Abnormalities of 12p are relatively rare in acute myeloid leukemia (AML), but are of particular interest because ETV6, an ETS family transcription factor that frequently acts as a tumor suppressor, maps to 12p13. In pediatric AML the recurrent t(7;12)(q36;p13) has been recognized in addition to other heterogeneous 12p abnormalities which results in the rearrangement or deletion of ETV6. Recently, a Medical Research Council study found that AML patients with 12p abnormalities have a poor prognosis. To date, abnormalities of 12p have been categorized as standard cytogenetic risk in Children’s Oncology Group (COG) AML trials, but there are limited data on their frequency, morphologic classification, and concomitant gene mutations or chromosome abnormalities.


To evaluate characteristics and prognostic significance of 12p abnormalities in AML in the COG trial AAML0531.


AAML0531 is a recently completed COG trial that enrolled patients between 2006 and 2010. Central review of cytogenetics was conducted on 981 of 1070 enrolled patients (non-Down syndrome) with de novo AML. Abnormalities of 12p were classified into 3 groups: 1) the t(7;12) or variants thereof; 2) abnormalities of band 12p13 other than the t(7;12) that were confirmed or highly suggestive of ETV6 rearrangement (other ETV6R); and 3) abnormalities resulting in loss of the band 12p13 (ETV6Loss). The entire group of children with 12p abnormalities (N=32) as well as the three subgroups were compared with children (N=949) without 12p abnormalities, with respect to demographics, frequency of gene mutation, additional cytogenetic abnormalities and survival.


Abnormalities of 12p were identified in 32 (3.3%) of patients. 8 patients had a t(7;12) after central review (6 were missed in the local laboratory’s initial submission); 10 had otherETV6R, and 14 had ETV6Loss. The median age of the 12p group was significantly lower (median, 5.8; range, 0.2–18.2) than the non-12p group (median, 9.9; range, 0.003–29.8; P=.018). Notable was the very young age of patients with the t(7;12) (median, 0.6; range, 0.2–6.4; P=.004), with 88% <2 years. In the 12p group, there were no CEBPA mutations; 1 FLT3-ITD and 2 NPM1 mutations (in the other ETV6R and ETV6Loss groups). These frequencies of genetic mutations were not significantly different from those in the non-12p group. Patients with 12p abnormalities were significantly less likely to have associated low-risk cytogenetic abnormalities of inv(16) or t(8;21) compared to the non-12p group [t(8;21) 0% vs 15%, P = .016 and inv(16) 0% vs 11% , P=.041]. Further, the proportion of patients with t(9;11) was 0% in the t(7;12) group compared to 20%, 21% and 21% in the other ETV6R, ETV6Loss, and non-12p groups, respectively. Six of 8 patients with t(7;12) had +19 as a secondary abnormality which was absent in the other 12p groups. Six of 7 patients with ETV6Loss had complex karyotypes with >6 abnormalities, compared to 0% in the other 12p groups and 4% in the non-12p groups (P<.002). On the basis of karyotype classification strategy in AAML0531 no patient with 12p abnormalities was low risk, 29 were standard risk, and 3 were high risk [due to presence of -5/5q- or -7].

The 12p abnormality group had a worse 3-year OS (28±17% vs. 69±3%, P<.001) and worse 3-year EFS (14±13% vs. 51±3%, P<.001) compared with the non-12p group. For those patients who achieved CR, the RR from end of induction 1 was 37±4% for non-12p vs. 83±40% for the 12p group (P<.001). 3-year OS was 54±40% for the t(7;12), 35±32% for the otherETV6R and 11%±19% for ETV6Loss (P<.03); 3- year EFS rates were 15±28%, 13±3%, and 14±19%, respectively (P>.1).


Patients with abnormalities of 12p should be considered for categorization as high-risk with respect to cytogenetic stratification for AML. Given the subtle nature of the t(7;12) and other ETV6 rearrangements and their prognostic implications FISH for ETV6 is recommended upfront for the evaluation of any young patient with AML with a +19 as a secondary abnormality and in any patient suggestive of 12p deletion or rearrangement.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.