It seemed inevitable, I think.  Several immuntherapy agents known as PD-1 or PD-L1 checkpoint inhibitors have been shown to lead to better survival, higher response rates, and a better side effect profile than second line chemotherapy with Taxotere (docetaxel) in broad populations or subsets with tumors positive for the marker PD-L1 that correlates with a higher probability of a good response. At this point, both Opdivo (nivolumab) and Keytruda (pembrolizumab) are FDA approved second line treatments that are widely favored over Taxotere, though the approval for Opdivo doesn't require PD-L1 testing or a specific level, while Keytruda's approval is for only patients with PD-L1 expression (technically >50% expression, the high level seen in about 25-30% of NSCLC patients, though it also has (somewhat less) activity in patients with ANY degree of PD-L1 expression, which gets up to 70-75% of NSCLC patients).   And patients who progressed could cross over to receive Keytruda, so the trial didn't just compare patients who got Keytruda first line to patients who never got a chance to benefit from an immunotherapy.

Of course, as the song goes, "how ya gonna keep 'em down on the farm, after they've seen Paree?". Once these clearly active agents got an inroad, the gravitational pull into first line treatment has been tremendous.  There are literally dozens of trials that have been done comparing immunotherapy to standard first line treatments -- single checkpoint inhibitor vs. chemo, single checkpoint inhibitor + chemo vs. chemo alone, single checkpoint inhibitor vs. chemo vs. combination, combinations of a PD-1 or PD-L1 checkpoint inhibitor with a different immuotherapy agent such as a "CTLA-4" inhibitor, and on and on. Multiple checkpoint inhibitors that are remarkably similar are each the subject of a similar and largely redundant group of trials, and in patients with an EGFR mutation these trials are being done against and in combination with EGFR inhibitors, in patients with an ALK rearrangement these trials are being done comparing to and in combination with ALK inhibitors, etc.  In the coming years, there will be an avalanche of trial results on them.

But today, we got the first press release on results, here with Keytruda compared to first line chemotherapy alone in the subset of patients with high PD-L1 expression (>50%), which is the ~30% of patients most likely to do especially well with immunotherapy.  Though we don't have any numbers yet, we learned that the trial was positive for the primary endpoint of progression-free survival (PFS) as well as for the secondary endpoint of overall survival (OS). It enrolled 305 patients, randomized to either Keytruda at a fixed dose of 200 mg (a relatively low dose, compared with prior tested doses of 2 mg/kg or 10 mg/kg in other studies) IV every 3 weeks, compared with any of a range of platinum-based doublet combinations. Of note, the far more specific side effect data presented in the press release applies not to this particular trial but other data with the agent, and not at the dose tested, so I don't think we can say much about the efficacy or side effects of this regimen in this patient population until we see the actual trial data presented.

Where does this leave us? I certainly understand the compelling urge to move Keytruda into first line treatment, but I think it made be hard to get it covered without the actual data being presented and without an approval.  But in addition, I think it's important to clarify what this likely means and doesn't mean.

First, I think it's very valuable that survival is actually significantly improved. Though we generally favor giving our best and best tolerated treatments first, it's powerful to see a survival benefit from giving Keytruda first, showing that you can't just presume you'll make up the difference by giving it later.

Second, it's critical to note that this is only in the limited subset of patients who we identify as the ones most likely to benefit from immunotherapy. The results could be very different, and even the opposite, in patients with low or no PD-L1 expression. Here, I think it's valuable to think about EGFR inhibitors like Iressa (gefitinib), Tarceva (erlotinib), and Gilotrif (afatinib), which we know have great activity in patients with an EGFR mutation that is much more than that of first line chemotherapy. However, studies have shown that we harm patients by having an EGFR inhibitor move ahead of chemotherapy if you don't have an EGFR mutation or don't know the person's mutation status.  Immunotherapy may show the same pattern: for those identified as having the greatest probability of good response to immunotherapy, it makes sense to start with it, but for others, chemotherapy may be better, and starting people on immunotherapy may lead them to progress on it quickly and then be too sick to benefit from the chemotherapy they could have done well with if they had started with it. So these results do not apply to unselected patients (who haven't had PD-L1 testing) or those who test for no PD-L1 or even lower level PD-L1 (less than 50% expression), who may do better by starting with chemo. Other trials will give us the answers to these questions.

Also, this is Keytruda alone, compared with chemo alone. We can't presume that it's better or worse than immunotherapy combined with immunotherapy. We can't presume that it's better or worse than a combination of two immunotherapy agents.

This is with a fixed dose that is pretty low. That's great and convenient, but it may not be a dose sufficient to lead to good responses for patients who have tumors less sensitive to immunotherapy.  In a trial with Keytruda given to patients with any degree of PD-L1 expression, results were better at the higher dose of 10 mg/kg (much higher than used in the KEYNOTE-024 trial) than at the current FDA approved dose of 2 mg/kg. Even if Keytruda turns out to be effective as first line therapy in more than the 25-30% of patients with particularly high tumor expression of PD-L1, a broader population may need a higher dose to do well with immunotherapy. Again, we'll learn more with upcoming trials.

Finally, this trial marks the first clear evidence of a value in testing for PD-L1 in advanced NSCLC. Up until now, though PD-L1 testing could help predict who is more or less likely to benefit, Opdivo is approved regardless of PD-L1 status, and the benefit with it is seen in an unselected population (i.e., no testing required). Since just about every patient wants their opportunity on immunotherapy and are eligible, in the US at least, to receive it in second line, there hasn't been a clear answer to why you would do PD-L1 testing. But now, in contrast, testing for PD-L1 right at the time of diagnosis, can serve to distinguish between the patients best served by up front immuotherapy and those who may not be. So this will mark a new standard of care once Keytruda becomes available as first line therapy for patients with high PD-L1 expressing tumors, at least unless a different PD-1 or PD-L1 checkpoint inhibitor shows benefit and is approved in everyone, regardless of PD-L1 status.

This is just the beginning. Immunotherapy is moving into the first line, at least for some patients, and we will only know more than we do now about which patients, which agents, and how best to sequence all of the options.

What do you think?