CD34+ obtained from High Sokal Risk Chronic Myeloid Leukemia (CML) Patients (PTS) Expresses Gene Profiles (GEP) Significantly Different From CD34+ Obtained From Low Sokal Risk Patients.

Abstract 2174

Poster Board II-151

Chronic Myeloid Leukemia (CML) is a clonal myeloproliferative disease which typically presents in chronic phase (CP), whose malignant progenitor cells proliferate rapidly, still retaining their ability to differentiate. If left untreated the disease can rapidly progress to accelerated phase and blast crisis. Although the treatment has been dramatically improved with introduction of Glivec therapy, the use of the Sokal and the Euro prognostic scores has remained an essential clinical tool to stratify CML patients at diagnosis based on different evolutive risk and to guide treatment decisions. To further optimize the management of the disease it is critical to gain a better understanding of regulatory pathways involved in the intrinsic heterogeneity of CML and propensity to progress. To that end our effort has been focused on identifying a molecular signature associated with a risk of the disease progression.

Here we present data obtained from our study of gene expression profiles (GEP) aimed at identifying genes and pathways which could predict the disease course of CP-CML patients at the onset of the disease. The study was performed on highly enriched CD34+ cells from peripheral blood obtained from patients with untreated CML in CP. GEP was performed by using the Affymetrix HG-U133 Plus 2.0 platform. Raw data were normalized by using the RMA algorithm and filtered. Genes associated with Sokal risk score were selected by a moderate t-statistic (Limma package, p-value threshold = 0.01). Hierarchical clustering was performed with TIGR MeV. Overall, 34 pts were included in the present analysis. In the initial part of the study, the first 20 pts (the “training set”) were successfully assayed for global GEP and microarray data and were used to define a set of genes differentially expressed in high (H) (7 pts) vs. low (L) (13 pts) Sokal risk pts. We identified 84 probes sets and the clustering of their GEP showed an homogeneous pattern in H Sokal risk pts, where the most significantly involved process networks (as defined by GeneGo software) were: “Cell adhesion_Histamine H1 receptor signaling in the interruption of cell barrier integrity” (PLCB1, CALM1, PRKCA, PPPIR14A, MYL4), “Cytoskeleton remodeling_TGF and WNT” (ACTN1, CFL2, TCF7L2) and “Development_WNT signaling pathway” (WNT6, FZD3, TCF4, VEGF-A). Of interest, among the most significantly up-regulated genes, we identified PVT1, a non-coding gene located on chromosome 8, close to c-Myc gene, which encodes for several miRNA able to activate c-Myc gene transcription. Among the most significantly down-regulated genes is PLCB1, which we have recently described as being deleted in myelodysplastic syndromes and in acute myeloid leukemia. In the second part of the study, the 84 probes set were tested on an independent test set of 14 pts, including 4 H, and 10 L Sokal risk pts., thus showing that the GEP clustering displayed the same feature which we have observed in the training set. In conclusion, our study has identified a distinct array of genes at diagnosis which might be involved in driving the evolutive risk of CML and potentially, this approach could be of value to better define patients who may need an optimized treatment.

Supported by:Novartis Oncology, TOPS Correlative Studies, PRIN, AIRC, AIL, FIRB 2006, Fondazione del Monte di Bologna e Ravenna.