Langer Lends Insight on Burgeoning Lung Cancer Developments, Challenges

Corey Langer, MD

From optimal therapies for patients with non-driver adenocarcinoma, to studying the biology of squamous cell non—small cell lung cancer (NSCLC), to exploring novel combination regimens, to predicting the risk of mortality, the field of lung cancer is undergoing tremendous changes, explained Corey J. Langer, MD.

“The vast majority of patients with advanced NSCLC, unfortunately, do not have oncogenic drivers,” said Langer when discussing agents for patients with non-driver disease. “Those are confined to 20% to 30% of the group, and we have seen lots of promise for new agents based on the molecular target. However, the majority doesn’t have it.”

The expert shed light on all of these topics during the 2016 OncLive State of the Science Summit on Advanced Non-Small Cell Lung Cancer on September 17 in Philadelphia, Pennsylvania.

OncLive: Can you provide an overview of your discussion on squamous NSCLC?

In an interview with OncLive during the meeting, Langer, professor of Medicine at the Perelman School of Medicine, director of Thoracic Oncology, at the University of Pennsylvania, discussed non-driver adenocarcinoma, some of these new combination therapies in lung cancer, and plenty more.Langer: We’re still in the era of systemic cytotoxics, and here, too, we have made some headway. In the last 10 or 12 years, we have seen distinctions histologically with the role of pemetrexed (Alimta) and bevacizumab (Avastin) clearly secured for nonsquamous NSCLC phase III trials that have shown a survival advantage for those agents in combination with platinum-based agents or platinum therapies and other cytotoxics that have proven superior. Really, that is our platform for nonsquamous disease.

For squamous, it’s been a far greater challenge. There are phase III histology-based data that show, for instance, that gemcitabine and platinum is superior to pemetrexed and platinum. Therefore, gemcitabine is certainly one of our go-to agents for those with squamous histology. Taxanes have also proven superior to pemetrexed in the second-line setting in squamous, so taxanes are generally used upfront.

We do have very interesting data looking at nanoparticle albumen-bound paclitaxel in combination with carboplatin which, compared to conventional solvent-based paclitaxel, has led to a significant improvement in response rate, although that hasn’t really translated into either a progression-free (PFS) or overall survival (OS) advantage in squamous cell.

Are there other agents that may have a benefit in squamous NSCLC?

What are some of the toxicities of some of these agents, either in combination or as monotherapies?

It should be noted that in individuals over 70, there was an advantage. Again, it’s a retrospective analysis. It’s only 150 patients or so, but about an 8- or 9-month difference favoring the nanoparticle-bound taxane over the conventional standard co-solvent cremophor-based molecule.There are data in the second-line setting combining ramucirumab (Cyramza), which is an angiogenesis inhibitor that targets the receptor on endothelial cells, in combination with docetaxel. Clearly, there is a survival advantage compared with standard chemotherapy alone. This was observed not just in adenocarcinoma, but also in squamous cell. It’s certainly an option in the second-line setting, and we have compelling data for an EGFR monoclonal antibody necitumumab (Portrazza) in combination with the gemcitabine regimen, which has led to a survival advantage in the frontline setting. The options are expanding, but not as quickly as nonsquamous.Pemetrexed compared with a taxane is generally less toxic. We don’t see the hair loss that’s typically associated with taxanes—much less neuropathy—and there is virtually no neuropathy with pemetrexed. We do see some rash, and sometimes gastrointestinal toxicity or myelosuppression.

What are some of the ongoing studies in this space that you’re particularly excited to see the results of?

However, the routine use of vitamins, B12, and folic acid has really mitigated the toxicity of pemetrexed. Because necitumumab is an EGFR monoclonal antibody, rash is seen. You do see some hypomagnesaemia. There is some concern about increasing incidence of thromboembolic phenomena, although the incidence overall is quite low. For any drug such as ramucirumab or bevacizumab, you have to worry about bleeding. Endothelial cells are the targets, and we’ve seen both bleeding and blood clots with this class of agents. Patients need to be closely monitored for those effects.In the second-line setting, across the board for non-driver NSCLC, immunotherapy has really had a major impact. CheckMate-017 showed superiority for nivolumab (Opdivo) over docetaxel in squamous cell disease. CheckMate-057 showed a major advantage, the same advantage for nivolumab over docetaxel in nonsquamous, although the P values and hazard ratios weren’t quite as robust as seen in the squamous cell group.

What about pembrolizumab as a combination therapy?

We’ve seen similar data now for pembrolizumab (Keytruda) in the same setting in those with PD-L1 expression, and now it’s just a press release. We’ve not seen the data, at least as of mid-September 2016, but in those with 50% or greater expression by IHC, single-agent pembrolizumab appears to be superior to standard platinum-based combinations in the frontline setting, both with respect to PFS and OS.There are ongoing studies pairing pembrolizumab with other immunotherapy agents, including ipilimumab (Yervoy). The data are still very preliminary, but it does seem to combine fairly well with these other agents. Of course, you’re always worried about heightened immune-mediated toxicity.

Looking specifically at patients with early-stage lung cancer, what tests or tools are currently available for predicting which patients are at the highest risk of mortality?

There are frontline studies combining pembrolizumab with standard chemotherapy combinations. Dr Shirish Gadgeel and others presented early phase I/ II data for pembrolizumab combined with pemetrexed/carboplatin, paclitaxel/carboplatin, and paclitaxel/carboplatin/bevacizumab. There will likely be randomized data comparing chemotherapy plus various immunotherapies—not just pembrolizumab—in the frontline setting. Stay tuned. We’ve seen a survival advantage for single agent.Our standard approach for early-stage disease is still histology and stage-driven. Individuals who have nodal involvement are at higher risk for relapse after resection. For stage I, node-negative disease, the area is quite murky. The same meta-analysis that showed benefits for chemotherapy in node-positive disease actually showed a survival decrement for platinum-based chemotherapy in individuals with T1 or T2 or smaller T2 N0 NSCLC.

These folks, by and large, have 5-year survival rates of 65%, 75%, and 85%, depending on tumor size. It will be a lot harder to figure out who’s at higher risk within that group. There are some assays. Just as an example, Myriad Diagnostics is one group that’s looking at a gene-signature assay that may be potentially prognostic and may separate out those who are going to do really well from those who are at much higher risk for relapse in early stage.

In terms of determining risk, you said it’s a bit murky. What would you summarize as the current role of prognostic signatures, specifically those based on cell-cycle progression?

When you say a specific therapy should be sort of attached to it, what therapies do you have in mind?

The big question is, “What do we do with these data?” It’s one thing to predict prognosis, but unless we can attach some sort of therapeutic intervention to that prognostic marker, it doesn’t have much clinical use. One could envision studies in those with poor prognostic signatures, again, comparing the standard which, in that group, is observation, to a platinum-based regimen. In those with good prognostic signatures, maybe we can get by with less than a lobectomy and do a sublobar resection or segmentectomy in that group. That would be a study I would love to see, considering we are discussing relatively small tumors that seem to have potentially more favorable genetic signatures.Prognostic signatures based on cell-cycle progression, as far as I’m concerned, are still a research tool. They are not necessarily ready for primetime. They are of interest. Certainly, in people with a poor prognostic signature, you might be more vigilant about monitoring for relapse or progression. However, until we have a specific therapy attached to those signatures, I do not think they should be used routinely. I do anticipate a future where they may potentially be used down the pipeline.We are still in the era of cytotoxic chemotherapy, so going to immunotherapy or targeted therapy is premature. There are studies in resected NSCLC that are looking at the role of targeted agents and now immunotherapy in that setting. However, those studies are confined to individuals at higher risk. T2 N0 patients, tumors larger than 4 centimeters, and stage II/IIIA patients with N1 or N2 involvement get their chemotherapy, their 4 cycles of platinum-based combination and then, if they’re EGFR-mutant, the current ALCHEMIST trial compares erlotinib (Tarceva) to placebo. If they are ALK-positive, [they receive] crizotinib (Xalkori) versus placebo and the trial has just been modified.

Are there certain patients for whom this test is more ideal?

The other component of ALCHEMIST is now looking at nivolumab versus essentially best supportive care after the chemotherapy has been completed in the marker-negative population. This is essentially, at least within the United States, our go-to trial probably for the next 3, 4, 5 years in the adjuvant setting. How to treat folks with stage I/IA disease is quite controversial beyond resection.Until we have a therapy that can be attached to these tests, I don’t think they are necessarily ideal for any patient yet. In early-stage disease, it’s a tool that can predict prognosis, but until we can differentiate benefit based on that prognostic prediction, using different therapies—whether it’s chemotherapy or other approaches— these tests do not have a routine use from a clinical or patient management standpoint.

How does this test compare with other methods for determining risk? Once it does become more routine, could it be used in conjunction with other factors?

There was a recent FDA submission for brigatinib in ALK-positive NSCLC. What benefit might this drug offer for patients if it is approved?

There are ongoing studies that are starting to look at these tests, and many of the adjuvant trials are now retrospectively looking at these tests based on tissue samples that have been collected.These tests really are among the very few that look at risk of relapse in early-stage node-negative disease. Compared with other gene signature tests, they look a bit more promising. However, they still need to be validated prospectively in much larger trials that also ask a therapeutic question.Brigatinib is a highly active ALK inhibitor amongst the second-generation agents. It seems to be as effective as ceritinib (Zykadia) and alectinib (Alecensa) in this setting in individuals who have already been exposed to, have responded to, and then had disease progression on crizotinib—a very poor ALK inhibitor. Crizotinib really started out as a MET inhibitor.

We see a new generation of agents, which each, at least in the frontline setting, has better PFS outcomes. We’ve recently seen data from the J-ALEX trial, which compared alectinib to crizotinib in the frontline setting and showed major improvement in PFS. The hazard ratio is under 0.40 and really around 0.10 in individuals who had CNS involvement. This whole new generation of agents—brigatinib as well as alectinib—seems to have more CNS penetration than frontline treatment.

Brigatinib is on the fast track for drug approval based on phase I and now randomized phase II data, with response rates in the 50% to 60% range, PFS of 12 to 16 months, and major CNS penetration.

The biggest issue for brigatinib is the relatively low incidence of pulmonary toxicity, but it can still occur and often occurs almost immediately within 24 to 48 hours in about 3% to 6% of patients. Learning how to manage that toxicity—perhaps learning how to prophylax against it—is one of the challenges. From personal experience, brigatinib is extremely well tolerated. I have at least 2 patients now with ALK-positive disease who had progressed on crizotinib who have been on brigatinib now for over 4 years.