An oncologist explains his dosing strategies for lenvatinib in advanced clear cell RCC, with reference to the design of the KEYNOTE-426 trial.
Nizar Tannir, MD: Just to follow up on that dosing, in KEYNOTE-426, which you’ll present to us later on, there was a provision for patients who tolerate the 5-mg twice-daily starting dosage of axitinib to dose escalate to 7 mg twice daily and even 10 mg twice daily. We didn’t see the CR [complete response] rate of 16% and overall response rate of 70%. Even though you’re saying it’s really high dose, 20 mg, do you think lenvatinib is more effective and more potent? Not just the dose, but because it’s a different TKI [tyrosine kinase inhibitor]. Maybe not related to the BFGF, not in first line, because we know that clear cell is mostly driven by VEGF initially, even poor-risk patients. What would you say to that?
When you also discuss the CheckMate 9ER [trial], where the dose was 40 mg, when Scott said this would be a regimen for a patient who might not tolerate a higher dose of TKI, we’re not going back and rewriting the CheckMate 9ER phase 3 trial, but do you think that choice of 40 mg instead of 60 mg is the key? They used cabozantinib with 60 mg with a PD-L1, atezolizumab, in other phase 3 trials that we won’t discuss. Is it 40 mg vs 60 mg for cabozantinib, and here we’re using 20 mg?
Moshe Ornstein, MD, MA: Great questions. I’ll start with the comparison to KEYNOTE-426 and the fact that that trial allowed for patients to have dose escalation from the starting dosage of 5 mg twice daily for axitinib. We haven’t seen data in terms of the dose escalation, and we haven’t seen the percentage of patients who’ve dose escalated in the trial.
What I can tell you from my experience and from talking to other RCC [renal cell carcinoma] oncologists and generalists in the community is that when a patient is on an I/O [immuno-oncology]–TKI and they’re tolerating it well, the temptation and the desire to dose escalate is minimal. If a patient is tolerating an I/O–TKI, the odds that the dose is going to be escalated are very low. The fact that we haven’t seen dose-escalation data from KEYNOTE-426, despite the fact that we’ve seen a lot of dose-escalation data for axitinib in general, is telling. It probably means that there weren’t that many patients who had dose escalation or that the efficacy after dose escalation didn’t alter the results.
This goes back to the other question that’s listed on the slide about the rationale for combining lenvatinib and pembrolizumab. The rationale for combining all I/Os and TKIs is similar in that there’s the element of the efficacy of the TKI itself. But what’s more important isn’t 1+1=2 or the individual efficacies of the TKI and the I/O agent. It’s what happens in the tumor microenvironment relatively early, when they’re given in combination. When they’re given in combination or even when a TKI is given as a monotherapy, there are shifts within the tumor microenvironment. There are lower levels of myeloid-derived suppressor cells. There are lower levels of regulatory T cells. The TKI creates a more immune-permissive tumor microenvironment and a more likely chance that there will be a response to the immunotherapy. There’s synergy there. It isn’t just an additive benefit.
Giving a higher dose of a TKI up front may lead to increased synergy. Can I prove it? No. But can we prove that one TKI vs another head-to-head in the frontline setting is superior? Also no. This is more likely driven by the dosing, but I don’t think it’s provable. But I certainly think the hypothesis is there when you think about how this synergy works with the I/O and the TKI.
Nizar Tannir, MD: Thanks, Moshe. Scott, you didn’t show us the Kaplan-Meier curves for OS [overall survival] for the CLEAR study. We all know there was an early separation of the Kaplan-Meier curves for the lenvatinib-pembrolizumab vs sunitinib. But then the curves crossed over past the 2-year mark. What’s your take on that? How do you interpret that? Does that mean anything? Do you take into account that the 2 curves crossed?
Scott Tykodi, MD, PhD: In any Kaplan-Meier analysis, the right-hand side of the curve is subject to censoring and is based on a small number of events, so it’s inherently unstable. At an early data reveal time point, I don’t make too much out of the shape of the curve after the median follow-up time point. But it’s a little uncomfortable, and it’s one of the big unknowns across the frontline regimens.
As you pointed out, there’s also this odd time desynchrony. Each of the studies came out at a different time, and the follow-up is different. The I/O–I/O regimen ipilimumab-nivolumab had a life before renal cell. The melanoma data are more mature than renal, and they have a wonderful flat tail on the survival in melanoma that doesn’t decay after about 3 years. That’s how the regimen is beginning to appear in renal cell, that there appears to be a lovely plateau phase.
When you add a TKI, and if the TKI is doing some of the heavy lifting in the combination regimens, is it going to have a TKI phenotype? Which isn’t plateau phase; it’s a decay. It’s sometimes measured in years. Disease control can be decent, but I have a figure saved of the Kaplan-Meier curve for sunitinib patients achieving CR. It’s a straight line heading for the bottom of the graph. If you wait long enough, you recur. There’s no plateau phase. If that’s the biology of the late success of I/O–TKI regimens, you’re going to eventually fall below and fail relative to the I/O–I/O regimen. That’s what the shape of the CLEAR curve calls to mind.
We don’t know at 3, 5, or 10 years where we’re going. If the lenvatinib does a lot of the work and makes the early phase look good, does it make the late phase look bad because you become resistant and you fail? We need more follow-up. There’s no answer to that question. In a couple of years, we’ll have a much better sense of the behavior of these different regimens. The question is, will we care? Will there be a new frontline standard of care, such as a triplet, that we’re all using, and we won’t pay any attention to the late follow-up of the doublets because we’ve stopped using them? That’s going to be a fascinating issue for the field.
Nizar Tannir, MD: I agree with you on that note. We’ll have a chance to discuss the triplets. As you know, there has been a recent press release, and we can discuss that.
Transcript edited for clarity.