One of the trials presented at the Chicago Multidisciplinary Symposium in Thoracic Oncology last month was the TITAN trial, one of a pair of studies conducted in Europe to test the oral EGFR inhibitor Tarceva (erlotinib) in patients with chemotherapy pre-treated advanced NSCLC. The other trial, SATURN, was designed to test Tarceva as a maintenance therapy vs. placebo in patients who had shown a response or stable disease after four cycles of first line chemotherapy (without the VEGF inhibitor Avastin (bevacizumab)) has been summarized previously and ultimately led to the approval of Tarceva as a maintenance therapy in this patient population. But what happened to the significant fraction of patients who progressed by the time of the repeat imaging after four cycles of first line platinum-based doublet chemotherapy? They were directed to the TITAN trial, which was a head to head comparison of Tarceva vs. either Taxotere (docetaxel) or Alimta (pemetrexed), both well studied and commonly used second line agents for advanced NSCLC. The trial looked for an improvement in overall survival with Tarceva. (click on image to enlarge) The trial closed earlier than planned, due to slow enrollment, with just 424 patients, which leaves it quite underpowered to detect a difference even if there really is one between the two treatment approach. Still, there may be some conclusions that can be drawn from what they saw, even if limited by smaller numbers than needed to say anything definitive.

The results really showed comparability of Tarceva with chemo, with both arms showing the same median survival of 5-6 months, even when looking specifically at different subgroups as defined by performance status, smoking status, and histology. Because 30 patients with squamous cell NSCLC received Alimta (this trial was conducted before it became clear that Alimta isn't effective in this NSCLC subset), a separate analysis was also done that excluded those patients who we might presume would do worse because they hadn't received a good choice of treatment based on what we know now, but that didn't change the overall picture of essentially equivalent results. A subset of patients had tissue available for testing, nearly all of whom having EGFR wild type. In this group, the recipients of Tarceva happened to have a non-significantly longer median overall survival than the recipients of second line chemotherapy (6.6 vs. 4.4 months, HR = 0.85, non-significant). Given the limited size of the study, the findings can only be taken so far. Nevertheless, they corroborate the results of the INTEREST trial that directly compared Iressa (gefitinib) to Taxotere, further supporting the idea that an EGFR inhibitor is not an inferior second line treatment compared with standard chemotherapy. Moreover, like the molecular marker breakdown of the INTEREST trial, this work indicated that EGFR tyrosine kinase inhibitor therapy is comparable to chemo even in those patients who don't have an EGFR mutation. The second striking result is that if you compare these results to the sibling SATURN trial (same patient eligibility, except for separation depending on who achieved response or disease control vs. progressed), you can see a huge effect of selection bias: there is a huge effect of the characteristics of the patient population enrolled on your study. In the TITAN trial of patients who progressed within four cycles of first line chemo, the median overall survival was about half of what we see in the SATURN trial and other maintenance trials that filtered out patients with less favorable cancer biology. Clearly, the biology of the cancer (and probably the health of the patient) is more favorable overall in the patients who respond or at least show stable disease. The cancers that are more responsive initially are also the ones most likely to benefit from subsequent therapy, while unfortunately, the opposite is also true: we seem to have much less of an impact with our therapies, regardless of the next choice, in the patients who show earlier progression. Still, we can at least say that these data support the concept that either chemo or Tarceva achieves comparable results in the second line setting.