KMT2A-rearranged pediatric AML selectively overexpresses FLT3, and CD33/FLT3 represents a promising immunotherapy combination KMT2A-rearranged AML. (A-B) Eleven surface molecules were quantified by percentage of positive cells (A) and MFI ratio (B) in KMT2A-rearranged AML (n = 5) and KMT2A wild-type AML. The only surface molecule found to be significantly overexpressed in KMT2A-mutated AML was FLT3 (84% vs 36% by percentage, P = .02; 3.4 vs 2.0 by MFI ratio, P = 0,0007). (C-D) Eleven surface molecules were quantified by percentage of positive cells (C) and MFI ratio (D) on t(8;21) and non-t(8;21) AML samples. (E-F) Volcano plots of differential expression in KMT2A-mutated and nonmutated pediatric AML samples. Only FLT3 is significantly overexpressed in KMT2A-mutated AML. (G) tSNE analysis using 36 BM samples, including 9 healthy donors, demonstrates that KMT2A-mutated AML cells cluster within AML samples that themselves cluster from healthy hematopoietic progenitors. (H) When CD45^dim cells of 5 KMT2A-mutated AML samples were concatenated and CD33 and FLT3 expression was quantified, an almost homogenous coexpression of CD33 and FLT3 was observed. (I) Setting 50% concomitant expression of 2 target molecules as arbitrary threshold, 16 out of 21 primary AML samples and 8 out of 15 relapse samples could be addressed with CD33/CLEC12A targeting. CD33/HAVCR2 targeting was particularly effective in AML relapse samples that could not be addressed with CD33/CLEC12A targeting, as 4 out of 7 samples were >50% double positive. wt, wild-type.