Expert Hopes Nivolumab/CRT Combo Will Advance Head and Neck Cancer Care

Robert Ferris, MD, PhD, discusses the RTOG 3504 trial, as well as the utility of immunotherapy agents for patients with intermediate and high-risk head and neck squamous cell carcinoma.

Robert Ferris, MD, PhD

Investigators are evaluating the safety of chemoradiation in combination with nivolumab (Opdivo) in intermediate and high-risk head and neck squamous cell carcinoma (RTOG 3504, NCT02764593).

The trial, which is currently ongoing, randomizes 176 patients to cisplatin with radiation with nivolumab or placebo.

Nivolumab has shown an improvement in overall survival in patients with platinum-refractory, recurrent or metastatic head and neck squamous cell carcinoma, compared to standard therapy. The PD-1 inhibitor is approved by the FDA for this indication.

OncLive: Please provide an overview of this study.

In an interview with OncLive at the 2017 ASCO Annual Meeting, Robert Ferris, MD, PhD, vice chair for Clinical Operations, associate director for Translational Research, and co-leader of the Cancer Immunology Program at the University of Pittsburgh Cancer Institute, discussed the RTOG 3504 trial, as well as the utility of immunotherapy agents in this landscape.Ferris: At ASCO [2016], we demonstrated that anti—PD-1 therapy as a monotherapy was better than standard of care chemotherapy. Demonstrating that head and neck cancer was an immune-responsive disease led to the natural application of anti–PD-1 therapy, such as nivolumab, into earlier stages of disease.

Now in the locally advanced setting, the RTOG 3504 trial aims to look at an intensification by adding immunotherapy to the standard of care chemoradiation, using cisplatin with intensity-modulated radiation therapy. Therefore, in a randomized trial, we hope to add nivolumab to chemoradiation and improve progression-free and overall survival.

What is the significance of this trial?

At this stage, the RTOG 3504 is focused on a phase I safety evaluation. Adding nivolumab to chemoradiation either in the weekly setting at 40 mg/m2 or as a bolus cisplatin 100 mg/m2 every 21 days. There are 2 standards of care for adding cisplatin to radiation, so we needed to test the addition of nivolumab to both standard of care regimens for locally advanced head and neck cancer.In head and neck cancer, we have HPV-positive disease, which tends to do better, and HPV-negative disease, which really has not seen any improvements in survival despite intensification with multiple chemotherapies, or even better radiation techniques. So, what we’ve done in the RTOG 3504 trial is to ask whether adding an immunotherapy on top of concurrent chemoradiation—which is really the most efficacious therapy for locally advanced disease—improves survival for the intermediate and high-risk patients who have really not seen a benefit in survival. So, we can randomize nivolumab monotherapy versus chemotherapy in recurrent metastatic disease, but by applying it to earlier stages—such as the locally advanced setting—we have the potential to actually elevate the survival curve and cure more patients in the intermediate and high-risk setting, which have not seen benefits in some decades regarding survival.

Combining immunotherapy with radiation makes a lot of sense—according to preclinical models—but we have relatively little data from patients. Another complicating issue is that the dosing schedule, the fractionation schemes of radiation, are a little bit restrictive, so we have to use them in the approved and traditionally applied format. There is some data that suggests such a fraction scheme may not be the best partner with immunotherapy, but in the locally advanced setting we have a curative regimen and we are pretty uncomfortable moving away from that in a group that we are trying to cure more.

Is there any data to report yet?

What are the updates with CheckMate-141?

Adding immunotherapy in the recurrent metastatic setting with different fractionation schemes where we can be more creative, those settings will get a lot of developmental information by adding nivolumab or other anti—PD-1 therapies. And then perhaps move them earlier as we get more data in the locally advanced setting with immunotherapy.We are currently in the middle of the phase I component of RTOG 3504, but we are close to getting out of the dose-limiting toxicity phase. I can tell you that we would be very surprised if there was a dose-limiting toxicity, in part because there are some investigator sponsored trials, some with nivolumab and some with pembrolizumab (Keytruda). In fact, in the clinical science symposium and at the last ASCO, we saw safety data from other anti—PD-1 therapies with concurrent chemoradiation as well as at the SITC meeting in 2013. So, I cannot report on our phase I yet, but I will say that we would be very surprised if we saw something different. The CheckMate-141 trial was a randomized phase II trial of nivolumab randomized 2:1 versus investigator’s choice of either docetaxel, cetuximab (Erbitux), or methotrexate. That trial had an overall survival benefit from a sort of preliminary first data look, because it met its primary endpoint basically on the first interim analysis. We now have a longer-term follow-up—minimum of 11.4 months—to demonstrate that tail of durability.

Secondarily, almost half of the patients had progressed on cisplatin in the locally advanced setting and so we’ve really focused the 1-year update combining it with a specific focus on the subset of patients, the nearly half, that had progressed in the locally-advanced setting. So, it’s really a first-line platinum refractory population. They did particularly well to nivolumab, so it will be interesting to see that subgroup analysis.

In addition, I am reporting in particular on those patients who had prior exposure to cetuximab. We know that the CheckMate-141 trial was focused on those who had progressed within 6 months of cisplatin, but about 62% had also seen cetuximab. And this was a prespecified stratification factor. And, because cetuximab is an important therapy either in the locally advanced or recurrent setting for head and neck cancer, we thought that it was very important to understand how those patients who had progressed not only on cisplatin, but also had been exposed to cetuximab, would respond to an immunotherapy.

Could you speak to the utility of nivolumab and pembrolizumab in this landscape?

There is more to come, I think the CheckMate-141 trial will help us to understand how to apply nivolumab in different subgroups with biomarkers and then in the locally advanced setting. I think the field has gotten to a point where they regard pembrolizumab and nivolumab as generally interchangeable. Pembrolizumab got an accelerated approval in August of 2016 for head and neck based on a single-arm phase II trial. They were then required to run a trial that was very similar to CheckMate-141, where they randomized pembrolizumab versus single agent investigator’s choice. That was KEYNOTE-040, and we are still waiting for those data—I think it has completed enrollment.

Most oncology folks believe nivolumab and pembrolizumab behave relatively similarly, so we are trying to find trials that reflect that. Combinations—additions into earlier lines of therapy with chemoradiation—are also happening with pembrolizumab. So, I think we currently don't really select based on efficacy or adverse event profile; the only quality of life data that we have is with nivolumab, so I don’t think anyone thinks that pembrolizumab has a different quality of life impact—we simply just don’t have the data.

Nivolumab now has publication in Lancet Oncology, that we had one of our co-authors, Kevin Harrington, PhD, lead. So that paper is accepted and is emerging with quality-of-life data that is stable or improved with nivolumab. We would expect pembrolizumab to have the same effect but they simply did not add that to their trial.

Is there anything else that you would like to add?

The biomarkers should be useful—pembrolizumab has focused a little more on PD-L1, both in the tumor cell and the inflammatory cell, which is called the composite score for PD-L1. We have a similar finding that you can add to the predictive ability of tumor PD-L1, if you add the tumor-associated inflammatory cell. If you take the tumors that are PD-L1 negative, you can actually improve the hazard ratio substantially if the tumor had PD-L1 on its inflammatory cells. So, we are getting better at selecting patients that are most likely to benefit. I think we are also seeing moving the anti—PD-1 agents—both pembrolizumab and nivolumab—into the preoperative setting, the neoadjuvant setting, because we can get better tissue biomarkers and begin with a surgical specimen after 1 or 2 doses of anti–PD-1, to begin to understand the dynamic biomarkers. And that may help guide us, as to who should go on to receive postoperative anti–PD-1 therapy. Or maybe, who is not going to be responsive and needs to have a little boost to trigger that interferon gamma before an immune checkpoint therapy.

ML Gillison, RL Ferris, Q Zhang, et al. A randomized phase II study of chemoradiation (CRT) +/- nivolumab (Nivo) with sequential safety evaluations of Nivo +/- lirilumab (Liri) or ipilumumab (Ipi) concomitant with (C) RT in intermediate (IR) and high-risk (HR) head and neck squamous cell carcinoma (HNSCC) (RTOG 3504, NCT02764593). J Clin Oncol 35, 2017 (suppl; abstr TPS6097)