Video
KEYNOTE-629 Trial and Pembrolizumab
Dr In comments on data from the KEYNOTE-629 trial that focused on the use of pembrolizumab as a therapy option for patients with cutaneous squamous cell carcinoma.
Glenn J. Hanna, MD: There are also data for another PD-1 inhibitor that recently has emerged based on the KEYNOTE-629 trial: pembrolizumab, a similar PD-1 antibody for cutaneous squamous cell carcinoma. This isn’t a fair question, but it’s tempting in situations to think about the data comparatively even though these trials are independent. They share some overlap in populations, but they’re independent trials and investigations. I was just curious. Dr In, do you have the option of 2 different PD-1 inhibitors? Do you interpret the data differently? Are there things based on institutional preference that drive your decision-making? Do you think there’s even a difference clinically or practically speaking in terms of the response or outcomes?
Gino K. In, MD, MPH: In general, it’s not fair to compare across trials. As far as I know, there are no trials that compare active checkpoint inhibitors directly against one another. When I look at the KEYNOTE-629 data in comparison with the registrational cemiplimab data, what we saw at least at the first internal analysis was that the response rates did look lower with pembrolizumab compared with cemiplimab. We’re seeing overall response rates around 40% compared with the cemiplimab data, where it was 45% to 50%.
These populations are perhaps slightly different. This was more of a Europe-based study, and some of the definitions about the extended disease also matter. Patients with locally advanced disease seem to have better responses. That’s consistent across both trials. But a lot of patients derive clinical benefit, even beyond those who have responses. And their responses are durable. Another thing that’s not specifically presented on these figures is that a lot of these patients still have a fairly high complete response rate with single-agent PD-1. In the pembrolizumab and cemiplimab data, a complete response rate can be somewhere between 10% and 20%.
In our clinical practice, I don’t view 1 in favor of the other. I’ve used cemiplimab more because it was approved earlier. But in the real world, sometimes it comes down to which drug can I get approved for my patients by their insurance. They’re both exceptional drugs for this disease setting, well tolerated, and very good for patients, so I don’t view 1 as superior over the other in that setting.
Glenn J. Hanna, MD: I agree. Doing a deep dive into the data because there are people who want to compare, but you have to be careful. Have patients had X amount of prior treatments or are they subtly different among the groups. There was the issue with weight-based dosing that was switched to flat dosing in the cemiplimab trial and some of the cohorts, whereas the other trial was more flat dosing. All those nuances are well taken and well recognized. The key is that this category of drug has a clear signal that can be quite durable and well tolerated.
We should highlight quality of life related to other cytotoxic agents in this space or antibody therapies where cemiplimab and pembrolizumab perform quite well. We said this is a pretty delicate population. Dr Cohen, did you want to add anything about the first-line pembrolizumab and cemiplimab data?
Ezra Cohen, MD, FRCPSC, FASCO: No. I’d reiterate what has been said. It was a good point about the complete responders. We see a higher complete response rate in cutaneous squamous cell carcinoma than in other solid tumors. That probably goes back to the high tumor mutational burden and the fact that immunotherapy is active in these cancers. Clearly, we have 2 very good options for patients like this.
This transcript has been edited for clarity.
%20(2)%201-Recovered-Recovered-Recovered-Recovered-Recovered.jpg?fit=crop&auto=format)
%20(2)%201-Recovered-Recovered-Recovered-Recovered-Recovered.jpg?fit=crop&auto=format)
%20(2)%201-Recovered-Recovered-Recovered-Recovered-Recovered.jpg?fit=crop&auto=format)
