Rapid Capture Targeted Next Generation Sequencing (NGS) for Detection of Genomic Kinase- and Cytokine-Receptor Rearrangements in B-Precursor Acute Lymphoblastic Leukemia

Background: Next generation sequencing (NGS) applications have recently identified various recurrent kinase and cytokine receptor rearrangements in a subgroup of B-progenitor acute lymphoblastic leukemia (BPC-ALL) that is characterized by a Ph-like gene expression signature. Implementation of whole genome applications into diagnostics have so far been hampered by high costs and long turnover times. Therefore we developed a custom-made NGS based capture panel without the need for gene specific PCR amplification steps. With a rapid turnover time, pooling of multiple samples and standardized bioinformatics, this approach can be implemented in molecular diagnostics of BCP-ALL to identify patients who may benefit from an add-on therapy with specific tyrosine kinase inhibitors. Moreover, genomic breakpoints can be used as an additional patient specific target for minimal residual disease (MRD) monitoring. Here we report a feasible approach for the identification of kinase and kinase-dependent pathway aberrations in BPC-ALL utilizing a customized sequence enrichment capture panel for NGS.

Patients and Methods: Ninety three samples from BCP-ALL patients with detectable minimal residual disease (MRD) levels (≥ 5x10^-4) after induction therapy from COALL studies 07-03 or 08-09 were applied to the NGS custom panel. Genomic capture probes were specific for exonic and complete intronic regions of ABL1, JAK2, PDGFRB, CRLF2, EPOR, and for the IgH-JH region on chromosome 14. JAK1, JAK3, IKZF1 and SH2B3 were analysed for exonic mutations only. These predefined genomic regions from recurrently mutated or rearranged kinase and cytokine receptor genes were enriched from 50 ng of initial genomic DNA without gene specific PCR amplification steps, followed by paired end sequencing of the captured genomic fragments (Nextera Rapid Capture Custom enrichment protocol with paired end sequencing on a MiSeq benchtop sequencing platform; Illumina). Six positive controls with known rearrangements (BCR-ABL1 and ID4-JH) were included to verify this method.

Results: All breakpoints from 6 positive controls were successfully identified and verified by patient specific amplification of the genomic fusion sites. ABL1 and JH specific capture probes allowed the "fishing" of the genomic fusion partner gene, here BCR and ID4, by paired end sequencing of overlapping, breakpoint spanning reads. In addition we identified a genomic breakpoint for CRLF2, PDGFRB, EPOR, ABL1 and JAK2 in 19/93 BCP-ALL patient samples. One additional patient showed a SH2B3 5 bp insertion in exon 5 that leads to a premature stop codon. The 11 patients with either EBF1-PDGFRB (4), EPOR-JH (4), or CRLF2-JH (3) rearrangements showed the highest levels of MRD (≥ 1x10^-2) at the end of induction. MRD monitoring using the genomic breakpoint in 3 cases showed highly concordant data to our standardized IgH/TCR quantification assays. For the 3 patients tested so far, our breakpoint specific MRD assays clearly showed the association of these rearrangements with the dominant leukemic clone.

Discussion: In summary, we identified kinase- or cytokine receptor rearrangements in 20/93 patients analyzed with our NGS based custom enrichment panel. The majority of patients had high MRD level at EOI therapy and an age of onset ≥ 10 years. Rapid identification of kinase- or cytokine receptor rearrangements in BCP-ALL is necessary for timely therapeutic intervention with target specific tyrosine kinase inhibitors. Short turnaround times of less than 2 weeks and processing of multiple samples followed by a standardized bioinformatical workflow allows the identification of unknown fusions partners from frequently rearranged kinase- and cytokine receptor genes, providing the exact breakpoint location, irrespective of the gene expression signature. MRD guided NGS panel analysis is a feasible strategy for identification of patients with a high risk of therapeutic failure or relapse suitable for targeted TKI intervention.