New Concepts: Multimodal Approaches to Advanced NSCLC - Episode 8
Solange Peters, MD, PhD: In this question we all have about improving stage 3 outcome, and potentially giving something after radiochemotherapy, is, in the patients who are not surgical patients—patients who receive radiotherapy for any kind of decision-making process—the immunotherapy might look very attractive. It might look very attractive mainly because of the biology underlying this strategy. We have been showing, for a while in other diseases—mainly melanoma with our preclinical models—that when you deliver radiotherapy and also chemotherapy, you really induce cancer cell deaths that might somehow activate the immune system already. Some pieces of the cells dying are presented to the immune system and start to create something like a vaccine against the tumor. And if you can really mount this immune response or reinforce it, make it active, make the T cells traffic to wherever they need to traffic, you can really generate what we call immunogenic cell deaths, meaning the vaccine works and the T cells, or lymphocytes, are able to come or go wherever you want in the body to kill any new cancer cell. If this really works, setting after radiotherapy and chemotherapy would be the best one because you have these cancer cell deaths. You have this immune system being awake.
So, that’s really the setting of the PACIFIC trial. The PACIFIC trial is using standard radiochemotherapy as defined by a minimum of 2 cycles of chemotherapy, minimum 50 gray—so, they’re very normal standards—and proposes to randomize patients between immunotherapy—an anti-PD-L1, durvalumab—for a year or placebo for the other group. And the idea is to show you are able to prove this biological mechanism of preventing or vaccinating the patients against the tumors or preventing the relapse of the tumor. But not only locally; wherever you could have an antibody. And that was the question of PACIFIC. I must say, this is really unprecedented, in stage 3 where we have been working, unfortunately, unsuccessfully, in improving the prognosis. This really strongly impacts the outcome of these patients. It prolongs the progression-free survival of 11 months, which is really absolutely amazing, I must say. In the field of cancer, in general, 11 months’ PFS is really very big. But also, what I like in this trial is that it demonstrates the biological principle because it also prevents the emergence of new metastases in other organs. For the brain, you reduce from 11% of patients relapsing in the brain to 5%. So, you show that this story of vaccination works. You are creating a state where the patient can defend himself against the emergence of new lesions and new organs being affected by the disease.
The trial was proposing to take the therapy for a year, right? For the patients, most of them, because it’s well tolerated, can continue up to a year. I’m sure daily life data will tell us it is absolutely feasible to give it up to 1 year. Now durvalumab can even be given every 4 weeks, so it’s kind of an easy task. Afterwards, what we don’t know is the ideal duration. This trial was studying 1 year. Maybe the next trials will let us know if it’s maybe better to give 2 years or if it’s better to give a shorter time. But now we know it’s 1 year. So, then you stop and you control, you just have a situation when you wait and wait by doing CT scans and observing the patient over time. That’s probably an interesting question to know. In advanced disease, we imagine probably a bit more than a year might be useful, so this is a question for the next trials.
In the PACIFIC trial, there was an attempt to try to identify the patients who benefited the most from this strategy, which is a complex one. You have to come for a year regularly to the hospital. It was a failure. When I say that, in this trial, quite a lot of patients had stage 3b, which is known to be not a good disease. More or less half and half, stage 3 and stage 3b. And the benefit of giving durvalumab maintenance was met in both subgroups, and I would say to the same magnitude. So, you cannot select patients based on the real extent of the diseases initially. Then you can look at other subgroups. You can look at gender, you can look at histological subtypes—squamous, nonsquamous—you can look at the real lymph node invasion, you can look at age. None of these subgroups were showing different results, so when you look at this famous Forest plot, they all benefit, and they all benefit equally in the strategy.
But I would say the more important parameter in this subgroup is trying to define a biological biomarker. In advanced disease, we use PD-L1. It’s not a good biomarker but that’s the one we have for the time being. So, PD-L1 predicts to benefit—a higher benefit of immunotherapy—checkpoint inhibitors in advanced disease. But in that situation where you give radiochemotherapy, so you completely change the microenvironment of the tumor by giving radiochemotherapy, you really modify everything. You create this vaccination. And giving durvalumab afterwards, the predictive ability of PD-1 looks like it’s disappearing. So, it means that the strategy is not dependent or not even better in high—PD-L1 patients. It probably means that by giving this radiochemotherapy, you make all of the patients equal in the probability of developing an immune response. And it just means that at the time being, you are not able to select the best patients for that strategy as long as they don’t have contraindications and they have a good performance status.
Suresh Senan, MRCP, FRCR, PhD: I think this study represents a tremendous breakthrough for us because we’ve been trying a number of different strategies—higher doses of radiation, more chemotherapy, newer chemotherapeutic agents, alternative radiation fractionation schemes—none of which have made a big breakthrough. And now we have a treatment combination that increases progression-free survival by 11 months, and that’s really quite impressive. I think it’s a new paradigm and it’s a new platform for us to expand, to improve the results further.
What is particularly interesting to me is that we have been treating patients with immunotherapy with durvalumab. The local disease response in the radiated region is also improved; an increase from 16% to about 28%. It confirms the hypothesis that chemoradiotherapy is a sort of in situ vaccine, which, subsequently, the more patients exposed to immunotherapy, the more it actually improves the local response. That’s a very impressive finding for radiation oncologists who focus on local therapy.
And the second impressive finding is a marked decrease in distant metastases. Half of the patients in the placebo arm got disease progression in the first 12 months. It gives an idea of what the magnitude of the challenge facing us is. So, we thought better systemic therapy, whatever you do locally, may be a bit irrelevant. You now have durvalumab as an agent improving both local control and distant metastases.
A third implication for radiation oncologists is that perhaps their role of radiation is just to avoid toxicity, because higher doses are clearly not necessary. In the trial, the range was from 56 to 66 gray, with an average of about 60 gray. And that fits very nicely with the findings of the trial that said that 60 gray is the standard. So, we should not try and push the dose up because if patients get more esophagitis as a result of that, then it would delay their treatment. And that’s what’s in the quality-of-life data from the PACIFIC trial, which were presented at the World Lung Cancer Conference. It showed that patients may take quite a few months to recover from the esophagitis. So, more radiation is actually not necessary, and we could do well with 60 gray and bringing in the effective immunotherapy as soon as possible.
So, I think it has a lot of major implications for the way we think about chemoradiotherapy. I think the standard chemoradiotherapy block is now defined for patients. This chemoradiotherapy—to about 60 gray, followed by 12 months of immunotherapy—would be the new standard.
Transcript Edited for Clarity