Advertisement

The HLA-B −21 dimorphism impacts on NK cell education and clinical outcome of immunotherapy in acute myeloid leukemia

Alexander Hallner, Elin Bernson, Brwa Ali Hussein, Frida Ewald Sander, Mats Brune, Johan Aurelius, Anna Martner, Kristoffer Hellstrand and Fredrik B. Thorén

Data supplements

Article Figures & Data

Figures

  • Figure 1.

    NKG2A and granzyme B distribution and responses of NK cells in vivo and in vitro. (A) Distribution of NK cell subsets in AML patients before treatment start (n = 64) and in healthy donors (n = 24). Frequency of NKG2A+ NK cells in subsets expressing 0, 1, 2, or 3 KIRs in AML patients (n = 54) before or after cycle 1 (B) and in unstimulated and IL-2–stimulated (500 U/ml) healthy donors (n = 6) (C). (D) Fold change, compared with start, in messenger RNA expression of NKG2A after IL-2 stimulation. (E) Granzyme B expression (median fluorescence intensity [MFI]) in unstimulated and IL-2–stimulated NK cell subsets (n = 24). (F) Percentage increase in granzyme B response between unstimulated and IL-2–stimulated NK cells (n = 24). (G) LFS for patients dichotomized by median NKG2A expression at treatment start (low, n = 29 or high, n = 30). (H) LFS for patients dichotomized by median frequency of NKG2A+ NK cells (low, n = 29 or high, n = 30). *P < .05, **P < .005, ***P < .0005.

  • Figure 2.

    NKG2A+ NK cell responses toward HLA-E–expressing cell lines and AML blasts. (A) Degranulation responses of IL-2–stimulated NKG2A+ NK cells (n = 8) toward .221-WT, .221-AEH, T2 cells without HLA peptide (Ctrl), and T2 cells with HLA peptide (T2-E). (B) Comparison of remaining response of NKG2A+ NK cells with the presence of HLA-E compared with the response observed in the HLA-E control cell lines. (C) HLA-E expression in .221-WT and .221-AEH cells and titration curve with HLA peptide of HLA-E expression in T2 cells. (D) Percentage of CD107a+ cells among IL-2–activated NKG2A+ NK cells after exposure to T2 cells with indicated HLA-E expression. (E) Expression of MHC class I and HLA-E on PBMCs from healthy donors (n = 10) and on CD34+ AML blasts (n = 8). (F) Frequency of CD107a+ NK cells in indicated NK cell subsets against HLA-matched CD34+ AML blasts from 8 patients. (G) Frequency of CD107a+ NK cells (unstimulated, n = 10; IL-2 stimulated, n = 12) in indicated NK cell subsets against HLA-matched CD34+ AML blasts from 2 patients that triggered substantial degranulation in (F). n.s, not significant. *P < .05, **P < .005, ***P < .0005.

  • Figure 3.

    NK cell educational responses based on HLA-B −21. (A) Frequency of responding NK cells toward K562 cells in terms of CD107a, IFN-γ, and TNF-α positivity in unstimulated and IL-2–stimulated NK cells from donors with an M/x (red, n = 11) or T/T (white, n = 13) genotype. (B) Frequency of responding NK cell subsets in the K562 assay as in (A). (C) Frequency of CD107a+ NK cells against HLA-matched CD34+ AML blasts using unstimulated and IL-2–stimulated NK cells from M/x (red, n = 11) or T/T (white, n = 10) donors (left panel). Responses in IL-2–stimulated NK cell subsets (right panel). *P < .05, **P < .005.

  • Figure 4.

    Impact of HLA-B −21 genotype on outcome of AML patients. LFS (A) and OS (B) for patients with an M/x genotype (n = 38) or T/T genotype (n = 42) receiving HDC/IL-2 treatment. Five patients (6%) had a M/M genotype, 33 patients (41%) had an M/T genotype, and 42 patients (53%) had a T/T genotype. LFS (C) and OS (D) for AML patients undergoing transplantation with an M/x (n = 32) or T/T (n = 27) genotype.