Roche doubles down on Poseida

After another stumble in TIGIT, Roche is already looking towards the next thing: allogeneic Car-T. The company has spent $1bn up front on Poseida, building on a licensing deal signed in 2022.

One question is what Roche wants that it doesn’t already have from the original tie-up. The Swiss group highlighted Poseida’s manufacturing capability, as well as one Car-T in particular that was outside the original deal, P-MUC1C-ALLO1. Roche is also getting two early-stage non-viral gene therapy projects in hereditary angioedema (HAE) and haemophilia, which might have been a draw, as the big pharma already has a strong presence in the latter.

However, there are big questions around allogeneic Car-T – not least whether therapy can produce durable responses.

Allo allo

Poseida’s most advanced project is the BCMA-targeting P-BCMA-ALLO1; this was named in the $110m 2022 tie-up, along with the bispecific P-CD19CD20-ALLO1.

But the jury is still out on P-BCMA-ALLO1. Data at last year’s ASH meeting looked cherrypicked, while a more recent update, at the International Myeloma Society meeting in September, found an ORR of 54% across the intent-to-treat population. However, Poseida homed in on a cohort of patients who got what it termed the optimised lymphodepletion regimen, and here the ORR was 91% among 23 patients.

Either way, Poseida has fallen well behind; it had once hoped to get an autologous BMCA-targeting Car-T filed with regulators in 2020.

Meanwhile, the first clinical data on P-CD19CD20-ALLO1 are expected next year. This project is being tested initially in B-cell malignancies, although there are also plans for trials in multiple sclerosis and lupus, which might be more tempting for Roche.

As for the aforementioned solid tumour contender P-MUC1C-ALLO1, that has produced one partial response among four evaluable patients with various cancers. Since then Poseida announced plans to test higher lymphodepletion regimens in its phase 1 trial; an update is expected this year.

One project that might not go any further is P-BCMACD19-ALLO1, on which Roche turned down an option in September. Before Tuesday’s deal the big pharma still had a chance to opt in on a CD70-directed asset.

Poseida’s allogeneic Car-T pipeline

Source: OncologyPipeline & company presentation.

Poseida’s preclinical HAE and haemophilia assets, meanwhile, could be a bonus for Roche. The group already has a haemophilia A blockbuster in the form of Hemlibra, but a previous attempt to get into gene therapy, via the acquisition of Spark Therapeutics, has fallen flat.

Poseida’s hopefuls use non-viral lipid nanoparticle delivery, which could help avoid some of the problems of current gene therapies – although the assets still have a long way to go.

The deal terms include a non-tradeable contingent value right (CVR) of up to $4 per share. $2 will be payable on a pivotal trial start for P-BCMA-ALLO1 (by December 2028); $1 on the first sale of P-BCMA-ALLO1 (by December 2031); and $1 on an autoimmune pivotal trial start for P-CD19CD20-ALLO1 (by December 2034).

If all of these are hit the acquisition's value will creep up to $1.5bn. This is small change for Roche, but represents a huge premium for Poseida, whose market cap was just $279m on Monday. It will be a while before it becomes apparent whether this was a panic buy or a justified investment.