Hi again! You can think of this as a companion piece to my last post, examining some recent (but admittedly preliminary) evidence suggesting that Iressa (gefitinib) and Tarceva (erlotinib) may not be equivalent for patients with differing types of EGFR mutations. This is a slightly different topic, but one that has been quite contentious for several years: do KRAS mutations, found in 20% or more of patients with NSCLC, identify a group of patients who are resistant to EGFR inhibitors? KRAS is an oncogene that is mutated in a large percentage of pancreatic, colorectal, and non-small cell lung cancers. The significance of this gene has been known for a while, and anecdotal evidence has suggested that patients with tumors harboring KRAS mutations have a poor prognosis and tend not to respond to treatment (chemo or EGFR inhibitors) as well as KRAS non-mutated (wild type) tumors. Retrospective studies looking at KRAS mutant patients who are treated with EGFR inhibitors like Iressa and Tarceva have very consistently shown that the chance of significant tumor response (shrinkage) of tumors in this group is almost zero. However, we now know that there are 2 clear groups of patients who benefit from EGFR inhibitors: those who get a tumor “response”, and those who get tumor “stabilization”, in which the tumors simply do not grow for some period of time. Even disease stabilization can be associated with a prolongation of survival, and this phenomenon is commonly seen in patients with EGFR wild type tumors. EGFR wild type tumors also have a negligible rate of tumor shrinkage with EGFR inhibitors, much like the KRAS mutant tumors, but still seem to derive benefit. What has never been convincingly shown by the proponents of the “KRAS is bad” hypothesis is that patients with KRAS mutations (which are mutually exclusive of EGFR mutation) do not get the same level of disease stabilization and survival benefit from EGFR inhibitors as do all other EGFR wild type patients. Well, at the 2009 World Conference on Lung Cancer meeting we did get some indications that KRAS might not be so very bad after all. In the first example, let us return to the Jackman database that I described in my last post. Not all of the patients in the database had EGFR mutations, and many had KRAS mutations. Dr. Jackman showed that the time to progression and overall survival of patients with KRAS mutations, treated with EGFR inhibitors, had the exact some outcome as patients without KRAS mutations but also without EGFR mutations. Both groups did fairly poorly, but there was certainly no indication that KRAS mutation conferred a worse outcome that anyone else who did not have an EGFR mutation.

Secondly, there was Dr. Wolfram Brugger’s presentation of biomarker data from the SATURN trial. This has been discussed before, but briefly: the phase III SATURN trial randomized advanced NSCLC patients who had not progressed after 4 cycles of platinum-doublet chemotherapy to either maintenance Tarceva or the placebo. This trial showed an improvement in progression-free and overall survival for the maintenance arm (a topic for a post of its own), but Dr. Brugger’s presentation was focused on analyzing outcomes based on different biomarker subgroups (i.e. patients with EGFR mutations, KRAS mutations, etc.). Intriguingly, every single analyzed subgroup had a similar level of benefit from maintenance Tarceva, including smokers, non-smokers, squamous cell and adenocarcinoma histology, and (given the topic of this post) KRAS mutant versus wild type. Although Dr. Pasi Janne from the Dana Farber in Boston (an ardent proponent of KRAS resistance in EGFR therapy) correctly pointed out that the statistical analysis did not show a significant benefit to maintenance Tarceva in the KRAS mutation group, it was clear that the survival curves looked exactly the same in the KRAS mutant and wild type groups, and that the hazard ratio for benefit (a statistical measure of risk of death or progression in one group over the other) was identical in the two groups. The lack of statistical significance was probably (although of course not definitively) due to the very small number of KRAS mutant patients in the study. So what are we to conclude from this? I’m certain that this will not change the mind of the groups who truly believe that one should not ever ever give an EGFR inhibitor to a patient with a KRAS mutation, but I also think that the growing group of skeptics have new ammunition to say that perhaps Tarceva is about as (modestly) useful in treating patients with KRAS mutations as it is in any patient with EGFR wild type NSCLC, and I would certainly never withhold this medication from such a patient in the right situation.