Figure 3.

Frequencies of 3 tested additive KIR/KIR-ligand models. (A-C) Blue columns are displaying frequencies of patients with AML, and red columns are displaying the control group with the respective category of scores/counts of the considered KIR/KIR-ligand-models: inhibitory score based on the findings by Boelen et al,¹⁹ including KIR3DL1/HLA-Bw4 and considering KIR2DL2 and KIR2DL3 separately, with a cutoff of 1.75 (A); inhibitory count based on the findings by Boelen et al, including KIR3DL1/HLA-Bw4 and considering KIR2DL2 and KIR2DL3 separately, with a cutoff of 1.0 (B); and favorable and unfavorable score based on the criteria given by Rafei et al³⁵ combining inhibitory and activating scores (C). The truncated version of KIR2DS4 is not considered as an activating KIR; unfavorable is defined by the presence of ≥3 inhibitory KIR/KIR-ligand matches and no activating KIR/KIR-ligand match. (D) Inhibitory/missing-ligand KIR score based on the criteria given by Krieger et al³⁶ accounting for iKIRs and missing KIR ligands; adjusted P values are depicted above the pairs of columns.