APLAR 2026: CAR T in Rheumatology: Promise or Peril?

Hello everyone, Peter Nash reporting from APLAR, Seoul, Korea 2026 for RheumNow. It's 34° Fahrenheit outside, so it's good to be inside. But today I want to talk to you about the hottest topic in rheumatology, and that of course is CAR T therapy in immune-mediated inflammatory diseases.

They had a very helpful symposium with an oncologist, Georg Schett, the one who started the ball rolling in rheumatology, and a nice review by one of his people he was mentoring, who worked with him and is now a professor in Hong Kong, Dr. Chan.

To get the ball rolling — it's the hottest topic, a lot of interest. For example, at ACR, 92 abstracts on CAR T — that's next week. Looking forward to that. And the oncologist kicked things off talking about a little bit of the history. 1998, the first CAR T — the history of treating and curing things like lymphoma and leukemia. The first case he presented: the first lady who is now 14 years after the first CAR T, cured.

But some of the things he spoke about included the wait of two to six weeks for ex vivo CAR T. Autologous CAR T can be very problematic. There are cases where during that wait, where you withdraw the immunosuppression, the underlying disease has flared, and one lupus patient ended up on dialysis whilst waiting. I don't see how you could stop these people's prednisone given they've been on it for years and need adrenal cover at least.

But it's the lymphodepletion that is problematic. He was saying that with ex vivo CAR T, the actual number of naïve CAR T cells delivered is highly variable, and it's almost as if every patient receives a slightly different load, and that the CAR T cell count is associated both with the efficacy response but also with the toxicity. So that two-to-six-week wait is tricky, as is the lymphodepletion. The prolonged survival of their patients is determined by what number of naïve CAR T cells they can get into the patient. The lymphodepletion and how problematic that is is driven a little bit by the individual cytokine profile — and we'll talk a little bit more about that — and it's the fludarabine pharmacokinetics that is associated with efficacy.

So it's a very interesting oncological background point of view, but that lymphodepletion was the difficult area. I even asked Georg: do you need to have that lymphodepletion, given that our patients have a very small B cell load that you're getting rid of? And he said they started using the regimen that oncology had used, and now they are looking at tapering — reducing the amount of lymphodepletion — but they're testing and studying it to see what the minimum requirements are.

The fludarabine for me is the difficult area. It's genotoxic in these young women, and the immunosuppressive lymphodepleting half-life is quite prolonged, and it's associated with ICANS. And as you know, a couple of the Novartis studies are on hold. The lk CAR is now an abandoned program because of one grade five ICANS death. Those things weren't discussed at the symposium; we'll have to wait for ACR.

But the final point the oncologist made is that in vivo CAR T will be the way to go. It's off the shelf. It's reproducible. You can get it on demand. You don't have to wait weeks and weeks. The product can be uniform. The manufacturing is much simpler, not variable, and it reduces the cost. It can be shipped around the world to places where they're unable to make their own. So it's going to increase accessibility.

Then Georg got up and had a nice presentation. He showed his first few cases that are now cured — as in, he says the goal of this treatment is not an incremental improvement in lupus treatment and control; the prime goal is cure, which he defines as in remission and drug-free for an extended period of time. They're out over five years in some of their patients. He likened it to the early days, such as how we learned with infliximab and the first TB cases, how to manage those safety issues, and now the TNFs are standard treatment around the world.

He talked about the advantages of CAR T, where one cell can kill many. He thinks of it as a living drug treatment, but importantly it can migrate to all organs — so not just blood and bone marrow, but into the brain and into other tissues — and deplete B cells everywhere, so you're not going to have areas of resistance.

He highlighted that patient selection is critical. He recommended high-activity patients with low amounts of damage, rather than low disease activity patients who have already got a lot of damage, and that's the scleroderma population where selection is pretty important.

He talked again about lymphodepletion tapering and low-dose studies. All their patients get cyclophosphamide and co-trimoxazole during the lymphodepletion period. The amount of CAR T expansion, he said, is not associated with toxicity, but there's a hint from the Novartis studies — where they had three deaths related to macrophage activation syndrome — that it's the speed of the expansion that is associated with the problem. They're also querying whether the reagents they use are a cause of that problem. But I get the feeling it's not how big the expansion is, but how fast the expansion is that leads to a problem.

There's a window of a few months where vaccinations won't work, and all tissues are B cell depleted. The naïve B cells return and these are not autoreactive, hence the good long-term response. He also highlighted they've now had 14 pregnancies in people in remission off therapy. All babies were normal. Pregnancy did not flare their lupus — that's almost a test. And most importantly, the babies had normal B cells.

He then went on to talk about relapse rates. Across the autoimmune disease cohort, the relapse rate was 14.5%, but it differed with different indications. The lupus patients had about a 6% relapse rate, the scleroderma patients had about a 20% relapse rate, and the myositis patients had about a 37% relapse rate. He is using a CD19 CAR.

I asked him how you should treat the relapse. Do they respond to the same CAR T being treated again? And he said the experience is limited — they've only had two or three cases where they've retreated after relapse. If you use the same CD19 CAR, it does not work because T cell memory wipes them out. You have to use a different CAR target, and he has had success after a relapsed patient on CD19 CAR by giving them a BCMA CAR.

You'll recall that the CD19 takes out the B cells up to the plasmablast but leaves the plasma cell alone, so it leaves intact long-term memory — tetanus, COVID, chickenpox, Candida, etc. Whereas the BCMA takes out the plasma cell, and from what we saw at ACR last year, all those patients end up hypogammaglobulinemic and need to be given immunoglobulin monthly — a very expensive exercise. I don't think BCMA is a good target if you have the choice of taking out CD19 instead.

Dr. Chan talked about CRS and trying to develop predictors for who will get CRS, so people can be vigilant and treated immediately. She reviewed as big a cohort as she could find around the world — 108 patients who'd had CAR T therapy both in Germany, China, etc. Her predictors were the baseline B cell load and the baseline B cell inflammatory activity. So you could measure the B cell load at baseline, and she said IL-6 was a good marker of the baseline inflammatory activity. If you can't measure IL-6, CRP was a very good surrogate.

She showed a nice study where if the patients had high B cell load and high IL-6, they had a 100% chance of getting CRS. If you had low B cell load and low IL-6, a 50% chance of getting CRS — so that's still very common. She then defined an 80/20 rule: 80% get CRS, which is very common, but 20% get CRS with hypotension and hypoxia and have to be treated aggressively.

So it was a very helpful symposium. It talked a lot about the future being in vivo CAR T for reasons of cost and practicality. It talked a little about the upside of efficacy. We didn't hear a lot about the trials that have been put on hold because of ICANS toxicity and because of the development of macrophage activation syndrome.

One of my concerns is: if you roll out CAR T to sites that don't have experience in using CAR T, you're going to maybe miss the early clues that you need to get in there with aggressive therapy to protect people from bad outcomes. That is one of the reasons why I think this should all be done in primary academic centers with a multidisciplinary team that includes hematology, oncology, infectious disease, neurology, and rheumatology.

We receive about one application a week from companies to study CAR T or T cell engagers in phase one dose-escalating studies. I've approached our public hospital to try and put an MDT together. The nephrologists have stolen all our lupus nephritis patients and we only get to see them if they have a problem outside the kidney. When I tried to talk them into getting involved in such a team, the head of nephrology conferred with his colleagues and said, "Our patients don't need this kind of therapy — they're all doing very well." So they're ten years behind. They don't understand the advantages, but all their patients are living on steroids, mycophenolate, hydroxychloroquine, and Rituxan for years and years and years. So they need a lot of educating, and a lot of work needs to be done, particularly on the safety side, to reassure and to assist with patient selection for CAR T moving forward.

ACR has 12 different companies with autologous ex vivo CAR T, and there's a series of companies working on in vivo. So it's very much watch this space. Peter Nash signing off from Seoul, Korea 2026 for RheumNow. Thanks very much.