Video
Author(s):
Switching their focus to advanced non–small cell lung cancer, panelists identify parameters that may influence the use of immune checkpoint inhibitors in this setting.
Transcript:
Balazs Halmos, MD, MS: Let’s talk about metastatic non-small cell lung cancer. This is a case from my own practice. A 68-year-old woman with a prior smoking history comes in with a couple of weeks’ history of weight loss, progressive shortness of breath, has a workup that includes a chest CT that shows a fairly large pleural effusion. This is followed by thoracentesis. Cytology is consistent with adenocarcinoma histology. TTF-1 [thyroid transcription factor 1] and napsin positive. Has further workup including a PET [positron emission tomography] CT. The PET CT shows multiple pleural deposits, a right-sided lung mass as well as multiple liver lesions consistent with metastatic disease. A brain MRI is also done. This shows no CNS [central nervous system] metastases. She appears to be in fair shape, has well-controlled hypertension and hyperlipidemia. She wants to go on with aggressive treatment to get the cancer under control. She wants to make it to her granddaughter’s college graduation. That’s her specific goal.What’s the best approach to achieve that? Do we need more information? How do we get her to make the bestinformed decision?
We’ll start our conversation with that, but first, a phone call from the pathologist tells you that the patient’s TPS [tumor proportion score] score is exactly 50%. Dr Reuss, take us through some of the data talking about immunotherapy-based approaches.
Joshua Reuss, MD: PD-L1 is not enough to make this decision, even in a smoker. You need the NGS [next-generation sequencing] data. Patients who are smokers can have actionable driver mutations as well and we can see in enrichment, certain mutations, such as MET for example, but other mutations can be seen there. We can’t rely on just the PD-L1 in this scenario. We should letKRAS, STK11, KEAP1, ... and some of these other mutations affect our decision making as well.
The big picture perspective is if you get the NGS testing back and it does not have an actionable driver mutation, we will take the STK11 and keep 1 alteration. We will put a pin in that, but if you have 50% or greater, then there are possible treatment options for PD-L1 monotherapy. Right now, our options are pembrolizumab [Keytruda], atezolizumab [Tecentriq], and cemiplimab [Libtayo] based off your respective KEYNOTE, IMpower and EMPOWER studies.
Before considering monotherapy you do have to take a step back and say, is this patient symptomatic? What is their disease burden? Even if they are PD-L1 high, and we’ll throw 50% into that category, I want to know clinically how the patient is doing because to me it’smore important. If someone is clinically not doing well, if that patient with your liver metastases has a transaminitis or pain or shortness of breath with the pleural effusion in the lung nodules, that’s a scenario where irrespective of PD-L1, I’m inclined to include chemotherapy. I can go into more detail about the IP [intraperitoneal] monotherapy studies, but curious how my colleagues would approach this scenario as well.
Balazs Halmos, MD, MS: Do you have some initial thoughts as to would you be comfortable with single agent IO? Let’s presume no molecular targets in such an individual.
Martin Dietrich, MD, PhD: The PD-L1 caliber 50% is, by definition, intermediate on the cusp of high. In my interpretation, it falls on the lower side. I think all the criticism we give PD-L1 is not necessarily only in the way the biomarker performs, but also in the way we interpret the biomarker. The PD-L1 correlation of benefit is almost exponential as we get closer to 100%, and 50% has an indiscernible benefit compared to 40% or 30%. As we get close to 50%, and certainly below 50%, monotherapy is not appropriate in a patient that is medically fit.
You have a patient that is symptomatic with progressive shortness of breath, so, by definition,the addition of chemotherapy should be considered and offered to the patient with strong recommendation to turn around this symptomatic compromise in the beginning. The question is how to interpret the biomarker status and how to approach immunotherapy. We have the PD-1 single-agent regimens that are available whether they are a common combination or not. There is a lot of immune stimulatory effect by chemotherapy. I would say it is probably quite the opposite, but I think this is a paradigm that’s changing. The question is shall we use one of the new combination therapies? Is there any sort of challenging set up either genetically or immunophenotypically that would require consideration for a second immunotherapy agent? I agree with Dr Reuss here. The first step is an NGS panel to delineate the benefit of immunotherapy. This includesprimary drivers, secondary resistance mutations, tumor mutational burden, and PD-L1.
I have not heard the criticism that we are sequencing too many genes in a while. As we interpret them better and better and learn more and more, I think the refinement there is better. If a patient has PD-L1 low disease and 50%, I would consider the addition of a second immunotherapy agent. I think the anti-CTLA-4s are a worthwhile consideration. PD-L1, in our PD-1 center, brought the best biomarker that we have for the use of CTLA-4 in an inverted fashion where typically the higher, the more benefit of PD-1 and the lower, the more relative benefit of anti-CTLA-4. I would be leaning towards one of the quadruplets here that would include a CTLA-4 antibody.
Balazs Halmos, MD, MS: You mentioned that the higher the TPS[tumor proportion score] score the higher the confidence that you have in single agent immunotherapy. The EMPOWER study actually did look at that in a very specific detail and suggested excellent outcomes for TPS score in the 80% or higher patient population. Are these thresholds relevant in your practice, or is it a continuum in terms of decision making?
Martin Dietrich, MD, PhD:It’s a contextualized marker where genetic and immunophenotypic markers, in addition to the patient’s clinical situation, play a role. PD-L1 is not negative low and high. The data from EMPOWER-Lung 1 [NCT03088540] is very similar than the data found in the single agent atezolizumab studies. Percentage distribution in long-term survivors in KEYNOTE ... is nearly identical to the percentage of patients that have PD-L1 super high levels. There is a lot of data to support that PD-L1 is a very granular biomarker for usage.
Balazs Halmos, MD, MS: This year we have seen the 5-year update for some of those studies. It is really stunning to see that there is a certain percentage of patients who are doing really well off therapy for years without cancer progression or recurrence.
Transcript edited for clarity.