Category Archives: 19-28z CAR

Three high-altitude take aways from AACR14

The American Association for Cancer Research (AACR) 2014 meeting last week was high energy and high impact. We will dive into particular talks and specific pathways and indications in later posts, in the meantime I wanted to mention a few key themes.

1) Immunotherapy Versus The World.  That’s a deliberate overstatement of a subtle shift in emphasis from last year’s big meetings, where combinations of immunotherapy with just about anything else were the hot topic. This year there were several talks which emphasized the futility of chasing oncogenic pathways and all of their resistance mutations, one after the other, as opposed to letting the immune system do the work. However, it seems to me overly optimistic to believe that immune modulation can defeat a high percentage of patient  tumors on its own, as some speakers acknowledged. Combinations remain necessary although we will have to work past some notable failures in combo trials, such as the liver toxicity seen in the ipilimumab + vemurafenib combination phase 1, discussed briefly by Antonio Ribas               (see

2) Immunotherapy Versus Itself.  In the ultimate battle of the titans, we see different immunotherapeutic modalities squaring off. This is a theme we’ve touched on before in this space, but the  competition is getting heated. In some indications, the leukemias, lymphomas, perhaps melanoma and some other solid tumors, there is an abundance of therapeutic choices, and the hard question of which therapy best suits which patient will ultimately need to be addressed outside of the context of clinical trial enrollment. Several talks really brought this message home. Roger Perlmutter of Merck (and before that, Amgen) envisions an important role for multiple immune therapies including bi-specific antibodies, chimeric antigen receptors (CARs), and immune checkpoint modulators like Merck’s anti-PD-1 antibody MK-3475.  For B cell lymphoma for example, there is blintumumab (Amgen), a potent bi-specific that redirects T cells to CD19+ tumor cells (and normal B cells), and there is CTL019, a CAR therapeutic which does much the same thing. The therapeutic profiles and toxicity differ, but the general idea is the same. One big difference is that while CTL019 drives T cell expansion and the development of long term anti-tumor memory, the bi-specific does not. Which is better? We don’t know yet. He did not mention that one might do well trying a course of BTK inhibition plus anti-CD20 antibody therapy, perhaps with restricted chemotherapy first e.g ibrutinib plus rituximab and chemo (R-BR or R-F). That choice comes down to efficacy, then toxicity, and eventually cost. Efficacy seems to be a home run with the CAR therapeutics, although these may run into trouble in the area of toxicity and cost calculation. Renier Brentjens discussed the CAR therapies being developed under the Juno Therapeutics umbrella. Acute lymphoid leukemia (ALL) can be treated with CAR 19-28z modified T cells to achieve a >80% complete response rate with >70% of patients showing no minimal residual disease, an outstanding result. However, 30% of treated patients end up in the ICU due to cytokine release syndrome and other toxicity, and recently patients in the ALL trials have died from unanticipated tox causes. Juno stopped 5 trials of their CAR technology last week due to toxicity. Apparently one patient died of cardiovascular complications and another of CNS complications (severe uncontrolled seizures) – it was hard to nail down as Dr Brentjens had gone off his prepared talk for these remarks which were off the cuff, so comment please if you have better info on this. Carl June discussed Dr Brentjens’ presentation, noting that the clinical results were really quite striking, and contrasting the CD28 motif-based CARs with the 4-1BB-based CARs (as designed by Dr June with U Penn and licensed to Novartis). He also stressed that in chronic lymphocytic leukemia (CLL) they have had patients who have failed up to 10 prior therapies, including rituximab and/or ibrutinib, and these patients have responded to CAR treatment. That’s very impressive data. The roadblocks to widespread use of CAR therapy however are large and include the toxicity, the “boutique” nature of the current protocols, the cost. Perhaps, Dr June suggested, CAR will end up as third line therapy, reserved for salvage therapy. I for one hope not.

Also in the immunotherapy space were hot new targets (e.g. CD47, OX40, GITR), advances on the vaccine front, and a few surprises. We’ll update soon.

3) The Medicinal Chemists Have Been Busy.  Not to be drowned out by the Immunotherapy tidal wave, small molecule therapies targeting specific oncogenic pathways continue to be developed and show promise. Most readers will be aware of the high stakes showdown (so billed) between Novartis, Pfizer and Lilly in the field of specific CDK4/6 inhibitors – in addition to bringing forward some really nice phase 2 data (we’ll discuss these another time) this “showdown” also illustrates that current portfolio strategy drives a lot of overlapping effort by different companies. As expected, much of the action is moving downstream in the signaling pathways, so we saw some data on MEK1 inhibitors and ERK1/2 inhibition. There were some new BTK inhibitors, nice advances in the epigenetics space, and some novel PI3K inhibitors. All grist for the mill.

stay tuned.

Update from AACR14: Clinical Halt for Memorial Sloan Kettering/Juno Therapy in Non-Hodgkin Lymphoma

Yesterday we learned that the Memorial Sloan Kettering Cancer Center (MSKCC) and corporate partner Juno had stopped enrolling patients into 5 clinical trials of their chimeric antigen receptor (CAR) T cell therapies. Details are spare at this point, but unexpectedly, the cause of the clinical stop was severe cytokine release syndrome (CRS). I say ‘unexpectedly’ because it was just last month that MSKCC released an update on their ability to detect CRS early enough to initiate aggressive treatment. We commented on this update in a recent post on the CAR 19-28z technology.

According to the MSKCC update given in February, they had developed “guidelines for managing the side effects of cell therapy” including CRS, and “diagnostic criteria” for identifying at-risk patients using clinical lab tests. These tests were for a panel of cytokines and for C-reactive protein (CRP). To be fair these comments were made in reference to work ongoing in acute lymphocytic leukemia (ALL), but it was clear that the clinicians felt they were broadly applicable. It seems now that these comments were premature.

This is a critical issue in the CAR technology field, potentially holding back not just MSKCC/Juno but similar work from U Penn/Novartis and NCI and partners working with Kite Pharma. The syndrome characterized as CRS is a consequence of the massive immune response to the tumor, which is a designed consequence of the CAR technology. CAR-modified T cells are potent cytotoxic agents, and are designed to recruit unmodified T cells to the cause (the so-called bystander effect). This result is the triggering of the acute phase response, and then an outpouring of cytotoxic compounds, pro-inflammatory cytokines, and effector proteins. When allowed to proceed unchecked, the response begins to engulf normal cells and tissues, causing additional cell death, organ damage, and in the most severe cases, death.

The reality is that clinical responses leading to CRS seem to have caught MSKCC/Juno flat footed in at least one clinical trial of Non-Hodgkin’s Lymphoma (NHL) – we note here that stopping 5 trials does not mean that CRS was seen in all 5, but they are related by clinical indication, so it is an obvious precautionary step to take.

What will happen next we cannot know yet, as we have not yet heard the necessary detail. At the very least the MSKCC/Juno NHL programs are in for careful scrutiny. This will impact the clinical development of the technology and slow access for patients. The patients affected are those who are the most in need, nonetheless, the caution is warranted. More broadly, this unexpected turn of events may encourage us to look again at more established therapeutics for NHL, including small targeted molecule drugs, cytotoxic antibodies, antibody-drug conjugates (ADCs) and bispecifics, and of course, combinations of those therapies with one another or with current chemotherapeutics.

And this just in: In the immune checkpoint space we have just learned this morning of the potential for unexpected immune toxicity after long term treatment. Thankfully this appears to be very rare but this too will bear watching.

Hematological Malignancy Treatment Landscape. Part 3: My other CAR is a …

A few folks kindly emailed to point out that I had not mentioned the Memorial Sloan Kettering Cancer Center (MSKCC) in my previous post. The sin of omission. Apologies, but I’ve been trying to digest the press release that the MSKCC put out on February 19th. The presser served two purposes I think, one valid (here’s some data) and one just pure PR grandstanding. Plus, the release didn’t link to the actual study. That was published in Science Translational Medicine the same day as I found out after some text searching, only to run into a paywall.

So, here is the referenced paper:

Davila ML et al. 2014 Efficacy and Toxicity Management of 1928z CAR T Cell Therapy in B Cell Acute Lymphoblastic Leukemia. Sci Transl Med. Feb 19;6(224):224ra25. doi: 10.1126/scitranslmed.3008226.

The data from MSKCC investigators (the PIs are Renier Brentjens, Isabelle Rivière, and Michel Sadelain) is very impressive. In a small phase 1 study of patients with relapsed of refractory Acute Lymphoblastic Leukemia (r/r ALL), patient T cells were transduced with the CAR construct 19-28z CAR and reinjected into the patient. As seen with CTL019 (previous post), this is an effective strategy for producing a lasting T cell response to the CD19-positive leukemia. In the 16 patient trial, 88% of patients responded and just under half went on the receive allogenic HSCT, the standard of care for ALL. So thats great. Just to add some details, a complete response (CR) was reported for 14/16 patients. Two of these patients were already minimal residual disease (MRD)-negative when enrolled. The MSKCC study defines MRD- patients as CRm. The CRm is given as 75% so I’m guessing here that the 2 patients already MRD- at enrollment were not re-counted. Of these 75%, one third were further defined as CRi (CR with incomplete cellular recovery) meaning that they were cytopenic for one or more cell types. Other than this reference to toxicity, the paper generally focuses on cytokine release syndrome (CRS), noting that CRS can be accurately tracked using the routine clinical laboratory marker CRP. More importantly the investigators use a cytokine panel and CRP to define a severe CRS (sCRS) group of patients and a non-severe (nCRS) group of patients. Please note here that if you take the time to burrow through the supplemental tables you will find that both sCRS and nCRS patients had grade 3 and 4 adverse events including hypotension, febrile neutropenia, hyponatremia and altered mental state (CNS toxicity). sCRS patients (7/16 = 44%) further presented with fatigue, atrial fibrillation, sinus tachycardia, electrolyte imbalance, hypoxia and respiratory failure. It’s an important distinction, as the sCRS patients remained hospitalized, including ICU care, for an average of 57 days, versus 15 days for the nCRS group. Both groups are incurring very high health care costs post-treatment.

19-28z CAR differs from CTL019 in several interesting ways. The extracellular domain is a CD19-specific scFv. This is genetically linked to transmembrane and cytoplasmic domains from the costimulatory protein CD28, linked in turn to the CD3 epsilon signaling motif. CTL019 is similar overall, but uses 4-1BB domains instead of CD28 domains. One important consequence of the use of CD28 versus 4-1BB as the costimulatory effector domain is the persistence of the expanded T cell population after injection into the patient. For the 19-28z CAR T cells described in the present paper, the average duration was 3 months. The authors suggest that this is shorter than the duration seen with CTL019 treatment, and that this may be beneficial. That remains to be seen, as we really need duration of disease remission data to understand the benefit/risk of maintaining the T cells that control the leukemia. A final note on this: the MSKCC investigators go to some lengths to sketch out a clinical development plan that can be applied to various centers (not just theirs). This is a very important and useful development, which we would hope make these cellular therapies more widely available.

On the basis of this early data presented here and the phase 2 data coming from the U Penn group (see the prior post), its safe to say that we are watching the development of a new standard of care for ALL patients.

Back to the presser for a moment. The rest of the press release was amusing in a ballsy kind of way. There was the breathless prose and the very pointed “we were the first” claims that science journal editors (but not PR directors) generally avoid. As we all know, none of this science developed in a vacuum and the claim of precedence will be sorted out by patent courts as necessary. For the record I really have nothing against their PR campaign, it just makes me squirm a little. I happily support the MSKCC every year by donating through their Cycle for Survival Program (thanks to the strong legs of longtime friends John and Susan Canavari). Also I really look forward to watching this technology develop, along with technology from the Fred Hutchinson Cancer Center and the Seattle Children’s Research Institute, under the umbrella of Juno Therapeutics, which was formed in Seattle in December specifically to advance this and related therapies.

It’s pretty clear that this technology is very important and very very exciting.

Now, why don’t these cell based therapies work as well in other indications, like Non-Hodgkin lymphomas?

stay tuned.