Prognostic significance of activated cytotoxic T-lymphocytes in primary nodal diffuse large B-cell lymphomas


Clinical outcome in diffuse large B-cell lymphoma (DLBCL) remains unpredictable, despite the identification of clinical prognostic parameters. Here, we investigated in pretreatment biopsies of 70 patients with DLBCL whether numbers of activated cytotoxic T-lymphocytes (CTLs), as determined by the percentage of CD3-positive lymphocytes with granzyme B (GrB) expression, have similar prognostic value as found earlier in Hodgkin's lymphoma and anaplastic large-cell lymphoma and whether loss of major histocompatibility complex (MHC)-I molecules or expression of the GrB antagonist protease inhibitor 9 (PI9) may explain immune escape from CTL-mediated cell death. Independent of the International Prognostic Index (IPI), the presence of 15% activated CTLs was strongly associated with failure to reach complete remission, with a poor progression-free and overall survival time. Downregulation of MHC-I light- and/or heavy-chain expression was found in 41% of interpretable cases and in 19 of 56 interpretable cases PI9 expression was detected. We conclude that a high percentage of activated CTLs is a strong, IPI independent, indicator for an unfavorable clinical outcome in patients with primary nodal DLBCL. Although in part of DLBCL expression of PI9 and loss of MHC-I expression was found, providing a possible immune-escape mechanism in these cases, no correlation with clinical outcome was found.


Diffuse large B-cell lymphomas (DLBCLs), representing the most frequent type of adult non-Hodgkin lymphomas, are characterized by a marked degree of morphologic and clinical heterogeneity.1,2 Although a subset of DLBCL patients are cured with standard therapy, the majority will ultimately die of their disease.3 A clinical indicator of prognosis, the International Prognostic Index (IPI), has been constructed and successfully used to define prognostic subgroups in DLBCL.3 However, also the IPI fails to predict clinical outcome in a significant number of patients accurately, suggesting that this may also depend on intrinsic cellular mechanisms determining the sensitivity to therapy.

We have previously shown that the presence of high percentages of activated cytotoxic T-lymphocytes (CTLs), as determined by the percentage of CD3-positive lymphocytes with granzyme B (GrB) expression, in lymphoma biopsies of patients with Hodgkin's lymphoma (HL) and anaplastic large-cell lymphoma (ALCL)4,5 is strongly associated with a very poor prognosis. This was explained by postulating that in cases with many activated CTLs, the continuous immunogenic pressure selects for tumour cells that are best equipped to resist CTL-mediated killing. Similar to activated CTLs, chemotherapy induces cell death by the activation of the cell death-signalling cascade.6 Thus, the disruption of both of these pathways will not only result in resistance to CTL-induced cell death but also in decreased sensitivity to chemotherapy, thereby explaining poor clinical outcome in these patients.

CTL-mediated apoptosis is achieved by at least two pathways: (1) by ligation of CTL expressed Fas/CD95-ligand to Fas/CD95 on the target cell, resulting in apoptosis via the activation of caspase-8 and subsequently caspase-3, and (2) by release of cytotoxic granules containing among others perforin and GrB.7 GrB is able to activate caspase-3 directly8,9,10 or indirectly via the activation of caspase-9 through the cleavage of Bid11 and subsequent activation of caspase-3, again resulting in apoptosis. Thus, the complete disruption of CTL-induced apoptosis is expected to occur only if both apoptosis pathways are disrupted.

Major histocompatibility complex (MHC)-I molecules presenting the appropriate peptides are necessary to provide the specific signals for recognition and killing by CTLs.12,13 A principle mechanism that may render tumour cells resistant to CTL-mediated killing is the loss of MHC-I expression, as has been shown previously in part of DLBCL and other lymphomas.14,15,16,17 Another possible immune-escape mechanism is the expression of the recently identified GrB antagonist protease inhibitor 9 (PI9). PI9 efficiently inhibits GrB in vivo and in vitro, and PI9-transfected cells are protected from GrB-mediated apoptosis.18,19 We have recently demonstrated the expression of PI9 in neoplastic cells of various lymphoma types including 40 DLBCLs.20

Here, we investigated whether percentages of activated CTLs (ie the percentage of CD3-positive tumour infiltrating lymphocytes with GrB expression) in lymphoma biopsies are related to clinical outcome in patients with primary nodal DLBCL and whether these percentages are a useful prognostic marker, independent of IPI. Furthermore, we investigated whether loss of MHC-I expression or expression of PI9 in tumour cells may explain immune escape in cases harbouring many activated CTLs.

Material and methods


Consecutive patients with primary DLBCL (n=70) were selected from the files of the Comprehensive Cancer Center Amsterdam (diagnosed between 1986 and 1998) and the Department of Pathology of the University Hospital Nijmegen (diagnosed between 1991 and 1999), The Netherlands. Of the cases described in this study, 23 cases have been screened by us for PI9 in a previous study.20 Cases were reclassified according to the WHO classification.2 Cases were excluded when considered extra nodal, that is, presenting with localized stage IE or IIE disease.21 At first presentation, the extent of the disease was determined by physical examination, serum lactate dehydrogenase (LDH) concentration, full blood count, bone marrow aspirate and biopsy, chest X-ray and computed tomography of chest, abdomen and pelvis. Patient and tumour characteristics are summarized in Table 1. The study was approved by the institutional review board of the VU Medical Center. Informed consent was provided according to the Declaration of Helsinki.

Table 1 Patient characteristics in relation to percentages of GrB-positive CTLs


Immunohistochemical analysis was performed using the following antibodies: GrB-specific Moab GrB7 (Sanbio, Uden, The Netherlands);22 polyclonal anti-CD3 (DAKO); Moab HCA2, reactive with HLA-A locus products;23 Moab HC10 preferentially recognizing HLA-B/C locus products (kindly provided by Dr J Neefjes, NKI, Amsterdam, The Netherlands);23 polyclonal anti-β2-microglobulin (A072; DAKO); and the PI9 recognizing antibody PI9-17, developed in our laboratory.20

Stainings were performed as described previously.16,20 Briefly, formalin-fixed, paraffin-embedded 3-μm tissue sections were deparaffinized with xylene, endogenous peroxidase was blocked and sections were pretreated according to the appropriate protocol for the different antibodies (Table 2). Moreover, to identify the nature of the GrB-positive cells, double stainings were performed for GrB and CD8 or CD3, again as described previously.16

Table 2 Antibodies used in this study with used staining conditions

Antibodies were visualized using a standard three-step streptavidin–biotin complex method using diaminobenzidine as chromogen. Staining intensity was enhanced using the catalysed reported deposition method (Dako) for all antibodies except GrB and CD3.

Interpretation of immunohistochemical stainings

The quantification of CD3- and of GrB-positive lymphocytes in the reactive infiltrate was performed using a commercially available interactive video-overlay-based measuring system, giving absolute numbers of positive cells/mm2 (Q-PRODIT; Leica, Cambridge, UK), as described previously by us.4,24 Per tumour slide, 45–55 fields of vision were randomly selected using an automatic scanning stage. The percentage activated CTLs was determined by dividing the number of GrB-positive lymphocytes by the number of CD3-positive lymphocytes.

HCA2-, HC10- and β2-microglobulin-positive tumour cells were quantified as follows: cases were divided into four categories: less than 5, 5–50, 50–90 and more than 90%. Cases in which reactive lymphocytes were not clearly positive using the different anti-MHC antibodies were considered as not interpretable.

For PI9 expression dendritic cells, which were previously shown to express PI9,20 served as an internal positive control. DLBCL were considered either positive or negative irrespective of the number of positive cells.

Analysis of clinical data

For each patient the following characteristics were noted from the medical records: age at diagnosis, sex, Ann Arbor stage at presentation, the presence of B symptoms, erythrocyte sedimentation rate, serum LDH concentration, therapy, response, occurrence of relapses and cause of death. The IPI, comprising age, stage, number of extra nodal sites, LDH level and performance status, was also determined.3 The median follow-up time was 17 months (range 0–153 months). Survival time was measured from time of initial diagnosis until death related to DLBCL or until the end of follow-up. Patients who died of causes unrelated to the disease were censored at the time of death. Progression-free survival time was measured from time of initial diagnosis until time of disease relapse. Patients who did not enter complete remission were assigned a progression-free survival time of zero in the statistical analysis.

Statistical methods

Survival curves were constructed with the Kaplan–Meier method. Differences between the curves were analysed using the log-rank test. Multivariate analysis was performed using the Cox's proportional-hazards model.25 Qualitative variables were analysed by Pearson's χ2 test or by the Kruskal–Wallis test. The Mann–Whitney U-test was used to compare group means and the Spearman's test was used to test correlations between different variables. All P-values are based on two-tailed statistical analysis, considering P-values below 0.05 as significant. Analyses were performed using the SPSS statistical software (version 10.1 SPSS Inc., Chicago, IL, USA). The optimal cutoff value for the percentage of GrB-positive cells was determined by two methods: (1) using the log-rank test, by testing the prognostic value for each possible cutoff point (1, 2, 3%, etc for GrB); and (2) using Cox's regression analysis, including all cutoff points as categorical variables. Both methods gave identical results.


Patient characteristics

Patient characteristics are summarized in Table 1 and ranked according to the percentage of GrB-positive CTLs. Age distribution showed a peak, with a broad range, in the sixth decade as has been described in the literature.1,26 Over half of the patients (63%) presented with stage III or IV disease, showing either multiple organ involvement or bone marrow dissemination.

The majority of patients (n=63) received polychemotherapy, consisting of CHOP (cyclophosphamide, doxorubicin, vincristine, prednisone) regimens or variants, either alone (n=39) or in combination with involved field radiation (n=24), in four cases only involved field radiation was administered. Owing to old age and poor performance status three patients, all presenting with localized disease, were kept from therapy. In 41 of 70 cases (59%), complete remission was achieved; of these, however, 18 patients (44%) experienced a relapse, with a median progression-free survival time of 22.9 months (range 4–118 months).

DLBCL harbour varying percentages of activated CTLs

In all cases tested, GrB-poitive lymphocytes were found, displaying a characteristic granular cytoplasmic staining pattern (Figure 1a). Absolute numbers of GrB- and CD3-positive lymphocytes varied considerably between individual DLBCL cases. No correlation between numbers of CD3-positive lymphocytes and numbers of GrB-positive lymphocytes was observed (Spearman's test, R=0.07). In four cases tested, double-staining procedures showed that the large majority of GrB-positive lymphocytes were positive for CD3 (Figure 1b) and for CD8, and are thus activated CTLs (Figure 1c). When activated CTLs (CD3- and GrB-positive lymphocytes) were determined as the percentage of all CD3-positive lymphocytes, a range from 1 to 67% was observed, with a mean of 22% (Figure 2). Similar percentages of activated CTLs were observed in ALCL (mean 18%, range <1–61%),5 but in HL lower percentages were detected (mean 8%, range <1–38%).4

Figure 1

Detection and characterization of activated CTLs. (a) Biopsy specimen of a DLBCL patient with 15% activated CTLs, demonstrating granular cytoplasmic staining in brown. (b) Double staining for CD3 and GrB, the majority of activated CTLs express both CD3 (brown membranous staining) and GrB (black cytoplasmic staining). (c) Double staining for CD8 and GrB, the majority of activated CTLs express both CD8 (brown membranous staining) and GrB (black cytoplasmic staining).

Figure 2

Distribution of percentages activated CTLs among patients with primary nodal DLBCL. Histogram of the number of patients with different percentages of GrB-expressing CTLs.

High numbers of activated CTLs predict poor clinical outcome

The influence of percentage activated CTLs (ie the percentage CD3-positive lymphocytes with GrB expression) on overall survival time was estimated by Cox's regression analysis, with the percentage entered as a continuous variable. The prognosis declined with increasing percentages of activated CTLs (P=0.03). If patients were divided into a group with <15% and 15% GrB-positive lymphocytes (the threshold with the strongest discriminative power) the presence of 15% GrB-positive lymphocytes defined a group of patients with an unfavorable prognosis: 33 of 43 patients with 15% GrB-positive lymphocytes died during the follow-up period compared to 11 of 27 patients with <15% GrB-positive lymphocytes (log-rank test; P=0.0003) (Figure 3a). Moreover, 23 of 43 patients with 15% GrB-positive lymphocytes failed to reach complete remission in contrast to only six of 27 patients with <15% GrB-positive lymphocytes (log-rank test, P=0.01) (Table 1). Furthermore, the progression-free survival time was significantly shortened in patients with 15% GrB-positive lymphocytes as compared to patients with <15% GrB-positive lymphocytes (5-year progression-free survival 23%, respectively, 40%, log-rank test; P=0.002, data not shown). When entered as a continuous variable, absolute numbers of GrB-positive lymphocytes were also related to overall survival time (Cox's regression analysis, P=0.02), but not to progression-free survival time. The number of CD3-positive lymphocytes did not correlate with clinical outcome (Cox's regression analysis, P=0.7).

Figure 3

Comparison of overall survival time in primary nodal DLBCL. According to the percentage of activated CTLs (a); low and high IPI (b); and the percentage of activated CTLs combined with low (solid line) and high IPI (dotted line) (c).

Predictive value of the percentage activated CTLs is independent of the IPI

Consistent with the previous literature,3 we found that also in our group the IPI is a strong prognostic marker (log-rank test; P=0.02, see Figure 3b). None of the factors included in the IPI was independently related to percentages of activated CTLs. However, when stages 1 and 2 were considered as localized and stages 3 and 4 as advanced disease, a high percentage of activated CTLs tended to be more frequently observed in cases with advanced disease (Pearson's χ2 test; P=0.08, see Table 1). If percentages of activated CTLs and the IPI were entered as categorical variable in the Cox's proportional-hazards model for multivariate analysis, both indices remained independent, strong prognostic markers. Their independent prognostic value is further illustrated when the prognostic value of the percentage activated CTLs was determined by univariate survival analysis for low- and high-risk groups separately (P=0.0002, Figure 3c). The strongest predictive effect was observed in patients presenting with low (ie 1 and 2) IPI. In this group, 13 of 19 patients 15% activated CTLs died as compared to only three of 14 patients with few activated CTLs (log-rank test; P=0.006, data not shown).

Loss of MHC-I expression in part of DLBCL

For all tested markers, cases with less than 5% positive tumour cells were considered to be negative (Figure 4). In general, loss of both HLA-A and HLA-B/C was observed (P<0.001, data not shown). In addition, usually loss of MHC-I heavy-chain expression correlated with loss of β2-microglobulin expression (P<0.001, data not shown). Cases were considered MHC-I negative in the absence of β2-microglobulin expression or when both loss of HLA-A and HLA-B/C expression was observed. When β2-microglobulin expression was present, cases were considered MHC-I positive when either or both HLA-A and/or HLA-B/C expression was observed. Furthermore, the mean percentage of activated CTLs was significantly lower in MHC-I-negative cases (mean 14%, range <1–56%) as compared to MHC-I-positive cases (mean 26%, range <1–57%, P=0.004, Figure 5). Despite its correlation with numbers of activated CTLs, the expression of MHC-I on tumour cells did not correlate with clinical outcome, even if cutoff values other than 5% were tested (data not shown).

Figure 4

Expression of MHC-I heavy and light chains Left panel: biopsy specimens of the same patient with clear expression of both HLA-A and HLA-B/C and β2-microglobulin (a–c) is detected. Right panel: biopsy specimens of the same patient when low expression for both HLA-A and HLA-B/C and β2-microglobulin (d–f) is observed.

Figure 5

Box-plot depicting percentages of activated CTLs in relation to MHC-I expression. Box plots are based on the median, quartiles and extreme values. The box represents the interquartile range that contains 50% of the values. The thin lines and crossbars indicate the total range of values seen from the lowest to the highest. A line across the box indicates the median. The P-value is 0.004.

Expression of PI9 is detected in part of DLBCL

PI9 expression was observed as cytoplasmic staining in tumour cells in 21 of 56 interpretable cases (data not shown). When positive, the percentage of PI9-positive tumour cells ranged between 75 and 100%. No direct relation between PI9 expression and percentages of activated CTLs was found (Table 3). However, PI9 tended to be more frequently expressed in cases with intact MHC-I expression harbouring many activated CTLs (data not shown).

Table 3 Tumour characteristics in relation to percentages of GrB-positive CTLs

PI9 expression on tumour cells did not correlate with clinical outcome: 10 of 21 patients with PI9 expression died during the follow-up period as compared to 23 of 35 patients with no PI9 expression (log-rank test; P=0.3, data not shown). In addition, 15 of 21 patients with PI9 expression failed to reach complete remission as compared to 20 of 35 patients with no PI9 expression (log-rank test; P=0.2, data not shown).


In this study, we have demonstrated that the percentage of activated CTLs (ie GrB/CD3-positive lymphocytes taken as the percentage of all CD3-positive lymphocytes) is a strong and independent prognostic marker in patients with primary nodal DLBCL. Double-staining procedures revealed that in all cases tested, the large majority of GrB-positive cells are also CD3 and CD8 positive and should therefore be considered activated CTLs.

We found that the higher the percentage of activated CTLs, the lower the chance on reaching complete remission, the shorter the progression-free and overall survival time. Moreover, patients with 15% activated CTLs more frequently presented with B symptoms. These observations indicate that the presence of a strong CTL-mediated immune response is, at the time of disease presentation, not of benefit to the patient. Apparently, the tumour cells are relatively insensitive to therapy in these patients. This is especially illustrated by the fact that 23 of 44 patients with 15% activated CTLs did not reach complete remission, whereas only six of 26 patients with <15% activated CTLs did not. Assuming that activated CTLs have antitumour specificity, poor prognosis in cases with many activated CTLs can be explained by supposing that under the pressure of a strong CTL-mediated immune response, a selection for apoptosis-resistant tumour cells occurs. Since most chemotherapeutic drugs used in the treatment of DLBCL depend upon the induction of apoptosis,27 this acquired resistance for CTL-mediated apoptosis results in crossresistance to therapy-induced apoptosis as has been observed in many in vitro models and can be explained by the fact that a common cell death signalling pathway is used.28,29,30,31 This mechanism has previously been postulated for HL and ALCL in which similar results were obtained.5,6 In a preliminary report Felgar et al described a similar, although not statistically significant, prognostic effect of the percentage activated CTLs in T-cell rich B-cell lymphomas (TCRBCL), showing a 5-year overall survival time of 78% in 10 cases expressing <15% activated CTLs vs 43% in seven cases with 15% activated CTLs (see comment on Felgar et al32). They also conclude that tumour surveillance may be active in TCRBCL, but that the ability of CTLs to destroy tumour target cells may be impaired.

Apart from the above-mentioned interference with cell death signalling pathways, failure of CTLs to induce tumour cell lysis may result from several other mechanisms. Interference with the antigen presentation pathway by neoplastic cells may be one of these, since the downregulation of MHC-I proteins was observed in a significant minority of cases, consistent with a previous report by Riemersma et al.14 The number of activated CTLs was significantly higher in MHC-I-positive cases as compared to MHC-I-negative cases (Figure 5, Table 3), suggesting that the number of activated CTLs depends on an intact MHC-I-dependent antigen presentation and that the presence of few activated CTLs can be due to loss of MHC-I. However, in eight cases with many activated CTLs loss of MHC-I expression was observed. Possibly, these CTLs are activated via MHC-II-mediated presentation of tumour-derived antigens by dendritic cells,33,34 but by loss of MHC-I expression the tumour cells escape lysis by these activated CTLs.

Despite this significant correlation between the expression of MHC-I and numbers of activated CTLs, MHC-I expression did not have any prognostic value. This can be explained by the large variation in numbers of activated CTLs in both MHC-I-negative and -positive cases (Figure 5).

As described previously by us,20 a possible immune-escape mechanism in DLBCL is the expression of the GrB antagonist PI9. As PI9 is never expressed in normal B cells, the observed expression in 38% of cases supports the notion that in these cases a selective outgrowth of tumour cells occurs that can inhibit GrB-mediated CTL-induced apoptosis. The expression of PI9 in neoplastic B cells will be advantageous particularly in cases with an intact antigen presentation. Consistent with this notion, the expression of PI9 was found most frequently in cases with intact MHC-I expression harbouring many activated CTLs.

Alternative immune-escape mechanisms are the induction of local T-cell anergy by the expression of certain immune suppressive cytokines and the disruption of intracellular CTL-inducible cell death signalling pathways. As indicated above, we think that this last immune-escape mechanism will result in crossresistance to chemotherapy-induced cell death.

Using a combination of IPI and percentage activated CTLs as prognostic markers, it is possible to identify a group of patients accurately with fatal outcome within 2 years (IPI high, 15% activated CTLs) and another group of patients with expected favourable outcome (IPI low, <15% activated CTLs). Still, even in the group with a relatively few activated CTLs, a considerable number of patients died of disease within 2 years, with six patients not reaching complete remission. Possibly, these lymphomas have acquired resistance to chemotherapy unrelated to a CTL-mediated immune pressure. At present, we are investigating whether testing the functionality of both identified apoptosis pathways can more accurately predict clinical outcome in individual cases.

We conclude that the presence of many activated CTLs in tumour biopsies of DLBCL patients is a strong, IPI independent, marker for an unfavourable clinical outcome and is particularly helpful in identifying patients with poor prognosis while presenting with low IPI. These data suggest that in DLBCL patients with poor prognosis, tumour cells have become resistant to CTL-induced apoptosis resulting in resistance to chemotherapy-induced cell death. Although in part of DLBCL expression of PI9 and loss of MHC-I expression was found, providing a possible immune-escape mechanism in these cases, no correlation with clinical outcome was found.


  1. 1

    A clinical evaluation of the International Lymphoma Study Group classification of non-Hodgkin's lymphoma. The Non-Hodgkin's Lymphoma Classification Project. Blood 1997; 89: 3909–3918.

  2. 2

    Jaffe ES, Harris NL, Stein H, Vardiman JW . Pathology and Genetics of Tumors of Haematopoietic and Lymphoid Tissues. World Health Organisation Classification of Tumors. Lyon: IARC Press, 2001.

  3. 3

    The International Non-Hodgkin's Lymphomas Prognostic Factors Project. A predictive model for aggressive non-Hodgkin's lymphoma. N Engl J Med 1993; 329: 987–994.

  4. 4

    Oudejans JJ, Jiwa NM, Kummer JA, Ossenkoppele GJ, van Heerde P, Baars JW et al. Activated cytotoxic T cells as prognostic marker in Hodgkin's disease. Blood 1997; 89: 1376–1382.

  5. 5

    Ten Berge RL, Dukers DF, Oudejans JJ, Pulford K, Ossenkoppele GJ, de Jong D et al. Adverse effects of activated cytotoxic T lymphocytes on the clinical outcome of nodal anaplastic large cell lymphoma. Blood 1999; 93: 2688–2696.

  6. 6

    Johnstone RW, Ruefli AA, Lowe SW . Apoptosis: a link between cancer genetics and chemotherapy. Cell 2002; 108: 153–164.

  7. 7

    Barry M, Bleackley RC . Cytotoxic T lymphocytes: all roads lead to death. Nat Rev Immunol 2002; 2: 401–409.

  8. 8

    Darmon AJ, Nicholson DW, Bleackley RC . Activation of the apoptotic protease CPP32 by cytotoxic T-cell-derived granzyme B. Nature 1995; 377: 446–448.

  9. 9

    Darmon AJ, Ley TJ, Nicholson DW, Bleackley RC . Cleavage of CPP32 by granzyme B represents a critical role for granzyme B in the induction of target cell DNA fragmentation. J Biol Chem 1996; 271: 27109–27112.

  10. 10

    Andrade F, Roy S, Nicholson D, Thornberry N, Rosen A, Casciola-Rosen L . Granzyme B directly and efficiently cleaves several downstream caspase substrates: implications for CTL-induced apoptosis. Immunity 1998; 8: 451–460.

  11. 11

    Heibein JA, Goping IS, Barry M, Pinkoski MJ, Shore GC, Green DR et al. Granzyme B-mediated cytochrome c release is regulated by the Bcl-2 family members Bid en Bax. J Exp Med 2000; 192: 1391–1401.

  12. 12

    Monaco JJ . A molecular model MHC class-I-restricted antigen presentation. Immunol Today 1992; 13: 173–179.

  13. 13

    Rees C, Mian S . Selective MHC expression in tumors modulates adaptive and innate antitumor responses. Cancer Immunol Immunother 1999; 48: 374–381.

  14. 14

    Riemersma SA, Jordanova ES, Schop RFJ, Philippo K, Looijenga LHJ, Schuuring E et al. Extensive genetic alterations of the HLA region, including homozygous deletions of HLA class II genes in B-cell lymphomas arising in immune-privileged sites. Blood 2000; 96: 3569–3577.

  15. 15

    Frisan T, Zhang QJ, Levitskaya J, Coram M, Kurilla MG, Masucci MG . Defective presentation of MHC class I-restricted cytotoxic T-cell epitopes in Burkitt's lymphoma cells. Int J Cancer 1996; 68: 251–258.

  16. 16

    Oudejans JJ, Jiwa NM, Kummer JA, Horstman A, Vos W, Baak JP et al. Analysis of major histocompatibility complex class I expression in Reed–Sternberg cells in relation to the cytotoxic T-cell response in Epstein–Barr virus positive and negative Hodgkin's disease. Blood 1996; 87: 38–44.

  17. 17

    Murray PG, Constandinou CM, Crocker J, Young LS, Ambinder RF . Analysis of major histocompatibility complex class I, TAP expression, and LMP2 epitope sequence in Epstein–Barr virus-positive Hodgkin's disease. Blood 1998; 92: 2477–2483.

  18. 18

    Bird CH, Sutton VR, Sun J, Hirst CE, Novak A, Kumar S et al. Selective regulation of apoptosis: the cytotoxic lymphocyte serpin proteinase inhibitor 9 protects against granzyme B-mediated apoptosis without perturbing the Fas cell death pathway. Mol Cell Biol 1998; 18: 6387–6398.

  19. 19

    Sun J, Bird CH, Sutton V, Coughlin PB, De Jong TA, Trapani JA et al. A cytosolic granzyme B inhibitor related to the viral apoptotic regulator cytokine response modifier A is present in cytotoxic lymphocytes. J Biol Chem 1996; 271: 27802–27809.

  20. 20

    Bladergroen BA, Meijer CJLM, Ten Berge RL, Hack CE, Muris JJF, Dukers DF et al. Expression of the granzyme B inhibitor, protease inhibitor 9, by tumor cells in patients with Hodgkin's and non-Hodgkin's lymphoma: a novel protective mechanism for tumor cells to circumvent the immune system? Blood 2002; 99: 232–237.

  21. 21

    Gospodarowicz MK, Sutcliffe SB . The extranodal lymphomas. Semin Radiat Oncol 1995; 4: 281–300.

  22. 22

    Kummer JA, Kamp AM, van Katwijk M, Brakenhoff JP, Radosevic K, van Leeuwen AM et al. Production and characterization of monoclonal antibodies raised against recombinant human granzymes A and B and showing cross reactions with the natural proteins. J Immunol Methods 1993; 163: 77–83.

  23. 23

    Stam NJ, Vroom TM, Peters PJ, Pastoors EB, Ploegh HL . HLA-A and HLA-B-specific monoclonal antibodies reactive with free heavy chains in Western blots, in formalin fixed, paraffin embedded tissue sections and in cryo-immuno-electron microscopy. Int Immunol 1990; 2: 113–125.

  24. 24

    Polkowski W, Meijer GA, Baak JP, ten Kate FJ, Obertop H, Offerhaus GJ et al. Reproducibility of p53 and Ki-67 immunoquantitation in Barrett's esophagus. Anal Quant Cytol Histol 1997; 19: 246–254.

  25. 25

    Cox DR . Regression models and life tables. J R Stat Soc Br 34; 187: 1972.

  26. 26

    Armitage JO . Weisenberger DD for the non-Hodgkin's lymphoma classification project. New approach to classifying non-Hodgkin's lymphomas. Clinical features of the major histologic subtypes. J Clin Oncol 1998; 16: 2780–2795.

  27. 27

    Schmitt CA, Lowe SW . Apoptosis and therapy. J Pathol 1999, 127–137.

  28. 28

    Friesen C, Herr I, Krammer PH, Debatin KM . Involvement of the CD95 (APO-1/FAS) receptor/ligand system in drug-induced apoptosis in leukemia cells. Nat Med 1996; 2: 574–577.

  29. 29

    Los M, Herr I, Friesen C, Fulda S, Schulze-Osthoff K, Debatin KM . Cross-resistance of CD95- and drug-induced apoptosis as a consequence of deficient activation of caspases 9 ICE/Ced-3 proteases). Blood 1997; 90: 3118–3129.

  30. 30

    Friesen C, Fulda S, Debatin KM . Induction of CD95 ligand and apoptosis by doxirubicin is modulated by the redox state in chemosensitive- and drug-resistant tumor cells. Cell Death Differ 1999; 6: 471–480.

  31. 31

    Boesen-de Cock JG, Tepper AD, de Vries E, van Blitterswijk WJ, Borst J . Common regulation of apoptosis signaling induced by CD95 and the DNA-damaging stimuli etoposide and gamma-radiation downstream from caspase-8 activation. J Biol Chem 1999; 274: 14255–14261.

  32. 32

    Felgar RE, Greer JP, Macon W . T-cell-rich large-B-cell lymphoma (TCRBCL) as compared to diffuse large-B-cell lymphoma [letter]. Am J Pathol 1998; 153: 1707–1715.

  33. 33

    Wolkers MC, Stoetter G, Vyth-Dreese FA, Schumacher TNM . Redundancy of direct priming and cross-priming in tumor specific CD8+ T cell responses. J Immunol 2001; 67: 3577–3584.

  34. 34

    Heath WR, Carbone FR . Cross-presentation, dendritic cells, tolerance and immunity [review]. Annu Rev Immunol 2001; 19: 47–64.

Download references

Author information

Correspondence to J J Oudejans.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Muris, J., Meijer, C., Cillessen, S. et al. Prognostic significance of activated cytotoxic T-lymphocytes in primary nodal diffuse large B-cell lymphomas. Leukemia 18, 589–596 (2004).

Download citation


  • B-cell lymphomas
  • granzyme B
  • prognosis
  • MHC-I
  • PI9

Further reading