Three new studies highlight advancements being made toward individualized molecular classification of specific cancers and the potential for more targeted therapy. Two studies released in advance online publications of Science and the New England Journal of Medicine (NEJM) identified mutations of the epidermal growth factor receptor (EGFR) gene in certain non-small-cell lung cancer (NSCLC) patients that render their tumors sensitive to the drug gefitinib (Iressa). In the third study, published in the April 29 NEJM, investigators reported that they had identified a six-gene "signature" using microarray analysis that can be used to predict the response of diffuse large-B-cell lymphoma patients to standard chemotherapy.

Gefitinib was approved by the U.S. Food and Drug Administration last May as a third-line therapy for NSCLC, which accounts for 85 percent of lung cancer cases. However, previous clinical trials testing gefitinib have shown significant variability in response rates. For example, 10 percent of patients responded to gefitinib in a clinical trial with mostly patients of European ancestry, whereas a 27.5 percent response rate was demonstrated in a clinical trial with solely Japanese patients.

Though a minority of patients responds to treatment with gefitinib, confirmed Dr. Frederic J. Kaye of the NCI Clinical Genetics Branch, a coauthor of the Science report, these new findings will allow physicians to identify the subset of patients who are most likely to benefit from the drug.

The EGFR protein, the cellular target of gefitinib, is a member of the tyrosine kinase family of proteins - a class of enzymes involved in cellular signaling that is commonly mutated in a number of different cancers. These mutations are believed to disrupt EGFR's ability to regulate itself, thus triggering its signaling pathway to induce uncontrolled cell growth. However, these mutations also render "the mutant EGFR proteins much more sensitive to gefitinib," added Dr. Kaye.

In both the Science and NEJM studies, the presence of EGFR mutations correlated with clinical response to gefitinib. The mutations were more prevalent in Japanese patients than in Caucasian patients. Tumors from 25 new Caucasian patients were also examined: 14 that responded to gefitinib and 11 whose cancer progressed during treatment. Thirteen of the 14 tumors from responders were found to have EGFR mutations, whereas no EGFR mutations were detected in the 11 tumors from patients that progressed.

These findings should improve the outcome for many patients with NSCLC and may lead to the development of other molecularly targeted drugs to treat patients with other cancers that have known EGFR mutations, such as glioblastoma. "This data powerfully validates and reenergizes the approach of targeted therapy and should impact the design of future clinical trials," said Dr. Kaye.

In the NEJM study on diffuse large-B-cell lymphoma, instead of identifying genetic mutations that predict treatment response, researchers were able to identify six genes that, when present in patients, predicted their response to standard chemotherapy.

The goal of the study, explained authors Dr. Izidore S. Lossos and colleagues from Stanford University, the University of Miami, and Applied Biosystems, was to validate previous findings that identified genes and gene signatures for predicting treatment response and to formulate these into "a model that was technically simple and applicable for routine clinical use." Their model uses the molecular characteristics of lymphoma to divide patients into three pools of risk that reflect overall patient survival and can help determine the most appropriate treatment for individual patients.

"Lossos and coworkers have passed a milestone in the development of clinical diagnostic tests for cancer," wrote Dr. Sridhar Ramaswamy in an NEJM editorial. Their findings translate diagnostics, he continued, "from unbiased, genome scale surveys of gene expression in human tumors to the creation and initial validation of a novel diagnostic tool that should fit easily into clinical practice and might refine the currently available measures used for risk stratification."