Comprehensive Fusion Gene Analysis Of Pediatric Non-Down Syndrome Acute Megakaryoblasitc Leukemia

Pediatric non-Down syndrome acute megakaryoblastic leukemia (non-DS-AMKL) is a biologically and clinically unique subset of acute myeloid leukemia. Though non-DS-AMKL was thought to be a poor prognostic disease with a survival rate of 14-34%, recent improvements of diagnosis and intensive chemotherapy with effective supportive care have increased the rate to 50-70%. Little was known about the specific molecular marker in non-DS-AMKL other than t(1;22)(p13;q13)/OTT-MAL, however, new diagnostic and prognostic markers have been determined using new techniques such as high-resolution array CGH and next generation sequencing. Recent studies have identified new recurrent cryptic translocations, inv(16)(p13.3q24.3) and t(11;15)(p15;q35), which created CBFA2T3-GLIS2 and NUP98-JARID1A gene fusions, respectively. It has been reported that CBFA2T3-GLIS2 is a novel independent factor for dismal clinical outcome and CBFA2T3-GLIS2 positive cases should be classified as the high risk group, meanwhile, the significance of NUP98-JARID1Ais less clear at present. The aim of this study is to identify the more accurate frequency and clinical impact of such fusion genes in non-DS-AMKL, which will dissect the biology of this distinct subtype of leukemia.

We analyzed 44 newly diagnosed pediatric non-DS-AMKL cases enrolled in two Japan clinical trials, the AML-99 study and AML-05 study. The median age was 1 year of age, and median white blood cell count at diagnosis was 23 x 10⁹/l. Morphologic classification and karyotyping of the cases enrolled in AML-05 study were centrally reviewed. Molecular characterization included the analyses of CBFA2T3-GLIS2, NUP98-JARID1A, NUP98-NSD1, OTT-MAL, MLL-AF9, MLL-AF10, MLL-PTD, FLT3-ITD, KIT, RAS, WT1, and NPM1. They were determined by RT-PCR and/or PCR followed by direct sequencing, though only MLL-PTD was determined by MLPA method. Some survival analyses were restricted to the cases enrolled in the AML-05 study.

CBFA2T3-GLIS2 was identified in 12 cases (27.3%), NUP98-JARID1A in 4 cases (9.1%), and OTT-MAL in 11 cases (25%), MLL-AF9 in 2 cases (4.5%), and MLL-AF10 in 1 case (2.3%). Only CBFA2T3-GLIS2-positive cases had a high white blood cell count compared with other cases (median, 33x10⁹/l for positive cases vs 20x10⁹/l for negative cases, p=0.0807). Seven of the 12 CBFA2T3-GLIS2-positive cases died, while 4 of the 11 OTT-MAL-positive cases died. In the AML-05 study, 2-years Overall survival rates (OS) of non-DS-AMKL cases (n=34) were 36.4% for CBFA2T3-GLIS2-positive cases (n=11) vs 73.9% for negative cases (n=23) (p=0.0773), and 70.0% for OTT-MAL-positive cases ( n=10 ) vs 58.3% for negative cases ( n=24 ) (p=0.75). Regarding NUP98-JARID1A-positive cases, the number was too small to perform outcome analyses, although 2 cases relapsed and died. Among all cases, 4 of 5 cases with normal karyotype and 5 of 6 cases with acquired 21 trisomy had CBFA2T3-GLIS2, suggesting that those karyotypes may be useful features to suspect the presence of CBFA2T3-GLIS2 in the cases. Twenty-eight DS-AMKL samples did not harbor any CBFA2T3-GLIS2, NUP98-JARID1A, or OTT-MAL.

In addition to well-known OTT-MAL, CBFA2T3-GLIS2 and NUP98-JARID1A are new recurrent fusion genes in pediatric non-DS-AMKL, which demonstrates that pediatric non-DS-AMKL is a distinct subtype of pediatric AML. CBFA2T3-GLIS2 is an independent poor prognostic factor and OTT-MAL may be a relatively good prognostic factor, while the impact of NUP98-JARID1A is still unknown and further investigation is needed. Therefore, routine screening for these fusion genes at diagnosis will be essential for proper identification and stratification of pediatric AML cases.

No relevant conflicts of interest to declare.