Kites Fly: Effective CAR-T Therapy in Non-Hodgkin Lymphoma? Hematological Malignancies Part 4

Sorry for the slight delay getting this out. I was trying to account for each patient as even 1 or 2 misplaced will impact the response numbers in these small trials. Took a while.

Our last post focused on the CAR technology coming out of the MSKCC and affiliated institutions, being brought together under the Juno company umbrella. Juno was funded by ARCH Venture Partners and the Alaska Permanent Fund, through a partnership managed by Crestline Investors, along with Bezos Expeditions, and Venrock. We noted in closing that CAR T cell technologies were performing very well in acute lymphocytic leukemia (ALL), but not as well in the Non-Hodgkin Lymphomas (NHL). In early data sets response rates were not trending very high.

Recently I came across Kite Pharma’s JPM update on their version of CAR therapy. Kite is financed by Pontifax Ltd., Alta Partners, Commercial Street Capital, and individual investors, in partnership with the National Cancer Institute (NCI) Surgery Branch under a Cooperative Research and Development Agreement (CRADA). This reflects that the technology is coming out of NCI labs.

I was struck again by the duration and response rates reported and the indications they were pursuing. It seems that there is one extra patient in the JPM slide deck, so I went back to the ASH talk to get the right numbers. So lets review. Kite calls its lead CAR construct a very straightforward name: anti-CD19 CAR. Like 19-28z CAR from Juno/MSKCC, this CAR is built with a anti-CD19 scFv, followed by CD28 and CD3 signaling components. Quite unlike the 19-28z effort however, the lead here is NHL indications, specifically as seen here:

Screen Shot 2014-03-26 at 5.28.38 PM

So Kite is focusing on relapsed and refractory NHL. By my count Kite has enrolled 23 patients in refractory NHL. Eight patients were presented at ASH in 2012, and that data published concurrently in Blood (  These patients were lymphodepleted prior to receipt of CAR-transduced T cells, and received CAR-modified T cells plus IL-2. The results of this small trial were impressive, with a overall response rate (ORR) = 75%, including complete responses (CR) and partial responses (PR). Note that in the context of lymphoma and leukemia, partial responses can be highly therapeutic and durable.

Here is the data. Note I lumped the non-responders and the “not-evaluable” patients together for simplicity. CLL is chronic lymophocytic leukemia, FL is follicular lymphoma and MZL is marginal zone lymphoma:

Screen Shot 2014-03-28 at 9.39.06 AM

This is impressive: 6/8 patients responded and 4 of those responses were ongoing at the time the data were reported. The typical CAR-associated toxicities were observed. In the second study there were a modification in the T cell expansion protocol, and no IL-2 was given along with the modified T cells. Several different indications are added to the above list: PMBCL is primary mediastinal B cell lymphoma, DLBLC is diffuse large B cell lymphoma and NHL is non-hodgkin lymphoma:

Screen Shot 2014-03-28 at 9.51.19 AM

Again very impressive: 11/13 patients responded and all but one of these responses is durable as of the time the data were collected. There is no toxicity data in the ASH abstract (ASH 2013 #168). The data compare very well with that generated using the competing CAR technology CTL019 in B cell                   lymphoma  (see

An interesting twist to the Kite story is that they have contracted with Progenitor Cell Therapy (PCT), a subsidiary of Neostem, to provide CRO services for their CAR technology. PCT has facilities in Mountain View, CA and Allendale, NJ. Essentially the patient’s apheresis blood sample will be shipped for T cell isolation, transduction and expansion, then shipped back to to the clinical site as frozen purified cells ready for use. The use of a CRO is interesting, as it will provide Kite with flexibility in accessing markets. This stands in contrast to the Novartis approach of using its own centralized facilities (in NJ, Basel, Singapore) and perhaps the Juno approach, which may be even more decentralized than Kite. We’ll have to see how this develops.

There are a large number of CARs in development, including those targeting different antigens and those having interesting modifications to the expression construct. We’ll come back to this subject repeatedly as data continue to come out.

stay tuned.

Comments welcome. Email: Twitter @PDRennert