Recent advances in genetic tools for leukemic diagnosis have facilitated studies on genetic traits which in concert with a driving mutation if present may influence the phenotype of the disease. Flt3-ITD is present in about 30% of AML patients and tends to be associated with a poor prognosis. Currently the use of inhibitors of mutated and in consequence activated kinases is undergoing clinical trials. The dim prognosis of Flt3-ITD positive AML and the hope associated with kinases inhibitors prompted us to evaluate in some depth the genetics of AML cases in relation to the presence of Flt3-ITD .

Ninety-two AML patients (diagnosed according to the FAB/WHO criteria, F/M=49/43, age median: 57, range: 21-81 years) who were seen in our institution were recruited to the present study which was conducted as follows:

The patients underwent GTG karyotyping and/or FISH for X or Y deletion, inv (3), -5/5q-, -7/7q-,+8, MLL, RUNX1, PML/RARA or RARA, inv(16). 52 lacking GTG and/or FISH abnormalities were called as normal karyotype (NK) patients, among those, 20 were Flt3-ITD positive. The other 40 patients had karyotype abnormalities or were FISH positive for AML-associated abnormalities among which 10 had Flt3-ITD .

In all 92 patients microarray analysis was performed using Agilent - Catalog Agilent Cancer CGH+SNP 180K or Roche - WG Catalog NimbleGene 12x270K microarrays.

We found (Partec Genomics Suite and Biodiscovery Nexus 8 software) that NK Flt3-ITD negative had, when the whole genome was analyzed, a higher number of total copy number variation (CNV) aberrations (amplifications and/or deletions, vs 32.21±8.70 vs 16.15±2.60, p=0.008) including one copy losses, (2.79±0.57 vs 1.20±0.35 p=0.019)as opposed to NK patients having Flt3-ITD mutation .

During the study, an interim analysis was performed at the time we switched from Roche to Agilent microarrays which revealed the significant prevalence of CNVs in AML Flt3-ITD negative patients over Flt3-ITD positive ones within the region 17q21.31 specifically in the positions from 44281092 to 4476549, 44702953 to 44745272 and 44745272 to 44765496. The same region appeared again as significantly discriminating Flt3-ITD negative and positive patients at the final analysis. Therefore, this region was analyzed in some depth. Hg19 annotation was used and according to that the region encompasses the KANSL1 gene among two others (NSFP1 and NSF). The NSF gene encodes a protein involved in transport to the Golgi and NSFP1 is a pseudogene. KANSL1, a component of the MLL1 complex (thought to contribute to leukemogenesis), was the focus of our interest, and the gene was further investigated to validate the microarray-based platform results and to localize the CNV area more precisely. In the region 44281092 to 4476549 within the KANSL1 gene Flt3-ITD negative patients had more deletions and amplifications as compared to those Flt3-ITD positive (31/62 vs 7/30, p=0.012). The same was valid when only deletions were considered (19/62 vs 4/30, p=0.058).

Then to validate microarray platform results the same 80 samples were examined using the TaqMan Copy Number Assay (ThermoFisher) for the KANSL1 gene (primers and probe designed for the region 44297111-44297201 lying in chr17), and as a reference the RNaseP gene was employed. The results obtained using the microarray platform correlated well (r=0.671, p<0.001) with those obtained using RT-PCR copy number assay for KANSL1 .

61 DNA samples obtained from the patients' solid organ biopsies were examined using the Copy Number Assay for KANSL1 gene CNV. In all instances except one the same CNP was observed in the marrow as that in the solid organ biopsies. It lends credit to the technology employed in this study, but still we do not have enough data to exclude the somatic origin of the microdeletions.

The microdeletions are of biologic significance, as suggested by the following findings:

  1. All three patients with FISH-documented MLL amplification had deletions within the KANSL1 gene (chr:17 44281092 to 4476549).

  2. Microdeletions within the KANSL1 gene were more frequently observed in the group M0-M2 patients than in other subtypes of AML (18/51 vs 5/35, p=0.026).

In conclusion, a lack of Flt3-ITD and normal karyotype was associated with the frequent presence of microdeletions within the chr17 cytoband q21.31. The latter was identified as (i) present within the KANSL1 gene and (ii) influencing the phenotype of AML.

Supported by the INNOMED/I/1/NCBR/2014


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.