NLNK : NASDAQ : US$26.94
NewLink Genetics is a biotechnology company devoted to
the development of cell-based cancer vaccines and other
All amounts in US$ unless otherwise noted.
Life Sciences — Biotechnology
NICE ODDS FOR IMPRESS, IDO IN THE MIX; RAISING PRICE Target to $35
We are more optimistic
that -the P3
IMPRESS trial of HyperAcute Pancreas (HAP) in resected pancreatic cancer
could meet its primary endpoint of an improvement in overall survival vs.
placebo based on the delay in the first interim analysis suggesting a
beneficial effect on survival (median OS >25 months) and per the P2 data.
Yet, we view the likelihood that the trial will be stopped on the first interim
look in Q1+ as small given the high bar (45% improvement in OS) – but note
considerable upside potential and the lack of a futility analysis as reasons to
be involved in the story. In addition, the IDO inhibitors, indoximod and
NLG919 (possible partnerships) provide another leg to valuation, with P1/2
data in Q4-Q1/15.
Why we like the Phase 3 HAP trial: Although the first interim analysis
(triggered by the 222nd death) was initially projected to occur in H1/13, it is
now projected in Q1/14+, with management noting in Nov. that the pooled
OS was >25 months (vs. 21 months initially projected). They noted several
factors that were hypothesized to cause the delay in events: 1) healthier
population (unlikely given prognostic factors are in line with prior trials:
e.g. 70% nodal status N1 and 9% elevated CA19-9, vs. 68% and 9% in
comparator RTOG-9704); 2) new salvage treatments (Abraxane
demonstrates 1.8mo OS benefit in first-line pancreatic cancer treatment –
cannot explain delta to 21 months particularly given smaller benefit
expected in resected patients); 3) enrollment issues (trial completed
enrollment in Sep.); 4) improvements in surgical techniques and
neoadjuvent therapy (allows more advanced/severe patients to be resected,
which decreases overall OS); and 5) issues with reporting patient deaths
(NLNK has hired investigators to determine patient status).
Net-net, we agree with management that the best explanation for the delay in the
222nd event is a treatment effect and we look to the first analysis in Q1+,
with a potential subsequent analysis ~9 months post.