Read on Twitter
Interesting Merck valued VelosBio's ROR1-based antibody technology at 2.75 billion with only preclinical data available. VelosBio's VLS-101 is an antibody carrying a spindle poison payload, which is toxic to all cells. 101 achieves cancer specific toxicity by targeting ROR1 (1/n)
ROR1 is a Wnt pathway receptor that is only expressed during embryonic development in normal pathology, but can be 'reawakened' in certain cancers. Thus, it is a cancer specific target. The efficacy of VLS-101 likely correlates with ROR1 expression levels in cancer. (2/n)
Thus 101's valuation likely reflects how Merck views ROR1 as a target, both in terms of cancer specificity as well as how much ROR1 is on cancer cells. VelosBio's gameplan was to test this drug in hematological malignancies, breast, and lung cancer, given these were likely (3/n)
the highest expressing ROR1 subtypes (I assume). Was interested in comparing ROR1 expression levels to CD47 expression levels in these cancer subtypes by mining RNA-seq data from DepMap. Results are attached. (4/n)
You can tell from the graphs that CD47 has clearly higher RNA expression levels across these cancer subtypes relative to ROR1 in cell lines that have been cultured from patient tumors. Some big caveats are that these are cells that have been in culture as opposed to the (5/n)
physiological setting, as well as that RNA levels do not necessarily reflect protein levels (although typically do). While CD47 is more ubiquitous than ROR1, it appears to be expressed at a higher volume on tumors. Although CD47 interacting drugs might face an antigen sink (6/n)
that ROR1 interacting drugs do not, there is likely a lot more opportunity for tumor infiltration w/ the CD47 class. Drugs like $TRIL 621/622 which limit off-target CD47 binding and deliver a cytotoxic signal could very well be more efficacious across a wider range of (7/n)
cancers than VLS-101. The 2.75b sticker price for VelosBio I think gives an interesting benchmark for how these big pharmas are valuing cancer-targeting agents where the value is tied to the target. I'd have to imagine Merck would happily buy $TRIL for 2.75b at this point...
