Clinical screening of gene rearrangements in childhood leukemia by using a multiplex polymerase chain reaction-microarray approach.

PURPOSE Currently, many forms of leukemia are considered potentially curable, with prognosis and clinical outcome strongly dependent on the underlying molecular pathophysiology. A substantial number of leukemia patients harbor nonrandom karyotypic abnormalities that define subgroups with unique biological and clinical features. For detection of these types of gene rearrangements, a combination of multiplex RT-PCR with hybridization on oligonucleotide gel array was presented previously, which identified five chromosomal translocations with fusion variants. In the present study, additional clinically relevant translocations were included in our analysis using a second generation of microarrays. We also expanded significantly on the clinical correlation of our findings. EXPERIMENTAL DESIGN An oligonucleotide microarray was designed for hybridization with products of a multiplex RT-PCR to identify the following translocations: t(9;22)p190, t(4;11), t(12;21), t(1;19), typical for acute lymphoblastic leukemia; t(9;22)p210 for chronic myeloid leukemia; and t(8;21), t(15;17), inv16, typical for acute myeloblastic leukemia. RESULTS To demonstrate the potential clinical application of the method, 247 cases of childhood leukemia were screened, and the above-mentioned gene rearrangements were found in 30% of cases. The sensitivity and specificity of the assay is comparable with the RT-PCR technique, so that it can be used to follow minimal residual disease. The feasibility of an additional refinement of the method, on-chip-multiplex PCR, has been successfully demonstrated by identifying a common translocation, t(9;22), in chronic myeloid leukemia. CONCLUSIONS Our data suggest that the microarray-based assay can be an effective and reliable tool in the clinical screening of leukemia patients for the presence of specific gene rearrangements with important diagnostic and prognostic implications. The method is amenable for automation and high-throughput analysis.