Levy Lends Insight on Evolving Treatment Strategies in Lung Cancer

Benjamin P. Levy, MD, Johns Hopkins Sidney Kimmel Comprehensive Cancer Center

Benjamin P. Levy, MD

The non—small cell lung cancer (NSCLC) paradigm has seen updates with molecular drivers, and research continues to focus on further advancing treatment for patients who express these abnormalities, as well as those with nondriver disease, explained Benjamin P. Levy, MD.

“There are a lot of different treatment options for patients with oncogenic drivers. It is contingent on the molecular driver that is identified,” said Levy. “We have learned that the list of oncogenic drivers is getting longer, so we have more therapies available.”

In an interview with OncLive during the 14th Annual New York Lung Cancers Symposium^®, Levy, clinical director of medical oncology and associate professor of oncology at Johns Hopkins Sidney Kimmel Cancer Center in Sibley Memorial Hospital, discussed therapeutic strategies for patients with oncogenic-driven and nondriver NSCLC.

OncLive®: How have molecular drivers evolved in lung cancer?

Levy: We've come a long way with the molecular driver identification in NSCLC. It began with EGFR and ALK in advanced adenocarcinoma of the lung, but we now have other mutations that we can test for, including RET, ROS1, BRAF, HER2, MET exon 14 skipping, NTRK, and KRAS. These are all identifiable mutations and, if we identify them, we can either lead [patients] to a targeted therapy that is approved or under the umbrella of a clinical trial.

The list is getting longer of all of the actionable mutations that we need to test for. It's a call for comprehensive genomic profiling for lung cancer. You can't just do sequential EGFR, ALK, ROS1, and BRAF. You have to do a larger panel—preferably a next-generation sequencing panel—so that we can identify all of these mutations and improve the outcomes for patients with these genetic alterations.

How have treatment options evolved for patients with oncogenic drivers?

Even within a particular oncogenic driver, such as EGFR, there are more therapies available—even within that molecular niche.

Within each genotype, there are expanded therapies. Around 5 or 6 years ago, we had gefitinib (Iressa) or erlotinib (Tarceva) [for patients with EGFR mutations], but now we have moved forward with other drugs, such as afatinib (Gilotrif), dacomitinib (Vizimpro), and now osimertinib (Tagrisso). These are next-generation therapies that are more efficacious and better tolerated.

The same goes with ALK. If we look back just 5 years ago, we were treating ALK-positive patients with crizotinib (Xalkori). Now, we have alectinib (Alecensa), brigatinib (Alunbrig), and lorlatinib (Lorbrena). The list [keeps growing] and it's giving our patients more options and extending their survival.

What is your frontline therapy choice for patients with nondriver NSCLC?

My first-line option for patients without an oncogenic driver is generally chemotherapy with immunotherapy, if they're not being considered for a clinical trial. [In my] practice, I have [a number of patients who are] enriched for oncogenic drivers, but about 50% of my practice does not have a driver that we can give a targeted therapy to. For those patients, I generally offer them chemotherapy with immunotherapy. That is for patients with adenocarcinoma and squamous cell lung cancer; we don't normally test for genomic alterations in squamous cell disease. We're doing more of that, but if the patient is a smoker with squamous [histology], we generally don't [test for genomic alterations].

If patients are eligible for immunotherapy, I offer them a combination of chemotherapy with immunotherapy. The one nuance in this is patients who have PD-L1 expression >50%. In these patients without an actionable mutation, you have the option to either to give them single-agent immunotherapy or the combination of a platinum doublet with immunotherapy.

What therapies are available in a subsequent setting for patients without oncogenic drivers?

There are very challenging and nuanced decisions for what to do in the second-line setting and beyond for patients without an actionable mutation who have been potentially treated with either immunotherapy alone or chemotherapy and immunotherapy. The choice is pretty clear for patients who have PD-L1 >50% who received first-line pembrolizumab (Keytruda). Generally, the second-line choice should be chemotherapy.

However, with patients who already received platinum doublet with immunotherapy, what do we give upon progression? Progression is very heterogeneous, and we have to define what that is. In some cases, if there is local growth in one area, one could consider radiation, local ablative therapies, and continuation of the treatment they are on. There is some scientific rationale to the synergy of radiation with immunotherapy. This is really a call for clinical trials in this space, because we don't know what to do.

Therefore, novel immunotherapeutic strategies are being looked at. One strategy we are looking into at Johns Hopkins Medicine is an epigenetic priming study; we are trying to prime the tumor to be more receptive to immunotherapeutic approaches.

Docetaxel with or without ramucirumab (Cyramza) is probably the standard here. We need to define what resistance is following chemotherapy plus immunotherapy and look at important mechanisms of resistance that can help define or inform treatment decisions. We're really in a dead space here. We don't know what to do right now. I hope if you were to ask the same question next year, we'll have some data to guide us on the best approaches for patients who receive chemotherapy/immunotherapy who don't have a driver, and what to do in the second-line setting for them.

How is bevacizumab (Avastin) being utilized to treat patients with nondriver disease?

Prior to immunotherapy, we were using a lot of bevacizumab. With the advent of KEYNOTE-189 and KEYNOTE-407, these are the 2 trials that cemented immunotherapy with chemotherapy in the adenocarcinoma and squamous cell patient population, respectively. Immunotherapy has taken over in combination with platinum doublet, and that has replaced bevacizumab.

There is still a potential use for bevacizumab. I would use bevacizumab in the IMpower150 regimen of carboplatin, paclitaxel, bevacizumab, and atezolizumab (Tecentriq). It is FDA approved and I tend to reserve that regimen for patients who may have an EGFR mutation and are progressing on osimertinib, as well as in patients who don't have a driver but do have a heavy symptom burden.

What research is ongoing to further identify molecular targets?

We are expanding our list of relevant genomic drivers that we can target. "Leave no gene behind," is one thing that I will say. We try to find the gene of interest, because there are multiple trials that are looking at targeted therapies for these particular molecular alterations.

Outside of patients who have drivers, the question is where the science is heading. Right now, for patients with advanced adenocarcinoma, we have carboplatin, pemetrexed, and pembrolizumab. How do we amplify the efficacy of these patients? Do we add a fourth drug to that 3-drug regimen? The jury is still out on this.

Many of these [studies] are umbrella trials where, for patients who are progressing on KEYNOTE-189, we try to identify a particular alteration or mechanism of resistance and put them on a particular therapy. Epigenetic priming studies are important because there is this idea that tumors no longer respond to immunotherapy because they go cold. How do we turn [tumors] back to being hot and inflamed, in order to really amplify neoantigens that allow immunotherapy to work again?

How would you describe the evolution of immunotherapy in patients with advanced NSCLC?

We have come a long way with therapeutic approaches. Right now, we have PD-1/PD-L1 inhibitors in combination with chemotherapy, but we have learned more about potential dual checkpoint blockade, such as CTLA inhibitors in combination with PD-1/PD-L1 inhibitors. In the CheckMate-227 study, one of the arms came out to be positive for this combination approach.

Will the conclusion of predictive biomarkers of response to immunotherapy involve 1 biomarker, or will it require a combination of assays or biomarkers?

It will be a combination of markers that we are looking into for immunotherapy. Currently, PD-L1 is the biomarker that we test for that may help drive decisions. There's a lot of interest in looking at markers in the plasma, such as tumor mutational burden (TMB), which may predict efficacy to immunotherapy.

It's important to identify negative predictors of response. Some data have looked at STK11 and KEAP1 mutations. If you have [one of those mutations], you may not do that well with immunotherapy—either as a single agent or in combination. While it's great to define predictive markers that correlate with better outcomes, it is also important to define patients that we know upfront may not respond to immunotherapy.

How are liquid biopsies best utilized in lung cancer?

Liquid biopsies have altered the diagnostic algorithm for our patients with NSCLC. Who would have thought that, even 5 years ago, we could use a simple 10 cc of blood to diagnose and genotype a patient with lung cancer? We are moving into a space where there should be a consideration on using liquid biopsy in treatment-naive patients [in whom you are able] to diagnose lung cancer, but you don't have enough [tissue] for molecular testing. This is a space that liquid biopsy should be done.

In fact, there are data that suggest that perhaps liquid biopsy should be done in all patients with advanced adenocarcinoma of the lung. I'm not willing to make that statement yet for all patients, but this is something that we need to at least consider for our patients. If we look historically at how successful we are at getting all the information we need off of tissue, we're not that good. Most of the data suggest that we don't have a great success rate of completing all tests needed in tissue. If we layer in the liquid biopsy, we do a much better job of detecting all the alterations.

In my practice, for every patient who walks through the door with advanced adenocarcinoma of the lung, if think they may not have enough tissue for molecular testing, I will order liquid biopsy. Right now, that's where the space is headed.

The other area of interest [we are looking at for liquid biopsy] is in the resistance setting. With patients who progress on osimertinib or alectinib, do we order a liquid biopsy on these patients to see what the mechanisms of resistance are? There is also a real utility of liquid biopsies longitudinally. We're not there yet, but there are a lot of data coming out, which show that looking at longitudinal plasma and variant allele frequencies and the changes over time may correlate better with how the patient is doing on a particular targeted therapy versus a scan.

In looking at minimal residual disease, patients who are cured are typically followed with scans, but perhaps we need to follow them with cell-free DNA (cfDNA). That could be an earlier indicator versus a scan to show [cancer recurrence] if that cfDNA goes up. These are all potential applications.

References

1. Gadgeel SM, Garassino MC, Esteban E, et al. KEYNOTE-189: updated OS and progression after the next line of therapy (PFS2) with pembrolizumab (pembro) plus chemo with pemetrexed and platinum vs placebo plus chemo for metastatic nonsquamous NSCLC. J Clin Oncol. 2019;37(15 suppl; abstr 9013). doi: 10.1200/JCO.2019.37.15_suppl.9013.
2. Paz-Ares LG, Luft A, Tafreshi A, et al. Phase 3 study of carboplatin-paclitaxel/nab-paclitaxel (Chemo) with or without pembrolizumab (Pembro) for patients (Pts) with metastatic squamous (Sq) non-small cell lung cancer (NSCLC). J Clin Oncol. 2018;36 (suppl; abstr 105). doi: 10.1200/JCO.2018.36.15_suppl.105.

<<< View more from the 2019 New York Lung Cancers Symposium