Abstract

Purpose: Precursor T-cell lymphoblastic leukemia (T-ALL) accounts for approximately 15% of pediatric ALL and represents a particular clinical challenge, because relapses are usually fatal. In T-ALL, monitoring treatment response (prednisone prephase; MRD kinetics) provides a powerful tool for risk assessment. We now aimed to identify DNA-based predictive biomarkers that allow risk assessment and treatment stratification at the time of diagnosis. We thus performed a genome-wide analysis for DNA copy-number aberrations to detect circumscribed gains and losses that are specific for molecular and prognostic subgroups.

Experimental design: Samples of 77 T-ALL patients enrolled in BFM-ALL protocols were analyzed by array-based comparative genomic hybridization (array-CGH) on an 8k BAC array allowing for an average resolution of 0.5 megabases. After using a Hidden Markov Model for bioinformatic analysis, our findings were correlated with minimal residual disease (MRD) data. Frequently affected pathways were identified using Babelomics software. This approach was complemented by genomic sequencing and the analysis of mRNA abundance for interesting candidate genes in a subset of leukemias.

Results: Several circumscribed genomic alterations affecting key regulators and downstream intermediates of TGF-β and PI3K-AKT signalling were identified thus suggesting an important role of these pathways in T-ALL leukemogenesis. Furthermore, deletions at chromosome band 6q15-16.1 haboring the caspase 8-associated protein 2 (CASP8AP2) gene were overrepresented in patients with unfavorable MRD. QRT-PCR revealed a marked gene-dosage effect resulting in reduced expression of CASP8AP2 mRNA in samples with 6q15-16.1 deletions (p=0.01). Fisher’s exact test revealed that deletions at 6q15-16.1 constitute a powerful marker for poor early treatment response (p=0.03), which is likely related to the reduced expression of the pro-apoptotic CASP8AP2 gene.

Conclusions: Array-CGH analyses indicate a novel role of TGF-β and PI3K-AKT signalling in T-ALL leukemogenesis. Moreover, deletions at 6q15-16.1 resulting in the reduced expression of the pro-apoptotic CASP8AP2 candidate gene identify a subgroup of T-ALL patients with poor treatment response. Prospective studies will now have to validate the usefulness of DNA-based molecular markers to clinically stratify the individual risk profile in pediatric T-ALL.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author