Brahmer Highlights 5-Year Nivolumab Findings in NSCLC

Julie R. Brahmer, MD

Impressive, long-term findings from the PD-1 inhibitor nivolumab (Opdivo) in patients with metastatic non—small cell lung cancer (NSCLC) are a signal of hope, according to Julie R. Brahmer, MD, who is the lead author of the CA209-003 trial.

The 5-year results were presented during the 2017 AACR Annual Meeting, and demonstrated that the estimated 5-year overall survival (OS) for these patients was 16% versus the historically seen 4% OS with standard chemotherapy.

“This is [a sign of] hope for patients with NSCLC that there is a possibility that even if you have metastatic disease, you can be alive and well off therapy, with your disease under control, even 5 years from now,” she said.

OncLive: Can you give an overview of the CA209-003 data you presented?

In an interview with OncLive, Brahmer, associate professor of oncology, Bloomberg Kimmel Institute for Cancer Immunotherapy, and interim director of the Sidney Kimmel Comprehensive Cancer Center at Johns Hopkins Medicine, discussed the 5-year findings, their significance, and expected next steps for this research.Brahmer: I presented the 5-year follow-up data for the trial CA209-003, which was a phase Ib cohort expansion study. I specifically reported on patients with NSCLC that were entered in the study. The study included approximately 129 patients with NSCLC with metastatic disease who had been previously treated to get onto this trial. They could have been treated with 1 to up to 5 prior lines of therapy, and they were randomized to 3 different doses of nivolumab given once every 2 weeks for up to 2 years. The previous efficacy results and the 3-year follow-up have been previously published, but I presented on the 5-year follow-up data.

What is the significance of these findings?

The estimated 5-year OS for this group of patients was 16%. Compared to the historical data we have, this is an improvement. Previously, we saw patients with metastatic disease treated with standard treatment have a 5-year OS of approximately 4%.This is the first time that we have 5-year follow-up data, and there is a group of patients that can live 5 years or beyond, even in this trial where we stopped therapy after 2 years.

What are the next steps with this research?

We looked at the various characteristics to try to figure out what clinical characteristics could predict who would live for 5 years or beyond and there were no clear-cut characteristics where we could look at the patient and predict if they would be in the 5-year survivor group.The next steps are to look at the pivotal phase III trials and see if are we seeing very similar data to their 5-year follow-up that show consistent plateauing of the curve. Potentially we could also present even longer-term data on this group of patients. We certainly want to look for biomarkers for long-term OS, as well as look at being able to stop therapy early and see that that response can be maintained even off treatment.

What else would you like to highlight about this study?

Why have long-term follow-up results not been more common in lung cancer?

Which biomarkers are being looked at right now?

Right now, based on the approval, nivolumab is given every 2 weeks until the cancer progresses, but now we have data that there is a group of patients where, if we stop therapy, their disease control will be maintained for a very long period of time.I think this is a study that has the longest follow-up thus far in patients with NSCLC and there was no clear disease characteristic that could point us to having a longer OS, and this just behooves us to look further at biomarkers to be able to predict how well patients do, as well as look at patients and laboratory correlates to predict those patients for whom we can stop therapy.This is one of the first trials that have reported that there was activity with a PD-1 antibody. In the past, for lung cancer, when we’re treating patients with chemotherapy, there just weren’t patients around after 5 years of treatment. With a 5-year OS of 4%, the likelihood of being able to follow patients this long was not very common back then.First, the PD-L1 biomarker—looking at PD-L1 status on a patient’s tumor is being looked at, and we looked at that in this trial as well. There was no clear-cut association with long-term OS. About 50% of our patients had PD-L1 or a tumor that stained for PD-L1, so out of the 129 patients we could only do the testing on about 60 patients.

Looking at the different subgroups, though, between PD-L1—negative (less than 1% staining) and PD-L1–positive (1% or greater staining), there was really no difference in OS, it was approximately 20%.

What other ongoing lung cancer research are you interested in?

If you looked at PD-L1 staining using a cutoff of 50% (50 out of 100 cells stained for PD-L1) the 5-year OS rate was 43%, but there were only 13 patients in that group.I think some of the different signatures looking for biomarkers is very exciting, and I think that will leapfrog us into the next phase of studies to look at these different signatures in NSCLC, and hopefully some of the different combinations as well. Potentially not all patients will respond to single-agent PD-1 or PD-L1 therapy. The hope is that there are different combinations, including the CPI-444 plus atezolizumab (Tecentriq) combination also presented at the AACR meeting.

Brahmer JR, Horn L, Jackman D, et al. Five-year follow-up from the CA209-003 study of nivolumab in previously treated advanced non-small cell lung cancer: clinical characteristics of long-term survivors. Presented at: 2017 AACR Annual Meeting; April 1-5, 2017; Washington, DC. Abstract CT077.