Dr. Tony Mok, a friend who lives and works in Hong Kong and is an international leader in lung cancer, has just published not only the IPASS trial in the New England Journal of Medicine but also another trial called FAST-ACT (for First Line Asian Sequential Tarceva and Chemotherapy trial) in the Journal of Clinical Oncology. While the IPASS trial directly compares first line chemo alone to the EGFR inhibitor Iressa (gefitinib), the FAST-ACT trial was a randomized trial (phase II, so smaller and less definitive than a phase III randomized trial) that directly compared chemo alone to chemo with the oral EGFR inhibitor Tarceva (erlotinib). Importantly, however, in recognition of the justifiable concern that chemo and oral EGFR tyrosine kinase inhibitors given concurrently may well interfere with each other, the investigators gave chemo and either Tarceva or placebo on an alternating basis during a 28-day cycle, in which cisplatin or carboplatin was given on day 1, gemcitabine on days 1 and 8, and then Tarceva or a placebo was given on days 15-28, before chemo restarted. This way, chemo and Tarceva didn't overlap, which should reduce the risk of a negative interaction between them. (click on image to enlarge)

Unlike the IPASS trial, this trial was not done exclusively in Asia (~6% of patients were Caucasian), and only a third of patients were never-smokers, so this is a population not as likely to receive a profound benefit from an EGFR inhibitor. In a nutshell, the study found that the patients who were assigned to an alternating combination of chemo and Tarceva had (non-significant) trend toward a higher response rate (35.5% vs. 24.4%; p = 0.12) and also a better median progression-free survival (PFS), best shown in the shift of the combination curve to the right, below: That's encouraging, but there was no improvement or even a hint of one in overall survival. These results were comparable between never-smokers and current or former smokers. And this is despite the fact that all of the patients on the Tarceva arm got the active drug, while only 73% on the placebo arm were later given Tarceva. We see this problem over and over. It would be more appropriate to ensure that everyone got the same access to the drugs, with the difference being the timing/schedule of treatments. This brings up an issue that comes up over and over in lung cancer, and frankly in cancer in general: namely, is there really a benefit to improving PFS without improving overall survival (OS)? This is a debatable point, but personally, I'm pretty dubious about the benefit of PFS with no OS benefit. Essentially, one set of very sensible conclusions to draw from the IPASS trial is that EGFR mutation positive patients get more upfront benefit from an EGFR inhibitor and EGFR mutation negative patients get much more upfront benefit from chemo, but there's no survival difference if you end up giving both sequentially, whether you start with chemo or EGFR inhibitor therapy. This is shown in the schematic below, really highlighting the difference between order A and order B (sorry for the stark-ness at the right side): Here, if you happened to get the "right" first line treatment (based on your EGFR mutation status here), you'd follow line B instead of A and have a longer PFS, but there's essentially no difference between getting 18 months of benefit starting in 2008 or 2009 if you have an EGFR mutation. You can add the value of chemo before or after and get the same total benefit over time. This likely applies to plenty of situations, including timing of starting treatment for slowly growing cancer, etc. My concern with the approach described in FAST-ACT is that there isn't any established value in pushing a limited number of treatments upfront, so that you've now burned up two effective treatment strategies and condensed them into one. If all patients on the placebo arm had actually received tarceva, perhaps they would have done better still. In other words, I think there's a real possibility that the two compressed together may produce less than the some of their parts, even if it produces a longer initial PFS (line C in the figure above). Overall, this trial doesn't tempt me at all to combine chemo and an EGFR inhibitor, even if they don't overlap. In fact, my general approach is to spread out the available effective treatments over time, rather than to use any more of our limited treatments than necessary at a given time to keep the cancer from progression. A front-loaded combined approach is like a "sprint", and while we aren't yet at a point where managing advanced NSCLC is quite the "marathon" of a chronic disease we'd like to have (short of a cure, of course), I do believe it's important to try to ensivion a treatment plan over the long haul, which includes distributing your effective treatment options over time.