Measuring gene expression in leukemic stem cells (LSCs) may have prognostic value for patients with acute myeloid leukemia (AML), researchers found.
Among patients free from chromosomal abnormalities, increasing LSC score -- indicating greater expression of 31 genes -- was associated with a higher mortality risk (HR 1.15, 95% CI 1.08 to 1.22), according to Ash Alizadeh, MD, PhD (pictured), of Stanford University in Palo Alto, Calif., and colleagues.
A high score was also associated with worse event-free and relapse-free survival.
The findings were confirmed in independent validation cohorts, and "support the clinical relevance of the cancer stem cell model for AML," in which specific subpopulations of cells -- LSCs in this case -- appear to be able to initiate and maintain tumors, the researchers wrote.
"Ultimately, this model has major implications for cancer therapy, most notably that in order to achieve cure, the cancer stem cells must be eliminated."
The prognostic model, however, still needs to be validated in a prospective study.
"If prospectively validated, the described LSC score may be incorporated into routine clinical practice for predicting prognosis in patients with AML, and used in clinical trials incorporating risk-based stratification or randomization strategies," Alizadeh and his colleagues wrote.
Using primary AML samples obtained from Stanford from April 2005 to July 2007, the researchers looked at gene expression profiles of LSC-enriched subpopulations using microarray analysis. They identified 52 genes whose expression distinguished LSC-enriched subpopulations from other subpopulations -- 31 were more highly expressed in LSC-enriched samples and 21 were more highly expressed in other samples.
The LSC score summarized the expression of the 31 genes more highly expressed in LSC-enriched samples.
In a training set of 163 patients with normal karyotypes, a high LSC score was associated with a greater overall mortality risk compared with a low score (HR 1.85, 95% CI 1.25 to 2.74). The absolute risk of death at 3 years was 57% for patients with a low LSC score and 78% for those with a high score. The finding was confirmed in three other independent patient cohorts.
A high LSC score was also associated with worse event-free survival in two cohorts (HRs 1.69 and 2.39) and worse relapse-free survival in one cohort (HR 1.78) (P<0.05 for all).
As an indication of initial therapeutic response, the rate of clinical remission was better for patients with low LSC scores in both an older cohort with a median age of 65 (56% versus 29%) and a younger cohort with a median age of 43 (88% versus 76%).
Even after adjusting for traditional risk factors -- age, mutations in FLT3 and NPM1, and cytogenetic abnormalities -- in a multivariate analysis, LSC score had prognostic value, with significant associations with mortality risk in the three validation cohorts (HRs of 1.07, 1.10, and 1.17, P<0.05).
"Future work is needed to prospectively validate the prognostic ability of the LSC score by evaluating its component genes using reverse transcription-polymerase chain reaction in an independent patient cohort," Alizadeh and his colleagues wrote.
"It will also be pertinent to examine the relationship of the LSC score to other gene expression signatures that have been proposed for predicting survival in patients with AML."
Illustration: Stanford University.
Abstract (The Journal of the American Medical Journal; 2010;304(24):2706-2715)