[Music] My name is Andrew Fez Mack. I'm an assistant professor of clinical pathology here at the University of Pennsylvania, and I also serve as the director of manufacturing in our cell and vaccine production facility here on campus. I have nothing relevant to disclose today. I wanted to provide a broad overview of CAR-T cells, and our learning objectives are to gain a basic understanding of our T cell biology and design, talk about current available targets and common approaches to manufacturing these cells, and finally, I'll touch on future directions for the field. Let's begin in the beginning. What are CAR-T cells? So, CAR is an acronym that refers to chimeric antigen receptors. These are artificial molecules that combine B cell-derived and T cell-derived domains and function. To understand why you might want a receptor like this, we have to go back to the basic biology of B and T cell activation. But you'll recall that a B-cell receptor derives its specificity from its variable domains shown here in blue at the end of the B-cell receptor. So, when the B-cell receptor encounters

an antigen to which it has sufficient affinity, it combined directly to this antigen and then the B-cell to have a number of downstream functions. That can produce cytokine that befell clone can proliferate and ultimately that B-cell can differentiate into a plasma cell which can make soluble antibody. The antibodies can then double back and through optimization neutralization or even fixation of compliment cause cytotoxicity against the target positive agent. Comparative to T-cell activation here we have a T-cell receptor again with variable domains that confer specificity. However, this T-cell receptor when it binds the cognate antigen testifying to that antigen in the presence and the context of MHC. So if the T cell T cell receptor is properly engaged and the T cell can have downstream effects much like the B cell it can collaborate with cytokines. The antibodies can then double back and through optimization neutralization or even fixation of compliment cause cytotoxicity against the target positive agent. So relying on MHC to present a piece of cognate antigen actually has

its downsides any virally infected and tumor cells downregulate MHC, rendering these cells resistant to T-cell clearance. So if we were to design an artificial redirected killer cell from scratch, we want to combine the ability of a cell to bind to a target totally independent of MHC, but at the same time cause direct cytotoxicity T-cell. And that's exactly what Akamai chimeric antigen receptor does. Here in this figure, you can see chimeric antigen receptors as they evolved over the years. All the way on the left side of the figure, you can see that there is a B-cell-derived extracellular portion of a chimeric antigen receptor. Here is depicted in blue, and they call this a single-chain variable fragment. This extracellular portion was actually derived from the heavy and light chain variable domains of a B-cell receptor in a single molecule fused to the intracellular T-cell signaling domains CD3 data. So what was originally found was the T cells expressing this chimeric antigen receptor showed an initial burst

of response and then slowly lost efficacy. Of course, those cells were becoming an urgent later, adding co-stimulatory domains like CD28 and for one BB overcame this entered energy to allow themselves to become durably active. These are what would allow the cells to be put into clinics so quickly. You can see that by basing your extracellular portion on B cell-derived highly specific receptors, you can achieve high specificity and MHC independence. And it was the addition of co-stimulatory domains insists that allow for durable activation. Knowing that the platform works, what are we able to target? You can see here that there are a number of different targets that are under investigation: target spent mark multiple myeloma, two myeloid leukemia, as well as other B-cell lymphoma. But it's not just hematologic malignancies. Party cells are now moving into target different fella tumors as well. So you can see but there's a wide variety of targets that are now being investigated and what can we say that's

common amongst these so they're commonly extracellular because that's as you recall from the design of the car. That single-chain variable fragment has to bind to target, and that target is an extracellular domain on your target cell. In addition, especially with respect to solid tumor targets, it becomes clear that there has to be a balance struck between sensitivity and specificity. So many highly specific solid tumor markers are not conserved either between different cases of a given tumor type or even within a heterogeneous single tumor itself. But choosing the right target for a specific solid tumor can be quite challenging, striking that balance. The proper balance might actually be tumor-specific. Let's shift gears a little bit and talk about how we actually make these cells. Each protocol differs slightly, so there are common elements to all manufacturing protocols for CAR T cells. By and large, it starts with large volume apheresis from a non-mobilised autologous donor. Might actually be windfall Teaneck at the time of collection or even at circulating tumor, so the product itself might have.

few T cells or lots of non-T cell contaminants and therefore is likely to undergo T cell enrichment. ESL enrichment might take the form of a density-based separation like by call or elutriation or even an antibody-magnetic bead-based separation. Once you have your enriched T-cell product, that can be activated and modified. So, the modification often occurs via viral transduction or transfection of DNA and RNA, and the activation can occur via beads or artificial antigen-presenting cells that provide an activating signal to these T cells to proliferate. The proliferation occurs either in a standard culture vessel or in a bioreactor to reach a clinically meaningful cell number for infusion. Of course, these products are then heavily tested for quality control. We're testing for things like making sure that they've retained their T-cell phenotype, that they are appropriately modified, and that the product isn't contaminated with bacteria or fungus. Then the product is then reused, completing the autologous cycle. But the major advances have been

made in car T cell manufacturing technologies. The original manufacturing protocols were almost entirely based on adaptation of legacy instruments. These are instruments that weren't designed specifically for cell therapy purposes or certainly not for gene-modified T cells. Growing, and they were modified or adapted to suit the purposes of their original protocols. If in a quantum leap forward where the Weisse manufacturers are now recognizing that there's a significant and growing demand for cell therapy-specific equipment and designing this equipment for these purposes. But even with these new technologies, we still have significant challenges achieving a meaningful dose, and really the biggest challenge is derived from the patient. So we have the autologous Carty cells that we're manufacturing, and our starting material comes from patients with a variety of underlying diseases in different states of disease at the time of collection and having been treated with often years of different modalities. What this means is our starting material demonstrates significant variability, and that leads to some significant run-run variability. This is unique to engineered cell therapy manufacturing, the amount of

starting material variability that we can tolerate. It's really unheard of in most other bio manufacturing processes. In addition, I showed some technological improvements that we've seen over the years, and the field continues to develop very rapidly. However, obviously, these new technologies come to come in significant cost. So finally, I wanted to touch on where the field is going and what the next generation of Carty cells might actually look like. And looking across all of these different types of T cells, you'll recognize that the next generation of Carty cells is going to be car plus. So I say that to mean that there will be best cell will be engineered to express a car, but either a heavily modified car platform or a car platform plus some other type of modification. And these additional modifications are going to lead to some significant increase in car capabilities, but specifically I believe they'll increased efficacy, increase safety, and maybe even allow us to start with non-autologous material, particularly when we encounter solid tumors. Car T-cells encounter a number of limitations, specifically

accessing or activating in the presence of children. So-called T-cells redirected for universal cytokine-mediated killing, or trucks, are able to produce a large amount of cytokines such as IL-12 within target positive tumor tissue. And this, in turn, recruits innate immune cells which can destroy local tumors. From other CAR T-cells have been engineered to express chemokine receptors to enhance their homing into the tumor. Finally, checkpoint receptors have actually been ablated from CAR T-cells, rendering them resistant to tumor-mediated checkpoint inhibition. Other CAR T-cells have been reimagined to enhance safety through Boolean gating. Tandem cars and dual cars require the presence of two targets on the tumor cell to fully activate. Pressive cars, on the other hand, have a traditional car that generates a full activation signal but also are transduced with a single chain that's used to an inhibitory signal. But you can imagine that if normal tissue is expressing a benign marker, this car would not kill that tissue. Together, the goal of this advanced approach is

to enhance specificity and limit off-tumor effects, ultimately making a safer product, CAR T cells have been called living drugs. They pose particular challenges when you try to control their in vivo behavior, so some next-gen CAR T cells will most likely include switches to enhance their safety. These might include a unique marker to which a monoclonal antibody could bind and clear the cells, or expression of a prodrug suicide gene. Still, others are developing conditionally activated CARs, a very exciting development where full activation of the CAR through signal 1 and signal 2 only occurs when an adapter molecule is provided. This could alleviate symptoms of overactive CAR T cells in vivo. Finally, as you recall from earlier, many of our challenges stem from the fact that these are autologous products collected from patients who have been treated for decades with all kinds of cytotoxic chemotherapy. So, one attractive modification would be to work with allogeneic T cells. There are two major risks working with allogeneic

and car T cells. First, is that the endogenous TCR and an allogeneic T cell population might have affinity for the host and therefore transducing the cell population and infusing it into a host could cause severe graft versus host disease. Alternatively, highly polymorphic HLA molecules on the surface of these car T-cells might be recognized by the host immune system and therefore rejected. But if you wanted to use our genetic material, you may have to ablate or delete the endogenous T-cell receptor and or HLA molecules. This has been called the universal Carty approach. There's a great deal of activity in this area. This is actually first tried in humans and Great Ormond Street Hospital and now the efforts are largely being led by selectively so you can see that they've targeted both cd19 as well as CD 123 and they have other Universal car T-cells in the pipeline. This is a very exciting development given that it would change drastically how we approach manufacturing. So in summary, the next

generation of car t-cells is likely to combine the Carty platform to make more effective, safer, and maybe even allogeneic car t-cells. But here I summarized our objectives and what we talked about today, and I'll just close by saying that this is a very exciting technology, but it's still rapidly evolving. So from day to day, we're learning new information about how to optimally do this manufacturing. At this point, it is still a technical art, and it really requires significant flexibility in how you operate your facility, especially given the run-to-run variability in the autologous product. Finally, the platform has proven really effective in some settings, but again, particularly when targeting solid tumors. I believe that it will be the next generation of party cells combining both the proven platform of Carty with gene editing or third or fourth generation cars to really enhance efficacy in solid tumors. But finally, I'd like to thank all my colleagues at the clinical cell on vaccine production facility, as well as my colleagues in the division of

An Introduction to CAR T-Cell Therapy

An Introduction to CAR T-Cell Therapy