Re-analysis of data finds flaws in protein-based screen
TUESDAY, Feb. 15 (HealthDay News) -- For years, researchers have looked for an accurate means of detecting deadly ovarian cancer early on, and a 2002 report raised hopes that a blood test based on protein-based "proteomics" technology might work.
But now, researchers in Texas say that claim is simply not biologically plausible.
The original study, published three years ago in the British journal The Lancet, reported dramatic results using technology called mass spectrometry to search for a pattern of proteins in blood that might point to ovarian cancer. The study reported that this method correctly identified all of the people in the study who actually had ovarian cancer, while only misdiagnosing three healthy people.
However, researchers at M.D. Anderson Cancer Center, Houston, re-analyzed data from that study and now conclude the test is not biologically plausible. They say the problem stems from the way the mass spectrometry data was analyzed by the researchers.
The new findings appear in the Feb. 16 issue of the Journal of the National Cancer Institute.
"We view this as a cautionary tale. If you are not careful with this new technology, whose quirks we don't fully understand, you can find results that may be due to something other than biology," study lead author Keith Baggerly, associate professor in the department of biostatistics and applied mathematics, said in a prepared statement.
He said his team's finding "illustrates the need for researchers to set standards by which to conduct proteomics research." In other words, all laboratories doing proteomics research should use common protocols so that results from one lab can be verified by other labs.
"We are moving in that direction. The technology being used to develop a variety of proteomic diagnostic tests is getting better and we are getting more reproducible results," Baggerly said.
SOURCE: University of Texas, news release, Feb. 15, 2005
Copyright © 2005 ScoutNews LLC. All rights reserved.
