Pembrolizumab Monotherapy Shows Potential in Non-Clear Cell RCC

David McDermott, MD

Data from the KEYNOTE-427 trial with pembrolizumab (Keytruda) monotherapy may pave the way for the use of immunotherapy in patients with non-clear cell renal cell carcinoma (RCC)—an area for which there are no currently approved therapies available, said David F. McDermott, MD.

In the single-arm, phase II trial, investigators explored the use of single-agent pembrolizumab in patients with advanced clear cell RCC (cohort A) and non-clear cell RCC (cohort B). Data from cohort B showed that the agent had encouraging antitumor activity as a frontline treatment in this population—especially in those with papillary or unclassified histology.

Specifically, the overall response rate (ORR) via central review in the entire cohort of patients with non-clear cell RCC was 24.8%. In those with papillary histology, the confirmed ORR was 25.4%, while it was 34.6% in patients with unclassified histology. Those with chromophobe histology had a lower confirmed ORR of 9.5%. At a median 11.1 months of follow-up, the median progression-free survival (PFS) with pembrolizumab was 4.1 months, and the estimated 12-month overall survival (OS) rate was 72%.

“Getting these data give us a new path forward potentially for immunotherapy in non-clear cell RCC, which is interesting and surprising in the cytokine era,” said McDermott. “To have this signal of a 25% response rate gives us something to build on, and hopefully, it will be something [we will be able to] provide to these patients who obviously need new therapies.”

In an interview with OncLive, McDermott, staff physician, Hematology/Oncology, director, Biology Therapy, at Beth Israel Deaconess Medical Center, and professor of medicine at Harvard Medical School, provided insight on pembrolizumab in non-clear cell RCC and the next steps for research in the overall RCC paradigm.

OncLive: Could you discuss the KEYNOTE-427 trial and the findings from cohort B?

McDermott: There is a lot of excitement about combination therapy with PD-1—based combinations—whether it’s combining PD-1 with VEGF or with CTLA-4—for treatment-naïve patients with kidney cancer. However, less is known about single-agent PD-1 blockade in these patients. As such, the KEYNOTE-427 trial was designed to get more data on that group. We wanted to find out: how does pembrolizumab work in patients who have never received treatment, both in clear cell RCC and non-clear cell RCC?

In 2018, we presented the data on patients with clear cell RCC. We saw a 36% ORR with single-agent pembrolizumab in that group of patients with pretty good tolerability. Then, this year, we followed up that up with data on 165 patients with non-clear cell RCC. Most of these patients had papillary kidney cancer, but we also have patients with chromophobe and unclassified [histologies]. We did an independent review of their histology to confirm that they actually had non-clear cell RCC, and what we found was interesting. Our findings were better than we expected, just like the clear cell experience was better than we expected.

The ORR in that group of 165 patients was 25%. The [confirmed ORR was] 25% in the papillary group, and it was lower at around 10% in the chromophobe group, which has been seen in other studies. However, in the unclassified group, it was 34%. Interestingly, when you look at PD-L1 staining in the tumor microenvironment, for those who had PD-L1—high tumors or who had a composite positive score (CPS) score >1, that ORR was 30% compared with 10% in the PD-L1– or CPS-low group. This suggests that PD-L1 staining might be a way of enriching for benefit. From a safety point of view, things look good—pretty typical for pembrolizumab. We didn't see any safety signals. Only 6% of patients had to stop treatment due to toxicity. There were, however, 2 treatment-related deaths.

Essentially, we have a pretty tolerable regimen for a condition, non-clear cell RCC, where, up until now, there have been no specific FDA approvals of any drugs. Essentially the recommendations are to use sunitinib (Sutent), but there's no clear approval for sunitinib in non-clear cell RCC. Furthermore, all of the prior checkpoint inhibitor trials have excluded non-clear cell RCC. Therefore, these data are important.

What are the next steps with this research?

It is possible, given that it's such an unmet need, that we might be able to try to move toward regulatory approval with just this data set. However, oftentimes, the FDA will want a randomized trial to confirm the benefit [of the regimen], so those discussions will probably start soon.

Do we need to do another trial? Do we need to do a randomized trial? That is one way to go. Also, we have to try to improve how we select patients for this therapy. Many other correlative therapies were done on the tissue on this trial. Therefore, at future [meetings], we will get data on the DNA sequencing and the RNA sequencing from these tumors. Can we predict better than histology or immunohistochemistry who should get this treatment and who shouldn't?

Is PD-L1 is a reliable biomarker for immunotherapy in these patients?

It's never going to be a perfect marker, but it certainly does enrich for benefit. In this group, patients were 3 times more likely to respond if they had PD-L1—high tumors than PD-L1–low disease. I believe we can do better than that. Other studies suggest that RNA sequencing might be better than just PD-L1 alone, so we might be able to add things in to increase benefit for patients. However, PD-L1 expression did better in this group than it has performed in other places. As such, it should be explored further.

There is a lot of excitement regarding immunotherapy combinations in this space. What are some challenges to consider with that approach?

It's interesting. What the KEYNOTE-427 trial suggests is that a lot of the benefit that we're seeing with these combinations is driven just by PD-1 and that's an easier therapy for most patients to tolerate than combinations. The question is, “Can we move single-agent therapy to the right patient?” That's a direction we need to go in.

Combinations are generally well tolerated but they do add a certain amount of complexity, particularly with managing VEGF. There are chronic adverse events (AEs), such as fatigue and diarrhea, which are not necessarily dangerous, but they can be challenging. When you combine PD-1 inhibitors with CTLA-4 inhibitors, you can see interesting complete responses, which is very exciting because it can sometimes lead to remissions.

However, you need to educate your team on managing the immunotherapy AEs when you're combining PD-1 with CTLA-4. Hopefully, there are different groups of patients who will benefit from each approach—maybe 1 group should get a single-agent, the other should get a PD-1 inhibitor with a VEGF inhibitor, and the next should maybe get a PD-1 inhibitor and a CTLA-4 inhibitor. We’re a long way from that at this point, but hopefully that's where we're going.

What is your take-home message to your colleagues working in RCC?

It's clear that we're improving OS, which is very exciting. We're going to start seeing a median OS of 30, 35, or 40 months for our patients, which is so much better than where we were 10 years ago. By combining these PD-1—based therapies with some of these other approaches, we're definitely improving survival, which is a really great outcome. Hopefully, we're also generating more complete responses, because those will lead to remissions; with that, patients can not only live [longer], but they can live off of treatment. Now that we have shown that we can create remissions, we can work to make them more common for our patients.

Lee J-L, Ziobro M, Gafanov R, et al. KEYNOTE-427 cohort B: first-line pembrolizumab (pembro) monotherapy for advanced non-clear cell renal cell carcinoma (NCC-RCC). J Clin Oncol. 2019;37(suppl 15, abstr 4569). doi: 10.1200/JCO.2019.37.15_suppl.4569.