I am amazed they did not see it in animal model – Sutent for example got the management extremely worried at one time because the QTc prolongation showed up in animals (together with the adrenal necrosis and other stuff – and the drug has also an accumulation problem and a non-linear PK with a long-lived main active metabolite). Fortunately they were careful (and lucky) with dosing in the clinic.
Failed to meet pharmacokinetic objectives? Presumably their prediction of human PK was off target and it would be really interesting to know by how much. Interesting that the QTc prolongation didn’t show up before (this is reminding me of Derek’s quiz). Presumably the compound is clean in the hERG assays or was option C indeed an option? As Jarza asks, how many patients?
Something can (must) be problematic in their Phase-II clinical trial protocols. A big lesson learned for the antitumor filed (particularly for Merck) although, there, toxicity is a bit less sensitive in the trial. Marginal dose should be very carefully considered.
Wavefunction looking at the Vertex WO 2004/000833 patent there’s only a few compounds without the dibasic piperazine-esque moity in their preferred list. In the crystal structure with the Abl-mutant this appears to be just a solubilizing group pointing out into solvent.
Favourable PHA-739358 phase I data was presented recently; it will be interesting to see how this compound proceeds.
One has to be extra careful about the basicity of the side-chains. Oned should try to use amines with lower pKA whenever possible. Putting strongly basic side-chains often improves the in vivo potency through the organ accumulation – not quite something you would want.