KEYNOTE-826: Immune Checkpoint Inhibition in Metastatic Cervical Cancer


Considerations for the KEYNOTE-826 trial, which tested pembrolizumab in combination with chemotherapy in patients with metastatic cervical cancer.


Bradley Monk, MD, FACOG, FACS: We want to do more. The survival, Krish, for the GOG 240 trial was about, 16.5, 17 months? Is that what you said?

Krishnansu S. Tewari, MD:The median was 17 months.

Bradley Monk, MD, FACOG, FACS: That’s not much, particularly because this is a young population. On September 18, a study was presented called KEYNOTE-826, published simultaneously in the New England Journal of Medicine, which is challenging. Premal, tell us about KEYNOTE-826 and what it did.

Premal H. Thaker, MD, MS:As Warner said, it’s ironic to be telling you about this paper because you’re the senior author on it. It’s a phase 3 randomized controlled trial that was looking at giving paclitaxel, platinum, bevacizumab, plus pembrolizumab vs chemotherapy bevacizumab and placebo, with maintenance cycles afterward for up to 35 cycles of pembrolizumab. About 617 patients were enrolled. It was 1:1, and there were 3 groups within this: looking at a CPS [combined positive score] score greater than 1, at the intent-to-treat patient population, and at a CPS score greater than 10. This had a dual primary of PFS [progression-free survival] and OS [overall survival]. On the first interim analysis it met all the end points. For the PFS, we had 10.4 vs 8.2 months in the placebo group improvement. That was a hazard ratio of 0.62, which is incredible. We have these results within the first 3 months, and you see those curves diverge. It’s nice for our patients that we get responses quickly for them as well. The OS hasn’t been accomplished yet, and quality of life was improved on the pembrolizumab group. This was important. We also saw that the adverse-event [AE] profile had no new events. There was nothing different that we would have anticipated in terms of immune-related AEs with the addition of pembrolizumab to the combination. We discussed how often Avastin [bevacizumab] is given, that only 63% of the patient population receives it. It sounds like our institution is very similar, that we probably give 80% to 85% of our patients Avastin, or at least that’s what we think we do.

Bradley Monk, MD, FACOG, FACS: We’re very proud, Krish, of your hard work with bevacizumab. There are places like France where bevacizumab isn’t reimbursed according to the GOG 240 trial. I don’t know if you knew that. There are situations and there are contraindications, as Bhavana said, and certainly fistula, GI [gastrointestinal] perforation, nonhealing wounds, proteinuria, uncontrolled hypertension, inflammatory bowel disease, a history of diverticulitis. Krish, because only 60% got bevacizumab, what was the performance of the control arm? Is this an example where the control arm underperformed, or did the control arm—even with 60% bevacizumab—look similar as what the GOG 240 trial did?

Krishnansu S. Tewari, MD:It looked slightly better. I’m not sure what the issues are with the uptake of bevacizumab in the study, but it was done in places where it’s not available regardless of whether it was reimbursed or not. What I can’t tease out from the data is if you need bevacizumab for the pembrolizumab to work, and I’ve read the paper.

Bradley Monk, MD, FACOG, FACS: You’re a coauthor.

Krishnansu S. Tewari, MD:I know. It’s interesting.

Bradley Monk, MD, FACOG, FACS: That’s what we’re here to talk about. I have 2 points, we’ll talk about this 1, and then we’ll move on to the second 1. Bhavana told you when to use bevacizumab. This trial is not about bevacizumab. It’s about pembrolizumab, and you should use bevacizumab when it’s right to use it, and pembrolizumab shouldn’t confuse that decision. You can look at an unplanned subset, make a definitive conclusion—please don’t do that—and you could say the same thing with stage IVB patients. In the GOG 240 trial, the de novo stage IVs did well with bevacizumab, but pembrolizumab didn’t appear to do much. You can’t support your bias; you must follow the data. You told me, Krish, that GOG 240 didn’t work well in adenocarcinomas, and you said all those people are wrong. You wrote a paper on how bevacizumab and adenocarcinomas and the cervix is a good option. I invite you to use the same rigor here. This is a study about pembrolizumab and a study about bevacizumab. I want to get your opinions. Warner, Premal, and Bhavana, what do you think? Do you need bevacizumab in the KEYNOTE-826 trial, or do you need 4 drugs?

Bhavana Pothuri, MD:I interpret the data exactly as they were presented. Use bevacizumab when you use it. Pembrolizumab clearly has a benefit in this setting.

Bradley Monk, MD, FACOG, FACS: What do you think, Premal?

Premal H. Thaker, MD, MS:It’s nice that you can counsel your patient if they’re not a candidate for bevacizumab because they will still benefit. You can use it to your advantage for your patient; that was the take-home. If you have a patient, as you mentioned, with diverticulitis or proteinuria, and you feel you can’t, there’s a benefit for these patients.

Bradley Monk, MD, FACOG, FACS: Warner, will this become the new standard if it gets approved, or do you think people are going to be skeptical?

Warner K. Huh, MD, FACOG, FACS:No. There’s a lot of excitement and enthusiasm for this. What Krish has done with the GOG 240 trial and bevacizumab is truly remarkable. There are very few oncologists, let alone gynecologic oncologists, who can change the standard of care to the magnitude that he has. But people are very excited about this.

Bradley Monk, MD, FACOG, FACS: We’ve ridden the life cycle, right? Now bevacizumab is a biosimilar, so the cost is coming down. That’s important.

Transcript edited for clarity.

Related Videos
Experts on endometrial cancer
Experts on endometrial cancer
Experts on endometrial cancer
Experts on endometrial cancer