Sunesis Pharmaceuticals — Update 13 December 2016

Sunesis released expanded results from the ongoing Phase Ia study of SNS-062 for the treatment of chronic lymphocytic leukemia (CLL) at the American Society of Hematology (ASH) meeting in December 2016. SNS-062 is an inhibitor of Bruton’s tyrosine kinase (BTK), a protein important for the activation and proliferation of B-cells. This mechanism has been validated by the success of Imbruvica (ibrutinib, Janssen/AbbVie), which had sales of approximately $850m in Q316. Unlike Imbruvica, SNS-062 does not form a covalent bond with the protein and is active in patients with the C481S mutation that confers resistance to covalent inhibitors.

The company previously reported the results from the dose ranging portion of the study in healthy volunteers (see our previous report), which reported on the safety and anti-BTK activity in 32 heathy volunteers receiving either SNS-062 or placebo. In short, these results showed 85% or greater inhibition of BTK activity at all doses (50mg to 300mg) for at least 12 hours following administration. 33% of patients who received active drug had an adverse event, compared to 38% in the placebo arm. The adverse events observed in the trial were all mild (grade 1) except for a single patient on the 300mg arm who reported grade 2 fatigue and headache.

The current results expanded on this data set by examining the effect of food as well and drug interactions. In the food-effect study, 24 volunteers were given 200mg of SNS-062 in either a fed or fasted state. Food delayed the absorption of SNS-062 by approximately two hours, although the total abortion (as measured by the area under the curve, AUC) and elimination of the drug were unchanged (Exhibit 1). By comparison, Imbruvica shows a two- to four-fold increase in peak concentration and a two-fold increase in AUC when given with food. These results suggest that SNS-062 may be able to provide more consistent inhibition of BTK and therefore avoid the over- and under-dosing that Imbruvica is susceptible to.

The effect of the Cytochrome p450 3A4 (CYP3A4) inhibitor itraconazole was also measured. A large number of drugs are metabolized in the body through CYP3A4 dependent pathways, and inhibitors of this enzyme can significantly increase the effect of certain drugs. Moreover, a large number of pharmaceuticals as well as some naturally occurring molecules in food are inhibitors of this enzyme. The drug interaction study investigated 24 volunteers who received 25mg of SNS-062 in combination with itraconazole and found that the peak concentration of SNS-062 was increased approximately two-fold and the total exposure (AUC) was increased seven-fold. This indicates that the drug is indeed degraded by CYP3A4, and care should be taken by patients to avoid these interactions. We do not see this necessarily as a limitation for the success of this program considering that this is a relatively common finding, especially among drugs used to treat cancer, including Imbruvica.

The adverse event profile observed in these newest arms to the study were largely similar to what has been previously observed. Of note is that although a single case of asymptomatic supraventricular tachycardia was seen in the highest dose from the previous results, this event has not been subsequently observed.

The importance of developing therapies to address the emergent problem of ibrutinib resistance was highlighted at the ASH meeting. The role of C481 mutations in progressive CLL has been established, but there were a number of reports linking this mechanism to other hematologic malignancies such as diffuse large B cell lymphoma and Waldenstrom’s macroglobulinemia. Another poster at the conference presented the results of the follow-up of 308 patients who were enrolled in clinical studies of Imbruvica in CLL.1 Of the 136 patients evaluable at the time of the report (40.5 months median follow-up), 55 had progressive CLL. 36 of the 46 patients (76%) evaluable for DNA sequencing had C481 mutations.

The company also released an updated survival analysis of patients from the VALOR study of Qinprezo (vosaroxin) for the treatment of acute myeloid leukemia (AML). The company reported the overall survival (OS) of patients 60 years and older treated with Qinprezo and cytarabine has maintained superiority over placebo and cytarabine with a median follow-up time of 39.9 months (Exhibit 2).

Importantly the company also released an analysis of the effect of subsequent transplantation on the survival of patients over 60 (Exhibit 3). The analysis found that Qinprezo improved survival regardless of whether a patient was followed up with hematopoietic cell transplantation (HCT). This analysis is important because the approval of Vidaza and Dacogen in Europe depended on an analysis of the effect of subsequent therapy in both cases, because subsequent treatments such as HCT often reduced the separation between drug and placebo. Sunesis did not release a complete statistical analysis of the data, but we find the fact that a treatment effect is apparent both with and without transplantation as supportive.

The most prominent near-term catalyst for Sunesis is the completion of the EMA application process. The company is scheduled to receive the Day 180 set of outstanding questions by the end of 2016, at which point the company will have up to three months to respond, followed shortly by a CHMP opinion. This puts the approval decision for Qinprezo around the end of Q117. Another catalyst is the initiation of the Phase Ib/II study of SNS-062, which we expect in early 2017 following IND approval. And finally, we expect a readout from the TAK-580 Phase Ib dose ranging study being performed by Takeda in 2017. TAKI-580 is being investigated in combination with an array of other agents (including nivolumab, paclitaxel, irinotecan, etc) for the treatment of solid tumors. Data from this program was initially expected sooner in 2016, but due to delays given the manifold nature of the study, with many combinations and indications, dosing has taken longer than expected.

We have increased our valuation to $217m from $194m, although the value per share has dropped to $10.40 per basic share ($8.36 diluted) from $13.36 ($10.72) as a result of the October 2016 offering of shares ($26m for 5.7m common stock and preferred stock convertible into 1.6m shares). The increase in valuation is attributable to an increase in net cash (from $18.8 to $35.8m including the value of the offering) and advancing our NPVs to the most recent quarter. We are not adjusting any of our sales or risk assumptions at this point.

Sunesis reported an operational loss of $8.5m for Q316, with the majority of the expense associated with R&D ($5.3m), although this is a reduction from previous quarters ($6.2m in Q1 and $6.6 in Q2). We have slightly increased our 2016 spending estimates by $1.2m to $39.5m because SG&A cost control was lower than expected for Q3 ($3.9m). The company ended the quarter with $24.3m in cash and investments, offset by $14.4m in debt, which was subsequently supplemented with the $26m offering in October. The company has stated that the offering will go towards regulatory filings for Qinprezo, development of SNS-062, and general purposes. We have adjusted our future financing schedule in light of the recent offering and reduced the funding requirement from $95m to $65m: $10m in 2017, $20m in 2018, and $35m in 2020, in addition to the $87.5m in approval and sales milestones we model ($37.5m for TAK-580 approval, $50m in Qinprezo licensing payments and milestones).