I’ll go ahead and get started and thank my co-investigators for the trial up-front. It’s a multicentre trial sponsored by Celgene. Sorry, I’ll start over.
My name is James Kochenderfer from the NCI. I’m presenting a multicentre trial of anti-BCMA CAR T-cell product known as bb2121. The trial is sponsored by Celgene and upfront I’d like to thank my co-investigators in this multicentre trial who are all listed here.
First of all, BCMA is a promising target for multiple myeloma. Clearly multiple myeloma is a disease that is essentially incurable and new treatments are needed desperately. BCMA is a member of the TNF superfamily, it’s expressed universally on multiple myeloma cells and, critically, expression is restricted to plasma cells and only a very small subset of B-cells; most B-cells do not express BCMA. We’ve shown in a phase I study at the NCI initial proof of principle for the anti-BCMA CAR T-cell, this was published in Blood in 2016, Ali et al, and there’s abstract today at this meeting, Brudno et al, it’s actually abstract 524 and will be presented this afternoon.
T-cells transduced on these studies were transduced with a gamma retroviral vector encoding an anti-BCMA CAR with a CD28 costimulatory domain. Significant cytokine release occurred on these studies. The trial that I’m talking about now is with our multicentre trial; the T-cell product called bb2121 utilises a different CAR. This CAR is encoded by a lentiviral vector and contains the same single chain FVE which is right here, called 11e5-3. It contains a 4-1BB co-stimulatory domain and a CD3 zeta T-cell activation domain.
We showed in preclinical testing this CAR was highly effective in vitro with low tonic signalling and was able to eradicate solid tumours of multiple myeloma in mice. Our protocol design started out with low doses and escalated up in a 3 3 manner. We started with a dose of 50 million total flat dose CAR positive T-cells then we went to 150 million, 450 million then 800 million. The general protocol schema was for leukapheresis and then manufacturing which took ten days plus release testing. Chemotherapy was given as a conditioning regimen to enhance the activity of the CAR T-cells - fludarabine 30mg/m2, cyclophosphamide 30mg/m2. CAR T-cells were infused on day 0, follow-up was conducted with blood draws and bone marrow biopsies. First response assessment was done at week 4.
The data which I will present in detail today is about our dose escalation part of the trial. These are the data that we have prepared and certified for presentation. 24 patients, clinical deterioration occurred in three which led to us only dosing 21 of the patients. We had 21 of those patients, all 21, evaluable for response. Manufacturing success rate was 100%. Since these patients were treated and follow-up was completed to a certain degree, we have treated 12 more patients on an expansion cohort. I will not present these 12 in detail today because the data has not been fully analysed yet.
The baseline demographics of our first 21 patients, data cut-off October 2nd 2017. Median follow-up 35 weeks for these patients. Age 58, ECOG performance status was 0 or 1 for all patients because that was a protocol requirement. A large fraction of the patients had high risk cytogenetics, 43%. One very critical thing about this study is how heavily pre-treated these patients were, it was an incredibly heavily pre-treated group. Seven was the median prior lines of therapy. 100% of the patients had a prior autologous stem cell transplant. 100% had bortezomib and 67% were refractory to bortezomib. 100% had lenalidomide and 86% were refractory. 91% had pomalidomide and 71% were refractory to that. We have here the various combinations, common cumulative exposures and 100% were exposed to bortezomib plus lenalidomide and 67% refractory. You can see the other combinations – 91% were exposed to bortezomib, lenalidomide and carfilzomib with 48% refractory, so a very heavily pre-treated group.
Our adverse events, generally this was a very well tolerated CAR T-cell product, especially in comparison to other protocols that I have participated in. 71% had cytokine release syndrome but only 10% was grade 3 or above. We had some neurological toxicity, 24%, but 0% of that was grade 3 or above in the dose expansion cohort. The neurotoxicity was generally much milder and less prevalent than what I’ve seen in previous anti-CD19 CAR studies. We also had neutropenia, thrombocytopenia and anaemia. No dose limiting toxicities were observed in the dose escalation. Cytopenias were mostly related to the flu-cy conditioning. We did have some slow recovery of counts which is somewhat to be expected in a multiple myeloma patient with very low blood counts and extensive myeloma involvement in the bone marrow. We had five deaths, three of these were due to disease progression on the lowest dose level of five million CAR T-cells. We had two patients who had what we call the active doses, which was greater than 50 million CAR T-cells that died. One died of MDS and one died of cardiac arrest. 14 patients experienced one or more SAEs, many of these SAEs were due to our strict protocol SAE requirement where a patient who was admitted only with grade 1 or 2 CRS or fever was counted as an SAE.
This slide summarises our responses. You can see, I mentioned before, the patients on the lowest dose level of 50 million did not have good outcomes with progressive disease in all three patients but after that in the active doses of 150 million, 450 million or 800 million CAR T-cells we had 17 out of 18, 94% of patients, had an overall response. 10 of the 18 patients obtained a complete remission. So, just to look at this, the green is complete remission, 10 complete remissions. Another important point is at the 52 week mark we had five of these patients so far that have met the 52 week, one year, progression free survival standard. We have durable ongoing responses over a year now, ongoing in many patients. Our longest response that is ongoing is 68 weeks at this point. Responses continue to improve as late as month 15 where we had a very good PR to CR transition. Median progression free survival has not been reached in the active dose cohorts and, again, we call the active dose cohorts 150 million and above. So, in general, very impressive responses compared to my previous experience treating multiple myeloma.
I want to emphasise that in our patients now we have somewhat more mature data over the past six months which shows a deepening of responses over time. First of all, time to first response was one month median; time to best response was 3.74 months; time to complete response 3.84 months; duration of response and progression free survival not reached. Importantly, the durability is shown here with a progression free survival rate at six months of 81% and a progression free survival rate at nine months of 71%. These two graphs show the deepening of response over time. We can see here as of May 4th 2017 with n=15 we had a 27% CR rate and if we look at October 2nd 2017 where we locked our data for this presentation we had a 56% CR rate, so the responses deepened. We did lose one responder to progression during that time so we went from 100% response rate to 94% but the responses going on deepen with a shift from PRs and very good PRs to CRs, again, very indicative of the durability of these responses.
In summary, we have shown durable responses with bb2121 at active doses of 150-800 million CAR T-cells. We had durable responses in a very heavily pre-treated population. 94% overall response, 89% very good PR or better, 56% CR or better. Progression free survival not reached. MRD negative results were seen in 90% of the MRD evaluable patients. Disease progression in four patients occurred, three of three evaluable patients remain BCMA positive progression. To date the safety profile has been manageable with doses as high as 800 million CAR T-cells. Two reported events of grade 3 CRS resolved within 24 hours.
I’ve talked mainly about our dose escalation because we’ve locked our data on that but I do need to mention one patient in our dose expansion phase who had a delayed onset reversible grade 4 neurological toxicity. Again, this is on the dose expansion phase and we’re not fully reporting the dose expansion phase but in the interests of full disclosure we are going to disclose this one grade 4 event which occurred in our dose expansion phase. Grade 4 neurologic toxicity associated with tumour lysis syndrome and CRS, patient recovered completely, the neurologic toxicity resolved. The patient had the highest disease burden on the trial and the patient has now obtained a very good PR despite low BCMA expression on their myeloma cells.
Finally, we have a trial that we’re starting up, a global pivotal trial called CARMA which is open to enrolment and the bb2121 dose range will be 150-300 million CAR T-cells. Thank you.