Karen Titus
December 2024—There is no disguising the enthusiasm—not to mention expertise—that David Miklos, MD, PhD, brings to the topic of chimeric antigen receptor T-cell therapies. CAR T-cell products have been strikingly effective in treating large B-cell lymphoma and multiple myeloma, as well as other hematologic malignancies such as acute lymphoblastic leukemia, follicular lymphoma, and mantle cell lymphoma.
Sounding equal parts life coach and physician, Dr. Miklos, professor of medicine and chief of the blood and bone marrow transplantation and cell therapy division at Stanford University, sees nearly limitless potential in this immunotherapy, which entails genetically engineering T cells to express engineered molecules that can recognize and bind to specific antigens on cancer cells.
“So much is possible,” he says. “Our little minds are limiting this, not the potential of science. There are a lot of obvious and mind-blowingly effective strategies that are just coming home to roost. We have really blown open the opportunity to advance therapies more quickly, to apply multipronged approaches using logic loops that are essentially programmed into a cell. It’s really exciting.”
He pauses, then asks, “Can you sense my enthusiasm?”
Here is an unusually upbeat tale in oncology—call it “Of Men, not Mice.” As Dr. Miklos explains, clinical translational research with cell therapies is almost immediately applicable. Compared with drug development, “It doesn’t take much time to move from a preclinical mouse to the GMP process development and into a patient. We can do this in one or two years.” Currently there are six FDA-approved therapies for lymphoma and myeloma. Researchers are also exploring the use of chimeric antigen receptors, or CARs, to treat a variety of solid tumors. The products may also be a potent therapy in autoimmune diseases.
Dr. Miklos is equally energized when he talks about what these therapies will mean for clinical laboratories.
“My pathologist colleagues at Stanford are a crucial part of the clinical care team as well as the research team,” he says. Such support needs to become routine as CARs become more accessible outside of large academic centers. Having accurate, quantitative assessment of targets for new T-cell therapies and even monoclonal antibodies will become a critical part of practice within the next 10 years.
This would be a marked difference from current practice, Dr. Miklos says. “Your average pathology report will simply say, ‘B-cell lymphoma, restricted by kappa or lambda,’ and maybe include comments about CD19 and CD20 expression.”
As he and colleagues showed in a Nature Medicine article two years ago (Good Z, et al. Nat Med. 2022;28[9]:1860–1871), that’s insufficient. Says Dr. Miklos: “The problem is, when you use a CAR T cell in large cell lymphoma targeting CD19, 40 to 60 percent of the patients who are progressing are going to lose the antigen CD19 expression.”
At Stanford, Dr. Miklos and colleagues have moved on to flow cytometry incorporating “a bead that contains known fluorochrome amounts to create a standard curve in the daily operations of mean fluorescence intensity measurements,” he says. The system is multiplexed—by using five different fluorochromes on that bead, “we can do quantification of CD19, CD20, CD22, and others” on FNA lymph node biopsies, Dr. Miklos says.
“This works for T-cell engagers, CAR T cells, and monoclonal antibodies,” he says. “And we want all clinical pathology labs to provide this antigen quantification. I don’t think it’s good enough anymore to say ‘dim’ or ‘high expression.’ We need to be quantitative.” Efforts to quantify tumor antigens by immunohistochemistry can also predict big differences in CAR19 efficacy outcomes, based on the quantification of the CD19 antigen determined by IHC prior to the trial (Locke FL, et al. Nat Med. 2024;30[2]:507–518).
The treatments are not without their risks.
One has come to the fore only recently: the potential for a second malignancy. The FDA announced an alert in November 2023 and ultimately labeled CAR product with a black box warning for the potential increased risk of T-cell cancers, says Ash Alizadeh, MD, PhD, the Moghadam Family professor of medicine, divisions of oncology and hematology, Department of Medicine, Stanford Cancer Institute/Institute for Stem Cell Biology and Regenerative Medicine.
The move was triggered in part by an Australian group that reported on two patients with second tumors who had received a special vector called the piggyBac transposon. Subsequent reports to the FDA were linked to other vectors, but the numbers were small, Dr. Alizadeh says. Nor was it clear whether these T-cell lymphomas were incidental or caused by the CAR T-cell therapy itself.
The FDA warning launched a flurry of activity and led to multiple reports, including one from Stanford (Hamilton MP, et al. N Engl J Med. 2024;390[22]:2047–2060).
Several months after these reports were published and compared, “Something seems to be evident, which is that there are two apparent patterns of T-cell lymphomas after T-cell therapies,” Dr. Alizadeh says.
One appears to be caused by the vector integrating into the genome and “creating trouble,” as Dr. Alizadeh puts it. But that seems to be the exception rather than the rule, though this is admittedly based on small numbers.
“Most of the lymphomas, like the one we described in our New England Journal of Medicine article, don’t seem to have the CAR product integrated into the T-cell lymphoma,” he says. Instead, it became manifest after the CAR therapy. It’s possible that this second malignancy is linked to the inflammation associated with CAR-based immunotherapy, or it could be due to some sort of clonal baseline premalignant or early malignant clones that were already present, though this remains to be determined.
In the latter group, he continues, it’s worth noting that the incident T-cell lymphomas often seem enriched for specific mutations typically associated with myeloid clonal hematopoiesis, including in TET2 and DNMT3A. That has led some to ask about the usefulness of screening patients before they receive CAR T cells, and whether they should be excluded from therapy or monitored more closely.
“This is a tough proposition,” Dr. Alizadeh says, given the pervasiveness of clonal hematopoiesis. “Using that as a litmus test for patient selection seems to be too heavy-handed, in my opinion and in the opinion of a lot of folks who’ve weighed in on this topic.”
The concern is legitimate, though it is a rare phenomenon, Dr. Alizadeh says.
“There are other people who say this is all way overblown,” he adds. Here’s how that math looks: “Count the number of CAR T-cell therapies that have been administered”—estimated by some to be 30,000 to 40,000 patients, he says—“and the number of T cells that are infused as a result—a few million/kilogram—and then the number of vector insertions per genome,” Dr. Alizadeh says. Assuming random integration, and effectively a saturation mutagenesis of the T-cell genome, “The number of these T-cell cancers seems so rare that it almost proves that T cells are almost immune to the effect of CAR T-cell-driven malignancy—that their genomes are kind of bulletproof.
“I’m not sure I entirely buy that argument,” he continues. “But it’s been made by some luminaries, including [Dr.] Carl June, who’s the father of the field.”
Nevertheless, as those in the field continue to scrutinize potential risks, the testing landscape could also evolve.
The Stanford team identified a lethal, Epstein-Barr virus-positive T-cell lymphoma in one patient (724 patients were included in the study) who had received CAR19 therapy with axicabtagene ciloleucel for stage IV EBV+ diffuse large B-cell lymphoma. Comprehensive profiling was done on both tumors, including flow cytometric immunophenotyping, targeted and whole-exome sequencing, single-cell RNA and single-cell DNA sequencing, and cell-free technologies to characterize the contribution of clonal hematopoietic mutations.
As their article demonstrates, “It’s complicated,” says Dr. Miklos. But there was no axi-cel expressed in the T-cell lymphoma itself.
Clonal hematopoietic mutations, known as CHIP, occur in everyone in an age-dependent manner and increase in frequency in patients receiving chemotherapy. “But now imagine that your hematopoietic stem cell has certain mutations passed onto lymphocytes, not just the myeloid cells,” Dr. Miklos says. While normally there is little proliferative demand on lymphocytes, “nothing drives T-cell proliferation the way that a CAR” does, with its signal antigen causing T-cell duplication every seven to 12 hours, essentially, until exhaustion. “We can drive the T cell that expresses the CAR to represent upward of 50 percent of circulating lymphocytes in the first seven days of therapy.”
This drive of antigen in a clonal-restricted cell population could have a detrimental effect if there are underlying clonal hematopoietic mutations. Dr. Miklos says his colleague (and coauthor of the New England Journal article) Mark Hamilton, MD, PhD, was scheduled to give a talk at the ASH meeting in December on the role of CHIP in causing second malignancies after CAR T cells are administered.
“If there’s a proliferative advantage offered by mutations in these cells that are either DNA stabilizing or DNA repair, or just absolutely necessary for clonal hematopoiesis, we can actually see that effect by monitoring the allele frequency of the mutations before CAR T, during, and after,” Dr. Miklos says. “And we can follow those as they track into the myeloproliferative malignancies themselves using single-cell DNA antibody paired testing.”
It’s unclear how much of this testing will be needed in clinical settings. But exploring the mechanisms of second malignancies will propel the field forward, and it could expand the role laboratories play in assessing the impact of these therapies.
Dr. Miklos lays out what he says will be a likely scenario for pathologists:
The first step is to understand the target antigen and be able to quantify it in ways that are clinically relevant for these so-called engaging therapies, be it CAR or T-cell engagers, he says.
Next, understand that patients will develop malignancies. “They’re going to have bone marrow biopsies because of cytopenias after CAR T, and you’re going to be left with this question of, is it MDS, AML, or is it inflammatory cytopenia that itself is not a cancer and will burn out as the CAR T cell goes away in six to 12 months?” He suspects many patients are being evaluated for MDS because of the cytopenia. “These inflammatory cascades do make the bone marrow cells look funny,” he acknowledges, but most will not turn into MDS; most patients will improve within one to two years.
“Cytopenia after CAR T was not an anticipated toxicity,” he says. “But it is something that pathologists are working on every day.”
Unfortunately, some patients do develop MDS and AML in the age-dependent and chemotherapy-related population. He and his colleagues are using Heme-STAMP (Stanford Actionable Mutation Panel for Hematopoietic and Lymphoid Malignancies) next-generation sequencing to identify contributing CHIP mutations.
CAR therapies will, in short, deliver plenty of questions to pathologists’ doorstep: Is this cancer? Is it a myeloproliferative disorder? Is it a bone marrow failure? Is it a temporary problem that will resolve with time?
Dr. Miklos recognizes that this is a very tall order. “It’s not fair to put the pathologist in that spot, but that’s where we are right now.”
In the meantime, the bulk of patient testing will be up front, prior to administering CAR T. “We want to use the right therapy for the right patient,” Dr. Miklos says. And on the horizon lie multitargeted CARs. “It’s clearly coming,” he says, noting that within the next two years the FDA will likely approve agents that target both CD19 and CD20, for example.
The six currently approved CAR T-cell products, all autologous, are arguably first-generation products, Dr. Alizadeh says.
“But many, many more are in various stages of development, for both liquid and solid tumors. There’s a lot of fancy engineering and more complex versions of these therapies that are emerging.”
That includes developing allogeneic products. Dr. Alizadeh and his Stanford colleagues are using these off-the-shelf products in clinical trials. If they work, the gains would be tremendous, he says.
Outside of academic institutions, only a minority of patients with aggressive lymphoma have their disease treated with CAR T cells, despite their eligibility for the therapy, due in no small part to logistical hurdles. The oft-quoted figure is as low as one to two out of 10, says Dr. Alizadeh, though he suspects it’s slightly higher. “So the off-the-shelf strategy would definitely bring more access to a cure.”
Allogeneic products could be tremendously helpful, Dr. Miklos says. “You could have cells from one individual, or maybe someday a pluripotent stem cell to make a single-batch product used by a hundred or a thousand patients.” Such a product would be akin to a drug, he suggests, whereas an autologous CAR T is more like a biological therapy, an individual precision medicine made repeatedly for one patient.
“Ongoing research suggests immunological resistance of the host eliminates this third-party CAR T cell despite many different stealth mechanisms being applied. So having persistent CAR T cells that result in long-term, durable remissions has been the challenge.” That researchers have been working on such allogeneic CAR T products for some eight years, with no FDA-approved product to show for their effort, “tells you the magnitude of this challenge,”
he says.
Where allogeneic CAR T cells might more easily shine would be in treating autoimmune diseases. It might be possible to dose patients with CAR T cells—perhaps every six to 12 months, Dr. Miklos hazards—and deliver therapeutic benefits without running the risk of developing resistance that a patient with cancer might. Nor would an off-the-shelf product need to eliminate every last cell, “the way we need to eradicate cancer,” he says.
Though off-the-shelf products could alter the supply landscape, most cell therapy laboratories and departments of pathology do not currently manufacture CAR T cells, says Aaron Wyble, MD, medical director of the blood bank and cellular therapy laboratory and assistant professor, University of Arkansas for Medical Sciences.
“Most of the critical manufacturing is done at a large pharmaceutical company, whether they’re doing a commercial product or a research product.”
That leaves gaps that pathologists can help fill.
One challenge hospital systems can face when bringing on CAR T, Dr. Wyble says, is that the apheresis center can be overseen by any number of different specialists, such as pathologists, nephrologists who often also oversee a dialysis center, or hematology oncologists. “How do you get all the clinical and laboratory folks on the same page?” he asks. At UAMS, the Department of Pathology’s transfusion medicine-trained pathologists oversee the apheresis unit to collect the material that will be sent to the CAR T manufacturers.
Regardless of who’s in charge, it’s essential to follow each manufacturer’s requirements for collection. “That product is going to be collected from the mononuclear cell layer of peripheral blood,” he says, with a connection to the apheresis machine either through a central line or peripheral large bore IV access. If the latter, “That center will, hopefully, have experience doing those types of collections with, say, allogeneic stem cell donors.” At UAMS, he and his colleagues rely on interventional radiology, which has its own challenges in terms of staffing and supply.
CAR T manufacturers, for their part, have questions of their own when they consider taking on a new hospital system client.
Being accredited by the Foundation for the Accreditation of Cellular Therapy (FACT) “can help move the process along if your hospital is interested in bringing CAR T on as a treatment modality for your patients,” Dr. Wyble says. But if the apheresis unit is overseen by hematology oncologists, what are the implications for pathologists? “It could be that they may only oversee the storage of the cryopreserved CAR T product,” he says. “They would oversee the processing facility requirements for FACT.”
That launches another round of questions, Dr. Wyble continues. Do they have experience with stem cells? Do they store autologous stem cells used for stem cell transplant? Do they receive fresh, allogeneic stem cells from the National Marrow Donor Program, or do they collect in-house and receive them from the apheresis unit? Even if that unit is run by a different group of physicians, are pathologists involved in testing or cryopreserving the product?
As more CAR T products become available, the challenges will only grow. UAMS performed its first collection and treatment in 2019–2020. “We had just one commercial product,” Dr. Wyble recalls, which gave him and his colleagues time to set up and tweak their standard operating procedures.
The following year they went from one commercial product to four; that number held through 2022. The number of patients grew from one patient in 2019 to six patients in 2020; by 2022, they were treating 50 patients.
Ramping up the number of patients is not inherently difficult, he says. “What is going to be a problem is if you see six patients in a week and all six of those patients have a different product.” For each patient, the lab may be working with a different hospital team, manufacturer, courier service, quality team, and SOP.
Just as important are the financial considerations. CAR T is a movement of money as well as science. And despite laboratories’ many roles in handling CAR T products, “they might not actually be receiving revenue that would be able to help with budgeting.” Dr. Wyble suggests that pathologists act sooner rather than later to collaborate with the clinicians who want to use CAR T therapy, since they’re the ones who can show the revenue it will generate. “You need to work with the clinical team so that all budgeting aspects can be addressed,” including equipment, staffing, and the aforementioned challenges of handling multiple products.
As he eyes the likely changes ahead, Dr. Wyble suggests there are lessons labs can learn from their pharmaceutical colleagues. “What is it about pharmacy that enables them to handle so many different, expensive chemotherapies and monoclonal antibodies,” each with their different handling and infusion requirements?
“The more we could learn from pharmacy, I think, the better we’ll do in the long run when it comes to the cell therapy lab,” he says.
Even with its rapid developments, CAR T is still a young field.
These therapies “require huge infrastructure,” Dr. Alizadeh says, and the cost runs into several million dollars per patient. “And so for any given patient where that investment has been made, we still have a relatively limited experience, considering the first approval of these agents was in 2017. It’s not an extremely long time.
“So we try to learn from every single patient,” Dr. Alizadeh continues, “through a battery of studies that we do on blood and tissue specimens, to learn from our successes and from our failures.”
What else do he and his colleagues need from their laboratory colleagues? They’re still learning that, too.
One open question is how to monitor patients. Most of the CAR T clinical trials had a 15-year requirement for long-term follow-up, though Dr. Alizadeh says it’s not clear what such monitoring will entail. Would the patient see an oncologist and hematologist and be monitored for adverse events through physical exam and history? “Or is it a more complex analysis of the blood—laboratory-based testing of persistence of CAR T cells, or clonal expansion of CAR T cells, or things like that?” he wonders.
It’s not clear how long CAR T cells persist. In some cases it’s been 10 years. “If that’s true in individual patients, then for the average patient, what should laboratory-based monitoring entail?” Dr. Alizadeh asks. “How do we monitor for the persistence of CAR and the associated sequelae?”
“What we don’t want to do is overburden the labs and the patients and the providers with tests that are not necessary,” Dr. Alizadeh adds. “It’s going to be a fine balance between doing nothing and doing something that’s meaningful.” It’s possible clinical studies will be needed to determine appropriate monitoring intervals, be it three, six, or 12 months. Likewise, pathologists and their clinical colleagues will need to figure out what type of testing is best. A quantitative PCR for viral persistence in the blood might suffice. “Or do you need to do integration site mapping and clonal analysis?” he asks, recalling the more extensive profiling he and his colleagues performed in their New England Journal of Medicine study.
This will all shake out over time, Dr. Alizadeh says. “But lab directors will need to be thinking about how to prepare for this.”
What does Dr. Miklos see in his crystal ball?
He suggests these therapies will eventually be applied to solid tumors—GI, pancreatic, esophageal, brain, and lung. The challenge thus far has been identifying targets that are expressed on tumor cells but not on healthy cells, he explains. While temporary ablation of B lymphocytes is manageable in many scenarios, that’s not as easily overcome in cases where, say, normal lung, pancreatic, or esophageal tissue becomes damaged. The search for true tumor antigens that are expressed only in high frequency on cancer cells has been challenging, and researchers might have to settle for a threshold effect of high expression rather than exclusive expression of the antigen on the cancer cell.
That goal is attainable, he says. “We just have to be more clever.” Until now research has been guided by one-drug, one-target approaches, testing agents in isolation that might become part of a drug combination down the road.
But with cell therapies, using AI may enable researchers to act more effectively. The information yielded by single-cell RNA and single-cell DNA analyses is tremendous, Dr. Miklos says, and needs to be put into a pipeline to allow development of AI strategies. “Once we have a strategy, we can engineer that into this wondrous thing called a cell—an autonomous agent that can carry multiple Boolean logic characteristics into the body. It’s a powerful concept. If we can capitalize on this capability to fight cancer, then what can’t we achieve?”
“Think of how much more clever this is than a small-molecule inhibitor that can do only one thing at a time,” he continues. The startling cures that have come in blood cancers will soon be applied to others. His Stanford colleague, Crystal Mackall, MD, has developed novel CAR T therapies now in clinical trials targeting B7H3 in glioblastoma multiforme as well as sarcoma and ovarian cancer. “We have another CAR that targets GD2 in childhood brain cancers with promising success,” he says. They’re also developing combination drugs for blood cancers with CD19, CD22, and BCMA, and multitargeted CAR Ts could potentially be applied to solid tumor cancers as well.
As Dr. Miklos surveys the rapid successes of CAR T therapies, he’s also struck by the elegant simplicity at work. Antigen loss is at the heart of the problem, and adding another antigen can boost survival dramatically and will increase overall survival an additional 20 percent.
For his part, Dr. Alizadeh hopes to see more data on allogeneic products that could allow broader access to cell therapies. That’s the first item on his wish list.
The second is more data on the key determinants of what he calls home run successes as well as strikeout failures, to help refine the next generation of products.
For nonblood cancers, much effort has been made on a range of targets for a variety of diseases, from brain tumors to soft tissue sarcomas to carcinomas. “The early data have been sobering in terms of the challenges of delivering these, and the cells reaching the targets and causing responses,” Dr. Alizadeh says. Nevertheless, the studies have yielded insights that may help launch the next group of agents. “I’m cautiously optimistic that in some diseases we will see activity. Whether they’re going to be anything like lymphoma or myeloma remains to be seen.”
He’s also keeping an eye on how these agents might be used in the nonmalignant setting, especially for targeting B cells, using the same agents that are approved for B-cell malignancies in a range of autoimmune syndromes. “Those early clinical studies show a fair amount of promise, and I suspect we’ll see them used in the not-so-distant future.”
Dr. Alizadeh sees a bright future, though he concedes he was at first skeptical that the early successes of CAR T would translate into long-term benefits.
“The biggest happy outcome is that I’ve been proven wrong. I can talk to my patients and use the words ‘potentially curative’ with a one-and-done kind of therapy. I hope to have more of those discussions with more patients as we get more of these agents that are better, sharper tools,” he says.
Karen Titus is CAP TODAY contributing editor and co-managing editor.