eP049: Discordant fluorescence in situ hybridization and RNASeq results in the identification of fusion partners in recurring translocations in hematological malignancies


      Reciprocal chromosomal translocations are implicated in neoplastic transformation in tumors. These can be identified by chromosome analysis, by fluorescence in situ hybridization (FISH) if cryptic, or by RNASeq/next generation sequencing. FISH probes are highly specific and sensitive, and have the advantage of faster turn-around-time, are less expensive, and can be used on a variety of specimen types. Therefore, several of these probes are commercially available, and are incorporated to screen diagnostic specimens. However, these probes may yield erroneous identification of fusion partners due to the proximity of genes covered by the FISH probe design. Here, we present two examples of this in hematological malignancies.


      Clinical material, bone marrow (BM) or peripheral blood (PB) was evaluated for hematopathology (morphology, immunohistochemistry, and flow cytometry), cytogenetics (karyotype and FISH) and RNASeq (next generation sequencing) methods using standard protocols.


      Patient one is a 62-year-old man was diagnosed with plasma cell myeloma (PCM) after he presented with anemia, hypercalcemia, lytic bone lesions, and 80-90% involvement by CD38+ plasma cells in bone marrow (BM). Chromosome analysis of the BM aspirate showed a highly complex karyotype with multiple related clones. FISH analysis with probes for FGFR3, MYC, CCND1, IGH, MAF and MAFB detected FGFR3 / IGH fusion. Therefore, a chromosomal diagnosis of FGFR3 rearranged PCM was rendered. In contrast, RNASeq data showed overexpression of NSD2 due to fusion with IGH. Patient two is a 15-year-old girl who presented with fever, elevated leukocyte count, B-lineage blasts in peripheral blood (PB) and in cerebrospinal fluid and was diagnosed with B-lymphoblastic leukemia (B-ALL). Chromosome analysis of a PB specimen at diagnosis showed an abnormal karyotype including a t(1;5)(q23;q33). A high risk ALL FISH panel of probes (Cep4/Cep10, ABL1/BCR, ETV6/RUNX1, ABL1, ABL2, PDGFRB, KMT2A) detected rearrangement in PDGFRB in 69.5% cells and loss of one signal for ETV6 in 19% cells. Therefore, a chromosomal diagnosis of PDGFRB rearranged B-ALL was rendered. Contrary to this, NGS analysis detected fusion between MEF2D (1q22) and CSF1R (5q33).


      FGFR3 and NSD2 located 6.25kb apart on 4p16.3 are involved exclusively in the t(4;14) in PCM. Since FGFR3 involvement is more frequent than NSD2, a commercial probe was designed to identify FGFR3 rearrangement, but design of this probe included NSD2. FGFR3/IGH fusion is routinely interpreted as FGFR3 positive PCM. Clinically patients positive for NSD2/IGH fusion have extremely poor prognosis compared to FGFR3/IGH positive patients. Moreover, the patients with t(4;14) with upregulation of FGFR3 may be treated with the novel pan-TKI, dovitinib, on a clinical trial. Patients with high-risk B-ALL are routinely screened for rearrangement in PDGFRB using a FISH probe, and cases positive for PDGFRB are rendered a cytogenetic diagnosis of PDGFRB rearranged B-ALL. Another infrequent genetic marker of extreme poor prognosis in Ph-like B-ALL is translocation of MEF2D (at 1q22) with several partners including CSF1R which is about 500pb apart from PDGFRB. Since PDGFRB and CSF1R are closely linked, PDGFRB FISH positive patients can be erroneously classified as PDGFRB positive B-ALL. Similar to FGFR3 positive PCM, some patients with PDGFRB rearranged B-ALL may benefit treatment with TKIs.
      These two cases highlight the necessity to exercise caution in the interpretation of FISH results, and the necessity for performing fusion screening using RNAseq or NGS methods for prognostication and therapeutic management of patients.