TREATING EGFR MUTANT NSCLC AFTER PROGRESSION

This is Mark Kris from Memorial Sloan Kettering, speaking on the growing interest and set of possibilities for patients with epidermal growth factor receptor (EGFR)-mutant lung cancers who have progressed on their first therapy. There are many very exciting, good drugs in testing right now. I would encourage patients to go on clinical trials now in testing these drugs because they present very good options for people with T790M mutations. Until they are available, however, we need to think about the general management of these patients.

There was a very good review article in the journal Cancer^[1] in August 2014. Dr Sacher from Dana-Farber was the first author, with coauthors Pasi Jänne and Geoffrey Oxnard. They included a very interesting table in that paper, and I would encourage you to take a look at it because it describes what happens when you treat an EGFR-mutant lung cancer with a tyrosine kinase inhibitor (TKI). They outlined three scenarios in a good framework that fits well with our National Comprehensive Cancer Network guidelines.^[2] The first—and perhaps the easiest—way to approach these problems at the time of progression, if you can, is to determine a specific type of resistance to the first-line TKI. If you find no pattern of resistance, you clearly need a new therapeutic strategy. Continuation of the TKI is not likely to be helpful. You need to either find a specific kind of resistance or move on to your best chemotherapy as quickly as possible. That probably represents about one fourth of patients.

In another 10% of patients, progression occurs in the central nervous system (CNS) but there is no evidence of progression systemically. When you look at the molecular patterns in these cases, the resistance is not related to second-site T790M mutations. The resistance is in the CNS and is likely the result of pharmacologic problems. The drug is simply not getting to the CNS and the clone is growing. People might say, "Well, that's easy; we will give the next generation of drugs that go after T790M." Please remember that in cases in which this has been looked at, most of the specimens taken so far do not have T790M. Our strategy is to continue the current TKI, treat the CNS disease, and if you have held the TKI during the treatment of the CNS disease, then restart it once that treatment is done.

The third and most difficult management issue is what to do for those patients who have growth of T790M. When you note growth that is characterized by the emergence of a T790M second-site mutation, the vast majority of cells in that patient's tumor are still sensitive to the original TKI. So it is very important to make sure that you are still providing the patient good coverage for the sensitive cells in that mixed tumor. That could be done either by continuing the TKI or by switching to an effective chemotherapy regimen. It's very important to remember that if you stop those drugs abruptly, the sensitive cells will grow and you will have a flare phenomenon. We see this very commonly when we are trying to put patients on clinical trials. In our experience, one fourth of the people in whom we stopped the TKI abruptly, to put them on a clinical trial of another agent for the treatment of progression, ended up in the hospital or died.

The T790M clone clearly needs to be treated, but generally it is not in most cells and it has a somewhat more indolent pattern. Exactly when to start a T790M-specific therapy is slightly uncertain. Recent data from the ASPIRATION trial^[3] suggest that many patients, at the time of progression, can be continued on their current TKI, particularly if it is very well-tolerated, and then we can reinstitute other therapies at some point down the line. Chemotherapy is always a good choice. It would work against the original TKI-sensitive and -resistant T790M clones. The responses to chemotherapy are generally in patients with a T790M resistance mutation. There is a place for local therapy, but these are very individualized decisions.

This is probably one of the toughest situations faced by oncologists. The good thing is that we have many options for these patients. These patients have a very long survival, and we will be faced with many different treatment decisions. When do you start local therapy? When do you continue the original TKI? When do you switch to chemotherapy? When do you go to a next-generation TKI? More data are going to come out in the months and years ahead about how to do this, but it's tough.

When it comes to progression in a patient with EGFR-mutant lung cancer, it is by no means one-size-fits-all. If you had to go to one thing it would probably be chemotherapy, but you have to remember that the cancer that you are fighting at that point is likely to be heterogeneous. Sometimes there are pharmacologic issues. Sometimes there is molecular emergence of new resistant clones. You need to put your game face on and work very hard to take care of these patients. It's a very rewarding field because these patients can live a long time with the right treatment. This is going to be our challenge in the years to come.