False positive NUP98 fluorescence in situ hybridization rearrangements in B-acute lymphoblastic leukemia

Gene fusions involving NUP98 have been reported in several hematologic malignancies yet have been very rarely reported in B-acute lymphoblastic leukemia (B-ALL). Two cases of B-ALL for which chromosome banding analysis (CBA) and fluorescence in situ hybridization (FISH) suggested apparent NUP98 rearrangements were further investigated with next-generation sequencing-based methodologies to verify the findings obtained with traditional cytogenetic methodologies. In the first case, CBA revealed a hyperdiploid karyotype with multiple structural abnormalities including additional material of unknown origin at 11p15; subsequent break-apart probe (BAP) FISH for NUP98 demonstrated 2 intact fusion signals and a single separate 5'NUP98 signal. However, whole-genome sequencing found no evidence of a NUP98 gene fusion. The results obtained with conventional cytogenetic methodologies were in fact attributable to structural variants (SV) with breakpoints not within NUP98 but within the 5'NUP98 BAP probe-binding sequence. In the second case, CBA revealed several structural and numeric abnormalities including a complex translocation between chromosomes 11 (at 11p15.4) and 19 (at 19p13.3) and an insertion of unknown material at 11p15.4. BAP FISH demonstrated a typical FISH signal pattern consistent with an apparent NUP98 rearrangement. However, no evidence of a NUP98 fusion was found on RNA sequencing. In conclusion, the two cases thus presented with clinical false positive NUP98 rearrangements by FISH. In the clinical laboratory, SVs in the vicinity of genes involved in recurrent rearrangements in hematologic malignancies may result in misleading results with conventional chromosome methodologies. This may preclude an accurate definition of the genetic attributes of malignancies with ensuing impacts on risk stratification and management. Higher-resolution testing methodologies such as whole-genome sequencing and RNA sequencing may be helpful in resolving unexpected results with conventional chromosome methodologies and enhancing the accuracy of genetic characterization of hematological malignancies in the clinical laboratory.