Anne Tsao, MD: Any new patient who comes in the door with metastatic non—small cell lung cancer does have to have their molecular profile. But if their genetics don’t reveal a mutation driver, we obviously are looking at PD-L1 [programmed death-ligand 1] immunohistochemistry [IHC]. If they have greater than 50% expression of PD-L1 IHC, then they are a candidate to receive pembrolizumab by itself. Certainly, though, for those who have less than 50% PD-L1 expression, they can receive chemotherapy, usually a platinum-based doublet, with pembrolizumab. And we know that there’s very good efficacy with that.

In terms of whether a patient who has high PD-L1 expression, whether they should receive the pembrolizumab alone or the chemotherapy-pembrolizumab, they usually make a decision about that depending on how much tumor or what tumor bulk the patient has. Are they imminently threatened by their disease? And if they are, and their performance status is good enough so that I can give them the triplet regimen, regardless of the PD-L1 expression, I might choose to do that if I need a rapid debulking.

Benjamin Levy, MD: There’s a lot of factors that go into the treatment decisions for a patient with non—small cell lung cancer without a driver mutation. We have to look at their performance status. We look, have to look, at their disease burden. We have to look at their motivation. We have to look at their comorbidities, all of those things kind of factor into a treatment decision. But then, of course, we have to look at other tumor factors. We have to look at their PD-L1 status.

So when a patient comes in, and we’ve made the decision or we’ve uncovered that they don’t—that we’ve identified that they don’t have a genetic alteration—the treatment decisions are really contingent, in my mind, on symptom burden and the PD-L1 level. If patients have a PD-L1 greater than 50%, and they don’t have a driver, these are the patients to whom I would offer single-agent pembrolizumab every 3 weeks. For a patient with a PD-L1 greater than 50% who is very symptomatic, this is an opportunity in which I would give pembrolizumab in combination with chemotherapy—carboplatin and pemetrexed.

For patients less than 50%, these patients, generally, without a driver mutation, I would offer the triplet therapy of carboplatin, pembrolizumab—carboplatin, pemetrexed, and pembrolizumab. I think we need to be mindful that not all patients are candidates for immunotherapy. So we have to factor those coexisting illnesses that they may have autoimmune phenomenon. We have to make sure that we’re not giving immunotherapy to patients who may have really significant autoimmune diseases. But for some autoimmune diseases we can give it, and it has to be a patient-by-patient decision on whether that’s done or not.

The factors that I look at include symptomatic burden, tumor burden, comorbidities, patient motivation, and…the PD-L1 to drive treatment decisions for these patients without a genetic aberration.

Roy Herbst, MD, PhD: Well, treatment of lung cancer is so revolutionized with the driver mutations, but we have to remember that 80%-plus of patients won’t have a driver. In that case, I really am, I’m looking for reasons to use immunotherapy. In almost all cases, immunotherapy, in my opinion, should be part of the first-line treatment, either alone or in combination with chemotherapy, different chemotherapy in squamous or nonsquamous disease. Of course, there are patients who are not candidates for immunotherapy because of rheumatologic or autoimmune-type processes. But these days, really, for a nondriver, I’m asking if I can get someone immunotherapy. Why? Because I know that the results with chemotherapy alone, especially in the metastatic setting, are quite disappointing. And certainly, while there are some other therapeutic options, none of them is even comparable to immunotherapy.

Well, you know, in our clinic here at Yale University, we get PD-L1 status on all patients. That’s important. In fact, it’s quite easy to get. It usually comes back before the mutational profiling, and it’s often having to wait for the mutational profiling before we start. But for PD-L1 high, with more than 50% cells PD-L1 positive, those patients can get immunotherapy alone, and we use that quite often. But for PD-L1 low or 0, so low being 1% to 49%—or 0 is of course no PD-L1—those patients usually, depending on their tumor type, squamous or nonsquamous, would get immunotherapy either with carboplatin or pemetrexed, that being more for the patients with nonsquamous. And for those with squamous, usually it’s platinum plus a taxane of sorts.

We’re a pretty early adapter here. We’ve been researching PD-1 [programmed cell death protein 1], PD-L1 as part of our score actually at Yale University. Lieping Chen is 1 of the first discoverers of PD-L1 if not the first. So we actually have been testing it for over 3 years now. We get that result and we use that to guide us. All patients can benefit from immunotherapy, even those who don’t have PD-L1, but we would usually get that result and we get it early on in the treatment’s course for a patient.