Advertisement

Article Figures & Data

Figures

  • Figure 1.

    Tumor biopsy samples. (A) Representative photomicrographs of HL tumor biopsy samples from different histologic subtypes ([i-iii] MC, NS, and NLP, respectively). The EBV status of tumor tissues was confirmed by EBER in situ hybridization (iv-vi) and LMP1 immunohistochemistry (viii-x). Tissue samples shown in panels iv, v, viii, and ix were positive, while vi and x were negative for EBV. (xii-xiv) LAG-3 protein on tumor-infiltrating lymphocytes in HL biopsies. The numbers of LAG-3+ cells in each tissue section were graded as described in Table 2. The tissue sections were graded +++ (xii), ++ (xiv), and negative (xiii). The relevant isotype controls are shown in panels vii, xi, and xv. (i-iii) Original magnification, × 10; (iv-xv) original magnification, × 40. (B) Representative lymph node sections from a patient with lymphocyte-rich EBV-positive classic HL. (i-ii) LAG-3 and FOXP3 protein on tumor-infiltrating lymphocytes, respectively; (iii) dual staining illustrates that the nuclear FOXP3 protein (red) and surface/cytoplasmic LAG-3 protein (brown) do not generally colocalize within the same lymphocyte. Photomicrographs were taken using a Nikon Coolpix 5700 camera (Nikon, Tokyo, Japan) and were acquired with Microsoft Office XP Photo Editor (Microsoft, Seattle, WA). Images were originally magnified under an Olympus CX41 microscope equipped with a 10×/0.65 NA (i-iii) or a 40×/0.25 NA (iv-xv) objective lens (Olympus, Tokyo, Japan).

  • Figure 2.

    LAG-3 is enriched within the CD4+ T cells expressing intracellular CTLA-4hi and GITRhi. (A) Representative data from an HL patient in remission showing LAG-3 staining and (B) HL patient PBMC samples (n = 14) showing mean fluorescence intensity for LAG-3 on CD4+ T cells with intracellular CTLA-4lo and CTLA-4hi and GITRlo and GITRhi cells. P values were generated using the Wilcoxon matched pairs test.

  • Figure 3.

    Cross-sectional analysis of Treg cell subsets in HL patients. Box and whisker plots summarizing the percentages of the mean (horizontal bar), standard error (box), and standard deviation (whiskers) of Treg cells in HL patients at diagnosis, in remission, and in randomly chosen healthy laboratory controls. The percentage of positive cells with each phenotype is shown on the vertical axis, and the P value and number of samples are indicated above and below each plot, respectively. P values were generated using the Mann-Whitney test.

  • Figure 4.

    Mean and standard error of ex vivo EBV-specific IFN-γ spot-forming cells following stimulation with peptide epitopes (SFC/106 PBMC). Responses were assessed before therapy (Before Rx) and 6 months (6m) and 12 months (12m) following diagnosis. The “6m” and “12m” columns include only data from newly diagnosed patients who went on to attain remission. These responses were compared with HL patients in long-term remission and healthy seropositive individuals. Panels A-C show data irrespective of EBV-tumor status. In panels A-B and D, “n” refers to the number of assays performed, with the number of patients shown in parentheses. (A) A cross-sectional analysis of LMP1/2-specific ELISPOT responses in HL patients at different time points and in healthy virus carriers. (B) A comparison of T-cell responses directed toward LMP1/2, EBNA3/4/6, and lytic epitopes in ND/RL HL patients and healthy virus carriers. (C) A matched-pair analysis of LMP1/2, EBNA3/4/6, and lytic epitope-specific CD8+ T-cell responses in HL patients both before and after receiving chemotherapy (Before Rx and After Rx). (D) LMP1/2-specific CD8+ T-cell responses in ND/RL EBV-positive and EBV-negative HL patients and ELISPOT responses in ND/RL HL patients with respect to clinical stage and Hasenclever/Diehl International Prognostic Score. In panels A and B, asterisks denote statistical significance.

  • Figure 5.

    Ex vivo enumeration of EBV-specific CD8+ T cells using MHC-peptide pentamers in PBMCs from HL patients. The values shown on the logarithmic axis indicate the percentage CD8+ T cells positive for HLA class I–peptide pentamers. There was no significant difference in pentamer frequencies between EBV-positive HL and EBV-negative HL cases.

  • Figure 6.

    Histograms showing mean and standard error demonstrating the association of intratumoral LAG-3 expression and LMP1/2-specific CD8+ T-cell responses in ND/RL HL patients prior to therapy. ELISPOT responses were compared in patients categorized as either low (0% to 25%) or high (26% to 100%) LAG-3–expressing tumor-infiltrating lymphocytes within HL lymph nodes.

  • Figure 7.

    Effect of HRS cell line supernatant on LMP effector T-cell function and expansion of Treg cells in vitro. PBMCs from 5 different healthy subjects were cultured in supernatants from HRS cell lines (L1236, HDLM2, L428, L540) for 3 days (A-B). A fibroblast line supernatant and serum free culture media (SFCM) alone were used as negative controls. (C) An LMP2A peptide–specific IFN-γ ELISPOT assay following 3-day incubation of PBMCs (from a healthy HLA-A2 EBV-seropositive subject) with supernatant from the L540 HL cell line. Results are representative of 3 separate experiments. (D) T-cell reactivity toward CLG epitope (HLA-A2 restricted, LMP2A) in PBMCs from 2 different HLA-A2–positive individuals (D1 and D2) with or without depletion of CD4+ LAG-3+ T cells. Data from fresh PBMCs and PBMCs cultured in HRS cell supernatant are shown.

Tables

  • Table 1.

    Patient characteristics

    CharacteristicNewly diagnosedRelapsedLong-term remission
    Total no. of patients 41 9 43
    Sex, F/M 14/27 5/4 17/26
    Mean age at diagnosis, y 31 41 30
    Clinical stage IIB/III/IV, no. (%) 20 (49) 5 (66) Not applicable
    Prognostic score 3 or more*, no. (%) 12 (29) 4 (44) Not applicable
    Timing of blood sample(s) Prior to therapy, 6 mo, 12 mo Prior to therapy, 6 mo, 12 mo Mean, 5 y; range, 2-33 y after diagnosis
    Histology, no. of total (%)
        Nodular sclerosing 27/41 (66) 7/9 (78) 18/36 (50)
        Mixed cellularity 7/41 (17) 1/9 (11) 6/36 (17)
        Lymphocyte rich 4/41 (10) 0/9 (0) 4/36 (11)
        Lymphocyte depleted 1/41 (2) 0/9 (0) 1/36 (3)
        Nodular lymphocyte predominant 2/41 (5) 1/9 (11) 3/36 (8)
        Unclassified Hodgkin 0/41 (0) 0/9 (0) 4/36 (11)
    Positive EBV serology, no. of total (%) 39/40 (98) 8/9 (89) 40/43 (93)
    Positive EBV tissue status§, no. of total (%) 15/41 (36) 1/9 (11) 9/27 (33)
    • * Hasenclever/Diehl International Prognostic Score.

    • Histology available in all newly diagnosed and relapsed cases and 36 of 43 (84%) remission cases.

    • Serology undetermined in 1 newly diagnosed case.

    • § Biopsy available for EBV testing in all newly diagnosed and relapsed cases and 27 of 43 (63%) remission cases. EBV tissue status correlated with histology, with 55% of mixed cellularity or lymphocyte-rich cases positive for EBV, against 27% of nodular sclerosing cases (r = 0.38, P = .013). In 2 of 67 cases the LMP1 and EBER stains were discordant (EBER positive/LMP1 negative). Both cases were classified as EBV tissue positive.

  • Table 2.

    Expression of LAG-3 by tumor-infiltrating lymphocytes in malignant HL lymph nodes

    Case no.Disease statusAge at diagnosis, y/sexHistologyLAG-3+ cells in HRS-positive regions*EBV tissue status
    2 ND 45/F MC +++ Negative
    8 ND 30/F NS ++ Negative
    18 RL 44/M NS ++ Negative
    22 Remission 48/M NS - Negative
    28 ND 21/F NS + Negative
    31 ND 76/M NS ++ Negative
    32 ND 30/M NS ++ Positive
    33 ND 26/M NS + Negative
    35 ND 48/F NS ++ Positive
    36 RL 34/F NS - Positive
    38 Remission 33/M NS - Positive
    40 ND 25/M NS + Positive
    44 ND 22/F NS + Positive
    47 ND 29/M MC ++ Negative
    48 Remission 35/F MC +++ Positive
    49 ND 45/F MC ++ Negative
    50 Remission 21/M NS + Negative
    51 RL 25/F NS - Negative
    53 ND 22/M MC ++ Positive
    54 Remission 24/F NS - Negative
    55 Remission 42/M NLP + Negative
    56 ND 20/M MC +++ Positive
    57 ND 21/F MC +++ Positive
    64 ND 43/M NS ++ Positive
    65 RL 56/M NS + Negative
    67 ND 13/F MC +++ Positive
    69 RL 34/F NS + Negative
    70 ND 24/M NS ++ Negative
    71 RL 31/F NS + Negative
    72 ND 26/M NS - Negative
    74 ND 34/M NS ++ Negative
    76 ND 23/M NS + Negative
    77 ND 36/M NS ++ Negative
    78 ND 21/M NS - Negative
    79 ND 26/M NS +++ Positive
    80 ND 34/M LR +++ Negative
    81 ND 28/M NS ++ Positive
    82 ND 36/F NS ++ Negative
    83 ND 29/F NS ++ Negative
    84 ND 23/F NS ++ Positive
    85 ND 59/M NS +++ Negative
    86 ND 20/M NLP ++ Negative
    88 ND 19/F NS - Negative
    • There was a significant correlation between percent LAG-3 staining and MC and/or LR histology versus NS (r = 0.555, P < .001) and between LAG-3 and positive EBV status (r = 0.31, P = .036).

      ND indicates newly diagnosed; MC, mixed cellularity; NS, nodular sclerosing); RL, relapsed; Remission, long-term remission; NLP, nodular lymphocyte predominant; and LR, lymphocyte rich.

    • * Percentage of non-HRS cells within HRS-rich areas that are positive for LAG-3; +++ indicates 74% to 50%; ++, 49% to 25%; +, 24% to 10%; -, less than 10%.

  • Table 3.

    Ex vivo longitudinal profiling of EBV-specific T-cell responses in HL patients and healthy individuals using ELISPOT assays

    EBV antigen, patient groupNo. of assaysMean SFCs (SE)*Median SFCs (IQR)*P vs pretherapyP vs 6 moP vs 12 moP vs long-term remission
    LMP1/2A
        ND/RL pretherapy 87 30 (8) 0 (0-24)
        ND/RL 6 mo 45 87 (24) 0 (0-106) .091
        ND/RL 12 mo 41 101 (25) 27 (0-169) .014 .494
        Long-term remission 136 153 (33) 30 (0-176) < .001 .07 .387
        Healthy 41 102 (40) 32 (0-94) .002 .329 .996 .442
    EBNA3/4/6
        ND/RL pretherapy 33 221 (77) 58 (0-187)
        ND/RL 6 mo 21 149 (42) 60 (40-178) .651
        ND/RL 12 mo 24 351 (93) 135 (8-135) .190 .311
        Long-term remission 39 290 (74) 106 (14-106) .258 .582 .645
        Healthy 43 186 (54) 60 (0-195) .941 .621 .144 .247
    Lytic
        ND/RL pretherapy 16 235 (63) 175 (0-476)
        ND/RL 6 mo 10 389 (165) 128 (49-916) .6
        ND/RL 12 mo 13 631 (161) 506 (35-1187) .076 .515
        Long-term remission 23 304 (78) 194 (80-294) .568 .71 .122
        Healthy 6 245 (51) 254 (116-367) .606 .492 .237 .536
    • SE indicates standard error; IQR, interquartile range; —, not applicable.

    • * Arithmetic mean and median values of γ-interferon—producing cells (referred to as spot-forming cells [SFCs] per 106 PBMCs) in response to HLA class 1—restricted LMP1/2, EBNA3/4/6, and lytic antigen peptide epitopes.

    • Significant P value.