After setbacks to Novartis and Juno, Kite is the most advanced player in the first wave of CAR-T projects, and now it must deliver the goods. Interim readout of the Zuma-1 trial in lymphoma should determine whether its lead asset, KTE-C19, has a chance of becoming the first commercial CAR-T therapy.
Given the importance of this dataset it is remarkable how unclear it is what the US FDA might deem acceptable. Indeed, given the wide range of readouts that the sellside thinks might move the stock, attempts at elucidating the goal of Zuma-1 look like guesswork.
The clearest thing right now are timelines. Zuma-1 – an open-label, single-arm trial – completed enrolment of 72 patients into its diffuse large B-cell lymphoma (DLBCL) cohort in July, meaning that 12-month follow-up will not be available until mid-2017. But it has a complex design, and the dataset in question relates to only some of the patients taking part.
The phase I stage of Zuma-1 enrolled seven refractory lymphoma patients, and phase II ups the ante, with 112 patients in two distinct cohorts. The all-important imminent readout relates to a three-month data cut, comprising 51 DLBCL patients, on the strength of which Kite wants to file KTE-C19 by the year end.
|% of market cap||59%|
|Event type||Interim data from pivotal Zuma-1 study|
It is likely that topline data will be press released before Kite’s October 18 investor day, with detailed results at December’s ASH meeting. Kite also has a Zuma-1 abstract at Esmo on October 7 (number 1048O), which likely relates to the phase I stage, but might be a placeholder for a phase II efficacy presentation if these data are generated in time.
At the AACR meeting in April Kite presented an update on six evaluable phase I patients, reporting three with complete remissions and three with progressive disease. One of the ongoing CRs was over nine months’ duration, and one death was deemed not be treatment related.
However, relapses are a key consideration, and it did not go unnoticed that two earlier responding patients – one CR and one PR – relapsed at three months. Relapses have plagued CD19-directed CAR-T therapies, with Novartis’s CTL019, for instance, seeing an initial 93% complete remission rate fall to 55% after a year.
However, that was in childhood leukaemia. Lymphoma patients have tended to yield less impressive initial responses that seem nevertheless to be more durable.
Guesses at efficacy
So how strong does efficacy need to be? Kite’s often-cited baseline is the Scholar-1 meta-analysis of 635 chemorefractory DLBCL patients, showing a 26% overall response rate, including an 8% rate of CRs.
Goldman Sachs analysts set a low benchmark for Zuma-1, saying that just doubling the efficacy seen in Scholar-1 would be enough for approval. Back in May Mizuho was expecting 70% ORR and 50% CR, but has since rowed back, now expecting a 30-40% CR rate, in line with RBC Capital and Sun Trust.
Then there is safety; Juno’s JCAR015 was recently set back by several months after a short clinical trial halt that was put down to its use of a fludarabine-containing lymphodepleting regimen. Zuma-1 also uses Cy/Flu preconditioning, and the KTE-C19 construct shares some key similarities with JCAR015 (Neurotoxicity only short-term toxic to Juno stock, July 13, 2016).
As well as being vague, it is not even clear whether any of the sellside’s efficacy guesses relate to interim three-month data or the much tougher 12-month readout. With such lack of clarity, one of the benefits of not being first to market is that Juno and Novartis will be able to follow in Kite’s slipstream.
Right now, whether the FDA shares Kite’s view that three-month data on 51 patients is enough for approval is anyone’s guess.
|Kite's Zuma-1 study (NCT02348216) of KTE-C19|
|Stage||Single-arm design||Primary outcome measure||Status|
|Phase I||7 refractory DLBCL/PMBCL/TFL pts||30-day dose-limiting toxicities||3 ongoing CRs in 6 evaluable pts.|
|Phase II||Cohort 1: 72 refractory DLBCL pts||12-mth overall response rate||Enrolment completed July 7; 3-mth interim data due on first 51 pts.|
|Phase II||Cohort 2: 40 refractory PMBCL/TFL pts||12-mth overall response rate||Enrolling.|
|Notes: DLBCL=diffuse large B-cell lymphoma; PMBCL=primary mediastinal large B-cell lymphoma; TFL=transformed follicular lymphoma.|
This story was amended to correct the one-year remission rate for CTL019.