More than a year ago, I wrote an introductory post about mutations in KRAS, one of the genes that contributes significantly to cancer cell growth and signalling, at least in many cancers. It's seen in around 20% of lung cancers, almost always in adenocarcinomas and not squamous NSCLC, and it's been implicated as being associated with a low likelihood of response to EGFR inhibitors.

I think KRAS is likely to be the focus of more discussion. One of the most important topics covered at ASCO 2008 was a clear demonstration that colon cancer patients with KRAS mutations (about 40-50% of that population) don't respond to intravenous EGFR monoclonal antibodies like erbitux (cetuximab) and vectabix (panitumumab) (plenary session abstract here). The evidence is conclusive enough that oncologists need to test for KRAS mutations in their colon cancer patients, because you run the risk of giving a therapy that isn't going to benefit them but has side effects. What does this mean in the lung cancer world? We don't know if KRAS mutations have the same meaning for lung cancer as for colon cancer. We really don't have any data yet on KRAS mutations with monoclonal antibodies against EGFR such as erbitux and vectabix (erbitux the more relevant one right now, with a positive phase III trial showing a small but statistically significant survival benefit for first line advanced NSCLC), but I hope and expect we will soon. With the oral EGFR inhibitors (tarceva and iressa), there is convincing evidence that responses are very rare among patients with KRAS mutations ("wild type" meaning normal, non-mutant form): That's shown in the middle of the table above. But I've highlighted in other posts that response rates don't equate completely with clinical benefit, and you can see on the right side of the table above that the difference in survival isn't that consistent. Also, you can see in the "waterfall plot" (in which bars going down from the horizontal line mean shrinking dimensions of the measurable cancer) from a Memorial Sloan Kettering-led trial of tarceva in BAC patients (abstract here) that several of the patients with KRAS mutations (blue bars in the figure below) who received tarceva achieved stable disease or even minor responses: So you can see that even if lung cancer patients with KRAS mutations almost never achieve a significant response, they can still have tumor shrinkage and may well achieve some improvement in survival with tarceva. I'd say, therefore, that it's an oversimplification and actually untrue that patients with KRAS mutations can't and don't benefit from oral EGFR inhibitors like tarceva and iressa. The association of KRAS mutations with benefit from erbitux and other EGFR monoclonal antibodies is still an open question. Finally, it's important to also point out that patients with KRAS mutations also tend to have a worse survival and get less benefit not only with EGFR inhibitors but also with standard chemo, so it's not clear that there's a better alternative treatment to suggest, except if we want to presume that patients would want to consider whether to accept treatment at all. I think most of my patients who aren't already potentially cured will want to try treatment, even if the likelihood of benefit isn't as high as we'd like to see it.