Management of CAR-T Toxicities: CRS, Neurological, Cytopenias, Infection

A roundtable discussion, moderated by Thomas Martin, MD, of the UCSF Helen Diller Family Comprehensive Cancer Center, focused on CAR T-cell therapy considerations in the treatment of multiple myeloma, including data on approved CAR-T options and a look at the pipeline. Dr. Martin was joined by a panel that included Sagar Lonial, MD, FACP; Peter Voorhees, MD; and Shambavi Richard, MD.

In the next segment of the roundtable series, the panel discusses side effect management with CAR T-cell therapy, including CRS, neurotoxicities, infections, and more.

—

Dr. Martin: I want to branch off to toxicity. Once a patient gets infusion of chimeric antigen receptor (CAR) T cells and in our two products, they’re a little bit different. [With] idecabtagene vicleucel (ide-cel), we try to give between 300 and 450 million cells, whereas ciltacabtagene autoleucel (cilta-cel) patients are given somewhere between 40 and maybe 60 million cells for a little lower. There are some differences in toxicities too. Let’s talk a little bit about CRS or cytokine release syndrome. Those are probably the most significant in terms of the incidences for toxicity after CAR T-cell therapy. Peter, what have you seen with ide-cel and cilta-cel in terms of CRS?

Dr. Voorhees: As far as the incidence, [they are] very similar. In the ide-cel experience, the overall rate of cytokine release syndrome was 84%. When you look at grade 3 and higher, it was 5%. The overall rate of CRS with cilta-cel was a bit higher at 95%. But the rates of grade 3 and higher CRS was very similar at approximately 4%. The median duration of CRS was also very similar between the two products, it was five days with ide-cel; it was four days with cilta-cel. It’s the timing of onset that’s different. The timing of onset with ide-cel is one day. It happens early, and this may in part be due to the fact that you’re giving a higher cell dose right from the get-go. Whereas perhaps there needs to be some degree of expansion with cilta-cel before you start to see those CRS manifestations declare themselves.

But with cilta-cel, the median onset is actually seven days. You have to keep that into consideration when you’re constructing your algorithms for inpatient versus outpatient CAR T-cell therapy, when to anticipate where things could go wrong. But thankfully, in the overwhelming majority of cases, both with ide-cel and cilta-cel, [it is] a manageable problem with majority being mild to moderate in severity, and tocilizumab was used in about half of the patients with ide-cel; it was used in about 70% of the patients with cilta-cel.

Dr. Martin: Interesting. In San Francisco, because we have bridges, it makes it challenging for us to do some of this as an outpatient; we will generally give low-dose chemotherapy if we can as an outpatient, but once we start with the infusion, we do inpatient. But how about at your guys’ center? Do you guys prefer to do one versus the other based on inpatient/outpatient strategies? Sagar, how about at Emory?

Dr. Lonial: No, not necessarily. I don’t think we’re quite ready. We have a 24-hour [immediate care center] for cancer that will help us to do both CARs and bispecifics; we’re not quite ready with the logistics yet I think to do them as outpatients.

Dr. Martin: Shambavi, how about in New York? Are you guys able to do anything outpatient or mostly inpatient?

Dr. Richard: No, we’re doing them all inpatient. Our outpatient, the challenges, again, [are] the traffic and the pathway through our incredible emergency room system.

Dr. Martin: Exactly, and there’s never a free bed. That’s the hard part. The other toxicity that’s of concern, and you’ve addressed it a little already, Shambavi, has been the neurologic toxicity. Can you just tell us about what your experience has been with ide-cel and cilta-cel in terms of the neurotoxicity?

Dr. Richard: Yeah, so there’s the immune effector cell-associated neurotoxicity syndrome (ICANS), which is the typical neurotoxicity that we talk about that’s in the setting of the CRS; [it] can happen at the tail end of CRS, right after the CRS, any time point, but generally tends to happen in that first two weeks of their hospitalization, most of the time. That seems to be pretty similar between the two groups. So, 18% with ide-cel, 16% with cilta-cel. In terms of grade 3 and higher, both under 5%. That seems to be more or less comparable. Not big differences there. But in terms of the delayed neurotoxicity, this is where there seems to be some differences. We’ve seen delayed neurotoxicities with cilta-cel; not that much reported with ide-cel.

In terms of the delayed neurotoxicity, or other toxicities—and these seem to take various formats, initially called motor and neurocognitive syndromes, which has now been renamed because they really look more like Parkinsonian-like syndromes—with the cilta-cel with CARTITUDE-1, there were six patients who had this particular unusual kind of toxicity. Then seven with other patients who had other kinds of neurotoxicities, which included things like cranial neural palsies or peripheral neuropathies, so sensory neuropathies, and things like that. These tended to be a little bit harder to manage, but the ICANS were easier to manage with steroids, generally took care of it the majority of the time. But with the delay, steroids didn’t seem to be that effective. Most of the time it was supportive care and just watching and managing, but the Parkinsonian ones were significantly harder to manage.

In fact, there have been three deaths with patients who had that, two from sepsis-related, one from just progression of Parkinson’s that couldn’t be controlled. [There are] two others that are either stable, not fully recovered, or slowly recovering, or recovered. What was seen with these were risk factors for these harder-to-manage and less-understood delayed toxicities. One was the grade of CRS. How quickly are you controlling the CRS? What was the disease burden? Higher disease burden, progressive disease at the time of CAR-T, any degree of ICANS, particularly higher grades of ICANS, all of these were risk factors for developing these. With the CARTITUDE-2, these mitigative strategies were set in place, where we were more aggressive with disease control. One of the biggest changes that we made was early and aggressive management of CRS, regardless of now the kind of product that we use. [We] really don’t wait for that grade 2 CRS before we start using our anti-L6 receptor antibodies and things like that.

We use those almost immediately with the first or second fever, especially if the fever is consistent, especially if it’s higher grade. Then we do that; we try to control the disease as much as possible. These have shown in the CARTITUDE-2 trials to reduce that risk from 5% to less than 1%. I think we’re on the right track with this. The others that are a little bit harder to control are high CAR T-cell expansion exposure and persistence, which seems to be somewhat arbitrary. We don’t quite understand who gets those high expansion numbers. Is it always relatable to high disease burden or do they happen independently? I think there’s a lot we need to learn with all of that.

Dr. Martin: Yeah, it’s a very interesting topic. In fact, we are also a believer of early intervention of CRS. We’re a first fever group. The minute the person has first fever, doesn’t matter which product, what time, how many days after the CAR-T, we give tocilizumab with the first fever and then if they have a second fever, we add steroids right away. We are very aggressive too. I’m just curious because your institution has published some data that BCMA may have some expression in the brain. Are you a believer that that may be the etiology of the late neurotoxicity?

Dr. Richard: I think it’s a very interesting question. One of the patients who died with this was from our institution. We did a fairly extensive workup with that patient, including autopsy studies and things like that. We did see BCMA at the basal ganglia at the nucleus on those autopsy studies, and looking at the database, BCMA expression has been seen. The problem is these are not very standardized in terms of detecting these on tissue outside of the bone marrow. I think that we need a lot better studies on this, really focusing on this, looking at healthy brain, looking at what happens with the CAR-T. We’ve done [lumbar punctures] and looked at cerebrospinal fluid (CSF), and of course we see CAR T cells in the CSF, but we know that with CRS or with ICANS, the blood/brain barrier is disrupted.

It’s not unusual to find CAR T-cells anyway in the CSF. I think there’s a lot to learn from all of this. While the science is very interesting, we are still having to act on what to do with patients while these things are still going on. One of the other things that we have done because, and there is preclinical data to support this, wherein monocyte- and macrophage-derived interleukin-1 (IL-1) are big drivers of these kinds of more unusual, either refractory CRS or neurotoxicities or maybe even hemophagocytic lymphohistiocytosis. And so we use anakinra, [an] anti-IL-1 receptor antibody, quite a lot, and sometimes it even becomes a second-line agent, wherein very quickly, if we are finding the ferritins are rising, we don’t wait for those tens of thousands. If the ferritins are rapidly rising, if they’re tripled or quadrupled, I just introduced the anakinra right away because it has pretty low toxicity and maybe it helps to mitigate all of these things.

Maybe it helps to even be steroid-sparing. I tend to use steroids much less because our patient in fact didn’t respond with any of these maneuvers with the Parkinson’s and got a lot of aggressive cytotoxics to try to control this, and ultimately died of an infection. I think all of these are very real problems in our patients and things to look at. The other interesting thing with this is that when we did the dopamine uptake studies, they were negative. I don’t think they have the same pathology as regular Parkinson’s patients. I think it’s Parkinson-like, but there’s some differences. The easiest thing to blame is that there is BCMA expression in these cells, but I think we have to prove that better.

Dr. Martin: It’s very interesting, and it’s very difficult to treat, as you had said, and they don’t really respond as well to Parkinsonian-like standard treatments, etc. The other toxicity that actually occurs quite frequently, probably the most frequent, is cytopenia. Sagar, what is your strategy at Emory in terms of treating cytopenias in people that have received CAR T-cell therapeutics?

Dr. Lonial: Yeah, I think a lot of it really depends on the timing. Obviously early you think about CRS-related or low-dose chemotherapy-related cytopenias. Again, the bystander effect with all those dying myeloma cells in the marrow sometimes suppresses normal hematopoiesis, and it takes a little bit of time to recover from that. The one thing that I think you really do need to make sure is being checked early on is antiviral issues. I think that really is critically important because certainly [with] cytomegalovirus (CMV), we’ve had one patient who dropped all of their counts and it was all related to CMV, so making sure that you do the appropriate infectious workup early on, a broad infectious workup. I think many of us do serial antiviral titers over time just to make sure, like you would in a post-allogeneic transplant patient, monitoring, not necessarily intervening, unless there are other signs or symptoms, but being aware of what’s going on with things like CMV and potentially Epstein-Barr virus.

Then when you get a few months out and you look in the marrow and there’s no myeloma but there’s also not a lot of normal marrow and you’ve gone through all the infectious issues and all the other things, that’s when I think about giving a stem cell boost if I’ve got cells left over, and [in] a lot of these patients, I do have cells left over. Ideally, if I’m going to do that, I want to do that in the first few months after the CAR infusion because if there are contaminating tumor cells in the graft, then you feel a little bit better if they’re circulating CARs. But if you’re a year out and their CARs are gone, then you worry a little bit more about introducing something.

But I’ve had patients who I didn’t have extra cells on, had a very empty marrow and were thrombocytopenic with platelets in the 3,000 and 4,000 range for a year, year and a half after CARs. I think it is something you got to pay attention to. I don’t have any hesitation with growth factor across the board. I think we all feel pretty comfortable using whether it’s white cell, red cell, or platelet growth factors to try and make them transfusion independent.

Dr. Martin: You can use those long-term no problem?

Dr. Lonial: Yeah.

Dr. Martin: Very good. You mentioned infections. I’ll just briefly say with infections, there are periods of infections, just like after autologous or allogeneic transplant. There’s the early period where patients actually develop neutropenia from the low-dose chemotherapy and from the CRS. We at UCSF, we use prophylaxis for that. We use bacterial and antifungal prophylaxis. If patients have CRS, they do get therapies like tocilizumab and sometimes steroids. There is an incidence of bacterial infections and fungal infections during that period of time. If they have long-term neutropenia, we would consider switching the antifungal from a Candida preventative to more of a mold-preventative therapeutic. But it’s also about the longer-term risk. Peter mentioned this earlier, as they develop severe hypogammaglobulinemia that can sometimes last six months, 12 months, or even longer.

And we do give intravenous immunoglobulin (IVIG) support to patients, especially if they have a serum immunoglobulin G level of less than 400, they get serial IVIG to try to keep them above 400. Patients are on varicella-zoster virus prophylaxis with acyclovir in our hands up for a year. Then we also use Pneumocystis jirovecii pneumonia prophylaxis with whatever therapeutic. We tend not to use the sulfa-based drugs just because it can cause cytopenias. We use other drugs like atovaquone, etc., and those continue until their CD4 cell count gets greater than 200, essentially in our experience. If somebody does come in with a febrile episode, like you said, Sagar, you have to look for all the opportunistic viruses and infections and so adenovirus, parvovirus, CMV reactivation, hepatitis B reactivation, you need to look at all those things depending on the clinical scenario, which is really important.

—

Continue on to watch the next roundtable segment.