Comparison of Fluorescence in Situ Hybridization (FISH) for EVI1 Rearrangement (3q26) and G-Banding in Patients with Acute Myeloid Leukemia with Inv(3) or t(3;3).

Acute myeloid leukemia (AML) with inv(3)(q21q26.2) or t(3;3)(q21;q26.2) is a distinct subgroup that may present de novo or arise form a prior MDS, characterized by thrombocytosis and marked dysmegakaryopoiesis. Since the presence of inv(3) or t(3;3) suggests poor prognosis in AML and needs for close monitor in MDS, detection of EVI 1 rearrangement (3q26) is crucial. We performed FISH for inv(3) or t(3;3) and compared those results with those of G-banding in AML, both at initial diagnosis and follow-up.

We retrospectively reviewed the results of G-banding in 550 patients with AML (2000 Jan-2012 June). Among them, patients who showed 3q abnormalities by G-banding were subjected to FISH analysis for inv(3) or t(3;3) with frozen bone marrow cells. In patients who showed EVI 1 rearrangement by FISH, we performed EVI 1 rearrangement FISH on serial follow-up bone marrow of those patients, for detection of minimal residual cells. We applied 3 kinds of FISH probes (A company, B company, C company) for the detection of EVI 1 rearrangement.

By G-banding, the frequency of 3q aberrations involving 3q26 was 1.5% (8/550 patients) in AML and the frequency of AML wit recurrent inv(3) or t(3;3) was 0.2% (1/550 patients) in AML. Among 17 consecutive specimen from 8 patients with 3q26 aberration by G-banding, 11 specimen showed EVI 1 rearrangement by FISH, while 8 specimen showed EVI 1 rearrangement by G-banding. Of note, among 8 patients with 3q26 aberration by G-banding, FISH for EVI 1 rearrangement revealed EVI 1 rearrangement in 5 patients. Furthermore, EVI 1 rearrangement (3q26) FISH detected EVI 1 rearrangement in 1 patients among 30 patients with 3q aberrations other than 3q26 by G-banding. Comparison of FISH results among 3 kinds of FISH probes showed marked discordances. Among 26 consecutive specimens in patients with 3q26 aberrations, probe of each company detected 11 (A company), 4 (B company) and 4 (C company), respectively.

Incidence of AML with inv(3) or t(3;3) involving EVI 1 rearrangement was rare (0.2%). One third of AML with inv(3) or t(3;3) by G-banding did not harbor EVI 1 rearrangement, suggesting involvement of other gene in the same chromosomal region. Comparison of FISH and G-banding during follow-up in the present study revealed that EVI 1 rearrangement FISH is essential for monitoring for minimal residual disease in AML with inv(3) or t(3;3).

No relevant conflicts of interest to declare.