I wrote in a post several months ago about the ongoing study of the monoclonal entibody against EGFR erbitux (cetuximab) in lung cancer, where it's role is still up in the air. Unlike the EGFR tyrosine kinase inhibitors (TKIs) iressa and tarceva, which showed no benefit when given concurrently with standard chemo, erbitux has a different mechanism and may still be useful when given along with chemo. As I mentioned in my last post, a phase III trial of carbo/taxane chemo with or without erbitux failed to show any improvement in progression-free survival, but it did actually show a higher response rate in the group that received chemo with erbitux. But perhaps there are particular patients who are likely to gain a lot more with erbitux than others, just as we've found that certain groups, such as those with EGFR mutations and never-smokers benefit most consistently from the EGFR TKIs.

I previously described the early results on SWOG Trial 0342 (abstract here), in which over 200 patients with previously untreated advanced NSCLC were randomized to two arms. The sequential treatment arm received 4 cycles of carbo/taxol chemotherapy followed immediately by weekly erbitux until progression of disease, while the concurrent arm received the same 4 cycles of chemo along with concurrent weekly erbitux, then followed by weekly erbitux alone. The schema is shown here:

(Click to enlarge)

The early results (abstract here) suggested that perhaps the concurrent arm did a little better, but neither arm did spectacularly, and it was worth debating whether the results were promising enough to commit to larger future studies. This year, Dr. Roy Herbst from MD Anderson Cancer Center presented more updated results (abstract here). With longer follow-up, the two arms converged together, both looking pretty good, with a median progression-free survival (PFS) of 4 months and overall survival (OS)of 11 months. You can see that the survival curves for the two groups are basically on top of each other:

Those results are perhaps a little better than the typical results of our larger trials with chemo like carbo/taxol alone, but we'd expect a median OS of 9 or 10 months, and it's not unusual to see results that are better in smaller phase II studies (which tend to be run in more specialized centers and often feature a little more cherry-picked patients -- which is technically called selection bias as a cause for more favorable results).

The most interesting part of this trial was the difference in results for patients who had tumors with EGFR gene amplification as measured by fluorescence in situ hybridization, or FISH. This basically measures whether there are extra copies of the gene that is the blueprint for the EGFR protein. The test was only able to be done for a limited subset of patients who had tissue available and sent in for analysis, but the results showed that the survival for the folks who had EGFR gene amplification by FISH had twice as long of a median PFS (6 vs. 3 months) and OS (16 vs. 8 months) of those who didn't have EGFR gene amplification:

I think that's very interesting, but it's just a single study, and this analysis includes tissue from just 44 patients. Importantly, though, the groups were pretty evenly split, with about 60% (26 patients) in the gene amplified group and 40% (18 patients) in the non-amplified group. Other limitations are that EGFR FISH testing is still just being explored and is not widely accepted by the entire lung community as being critical, that it requires tissue that is not always readily available (especially in patients with metastatic disease, who are often diagnosed with a very tiny amount of tissue like a fine needle aspiration or cells from a malignant pleural effusion), and also that EGFR FISH is a test that is only currently done at a small number of labs around the world.

But this angle provides the most interesting aspect of this study, and it gives a hint that maybe erbitux can be a valuable contribution to lung cancer if we can properly select the patients who will gain the most from it.