Gene expression profiling in acute myeloid leukemia

Publisher Website
Google Scholar
Abstract
This review deals with the emerging promises of gene expression profiling (GEP) and the currently accumulating knowledge about the classification and the discovery of novel disease entities in clinical acute myeloid leukemia (AML). Gene expression profiling studies in AML have shown that known and novel classes of disease can be recognized by unsupervised analyses. Prognostically informative molecular signatures can be deduced. Supervised analyses show that particular clinically relevant subsets of AML can be predicted with high accuracy with minimal sets of genes. The AML GEP studies published to date show a remarkable level of concordance in findings, especially for similar GEP platforms. This confirms the robustness of the methodology and the promise for future applicability of GEP in clinical diagnostics. For the time being, certain technical hurdles remain to be overcome. These relate, for instance, to the conversion of data between different GEP platforms, the effect of differences between various statistical clustering methods, and the still incomplete understanding of the effect of biologic (eg, morphology) and genetic factors on the expression signature. GEP analyses, perhaps in combination with high-throughput mutation analysis and proteomic approaches, may ultimately result in the establishment of a comprehensive diagnostic approach that will yield a key to the precise pathobiologic nature of AML.