Managing Neurotoxicity and Cytokine Release Syndrome Associated With CAR T Cells

Bianca D. Santomasso, MD, PhD

Bianca D. Santomasso, MD, PhD

At the 2019 Society of Hematologic Oncology (SOHO) Annual Meeting, experts addressed the role of chimeric antigen receptor (CAR) T-cell therapy in patients with hematologic malignancies, including lymphomas. Although these treatments are exciting as they enter the treatment landscape, they come with new toxicities that physicians must learn to monitor and treat appropriately.

The 2 most common toxicities that can occur following treatment with a CAR T-cell therapy include neurotoxicity and cytokine release syndrome (CRS). According to Bianca D.Santomasso, MD, PhD, are distinctly different. CRS can appear as early as days 2 or 3 of treatment, but neurotoxicity can appear around day 7 or later. In some cases, neurotoxicity can develop at the same time as CRS, though.

“Historically, we used to lump neurotoxicity and CRS together,” said Santomasso, “but now we understand them as being 2 distinct processes with distinct management.”

The current treatment for neurotoxicity remains corticosteroids, while CRS can be treated with tocilizumab (Actemra), but tocilizumab is unable to treat or prevent the development of neurotoxicity. In some cases, neurotoxicity may disappear on its own, but corticosteroids are still recommended. Physicians should monitor their patients receiving CAR T cells closely to observe the potential development of either of these toxicities, said Santomasso.

“I believe there is a learning curve, but just having a team that is seeing a lot of patients get this therapy makes it easier and less scary that there are these new toxicities,” Santomasso added. “However, the overwhelming majority are reversible, and patients do very well and are very grateful for having the opportunity to receive it.”

In an interview withTargeted Oncology, Santomasso, assistant attending neurologist at Memorial Sloan Kettering Cancer Center, discussed the challenges with treating patients who develop neurotoxicity following treatment with new CAR T cells following her talk at the 2019 SOHO Annual Meeting. She also highlighted how the CAR T cells are changing the treatment landscape for patients with lymphomas.

TARGETED ONCOLOGY: Could you provide an overview to your talk on the toxicities associated with CAR T-cell therapies?

Santomasso:In my presentation, I will be talking about 1 of the principle toxicities that can occur with CAR T-cell therapy, neurologic toxicity specifically. There are 2 main toxicities associated with CAR T cells, including CRS and neurotoxicity, but I am a neurologist, so my talk is about what we know about neurotoxicity and how to identify and manage it. Historically, we used to lump neurotoxicity and CRS together, but now we understand them as being 2 distinct processes with distinct management.

TARGETED ONCOLOGY: What research is currently being done to combat the toxicity of CAR T-cell therapy?

Santomasso:Many investigators have tried to describe what toxicity is and what it looks like clinically. Neurotoxicity usually develops after CRS. It can start around the same time, but it usually occurs after the patients no longer have a fever. It is important that people be aware that can happen and that the timing can be different. CRS usually happens around day 2 or 3, and neurotoxicity can happen around day 4 but as late as day 7 or 10, and even a little bit later.

It is primarily reversible. It usually takes place in the first couple of weeks after therapy, and, interestingly, the patients develop, predominantly, what could be identified as encephalopathy, meaning they have disordered thinking and interactions. The most specific symptom that we have seen is what we call expressive aphasia. People have difficulty naming, speaking, their speech can be a little bit halting, and that seems to be the specific first sign. Then there can be disorientation as well, difficulty with doing things we normally have an easy time doing, such as writing a sentence. That has helped us develop screening tools that can help to administer these therapies safely. We can screen using something called an immune effector cell encephalopathy (ICE) score by doing a 10-point task of basically naming, orientation, following command, counting backwards, and writing a sentence. With this, we can determine whether someone might be developing neurotoxicity. That is great because that then allows physicians to monitor a patient more closely and know perhaps when to intervene.

TARGETED ONCOLOGY: In your practice, how are you managing neurotoxicity?

Santomasso:We typically use ICE screening. We perform it a couple of times per day for patients but at least once a day. A perfect score for ICE is 10, but once the score falls to 9, we become more vigilant and start monitoring more closely. When the toxicity becomes more severe, it can progress beyond language changes and disorientation, and can even [result in] seizures. We also lookout for a level of consciousness.

When someone falls into a severe toxicity case, we give them corticosteroids. It’s unclear how well the corticosteroids reverse the toxicity. I think it prevents the most severe progression, but the toxicity also runs its course. We’ve had patients who have spontaneously recovered without steroids. I think we practice to be on the safe side, so we do give corticosteroids, but we find that giving a short course can help protect the patient. We have not had, in our experience, a change in efficacy, but that is still being looked at and researched actively.

TARGETED ONCOLOGY: What strategies are in the works to better manage neurotoxicity?

Santomasso:We know now that the anti-IL-6 blocker, tocilizumab, which is used to manage CRS, doesn’t work for neurotoxicity. People are interested in understanding whether there are other strategies besides corticosteroids that can be used for the neurotoxicity. For example, could you block IL-6 itself? Could you block some other cytokines, such as IL-1 or GM-CSF? Those are some interventions that people are looking at.

Another question people have is if there are any long-term effects of the transient toxicity. People suspect not, but we see long-term toxicities from lots of other therapies, including chemotherapy and stem cell transplant. It is important since this is a therapy that is moving into the mainstream, that we investigate that because if we were to find that there were long-term effects, it would be something we could maybe develop interventions for to help patients and prevent.

TARGETED ONCOLOGY: How is CAR T-cell therapy currently used for diffuse large B-cell lymphoma (DLBCL)?

Santomasso:There are 2 commercially approved products for DLBCL. They are CD19 CAR T cells, which are expressed by the lymphoma. Again, both of these therapies are associated with both CRS and neurotoxicity, but we have seen very promising response rates. I think now people are looking to see how durable they are, and they appear to be quite durable. We are also looking at what the risk factors are for relapse. If patients do relapse, we want to understand why.

One point I would like to make regarding toxicity is that we have started to understand that even before the patient gets the therapy if they have a high burden of disease, they are more likely to get severe toxicity. It does appear that the lower disease burden of both leukemia and lymphoma are associated with less CRS and less neurotoxicity. Investigators are trying to figure out what the best time during the course of therapy to give CAR T cells in order to be safe and effective.

TARGETED ONCOLOGY: What challenges exist for CAR T-cell therapy in DLBCL, and how are these challenges being addressed?

Santomasso:Some of the challenges are the potential for relapse and the toxicities themselves, so trying to minimize that. Currently, some of these therapies are administered in patients because of the risk of toxicity, and it would be great if we could move some of these therapies to the outpatient setting. There is 1 CAR T-cell therapy that has been given in the outpatient [setting] in clinical trials, and I think if more of that could be done, that would be great.

There is also the issue of manufacturing time and moving these therapies quickly to patients. Often, we are talking about patients who have relapsed/refractory disease. Sometimes, they progress to the point where they can’t receive CAR T cells, even while they are waiting. Faster manufacturing is a better thing. People are interested in off-the-shelf CAR T cells that could be delivered more quickly with less manufacturing interventions, but that is a way off.

TARGETED ONCOLOGY: What is the main message from this talk?

Santomasso:The key takeaway, as far as neurotoxicity and CRS is concerned, is to be aware and know your patient and product. All products seem to have some risk of [these toxicities], but there are different risk factors according to the product or patient. Be aware and monitor patients.

The other point I emphasize is grading. The American Society for Transplantation and Cellular Therapy (ASTCT) has made a very nice grading guideline that should be able to help us to all grade on the same scale so that we understand the toxicities across centers. Previously, there were many different grading scales, so it was hard to pair what was going on in 1 institution with 1 product to another institution.

I believe there is a learning curve, but just having a team that is seeing a lot of patients get this therapy makes it easier and less scary that there are these new toxicities. However, the overwhelming majority are reversible, and patients do very well and are very grateful for having the opportunity to receive it. They do well as far as their disease, too. It’s important to be vigilant but to not be afraid of the toxicity. I think we have some good steps in place already to manage it.

With this ASTCT Grading System, there are 5 components to it. There is the encephalopathy score, which is a 10-scale that can be done at the bedside. It only takes a couple of minutes, and it’s like a paired down mini-mental status exam. That is something that is used classically. It’s a 30-question [exam] and takes little time. This is extracting down to the very basic core.