Here's the first of a series of posts on key presentations on lung cancer from ASCO 2010, as reviewed by myself and Dr. Nate Pennell of our faculty here several weeks ago. The first topic we covered was the very interesting if troubling Canadian BR.19 trial of post-operative Iressa (gefitinib) vs. placebo, as summarized by Dr. Pennell. Dr. Pennell: So this is certainly one of the most important presentations at ASCO this year, I think, if nothing else to show us that we aren't nearly as smart as we think we are. Gefitinib, as I'm sure most of our GRACE community knows, is an EGFR tyrosine kinase inhibitor very similar to Tarceva, which is the approved drug here in the United States. And back in the early 2000s there was a lot of enthusiasm for exploring the potential for gefitinib in the adjuvant setting for patients with early stage disease to see if it improved cure rates essentially. And so the BR.19 trial was launched in the first half of that decade in resected patients with stage I-B through IIIA non-small cell. These patients could receive adjuvant chemotherapy as appropriate, I think, once the adjuvant chemotherapy trials in 2003 and 2004 came out and then were randomized in one-to-one fashion to either two years of daily gefitinib or to placebo. (click on image to enlarge)

Now, this trial was halted early. I don't remember what year it was halted in: 2005 or so. Dr. West: Yeah, about in 2005, it was halted after the ISEL trial came out negative and then they looked at the SWOG study and saw that that was negative and even potentially harmful with Iressa. Dr. Pennell: Right, the SWOG 2023 study, which was for locally advanced lung cancer after chemotherapy and radiation. Patients were randomized in a very similar fashion to gefitinib or placebo. And when the results first came out, it showed that there appeared to be a worse survival in the gefitinib arm. That, in addition to the ISEL trial, which was a randomized trial of gefitinib versus best supportive care in second-line treatment of lung cancer failed -- then this trial was halted. However, about half of the patients that were planned had been enrolled. It was supposed to be over a thousand and only about 500 were enrolled, but we do have the survival results from those 500 patients, and here are the curves. So, importantly, this yellow line, the one here on the bottom is the gefitinib arm; so most of your trials you really want to see your active treatment arm be up here above the control arm, but statistically this was a negative trial, with no difference in overall survival between the two arms. However, at least numerically, as you can see, the line for the gefitinib treatment was clearly below the control arm, with a hazard ratio of 1.23 which statistically means that there was actually a 23% worse risk of death in the gefitinib arm, although this was not statistically significant. While they are not statistically different, it certainly is troubling, especially in the setting of knowing from the SWOG study, that there also appeared to be a worse survival in patients treated with gefitinib compared to placebo. In disease-free survival, again, there was no difference, with perhaps numerically slightly worse outcome in the gefitinib group. On multivariate analysis, gefitinib treatment didn't fall out as one of the potentially harmful prognostic factors. Only advanced age and large tumor size were significantly associated with shorter survival, although there actually was a numeric trend towards the gefitinib being a risk factor for a worse outcome. One potential explanation for this could be, well, perhaps the patients on gefitinib were dying because of toxicity from the drug. If you kind of scan through this, there weren't a whole lot of significant toxicities in either of the arms and did not appear to be anything significantly different between them. The one major toxicity we always worry about with the EGFR TKI-induced pneumonitis, which is a serious inflammation of the lungs, which again really was not different between the placebo and gefitinib arms. So it doesn't appear the toxicity is necessarily a reason for why there might be some harm. Now, this was a very interesting slide from this presentation. At the very least, if there was no benefit in unselected patients from adjuvant gefitinib, you would hope that patients who had sensitizing EGFR mutations at least would have a better outcome, and that's certainly what I would have expected to see. And we do have to take into account that there are only 76 patients with EGFR mutations included in this analysis, only 40 in the placebo arm and 36 in the gefitinib arm. However, again, as you can see the yellow line here is down below the red line, the gefitinib arm numerically, again, appeared to have a worse overall survival than the placebo arm. Again, not statistically significant, but troubling and I would have expected this yellow line to be pretty far up here and it wasn't. I don't have a good explanation for that. So, to put this in perspective, I think it's important for us to recognize that the drugs that we use are not completely benign, and we are often presented with results that we don't expect, which potentially have put patients in harm's way. And that's why it's very important that we actually do these studies before we move these treatments into regular therapy. Five years or so ago at ASCO there was a show of hands about how many people would give adjuvant gefitinib or erlotinib (Tarceva) in patients if you knew that you had an EGFR mutation. And though I wasn't there, I was told that many of the people in the room raised their hand, and would have treated someone without knowing the results of these trials with adjuvant tyrosine kinase inhibitors, and I would say until we have definitive evidence that it shows a benefit, I would not treat someone with adjuvant EGFR TKIs, even if necessarily they had a mutation, unless it was on a trial. Dr. West: Well, we've certainly seen in the last several years from the IPASS study and many others that it's a completely different question in patients with versus without an EGFR mutation. So even the detrimental effect seen on the SWOG 0023 trial was in an unselected population, and just about everybody would say, "Well, even if that occurred, we could not anticipate that this would be the same in patients with an EGFR mutation." And that is part of what is so shocking about this is that, as you had mentioned, the patients with the mutations being harmed is even more inconceivable than the rest of the study, and yet that's what it showed. Moreover, the patients who were on the study stopped taking either placebo or gefitinib as soon as the news broke about the potential detrimental effect, and the median duration of treatment was under five months with either of these. And so, not only did you see these differences, but this wasn't even with years and years of therapy, but only in five months -- and we just cannot explain it. At least I can't, but I don't think anybody can. Dr. Pennell: No, and the question came up today when I was giving a talk: why is this different than in the IPASS study? Why are these drugs potentially harmful in the setting when they are so dramatically better than chemotherapy in EGFR-mutant patients in the metastatic setting? And there's really no good explanation for it. You can hypothesize that perhaps if there is micrometastatic disease out there that you're trying to eradicate, that if this was more of a static drug that was sort of just holding it in check rather than killing it. Sometimes you can hypothesize perhaps some kind of a rebound, where once you've stopped the drug, it might later be unleashed and be even more aggressive than before. But that's just a hand-waving possible answer. I don't really have a good explanation. ------------------------------------------------ We welcome your comments and questions, and especially a good explanation if you have one.