches of billions of cells.
Finding the precise transcription factor combinations required to produce a particular cell type has proven to be one of the most difficult tasks. “There’s about 2,000 transcription factors, and we’ve found that it takes anywhere typically from one to six transcription factors to define a specific cell type,” said Kathryn Corzo, president and chief operating officer of bit.bio. “If you do the math, it’s significantly complex beyond the scope of humanity. Our approach is with high throughput experimentation and bioinformatic strength.”
Earlier this year, bit.bio announced that their efforts would be concentrated on cells for immunology, neurology, and endocrinology. Yet even by whittling down the candidates down to a handful of cell types, Corzo said the challenge for bit.bio is finding the capital to take all of their iPSC-derived cells all the way to the clinic. “We are working on creating pre-clinical data sets, which we call an IND Enabling Package, or CTA Enabling Package, that would allow a partner to take over and develop that cell into a therapeutic solution,” said Corzo. “We could also arrange the agreement to leverage our proprietary opti-ox technology and manufacture the cells at scale for further clinical development with their own GMP master cell bank.”
According to Corzo, bit.bio’s therapeutic iPSC-derived cell making the most progress is a human hepatocyte. “We’ve been able to scale [the hepatocytes] in 3D, which means we’re able to manufacture billions of cells required to treat patients,” said Corzo. “I think the most important thing is that TxHep not only has the morphological features of a hepatocyte but also exhibits functionality comparable to that of a human hepatocyte.” Corzo said that the lead therapeutic program, bbHEP01, is focused on delivering hepatocytes, transiently to treat acute and chronic liver failure for patients waiting for transplantation.
It will take much work and exploration before bit.bio can get the FDA to approve even one of their iPSC-derived cell types. But it is no longer a matter of if and how, but when.
Read “The End of the Beginning for Cell and Gene Therapy, Part 1” here.