Cryptic NUP214-ABL1 fusion with complex karyotype, episomes and intra-tumor genetic heterogeneity in a T-cell lymphoblastic lymphoma

Introduction

Lymphoblastic lymphoma (LBL) is a rare and aggressive form of non-Hodgkin’s lymphoma (NHL). LBL develops from immature B cells committed to the B- (B-LBL) or T-cell lineage (T-LBL). LBL is morphologically indistinguishable from acute lymphoblastic leukemia (ALL) and 90% of it have a T-cell phenotype. LBL also accounts for approximately 2% of all NHL cases and occur in adult, children and adolescent, with a male predominance (three time more male are affected)^[1-2].

Chromosomal abnormalities in T-LBL are not well defined and cytogenetic data in T-LBL is limited. However, a few published cytogenetic studies revealed that typical chromosomal aberrations identified in T-cell ALL (T-ALL) are also present in T-LBL. These include translocations of T-cell receptor (TCR) gene to genes encoding transcription factors such as TAL1, TLX1, LMO2, and LYL1. In particular, the translocation t(9;17)(q34;q22~23) is typically found in T-LBL^[1-4]. However, no single recurrent and typical genetic alteration for T-LBL could be identified. This is in contrast to other malignancies like translocation of ALK gene in anaplastic large cell lymphoma, MYC gene in Burkitt lymphoma or BCL2 gene in follicular lymphoma.

Here we present the comprehensive analysis of a T-LBL case with a normal karyotype, according to standard G-banding with trypsin-Giemsa (GTG)-banding, using high resolution molecular methods, identifying also some intra-tumor genetic heterogeneity besides unusual acquired genetic alterations. Also here we report NUP214-ABL1 gene fusion in this patient, which appears cryptic due to its localization in episomes.

Case report

A seventeen-year-old female patient, who was initially diagnosed in South Africa with T-ALL, presented in the clinic in Poland with abdominal pain, accompanied by diarrhea and vomiting; she was here initially treated only symptomatically. A few days before, a blood test already revealed hyperleukocytosis (589 × 10⁹/l) with presence of 94% lymphoblasts in blood smear, hemoglobin 8.5 g/dl, and platelet count 53 × 10⁹/l. Bone marrow findings were: hypercellularity with 95% lymphoblasts, lack of megakaryocytes and Periodic-Acid-Schiff (PAS) staining identified in 70% of the blasts thick grains (data not shown). Ultrasound of abdomen showed enlargement of the spleen to 152 mm, and presence of fluid in the lower pelvis. Cervical lymph nodes were bilaterally enlarged with diameters of 3-4 cm, and small submandibular nodes were bilaterally enlarged to 2 cm in diameter.

Cytogenetic and immunophenotypic analyses were done. The latter characterized a T-LBL due to high expression of CD45 (100%), CD2 (96.6%), CD4 (97.3%), CD8 (90%), CD7 (77.1%), CD5 (76.0%), sCD3 (71.2%), CD1a (70.0%) and the lack of TdT, CD19, CD34 and CD38.

Banding cytogenetic analyses were done in unstimulated bone marrow cells according to standard procedures^[5] from the material taken at initial diagnoses. A total of 20 metaphases were available and analyzed on a banding resolution of 300 bands per haploid karyotype, revealing a normal female karyotype. Molecular diagnostic polymerase chain reaction (PCR)-based tests for presence of gene fusions BCR/ABL (p190 and p210), TCF3/PBX1, MLL/AF4 and SIL/TAL1 were negative (results not shown).

Also genomic DNA isolated from cells fixed in acetic acid-methanol (1:3) was subjected to array-comparative genomic hybridization (aCGH) as well as the multiplex ligation probe amplification (MLPA) studies, as previously reported^[6], Finally, fluorescence in situ hybridization (FISH) was done^[6-8], revealing a highly complex karyotype [Figure 1 and Table 1] with gene-amplification due to episomes (abbreviated here as epi), which can be reported as:

Figure 1. Result of multicolor banding (MCB) probesets for chromosomes 4, 7 and 10 are shown. MCB 10 showed the founder clone and subclone. Locus-specific probes (LSPs) for chromosomes 2, 7 and 10 characterized the breakpoints in 2q37.3, 7p34, 10q24.3 and 10q25.1 [Table 1]. The final karyotype after application of all approaches is summarized in the text. der = derivative chromosome

46,XX,der(2)t(2;7)(q37.3;q25.1),del(4)(p14p16),t(7;10)(q34;q24),del(9)(p21.3p21.3),epi(6;9)(q23.3;q34.12)x20~30[20%]/46,XX,der(2)t(2;7)(q37.3;q25.1),del(4)(p14p16),der(7)(7pter->7q34::10q24.1->10q25.1::2q37.3->2qter),del(9)(p21.3p21.3),der(10)t(10;7)(q23;q34),epi(6;9)(q23.3;q34.12)x20~30[40%]/46,XX[40%].

In the FISH-studies done here, between 15 and 25 metaphases were evaluated per applied probe-set, thus in the final karyotype overall percentages are given for the observed clones.

NUP214-ABL1 fusion could be deduced from aCGH data - the region being amplified ends on one side at NUP214- and on other side at ABL1-gene - as the amplified region is present as episomes, which are circular, there must be NUP214-ABL1 fusion.

The patient was treated according to the Polish Adult Leukemia Group (PALG) protocol, with induction therapy consisting of prednisone, daunorubicin, vincristine and PEG-L-asparaginase. No remission was achieved and the patient was re-treated according to fludarabine, cytarabine, and mitoxantrone (FLAM) with consolidation course (metrotrexate, cyclophosphamide and PEG-L-asparaginase) and maintenance treatment. After ten months, the patient relapsed and was now treated according to Hyper-CVAD protocol (cyclophosphamide, vincristine, doxorubicin/Adriamycin and dexamethasone). Still, one month later the patient unfortunately died.

Discussion

Recurrent acquired genetic lesions play a key role in predicting and assessing risks, so are the treatment protocols to be applied. Still, little is known about the copy number alterations (CNAs) accompanying structural abnormalities in T-LBL, such as the NUP214-ABL1 fusion gene. ABL1 fusion proteins are sensitive to tyrosine kinase inhibitors, which potentially can be included in future treatment strategy and NUP214 is a component of the nuclear pore complex, which mediates nucleocytoplasmic transport. NUP214 is widely expressed and is involved in the pathogenesis of acute myeloid leukaemia associated with the translocation t(6;9)(p23;q34) as DEK-NUP214 fusion^[9-11].

To the best of our knowledge, a cryptic NUP214-ABL1 fusion yet has only been identified in 6% of individuals with T-ALL and is the second most prevalent fusion gene involving ABL1^[12-15]. Here we report this for the first time in a T-LBL case, and even detected it as a high level amplification; most probably after inversion, duplication or translocation, gene fusion, circularization and amplification happened. As ABL1 is one of the best targetable tyrosine kinases, identification of ABL1 gene fusion is clinically important, as patients may potentially benefit from tyrosine kinase inhibitors^[16-17].

Episomes are submicroscopic, circular and large acentric DNA fragments that can replicate autonomously. One of the common formation-mechanisms for extrachromosomal elements in cancer cells is episome-replication and unequal segregation during cell division, resulting finally in an increase of copy numbers. Still they that are invisible in banding cytogenetics; this is because episomes are composed of only several hundred kilobases of amplified oncogenes and/or drug-resistance genes, and thus are too small to be visualized by light-microscopy^[18-20]. Of interest, we detected a variable number of episomes (20-30) in different cells. However, we suggest that c-MYB is also present on the same episomes [Figure 2]; due to lack of material we could not confirm this by FISH.

Figure 2. A: Array comparative genomic hybridization (aCGH) analysis of chromosome 6 revealed high level of 6q23.3 amplification containing MYB gene (arrow); B: MYB Dual Color Break Apart Probe was applied and showed high level of amplification more than 20 copies/per cell; C: aCGH analysis of chromosome 9 revealed biallelic deletion of CDKN2A at 9p21.3 and high level of 9q34 amplification contains ABL1 and NUP214 (arrow); D: FISH confirmed the homozygous deletion of CDKN2A in metaphase; E: BCR, ABL Dual Color Probe was applied and showed variable number of episomes (20-30) in spread metaphases. aCGH: array-comparative genomic hybridization; FISH: fluorescence in situ hybridization; wcp: whole chromosome paint

Additionally, recurrent acquired CNAs in different chromosomal regions were also identified besides unique ones for this case [Table 2]. Taken together, the genomic abnormalities in T-ALL and T-LBL are so similar that they could be considered as identical diseases in the future^{[1,4,12,14,15,21-24]}.

As shown in our case, NUP214-ABL1 is accompanied with loss of cyclin-dependent kinase inhibitor 2A (CDKN2A), which encodes the tumorsuppressors p16INK4A and p14ARF, and affects cell cycle progression. CDKN2A gene deletion can be detected at initial diagnosis or acquired at relapse, suggesting that CDKN2A gene deletion is a secondary genetic event and associated with chromosomal rearrangements. This may as a result lead to the aberrant expression of a diverse group of T-cell-specific transcription factors, which again can function as oncogenes, such as TLX1 and TLX3^[2,21,25]. The translocation t(7;10)(q34;q24), resulting from the TRB/TLX1 fusion gene, has been reported in several studies, and is present in 5% of pediatric and 30% of adult with T-cell ALL^[25-28].

Overall, in the present T-LBL case we identified substantial intra-tumor genetic heterogeneity and complexity. The founder clone has TRB/TLX1 fusion gene and the subclone has TRB/TLX1 fusion gene plus complex karyotype involving three-way translocation t(2;7;10)(2q37.3;7q34;10q25.1), further developing into a more complex subclone. Interestingly, the breakpoints at 2q37.3, 7q34, 10q24.3 and 10q25.1 were not previously reported in T-LBL^[29]. Thus, this data provides genetic support for a multi-step pathogenesis: deletion of a tumor-suppressor gene (CDKN2A), deregulated expression of a transcription factor TLX1 and most likely overexpression of a constitutively activated tyrosine kinase (NUP214-ABL1) and oncogene c-MYB due to episome amplification and the unique phenotypes of the T-LBL case mentioned above.

To conclude, this study demonstrates the power of high resolution molecular approaches. It may be considered that the use of such approaches is the most efficient and future standard method for screening ABL1 alteration. Particularly in T-LBL patients this may be advantageous, as ABL1 modulates T-cell development and plays a role in cytoskeletal remodeling processes in T-cells. Besides, the intra-tumor genetic heterogeneity in cancer has important implications for reservoirs of cells involved in progression of disease and drug resistance therapy. As NUP214-xABL1 fusion is sensitive to the tyrosine kinase inhibitor, this suggests that new therapeutic approaches in T-LBL may improve outcome and/or decrease treatment-related morbidity.