Gene expression profiling could improve risk assessment of breast cancer

Patterns of tumor gene expression can predict clinical outcomes in breast cancer patients with about 90% accuracy, according to a report published in the May 10th issue of The Lancet.

Such testing could lead to more appropriate therapy for some patients and, ultimately, it could obviate lymph node dissections.Dr. Joseph R. Nevins, from Duke University Medical Center in Durham, North Carolina, and colleagues analyzed DNA microarray data from 89 primary breast tumor samples to identify aggregate patterns of gene expression that were associated with clinical outcomes.The researchers identified certain expression patterns, known as metagenes, that were linked to lymph node disease and cancer recurrence. Moreover, such patterns were about 90% accurate in predicting outcomes in individual patients.”Involvement of lymph nodes is the most important predictor of breast cancer outcomes,” Dr. Nevins told Reuters Health. “One of the driving forces behind our research was to come up with [a test] that could possibly eliminate the need for lymph node dissection,” he added.”The accuracy we achieved is not high enough to justify the use of expression testing for clinical decisions,” Dr. Nevins noted. However, with additional studies involving larger patient numbers, “I think we can get the accuracy to a point where such testing will be usable,” he noted.Dr. Nevins added that gene expression profiling could ultimately lead to novel therapeutics. In the meantime, however, such testing could still promote better outcomes by improving risk group classification, and, thereby, increasing the chances that a particular patient receives appropriate therapy, he said.(Source: Lancet 2003;361:1590-1596: Reuters Health: Anthony J. Brown, MD: Oncolink)