Crispr Therapeutics AG (CRSP)

[jondoeuk]

This will be an oral presentation (of preclinical data) at AACR next month.

Using AI and machine learning, they did a very simple set of experiments in a robust fashion, which is to take thousands of different edits without any bias and to do them pairwise to see what was going to increase the potency of the CAR-T cells. One of those edits is TGFBRII knockout and the other one is a Regnase-1 knockout. Not only did the latter come out of a large empirical screen in one of the partnerships the company had, but other groups have shown each can improve antitumour activity https://insight.jci.org/articles/view/133977 https://www.nature.com/articles/s41586-019-1821-z

So that combination of edits in preclinical testing, is at least tenfold better, if not more. Also, in their manufacturing process, the expansion of these cells is a lot greater, and the CEO has said they have seen a greater proportion of the central memory genotype phenotype https://ashpublications.org/blood/article/127/24/2980/35453/Phase-1-studies-of-central-memory-derived-CD19-CAR

When they start dosing patients, will know quickly what the PK/PD profile looks like for CTX112 and CTX131. If the expansion at DL1, let's say, is tenfold higher for these CAR-T's, the therapy should be more potent and work better.

Using CTX131 in R/R ccRCC, the first thing will be those expansion kinetics and the second will be responses. Assuming those responses are at low doses than with CTX130 it would inform them as to how they will think about investments, because if they start seeing those signs, will be going to go pedal to the metal on other CAR constructs and targets, with the off-the-shelf platform, especially now they have their own manufacturing.