CRIS are a bunch of nincompoops and just did what their LI suggested (see comments from R&D Day, below). Further, there were other patients in this current trial on higher dosing that did not experience any such issue or a DLT. Clearly, IMO, the LI wasnt bright to enroll a patient with liver impairment when in the previous Ph1, there were 2 incidence of liver enzyme issues at 450-600 mg. I am not even sure they need to dose this differently, it may be more stringent enrollment requirements.
So in summary, GDC-0917 had a favorable safety profile. We didn't actually determine the maximum tolerated dose, had, what we consider, a reliable PK profile. We stopped at a dose of 600 milligrams once daily for 14 days, every 21 days, based on the dose that portended the effective concentration for 90% inhibition. We saw, as Ali showed, that you can actually down-regulate cIAP in tumor biopsies, and we saw encouraging antitumor activity. Now this agent then subsequently was in-licensed by Curis. And so, we had an opportunity, like, say -- to say, okay, what needs to be done to explore this and try to find and determine what is the optimal dose, what is the optimal schedule, what are the optimal patient populations? So the first point. In the studies that we did in-house in start, we tried to determine, with the help of Curis, whether indeed there's dose-dependent antitumor activity in preclinical model.
Since the elimination of half-life range between 4 and 8 hours on averaging about 6 hours, the question arose: Can you deliver the drug twice daily and hence, improve the antitumor effectiveness? Is it possible to give the drug continuously? The 2 weeks on, 1 week off, dose would base somewhat empirically, and so one wanted to see, can you give the drug safely on a continuous daily dosing? And finally, if you saw antitumor activity in ovarian cancer patients, can we actually try to determine what might separate those patients who respond versus those that don't?
So let me address first and foremost, the dose-dependent nature of CUDC-427. And this is the same cell line that's a triple negative breast cancer model. And this is actually performed independently in our lab, not in Genentech or at Curis. And as you can see, we examined the dose schedule. This is a dose that was actually recommended by -- for preclinical studies, by Genentech, 10 milligrams per kilogram every day for 28 days. And we examined both increase of dose 20 30 and also a BID schedule. And I'll bring your attention to the fact that the recommended dose from Genentech was actually 10 milligrams, which is the blue line. But we found that as you increase the dose, you were then more able to improve upon the antitumor activity.
If you compare BID to once-daily dosing in the animal model, it's not quite so apparent because, of course, the overlapping error bars that are present. But what this really suggests, first and foremost is, that we can't be content with just a dose at the ED 90. We really need to explore how high we can go to try to get maximum antitumor activity out of this molecule.
So to address this, we've launched a Phase I study right now where we give the agent twice daily to patients with advanced malignancies, and the cycles are every 21 days. And so schematically, this is how this new schedule appears.
Now I've addressed dose, and I've addressed schedule. The most important question in targeted therapies is how do you identify the right patient? Because if you can get the right dose and right schedule, but you don't have the right patient, the whole system falls down.
AI
