Expert Financial Analysis and Reporting

Overview

Kite released a press release with topline data on the interim look at the ZUMA-1 trial. This note interprets information presented in that press release and I would urge you to read it before going on.

Key Points:

• The CR rate at three months for 51 r/r DLBCL patients was 33%. Kite has said that KTE-C19 could be approved on this interim look with a 40% CR rate at six months so this result did not meet the 40% hurdle.
• I think that the CR rate at six months will decrease to 25% to 30%. Key opinion leaders feel that a CR of 25% at six months is clinically meaningful as CRs for other current treatment regimens are estimated at only 8%.
• I do not think that Kite will gain approval in mid-2017 based on this interim look as they have guided investors. I think that the FDA will want to see results from the full trial which is scheduled to report out in March 2017. I think that KTE-C19 will likely be approved in late 2017 or early 2018.
• With this interim look strategy, Kite was hoping to be the first CAR-T drug approved for r/r DLBCL. Novartis is running a very similar trial with its very similar product CTL019 with the same clinical guidelines and timelines. I believe that both drugs likely will be approved at the same time in late 2017 or early 2018.
• Results in the second cohort of ZUMA-1 which enrolled r/r PMBCL and TFL patients were very impressive as the CR rate was 64% in 11 patients treated. PMBCL has a limited prevalence of perhaps 600 or so patients and TFL is somewhat higher. The big market opportunity is r/r DLBCL.
• There were two deaths in the trial and a very high incidence of serious side effects, many of which required hospitalization. Kite did not provide clear data on this. My view is that Kite has not been transparent in presenting data on side effects and has focused investors on efficacy. The high incidence of serious side effects is the Achilles heel of CAR-T therapy.

Investment Opinion

Kite was very enthusiastic in a conference call that discussed these results and said that it plans to file a BLA on this data. It almost has to say this because it has set that expectation among investors. Analysts on the call were similarly enthusiastic. The initial result was that the stock traded up on the news. I seem to be in the distinct minority in having concerns that the data will not lead to approval in mid-2017 and that Kite and Novartis will likely gain approval at the same time.

I think that it is difficult for the FDA or investors to make a meaningful judgment on the risk to benefit ratio of KTE-C19 at this time. We do not know the CR rate at six months or nine months which is very important to assessing benefit. Kite is also being very opaque and minimizing the side effect profile of KTE-C19. I would not want to own the stock at this price.

The Most Important Number in the Data Set

Kite’s plan in designing ZUMA-1 was to file a BLA on the basis of the interim data from 51 patients treated in the r/r DLBCL arm of the trial. Kite has indicated in its presentations that it believed that a CR rate of 40% with a duration of response of 5 to 6 months would be sufficiently compelling to file for a BLA on this data by year end 2016 and to receive approval by mid-2017. In my report “A Preview of Highly Anticipated Results from Interim Look at ZUMA-1”, I projected that the CR rate at this interim look would be 25% to 35%. So what was it?

Three months after cell infusion the CR rate was 33% so that one out of three patients had a CR. Kite says this is sufficient to execute its plan and file the BLA on the interim data. I think this is at the very low end of the point at which the FDA might accept a BLA filing at this time. Here is the critical issue. It is the CR at six months or perhaps nine months, not three months, that matters and it is highly probable (almost certain) that some of the patients with CRs will progress. The question is how much. My guess is that the CR rate at six months will be 25% to 30% in these 51 r/r DLBCL patients.

We know that the CR rate at six months can be no better than 33%. Key opinion leaders I have listened to have said that the CR should be at least 25% to be medically meaningful. There is some reasonable chance that the CR could be close to or even less than 25% for these 51 patients at six months and nine months.

What Will The FDA Do?

Kite has set the expectation that it will file on this interim look and in its conference call, it said it would do so. The plan would be to perhaps supplement this data on 51 patients with additional data from the 21 other patients in the r/r DLBCL cohort. The full trial is scheduled to complete in 1Q, 2017 and I think that the FDA will want to see the full data set. The key question for the FDA is what the CR rate is at six or nine months. I give less than a 10% chance that the FDA will approve the drug based on this interim look. I think that given the uncertainty on the percentage of CRs and their duration that it will not rush an approval.

I think that if the CR rate is above 25% at the end of nine months after cell infusion or possibly after six months that the FDA will approve the drug for r/r DLBCL. The BLA for the full trial could be filed in mid-year and approval could come in late 2017 or early 2018. While this was an open label trial with no control group, historical data from the SCOLAR-1 meta-analysis suggests that the CR with current treatments is on the order of 8%.

The CAR-T approach represents a dramatic new approach to cancer therapy and even a 25% rate is clinically meaningful. More importantly, an approval would allow the medical community to experiment with different doses and combinations with other drugs and treatments. This is just the start of the CAR-T era and over the next two decades there is a high probability that responses will be significantly improved and side effects reduced.

Kite Doesn’t Want to Talk About Side Effects

In its press release Kite published on its website Kite did not even mention side effects which is more than curious, but in a separate business wire release it did release data on side effects. It said that across the combined 62 patients, the most common grade 3 or higher adverse events included neutropenia (66%), anemia (40 %), febrile neutropenia (29 %), thrombocytopenia (29 %), and encephalopathy (26 %).  Grade 3 or higher cytokine release syndrome (CRS) and neurological toxicity was observed in 18 % and 34 % of patients, respectively. Two patients died from KTE-C19 related adverse events (hemophagocytic lymphohistiocytosis and cardiac arrest in the setting of CRS).

I believe that Kite has purposely downplayed side effects in its interaction with the investment community and continued to do so in this instance. It reports serious side effects as grade 3 or higher which include grade 3, 4 and 5 in one number. Here is the problem with this number. Grade 3 side effects are serious side effect, grade 4 are serious side effects that require hospitalization and grade 5 is death. Kite should break each of these grades out individually. The percentage of grade 4 and 5 side effects will have an important effect on the benefit to risk assessment that determines how the product will be used if approved. Based on data I have seen from other trials, I would not be surprised if as many as 30% or more of patients had a grade 4 or higher side effect.

In all past releases, Kite has only reported on side effect due to cell infusion and not on the chemotherapy pre-conditioning program. The business wire release is the first time I have seen them report on side effects from the pre-conditioning regimen which are neutropenia, anemia, febrile neutropenia and thrombocytopenia. The encephalopathy, cytokine release syndrome and neurological toxicity are due to the cell infusion.

Other Points of Interest

• Kite released data on 11 patients enrolled in the second cohort that is enrolling 40 r/r PMBCL and TFL patients. This is much more impressive as the CR rate at three months in this patient group was 64%. PMBCL has a limited prevalence of perhaps 600 or so patients and TFL is somewhat higher. The big market opportunity is r/r DLBCL.
• Kite was able to manufacture KTE-C19 for 99 percent of patients enrolled in the study, and successfully handle the study logistics and adverse event management at over 20 sites, most of which had no prior experience in CAR-T therapy.”