Presence of MLL Gene Rearrangements in Infant Acute Leukemia Could be Predicted By Tumor cells’ Immunophenotype

Acute leukemia (AL) in children less than 1 year old is the relatively rare disease with specific biological features and poor outcome. It is also characterized by high incidence of MLL gene rearrangements. Immunophenotype of infants’ leukemia varies due to presence or absence of MLL gene rearrangements.

Aim of the study –description of immunophenotype in infant acute lymphoblastic and acute myeloid leukemia (ALL and AML respectively) due to presence of MLL gene rearrangements.

Methods. Totally 540 cases of pediatric AL were studied. 113 patients (59 boys and 54 girls) aged from 5 days to 11 months were included in the study group. Their data was compared to 427 cases of acute leukemia in older children. Tumor cells immunophenotyping was performed by 6-8-color flow cytometry. Detection of various types of MLL-gene rearrangements was done by fluorescence in-situ hybridization, reverse-transcriptase polymerase chain reaction (PCR) and long-distance inverse PCR.

Results. ALL was found less frequently in infants than in older children (68.1% and 86.9% respectively, p<0.001) while percentage of acute myeloid leukemia cases was higher in infants (27.4% and 11.5% respectively, p<0.001). Significant immunophenotypic differences were observed in patients with and without MLL gene rearrangements in both ALL and AML. Number of ALL cases in those tumor cells expressed CD10, CD20, CD45, CD133, CD15, CD65 NG2 significantly varied between MLL-positive and MLL-negative groups (p<0.001, p<0.001, p=0.002, p<0.001, p=0.004, p=0.019 and p<0.001 respectively). NG2-positivity represented the highest overall correct prediction (OCP) rate for presence of MLL-rearrangements (90.6%). Diagnostic accuracy of CD20-negativity and CD45-positivity was lower (81.2% and 81.9% respectively) while OCP for CD10-negativity (76.4%), CD133-positivity (76.5%) CD15-positivity (67.7%) and CD65-positivity (53.7) was not sufficient enough. Nevertheless CD10-positive BCP-ALL with MLL-rearrangements differed from CD10(+) cases in MLL-germline group. CD10 homogeneous expression was noted frequently in MLL-germline cases than in MLL-rearranged ones (p=0.001). Although there were found no significant differences in CD22-positive patients’ number, CD22(+)-cells percentage was significantly lower in MLL-positive cases (median 89.9%, range 25.2-99.7% and median 99.9%, range 96.0-99.9% respectively, р=0.003). Thus CD20-negativity, CD10-negativity/low expression, high CD45, CD15, CD65, CD133 and NG2 expression, decreased CD22-expression are immunophenotypic signatures of MLL-rearranged infant ALL, although NG2 has the highest diagnostic efficacy. Interestingly CD10-negativity and positivity for CD34, CD15 and CD65 could be able to distinguish MLL-AF4-positive cases from patients carrying other types of MLL-rearrangements. Number of AML cases in those tumor cells expressed CD99, CD133, CD15, CD65, CD4, CD11b, CD61, NG2 varied between MLL-positive and MLL-negative groups ((p=0.019, p=0.012, p=0.002, p=0.004, p=0.005, p<0.001, p=0.015 and p<0.001 respectively). Thus CD61-negativity, high CD99, CD15, CD133, CD65, CD11b, CD4 and NG2 expression were immunophenotypic signatures of MLL-rearranged infant AML, although CD11b and NG2 had the highest diagnostic efficacy (95.0% and 89.7% respectively).

Conclusion. Thus immunophenotype of AL in children less than 1 year old differs significantly from patients of older age groups. Infants’ ALL and AML immunophenotype varies greatly due to the presence of MLL gene rearrangements. Complex diagnostic immunophenotyping of infants’ AL allows predicting presence of MLL rearrangements while NG2 and CD11b are the most applicable single markers for ALL and AML respectively.