We've seen clear evidence that patients who have tumors with certain mutations in the EGFR gene are highly likely to respond to oral EGFR inhibitors like tarceva (erlotinib) or iressa (gefitinib) -- with response rates that are in the 70% range and often last for many months or even a few years (see prior post). On the other hand, K-Ras mutations are associated with a very low probability of responding to EGFR mutations (see prior post).

But the overall favorable or unfavorable results of these mutations may not be limited to their associations with how patients respond to EGFR inhibitors and/or other systemic therapies. I've noted how patients with advanced NSCLC and EGFR mutations have a superior survival even if they don't receive an EGFR inhibitor. Another approach to assessing the prognostic value of mutations is to look at survival of patients with resected earlier stage NSCLC tumors.

Last year, a group from Memorial Sloan Kettering Cancer Center (MSKCC) in NYC, which has been at the forefront of studying the implications of molecular variables in lung cancer, published their analysis of EGFR and KRas mutations among patients who underwent surgery for stage I - III NSCLC (abstract here). Among 296 patients, EGFR mutations were detected in 40 (14%), while K-Ras mutations were detected in 50 (17%). The profiles of patients with these tumors were quite different, in that 48% of patients with EGFR mutations were never-smokers, and these patients also were quite likley to present with stage I disease (88%) and had a very favorable 3-year survival (90%); the patients with K-Ras mutations, on the other hand, were overwhelmingly current or former smokers (92%) who were more likely to present with higher stage disease at presentation and had a worse 3-year survival (66%):

(click to enlarge)

As shown above, the difference in survival was present for the entire resected NSCLC population or for stage I in particular.

Another similar analysis (abstract here) just came out of Japan, where investigators reviewed the tumors from 397 patients who had resections of lung adenocarcinomas over a 5 year period from 2000-2005; though most of these patients had stage I - III disease, a few (9) had resections even for stage IV NSCLC. They looked for mutations in EGFR, K-Ras, and p53 (another gene related to development of cancer) and assessed prognosis, though they removed the patients who were known to have received an EGFR inhibitor because they wanted to separate the overall effect of the tumor's biology from a particular interaction of these mutations with EGFR inhibitor therapy.

This population was very different from the one from MSKCC, with 48% of patients being never-smokers (seriously??), compared with 12% of the US population from MSKCC (though the Japanese study only included adenocarcinomas, which disproportionately include the majority of never-smokers). But the associations were the same, with patients with EGFR mutations (49%) showing a significantly superior survival compared with patients who didn't have an EGFR mutation, while the patients with K-Ras mutations (13%) vs. those without, and the patinets with the less well studied p53 mutations (38%) also had a

Taken together, these studies highlight that the biology of tumors with EGFR mutations or K-Ras mutations (and maybe also p53 mutations) have fundamentally different behavior and prognoses, independent of response to EGFR inhbitors. But these markers are also associated with different clinical variables, which also provide some prognostic information. We'll turn to the data about these clinical factors from both the MSKC and Japanese study of early stage patients next.