EGFR testing in lung cancer - which patients should receive kinase therapy?

Mutations in the epidermal growth factor receptor (EGFR) correlate with increased response in patients with non-small cell lung cancer (NSCLC) treated with EGFR tyrosine kinase inhibitors (TKIs). As a result, the blocking of EGFR to treat non-small cell lung cancer is becoming increasingly common, but there is considerable debate about when to initiate therapy and how best to select patients.

Recently, results from the iTarget trial were reported for gefitinib/Tarceva (OSI, Genentech/Roche) in the first-line treatment of 34 patients with mutations. The investigators, from the Massachusetts General Hospital Cancer Center in Boston, reported an overall response rate of 55%. This is similar to reports from other investigators, although no patient with atypical mutations achieved a response. As previously reported, mutations were seen more frequently in females, lifetime nonsmokers, and patients with adenocarcinoma. A median progression-free survival of 9.2 months and an overall survival of 17.5 months was also observed. Patients were treated with gefitinib 250mg per day until progression or unacceptable toxicity.

An editorial by Frances Shepherd, published in the Journal of Clinical Oncology, suggested that it is highly likely that a panel of tests will be used in the to determine which patients are likely or not likely to benefit most from therapy. Dr. Shepherd argued that the unselected non-small cell lung cancer population is definitely not the appropriate comparator.

There have now been several publications in which the survival rates of patients with EGFR mutations treated with chemotherapy have been reported and their survival has been significantly longer than that of patients with wild-type EGFR and the median survival has not been reached at 2 years. According to Shepherd, this tells us that the studies of first-line EGFR tyrosine kinase inhibitor therapy in mutation-positive patients likely have all been published prematurely, with median follow-up times less than half the expected survival time of patients treated with chemotherapy.

"With this in mind, does the 17.5-month median survival in the Sequist et al., study really compare favourably to historical controls?"

The problem is, we still don't know which patient sub-types would ideally benefit most and would, therefore, see an improved response rate. Further trials are likely needed before this crucial question can be adequately answered.