Abstract
Childhood B-cell precursor acute lymphoblastic leukemia (BCP-ALL) is a heterogeneous disease in which the 5-years event-free survival rates are currently above 80%. We recently identified a poor prognostic set of patients within unclassified BCP-ALL patients. These patients have a high risk of relapse and a gene expression profile similar to BCR-ABL1-positive ALL cases. This so-called BCR-ABL1-like (or Ph-like) group of patients showed similar high frequencies of deletion in B-cell development genes (den Boer et al, Lancet Oncology, 2009 and Mullighan et al. New Eng J Med, 2009). Here, we investigate the miRNA expression pattern in BCR-ABL1-like ALL. Further, we addressed whether altered expression levels of discriminative miRNAs of BCR-ABL1-like ALL has prognostic importance. MiRNA expression levels of leukemic cells of 91 newly identified children with BCP-ALL: i.e. 15 BCR-ABL1-like, 14 BCR-ABL1-positive, 9 TCF3-rearranged, 14 hyperdiploid (>50 chromosomes), and 15 B-other patients (negative for BCR-ABL1-like ALL and other known genetic lesions) were measured by Taqman® Array Human miRNA cards (TLDA cards, Applied Biosystems). Using the R statistics program (R-2.15, release June 2012), the Limma package was applied to compare the expression levels of miRNAs between the groups. Each genetic subtype was compared to the remaining cases to identify subtype-specific miRNAs.
Our data revealed that children with ETV6-RUNX1-positive, hyperdiploid, TCF3-rearranged and MLL-rearranged ALL demonstrate a subtype-specific miRNA signature. In contrast, BCR-ABL1-positive and BCR-ABL1-like cases showed a more variable miRNA expression pattern, resulting in 2 clusters of patients. The majority of the children with BCR-ABL1-like ALL (11 out of 15) showed a miRNA expression pattern different from that of BCR-ABL1-negative genetic subtypes of pediatric BCP-ALL and cluster with 43% of BCR-ABL1-positive patients (6 out of 14, cluster-I). The remaining 8 BCR-ABL1-positive and 4 BCR-ABL1-like cases showed more heterogeneous expression patterns (cluster-II). The prognosis of patients in both clusters, however, did not differ. Similarly, there was no significant difference in IKZF1, PAX5, JAK2 or CDKN2A/B status between these two clusters. Top-10 most differentially expressed miRNAs of the children with BCR-ABL1-like ALL were miR-324-5p, miR-345, miR-190, miR-130a, miR-545, miR-152, miR-103, miR-191, miR-197 and miR-101. None of these discriminative miRNAs was predictive for clinical outcome of BCR-ABL1-like patients. In conclusion, our data suggest that a miRNA signature is not suitable to dissect good and poor prognostic BCR-ABL1-like cases.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.