Research into immunotherapy for non–small cell lung cancer (NSCLC) is progressing rapidly according to Naiyer Rizvi, and is unlikely to slow any time soon.
"The field is changing so fast,” says Rizvi, director of Thoracic Oncology and Immunotherapeutics at Columbia University Medical Center. “Soon, we will have a better understanding of the first-line use of PD-1 agents, Then, maybe a year later, the data on the combination of PD-1/PD-L1 and CTLA-4 will come out. It is going to be a busy year. The NCCN is going to be busy rewriting their guidelines every six months at this rate.”
Several immunotherapeutic agents are also being investigated in new combinations and earlier settings.
The ongoing KEYNOTE-024 study is currently investigating the anti–PD-1 agent Keytruda (pembrolizumab) in the first-line setting for patients with stage 4 metastatic NSCLC whose tumors express PD-L1.
The phase 2 POPLAR study, which compared the PD-L1 inhibitor atezolizumab with docetaxel, found that in previously treated patients with NSCLC, those patients with the highest level of PD-L1 expression experienced a median overall survival of 15.5 months with atezolizumab versus 11.1 months with docetaxel.
Despite these data, the significance of PD-L1 as a biomarker remains unclear. Phase 1b data recently published in The Lancet Oncology1 showed that the combination of the PD-L1 inhibitor durvalumab and the anti–CTLA-4 agent tremelimumab induced a response rate of 23 percent in patients with advanced NSCLC. Responses in the trial were observed regardless of PD-L1 status.
Immune agonists such as OX40 and 4-1BB are also an exciting possibility, says Rizvi. In an interview with CURE, Rizvi discusses the role of PD-L1 testing, which PD-1/PD-L1 agents have the most potential, and what is on the horizon for the use of immunotherapies in NSCLC.
What role do you predict for PD-L1 testing in NSCLC going forward?
The data with Opdivo (nivolumab) in second-line lung cancer is pretty compelling due to the survival advantage that exists, so I think that is becoming the standard of care in all second-line NSCLC. If that is the case, the question then becomes, “How is the biomarker relevant?"
I think the biomarker is useful in understanding who will be more likely to respond in the context of clinical trials or at academic centers. If you have a patient who you are testing for PD-L1 upfront, and they come back as PD-L1 “zero,” they are more likely to benefit from a combination trial or something of that sort if they progress on chemotherapy.
The KEYNOTE-024 trial is looking at Keytruda in the first-line patient population versus chemotherapy. If that is positive, then I believe everyone will get PD-L1 testing upfront and then eventually receive first-line Keytruda.
What factors should be considered when choosing between Opdivo and Keytruda for NCSLC?
For newly diagnosed patients with known PD-L1 status, I think you could make the argument for either Opdivo or Keytruda in the first-line setting. I don’t think there is reason to pick one over the other.
I do think, however, giving Keytruda every three weeks is more favorable for patients than giving Opdivo.
Do you see a role for atezolizumab?
There is a similar trial to KEYNOTE-024 looking at atezolizumab versus chemotherapy in the first-line setting. Atezolizumab is a little bit late to the game to capture that space in lung cancer. I am not sure what the uptake is going to be, because the PD-L1 positivity requirement with that agent is less. Their biomarker requires 15 percent positivity, whereas for Keytruda, it is around 25 percent.
From a treatment standpoint, you will have a lower hit rate with the atezolizumab biomarker versus the Keytruda biomarker. They have multiple first-line chemotherapy combination trials, but they are not as far along as some of the immunotherapy combination trials with other PD-1/PD-L1 agents.
What are the biggest questions that remain regarding the use of immunotherapy?
I think the biggest question is, “How do we turn the 75 percent to 85 percent of patients who don’t respond to immunotherapy into responders?” We have made some process. We published preliminary data in The Lancet Oncology looking at the combination of durvalumab and tremelimumab that showed a response in patients who were PD-1–negative as well as positive.
That is the right direction; however, as we move forward, we need to determine exactly why these patients don’t respond and how we can convert those nonresponders to responders.
What should oncologists take away from the durvalumab and tremelimumab preliminary study results?
It is a very active combination. The schedule is every four weeks for the durvalumab and tremelimumab regimen, which is attractive. Anytime a CTLA-4 antibody is used in combination with a PD-1/PD-L1 agent, you are going to have more toxicities than you would with an immunotherapy alone. You have to find the balance between better response rates versus more toxicities.
Is there a learning curve for oncologists utilizing immunotherapies?
I think doctors, if they haven’t already learned, will be learning how to manage these toxicities. There is a learning curve — not a steep learning curve — but it will take time.
If an oncologist’s only experience with immunotherapy is with melanoma, they may not have a lot of experience because you don’t see melanoma that often. There is a different set of toxicities that are less predictable, and are ones that don’t necessarily occur in the first week. It is something that oncologists will gain experience with as they treat more patients.
What’s on the horizon for immunotherapies in NSCLC?
There is a lot of interest in looking at combinations with immune agonists such as OX40 and 4-1BB. There are different ways these agents could be used to remove the immune suppression within the tumor. There are hundreds of ongoing trials right now aiming to better understand this.
We don’t know if anything is going to be better than CTLA-4 with PD-1/PD-L1. We need something we can give that has a more favorable toxicity profile or that is more active. Or, we need to figure out how to select patients for different combinations, based on what their profiles are.