I appreciate the opportunity to present these data on behalf of my colleagues at the Children’s Hospital of Philadelphia and the University of Pennsylvania talking about CAR T cells for ALL; this is a paediatric study. For those of you who aren’t familiar with the concept, a chimeric antigen receptor is depicted here and what you see is an ability to target a cancer cell based on an antigen that is expressed on the surface, in this case CD19 which is on B cell malignancies. The external targeting domain is provided by a piece of an antibody and then there are internal activation domains that allow for full T cell activation with both the signal 1 activation signal and the signal 2 costimulatory signal.
So these genetically engineered T cells are used to target cancer. In theory anything that you can build a monoclonal antibody to you might be able to target with a CAR. The idea here is that we have to collect T cells from each patient so this is very much an individualised therapy. We collect these T cells by apheresis, we then bring them to the lab where we genetically engineer them using a lentiviral vector which is an inactivated form of the HIV virus that cannot cause disease but retains the ability to genetically modify the T cells. This genetic modification then allows the expression of a CAR protein on the surface of the T cell. So the original T cell receptors of the T cell are still there but it gains this new recognition capability through the CAR protein. This allows interaction with the cancer cell and hopefully killing the cancer cell and also, as you’ll see in the data, it allows activation of the T cell which causes very significant proliferation. I think that’s a very important point.
So our experience with this CD19 directed CAR T cell product, which is called CTL019, is that at the University of Pennsylvania and at CHOP over 130 patients have been treated with CTL019. The group that we’re reporting on today is a group of paediatric patients with ALL. 39 patients with ALL of which 36 were in a complete response 28 days after a T cell infusion. There have been relapses on the trial, 10 of them, half of which are related to disappearance of the T cells and these patients have experienced a CD19 positive relapse, and half of which are related to actual antigen escape and these patients have undergone a CD19 negative relapse. There is some correlation between the exposure to blinatumomab, the drug that you just heard about which is also directed at CD19, and this risk of CD19 negative relapse.
We have a median follow-up of 6 months but there are patients who are out a year and even as far as 31 months was the first patient that we treated. So 15 of our patients are out a year or more. We have not seen events in patients who remain in remission after 12 months and I think it’s important to point out that this has not been treated as a bridge to stem cell transplant, only 3 of the patients that we’ve treated have then subsequently gone on to stem cell transplant.
This very complicated slide basically shows you the amount of disease that each of these patients have so it’s important to note on the right side of the slide that there were a small number of patients who had responsive disease to their lympho-depleting chemotherapy and were at 5% or fewer blasts. Then I also want to point out most of the patients had refractory disease at 5% or even 50% blasts. So if you look at the response rate it’s statistically similar in each of these groups, so 5% or more leukaemia and 88% response rate and 50% or more leukaemia and 82% response rate. So the response rate seems to be within the limitations of the statistical design of the study independent of the disease burden which I think is a key point here.
Patients do get ill if they come into the study with high disease burdens and this slide illustrates the impact of disease burden on this toxicity which was mentioned in the previous talk which is cytokine release syndrome. Patients who have less than 50% bone marrow blasts essentially do not have significant degrees of cytokine release syndrome; they have one but it’s not a severe one. Patients who have more than 50% of blasts have a very high likelihood of a significant cytokine release syndrome. So this is a clear difference that we’ve seen in this large group of patients.
Persistence of the cells is a key point here and so these cells, after we give them, stick around for many months in many patients. So here we see the likelihood of persistence from a recent paper that was just published in The New England Journal showing that about two-thirds of the patients retain their T cells for six months or longer. So I think that’s a key point in terms of maintaining remission in these patients.
This very busy slide, I apologise for it, essentially makes one point and that is that the proliferation of these cells is enormous. This is a detection of CAR positive cells by a PCR technique and you have to see that this is a semi-log plot so that up to 10- or 100,000-fold increase in CAR T cells immediately after infusion as these cells go on the hunt looking for cancer cells.
Duration of response has been very favourable so 76% of the patients remain in remission if they go into remission at 6 months and the event free survival rate in this study at 6 months is 70%, again emphasising that very few of these patients went to stem cell transplant. Toxicities exist: B cell aplasia which requires IVIG therapy, cytokine release syndrome, which can be severe in some patients, a macrophage activation syndrome and neurotoxicity but I would say the significant risk is really cytokine release syndrome.
So in summary here we see very high rates of remission induction, 92% in this 39 patient paediatric cohort. We see amazing growth of these cells which I think is really correlated to the response. We see persistence which I think is the key to patients remaining in remission, I think that’s extraordinarily important. We see a cytokine release syndrome, we don’t see it in patients with low disease burden, we do see it in some patients with high disease burden.
So in the end I think that we’re beginning to see the picture of longer term persistence that may allow longer term disease control.