The identification of key genetic pathways and mutations in NSCLC will be critical to furthering our understanding of this malignancy and to developing new molecularly targeted therapeutics. The ability to sequence the entire cancer genome has shown how molecularly complex lung cancer is, and the challenge for now is to identify which mutations are drivers, which are prognostic, and which are predictive of treatment response.

KRAS: KRAS mutation in NSCLC is a modest prognostic marker for poorer outcome. However, its potential role as a predictive marker for resistance remains unclear with few studies showing statistically significant interaction. This is so for chemotherapy, EGFR TKIs and EGFR monoclonal antibodies. The role of KRAS mutation on different codons or with different amino acid substitutions is only now being explored in NSCLC. At this time, unlike colorectal cancer, KRAS mutational status should not be used to select patients' therapy.

EGFR: There is probably no more contentious or hotly debated topic than the question of how EGFR markers should be incorporated into clinical decision-making. EGFR gene copy, mutation and protein expression and proteomics have been explored and all, to some degree, have been shown to have some prognostic and predictive effects.

EML4-ALK: No other molecularly targeted drug has been approved for NSCLC as quickly as crizotinib, which was approved based only on phase II trials. Phase III post-commitment studies are ongoing comparing crizotinib to chemotherapy in EML4-ALK mutated NSCLC. As quickly as EML4-ALK was discovered, however, resistance mutations have been identified as well as a new sensitising rearrangement mutation in ROS-1.

Most mutations to date have been identified in adenocarcinoma. Mutations have now been reported in squamous NSCLC, but their prognostic and predictive powers remain to be clarified. This will be the goal of a SPECS grant recently awarded to a North American squamous cell lung cancer consortium.