Background: Leukemia is the most common pediatric malignancy, accounting for nearly 40% of all new childhood cancer. The cure rates for pediatric ALL have increased to more than 80% and the cure rates for pediatric AML now approach 50%. Much of this progress can be attributed to cytogenetic and molecular risk stratification with subsequent randomized control trials. Genomic instability events may also serve as relevant prognostic biomarkers. A novel high-throughput genomic technology called Molecular Inversion Probes (MIPs) quantifies genomic instability, gene copy number and allelic imbalances at the highest genomic resolution. MIPs can analyze genetic target sequences in parallel with high specificity and sensitivity. Further classifying leukemic blasts using more precise molecular techniques could prove useful in further risk stratifying and developing new treatment strategies.Objective: To adapt MIPs to characterize and define unique molecular subtypes of childhood leukemia.Methods: DNA was extracted from AML and ALL clinical samples obtained at diagnosis. 400 ng of leukemic DNA was mixed with a previously synthesized MIP cancer probe set (667 probes representing all human chromosomes from exon sequences of 298 cancer genes); each probe contains two unique recognition sequences to targeted genomic DNA and a unique barcode (tag) sequence. A thermostable polymerase and ligase were added, and the mixture was denatured. Probe homology sequences were then annealed to complementary genome sites. Standard PCR reagants were added including one fluorescent-labeled primer and unlabeled primer. The reaction was thermocycled and probes circularized in the nucleotide specific “gap fill” reaction were amplified. The PCR mixture was hybridized to a barcode microarray overnight. The chip was stained and washed, and then interrogated by an Affymetrix scanner via argon laser excitation at 488-nm. ‘Cluster along Chromosomes' analysis was performed to detect significant amplifications or deletions.Results: Childhood ALL and AML samples had unique patterns of multiple gene deletions. Additionally, overlap was found in several of these deleted genes. Significantly deleted genes in common between ALL and AML samples included: RAP1GA1, HYAL1, HSPA9B, CSF1R, TNF, NOTCH4, FANCE, CCND3, BAG2, PLAGL1, MAD1L1, POLM, POLD2, IGFBP3, PSD, SUFU, CYP17A1, MXI1, HRAS, IGF2, RRM1, BRCA2, TP53, ERBB2, ERCC2, and POLD1.Conclusions: MIPs represent a novel technology for highly sensitive and specific gene copy analysis of childhood leukemia. Unique and distinguishing signatures of allelic imbalance can be determined between ALL and AML clinical samples using MIP technology. The unexpected overlap of deleted genes in both ALL and AML may represent a common molecular mechanism that requires further investigation. We currently are adapting MIPs for an expanded cancer probe set and are exploring new techniques to determine loss of heterozygosity (LOH) in pediatric leukemia.