Dusting off genetic fingerprints

Gene expression patterning can predict the outcome of acute myeloid leukemia

By Brian Hoyle

Thirty-year-old Sandra Raymond was recently diagnosed with acute myeloid leukemia (AML). Unlike many AML-sufferers, Sandra's chromosomes appeared normal, making it more difficult to predict the outcome of the disease. Chromosome pattern is often used to determine patient prognosis in AML cases, but patients, like Sandra, whose patterning is normal, pose a big problem, since this can mask AML-related genetic abnormalities.

Sandra and patients like her may soon be able to get more accurate prognoses. Two articles in the April 15 New England Journal of Medicine report that genetic fingerprinting can more accurately pinpoint the nature and outcome of AML than simply looking at chromosome patterns and rearrangements.

The first study, led by Dr Jonathan Pollack of the Stanford University School of Medicine, used gene expression profiling to find different patterns of gene expression in adults with AML. The other study, authored by a team of Dutch researchers headed by Peter Valk, PhD, of the Erasmus University Medical Center in Rotterdam, used another version of the profiling technique to locate genetic hot-spots in AML.

Both studies aimed to pinpoint genes that are active in AML. The results of both teams "are surprisingly robust," write Drs Edison Liu and Krishna Karaturi of the Genome Institute of Singapore in an accompanying editorial.

Until now, the treatment AML patients receive has been guided by the microscopic detection of abnormalities like incomplete or missing chromosomal material. But, since some forms of AML don't result from gross chromosomal alterations, such an analysis can be useless. "Clinicians are not sure what to do with intermediate risk patients who have a normal [chromosome pattern]," Dr Pollack told Reuters Health.

Dr Pollack and his colleagues showed that there are differences in genetic activity among 116 AML patients. In particular, the 45 patients with normal chromosome arrangements could be separated into two groups based on which genes were turned on. The researchers hope that separate treatment strategies may be devised for the two 'different' diseases.

The Dutch researchers pinpointed different patterns of gene activity in 285 patients with AML. The patients could be grouped into 16 'clusters' of gene activity. Three of the clusters had a 57-72% survival rate after five years. A fourth cluster had a five-year survival rate of only 32%. Finally, the survival rate of a fifth patient cluster was a dismal 18%.

To Dr Anthony Komaroff, professor of medicine at Harvard Medical School, this new information means that "AML now joins a growing list of malignancies ... for which gene-expression profile fingerprints provide powerful prognostic information." And a better understanding of the significance of these genetic patterns could help determine the most effective therapy that will fit each individual with AML.