Pediatric AML still represent an unfavourable disease resulting from the heterogeneous clonal expansion of malignant transformed haematopoietic stem or progenitor cell. The leukemia cell population is continuously replenish by rare, functionally distinct “leukaemia stem cells” (LSC) endowed with the capacity to self renew as well with the ability to generate clonogenic leukemic progenitors The AML-LSCs have been well documented and seem to behave like quiescent or slowly dividing hematopoietic stem cells. Therefore, LSC are considered less sensitive to treatments, which rather target actively dividing cells, and responsible for relapse. Recently, Y. Ma et al. suggested a major role of SALL4 gene both in stemness activity and leukemia transformation of normal hematopoietic stem cells. We sought to evaluate the expression of SALL4 gene in a panel of 88 pediatric AML, 60 Acute Lymphoblastic Leukemia (T and B ALL) and a few hematopoietic normal tissues. SALL4 expression was determined by quantitative RT-PCR in pre-treatment bone marrow samples (BM) (median blasts: 80%) and in normal tissues. SALL4 expression was much higher in AML compared to ALL (p<0.0001) and normal tissue (p < 0.0001). Values varied greatly among AML samples (range: 0 to 144; median 0.3). AML samples were dichotomized at SALL4’s median expression into“low” (n=44) or “high“(n=44) expressers. High expression was correlated to cytogenetic risk, as defined by MRC classes (p= 0.0008) but not to FAB categories, WBC, MLL rearrangements or FLT3, N RAS and NPM1 genotypes. Low expression was significant associated to t(8;21) translocation (P=0.02). SALL expression progressively increased from MRC1 class (low risk) to MRC3 (high risk) (p=0.002). High SALL4 patients had a shorter OS (OS: 46.8%±8% vs 74%±7%, p= 0.035) and a shorter EFS (EFS: 40% ± 8% vs 58% ±8%; P= 0.2) compared to low SALL4 expressers, although the latter association was less strong. As the highest values of SALL4 correlated with intermediate (MRC2) and high (MRC3) cytogenetic risk, we looked at fitting a model based upon the 4 quartiles of SALL4 expression. The upper quartile (values > 16, n= 12) had the worst outcome compared to the three others. Once stratified on MRC groups, MRC2 patients in the upper SALL4 quartile had 3.2 times more risk of relapse (HR= 3.2, CI95%: 1.3–7.8, P=0.02) and 5.4 more risk to die (HR=5.4, CI95%: 1.8–7.6; P= 0.0005) than MRC2 patients in the three others quartiles. In conclusion, SALL4 expression level may define an important risk factor in AML, particularly among patients with cytogenetic intermediate risk.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author