Cryptic insertion of CBFB into MYH11 leading to a type D fusion in acute myeloid leukemia with normal karyotype

概要

Acute myeloid leukemia (AML) with inv(16)(p13q22)/t(16;16)(p13;q22) has distinct clinicopathologic features, including myelomonocytic leukemic cells, abnormal bone marrow eosinophils, and a relatively favorable prognosis. 1,2 Most cases are classified as AML M4 with abnormal eosinophils (M4Eo) in the French–American–British (FAB) classification. The inv(16)/t(16;16) leads to a fusion gene between the core-binding factor beta subunit (CBFB) at 16q22 and myosin heavy chain 11 (MYH11) at 16p13. Over 10 types (types A–K) of CBFB/MYH11 fusion transcripts have been characterized.3 Furthermore, in addition to inversions and translocations, a few cases of insertions leading to the CBFB/MYH11 fusion gene, namely, insertion of CBFB into MYH11 or insertion of MYH11 into CBFB, have been reported.4-10 We describe here a rare case of AML with a normal karyotype and type D CBFB/MYH11 fusion transcript. Fluorescence in situ hybridization (FISH) revealed an atypical signal pattern by cryptic insertion.

A 53-year-old man was admitted due to fever and leukocytosis. He had no history of chemotherapy or radiotherapy for malignancies. Peripheral blood showed hemoglobin 52 g/L, platelets 21×109/L, and white blood cells 41.0×109/L, with 5% neutrophils, 16% lymphocytes, 34% monocytes, and 45% blasts. Bone marrow was hypercellular, with 14.8% monocytes, 5.6% eosinophils, and 74.8% blasts staining positive for myeloperoxidase (Figure 1A, 1B). Flow cytometric analysis revealed blasts positive for CD13 (77.0%), CD33 (81.5%), CD34 (29.7%), and HLA-DR (85.8%).

G-banding analysis of bone marrow cells showed 46,XY[20] (Figure 2A). Reverse transcription polymerase chain reaction (RT–PCR) screening detected a CBFB/MYH11 fusion transcript, whereas other fusions, including BCR/ABL1, RUNX1/RUNX1T1, PML/RARA, and DEK/NUP214, were negative. Furthermore, NPM1, FLT3-D835, and KIT (exons 7 to 17) mutations, and FLT3-internal tandem duplication were also negative. To perform RT–PCR for CBFB/MYH11, we created a forward primer, CBFBF (CBFB exon 4, 5’-CTCCAAAGACTGGATGGTATGGGC-3’, cDNA 605–628 in NCBI NM_001755.3), and reverse primer, MYHR (MYH11 exon 34, 5’-CTTGGACTTCTCCAGCTCATGG-3’, cDNA 4707–4728 in NM_001040113.2). One PCR band of 894 bp was generated in the bone marrow cells of the patient (Figure 1C) and was larger than the type A fusion transcript that is most commonly found. Nucleotide sequencing revealed CBFB exon 5 fused with MYH11 exon 30 indicating a type D fusion transcript since sequences were identical to those of GenBank accession number AF249897.1 (Figure 1D). Namely, rearrangements of CBFB and MYH11 occurred although G-banding analysis showed a normal karyotype but not inv(16)(p13q22)/t(16;16)(p13;q22).

The inv(16)/t(16;16) is a subtle chromosomal rearrangement that may be overlooked by G-banding analysis when metaphase preparations are not optimal. 2 Thus, to confirm the mechanism of the rearrangement, we carried out FISH using a Vysis LSI CBFB Break Apart Rearrangement Probe (Abbott Molecular, Abbott Park, IL) on interphase nuclei. This is the most commonly used probe for detecting inv(16)(p13q22). 8 The expected pattern in a nucleus containing inv(16)(p13q22) is separation into red (5’ CBFB) and green (3’ CBFB) signals besides one fused red/green (yellow, 5’ and 3’ CBFB) signal. However, FISH detected an atypical split signal pattern in 83 of 100 interphase cells: two fused yellow signals and one small red signal (Figure 2B). We next performed FISH using an LSI CBFB probe on metaphase spreads. A small red signal was located at 16p13 in addition to two fused red/green (yellow) signals at 16q22 in 18 of 20 metaphase spreads (Figure 2C) indicating only a part of the 5’ CBFB probe moved to 16p13. That is, this suggests the 16q22 fragment, including CBFB exons 1 to 5, was inserted into the MYH11 intron 29, resulting in the generation of a type D CBFB/MYH11 fusion gene at 16p13 (Figure 2D). Accordingly, the karyotype was finally interpreted as ins(16)(p13q22q22).

With respect to these laboratory findings, we diagnosed the patient as having AML M4Eo and AML with CBFB-MYH11 in FAB and World Health Organization classifications, respectively. 2 The patient received induction therapy, including daunorubicin and cytarabine, followed by three courses of high-dose cytarabine and remained in complete remission (CR). However, 16 months after the initial diagnosis, a bone marrow aspiration detected 23.4% blasts. G-banding showed 46,XY[20]. Subsequent RT–PCR showed an increased level of CBFB/MYH11 fusion transcripts while FISH revealed similar signal patterns to those at initial diagnosis in 18 of 20 metaphase spreads. Consequently, the patient underwent re-induction therapy, three courses of consolidation therapy, and an unrelated allogeneic peripheral blood stem cell transplantation. The patient has been in a second CR for more than 12 months.

An insertion, cryptic or not, is an alternative mechanism of chromosomal rearrangement producing a CBFB/MYH11 fusion gene.5 To our knowledge, only two other cases of cryptic insertion of CBFB into MYH11 have been reported to date (Table 1). 4,5 The diagnosis for both cases was AML M4Eo. The first case presented with a normal karyotype and type D fusion transcript. 4 FISH using an ON CBFB t(16;16);inv(16) break-apart probe (Kreatech Diagnostics, Amsterdam, The Netherlands) displayed a normal pattern, but FISH with a “homemade” CBFB/16q22 BAC clone on interphase cells yielded two fused red/green signals and an extra small red signal. In metaphase spreads, the extra red signal (3’ CBFB) was evident at 16p13, indicating a part of CBFB was inserted into MYH11. An LSI CBFB probe was not used although the expectation was that a clear insertion into a 16p13 signal was produced by this probe. The second case presented with type A fusion and a normal karyotype. 5 FISH using a CBFB/MYH11 Translocation Dual Fusion Probe (Cytocell, Cambridge, UK) showed normal signals, whereas FISH with BAC clones showed co-localization of CBFB and MYH11 signals at 16p13, indicating a possible insertion of CBFB into MYH11. An LSI CBFB probe was not used, but it was suspected this insertion could be missed by this probe due to the level of resolution of FISH. In contrast to these two cases, we demonstrated that cryptic insertion of CBFB could be detected by FISH using a widely used, commercially available LSI CBFB probe. All three cases showed normal karyotypes, suggesting the rearrangement, ins(16)(p13q22q22), cannot be identified by G-banding analysis.

In comparison, five reported cases exist of AML with a cryptic insertion of MYH11 into CBFB. 5-10 In addition to two cases identified by conventional cytogenetics as ins(16)(q22p13p13), FISH detected the localization of MYH11 on 16q in all cases. FISH with an LSI CBFB probe was performed in three cases. 8-10 In one case with ins(16)(q22p13p13), 5’ and 3’ CBFB signals were separated probably by an insertion. 9 As suspected from Figure 2D, 5’ CBFB and 3’ CBFB signals become separated on ins(16)(q22p13p13) if a large 16p13 fragment containing MYH11, which is visible by G-banding, is inserted between CBFB exons 5 and 6. However, FISH with an LSI CBFB probe yielded normal results in two cases with normal karyotypes. 8,10 Namely, the cryptic insertion of MYH11 into CBFB might not be detected by an LSI CBFB probe. These findings indicate that knowing the exact location of FISH probes is very important for detecting and interpreting cryptic insertions of CBFB or MYH11.

Interestingly, two of the three cases with an insertion of CBFB into MYH11 displayed type D fusions, whereas all four cases examined with an insertion of MYH11 into CBFB showed type A fusions. In AML with inv(16)/t(16;16), more than 85% of fusions are type A, and two fusions (types D and E) are approximately 5% each. 1,3 It has been shown that non-type A fusions were correlated with distinct clinical and genetic findings, such as an absence of KIT mutations and a distinctive gene expression profile. 3 The present case with a type D fusion also did not show a KIT mutation. Considering the low frequency of a type D fusion in all AML with inv(16)/t(16;16), the cryptic insertion of CBFB into MYH11 might be associated with a type D fusion.

Patients with AML harboring CBFB/MYH11 achieve a longer CR once they are treated with high-dose cytarabine as consolidation therapy, but those with KIT mutations have a higher risk of recurrence and worse survival.2 KIT mutations, not the type of fusion transcript, could affect clinical outcome. 3 That is, overall survival was similar between non-type A patients and type A patients without KIT mutation. Our patient was treated with high-dose cytarabine and showed a first CR of relatively short duration although the KIT mutation was negative. It might be that a cryptic insertion was associated with early relapse. In any case, it is critical not to miss a CBFB/MYH11 fusion since this is a prognostic marker. This case highlighted the importance of FISH with an LSI CBFB probe and screening by RT–PCR to detect CBFB/MYH11 in AML with a normal karyotype.

この論文で使われている画像