The IPASS trial that randomized never-smoking Asian patients with a previously untreated advanced lung adenocarcinoma to either standard chemo with carboplatin/Taxol (paclitaxel) or the oral EGFR inhibitor Iressa (gefitinib) was a pivotal study that changed how many of us thought about NSCLC. Clinical factors such as patient race, smoking status, tumor histology, and potentially patient sex have historically been useful in predicting which patients are most likely to benefit from an oral EGFR inhibitor (with Asian, never-smoking status, adenocarcinoma and especially bronchiooloalveolar carcinoma (BAC), and women being prevalent features of major responders). However, the IPASS study showed that the molecular marker of EGFR mutation is clearly more important than these clinical factors: in those patients who have an EGFR mutation, Iressa was associated with a far better response rate (RR) and progression-free survival (PFS), as well as a trend toward a more favorable overall survival (OS). On the other hand, in those who don't have an EGFR mutation, even among Asian never-smoking women with an adenocarcinoma, chemotherapy was a clearly superior option. It is worth noting that this is specifically for the question of first line therapy, for which chemotherapy is the default standard therapy -- the results comparing chemo to EGFR inhibitor therapy as second line treatment have been very comparable in broad populations. Eligibility on the IPASS trial was based on clinical selection for patients more likely to benefit from EGFR inhibitor therapy, and it was only in a planned retrospective analysis that the importance of the very different results by EGFR mutation status became apparent. A slightly different question emerges if you know that a patient has an EGFR mutation before you start treatment. The IPASS trial strongly suggests that patients with an EGFR mutation will be better served by receiving first line EGFR tyrosine kinase inhibitor (TKI) therapy, and this has now been confirmed in three independent prospective randomized trials. One pivotal study of this concept was done by the North-East Japan Study Group, performing a remarkably similar trial to that done by IPASS: carbo/Taxol or Iressa as first line therapy in patients with a prospectively identified EGFR mutation, using PFS as the primary endpoint. Though it was designed to enroll 320 patients, it closed early, when an interim analysis showed such a remarkably superior improvement in PFS that the Data Safety Monitoring Board considered that continuing to randomize patients on the trial would be unethical: (click on image to enlarge)

In every clinical endpoint, Iressa emerged as superior: RR was 73.7% vs. 30.7% (p < 0.001), median PFS was 10.8 vs. 5.4 months (p < 0.001), and median OS was 30.5 vs. 23.6 months (p = 0.31). It's worth noting that while the median OS is numerically quite a bit longer in the Iressa arm, when you look at the curves in the bottom right of the figure above, you can see that there happens to be a bubble right at the 50% point, but there isn't a major difference overall (hence the non-significant difference). Still, there is nothing to recommend first line chemo if you know someone has an EGFR mutation, especially since the side effect profile of Iressa was nearly limited to the expected rash and diarrhea, compared with significant drops in blood counts, fatigue, and several other side effects being more common with standard chemotherapy. Remarkably similar results were also seen in another Japanese study performed concurrently, this one done by the West Japan Thoracic Oncology Group, including PFS as the primary endpoint and with the same design except that the chemo arm received cisplatin and Taxotere (docetaxel). Here, 172 eligible chemo-naive patients were enrolled at the time that the IPASS trial and preliminary results of the North-East Japan trial were initially presented, and this trial was also closed early for concerns of further randomization being unethical. Similar to the trial done by the North-East Japan Study Group, this study demonstrated a very significant improvement in PFS among recipients of first line Iressa (median PFS 9.2 vs. 6.3 months, p < 0.0001), though there wasn't any evidence of any OS benefit from earlier EGFR inhibitor therapy in this trial (median OS 30.9 months for Iressa, and not yet reached with first line chemo). If anything, on this trial, the group that appears to be doing better over time (right side of the curves) on the survival plots in the bottom right of the figure above is the recipients of first line chemo. The authors didn't present data on response rate, and there were no surprises with regard to side effect profiles of the two strategies. The final trial testing this approach that we now have evidence on is the so-called OPTIMAL trial from China, with preliminary data coming out of the Chicago Multidisciplinary Thoracic Oncology Conference in December, 2010. Here, Dr. Caicun Zhao outlined a trial in which 154 patients with advanced NSCLC and a known EGFR mutation who had received no prior systemic therapy were randomized to chemotherapy with carboplatin/Gemzar (gemcitabine) or the EGFR TKI Tarceva (erlotinib) as first line therapy. The primary endpoint was again PFS, with a very significant nearly three-fold incraese in median PFS in the Tarceva arm (13.1 vs. 14.6, p <0.0001): Tarceva also produced a very significantly higher RR of 83%, vs. 36% with chemotherapy (p < 0.0001), about as high as we've seen for first line therapy in advanced NSCLC in any circumstance. The disease control rate, which includes minor responses as well as stable disease along with major responses, was all the way up to 96% in the Tarceva arm, vs. 82% with chemo (p = 0.002) (the latter being very respectable for chemotherapy as well, and reminding us that patients with an EGFR mutation also tend to do well with conventional chemotherapy). Fewer patients on Tarceva discontinued treatment than on standard chemotherapy. There was no presentation of survival, which is obviously a very relevant issue. With these three trials, there is no question that those advanced NSCLC patients with a prospectively identified EGFR mutation experience a significant improvement in PFS with initial EGFR TKI therapy, a result consistently seen with either Iressa or Tarceva. Accordingly, this approach of first line EGFR TKI is now recommended by the National Comprehensive Cancer Network in such patients. However, the argument can be made that the PFS benefit is related to the fact that the recipients of EGFR TKI received it on an ongoing basis, while those who received chemo stopped it after a fixed course of treatment and that some patients may have had a longer PFS if they had either continued with the same first line chemotherapy, dropped an agent and continued on single agent chemo (potentially also with the anti-angiogenic agent Avastin (bevacizumab)), or switched to a new chemo agent. Also very important is the looming question of whether there is really a durable advantage, as it may arguably be just as good to get the EGFR TKI as maintenance therapy, second line treatment, or later. Though there are trends toward an OS benefit in the mutation-based subset analysis of the IPASS trial, as well as on the North-East Japan trial reported above, the West Japan trial suggested a trend in the other direction, and the OPTIMAL trial has reported no OS data either way. Meanwhile, we know that maintenance Tarceva on the SATURN trial was associated with a HUGE improvement in PFS for the subset of patients with an identified EGFR mutation (hazard ratio of 0.10, which means a 90% improvement) but absolutely NO improvement in OS, presumably because the crossover of all of these patients to Tarceva at progression led them to do every bit as well as the people who benefited profoundly from Tarceva a little earlier. Whether the timing of getting an EGFR inhibitor is really important for someone with an EGFR mutation over the long run is still an open question, with the clearest result being that getting one of these drugs at some point is the most important factor. However, given that there is always a potential for unexpected problems to intervene and keep someone from getting the next intended treatment, there is a growing consensus that we'd really want to push toward giving the most effective treatment early. We'll hope to get longer-term survival information from these trials in the future, and in addition, there is another European trial of first line chemo vs. Tarceva called EURTAC that we expect will help provide further valuable insights. In the meantime, most experts would favor initiating treatment with a first line EGFR inhibitor in someone with a known EGFR mutation on the basis of what we know now.