In a Targeted Oncology case-based peer perspectives live discussion, Nathan A. Pennell, MD, PhD, associate professor of the Department of Medicine at Case Western Reserve University and a medical oncologist at Taussig Cancer Center and the Cleveland Clinic Cancer Center discussed treatment options for EGFR-mutant non–small cell lung cancer, based on the case of a real patient.<br />
Nathan A. Pennell, MD, PhD
In a Targeted OncologyTMcase-based peer perspectives live discussion, Nathan A. Pennell, MD, PhD, associate professor of the Department of Medicine at Case Western Reserve University and a medical oncologist at Taussig Cancer Center and the Cleveland Clinic Cancer Center, discussed treatment options for EGFR-mutant nonsmall cell lung cancer, based on the case of a real patient.
TARGETED ONCOLOGY:Would you initiate immunotherapy before molecular panel test results come back?
PENNELL:This is a patient with metastaticEGFR-mutant lung cancer [who] does have a brain metastasis and a good performance status.
Let’s say they have a 5-pack-year smoking history back in college but haven’t smoked in the last 25 years. PD-L1 status comes back at 55% expression [on tumor cells] and the [notation] on the bottom says “hot spot panel pending.” The patient has a lot of pain. They’re sick; they’re losing weight; you’re worried about them. Would you act upon this information?
You know what you’re supposed to say, but this is a real scenario that a lot of people are going to be confronted with. It’s just an IHC [immunohistochemistry] test for PD-L1. It’s almost always going to come back first, before anything else, but lots of [clinicians] are going to be confronted with a sick patient with a high PD-L1 and no genetic testing.
I don’t know what the timing is of the [onset of] pneumonitis, how fast it happens, and how much they got. We do know that the half-life of pembrolizumab [Keytruda] is months. Even 1 cycle could potentially put them at risk. My personal belief is, I probably would just give them chemotherapy and leave it out. Then, if the mutation testing came back negative, I’d just add immunotherapy with cycle 2, but I think most of the time you’re going to want to wait.
Some people are so sick that they need to be inpatients. I usually don’t even give chemotherapy, but for small cell lung cancer [SCLC], we do that all the time.
Because we test so well and quickly, there are not that many patients for whom you are thinking, “I can’t even wait 1 week to get treatment.” Most of the time, when you schedule them for chemotherapy, it’s going to take a week to get them scheduled anyway, so it’s not a big deal.
This patient has anEGFRexon 19 deletion. We know immune checkpoint inhibitors don’t work as first-line therapy inEGFR-mutant disease. That is an important point. Even with 100% PD-L1positive,EGFR-mutant tumors, it is very unlikely to respond to a checkpoint inhibitor. It is not being driven as a reaction to local cytokine release from an inflammatory response. It’s a constitutive upregulation of everything throughEGFRupregulation. It’s not PD-L1 in a meaningful biomarker way.
There is a single, asymptomatic brain metastasis, no significant edema, andEGFRexon 19 mutation.
Would you refer the patient for gamma knife radiation or would you treat with osimertinib (Tagrisso), rescan in a couple of months, and reserve gamma knife for later?
I’ve [heard varying] comments about this from radiation oncologists. If [the disease is] progressive, they may say the risk of further adverse effects to the patient can [be detrimental] by giving a larger gamma knife target if [there is no response, and that this is] mistreating the patient. It is really interesting. I don’t mention it to them until they’ve progressed, and then I refer. But the response rate in the brain is basically identical to the systemic response to osimertinib.
Discuss how the FLAURA trial data apply in this case.
This is a frontline trial inEGFR-mutant nonsmall cell lung cancer. Only exon 19 and 21 were examined, which is about 90% ofEGFRmutations. Patients could be treated with [standard of care therapies of] erlotinib [Tarceva] or gefitinib [Iressa] or [the experimental agent] osimertinib. This was the randomized phase III study.1
Osimertinib was approved based upon a dramatic improvement in median progression-free survival [PFS] with osimertinib, 18.9 months versus 10.2 months [HR, 0.46; 0.37-0.57;P<.0001],2which is right on with what you would expect with the first-generation tyrosine kinase inhibitors [TKIs]. These data have been updated, and they also have a statistically significant improvement in overall survival [OS; 38.6 vs 31.8 months; HR, 0.799; 95.05% CI, 0.641-0.997;P= .0462].3
Some of the arguments raised against preferential use of osimertinib are about sequencing. What are your thoughts on sequencing it rather than using it in the first line?
If you give a first-generation TKI in the front line, then when they progress and they haveEGFRC797S, you [cannot] just give them osimertinib later on.
Of the people who went on to second-line therapy, close to 50% [in the standard TKI arm] went on to osimertinib. This is actually a good real-world [example of]…what would happen if patients go on [to subsequent therapy]. That’s about as many as you would get. But 30% of patients got no subsequent therapy [after] first-line therapy, so that’s the rescue run of not using it first. It’s also better tolerated.
What would happen if this patient received an immune checkpoint inhibitor instead of an EGFR TKI?
This is the trial that was presented at the American Society of Clinical Oncology Annual Meeting a couple of years ago. These were patients withEGFR-mutant tumors who were treated in a phase II trial of single-agent pembrolizumab and not an EGFR TKI. Seventy-three percent of the patients on the trial had high PD-L1 expression, and only 1 patient responded to therapy. Retrospectively, that patient didn’t have anEGFRmutation. There was a 0% response rate even though most of the patients had high PD-L1 [expression]. In fact, several of the people died without going on to a TKI. That’s the risk you run. Do not treat based on the PD-L1 until you get the mutation [analysis] back. Yes, [you may try it for] second-line treatment, but it just doesn’t work for a while as a single agent.4CASE (continued)
Following osimertinib in the first line, there is disease progression. Would you rebiopsy?
You could, but what would you do with that information? Let’s say you got anEGFRC797S, the most common resistance mutation. What would you use then? They’re progressing, so you’re probably not going to continue on osimertinib regardless.
There are case reports of people going back to first-generation EGFR inhibitors, but overall, the first-generation inhibitors are not broadly active against that mutation. That’s not something you’d probably do. [Compared with] the 50% to 60% of people who getEGFRT790M following first-generation inhibitors, it’s only about 15% to 20% of people who get a secondary mutation after osimertinib, so it’s a smaller group.
I would say the 1 reason to biopsy is that about 10% of people have transformation to SCLC. I’ve now had about 10 of these, so it’s real, and they do respond to chemotherapy. You can’t pick that up on the blood test. You have to do a tissue test in order to do that.
I’ve only had 1 outpatient that transformed to SCLC. I’ve heard of other people who’ve had it. I think it’s probably just a reflection [of the fact that] thatEGFRmutations are 3 times as common asALK, and therefore we hear a lot more about it. I wouldn’t be surprised if that happens in any of the oncogenic driver genes.
Let’s say you do the biopsy and it’s not SCLC, but they do haveEGFRC797S. How would you treat this patient?
We’re going to go to carboplatin, pemetrexed, and pembrolizumab after all. All patients with driver oncogenes, although most of them wind up beingEGFRpositive, who progress after TKI [are going end up with this treatment]. But generally, there is not much out there in terms of data for this. So definitely chemotherapy is the right choice.
Then we had that subset [of patients] from the IMpower150 studythose patients who received EGFR or ALK TKIs that were on that trial had failed prior therapy, so it wasn’t first-line.5All of those patients had had a prior TKI, so I think that suggested that at least the PFS and maybe OS were better with the addition of immunotherapy [of atezolizumab (Tecentriq) to chemotherapy]. I think carboplatin, pemetrexed, and pembrolizumab or carboplatin and paclitaxel would be reasonable in this group. But there are some people who don’t believe that small subset [and say] “Just do chemotherapy.” I can’t say that’s wrong.
Do you continue or stop the TKI?
Well, if you’re going to give immunotherapy, you can’t use a TKI, because there is a risk of toxicity. However, I just had a conversation…about a patient withROS1and whether to give him immunotherapy or continue on priorROS1-targeted agents. You know, the consideration is that this patient has brain metastases, and chemotherapy doesn’t work well against brain metastases. If you had central nervous system control with your agent and they were systemically progressing, I know some people might suggest continuing the TKI just for brain control [during] chemotherapy.
There is a trial in patients withEGFR-mutant disease in which patients who progressed on a TKI were treated to see whether therapy should be stopped to switch to chemotherapy or if they should continue it and add chemotherapy. This was with first-generation gefitinib. There was no difference. There was no improvement in survival with adding a TKI. Numerically, it was actually worse to continue the TKI than it was to stop it.6
I don’t think there are a lot of efficacy data to support continuing it after progression, but certainly [continue] in people for whom you’re really worried about brain progression if they have a lot of brain metastases that are well controlled. I don’t think it’s wrong to consider whether you get that covered by insurance to add chemotherapy and continue it or not.
Rechallenge with a TKI definitely has a better chance of helping if you’ve got a long interval between therapies. I’ve had patients withALK- andEGFR-mutant tumors who went on carboplatin and pemetrexed for 18 months or 2 years and then go back to a TKI after and had had good response.