Time to ditch the rulebooks?

Last week Synta/GSK announced a deal that saw GSK pick up rights to STA4783 (release), with a headline figure of $1.1bn for the deal. The structure of STA4783 can only be described as unusually nasty - in these times of molecularly targeted therapeutics, SBD, high-throughput structural biology etc. this compound looks like a real throwback.

How do you envision going about deciding to make such compounds? The Synta presentation even advertises the fact that this compound came from a “Synta chemical compound library”. Thiohydrazides? Sheesh! But then again: “That is best which works best”. The mechanism of action is apparently all down to the increased generation of reactive oxygen species (ROS), which causes cytochrome C release and activation of Caspase-9, leading to apoptosis. Apparently, many cancer cells operate under conditions of increased stress due to higher instrinsic levels of ROS and are thus more sensitive to an increase in these species. I am at a bit of a loss in understanding how the increase in ROS happens - is it due to the molecule itself or one of its metabolites? Perhaps it (or one of the metabolites) inhibits superoxide dismutase?  The two principal metabolites detected in vivo are the desulfurised versions, and the product of hydrazide hydrolysis (although exactly whether this is the hydrazide or thiohydrazide is not specified in the sources that I have looked at).

Whilst not effective in animal models as a single agent, they noted significant synergy with other chemotherapeutics - hence the megabucks deal. I imagine that GSK see significant potential in this molecule - combining it with paclitaxel (as they are doing with metastatic melanoma initially) will open up a large number of the most commonly occuring tumour types to this combination therapy.