Just prior to ASCO, I mentioned the early results of the Cancer and Leurkemia Group B (CALGB -- Group A long-since defunct) 30406 trial. This study enrolled 181 people with advanced lung adenocarcinoma who had either never-smoked or had a rather minimal smoking history of 10 "pack-years" or fewer, and who were randomized to receive the epidermal growth factor receptor (EGFR) tyrosine kinase inhibitor (TKI) Tarceva (erlotinib) alone or in combination with standard chemo of carboplatin and Taxol (paclitaxel) as first line treatment. The idea of this trial was that the selection of patients by the clinical factor of smoking status was meant to enrich for a higher probability of such patients having an EGFR mutation, and it had also been recognized that never-smokers (less than 100 cigarettes in their lifetime) or "oligo-smokers" (oligo meaning few, so the term means those ex-smokers who had a light smoking history but didn't qualify as never-smokers) in many clinical trials of EGFR inhibitors appeared to be the greatest beneficiaries of an EGFR inihibitor like Tarceva or Iressa (gefitinib). At the same time, there was reason to question whether combining chemo with an EGFR inhibitor might provide additional benefit, have no effect, or might even be detrimental. (click on image to enlarge)

The patient population enrolled on this trial was about what you would expect for this subset within NSCLC: it skewed young (median age 58-60 in the two groups), about 60% women (vs. more men than women in a broader lung cancer population), and on this trial was about 80% never-smokers and 20% with a light former smoking history. The authors also determined EGFR mutation status for nearly everyone on the trial and found that about 40% of the patients in this predominantly Caucasian (~80%), US-based trial of never- or oligo-smokers had an EGFR mutation, the remaining 60% being EGFR "wild type" (no mutation). This ratio is the reverse of the prevalence of an EGFR mutation in the IPASS study of Asian never- or light former smokers with a lung adenocarcinoma. When we look at the overall results, independent of EGFR mutation status, it's hard to get a sense of any meaningful differences between the two approaches. Results in terms of both progression-free survival (PFS) and overall survival (OS) are quite similar. There is a higher response rate (RR) in recipients of chemo with Tarceva, but it's not clear that this is better than giving one followed by the other, since we wouldn't presume that you only have one opportunity to treat these patients and need to give every treatment at once. The authors then broke down the results by EGFR mutation status, but they're presented separately for the two treatment approaches. Not surprisingly, with Tarceva alone, patients with an EGFR mutation do remarkably better than those without one, in terms of PFS, OS, and RR, all highlighted below: None of this is news. We'd already seen these same conclusions clearly illustrated in the IPASS trial and others. Never-smokers who don't have an EGFR mutation who receive don't EGFR TKI don't get anything close to the benefit seen in those seen in those with an EGFR mutation who receive an EGFR TKI. It's not the clinical variable, but rather the molecular marker that counts. When looking at the results in patients who received chemo/Tarceva, broken down by EGFR mutation status, we really just see the same result of far better PFS, OS and RR results for those with an EGFR mutation. The trial wasn't designed to compare the two treatment strategies, and the investigators didn't go out of their way to provide comparative results. Nevertheless, there are some conclusions that we can make. The RR of 33% in EGFR wild type patients is clearly better with chemo/Tarceva than the 9% noted above for similar patients who received Tarceva alone, and PFS was also clearly superior for EGFR wild type patients who received chemo/Tarceva (4.8 mo vs. 2.7 months). However, when we look at the more longitudinal question of OS, we see that the median OS for EGFR wild type patients who received chemo/Tarceva was 13.7 months, compared with 18.1 months for those EGFR wild type patients who started with Tarceva alone. It's hard for me to conclude that there's an advantage to combining EGFR TKI therapy with standard chemo for those patients without an EGFR mutation, even if it's clear that such patients are better served by starting with chemotherapy rather than deferring on chemo while pursuing a less effective EGFR-based initial treatment. Looking at the patients with an EGFR mutation, I would contend that the results are unclear. The difference in RR is pretty negligible: 67% with Tarceva, and 73% with chemo/Tarceva, and both results right in the typical range what you see in the various other trials of patients with an EGFR mutation who receive an EGFR TKI alone. Both median PFS and median OS are numerically a little higher with chemo/Tarceva (15.7 vs. 17.2; and 31.3 vs. 39.0), but these results are based on just 33 patients per arm, which makes a median a rather insensitive variable. But it's clear that there isn't a penalty in giving concurrent chemo/Tarceva here. My personal view is that these results wouldn't lead me to want to give chemo concurrent with Tarceva as first line therapy to people with advanced NSCLC and an identified EGFR mutation, but it also alleviates my previously expressed concerns that chemo and EGFR TKI therapy are potentially antagonistic, at least in those with an EGFR mutation. It makes me more inclined to consider continuing the EGFR inhibitor and adding chemotherapy for patients who show modest progression after a good initial response to an EGFR TKI, though this is certainly a setting in which we just don't have any clear answers. Having a couple of patients who have developed a nice response upon re-initiation of an EGFR TKI after a prolonged hiatus from EGFR TKI therapy after initial progression, the latter approach is definitely a viable alternative to just continuing the EGFR TKI indefinitely while adding and changing out several concurrent systemic therapies. On the other hand, these results corroborate that smoking status is a far inferior variable compared with EGFR mutation status for predicting benefit from an EGFR TKI. This study also reinforces that those with EGFR wild type are better served by initial chemotherapy and that there is nothing to recommend concurrent chemo and EGFR TKI therapy in those without an EGFR mutation. Lest anyone think that it's only my view that the results of Tarceva alone vs. chemo/Tarceva are overall comparable and with no clear finding to definitely recommend the combination, I'll close by saying that the CALGB is now planning their subsequent trial for first line treatment of patients with a prospectively identified EGFR mutation. It will be Tarceva with either concurrent Avastin (bevacizumab) or placebo, and it doesn't include chemo concurrently.