I’m an adult haematologist and one of the areas of unmet need in our specialty is relapsed and refractory acute lymphoblastic leukaemia. Essentially, whenever patients relapse after having had a bone marrow transplant there are very few treatment options available to them. The reason we were looking at CD19 CAR is because, of course, there has been huge success with CD19 CAR T-cells for paediatric and young adult ALL. So we thought that it would be a good population to look at for this type of therapy.
The problem with adults who have relapsed and refractory ALL is often they are older and they have quite a lot of comorbidities. They can be particularly vulnerable to the side effects of treatment and we all know that CAR T-cell therapy can be quite challenging to withstand – there are toxicities associated with this treatment. So what we’ve done at UCL is we’ve designed a type of CAR T-cell that has a fast off-rate from CD19. By that I mean that it has a different binding kinetic to some of the other conventionally available CARs. In having that rapid binding kinetic and the fast off-rate it means that essentially it behaves in a more physiologic manner. It means that it doesn’t necessarily secrete quite as many cytokines as the other available CARs. This, we believe, has a knock-on effect on the risks that the patients will have of cytokine release syndrome from this CAR T-cell. Actually cytokine release syndrome is probably one of the biggest problems in ALL, it’s one of the most challenging things to manage.
So that was the ethos behind why we felt a novel CAR was indicated and a low tox CAR, if you like. It was with that that we developed the study.
What were your methods?
What we did was we’ve got a twenty patient study. It’s a phase I study and the patients are enrolled in, and it’s an autologous CAR T, second generation 4-1BB.zeta CAR. The manufacturing takes approximately four weeks so the patient is leukapheresed and then they undergo bridging therapy. Then once the product is QP released and it meets criteria the patient is admitted to UCLH for treatment. Effectively what we do is we give a period of lymphodepletion treatment, so that’s chemotherapy, to effectively create space for the incoming CAR T and that’s fludarabine and cyclophosphamide.
Then we designed the study in such a way that safety is really at the heart of it. So the first thing that we do is we look at the burden of ALL that the patient has before they come into us. If they have high burden disease, and by that we’ve defined that as more than 20% of the leukaemia blasts in the marrow, then we give a lower initial dose of CAR T-cells.
The second thing that we’ve done is that we’ve split the dosing, it’s just a safety design, so that if the patient gets any sort of toxicity over the first nine days following the first dose that we have the option to withhold that second dose and not contribute to any further toxicity. We felt that that design was protecting this at risk patient population. So those are the two safety design features.
Then the patient essentially we evaluate their disease burden at month one. So we look and see what the CAR T-cells have been able to do at that point. Then the follow-up is ongoing for the next two years on the study.
The kind of patients that we recruited into this study, again they were patients with adult B-ALL. The median age of our patients was about 43 years and they were a population of patients pretty heavily pre-treated with chemotherapy and transplantation before. So 47% of our patients had actually had inotuzumab before and 26% had been exposed to blinatumomab. 63% had had a prior transplant so these patients had been through a lot of treatment. Actually they also had quite a heavy burden of disease when they came to us so 42% of them had more than 50% blasts and 21% had between 5-49% blasts. The more disease you have the more high risk you are of toxicity from treatment so we were aware that we were going into this with quite a high risk profile patient population.
In terms of the manufacturing, again this is a phase I study so it’s not a commercial product. We manufactured this in house and we used two different types of manufacturing platform – we had an open process which we found to be quite labour intensive and it was very difficult to scale to the number of patients that we wanted to treat. So through the study we switched over to a semi-automated closed manufacturing process using a platform called the [??]. We found that to be hugely helpful to us because we were able to turn around patient products much more quickly. We found that actually the characteristics of the products were very heavily skewed towards memory T-cell populations and actually that seems to be pretty important for CAR T-cell therapy. You want to have a good memory population to enable your CAR to persist in the patient and that is important for B-ALL where persistence is really pretty key.
So that was the manufacturing. Then the next thing we checked on the study was whether or not the CAR T-cells did persist. We were able to do that just looking at the DNA for the CAR in the patient blood. The nice thing about this was whenever we got patients over the twelve month mark we saw that actually the CAR was still present in the blood using this DNA analysis. We were also able to look at the pharmacokinetics of the expansion of the CAR T in the blood as well. We noted that it was very comparable to that that was reported on the ELIANA study, that’s the paediatric and teenage/young adult pivotal trial of CD19 CAR for B-ALL. We were very happy with that, indicating that our CAR can expand equivalently to and perhaps even slightly more than what’s perceived to be the gold standard CAR. Not only that but it persists in the blood.
Probably the most important thing, really, is the fact that when we used this CAR the safety profile was really excellent. We’ve already defined the patient population as being high risk for toxicity by virtue of the amount of the disease that they have in their bone marrow. Actually for cytokine release syndrome we didn’t actually see any grade 3 CRS on our study and that compares really favourably with what’s out there from other CD19 CARs for ALL. The fact that we didn’t see any cases is really quite… we were very pleased with that. We did see grade 2 CRS, we saw that in seven of the patients and that usually defines patients who have got a high fever and maybe they’ve had a drop in BP, in their blood pressure, associated with that. But that was really the worse extent of CRS that we saw.
We did use tocilizumab in a proportion of our patients, in 32%, to mitigate CRS. Those patients that did get the grade 2 CRS, they all had heavy burden disease. It’s very easy to predict who may potentially develop that complication from the outset based on the blast percentage. So that was that side of the toxicity and then in terms of neurotoxicity, that’s another big problem with CAR T-cell therapy and particularly neurotoxicity that’s very slow to recover with appropriate therapy. So what we actually saw, we saw this neurotoxicity arise in four out of the nineteen treated patients and we saw one patient with a grade 2 event and three patients with a grade 3 event. But, again, the important thing to note is that the neurotoxicity responded really swiftly to the introduction of steroid treatment. So we had a couple of those patients who had the grade 3 events they resolved to grade 1 in less than 24 hours with steroids and the other one within 72 hours. So again it’s quite easy to manage that syndrome in some ways and it was also very easy, again, to predict who was going to get it because it was the patients who had the high burden of disease who went on to get the neurotoxicity. All of those cases, indeed, were preceded by CRS as well. So you’ve got two pieces of information that help guide and you can be more vigilant with these patients knowing that they are potentially at higher risk downstream.
So that was the kind of incidence of the common toxicities that we see with CAR T. There were cytopenias as you see with all CAR Ts but most of them did respond by about month three and had normalised. So we were very heartened by that tox profile on study. What we did see was that in our current cohort eleven out of our nineteen patients, that’s 58%, are currently disease free and that’s at a median of 12.2 months of follow-up. I think we just have to put this in the context of this is a patient population for whom there really are no other effective therapies. So it’s really quite an important piece of work.
The other thing to say about this is the fact that these patients are in remission but without having had a further line of consolidation. So some people argue that maybe a bone marrow transplant or a second bone marrow transplant might be indicated if you do manage to achieve a remission with a CAR T. But in our hands to date there have been very few of these patients have actually had a bone marrow transplant and we’ve got five or six of them who are over the twelve month mark without having had another line of treatment. So watch this space, I guess we’re watching the patients’ progress on this but we’re very encouraged to date by what we see.
What we’ve learned from this study to date is the fact that the manufacturing method, the in-house semi-automated manufacturing method seems to be pretty good. It makes a good product, it’s got lots of memory T-cells and it seems to expand really nicely in the patients. So, really nice persistence of the CAR Ts which we think is probably helping to protect patients against the risk of relapse downstream. Really, of note, the fact that it was very tolerable from the safety perspective so patients weren’t getting very severe cases of CRS and neurotoxicity. In fact, when neurotoxicity arose it responded really quickly to steroid treatment and that’s very reassuring for us.
From the point of view of efficacy, 84% of our patients achieved a complete response to treatment at one month. It just appears to be durable – as I say, 58% are disease free at a median of twelve months. So with this information we hope to invoke a global phase II study. We’re looking at taking this to a bigger population of patients where the manufacturing will no longer be done by my small team at UCL but will be done on a bigger scale. Hopefully we’ll just get to confirm these findings and perhaps be able to offer something to adult patients with ALL who really don’t have many other treatment options available.
What are you looking at working towards in the future?
We’re also looking at CD19/22. Because the thing with ALL, one thing that people always want is to optimise the CAR T-cell product itself and we’re looking at lots of different manufacturing technologies and additives to be able to make the CAR expand even more and to essentially enable even greater persistence. We’re aware of the fact that on this study, and like any study targeting a single antigen, that antigen escape is one of the Achilles heels of the treatment. So there have been three instances of CD19 negative relapse in this study so my next plan is to try to create a CD19/22 study to see whether or not we can try to prevent those CD19 negative relapses from happening. So we’ve got manufacturing progression and we’ve also got this antigen targeting approach to try to minimise CD19 negative, that’s our big next agenda point.