Current and emerging strategies for patients with non-small cell lung cancer (NSCLC) harboring an EGFR exon 20 insertion mutation

DR PIOTROWSKA: Well thanks, Neil, for having me to talk about this, I think, interestingly and timely topic of current and emerging strategies for patients with non-small cell lung cancers harboring an EGFR exon 20 insertion mutation. So EGFR exertion mutations are an important entity to be aware of and recognize. And as you know, EGFR mutations are one of the conical driver mutations that we typically look for in non-small cell lung cancer. They were one of the first types of mutations in non-small cell lung cancers that were described back in 2004.

But what we’ve learned over the ensuing, gosh, 16 years now is that not all EGFR mutations in lung cancer are created equal. What we know is that most commonly the types of mutations that we see in non-small cell lung cancer include the exon 19 deletions and L858R mutation in exon 21. As you can see on the right-hand side, on the pie chart here, together these make up about 85% of all EGFR-mutant lung cancer. And these, of course, are well known to be highly sensitive to various different EGFR inhibitors. Currently, the preferred first-line strategy for patients with exon 19 and L858R is osimertinib, but there are many other TKIs and targeted therapies approved and in development for those mutations.

Exon 20 insertions make up about 4% to 10% of EGFR mutations in non-small cell lung cancer, which makes them actually the third most common type of EGFR mutation that we see. And so I think important to recognize and identify.

You can see on the figure here on the left-hand side of the slide here a schema of EGFR. And you can see that the exon 20 insertions occur within this area called the C-helix and the loop following the C-helix within exon 20. This is still within that critical tyrosine kinase domain of EGFR. And these occur between amino acid positions approximately 761 to 774. And for our audience here, you don’t have to memorize this. I’ll give you kind of a cheat sheet that you can refer back to on the next slide. But suffice it to say that these are a heterogeneous group of mutations. There have been over, I think it’s 50 now, exon 20 insertion mutations that have been described, with varying frequency as you can see here.

And if we look at EGFR exon 20 insertions as a driver in and of themselves, you can see that within the United States and China these seem to occur in up to 2%, or maybe a little bit more that 2%, of non-small cell lung cancers. And so I think it’s important to recognize these because this frequency is on par with some of the other rarer targets that we’re used to looking for, like ROS1 or RET or others. So again an important entity to be aware of and recognize.

I think for oncologists in practice this landscape of identifying EGFR mutations in lung cancer is getting increasing complex, and I often get emails and phone calls from colleagues, not just at my own institution but elsewhere, saying the NGS report came back with an EGFR mutation but I don’t know what this is. And so I’ve included this table here as really kind of a reference you can come back to. Again on the left showing the sensitizing mutations, most commonly the exon 19 deletions and L858R, the T790M resistance mutation which is in exon 20 but is highly sensitive to osimertinib, and then some of these rare but sensitizing mutations like L861Q, G719X, S768I, which are also important to be aware of but generally are treated with TKIs like afatinib, and there’s some data for osimertinib now.

The exon 20 insertions are shown here on the right. As you can see, there’s a number of different insertions and duplications that we see, and these can be challenging to identify. Many NGS reports now will say this is an exon 20 insertion, but what I’ll say is if you see a sequence like this and you’re not sure what it is, it’s very fair to reach to the pathologist who sent out that report to the company and say hey, what is this. It’s also fair to reach out to myself or someone else in the community to say I’m not sure what to do with this EGFR mutation because these can be quite complicated. But I hope this list will serve as a reference.

Why are exon 20 insertions so important to identify? Historically these have been resistant to first- and second-generation EGFR inhibitors. They have been thought to be difficult to target. And part of the reason has to do with the structural properties of the exon 20 insertions. The cartoon here on the left kind of depicts that, but just to simplify it I’ll say that the specific insertion at exon 20 will lead to steric hindrance. It pushes that C-helix into an active confirmation in a way that leads to steric hindrance, which makes it difficult for first- and second-generation EGFR inhibitors to inhibit that EGFR mutation.

We can see the same here on the viability plot on the right, where you can see a cell line expressing a number of different exon 20 insertions. You can focus your eyes on the yellow, gray, and blue dots here. You can see that much higher doses of erlotinib are required to inhibit these cells lines than they are for the exon 19 deletion shown in black.

And so in the clinic what this translates to is clinical resistance of patients with EGFR exon 20 insertions historically to first- and second-generation EGFR inhibitors. Generally the median PFS of patients with EGFR exon 20 insertions has been less than 3 months, and you can see that on the waterfall plot on the right here. I’ll just briefly point out that the FQEA insertion at position 763 is kind of unique among the exon 20 mutations because it occurs very early in the exon, and it is sensitive to first-, second-, and even third-generation EGFR inhibitors. But for the vast majority of these chemotherapy remains our current standard of care. I think the role of immunotherapy and chemoimmunotherapy for these patients is still emerging. This is a relatively rare group, and so we don’t exactly know yet how important immunotherapy will be for these patients.

But importantly, and what will be the focus of the rest of today’s talk is really the fact that there are some new and exciting emerging targeted therapies for the exon 20 insertions in EGFR that are important to be aware of.

So here I’ve summarized some of the new and exciting emerging targeted therapies for EGFR exon 20 insertions. You can see here the number of different oral EGFR inhibitors, poziotinib, mobocertinib, osimertinib, and CLN-081 are all oral drugs. Amivantamab is a new and exciting compound that’s an EGFR and MET bispecific antibody. And I’ll go through the data for each of these drugs in a little bit more detail on the following slides.

So starting off first with poziotinib. This was one of the first oral EGFR inhibitors tested for the exon 20 insertions in EGFR. Poziotinib was selected based on structural modeling and preclinical models showing that it had activity against the EGFR exon 20 insertions. And what we saw in an early Phase II single-institutional investigator-initiated study out of MD Anderson was some early promising efficacy results, with a response rate of around 40%.

This led to a larger Phase II study called the ZENITH trial, which I’ve shown you the results of here. ZENITH enrolled over 100 patients with EGFR exon 20 insertions, and suffice it to say that the results of the ZENITH trial have been fairly disappointing, with a response rate to poziotinib among these exon 20 patients of about 15%.

And we think some of that limited efficacy may have to do with the safety profile of this drug, which is summarized in the table on the right. You can see that 82% of these patients had diarrhea of any grade, including 26% Grade 3 diarrhea. 68% of these patients had rash, including about a third of them having Grade 3 rash. Stomatitis and paronychia were also common. And all in all 78% of patients required dose reductions due to toxicity.

And so we believe that perhaps some of that dose exposure may have limited the efficacy of poziotinib in this population. And while these studies are still ongoing, and the company is evaluating different dosing strategies, I would say overall enthusiasm for poziotinib has fallen somewhat, particularly in light of some of the other therapies that are now emerging, as we’ll go through next.

So next to mention mobocertinib, another oral EGFR inhibitor, formally known as TAK-788, also designed and developed specifically to inhibit EGFR exon 20 insertions. In a Phase I/II study mobocertinib was found to have a recommended Phase II dose of 160 mg daily, and that study included a cohort of 28 patients with EGFR exon 20 insertions, for which the efficacy data is shown here. You can see that these patients had a confirmed response rate of 43%, and as you can see in the swimmer’s plot on the right some of the responses were quite durable, with a median duration of response of 14 months and a median progression-free survival of 7.3 months.

Based on these data, mobocertinib has actually received breakthrough therapy designation by the FDA, and it’s continuing to be explored in clinical trials. I think it’s again important to note some of the toxicities of mobocertinib, which like poziotinib has caused a fair bit of particularly GI toxicities and some rash. Again, here you can see 85% of patients having any-grade diarrhea and 18% of patients having Grade 3 or higher diarrhea, and 36% of patients having rash, although less Grade 3 or higher rash in this study. All in all, about a quarter of these patients required dose reduction due to an AE, and 14% discontinued treatment due to AEs. So while we do see some activity, I think it is important to note that this is a drug with a notable toxicity profile as well.

It’s also important to highlight that there have been several cases of cardiomyopathy seen with mobocertinib. And indeed now an ongoing Phase III study is evaluating mobocertinib versus chemotherapy as a front-line treatment strategy for patients with EGFR exon 20 insertions, and cardiac surveillance is being included in part of that study. So we’ll see what those outcomes look like. But again, I think an exciting and important drug to be aware of for these patients.

Moving on then to amivantamab. This is a drug which is distinct from the first 2 we discussed because it has a different mechanism of action. This is a bispecific antibody targeting both EGFR and MET. This drug, which is also known as JNJ-372, has shown activity not just in the exon 20 patients, but also in the more classic EGFR mutations, including in the setting of osimertinib resistance. And in the Phase I study of amivantamab, which is called the CHRYSALIS trial, which established the recommended Phase II dose of this drug as 1050 or 1400 based on weight administered IV weekly. 39 patients with EGFR exon 20 insertions were included. And at ASCO we saw the results of that EGFR exon 20 cohort with a response rate of 36%, a median PFS of 8.3 months, and again, as you can see in this spider plot here on the right, some durable responses.

This is another drug that has received FDA breakthrough therapy designation specifically for the EGFR exon 20 population. But again I think it’s important to talk about toxicities with these drugs. Here, 72% of patients had any-grade rash, so a common toxicity similar to others that we discussed. And then unique to this drug is 60% of these patients had infusion-related reaction.

We saw that actually the majority of these were Grade 1 or 2 mild infusion-related reaction.

They predominantly occur with the first infusion, cycle 1, day 1. Symptoms include chills, shortness of breath, nausea, flushing; pretty standard infusion reaction kind of side effects. And what we found so far is that this seems to markedly improve with split dosing, so the first dose being given over cycle 1, day 1 and day 2 and does not seem to recur with subsequent infusions. And so I think this is an important toxicity to be aware of to warn patients about what seems to be quite manageable so far, and I think, again, amivantamab is a drug that I’m really excited to see move forward, and I look forward to seeing more data in the exon 20 population.

Next I wanted to touch on osimertinib, which of course is an important drug in this context because it’s widely available. It’s FDA approved for exon 19 deletions and L858R, and I often get the question well how does this work for the exon 20 patients. What we know from preclinical models, and you can see it in the figure here, is that osimertinib has kind of intermediate activity compared to other EGFR inhibitors against exon 20 cell lines. But a key feature of osimertinib is that it spares wild-type EGFR, which means that it has a wider therapeutic window. It can potentially be dosed at higher levels in patients.

And based on these data we designed the ECOG status-ACRIN 5162 trial through the cooperative group system, which is a Phase II study of osimertinib given a dose of 160 mg daily, so twice the standard dose of 80 mg that we usually use in exon 19 deletion and L858R patients. And this was a study for patients with EGFR exon 20 insertion mutations who had to have at least 1 prior line of therapy.

We presented the preliminary results of EA5162 at ASCO and we saw, as you can see on the waterfall plot here, 4 out of 17 confirmed responses, so response rate of 24% in this admittedly small cohort, and a median progression-free survival of 9.6 months. Generally osimertinib 160 was fairly well tolerated. We did see some diarrhea, fatigue, thrombocytopenia, but Grade 3 toxicities were very rare in this cohort, and generally osi 160 was fairly well tolerated.

And so this is I think a strategy which again holds promise in this small cohort of patients, and we are hoping to expand the study to further evaluate the efficacy of osimertinib in this population.

In the last few slides a few more new and emerging therapies that are important to be aware of. First is CLN-081, also known as TAS6417. This was another oral drug. As you can see, there’s lots of new oral therapies being developed specifically for EGFR exon 20 insertions in this day and age. And CLN-081 is certainly an important one to be aware of among them. It’s currently still in Phase I testing and dose escalation. And preliminary data has shown that it does cause some rash, stomatitis, dry skin, but no dose-limiting toxicities seen so far.

And as you can see, there have been responses seen; 6 out of 17 patients have had a response, although not all of these have been confirmed in this early report. And importantly we’ve seen some patients who had previously been treated with other TKIs, including mobocertinib, poziotinib, and osimertinib who then responded to this drug. And so I think that’s exciting to see, and we look forward to seeing more data with CLN-081 in the future as the study continues to accrue patients.

Finally, I wanted to mention tarloxotinib, which is an interesting drug with a unique mechanism of action. It’s hypoxia-activated TKI, which is thought to have more specific activity within the cancer cell environment based on its hypoxia-activated mechanism. Again, preclinically this was a drug that was predicted to have activity in EGFR exon 20 insertions, and I think this is an example of a drug where we had hoped it would have good preclinical activity, but unfortunately so far that has not borne out in the clinic.

At ESMO we saw some preliminary results with tarloxotinib in the EGFR exon 20 space, and unfortunately there were no responses in 11 patients with EGFR exon 20 mutations. There were some responses seen among HER2 patients, and that is where this drug is being developed right now going forward.

And then finally BDTX-189 is an oral allosteric ErbB inhibitor which is currently in Phase I testing, and we look forward to seeing data with that drug.

So I hope that this whirlwind tour of some of the new and emergent therapies for EGFR exon 20 patients has shown you that there are some signs of promise for these patients. We now see that there are both oral TKIs, mobocertinib, osimertinib, and some others, as well as antibody-based therapies which are in development, and some new drugs coming down the pipeline as well. I think it’s fair to say though that although we see good signs of activity so far, these are certainly not the outcomes that we’re used to seeing for patients with exon 19 deletions and L858R, and we really do need to do better. We need even better and more effective therapies for these patients, ones that not only yield higher response rates, but hopefully yield more durable responses as well. And I think a key issue for drug development going forward will really be to optimize the therapeutic window of these drugs, minimize safety concerns. As you can see, many of these drugs have more toxicities than we’re used to seeing for some of the drugs, like osimertinib in the exon 19 and L858R space. It will be important to identify the optimal sequence of these therapies. Perhaps these drugs can be used in sequence to induce more durable responses and prolong survival for these patients. And then always critical is CNS benefit for these patients. It remains to be seen which of these drugs will be active in the CNS and how we can utilize them there.

Case: A woman in her mid-50s with metastatic NSCLC and an EGFR exon 20 insertion receives mobocertinib as second-line therapy on a clinical trial

DR PIOTROWSKA: I will move onto a couple of cases that I had from my clinic that I thought would be useful to illustrate how we use some of these therapies.

So the first case is a patient from my own clinic. This is a 55-year-old woman, a never smoker, who presented months of cough and right-sided pleuritic chest pain. She had no significant past medical history. She was really quite healthy and was of course surprised to hear that she had a large right lower lobe lung mass encasing the right lower lobe bronchus. By PET scan she not only had this right lower lobe lesion, as well as bilateral pulmonary nodules, significant adenopathy extending up to the cervical lymph node chains, and some bone mets, including the lumbar spine and ribs. Brain MRI here was negative.

She underwent bronchoscopy and endobronchial ultrasound with biopsy of evident endobronchial disease showing a TTF1-positive adenocarcinoma. The patient PD-L1 score here was 0, and next-generation sequencing revealed this V769_D770 insertion of ASV, which is an EGFR exon 20 insertion. And so this patient was referred to me with the question of what should the optimal first-line therapy for this patient be.

This patient was referred to me at a time when I did not have a clinical trial available for her. Today I would, if I could, enroll her on the first-line trial of mobocertinib versus chemotherapy, which I mentioned and I think would be a great choice, but for her I did not have this option. And so we discussed chemotherapy. We discussed chemo and immunotherapy. We ultimately ended up selecting carboplatin/pemetrexed chemotherapy for this patient.

We did not use immunotherapy for a couple of reasons. One was the low PD-L1 score, and her never-smoking history, which made me think the benefit of immunotherapy may be a little bit less, particularly in light of the EGFR mutation. And secondly, and importantly, because as I showed you, TKIs like osimertinib may have some activity for these patients and are sometimes something that I will reach for in an off-label fashion if I don’t have a clinical trial available. We know that TKIs such as osimertinib have higher rates of autoimmune toxicities, and in particular pneumonitis, when used after immunotherapy when being used, for example, after carboplatin, pemetrexed, and pembrolizumab in this case. So for those reasons we decided to treat her with carboplatin/pemetrexed alone and reserve immunotherapy for potentially a later line of treatment.

So this patient went on to receive carboplatin/pemetrexed. She had an initial response to treatment. Her right lower lobe lung mass shrank. Her bone mets became sclerotic, but unfortunately after about 9 months on the carbo/pem, she got 4 cycles of doublet followed by pemetrexed maintenance, and then she developed progression. She developed an increased right lower lobe lung mass. She had developed a new 3 cm liver lesion, as you can see, and here again her brain MRI was negative.

At this point, I was able to explore clinical trial options for her. What we talked about, what we had at the time, was the mobocertinib Phase II study, which was enrolling patients with EGFR exon 20 insertions, and I thought this would be a good option for her. We also talked about some of the preliminary activity that had been seen with osimertinib, and that is an option as an off-label indication.

She opted to go onto the mobocertinib trial and did actually quite well on that. She did have a partial response. She had a nice shrinkage in her lesion. She did have some toxicities, consistent with what we saw in the trial and what we discussed earlier, with Grade 2 rash and some diarrhea. But overall she had a fairly good quality of life. She was happy to be taking an oral therapy for the time that she was on treatment.

And unfortunately after about 8 months she developed progression once again, with both CNS and systemic liver and lung progression at that point. So here, once she progressed there again, we talked about several options. We talked about chemotherapy. We talked about immunotherapy as an option. And here we talked actually about really the dearth of data for the EGFR exon 20 insertions and how they do with immunotherapy.

There’s really been very little data looking specifically at the exon 20 insertions in particular. Of course we know that as a whole category EGFR mutations don’t respond very well to immunotherapy, particularly with monotherapy, and we presume that the exon 20 insertions fall into that same category. But I think this is an area where we need more options. And so she, at this point, ended up getting her brain mets radiated and going onto docetaxel chemotherapy as her next line of treatment.

Case: A man in his late 60s with advanced NSCLC and an EGFR exon 20 insertion receives amivantamab as second-line therapy on a clinical trial

DR PIOTROWSKA: The next case I wanted to discuss was another patient from my clinic. This was a 67-year-old male former smoker, 20-pack-year smoking history, quit about 10 years ago. He had a history of hypertension and hyperlipidemia. He presented initially with dyspnea and was found to have a large left-sided pleural effusion.

On chest CT he was also found to have a 1.8 cm left upper lobe mass. He had extensive pleural nodularity in the effusion, which was tapped. And on thoracentesis the pleural fluid cytology showed an adenocarcinoma. In this case, his PD-L1 score was 60%. And again, the next-generation sequencing also identified a different EGFR exon 20 insertion. This one at position 772.

This patient, we had a similar conversation to the one with the first patient, and we decided again to treat him with carboplatin/pemetrexed. After an initial response and 6 months on treatment, he progressed, with an increased pleural effusion. And in this case, actually, we were able to enroll this patient onto amivantamab as a second-line therapy.

He had stabilization of his disease for about 7 months on amivantamab. He did have a classic Grade 1 infusion reaction with the first treatment, but that did not recur, and some manageable rash on amivantamab. But unfortunately when he progressed again also had multiple brain metastases, as you can see in these figures here.

So I thought this was a good case to illustrate the fact that brain penetration is really an important consideration with these therapies and something that we’re going to really have to look at carefully with these new and emergent therapies. We really don’t have much data yet on the CNS penetration of oral TKIs like mobocertinib. There was some data with mobocertinib presented at I believe it was ASCO, where they showed the response to be a little bit lower when you looked at patients with brain metastases at baseline than those without. But we haven’t seen a great deal of data about CNS penetration with that drug. With amivantamab, again we really don’t know yet, but with a large antibody-based therapy we worry that that may not have as much CNS activity.

So in this case, actually, I thought it would be worth applying for off-label use of osimertinib. Osimertinib has good CNS activity in the setting of exon 19 deletion and L858R patients, of course, and so in that context we will often use it for CNS penetration. In this situation, this patient did have these 3 brain metastases treated by stereotactic radiosurgery, but thereafter we applied for him to get off-label osimertinib, and that is the therapy that he’s just recently started.

So with that I’ll stop, and move onto the discussion.

Activity of mobocertinib in patients with advanced NSCLC and EGFR exon 20 insertions in the first-line setting; mechanism of action of amivantamab, a bispecific antibody targeting EGFR and MET

DR LOVE: Let me just begin with some follow up questions.

DR PIOTROWSKA: Sure.

DR LOVE: So on your first case you brought up the possibility of putting the patient on a trial of a TKI, and I just kind of wonder whether or not, I don’t know how symptomatic the patient was at that point, but I wonder whether or not the response rates are high enough to justify using it first line at this point.

DR PIOTROWSKA: It’s a great question.

If we look at mobocertinib, which is the TKI that’s in the first-line trial right now, the response rate that we saw so far was about 40%, which I think is actually probably comparable to what we might expect with chemotherapy. It’s interesting, I think with some of the newer targeted agents, the RET-targeted therapies for example, there’s a lot of debate as to whether we need a confirmatory Phase III study to compare to chemo because the response rates are so high, and the duration of response is so high. I think with these exon 20 drugs, actually these comparative studies are going to be really important to tell us what is the best first-line strategy because the outcomes aren’t necessarily dramatically different between chemo and TKIs.

DR LOVE: Another question about your second case, and also questions in general about this bispecific. So this is bringing MET and EGFR together, but when we hear about bispecifics in hematologic oncology you hear about cytokine release and neurologic syndrome because they activate T cells. I gather that’s not what happens here?

DR PIOTROWSKA: No. This is a monoclonal antibody just with 2 targets, targeting MET and EGFR. So I think it’s different than like the BiTE kind of therapies that they’re using in the hematologic malignancies. In this case, actually, it’s been developed because we know MET is an important resistance mechanism in the setting of EGFR, and so the thought is targeting both of those at the same time may prevent that development of MET as a bypass pathway.

DR LOVE: But it’s like a bispecific antibody? It just presents 2 things to the cell. Is that the way to think about it?

DR PIOTROWSKA: I think so, yeah. It’s presenting both the MET and the EGFR, both of them.

DR LOVE: Another thing that you mentioned with the second patient that you were thinking about, or the patient was going to go onto docetaxel. And I was curious about in general whether you use ramucirumab in that situation, and your general thoughts about antiangiogenics in exon 20 because I have heard, John Heymach for example, has talked about the biologic basis and the fact that anti-angiogenesis should be very effective in the classic EGFR situation. What about in exon 20?

DR PIOTROWSKA: I don’t know yet. I think I am someone who believes that antiangiogenics probably will have some role in the classic EGFR mutations. I think we haven’t yet learned exactly what that role should be. Should they be combined with first-line TKIs for example? Or is it going to be in later-line setting in combination with chemo? But there certainly seems to be some magic there. I haven’t seen much data. I think one of the things to know about exon 20 mutations is these trials have only really been going on for a few years. Prior to that, we really didn’t recognize these as a unique target that was being studied. So I think that will be an important question for us as a field to look at going forward.

Specifically for that second case, for a patient like that, where if the brain mets can be treated, and I think the patient is eligible for a VEGF targeted therapy, that patient in particular was young and had a good functional status, I think it is reasonable to consider docetaxel/ramucirumab combination for patients where we think that they’re able to tolerate that increased toxicity and don’t have a contraindication to the VEGF therapy.

Cardiac toxicities associated with EGFR tyrosine kinase inhibitors

DR LOVE: When you were talking about mobocertinib you brought up the issue of cardiomyopathy, and I was wondering if you could just talk a little bit more about what known about cardiac effects of TKIs in general. Is this specific to this agent or do you see some of that with osimertinib and the other TKIs?

DR PIOTROWSKA: It’s a really important point, Neil. I think it’s a really important point to be aware of. So we have seen this. So osimertinib in particular can have some cardiomyopathy associated with it. I think of the effect of osimertinib on the heart in 2 categories. One is it can certainly cause QT prolongation, and that we’ve known about. And in my practice I will often try to make sure that a patient has a baseline EKG when I’m starting osimertinib. I don’t routinely monitor the EKG, but I will if I’m adding any QT-prolonging meds, antibiotics, antidepressants, things like that I will check another EKG to monitor.

The other category is cardiomyopathy, decrease in EF. And that’s rare. It’s only seen in a few percent, I think it’s just about 1% to 2% of patients treated with osimertinib. But it can happen, and it’s really important to be aware of because when it happens it can be profound and cause highly symptomatic CHF. And I think the label now, the package inserts, recommend monitoring only in patients who are at increased risk of cardiac complications.

Again, when I’m starting patients on osimertinib I generally think is this someone who’s at risk? Do I have an echo in the system? Do I know what their EF is? If not, I might order it. And then I just keep that in my differential diagnosis if a patient’s coming in with dyspnea, leg swelling, things like that, and I certainly will have a low threshold to repeat the echo. And if patients do have a clinically significant drop in ejection fraction I would be very wary to rechallenge them, with osimertinib in particular. This cardiac toxicity seems to be more common with osimertinib that we saw, for example, with erlotinib, gefitinib, some of the first-gen drugs.

DR LOVE: So I noticed your second patient was a smoker, and I was curious about the phenotype of the exon 20 patients. What fraction are smokers?

DR PIOTROWSKA: It’s generally pretty similar to the standard EGFR mutations. The minority are smokers. It can occur, just like the standard EGFR mutations can occur in former light smokers and even in heavier smokers rarely, but in general I would think of the phenotype of these patients to be similar to the standard mutations.

Efficacy of osimertinib in patients with NSCLC and EGFR exon 20 insertions

DR LOVE: Another thing that you said that was very interesting was the idea that some of these exon 20s are actually sensitive to first-generation TKIs, osimertinib. Is that generally put on the NGS report or do you have to kind of figure that out?

DR PIOTROWSKA: I think you really bring up a really important point here. I think it varies on the NGS reports. Some of them will get to the granularities of saying oh this is a FQEA insertion, this a sensitizing one, but I think many don’t. And I think in some NGS reports they don’t even tell you it’s an exon 20 insertion, and that’s where this gets really complicated. And for a general oncologist in community practice I think this is a huge challenge, and a real challenge to interpreting NGS reports. And so I think the onus is on the companies issuing these reports to try to be as transparent as they can be and not only saying this is an exon 20 insertion but helping to identify those rare ones that do respond differently.

Another key point is these mutations are very variable. They’re heterogeneous, even the exon 20 insertions, and we don’t yet know whether maybe these will respond differently, a more proximal one versus a more distal one, whether those may have different responses to treatment. And as we get more granular, as we get that kind of information, I think these reports are going to have to be very transparent and very clear on what the best treatment for this patients is based on what we know so far.

DR LOVE: I’ve got to say, we’ve talked about exon 20 with a number of people, and that’s the first time I’ve heard that, and I find it a little bit scary that we might walk by that. What about the mutations that afatinib is approved to? I realize they’re not exon 20, but do you use afatinib at all in the first-line situation? For example, for those mutations? I guess we don’t know very much about osimertinib with those mutation.

DR PIOTROWSKA: We’re starting to get some data. So the mutations you’re referring to, the G719 mutations, the S768I, which actually is a mutation in exon 20, and I’ve seen that sometimes be a point of confusion in patients I’ve seen in clinic, and the L861Q. Afatinib is the FDA-approved therapy, that’s the drug that’s approved for these mutations, but we are now starting to see emerging reports with osimertinib, which seems, at least from the data we have so far, to be comparable. So I think based on the safety profile of osimertinib I’ve started to use it more and more for those patients as well.

Therapeutic approach for patients with NSCLC and HER2 alterations

DR LOVE: So here’s a question I have asked a bunch of people, except so far I haven’t exactly understood anybody’s answer. So maybe you can explain it to me. Because I’m coming from the perspective of pragmatics, and one of the things that you hear a lot about, particularly this past year, is HER2. So particularly, of course, trastuzumab deruxtecan, and so can you take a shot at explaining to me the biology of HER2 biologically and how it overlaps with EGFR. And really for the most important practical point of view what kinds of patients with lung cancer would you think might respond to an anti-HER strategy, for example exon 20, just as an example. Is that connected to HER2 at all?

DR PIOTROWSKA: Yes.

DR LOVE: I’ve heard people talk about exon 19 insertions, which I don’t know what those are exactly, but I’ve heard people say that. So okay, take a shot.

DR PIOTROWSKA: It’s a very complicated question. In general I would say there is some overlap. Of course HER2 and EGFR are part of the same family. They’re all part of the HER family, so that’s why there is overlap. And there are some structural and biochemical similarities between EGFR exon 20 insertions and HER2 exon 20 insertions, but it is critical to know that these are not the same thing.

And in fact I saw a patient in clinic a few weeks ago that was referred to me for EGFR exon 20 clinical trials, but when I read the report the patient had a HER2 exon 20 insertion. And the referring doctor, I think very appropriately, said well I thought all these exon 20 insertions were the same. And so I think the key practical point is not all of these therapies are going to work for both.

But in general if you find a patient who has a HER2 alteration in the 2 broad categories — the most common ones are exon 20 insertion. The second most common are these transmembrane domain mutations that we see, like the exon 19 insertions and things like that, those are very rare. I think we can put those aside as uncommon. But if you see a HER2 alteration I think it’s reasonable to consider a HER2-targeted strategy either with, there’s some activity with T-DM1, there’s some activity with trastuzumab deruxtecan, and then some of these oral therapies are also being developed in the HER2 space. In particular poziotinib has trials ongoing with HER2. Tarloxotinib, which I mentioned, has studies ongoing with HER2.

And so if you see a patient with a HER2 alteration, if it’s a young never smoker, someone that you think probably has an oncogenic driver, it’s worth trying to treat that or refer that patient for clinical trials.

DR LOVE: But again, like the report would say what? A HER2 mutation?

DR PIOTROWSKA: Yeah.

DR LOVE: Or a HER2 amplification? Would it say amplification? I mean what would the report say?

DR PIOTROWSKA: So amplifications probably are a little bit trickier. HER2 mutations, a HER2 mutation, a HER2 insertion, I think those are things that are worth looking at and acting upon. And again, that might be a situation where you refer the patient for clinical trials targeting HER2, or you think about off-label use of some of these HER2 antibody-drug conjugates and other novel therapies.

DR LOVE: Any thoughts about trastuzumab deruxtecan? Have you used it? Again, I kind of just wonder where it’s heading in lung cancer.

DR PIOTROWSKA: I haven’t used it personally yet, but I look forward to using it when I have the right patient. These patients are relatively uncommon, so I haven’t been in that clinical situation yet. But I think the data looks really exciting. I mean the progression-free survival in that DESTINY lung study was I believe 14 months, really amazingly durable. And I think this is a drug that’s really — I hope that maybe it will get approval for lung cancer or at least that it will be available for off-label use. We’ve previously used T-DM1 for these patients, and I think that the data with trastuzumab deruxtecan looks at least as good if not better than T-DM1 for these patients.

DR LOVE: Yeah, the T-DM1 data, I don’t know if it wasn’t enough, but it kind of never seemed that exciting to me, but I don’t know. Have you used T-DM1 in a lung cancer patient at all?

DR PIOTROWSKA: Again, I haven’t used it, but I’ve seen patients in consultation who then have gone on to get it locally. I think you’re right, it’s not mind blowing, but at the same time we really have a dearth of HER2 options for these patients. And so I think any targeted therapy that gives us another option and something that’s targetable I think is worth considering. So that’s really the reason that I consider it. It’s not as exciting as putting an EGFR exon 19 patient on osimertinib, but it feels good to give these patients a targeted therapy and to have another option in our toolbelt.

DR LOVE: I mean to me and a lot of people the T-DXd waterfall plot looked pretty good, like really good.

DR PIOTROWSKA: Yeah. Yeah. I agree completely.

DR LOVE: They’re still trying to sort all that pulmonary toxicity.

Of course that’s going to really play out very differently in lung cancer than in breast cancer.

DR PIOTROWSKA: I agree, but I think again, pulmonary toxicity is something we’re used to seeing in lung cancer and something that we’re used to monitoring for. So I don’t think it’ll preclude us from using it, and I think in that small number of HER2-positive non-small cell lung cancer patients, the waterfall plot, the duration of response, they were very compelling. I’m really looking forward to seeing more data or at least starting to use it more clinically.

Role of immune checkpoint inhibitors for patients with NSCLC and targetable mutations

DR LOVE: So I want to bring up a couple clinical paradigms and see how they apply to exon 20, and some of these came up when you were presenting your cases. One is the issue of checkpoint inhibitors in patients with targetable mutations. Kind of where are you today in general with that? Oncologists in practice are getting a very strong message that they’re really paying a lot of attention to in delaying using checkpoint inhibitors regardless of PD-L1. Any comments in general? For example with the typical EGFR-activating mutations where you bring in checkpoint and how you’re thinking about it in exon 20.

DR PIOTROWSKA: That’s a great question. I’m glad to hear that oncologists are getting that message because I think in the front-line setting it’s really an important message. I think there is an impetus to put people on immunotherapy when we see high PD-L1 without necessarily always waiting for that molecular testing to come back. And I think there we know that that’s not the right thing to do because if the patient has an EGFR mutation you may be limited in your ability to safely give a TKI following that immunotherapy.

In my practice I’ve taken that data. I’ve taken the meta-analysis looking at the chemotherapy versus immunotherapy trials and the studies that showed that in EGFR patients even docetaxel was superior to immunotherapy monotherapy, to leave immunotherapy for later line, if at all, use. So many of my EGFR patients, I won’t use immunotherapy. I’ll try to keep them on various chemotherapies, TKIs, clinical trials. Sometimes if I’m really running out of options I’ll use it, but I would say often it’s something that I’m wary of because if want to put them back on a TKI after that I worry that I won’t be able to.

I think patients will often ask for that therapy. They want immunotherapy. They’ve gotten the message that this is something that can work well, and it takes a lot of conversation with them to explain that it might not work well for these patients.

For exon 20, I think it’s a little bit harder to be dogmatic about it because we have so much less data, and we don’t know yet how these new TKIs are going to interact with immunotherapy. We don’t necessarily know that those same concerns will apply. But as you saw in these 2 cases, my general bias, because we think that these patients are likely to go on to targeted therapies after their first line, because we don’t know how well immunotherapy works for them, I’ve haven’t used it in the front-line setting. I’ve used carboplatin/pemetrexed first. But I will be more willing to introduce immunotherapy in a later-line setting if I’ve exhausted my clinical trial options, for example.

DR LOVE: I’m curious, in trying to separate out the targetable mutations I can recall someone saying you do see benefit with patients with EGFR in some cases, that may be worth giving, but “you never see benefit with ALK.” Agree or disagree?

DR PIOTROWSKA: Agree. I think in ALK in particular, I’ve seen that there’s a preponderance of high PD-L1 in ALK. There seems to be something about ALK that activates PD-L1, and those patients tend to have high PD-L1 levels, but similarly very little activity. And really that’s not something I reach for for my ALK patients, even with high PD-L1, which is a common scenario.

DR LOVE: What about the patient with a history of smoking and a targetable mutation, for example BRAF, do you treat them differently?

DR PIOTROWSKA: Yes. I think BRAF, and probably even KRAS despite some of the emerging KRAS-targeted therapies, I think those are the 2 groups where I think about them a little bit differently. There was a nice series out of Europe called the IMMUNOTARGET Registry that looked at response rates to immune checkpoint inhibitors across different oncogenic drivers, which was a nice reference. And there we saw that really BRAF stood out as one of the groups with the higher response rate to immunotherapy.

And so certainly for a BRAF patient, particularly one with a smoking history, which is commonly the case with BRAF patients, I will use chemoimmunotherapy or immunotherapy more willingly than I would for, for example, a never smoker with EGFR-ALK.

Management of NSCLC in patients with actionable alterations who develop brain metastases

DR LOVE: So another clinical paradigm area I’m curious what your thoughts are about it, and again, not just about exon 20, but patients with targetable mutations in general, EGFR in particular, including exon 20, is brain mets. And we again see a real shift. I couldn’t believe it. I heard a case the other day where a patient had been sent to a radiologist, and hadn’t seen a medical oncologist, and the radiologist called the oncologist and said I think you better do some testing to make sure this patient’s a nonsmoker. So I was like woah man, if we’ve got to that point the message is definitely out. But there is a lot of debate about that, particularly in the untreated patient, whether to start targeted therapy and hold off on radiation. Do you approach it differently if there are a small number of lesions that can be treated stereotactically? What’s your current paradigm, in general, across the board, EGFR in particular, including exon 20, about brain mets?

DR PIOTROWSKA: So I think here when you talk about EGFR this is the space where you have to separate out the exon 20s from the exon 19s and the L858Rs because the exon 19s and the L858Rs are generally so exquisitely sensitive to a drug like osimertinib, which has good CNS penetration. For those patients, I think it’s always a decision that I try to make in collaboration with my radiation oncology colleagues. I think if patients have large or highly symptomatic brain mets we’ll usually treat those, of course, with the radiation up front.

But if they have smaller and asymptomatic or even minimally symptomatic brain mets we try to send the molecular testing as quickly as we can. If we find they have a classic EGFR mutation we will often put those patients on osimertinib and watch very closely, get an MRI after 4 weeks even, to make sure they’re headed in the right direction. And that can be reasonable.

I think here the exon 20 space is different because these new, even the TKIs that we have, we don’t know much about their CNS activity, but what we do know, frankly, is not that encouraging so far with mobocertinib, with poziotinib. I think they’re probably not going to have as much activity in the CNS as we’d like to see. And those patients, I would say, are more likely to treat with radiation, whether I’m considering a clinical trial or chemo up front.

DR LOVE: This paradigm of targeted therapy before radiation, how far does that extend into targeted therapy? I mean, for example, RET, MET, other alterations. Do you kind of apply the same paradigm?

DR PIOTROWSKA: I think it depends a little bit on the treatment and how CNS penetrant it is. So the new RET-TKIs, selpercatinib, pralsetinib, those have great CNS activity, and so there I would apply a similar paradigm. Some of the MET-targeted therapies we’re still, I would say, learning about. We know crizotinib doesn’t have as good CNS penetration, but capmatinib seems better. So I think you really have to take into account not just the patient and the target, but the CNS activity of the drug, and then also the size and what the radiation would entail, and make a decision together with radiation oncology to whether or not you basically have room, if they don’t respond, whether you’re going to lose the opportunity to treat them before patients become symptomatic or sick.

Development of resistance in patients with NSCLC and targetable alterations; use of liquid biopsies for monitoring patients

DR LOVE: Another broad clinical issue is the idea of repeat testing in patients with targetable mutations, particularly on progression, but also the issues of monitoring with liquid biopsy over time and how useful that may be. Can you kind of comment on that in general, and then more specifically as it relates to EGFR and exon 20?

DR PIOTROWSKA: Sure. I think in general while I find liquid biopsies to be a very helpful thing in a situation such as a patient that I’m meeting up front where they may not have enough tissue for molecular testing, I don’t want to wait for the full NGS to come back in a couple of weeks, I think liquid biopsies can be incredibly helpful, keeping in mind the caveat that if those are negative you can’t assume that the patient doesn’t have a target. And you have to remember that tissue really is the gold standard. The negative liquid biopsy could just be a situation where you didn’t have enough DNA to detect something.

I think serial monitoring is not something that I routinely do in the clinic yet. I don’t think it’s particularly clinically useful. Even if we see ctDNA going up, and the scans are stable and the patient’s going well, I don’t think that something that provides any clinically useful information right now, although that may change in the future.

I do think that at the time of resistance in the classic EGFR mutations, exon 19, L858R for patients progressing on front-line osimertinib, in that situation I try to get tissue biopsies when I can, particularly because of this issue of histologic transformation. So non-small cell lung cancers that can transform into a small cell carcinoma, or even now we’ve seen some cases where they take on a more squamous cell histology from a previously adenocarcinoma. And that, of course, you can’t pick up based on a liquid biopsy. If patients can’t wait, or can’t have a tissue biopsy, then I think liquid biopsy can be informative to look in particular for MET amplification for some of the tertiary EGFR mutations like C797S, which can drive resistance. Those may open up clinical trial options. So liquid biopsies can be useful in that setting.

In the exon 20 space, just lastly I will say we don’t really know yet. We’re still learning about how these therapies work in the first place. We have very little data about how resistance develops to drugs like poziotinib, mobocertinib, any of these. So right now I think we should be doing those studies on a research basis to learn. It’s an important question that is going to help us guide how we sequence these therapies, but clinically right now I don’t think they necessarily will guide what we do as next-line therapy for these patients. So I don’t use either tissue or liquid biopsies routinely in my exon 20 patients outside of a research setting, which is an important context that I do think we will need to do more in.

EGFR-mutated adenocarcinomas that transform to small-cell lung cancer and other neuroendocrine carcinomas

DR LOVE: I see that you and Lecia Sequist did a paper looking at histologic transformation in small cell, other neuroendocrine. What do you typically see with the driver mutations? Do you usually lose it? Do you keep the osimertinib going, or whatever you might be on? What goes on when you see a selective transformation? Do you usually lose the driver mutation?

DR PIOTROWSKA: It’s a great question. What we’ve seen so far is that these patients — so we’re talking about a patient, for example, who goes on osimertinib or even some of the older TKIs, and at the time of progression has a biopsy, was previously a non-small cell lung cancer, and adenocarcinoma, and now you’re seeing a small cell carcinoma. What we typically see in these is that they do still have the EGFR mutation. It’s still present. It’s a conical mutation that you can still detect on NGS. But interestingly what we’ve seen is that they lose protein expression of EGFR. They seem to lose dependence on EGFR in that setting.

So we typically treat these small cell transformed patients with platinum/etoposide chemotherapy. We have seen some interesting cases where later on down the road, if they progress again, if you get another biopsy, they can have the non-small cell, the adenocarcinoma clone, grow up again. And so we think that this is probably not that the entirety of the cancer is transforming into small cell, that there’s probably different clones at play. And so we will sometimes use osimertinib. For example, they get 4 cycles of platinum/etoposide, and then we’ll put them on osimertinib kind of as maintenance therapy. But that’s an area where we still need to learn a lot, and we don’t exactly know the optimal strategy for those patients in terms of the TKI specifically.

DR LOVE: I feel like maybe I’ve seen something about targetable mutations in small cell itself, which I thought really you don’t see, but do you see that?

DR PIOTROWSKA: Very rarely. You do sometimes see a never smoker with small cell, de novo small cell, with EGFR. I’ve had a handful in my practice.

DR LOVE: Wow.

DR PIOTROWSKA: The question, of course, becomes how do you treat them? And we think in those cases the small cell kind of predominates, and we treat them clinically like a small cell. But we do sometimes think about trying to incorporate the TKI at some point in their treatment.

DR LOVE: We were talking earlier about testing and the importance of making sure you get some driver mutation or NGS, particularly in a nonsmoker, but I think in everybody. And one thing I’ve been seeing, I’ve heard investigators talking about it, and then I started to see oncologists doing it, is in the patient who you really want to treat, and you just really feel uncomfortable waiting for your NGS or whatever, to give the patient chemo temporarily, but don’t give an IO, just keep them going. And then by the time your first cycle’s done you have back the NGS. Any comments on that strategy? And if you do that, and then you find a driver mutation, do you always switch, or do you ever keep the chemo going if the patient’s doing well?

DR PIOTROWSKA: Great question. I do do that. I’ve been a proponent of that strategy for patients who can’t wait. I think a lot of patients can wait. They may feel anxious, but they’re clinically stable, and they can wait for the molecular testing, but there are those rare patients where they’re really sick, and you can’t keep waiting. And I think using chemotherapy in that situation’s a great strategy.

What I will often do in that situation is if they’re clinically stable after having started the chemo I’ll restage them after 2 cycles, sometimes if they’re really anxious after 1, but I’ll give them 2 cycles, restage. If they’re doing well on chemo, what I might do is actually give them the 4 cycles of the doublet and then switch to the TKI kind of as a maintenance strategy. But it allows you, if their disease is not responding to the chemo, then you can make the switch to the TKI sooner.

Adjuvant therapy for patients with NSCLC and targetable mutations

DR LOVE: So again, I’m kind of just trying to sound you out on a few general mutation targeted therapy kinds of questions. And of course there’s always ADAURA out there, so I can’t go by this without asking you what you think about that. And also a question I’m hearing more and more from oncologists is, they’re kind of getting sold on it, they’re ready to go, but then they want to know well what about RET and ALK and everything else. We’re going to have 5 or 10 years for that. Any thoughts?

DR PIOTROWSKA: So I would say generally I’m an ADAURA believer. I’m in the camp of saying that I think that the data is compelling. It’s not perfect, and I certainly understand the viewpoint of those who want to wait for the overall survival data to come out. But I think that’s going to take years, and the disease-free survival benefit is so significant that I think it really is, to me, compelling enough to change my practice.

I’ve already been in the situation where I’ve actually discussed it in an off-label fashion with several Stage III patients, and they’ve wanted to take it when we’ve gone through it as well.

I think this question of how we extrapolate this to other targets is a really important one. It’s really challenging to do these types of studies for your RET patients and your MET patients, and all of these were targets. It’s going to take years, and it’s probably going to be practically impossible. I think that’s where the caveat of not knowing if this is truly a curative therapy, of taking into account one of the benefits of ADAURA is that osimertinib’s so well tolerated. Patients can go onto 3 years of osimertinib and have a very good quality of life, and that’s variably true with other TKIs.

So for now I’m not extrapolating the ADAURA data to other targets. I will try to get patients on the ALCHEMIST trial if I can, if they have ALK or other targets. But I think if we see an OS benefit, if we really believe that osimertinib is curing these patients, to me I think that would be the threshold that I might use to start to apply it more broadly to other targets, acknowledging that we may never have that primary data for those patients.

DR LOVE: I’ll tell you about a really interesting observation we’ve made ever since the ADAURA trial was presented, which is when we put up a poll question for an audience of general medical oncologists at one of our webinars, and we say outside a clinical trial and putting aside regulatory reimbursement issues, generally speaking how would you like to treat a Stage II or III patient with an EGFR activating mutation after surgery. And we give them the choice of just chemo, chemo followed by osimertinib, or osimertinib and no chemo. The general oncologists give very different answers than investigators. So you all say, well what do you say?

DR PIOTROWSKA: I say chemo and osimertinib.

DR LOVE: Okay. You know what most common — it’s about 50/50.

DR PIOTROWSKA: I’m going to guess it’s osimertinib alone.

DR LOVE: Exactly.

DR PIOTROWSKA: Yeah. I’ve heard that.

DR LOVE: They don’t want to give the chemo.

DR PIOTROWSKA: Yeah.

DR LOVE: So I’m curious what your thoughts are about it. I totally understand the scientific point of view, that you really don’t have the survival documentation. But I think their argument is if it really does play out the relative risk reduction is so much greater with osi, it’s like you don’t really have that much left for the chemo to work in, if that makes any sense. They’re older. They’re not talking about a 55-year-old patient. They’re talking about a typical lung cancer patient. So if you see somebody as a second opinion, and the first opinion was osimertinib without chemotherapy, do you say well that’s an acceptable alternative, maybe not what I would do, but I think it’s okay? Or you go this is just a good idea.

DR PIOTROWSKA: My perspective has been if they’re chemo eligible, taking that as the first question. If you think they’re a patient who can get chemo, I think it’s hard, even though I totally agree it’s a modest benefit, it’s not as good as we’d like it to be, I think it’s hard to put aside therapy with a survival improvement, a survival benefit. So that’s been my reasoning for still offering these patients chemo.

I think it’s a totally different situation where they’re borderline eligible for chemo, or they’re not eligible or chemo, where if you have osi to reach for I think it’s very fair to say okay, this gives me an out for not pushing chemo.

I’ll just briefly mention the Chinese ADJUVANT study, which is I think not exactly this point, but I think it is a relevant one in this discussion. So the ADJUVANT study was a study that compared gefitinib to chemotherapy. Patients were randomized with resected Stage II and III, EGFR-mutant cancers in China were randomized to either chemotherapy or gefitinib. When the study was first presented, actually, I think there was a lot of discussion that we were not sure that that study would have necessarily gotten through the IRB here in the US because of the fact that you’re foregoing therapy with a survival benefit in the experimental arm. And while there was a disease-free survival improvement in ADJUVANT, there was no overall survival benefit seen, comparing gefitinib to chemotherapy.

And it’s not the exact same scenario. It’s not osimertinib. It’s a different population. But I think that that has to give us some pause in terms of saying this TKI is going to be better than chemotherapy. And I think for right now if patients are eligible, my bias is to give them the chemo and then put them on the TKI afterwards.

Therapeutic approach for patients with locally advanced unresectable NSCLC and EGFR mutations

DR LOVE: So the other situation where you see a big difference between investigators and general medical oncologists is locally-advanced unresectable with EGFR tumor mutation. And there I hear investigators saying chemoradiation followed by durvalumab. And I hear medical oncologists bringing in osimertinib either after the chemoradiation or before or instead of. And again, of course these are generalities. They’re thinking about a typical patient, et cetera. But again, any thoughts? I know there was a trial at one point to give targeted therapy prior to chemoradiation, and it didn’t accrue, et cetera. But in general, outside a clinical trial setting, and understanding how the ADAURA data’s affected people so much, do you think that there should be an option to include something like osimertinib or targeted therapy in the locally-advanced unresectable situation?

DR PIOTROWSKA: It’s a great question. And there is a study called the LAURA trial, which is basically like PACIFIC but using osimertinib instead of the durvalumab. So I think that’ll be a really interesting and important study. Until we have the results of that study I have moved towards — I give my patients who have unresectable Stage III who have a known EGFR mutation, I give them the chemotherapy and radiation. I don’t tend to put them on durvalumab because I think the benefit there seems to be lower. These patients generally have low PD-L1. We know extrapolating from the metastatic setting that the benefit of durvalumab is likely less for those patients.

And again, going back to that concern about toxicity, if that patient recurs after durvalumab and you want to put them on a TKI, like osimertinib, you worry that they’re going to end up at increased risk of toxicity. So to me I have not used durvalumab in that setting.

I haven’t been in that situation since ADAURA came out, and 1 question is instead of giving them durva would you give them osimertinib? I think it is a bit of an extrapolation. I think it would be a discussion with the patient. I do worry about the risk of pneumonitis with having just gotten definitive thoracic radiation, getting osimertinib. So I might wait for them to recur, but I think that’s one where it would be worth having a discussion with the patient. But I do feel fairly strongly that I think durvalumab in that situation may not be quite as beneficial, and I look forward to the LAURA data to help guide us on the role of osimertinib in that setting.

Potentiation of the risk of immune-related adverse events in patients receiving immunotherapy followed by osimertinib

DR LOVE: We’ve talked a couple times now about this issue of prior IO and then osimertinib. Can you expand a little bit on that in general, to targeted therapy for example? I’ve heard people say maybe you get more hepatic toxicity with ALK inhibitors, et cetera. In general what do we know? And also what do we know about the time from the IO to the TKI in terms of causing these — how long do you need to wait?

DR PIOTROWSKA: So I would say in the EGFR space the best and largest data set was a data set that came out of Memorial Sloan Kettering a couple of years ago, where they looked at patients with EGFR mutations who had had either PD-1/PD-L1 inhibitors or I think they also included some patients who got PD-L1 and CTLA4 there as well. But they looked at those patients who were subsequently treated with EGFR inhibitors. And what they saw was that there was an increased risk of autoimmune complications when patients received EGFR inhibitors, and specifically osimertinib, following the immunotherapy. There was a particularly increased risk of pneumonitis. I think the rate of pneumonitis in that context was about 15%, which is higher than we’re used to seeing with osimertinib on its own.

The timeframe is really interesting. This was a retrospective analysis, so it’s not a perfect analysis. And we know that these immunotherapies can have really long half-lives. But in that study, what they highlighted was 3 months as kind of a timepoint before which that risk seemed to be particularly high and after which it seemed to fall. So that’s a line in the sand. I would say it’s not a black and white line that if you get past 3 months you’re totally safe. I think you still have to watch these patients.

Interestingly, that risk seemed to be a little bit lower with first-generation TKIs than with osimertinib. There does seem to be this particular kind of increased risk of pneumonitis with osimertinib, although you can see other toxicities, hepatotoxicity. And then you alluded to some of the data from my colleagues, Jess Lin and others, looking at crizotinib after immunotherapy for ALK-positive patients, and there there was an increased risk of hepatotoxicity.

So I think what we’re seeing is there does seem to be some potentiation of toxicities that we may see in and of themselves with these TKIs, but just are more frequent when patients have had prior immunotherapy. There’s some priming of the immune system that may raise that risk, and I think it’s something that we’ll probably have to watch for and learn more about with many of the new and emerging TKIs. We don’t know yet, pralsetinib, selpercatinib, capmatinib, all of these drugs, how this will play out. But I think it is important and noteworthy, if your patient’s had a recent immunotherapy, and you’re putting him on a TKI, to watch them closely for any autoimmune toxicities.

DR LOVE: Do you think that the pneumonitis that you see with TKIs, for example osimertinib, is immune regulated, and that’s why you see an increase? Or it’s like coincidental, occurring at the same time. You’re getting maybe delayed autoimmune pneumonitis from the IO at the time you’re getting — do you visualize pneumonitis from a TKI as being immune? Because I never did.

DR PIOTROWSKA: Well, if you think about the fact that it responds to prednisone, even if patients have never had an immunotherapy, if they get an EGFR inhibitor pneumonitis, prednisone is a main way that we treat it. To me, that tells me that it is an inflammatory condition, and the immune system is playing a role. Whether that’s the same as an autoimmune pneumonitis that we would see in a patient on pembrolizumab or nivolumab I think mechanistically I’m not sure if those are exactly the same, whether it’s the same type of reaction. But my general sense is it’s fair to say it is an inflammatory condition and therefore likely mediated by the immune system.

DR LOVE: Interesting.

DR LOVE: So even though it’s a little bit off topic I have to ask you about something else that I saw in your CV that I never saw before, so autoimmune hemolytic anemia with checkpoint inhibitors. I’d not heard about that. What’s the deal?

DR PIOTROWSKA: Well that was actually a project that I participated in with some hematologists at our center. I had a patient, it was an interesting case of a patient with CLL, longstanding CLL, who actually had had a positive direct Coombs test in the past, and a long time passed. We hadn’t really thought much of it. Developed non-small cell lung cancer and got treated with an immunotherapy, I think in that case it was nivolumab, and within weeks developed a raging autoimmune hemolytic anemia.

And in talking to some of my colleagues, and then sending her to hematology, we saw that this was something that seemed to be not an isolated event. And in particular these patients with underlying predisposition to autoimmune hemolytic anemia, like a CLL patient, that seemed to be a particular risk factor. And so the hematologist looked at a series of patients, and then they did see that that was an increased risk.

DR LOVE: That’s really interesting. So these patients had some other, like CLL or something, that was maybe the real cause of it, that it was getting amplified.

DR PIOTROWSKA: Yeah, it’s like you’re unleashing. They may be prime to it, but it’s not clinically significant until you put them on the immunotherapy, you give that kind of priming signal to the immune system, and then you unleash this toxicity that they are prone to.

Perspective on COVID-19 vaccination for patients with lung cancer; experience caring for patients with COVID-19

DR LOVE: So another topical question I can’t help asking you because it’s on everybody’s mind right now. And really it’s happened so fast that I haven’t had a chance to ask many people, is what you’re thinking about in terms of the vaccine for COVID in your lung cancer patients? Are there any medications you want them to try to avoid prior to getting vaccinated? Are you going to allow them to be vaccinated? What about people on checkpoint inhibitors, et cetera, any thoughts about that?

DR PIOTROWSKA: It’s a great question, and I think it’s one that we’re going to be having to answer within a matter of weeks. Generally I would say that I am going to tell my patients to get it. I think we’re going to have to make these decisions probably with a dearth of data, and some of it we’re going to have to make consensus decisions without necessarily having all of the perfect information. But given the unusual and really unprecedented times I think we’re going to have to make that decision. So I’ll be talking to my colleagues in thoracic and melanoma about what their thoughts are on immunotherapy. But I suspect that we’ll likely say yeah, it’s probably something that our patients in particular, who are at risk of COVID complications should get.

I did run into, actually, an interesting question in the clinic last week of a patient who is on immunosuppression for interstitial lung disease, an autoimmune condition. And we actually went back and asked the pulmonologist, she asked us, can I get the COVID vaccine. And we said well you’re getting chemo, we want you to get it, but we don’t know, with you being on prednisone with an underlying autoimmune pulmonary condition. We went back to the pulmonologist, and they said that we don’t know. That those patients have not been included in the studies. And so for someone like that we may opt to vaccinate everyone around her and wait for a little bit more data to come out.

DR LOVE: Final question. Just kind of curious. You were mentioning to me before we got started that you spent a little bit of time taking care of COVID patients who didn’t have cancer. Just kind of curious what that was like after your prior career in oncology.

DR PIOTROWSKA: My prior, and I will say I will say I was glad to get back to my career as an oncologist. But it’s a time that I will definitely look back on in my career as something that was really special and unique, I think for a number of reasons. This was during the COVID kind of surge. The numbers were really ramping up very rapidly here in Boston over the end of March, and basically the hospital was running out of general medicine doctors to take care of them, and so we set up these COVID teams within a matter of literally days.

The way that our oncology division chose to handle it was to form teams of oncologists so we would all work together, which was really a wonderful experience in terms of the camaraderie of working with my colleagues. I think it was really hard, but also fun to see how much of the general medicine knowledge came back to us. And most importantly, I think, what I’ll remember most about that time is the patients and the isolation that they were in.

What was really striking was at the time on our unit we had many families where multiple family members from a household were sick and admitted, in separate rooms, but we were dealing with navigating the challenges of 1 person would get sick and transferred to the ICU, and that turned out to be the sibling or the husband or wife of another person. And it was just a really emotionally hard and difficult time, but also something we all felt proud to be a part of, and it felt really important to do.

It was also really scary, not knowing, are we going to bring this home? Are we going to get sick? Are our kids going to be exposed? And thankfully that turned out not to be as much of a concern, but it was really a unique and special time. And as I was mentioning to you earlier, I think the numbers going up here in Boston and everywhere, there’s a good chance that we’ll be doing it again at some point in the near future. And I think that if that comes to be the case, I think we’ll all be proud and happy to help because what else are we going to do? I think these patients need us, and until we get this under control, ultimately we’re all just going to try to do what’s best for these patients, whether or not they have cancer.