Solange Peters, MD, PhD: In metastatic patients in the United States and in most of the European countries, immunotherapy has dramatically changed second-line treatment. We were used to docetaxel since 2000, when it was shown to be better than nothing, but docetaxel is difficult to deliver—it has lots of toxicities, it has very poor tolerability, and patients hate it. So, it was something we were doing because of guidelines but not really, because we were improving the fate. We were just prolonging survival, but the conditions were not very good. Since the 4 randomized trials showing that with immunotherapy, whichever compound you use is better than docetaxel, we were all very happy to adopt immunotherapy in that setting regardless of PD-L1. Docetaxel is such a bad treatment that whatever the category—PD-L1 low or PD-L1 high—anything would be better than docetaxel. Of course, the management is different. If PD-L1 is low, you have a small difference. If PD-L1 is high, you have a big difference. But it’s better than docetaxel. Dramatic change—most of the European countries have some access to immunotherapy in the second-line setting, 1 drug or the other one.

The next advance was to move it to the frontline setting. And, again, by selecting patients for being potential responders—and this was based on taking the very high PD-L1 expression, with more than 50% of the tumor cells expressing PD-L1—pembrolizumab was shown to be better than chemotherapy. But it’s not docetaxel, it’s platinum-based chemotherapy, which is a good chemotherapy. So, in this scenario, you can use pembrolizumab. It was granted approval in Europe in December last year, and countries are, time after time, starting to reimburse it in frontline in high PD-L1 expressers—so, second-line for all patients and frontline in a selected number of patients, maybe 20% to 30% of them. And, of course, every country has its own regulation and reimbursement processes. Unfortunately, in Europe, we are now in an imbalanced environment where the chances to get access to this kind of drug are not the same in every country. We have to work on it.

Marina Chiara Garassino, MD: All patients must be tested for PD-L1, because if you have a proportion score of PD-L1 more than 50%, you are allowed to treat patients with pembrolizumab. And we know that the results of the KEYNOTE-024, which compared the pembrolizumab with chemotherapy, showed, with a long follow-up, very good results for these kinds of patients, and the prognosis of patients with advanced non—small lung cancer treated with pembrolizumab and with a proportional score of more than 50% is about 30 months, which is very different from what we had before.

I’m not sure that I will give you the same answer in the near future, because the treatment of advanced non—small cell lung cancer in 2018 is challenging. We will have to wait and see for at least 2 trials, which are the BMS227, comparing the monotherapy and the chemotherapy plus the anti-CTLA4 in the metastatic setting, and the KEYNOTE-189, comparing the use of chemotherapy with chemotherapy plus the pembrolizumab again. We will have to see very carefully if, in the subgroup analysis, they will be independent or not about the expression of the PD-L1, because there is a strong rationale to combine both the anti-CTLA4 and also chemotherapy. So, it will be quite challenging to tailor the treatment for this kind of patient.

The story is that the older tests are not interchangeable. It is a quite interesting story, because all the companies developed the drugs with their own tests, and so we have the drugs with the test of the company, and there are several trials trying to harmonize all the tests. What seems clear is that the one test—the SP142, which is the Ventana test—seems to be different from the others. And another important piece of information comes from the harmonization test of the French group that says if you do a laboratory-derived test on your own, maybe you are not so reproducible as the label test. So, in my opinion, it is better to use the commercial tests that are proven to be the same—and, for example, 1 is that one of pembrolizumab, which is the 22C3 test.

Don’t do in-house testing because it can be not risk producible, and it is better to use commercial tests, with the exception of the Ventana SP142 assay, which can be quite different from the others.

Transcript Edited for Clarity