Two research teams have developed models for classifying the clinical outcomes of patients with non-small-cell lung cancer (NSCLC) using mass spectrometry techniques. Currently, clinicians do not have adequate methods for determining which patients will benefit from treatment with certain drugs. The new models could help physicians decide who will benefit from certain treatment options.
The studies are published in the June 6 Journal of the National Cancer Institute. In one study, an international team led by David Carbone, M.D., Ph.D., of the Vanderbilt-Ingram Cancer Center in Nashville, developed an algorithm to predict the outcomes of NSCLC patients treated with the drugs gefitinib and erlotinib, two tyrosine kinase inhibitors. The algorithm places patients into categories indicating "good" or "poor" survival before treatment with one of the drugs and is based on the pattern of a group of proteins in the patient's blood serum.
The authors developed the algorithm on a group of patients with known outcomes then tested it on pretreatment serum for independent validation and control groups. Patients who were predicted to have "good" outcomes survived for a median of 306 days, while those in the "poor" group survived a median of 107 days.
"In the clinical development of biomarkers for the individualization of therapy, it is important to distinguish between those who will benefit from an intervention from those independent of the planned intervention. Biomarkers predictive for survival benefit from an intervention are much more useful for guiding management"the authors said."Use of (Protein)signature will reduce rates of both overtreatment and undertreatment and improve survival for NSCLC patients," the authors added.
In the second study, Kiyoshi Yanagisawa, M.D., Ph.D., of Nagoya University in Japan, and colleagues analyzed protein patterns in NSCLC tumor tissue and normal lung tissue. The researchers identified a pattern that was associated with increased survival among NSCLC patients and may distinguish patients with poor prognosis from those with good prognosis. They also tested their model on independent validation and control groups.