Three Novel AML Cases Harboring the Semi-Cryptic t(7;21)(p22;q22) Translocation Expressing RUNX1-USP42 Fusion Transcripts Associated with Diploidy/Tetraploidy and/or 5q Alterations: a Probably Underestimated Combination

Abstract 2708

RUNX1 is implicated in numerous chromosomal abnormalities acquired in acute myeloid leukemia (AML). The most frequent one, the t(8;21) is associated with a particular morphology together with a favorable prognosis. This is not the case for other 21q abnormalities, that are much less frequent and for which the prognosis is quite different. Moreover, beside point mutations, conventional cytogenetics failed to detect some of chromosomal alterations involving RUNX1.

Recently 3 cases of the rare and semi-cryptic t(7;21)(p22;q22) translocation expressing the RUNX1-USP42 fusion transcripts have been reported, demonstrating the recurrence of this abnormality in AML. We describe here 3 additional cases with the same translocation and fusion transcripts, associated to 5q alterations leading to EGR1 and CSF1R heterozygous losses.

In all our patients, the t(7;21)(p22.1;q22.3) was initially detected by the systematic FISH evaluation of the blastic populations using ETO-AML1 Dual Fusion probe. Patient#1 bone marrow karyotype was characterized by a tetraploid clone (89,XXYY) with loss of chromosomes 15, 17 and 18 in addition to the t(7;21), and a unbalanced translocation der(5)t(5;13)(q23;q?) between long arms of chromosomes 5 and 13, resulting in a heterozygous loss of EGR1 and CSF1R. Patient #2 blood and bone marrow karyotypes revealed a diploid clone with a del(5)(q31q33) associated with the t(7;21). The FISH analysis confirmed EGR1 and CSF1R deletions. In patient #3, the bone marrow karyotype showed diploid/tetraploïd clones, both harboring the t(7;21)(p22;q22), confirmed by FISH experiments (WCP7, AML1 probes). In addition, a der(5)t(1;5)(q3?2;q21-23) was identified within the tetraploïd clone, resulting in the loss of EGR1 and CSF1R, confirmed by FISH.

In all three cases a RUNX1-USP42 fusion transcript was detected using RT-PCR, as well as the reciprocal transcript. Sequence analysis of RT-PCR products showed that the breakpoints occurred exactly in the same introns of USP42 and RUNX1 as in the previously described cases. For patient #1 and #3 a chimeric transcript was found formed of the RUNX1 exon 7 fused to the USP42 exon 3. In patient #2, a shorter chimeric transcript arised from the fusion of the RUNX1 exon 5 to the exon 3 of USP42. As already noticed in the previous reports, an alternative splicing of the RUNX1 exon 6 has been detected in these three cases.

The description of these 3 novel t(7;21) confirm the recurrence of this balanced translocation in AML, and shows that this chromosomal abnormality is often associated with diploid/tetraploid clones and/or 5q alterations. Special attention should be paid in karyotype analysis of AML with diploid or tetraploid clones harboring 5q alterations. In such cases RUNX1 rearrangements should be explored using FISH analysis, and RUNX1-USP42 fusion transcript should be searched by RT-PCR in positive cases. Prospective and retrospective studies of AML have now to be settled in order to assess the incidence and clinical relevance of this cryptic translocation.