Frequency of ETV6/AML1 fusion in adult acute lymphoblastic leukemia.

In a paper published in Leukemia in 2002, Jabbar Al-Obaidi and colleagues on behalf of the MRC Adult Leukaemia Working Party describe interesting findings on the frequency of ETV6/AML1 fusion as detected by fluorescence in situ hybridization (FISH) in 159 adults with acute lymphoblastic leukemia (ALL).1 In this large series, seven patients aged 15–50 years were shown to carry an ETV6/AML1 signal co-localization using commercial probes, giving a 4.4% incidence for this lesion, the highest so far reported in adult ALL patients.2,3 Interestingly, one of the two patients tested by RT-PCR did not show the expected ETV6/AML1 fusion product, raising the possibility that FISH may reveal cases with alternative breakpoints which are missed by PCR.1 We performed a centralized cytogenetic and molecular genetic analysis in 162 consecutive patients with ALL (71.6% B-ALL; 28.4% T-ALL) in the 14–85 years age range (median age 34), 135 of whom were enrolled in the 0496 GIMEMA trial. Fresh bone marrow (BM) samples were shipped by overnight courier and processed by established short-term culture methods for karyotyping. In 135 cases part of the cells were also submitted to standard procedures for the detection of the following aberrations by Southern blotting and by RT-PCR methods:4 MLL rearrangement, BCR/ABL fusion, E2A/PBX1 fusion, ETV6/AML1 fusion. All cases were also studied by FISH using the Vysis dual-color probe combination (Vysis, Downer's Grove, IL, USA) for the detection of the t(12;21) and of the corresponding ETV6/AML1 fusion. The expected normal signal configuration consisting of two separate red and two green signals was observed in >95% of nuclei in five normal controls, indicating a high hybridization efficiency with a <1.5% false red/green fusion rate. In this series no case with evidence of t(12;21) or ETV6/AML1 fusion by FISH analysis was found (at least 200 cells observed in >90% of the cases; >100 cells in the remaining cases); likewise, no patient carrying a hybrid ETV6/AML1 transcript was detected by RT-PCR. Thus, our findings do not support the suggestion that FISH may be more sensitive than molecular methods in the detection of this aberration, in line with previous observations on a limited number of cases which have shown concordant results by FISH and molecular genetics.2,5,6 The difference in the incidence of this genetic lesion between our series and that of Jabbar Al-Obaidi et al is statistically significant (P = 0.009). Since these experiments were performed by experienced personnel in referral laboratories, technical errors which may account for the different incidence of this aberration in the Italian and in the UK series are highly unlikely; we suggest that geographic heterogeneity in the genotype of adult ALL may be a reasonable explanation for this observation. Interestingly, the incidence of ETV6/AML1 fusion in adult ALL studied in France, Spain and Italy appears to fall in the 0–1% range,2,6 whereas a 3% incidence has been reported in 1996 by Aguiar and colleagues who analyzed 101 ALL in the UK.3 The incidence of ETV6-AML1 fusion in 334 unselected Italian and German children with ALL was 18.9%,7 as compared with 22.5% in 617 children in the UK.8 Unknown environmental factors may be involved in the generation of t(12;21)-positive ALL occurring in adults, and it is noteworthy that AML1 translocations may occur in therapy-related leukemias arising after exposure to topoisomerase II inhibitors,9 and that deletion and duplication events occur at the site of ETV6-AML1 recombination, suggesting that staggered double-strand DNA breaks may be at the origin of the translocation.10 Thus, the interplay between racial and environmental factors may account for the variable incidence of this translocation in different countries.6