Screening of novel genetic aberrations in pediatric acute myeloid leukemia: a report from the AIEOP AML-2002 study group

To the editor:

Acute myeloid leukemia (AML) is a heterogeneous disease with known specific recurrent genetic aberrations. The continuous and increasing identification of new genetic lesions has permitted the identification of new subgroups of patients with different prognosis.¹  In the present work, we evaluated the incidence of rare genetic abnormalities in pediatric AML such as del(4)(q12)FIP1L1-PDGFRA, t(16;21)(p11;q22)FUS/ERG, t(8;16)(p11;p13)MOZ/CBP, t(11;17)(q23;q12-21)MLL/AF17, t(4;11)(q35;q23)MLL/ArgB2, t(5;11)(q35;p15.5)NUP98/NSD1, t(3;5)(q25;q34)NPM1/MLF1, and MLLPTD in 306 children with newly diagnosed de novo AML other than acute promyelocytic leukemia enrolled in Associazione Italiana Ematologia Oncologia Pediatrica (AIEOP) centers from 2000 to 2009,²  all negative for known recurrent genetic abnormalities involving MLL, CBFB, and FLT3 genes (77 males and 77 females, median age at diagnosis 7.2 years, range 17 days to 17 years). RNA was extracted from fresh bone marrow at diagnosis, and multiplex RT-PCR was used. Sequencing by Sanger method was applied to all positive cases to characterize fusion breakpoints. We identified 1 patient each positive for t(16;21)(p11;q22)FUS/ERG, t(11;17)(q23;q12-21)MLL/AF17, and t(4;11)(q35;q23)MLL/ArgB2, respectively, suggesting that these rearrangements are extremely rare in pediatric AML. Two of the 306 patients had del(4)(q12)FIP1L1/PDGFRA, and 4 had t(8;16)(p11;p13)MOZ/CBP. Interestingly, 6 patients (2%) had t(3;5)(q25;q34)NPM1/MLF1, 6 (2%) had MLLPTD, and 6 (2%) were found to carry t(5;11)(q35;p15.5)NUP98/NSD1. In our pediatric cohort, the incidence of this last aberration is lower than that previously reported by Hollink et al.³  Subsequently, because a strong association of t(5;11) fusion with FLT3ITD has been described (91%),³  we extended the screening to 42 children with de novo AML harboring the FLT3ITD mutation, enrolled in the AIEOP-LAM 2002 protocol. We found that 6 of 42 (14%) had the NUP98-NSD1 fusion. So, 6 of 12 NUP98/NSD1-positive patients (50%) were FLT3ITD positive, showing a lower association in our pediatric cohort for these 2 aberrancies than that reported by Hollink et al.³  Then, we looked at the event-free survival (EFS) of patients with t(5;11)NUP98-NDS1 (n = 12) and found that it was worse, compared with patients negative for known molecular lesions and enrolled into the LAM 2002-AIEOP protocol (30.1% vs 57.1% at 3 years, P < .05).⁴  Furthermore, we did not find any difference in either clinical or biologic features between patients with isolated t(5;11) and those with t(5;11) + FLT3ITD (Figure 1). The 8-year EFS of FLT3ITD+ children who did or did not carry t(5,11) was 33.3% and 42.7% (P = .2), respectively. This finding suggested that NUP98/NSD1 fusion protein identifies a previously unrecognized subgroup of FLT3ITD patients with an even worse prognosis.

To test whether MLLPTD might also play a role in the occurrence of childhood AML relapse, we analyzed samples from 40 AML patients at relapse, never finding this abnormality. By contrast, 4 patients harbored at relapse the same MLLPTD found at diagnosis, suggesting the stability of this mutation.

In summary, we confirm that t(5,11) is not exceptional in pediatric AML, being frequently associated with FLT3ITD, and identifying patients at high risk of treatment failure. We also suggest a negative role of this translocation in FLT3ITD positive patients to be further considered in the risk stratification of patients. The putative role of the remaining rare abnormalities^5,6  in AML remains to be confirmed in prospective studies with larger cohorts of patients.