Expert Financial Analysis and Reporting

Key Points:

• Novartis will almost certainly gain approval for CTL-019 in r/r pediatric ALL early in 4Q, 2017. This would represent the first ever approval for a CAR-T product.
• Kite says that it will receive approval for Axi-Cel ((axicabtagene ciloleucel)) in r/r DLBCL and launch in the US in 2017 and will have first mover advantage in this cancer space. I disagree that it will have meaningful first mover advantage.
• Novartis has said that it will file a BLA for CTL-019 (tisagenlecleucel-T) in the US for r/r DLBCL in 2017, but has been no more specific.
• It is possible that KITE could receive approval for Axi-Cel in r/r DLBCL by a few months prior to CTL-019 approval. However, it is also (probably more) possible that the FDA will want to carefully review the data on r/r DLBCL from both companies and approve them at the same time.
• In the case of r/r DLBCL, current data suggests that about one of three patients treated may receive meaningful benefit. The cost of cell infusion alone could be $300,000 or more but the additional costs of for the chemotherapy treatment used to reduce T-cell populations before infusion and costs required to treat side effects (often requiring hospitalization) could add $300,000 and result in all-in costs of $600,000.
• As of yet, there are no biomarkers to prospectively identify which patients might benefit. Managed care could be spending $1.8 million to treat three patients in order to achieve one positive outcome. This suggests to me that negotiating reimbursement and working with clinical sites to establish protocol for treatment will take time. This points to a slow launch. Most bullish analysts believe the opposite. They think that Axi-Cel represents such a dramatic breakthrough that it will be quickly and widely accepted for reimbursement.
• I don’t see KITE as having a meaningful first mover advantage in r/r DLBCL even if Axi-Cel is approved a few months before CTL-019 in r/r DLBCL. I think they will be fighting it out toe to toe with Novartis. This does not seem to be reflected in consensus Wall Street thinking.

Investment Thesis

This note builds on the report that I published on Kite following the release of six month data in the ZUMA-1 trial. Analysis of Six Month Data from ZUMA-1 That Led to 40% Price Increase. I would urge you to read this report carefully before going on in order to understand my detailed thinking on Kite. From an investment standpoint, Kite appears to be priced for perfection at a $4.3 billion market capitalization as investors appear to emphasize the positives of the story while ignoring the negatives. One of the strongly held beliefs of investors is that Kite is the first mover in the CAR-T; management also claims this to be the case. This note suggests the contrary.

I am struck that the bullish on the stock and Kite management rarely, if ever, mentions Novartis as a competitor even though it should be the first company to gain approval for its CAR-T product CTL-019 (tisagenlecleucel-T) in pediatric r/r ALL. A case can be made that Kite has a slight lead in r/r DLBCL over Novartis (a few months), but a good case can be made that Novartis’s CTL-019 and Kite’s Axi-Cel (axicabtagene ciloleucel ) will be approved at or nearly at the same time.

I continue to believe that Axi-Cel will be approved for r/r DLBCL, r/r PMBCL and TNF in late 2017 or early 2018. I think that it is a meaningful advance for the treatment of relapsed and refractory B-cell lymphomas and leukemias, but its role in solid and other tumor types as well as milder forms of B-cell lymphomas and leukemias is questionable. I believe that KITE is priced for perfection and both the Company and the analysts who support the stock have glossed over negative while over-emphasizing the positives. This note focuses on a case in point; i.e. that KITE does not have a significant first mover advantage over Novartis. At significantly lower prices, I would be positive on the stock.

Recent AACR Presentation by KITE

At a recent AACR presentation, KITE presented data that showed that the CR rate for Axi-Cel in r/r DLBCL remains unchanged at 31%, the same as was presented in late February. There was an improvement in the CR rate r/r PMBCL-TNF cohort, but as I explain later in this report, r/r DLBCL data is key.

Median progression free survival data was as follows:

• Median duration of response had not been reached for patients achieving a CR,
• It was 1.9 months for those achieving a PR,
• It was 8.2 months for patients achieving either a CR or PR.
• For the trial as a whole, mPFS was 5.9 months.

Kite emphasizes that mOS has not been reached and the lower bound of the confidence interval now rests at 10.7 months so that mOS should at least surpass this number. There is no control group in the trial, KITE did a large meta-analysis of other trials called SCHOLAR-1 that suggests that mOS in this patient group is 6.6 months. Generally, a 4.5 month in mOS for a new drug over standard of care is considered a major advance.

The Company did not present any data for r/r DLBCL alone. My guess is that both mOS and mPFS are less (perhaps meaningfully) in this group than in r/r DLBCL and TNF patients. In my opinion, the mPFS data and mOS data for r/r DLBCL patients is extremely important because DLBCL patients represent 85% of the addressable population and we know that they do less well than r/r PMBCL and TNF patients. KITE presented very detailed data on other subsets of the trial so it is almost certainly true that they have this data. This continues a disturbing pattern (in my opinion) in which KITE appears to present positive data and omit less positive data.

FDA Accepts Novartis BLA for CTL-019 in Pediatric r/r ALL

Novartis reported on March 29, that the FDA had accepted the BLA filing of CTL-019 in pediatric r/r ALL and granted priority review designation. This may potentially shorten the time for approval to within six months of the filing acceptance. Hence, CTL-019 could be approved before October 2017. Kite has been extremely active in presenting data on Axi-Cel to fire up investors while Novartis has been quiet. This is understandable as Kite needed to create excitement in order to raise enormous amounts of capital to fund the Company; since going public it has raised over $1 billion. Novartis obviously has no need for capital and has not provided much data. In its press release. Novartis provided some interesting data.

The filing in r/r pediatric ALL was based on results from 50 patients treated in the ELIANA study. This was a phase 2 study that enrolled patients at 25 centers in the US, EU, Canada, Australia and Japan. Data presented at ASH in December 2016 showed that 41 of 50 patients (82%) achieved complete remission or complete remission with incomplete blood count recovery at three months post CTL019 infusion; Novartis offered no update on this data in this press release. This was comparable to data from earlier US multicenter trial and an earlier single site trial led by the Children's Hospital of Philadelphia.

Novartis reported that 48% of patients in the ELIANA trial experienced grade 3 or 4 cytokine release syndrome (CRS). In the case of grade 3 side effects hospitalization is possible while with grade 4 side effects, hospitalization is very likely. There were no deaths due to CRS. Fifteen percent of patients experienced grade 3 neurological and psychiatric events including confusion, delirium, encephalopathy, agitation and seizure. No cerebral edema was reported and no grade 4 neurological and psychiatric events were observed. Remember that there were deaths due to cerebral edema reported in prior NCI trials and in the Juno trial of JCAR-015 in adult r/r ALL.

Acute lymphoblastic leukemia (ALL) makes up approximately 25% of cancer diagnoses among children under 15 years old and is the most common childhood cancer in the US. The annual incidence of ALL is about 4,300 in children under 15. The incidence of r/r ALL is about 900 patients and treatment options are limited.

Novartis said that it plans additional filings for CTL019 in the US and EU markets later this year, including a BLA with the FDA for treatment of adults with r/r diffuse large B-cell lymphoma (DLBCL) and applications for marketing authorization with the European Medicines Agency in r/r B-cell ALL and r/r DLBCL.

Kite Announces Submission of BLA for Axi-Cel

Kite announced on March 31 that it has completed the rolling submission of the BLA for Axi-Cel in r/r DLBCL. The FDA still has to review and accept the BLA which could take a month or two so that acceptance of the filing could occur in April or May. It is almost certain that if the BLA is accepted that it will be given priority review so that approval could potentially come in October or November of 2017. Kite says that it is preparing for approval and launch in 2017.

The Race for Approval and Commercialization in r/r DLBCL between Novartis and Kite

Kite is seeking approval in r/r DLBCL based on 71 patients enrolled in one arm of the phase 2 ZUMA-1 trial. A second arm enrolled 41 patients with r/r PMBCL and TFL. There are about 8,800 patients with r/r DLBCL and perhaps 900 with r/r PMBCL and 600 with TFL. Because of the much greater incidence of r/r DLBCL, I think that FDA will analyze the data from ZUMA-1 separately rather than comingling the results. This is important because the CR rate of Axi-Cel at six months is 31% for r/r DLBCL; 50% for r/r PMBCL and TFL; and 36% for the two groups combined.

Novartis has not provided much information on its JULIET phase 2 trial of CTL-019 in r/r DLBCL and how its structure and timelines compare to ZUMA-1. JULIET is scheduled to enroll 130 patients according to ClinTrials.gov. Novartis has given no indication of the split between r/r DLBCL patients and TFL/ r/r PMBCL patients.

JULIET started about five months after ZUMA-1 which suggests KITE may have only a small lead over Novartis. Novartis says that it will file a BLA in the US later this year but has not been specific. Based on this statement, Novartis could file as late as December 2017 in the very best case for Kite. Perhaps, the filing will lag Kite filing by five months which was the lag between Kite beginning enrollment in ZUMA-1 and Novartis enrolling in JULIET. Conceivably, it could be shorter than five months; without input from Novartis we just don’t know.

What Action Might the FDA TAKE?

FDA approvals are not just based on clinical trial data. The FDA also requires extensive information on quality assurance and manufacturing. In the case of CAR-T cells, it will also want extensive information on logistics. The product manufacturing process requires T-cells to be removed from the body and frozen at cryogenic temperatures and delivered to a manufacturing site. There, they are thawed and genetically engineered to express a chimeric antigen receptor. They are then frozen again, returned to the clinical site, thawed and infused into the patient. This is an extremely complex process. Note that a company called Cryoport is handling the logistics for both Kite and Novartis. There remains a not insignificant risk that questions on quality assurance, manufacturing and logistics could delay approval for Kite or Novartis or both.

The apparent 80% complete response rates in pediatric r/r ALL make this an easy decision for the FDA. Barring a potential problem as described in the previous paragraph, I give an extremely high probability of approval of CTL-019 in pediatric r/r ALL in early 4Q, 2017. The results in r/r DLBCL are not as impressive as Kite reported a 31% response rate in ZUMA-1 at six months. These results are not mature and could change over time. We don’t know that the results with CTL-019 might be.

I think that it is quite possible that FDA will want to consider the results in r/r DLBCL for Axi-Cel and CTL-019 at the same time in order to gain as much information as possible. Investors must remember that the FDA decision will have very significant impact on medical practice. However, it is looking at data from 200 to 230 patients in phase 2 trials without a control group. Information from the ZUMA-1 and Juliet trials may help inform on both products.