Frontline Data from Keynote 426 - Sandy Srinivas

May 25, 2019

Sandy Srinivas, MD discusses the frontline data from the KEYNOTE 426 trial comparing pembrolizumab plus axitinib to sunitinib monotherapy in clear cell kidney cancer. 800 patients were randomized and the combination demonstrated benefit across all three outcomes measured including response rates, PFS and overall survival. Dr. Srinivas believes the tolerability for pembrolizumab plus axitinib is more favorable than nivolumab plus ipilimumab although the nivo/ipi combination does provide CRs in the 20-29% range. Dr. Srinivas and Dr. Pal also discuss the JAVELIN 101 dataset which is similar in design to KEYNOTE 426 except that the combination is axitinib plus avelumab versus sunitinib. While the data favors the combination arm compared to sunitinib alone, the PFS was around 13 months which was lower than the 15.1 months observed in KEYNOTE 426. Dr. Srinivas advises caution when comparing the data across trials since these were not designed head to head. The JAVELIN 101 trial also has not met the OS endpoint as of date.


Sandy Srinivas, MD, Professor of Medicine, Medical oncologist, Urologic Specialist, Genitourinary Specialist, Stanford University Medical Center, Palo Alto, California

Sumanta (Monty) Kumar Pal, MD, is an internationally recognized leader in the area of genitourinary cancers, including kidney, bladder, and prostate cancer. He is the Co-director of City of Hope's Kidney Cancer Program and is the head of the kidney and bladder cancer disease. Dr. Pal sits on the Editorial Board for clinical genitourinary cancer and is a reviewer for multiple journals including The Journal of Clinical Oncology, The Journal of Urology, European Urology, and many others.

Read the Full Video Transcript

Monty Pal: Welcome to UroToday. My name is Dr. Monty Pal, and I'm a Medical Oncologist at the City of Hope Comprehensive Cancer Center. I'm delighted today to have Dr. Sandy Srinivas, who is a dear friend and colleague, and Professor at the Stanford University School of Medicine in their Department of Medical Oncology. She's world renowned in the area of kidney cancer and, today, Sandy, welcome to the program. 

Sandy Srinivas: Thank you, Monty. It's a pleasure to be here. 

Monty Pal: Thanks. I'm going to be probing you a little bit about some of the frontline data that we're seeing from ASCO GU 2019. I would say one of the highlights of this meeting, really across all the different domains, is the KEYNOTE-426 dataset, which pits axitinib with pembrolizumab against sunitinib.

Sandy Srinivas: Sunitinib

Monty Pal: Any thoughts on that? 

Sandy Srinivas: Well, it's really exciting. I think first line compared to where we were even two years ago, it's rapidly changing. KEYNOTE-426 randomized about 800 and so patients to clear cell kidney cancer patients to either, as you said, pembrolizumab plus axitinib versus sunitinib, and they've had benefit across all three outcomes, response rates, PFS, and overall survival superior to sunitinib. 

So, it becomes a little challenging now. What do we do for our frontline patients with this positive trial? Up 'til now, at least in my mind, I've been thinking that for patients with intermediate and poor-risk patients we would be using ipi-nivo. Certainly, I think, compared to ipi-nivo, I'd be looking forward to seeing what the discontinuation rates are. But, my sense is that the toxicity profile is definitely going to be more tolerable compared to ipi-nivo. 

But, I would be curious. I'm always interested in looking at what the complete response rates are, so one of the attractive things about ipi-nivo has been the CRs. I think I'm looking forward to seeing an abstract from McDermott looking at what the treatment-free interval would be. I think for ipi-nivo it's almost about 20 to 29% of patients who get a good response are able to stay off any additional therapy. So, I'm looking forward to whether the same thing would hold true with the VEGF and IO combinations.

But what this does is, I think, it opens the door to all patients. I mean should every patient with kidney cancer, metastatic clear cell, get treated with immunotherapy? I think the tolerability of pembrolizumab with axitinib makes it a little bit more attractive, and perhaps this would be applicable to the good-risk patients as well.

Monty Pal: I think that's a beautiful distillation of the dataset.

The other dataset that's been sort of hanging over us for a couple of months now is the JAVELIN 101 dataset. This axitinib and avelumab versus sunitinib, and the two trial design is so similar, right, JAVELIN 101 and KEYNOTE-426. Can you help us decipher the differences? Clearly, the result is different a little bit as well.

Sandy Srinivas: Right. Well, you know, I mean the whole concept about PD-1 and PD-L1 inhibitors we haven't really sorted that out. In our mind, we view them about the same, at least in other cancers. In bladder cancer, there are five of them that are approved, and it's been hard to distinguish between efficacy and adverse events. But, JAVELIN is a PD-L1 antibody with axitinib versus sunitinib, and they had a readout with a pretty good response rate, which was double that of sunitinib. The PFS was also around 13 months compared to KEYNOTE at 15.1 months. 

It's hard to compare across trials, but I think they don't have overall survival yet. It might be a little premature, but certainly axi-avelumab is an option for patients, and we just have to wait to see what the long-term survival turns out to be.

Monty Pal: I guess that makes sense, yeah, so two very similar options.

But, let's just go with the hypothetical scenario. You have, at this moment, axi-avelumab. You have also at your disposal axi-pembrolizumab. What are you going to be picking for your patients between the two?

Sandy Srinivas: I think if you think about level 1 evidence, always it helps to have all three. Having a high-response rate, having a PFS that's around 15 months, and having a hazard ratio of 0.5 for overall survival was pretty impressive for KEYNOTE, so I that's almost ... If I go back to clinic, that's definitely something that's going to be high in my mind. 

I think ipi-nivo would still hold true, again for the reasons that I said, that our bar, as kidney cancer doctors, I think, have gone up much higher. No more are we just looking for some minor tumor shrinkage. We want that home run for a complete response. The bar with ipi-nivo was around 9%, but JAVELIN, I think, was around five. I'd definitely be interested in looking to see what the KEYNOTE-426 data holds up.

Monty Pal: I agree with you, a very important bit of data. Absolutely. 

But, let me paint a clinical picture that I've sort of been struggling with in this decision-making process. This goes back to the question nivo-ipi versus a VEGF-IO combination. You have a patient who comes to your clinic with de novo metastatic disease, primary intact, multiple metastases to the lungs. In that specific scenario, would you consider nivo and ipi upfront? Would you consider the combination of VEGF and IO with some of the surgical considerations we have with that modality?

Sandy Srinivas: Yeah. I think prior to this year’s meeting, I used to use this whole risk model for intermediate and poor risk and favorable risk. The updated ipi-nivo, it looks like the favorable is becoming less significant, even though the hazard ratio is still greater than one. 

Maybe with time, the favorable will be different, but I'm going to look at the patient in front of me, see what his overall performance status is. You always worry about the ipi and ability of patients to tolerate the side effects of ipi. So, if I find somebody somewhat marginal, I think going with the VEGF plus an IO is definitely more attractive. At least we perceive that to be less toxic compared to ipi-nivo.

In the future, maybe there might be things that might help us, like are there some groups of patients who are more VEGF-dependent than others? Some of those might help us, but for now, I think you're left with thinking about the patient's functional status and what you think their ability to tolerate ipi would be.

Monty Pal: This is great. This is really helping me as I think about how to apply all this data in the clinic come Monday. Thanks a lot for that, Dr. Srinivas, and thank you very much to the UroToday audience for tuning in. We'll have more of Dr. Srinivas later. Thank you.

Sandy Srinivas: Thank you, Monty.