Biotech Values

[DewDiligence]

IMDX -19%—my take from today’s CC…

The planned phase-3 trial of CBM305 in a subtype of soft-tissue sarcoma, which triggered a 10% pop in the share price during yesterday’s AH session, will be agonizingly hard to pull off for several reasons:

• Only 0.125 of alpha is allocated to the PFS co-primary endpoint and 0.125 of alpha to the OS co-primary endpoint, so the total alpha spend is smaller than usual for a pivotal trial. (This makes sense insofar as there is no second pivotal trial in the indication in question; see #msg-135444482 for an explanation of the setting.)

• Officially, the trial has to hit both co-primary endpoints (PFS and OS) to be deemed a success, although management says the FDA may consider full approval based on only PFS if the numbers are really impressive. (Again, see #msg-135444482.)

• The modeling assumption for PFS in the placebo arm is a median value of 4-6 months, which is an oddly vague description (i.e. there’s a colossal difference between 4 months and 6 months for modeling purposes). In any case, the 4-6-month range seems rather short for a trial where all patients will have stable disease or an objective response following first-line chemo, which suggests that the observed PFS in the placebo arm may be longer than what is being modeled, hurting the comparison of the two arms.

• The PFS analysis will occur after 141 events, and management says the trial has 90% power to be statsig at a PFS HR=0.55. However, HR=0.55 for PFS is a tall order, especially where the treatment arm is an I-O agent that may require a few months to ramp up. (Management says they can hit the p<0.0125 threshold even with PFS HR as high as 0.68, but they don’t say how likely that is according to their model. It would seem to be rather unlikely based on the other revelations about the trial design.)

• Enrollment speed may be a problem insofar as there are only a few hundred patients in the US who currently meet the eligibility requirement; the trial is seeking to enroll 248 patients.

• Management acknowledges that the readout for the OS co-primary endpoint will take at least 5 years from the start of the trial, even with speedy enrollment.

• The dose of VBM305 to be tested in the phase-3 trial is 4x the dose used in phase-2. IMDZ will run a short safety trial before starting phase-3, but such an abbreviated test of the higher dose won’t rule out unexpected SAEs during the phase-3 trial itself.

Feedback welcome.