Stephen Liu, MD:It’s not currently standard to order genetic testing for someone with locally advanced nonsmall cell lung cancer. In my own practice, we do it, and it influences my surveillance in different ways. For example, if a patient is known to have anALKrearrangement, I know that that particular cancer has a very high tropism for the brain, a much higher risk for relapse within the CNS [central nervous system]. Part of my surveillance for that patient will include a contrast-enhanced MRI [magnetic resonance imaging] of the brain. It doesn’t really influence my management, but I also have that information during treatment. It brings up certain interesting subsets within the PACIFIC trial data.
If we look at patients in PACIFIC who received durvalumab, patients with a sensitizing mutation withinEGFRdid not seem to derive the same benefit. One could argue that for those patients who may not be getting as much benefit from durvalumab, that perhaps we should withhold durvalumab from patients with anEGFRmutation.
These are unplanned subset analyses, but they are quite compelling to discuss because we also know that if patients receive a checkpoint inhibitor, such as durvalumab or a PD-L1 [programmed death-ligand 1] inhibitor and then immediately afterward receive a TKI [tyrosine kinase inhibitor], there’s quite a bit of toxicity. The rate of severe immune-related adverse events when TKIs follow immunotherapy is quite high, about 15% in some retrospective series.
If you started immediately after a checkpoint inhibitor, within the first 3 months, that number goes up to 24%. Whereas, if you start with a TKI and then go to immunotherapy, we don’t see that toxicity. So the stakes are higher. If I deliver durvalumab to someone with a sensitizingEGFRmutation, complete durvalumab therapy, and they have progression of their disease, I have now made administration of a kinase inhibitor, such as osimertinib, which is much more difficult and much more toxic.
There are 2 different sort of different camps. There is 1 saying this is going to make TKI delivery higher, and therefore I should not give that checkpoint inhibitor. The other is saying I’m going to give the checkpoint inhibitor and hope that obviates the need to use a kinase inhibitor, hope that the durvalumab here improves the chance of cure, so that I wouldn’t need to bother with kinase inhibitor in the future. I don’t know which strategy is optimal. These are relatively small numbers, and these are subsets.
In my own practice with potential cure, I do administer durvalumab, but it’s something I discuss with the patients. I go over the data that are there, the unknown answers and the uncertainty in that setting, and try to share that with the patient. I am somewhat conflicted about how to manage those patients ideally. In my current practice, though, I still do not withhold durvalumab from patients with a sensitizing mutation. I still deliver it, hoping that synergy really gives us long-term survival, that cure that we’re hoping to achieve.
If the patient was known to have a tumor that did not express PD-L1, I would still administer durvalumab. We know from some post hoc subset analyses, patients with a PD-L1 score of less than 1% did not seem to derive benefit from the addition of durvalumab. It’s important to understand these were unplanned subset analyses. Many patients did not have tumor evaluable for PD-L1.
While the EMA, the European Medicines Agency, did not approve durvalumab for PD-L1negative non–small cell lung cancer, it is approved in the United States, acknowledging that these are subsets, these are unplanned, and PD-L1 expression is somewhat heterogeneous. This is a stage III non–small cell lung cancer, which is much more likely to be diagnosed by fine-needle aspirate or cytology. But PD-L1 expression may not be quite as reliable and may account for the fact that large numbers within PACIFIC were not evaluable for PD-L1 expression.
It is an area of uncertainty. In that setting, I wouldn’t withhold durvalumab. Overall, it’s a relatively safe drug. There’s potential synergy between durvalumab and radiation, which can change PD-L1 expression. The biopsy may not reflect the PD-L1 status of the tumor as a whole, and because of the overall survival benefit seen in the all-comer approach with the addition of durvalumab, it would still be my practiceregardless of PD-L1 expression—to offer durvalumab to patients with stage III nonsmall cell lung cancer.
The current standard for locally advanced nonsmall cell lung cancer is definitive chemoradiation with platinum-based chemotherapy followed by consolidation durvalumab, a PD-L1 inhibitor for 1 year. That would be our current standard, and that is how I approach all patients with an unresectable non–small cell lung cancer.
I think where the field is going is looking at patients with resectable nonsmall cell lung cancer. Immunotherapy certainly would have similar, if not even greater, effects in that setting. Can we incorporate the use of checkpoint inhibitors in patients with resectable tumors in the neoadjuvant setting? This is regarding neoadjuvant immunotherapy, neoadjuvant chemoimmunotherapy. We’ve seen from some early, relatively small studies that the rate of major pathologic response with chemoimmunotherapy approaches 80%. These are quite high responses.
Does this translate to better survival? When we think of a stage I or a stage II resectable lung cancer, our cure rates are nowhere near what they are for breast cancer, prostate cancer, and colon cancer. A significant number of patients still dying from cancer, even with a stage I lung cancer that’s resected. For a stage I, the smallest nonsmall cell lung cancer removed, we still have 30% of patients dying from lung cancer. We clearly need to do better. Administration of checkpoint inhibitors in that setting may get us to the cure rates that we really want. I expect over the next year or so we’ll start to see some of these neoadjuvant studies promising results being discussed and published.
If those will translate to long-term survival will take a little more time. But seeing that early efficacy is something I’m really looking forward to. I think unmet needs are figuring out how to manage patients with driver mutations, driver fusions. AnALK-positive tumor,ALKnonsmall cell lung cancer, is typically not an immune-responsive tumor, but is there a role for immunotherapy in the earlier-stage setting? Is there a role for TKI therapy in that setting, either before or after surgery or after chemoradiation? Studies exploring targeted therapy for earlier-state locally advanced non–small cell lung cancer are things we’re really looking forward to in the coming years.
Transcript edited for clarity.
Case: A 62-Year-Old Male With Stage III NSCLC