The Regorafenib Dose Optimization Study

Transcript:

Tanios Bekaii-Saab, MD, FACP: Both CORRECT and CONCUR, the 2 phase III studies, suggested that regorafenib was an active agent. However, there are toxicities to this agent, which limit its use. At the time, where we thought about ReDOS, if you asked a number of practitioners who took care of patients, “What dose do you start with?” some would tell you 120 mg. Some would tell you 80 mg. Others would tell you that they treated every other week, or every other day. Everyone had their favorite regimen. There was no detail.

So, we thought we had to go back and interrogate whether there was a dosing strategy that we could elicit, that would certainly be an alternative to starting at 160 mg. That’s how we came up with the design of ReDOS. ReDOS had 2 arms. ReDOS stands for “Regorafenib Dose Optimization Study.” So, we wanted to really optimize the dose strategy.

Arm A of the study was an arm where patients started at 80 mg, came back after a week, and got evaluated. If no significant toxicities related to regorafenib were found, the patients went from 80 mg to 120 mg. Then, they came back the week after (before, again, week 3), and were evaluated. If they had no significant dose-limiting toxicity from regorafenib, they went up by an extra 40 mg to 160 mg. So, it’s a strategy to optimize the dose in the first cycle from 80 mg to 120 mg, to 160 mg. Then, you give a week off. And then, in cycle 2, you start with whatever dose was optimal for the patient in cycle 1.

Arm B was just the standard 160-mg dose. It was given as 3 weeks on, 1 week off—per the label, and as per CORRECT. Patients were essentially randomized to the 2 arms. In fact, the randomization was a little different. It was 1:1:1, because we had the second randomization that essentially was designed to help us understand whether preemptive Clobetasol (steroid cream), versus reactive, would improve the risk of hand-foot syndrome reaction.

The study had a statistical design that was powered to show a difference between the 2 arms, superiority for arm A versus arm B, and what we decided to be a primary endpoint (a mix between efficacy and toxicity measures). This was essentially a primary endpoint where patients would go through 2 cycles of treatment. They would make it through the 2 cycles and get to the scan with the intent to start with cycle 3. The percent of patients who were able to get to that point on arm A should have been at least 15% superior to arm B to declare the study positive. The study did reach its primary endpoint. It was a positive study. More patients on arm A, a little bit less than double the number of patients, were able to achieve that goal, versus those that were on arm B. So, the study was positive for the dose escalation strategy—from 80 mg to 120 mg, to 160 mg. So, 160 mg was the goal.

In addition to the primary endpoint, we had secondary endpoints of overall survival and progression-free survival. We also had quality of life measures. What we’ve observed is that overall survival was superior in arm A versus arm B—it was 9-months on arm A versus 6-months on arm B. Now, this was not statistically significant. But, if you look at the curves, they really separate early and continue all the way beyond the years. There is a nice separation between them. So, there’s a drug or exposure effect, we think. There may be other factors, follow-up therapy, etcetera, that we’re still digging into. But, interestingly, if you look at the 160-mg dose (arm B), and you overlap it over the CORRECT arm (the same dose), they literally hug each other very closely. Arm A, which was the dose escalation dose, historically stands on top of both.

So, our control population (the 160-mg standard dose), did as well as we would expect it to do. On the other hand, I think that arm A performed better. When we looked at progression-free survival, it was a little bit better with arm A—2.5 versus 2 months. But, it was not statistically significant.

Interestingly, when you look at quality of life—a higher number is better—patients on arm A did not lose any benefit. Their quality of life was flat. They did not suffer from any deterioration.

On the other hand, when you look at arm B, at 2 weeks, you see a drop in their quality of life. Parameters in their quality of life suffered, pulled it down, and then recouped to improvement after dose adjustments, we assume, and other factors, as well. So, all of the primary and secondary endpoints seem to mostly align, suggesting that a dose escalating strategy is at least as good, perhaps better, than starting with 160 mg oral, every day.

In fact, we asked our biostatisticians if there was any chance that arm B, the 160-mg arm, could be any better than arm A, just by pure chance, alone. The answer was that there was a 0% chance that the 160-mg dose would be better than the 80-mg to 120-mg, to 160-mg regimen. This essentially tells you that we have an arm that is at least as good, and probably better. In my viewpoint, this defines a new option, a new standard, for most of our patients. In fact, in my own clinic, the strategy will be to start with 80 mg, 120 mg, 160 mg, on a weekly basis, to try to achieve the optimal dose for every patient, in the first cycle. And then, to go on to the second cycle, and from there, move on.

Transcript Edited for Clarity