Advertisement

Natural killer receptor ligand expression on acute myeloid leukemia impacts survival and relapse after chemotherapy

Sara Mastaglio, Eric Wong, Travis Perera, Jane Ripley, Piers Blombery, Mark J. Smyth, Rachel Koldej and David Ritchie

Data supplements

Article Figures & Data

Figures

  • NKRLs are variably expressed by AML blasts. The expression of 6 activating (MICA, MICB, CD155, CD112, ULBP1, and ULBP2/5/6) and 3 inhibitory (PD-L1, PD-L2, and HLA class I) NKRLs was detected by flow cytometry on cryopreserved bone marrow blasts harvested from 66 patients with a new diagnosis of AML. Side scatter (SSC-A) and CD45 were used to gate the blast population. CD33 and CD34 were used as additional markers of disease. The RFI was calculated for each ligand as follows: mean fluorescence intensity (MFI) of the ligand/MFI of the unstained control. (A) NKRL RFIs calculated and plotted for all 66 patients. Each replicate is shown; black horizontal line corresponds to median values. (B) Representative plots of a patient expressing only inhibitory ligands (HLA-I and PD-L2). (C) Representative plots of a patient with high expression of both activating and inhibitory NKRLs. (D) Representative plots of a patient with a double blast population (CD33posCD34pos and CD33posCD34dim), where the expression of ULBP1 was strikingly different: the CD33posCD34pos blasts expressed low levels of ULBP1 (RFI = 5.5), whereas the activating ligand was highly expressed on CD33posCD34dim cells (RFI = 69.5). Blue histograms correspond to unstained control MFI; red histograms correspond to NKRL MFI. (E) An RFI score ranging from 0 to 2 was calculated for each ligand as follows: (1) samples with RFI <2 (ie, less than twofold increase of median fluorescence above background) were considered negative and assigned score 0; (2) Samples with an RFI value between 2 and the III quartile were assigned score 1; (3) Samples with RFI equal to or above the III quartile were assigned score 2. Panel E represents the distribution of patients according to the RFI score for each NKRL.

  • NKRLs are variably expressed by AML blasts. The expression of 6 activating (MICA, MICB, CD155, CD112, ULBP1, and ULBP2/5/6) and 3 inhibitory (PD-L1, PD-L2, and HLA class I) NKRLs was detected by flow cytometry on cryopreserved bone marrow blasts harvested from 66 patients with a new diagnosis of AML. Side scatter (SSC-A) and CD45 were used to gate the blast population. CD33 and CD34 were used as additional markers of disease. The RFI was calculated for each ligand as follows: mean fluorescence intensity (MFI) of the ligand/MFI of the unstained control. (A) NKRL RFIs calculated and plotted for all 66 patients. Each replicate is shown; black horizontal line corresponds to median values. (B) Representative plots of a patient expressing only inhibitory ligands (HLA-I and PD-L2). (C) Representative plots of a patient with high expression of both activating and inhibitory NKRLs. (D) Representative plots of a patient with a double blast population (CD33posCD34pos and CD33posCD34dim), where the expression of ULBP1 was strikingly different: the CD33posCD34pos blasts expressed low levels of ULBP1 (RFI = 5.5), whereas the activating ligand was highly expressed on CD33posCD34dim cells (RFI = 69.5). Blue histograms correspond to unstained control MFI; red histograms correspond to NKRL MFI. (E) An RFI score ranging from 0 to 2 was calculated for each ligand as follows: (1) samples with RFI <2 (ie, less than twofold increase of median fluorescence above background) were considered negative and assigned score 0; (2) Samples with an RFI value between 2 and the III quartile were assigned score 1; (3) Samples with RFI equal to or above the III quartile were assigned score 2. Panel E represents the distribution of patients according to the RFI score for each NKRL.

  • NKRLs are variably expressed by AML blasts. The expression of 6 activating (MICA, MICB, CD155, CD112, ULBP1, and ULBP2/5/6) and 3 inhibitory (PD-L1, PD-L2, and HLA class I) NKRLs was detected by flow cytometry on cryopreserved bone marrow blasts harvested from 66 patients with a new diagnosis of AML. Side scatter (SSC-A) and CD45 were used to gate the blast population. CD33 and CD34 were used as additional markers of disease. The RFI was calculated for each ligand as follows: mean fluorescence intensity (MFI) of the ligand/MFI of the unstained control. (A) NKRL RFIs calculated and plotted for all 66 patients. Each replicate is shown; black horizontal line corresponds to median values. (B) Representative plots of a patient expressing only inhibitory ligands (HLA-I and PD-L2). (C) Representative plots of a patient with high expression of both activating and inhibitory NKRLs. (D) Representative plots of a patient with a double blast population (CD33posCD34pos and CD33posCD34dim), where the expression of ULBP1 was strikingly different: the CD33posCD34pos blasts expressed low levels of ULBP1 (RFI = 5.5), whereas the activating ligand was highly expressed on CD33posCD34dim cells (RFI = 69.5). Blue histograms correspond to unstained control MFI; red histograms correspond to NKRL MFI. (E) An RFI score ranging from 0 to 2 was calculated for each ligand as follows: (1) samples with RFI <2 (ie, less than twofold increase of median fluorescence above background) were considered negative and assigned score 0; (2) Samples with an RFI value between 2 and the III quartile were assigned score 1; (3) Samples with RFI equal to or above the III quartile were assigned score 2. Panel E represents the distribution of patients according to the RFI score for each NKRL.

  • Figure 2.

    ULBP1 positivity at diagnosis correlates with better outcome. Kaplan-Meier estimates of OS (A), CIR (B), and RFS (C), according to ULBP1 expression: patients with blasts expressing and nonexpressing ULBP1 at diagnosis are shown by red and blue lines, respectively. P values are shown.

  • Figure 3.

    NKTL net score and group stratification. A flow cytometry net score ranging from −2 to 8 was calculated for each patient as follows: (sum of each activating ligand RFI score) − (sum of each inhibitory ligand RFI score). The histogram represents the distribution of patients according to their net score. A further stratification was performed in order to correlate NKRL expression to clinical outcome: patients were split into 2 homogeneous groups according to their flow cytometry net score: group 0 (G0) with score −2 to 1 and group 1 (G1) with score 2 to 8.

  • Figure 4.

    The activating pattern of NKRLs correlates with better outcome. Kaplan-Meier estimates of OS (A), CIR (B), and RFS (C), according to flow cytometry group stratification: patients belonging to group 0 (NKRL inhibitory pattern) and 1 (NK receptor ligand activating pattern) are represented by blue and red lines, respectively. P values are shown.

  • Figure 5.

    NKRL groups stratify patients with intermediate prognosis. Only patients belonging to the intermediate cytogenetic (A) and intermediate molecular risk according to ELN criteria32 (B) were analyzed for OS, CIR, and RFS. Patients belonging to group 0 and 1 are represented by blue and red lines, respectively. (C) Patients were stratified by cytogenetic prognosis, and subjects with intermediate cytogenetic risk were further classified according to their NKRL group: purple line corresponds to favorable cytogenetics; red line corresponds to intermediate cytogenetics and NKRL group 1 (activating pattern); light green line defines intermediate cytogenetics and NKRL group 0 (inhibitory pattern); blue line corresponds to adverse cytogenetics. Kaplan-Meier estimates for OS (left panel) and RFS (right panel) are shown. P values are shown.

Tables

  • Table 1.

    Baseline patient characteristics

    AllGroup 0Group 1P
    Patients, n (%)66 (100)32 (48)34 (52)
    Age at diagnosis, median (range), y56 (16-78)48 (16-78)60 (26-77).3 (>60 vs <60)
    Sex, M/F37/2917/1520/14
    % BM blasts at diagnosis, median (range)78 (20-97)81 (21-94)70 (20-97)
    Cytogenetic risk, n (%).7 (favorable vs intermediate/adverse)
     Favorable7 (10.6)4 (12.5)3 (8.8)
     Intermediate46 (69.7)20 (62.5)26 (76.5)
     Adverse13 (19.7)8 (25)5 (14.7)
    ELN, evaluable, n (%)60 (91)30 (93.8)30 (88.2).08 (favorable vs intermediate/adverse)
     Favorable20 (33.3)7 (23.3)13 (43.3)
     Intermediate15 (25)7 (23.3)8 (26.7)
     Adverse25 (41.7)16 (53.3)9 (30)
    Response to iCT, n (%)
     CR45 (68.2)22 (68.8)23 (67.6)
     NR11 (16.7)6 (18.8)5 (14.7)
     n.e./death10 (15.2)4 (12.5)6 (17.6)
    AlloHSCT, n (%)
     No46 (69.7)20 (62.5)26 (76.5)
     Yes16 (24.2)11 (34.4)5 (14.7)
     n.a.4 (6)1 (3.1)3 (8.8)
    • All patients (All) and patients stratified according to their flow cytometry NKRL group (Group 0, NK receptor ligand inhibitory pattern; and Group 1, NKRL activating pattern) are shown separately.

    • BM, bone marrow; CR, complete remission; F, female; iCT, induction chemotherapy; M, male; n.a., not available; n.e., not evaluable; NR, nonresponse.