Joshua Bauml, MD, speaks on the treatment options and considerations he makes when treating patients with non–small cell lung cancer. Bauml explained his treatment decisions based on 2 case scenarios during a <em>Targeted Oncology </em>live case-based peer perspectives presentation.
Joshua Bauml, MD
Joshua Bauml, MD, spoke recently on the treatment options and considerations he makes when treating patients with nonsmall cell lung cancer (NSCLC). Bauml, assistant professor of medicine, Perelman School of Medicine, University of Pennsylvania, explained his treatment decisions based on 2 case scenarios during aTargeted Oncologylive case-based peer perspectives presentation.
A 63-year-old man presented to his primary care physician with intermittent cough and difficulty breathing on exertion. His past medical history showed hyperlipidemia, well managed on simvastatin (Zocor); hypothyroidism, managed on levothyroxine (Synthroid); chronic obstructive pulmonary disease, managed on inhalers. He recently quit smoking but had a 40-pack-year history. A physician exam showed intermittent wheezing and his ECOG performance status was 1. His creatinine clearance levels were within normal limits
A chest CT revealed a 3.1-cm spiculated mass in the right upper lobe and 2 enlarged right mediastinal lymph nodes measuring 1.5 cm and 1.7 cm; moderate emphysema noted. A PET later confirmed the lung lesion and mediastinal lymphadenopathy without evidence of distant metastases. His brain MRI was negative.
A pulmonary function test was performed and revealed a forced expiratory volume in 1 second, 1.2; diffusing capacity of the lung for carbon monoxide, 52%. A bronchoscopy with transbronchial lung biopsy and lymph node sampling revealed adenocarcinoma with positive nodes in stations 4R and 7; level 4L was negative. He was later staged with T2aN2M0, stage IIIa. Genetic testing was negative for known driver mutations.
Based on the extent of mediastinal disease and emphysema, the patient’s cancer was deemed inoperable, and he was referred for consideration of concurrent chemotherapy and radiation.
Do you typically order genetic testing for locally advanced lung cancer?
My preference is to do it on everyone, which may be a little too extreme. Even my patients with early-stage lung cancer, I will tend to check because I would like to know. Our cure rates in localized lung cancer are very poor. But I don’t know whether it is right. Certainly, in the metastatic setting, most people are doing it.
I’ve yet to have insurance yell at me when I’ve done it. But we’re also using an internal panel, and the patients are not getting stuck with the bill. That makes me feel more comfortable. If my patients were getting stuck with the bill in stage I disease, I wouldn’t do the testing. Therefore, if someone is coming into my office, I’m sending them for testing.
For adenocarcinoma, I personally do not do any polymerase chain reaction (PCR)based testing. My exception would be for someone who, for example, is in the medical intensive care unit and a never-smoker 40-year-old woman. Then I’ll do a PCR-basedEGFRtest because it is fast. But from an efficiency of tissue standpoint, I don’t think there is any comparison between next-generation sequencing and a PCR-based test.
In multipleNew England Journal of Medicinepapers, I believe there are now more NTRK inhibitors being evaluated that actualNTRK-positive patients. I think that is an issue. But this is a target that is seen in a wide variety of tumors. I check it on all my salivary gland tumors and acinic cell tumors, but I have yet to see anNTRK.NTRKis an interesting space, and it is one of the reasons why, even for some non-adenocarcinomas, I do check.
What are the options for treatment?
I think this varies by institution, whether you want to use neoadjuvant chemotherapy or trimodality approaches, although I haven’t seen a lot of institutions using trimodality anymore. It is technically an option.
One of the things that I think is important to remember when you are thinking about concurrent chemoradiotherapy is that the incremental benefit of the concurrent chemotherapy is relatively modest. If you have a patient who is not very fit, I will give sequential therapy. I think that it is reasonable.
He underwent therapy with cisplatin/etoposide and concurrent thoracic radiation therapy and follow-up imaging showed a partial response with shrinkage of the primary and nodal lesions.
Can you comment on the design of the PACIFIC trial?
The PACIFIC study looked at patients with locally advanced, unresectable NSCLC who had completed concurrent chemoradiotherapy with a platinum-based doublet with at least 2 cycles.1They started 1 to 40 days post chemoradiotherapy, which is interesting. That is very fast. Initially, it was even tighter, and then they expanded it because they couldn’t get people on.
The study was randomized 2:1 to durvalumab (Imfinzi), which is a PD-L1 inhibitor, versus matching placebo every 2 weeks for 12 months. There were 476 patients in the durvalumab arm and 237 patients in the placebo arm. The coprimary endpoints were progression-free survival (PFS) and overall survival (OS).
The baseline characteristics were well balanced, and I didn’t have any concerns there, although the amount of induction chemotherapy was a little bit higher than we would generally use in our practice. They were allowed to have induction chemotherapy, but they were not allowed to have consolidation chemotherapy. That is a slight disparity from our approaches.
What were the outcomes of this trial and their significance?
Most patients did receive therapy, but many of them did not complete the full 12 months of therapy. The PFS was much better for the use of durvalumab. What is most remarkable here is that if you think about PD-1 and PD-L1 inhibitors, the response rate is generally 15% to 20% for select patient populations. You would expect that if all we were dealing with was the effective PD-1 inhibitors given early, you would expect a 1:1 ratio in increase in therapy versus improvement in PFS. This is 12 months of therapy, and nearly 12 months of improvement in PFS. That is impressive and implies that there is some sort of synergy going on. If you take a look at the 12-month and 18-month PFS rates, it is quite impressive the difference that we are seeing here.
Taking a look at the subgroups, you can see that there was benefit regardless of histology, PD-L1 status. The PD-L1 assay used here is not the Dako 22C3 assay associated with pembrolizumab (Keytruda). Therefore, this was the assay developed with durvalumab, where the cutoff is 25%. The correlation between that assay and 22C3 is pretty good. So I feel relatively comfortable on this analysis. Interestingly, [the hazard ratio for]EGFR-mutant patients did [favor placebo], implying the incremental benefit was not as substantial. But these numbers are very small.
The overall response rate (ORR) is sort of a strange outcome, in my opinion. After chemotherapy, you have already radiated, so what are you actually measuring in terms of response? We don’t have OS data, so what they did was a modified analysis of time to distant metastasis or death, assuming that if you have distant metastasis, you are eventually going to die. You can see a clear, continued benefit for the durvalumab arm versus placebo. This was statistically significant.
What are your thoughts on the trial’s choice to begin immunotherapy so quickly after chemotherapy?
They found that if you get treatment within 2 weeks, the incremental benefit is huge. And if you have it in more than 14 days, it is much more modest. What can we take from this? Our numbers here are actually pretty substantial, with 120 people getting therapy in 2 weeks. What I do is when I have a patient who completes chemoradiotherapy, I will see them 2 weeks after treatment with a repeat CT scan of their chest. Provided they don’t have progression of their disease and they are ready to go, I will start them on durvalumab that day. I am a little uncomfortable with this because as you know, radiation pneumonitis tends to occur at 4 weeks. So, I am taking that time point and hammering it home. But if you take a look at the safety [of the PACIFIC trial], even with those numbers, the rate of pneumonitis was not substantially different between the durvalumab and placebo arms. It seems to be a safe approach, and I feel relatively comfortable with that.
Durvalumab after 2 weeks is tricky and a very aggressive move, but I talk with the patient about it at day 1. When I talk to them about the chemoradiotherapy, I say, “You’re going to get chemoradiotherapy, and then something else is going to happen. And I am not going to talk to you about it right now because it is overwhelming. But just so you know, it is coming.” The nurse practitioner I work with will also counsel them continuously and say, “Next week you’re going to see Dr Bauml, and he’s going to talk to you about immunotherapy.” For better or for worse, our patients hear about immunotherapy through direct-to-consumer advertising. Mostly for worst, but they at least know about it. It is not a completely novel approach.
The underlying principle of why this study might work, and why immunotherapy after radiation therapy might work, is that when you give radiation, you are causing apoptosis of cells and necrosis of cells, release of neoantigens, which can prime the immune system. Those neoantigens are cleared from the body very rapidly. The closer you are to that approach, the more likely you are to benefit. To me, biologically, that makes sense. There is a lot of interesting biology regarding radiation plus immunotherapy; whether it is going to be smoke and no fire, I don’t know. But there is reason to think that biologically, starting it earlier is better.
Are there any other data to support starting immunotherapy earlier?
We did a study where we looked at patients with oligometastatic disease, radiation, stereotactic ablation, and then gave them immunotherapy. We saw very long and durable responses there. It is a similar idea, this sort of amplification of neoantigenicity with immunotherapy. I don’t know the right answer, and further data are definitely needed to see what the safety is of doing it so quickly. I find this very intriguing.
An 81-year-old man presented with symptoms of coughing, dyspnea, upper back pain, and fatigue requiring frequent rest. His past medical history revealed hypercholesterolemia, controlled on pravastatin (Pravachol); hypertension, controlled on verapamil (Verelan); psoriatic arthritis, not on treatment for 3-plus years. He was a former smoker, but physically active and played golf most weeks with an ECOG performance status of 1
A Chest CT revealed a 2.5-cm solid mass in the left upper lobe and lymphadenopathy in the left hilar and bilateral mediastinal nodes and bilateral, small pulmonary nodules, with the largest one 8 mm. PET/CT imaging showed 18F-FDG uptake in the lung mass, left hilar and both mediastinal lymph nodes, and thoracic spine (T5/T6).
A Bronchoscopy and transbronchial lung biopsy were performed and pathology showed grade III squamous cell carcinoma; PD-L1 expression by immunohistochemistry 22C3, tumor proportion score of 65%. He was later diagnosis stage IV squamous NSCLC.
Are you routinely testing for PD-L1 expression in patients with newly diagnosed disease?
This is an important question. One of the things I’m finding as an oncologist is that it is more and more important to get into the nitty-gritty with my pathologist to find out exactly what is happening. For instance, pathologists used to send for immunohistochemistry to figure out which subtype of adenocarcinoma it was. But that doesn’t make an impact on our clinical decision making, and they would exhaust tissues so we couldn’t check forEFGRmutations. That requires a conversation to say, “Look, I need this. That is fine, but I need this.”
Similarly, it is very important with PD-L1 to know which assay is being used. The blueprint assay compared to the different PD-L1 assays that have been developed. What they found was that correlation on the tumor is pretty good, but correlation on the immune cells surrounding is pretty awful. One of the assays, the VENTANA assay, associated with atezolizumab (Tecentriq), had a 30% lower sensitivity than any of the other assays. When, for instance, you read the atezolizumab papers, and it says, “We have activity in PD-L1negative tumors,” you probably don’t. It is probably just that 30% of those assays were PD-L1 positive and were just not being picked up.
The patient was started on pembrolizumab.
How does his older age and good performance status factor into this choice?
The patient here was started on pembrolizumab. He was older but had a good performance status. For a patient with adenocarcinoma, I would say that it matters, but I don’t see how it makes an impact because pembrolizumab is so well tolerated.
Do precautions need to be taken considering his psoriatic arthritis?
The psoriatic arthritis is sort of a red herring thrown in there. I would start the patient with pembrolizumab, but I would need to monitor the patient closely. You need to have a close conversation with the rheumatologist and say, “How bad is this, and what are we dealing with?”
For example, I had a patient who had a PD-L1 tumor [with 80% expression]. Additionally, she had lupus and vasculitis. Her legs were almost black from multiple episodes of thrombosis. This was a young woman, with a horrible tumor. I struggled with her, and I ended up deferring the use of immunotherapy because I thought if I gave it to her, her legs might falter. But, the studies that have done the retrospective analysis showed that our anxiety around this is probably overblown. It is not as real as we think it might be, but it is something that we have to consider.
Can you explain the design from the KEYNOTE-024 trial?
This was a study that answered the question we just asked.2The trial looked at patients who are stage IV, greater than 50%, good performance status, but they excludedEGFRandALKtranslocations. If you have a patient with PD-L1, I don't care what it is, and they have anEGFRmutation, they should receive an EGFR tyrosine kinase inhibitor (TKI) first. The benefit of immunotherapy in that space is unproven at best, and absent at worst.
The patients were randomized to pembrolizumab at 200 mg every 3 weeks for 2 years or platinum-doublet with transition to pembrolizumab after progression. Note that there wasn’t maintenance here. The rate of crossover was a little bit disappointing. Of the 150 patients treated with chemotherapy, only 82 patients crossed over to pembrolizumab. It is lower than I would have hoped it to be. But compared with other trials, it is better.
What are the outcomes of the trial?
Julie Brahmer, MD, presented the updated OS analysis at the World Conference on Lung Cancer in Japan last year. The median OS was 30 months for pembrolizumab and 14 months for chemotherapy. This is not subtle; it is a huge improvement. Again, with immunotherapy, you have to look at the tail of the curve. If you look at 30 months, and 50% of these patients have OS, that is a remarkable improvement for these patients. They have a very good quality of life during that time period.
If you look at the hazard ratios, you can see that pembrolizumab benefited nearly all subgroups, although there was a crossover at the never-smokers. But I think that hazard ratio covers every possible number. That is an impressively large hazard ratio. Nearly everyone else benefited from pembrolizumab.
The ORR was higher with the use of pembrolizumab than with chemotherapy. What I think is most interesting about this is that the time to treatment response was the same in both arms. At the first assessment, the median time to response is identical for chemotherapy and immunotherapy. We’ve seen this. If you have a patient who is going to respond to immunotherapy, they respond. There is a small subgroup that will respond at 6 months, but if they are going to respond and have these beautiful responses, they respond.
The duration of response has not been reached for pembrolizumab, and it is 6.3 months for chemotherapy. This is within the range of what I would expect for platinum-doublet chemotherapy. I think this is reassuring. We are seeing a nice benefit from pembrolizumab, and chemotherapy is functioning as we would expect it to.
Is there any significance to the toxicity profiles that were noted in the study?
The toxicity is not surprising to anyone who has given pembrolizumab. The rate of grade 3 to 4 adverse events (AEs) is much lower for pembrolizumab than it is for chemotherapy. The toxicity should correlate with the degree of exposure, but it doesn’t. The patients who had a longer exposure to pembrolizumab had less toxicity. That is pretty exciting. Certain AEs are simply absent for pembrolizumab. There is no neutropenia or thrombocytopenia, and everything is lower with the exception of pyrexia with pembrolizumab. But even that has a very low incidence.
The immune-mediated AEs show that toxicity happens, and it needs to be addressed and treated. The most common immune-mediated AE is hypothyroidism, which is not a big deal. Even with hyperthyroidism, usually you ignore it, and then they burn out, and then you put them on levothyroxine. Pneumonitis is the one that I worry about. Any patient who has shortness of breath when they are on immunotherapy, I will tend to do a collection tree protocol to make sure they don’t have a pulmonary embolism. If they do have pneumonitis, I start them on high-dose steroids. The key thing is that after I tapper them off steroids, I can reexpose them to immunotherapy. About 75% of patients will be able to be safely reexposed to immunotherapy after a pneumonitis episode. In contrast, if you look at some of the other AEs, the rate of recurrence is higher. The rate of recurrence for colitis is almost 80%, and the rate of recurrence for muscle pain is the highest. If you have someone who is having strong AEs, you have to pay attention to what is going on.
How do you manage the myositis?
Usually they have to come off the drug. I’ll try to manage it with supportive care, but if they have serious myositis, they have to come off it. I had a patient who was having a beautiful response, but she couldn’t move her hands. I had to stop it and give steroids. Generally, with myositis you are not going to get them back on the drug. I always try, but I don’t expect that it is going to work. Usually it presents as awful arthritis, and it can be very bad.
The other big one that you have to worry about when someone presents with shortness of breath or chest pain with one of these immunotherapy agents is myocarditis. It is much higher in combination therapy, but it is very serious and must be thought about with these patients.
How would you treat the patient if he progresses on pembrolizumab?
I give them a platinum-doublet. They are basically platinum naïve. This is when it gets confusing; people start talking about line of therapy. They say, “You can’t use platinum-doublet in second line.” But it’s first-line chemotherapy, so it could be considered the same as if someone is on an EGFR TKI and they progress and then I put them on another EGFR TKI and they progress, and then I put them on platinum-doublet chemotherapy. It is first-line chemotherapy. That is what I give them in that space.
The complicated thing is, what are we going to do for these patients who progress on durvalumab? They have been given a year of durvalumab, and then they progress after that. What do you do with that person based on the PD-L1? That is very complicated, and I don’t have a great answer for that.