Stimulation of T-cell cytokine production and NK-cell

The Hematology Journal (2003) 4, 336–341 All rights reserved 1466-4680/03 $25.00

& 2003 The European Hematology Association

www.nature.com/thj

Stimulation of T-cell cytokine production and NK-cell function by IL-2, IFN-a and histamine treatment during remission of non-Hodgkin’s lymphoma Richard A˚hlberg1,4, Barbara MacNamara2,4, Margareta Andersson1,4, Chengyun Zheng1, Anna Svensson3, Go¨ran Holm1, Mona Hansson3, Anna Porwit-MacDonald2, Magnus Bjo¨rkholm1 and Anne Sundblad*,1 1 Division of Hematology, Department of Medicine, Karolinska Hospital and Institutet, Stockholm, Sweden; 2Division of Pathology, Department of Oncology and Pathology, Karolinska Hospital and Institutet, Stockholm, Sweden; 3Division of Laboratory Medicine, Department of Clinical Immunology and Transfusion Medicine, Karolinska Hospital and Institutet, Stockholm, Sweden

Limited therapeutic options remain for patients with relapsing lymphoma following chemotherapy and autologous stem cell transplantation (ASCT), hence motivating investigations of complementary treatments. The aim of the present study was to evaluate feasibility and immunological effects of an immunotherapy schedule administered during chemotherapyinduced remission of aggressive non-Hodgkin´s lymphoma (NHL). Repeated cycles of rIL-2, rIFN-a and histamine were administered to a patient with a grade III follicle center cell lymphoma, following relapse and high-dose chemotherapy with stem cell support. T-cell cytokine production and repertoire alterations were monitored by flow cytometry together with assessment of natural killer (NK) cell-mediated cytotoxicity. The treatment schedule induced significant increases in frequencies of CD4 þ T cells expressing intracellular IFN-g or IL-4, thus a T helper (Th) 1 and Th 2 type of response were observed. CD8 þ T cells showed enhancement mainly of TNF-a production. Such induction of T-cell effector functions was accompanied by an augmentation of NK-cell cytotoxicity and a pronounced reduction of possibly regulatory CD57 expressing lymphocytes. The results indicate synergistic T- and NK-cell activation by tolerable doses of the combined immunotherapy, administered during remission after chemotherapy and ASCT in NHL. The Hematology Journal (2003) 4, 336–341. doi:10.1038/sj.thj.6200320 Keywords:

IL-2; IFN-a; histamine; immunotherapy; non-Hodgkin’s lymphoma

Introduction High-dose chemotherapy (HDT) in conjunction with autologous stem cell transplantation (ASCT) is increasingly implemented in the treatment of lymphoma.1 The results of HDT and ASCT in progressing/relapsing follicular non-Hodgkin’s lymphoma (NHL) after initial chemotherapy treatment show high response rates; however, the majority of patients relapse.2 As potential complementary treatments, immunotherapy strategies could be considered. Such approaches aim at inducing or enhancing immune-mediated effector mechanisms against residual tumor cells after treatment, however, within the limitations imposed by the impact of prior chemotherapy on the immune system. It is thus of *Correspondence: A Sundblad, Centre for Molecular Medicine, L8:03, Karolinska Hospital, SE-171 76 Stockholm, Sweden; Tel: 468 51771616; Fax: 468 51773054; E-mail: [email protected] 4 The first three authors contributed equally. Received 19 January 2003; accepted 24 June 2003

importance to investigate quantitative as well as qualitative aspects of lymphocyte responses upon immunostimulation in candidate cases with a condition of treatment-induced immunosuppression, which was the specific aim of the present study. Cytokines that have been implicated in antitumor activity in NHL include recombinant IL-23 (rIL-2) and interferon-a (rIFN-a). IL-2 may activate both natural killer (NK)- and T-cell mediated cytotoxicity. NK-cell activity can also be enhanced by IFN-a, a cytokine that has been shown to be involved in antitumor response in NHL of B- and T-cell origin.4–7 Synergistic effects of IL-2 and IFN-a on NK cytolytic function have been demonstrated both in vitro and in vivo in animal models. Clinical trials indicate that combinatory immunotherapy with rIL-2 and rIFN-a may prolong remission in high-grade NHL following ASCT.8 Histamine has been shown to further potentiate IL-2- and IFN-a-mediated antitumor effects in vitro and in experimental tumor models9,10 which may be due to that histamine enhances

Cytokine stimulation of T- and NK-cells in NHL ˚ hlberg et al RA

337

NK- and T-cell cytotoxic activity by inhibiting the generation of reactive oxygen species from monocytes.9,11 Treatment trials based on combinations of histamine, rIL-2 and/or rIFN-a have previously been performed with promising results in AML12 and melanoma.13 The potential of histamine to potentiate actions of IL-2 and IFN-a could be of interest also in NHL, since it would allow for dose reductions of the administered cytokines causing frequent toxicity in the doses that have been utilized in reported trials. The present study describes specific immunological effects induced in a previously chemotherapy-treated NHL patient in remission, by a combination treatment of histamine, rIL-2 and rIFN-a administered within dose ranges that should be well tolerated according to previous studies.

Materials and methods Patient and treatments A 56-year-old Caucasian male, diagnosed 2 years prior to the study with a grade II follicle center cell lymphoma with bone marrow and nodal abdominal involvement was selected. The patient received initially eight courses of CHOP (cyclophosphamide, vincristine, doxorubicin, prednisone) and reached partial remission of abdominal lymphoma manifestations and complete clearance of previous bone marrow involvement. After 6 months, CT showed signs of abdominal tumor progression, and aspiration cytology of an enlarged lymph node in the neck revealed a transformation to a grade III follicle center cell lymphoma. After two courses of DexaBeam, the patient received HDT (cyclophosphamide and etoposide), total-body irradiation and autologous stem cell support.2 At 8 months after HDT, only minor enlargements of abdominal lymph nodes were remaining and the patient had recuperated well including blood chemistry values with the exception of a lymphopenia (lymphocyte counts fluctuating around 1 109 cells/l). The patient received at this time three consecutive cycles of immunotherapy consisting of: three million units (MU) rIFN-a (Schering-Plough, Brinny, Ireland), 1.5 mg/kg rIL-2 (Chiron, Amsterdam, Netherlands)

1

Cycle n˚ Histamine rIFN-α rIl-2

Weeks

0

*PBMC

*

■

■ ●

■ ■ ●

1

2

3

*

■ ■ ●

■ ■ ●

■ ■

■

4

5

6

7

*

2 ■ ■ ●

■ ■ ●

■ ■ ●

■ ■

3 ■ ■ ●

Follow-up

■ ■ ●

■ ■

… 14 15 16 17 18 19 20 21 22 23 … 28 42

*

* * *

■ Histamine

0.5 mg s.c. twice a day, 5 days a week

■ rIFN-α

3.0 MU s.c. once a day, 5 days a week

● rIL-2

1.5 µg/kg s.c. twice a day, 5 days a week

* *

Figure 1 Treatment and sampling schedule. The three cycles of the combined treatment are described. An asterix (*) denotes the time point (week number) in relation to the treatment cycles, when phenotype and functional assays of T- and NK-cells of were performed on freshly collected PBMCs.

and 0.5 mg histamine (Apoteket’s Production and Laboratory Units, Umea˚, Sweden), self-administered subcutaneously 1–2 times daily, 5 days a week (the schedule is shown in Figure 1). The combined immunotherapy was well tolerated, without any side effects or signs of toxicity. Informed consent was obtained from the patient, and the study was approved by the Karolinska hospital ethics committee.

Blood samples and cell separation Peripheral blood samples were repeatedly collected from the patient before, during and after treatments as denoted by an asterix (*) in Figure 1; before treatment (week 0), during cycle 1 (week 6), before (week 14) and during cycle 2 (weeks 17 and 18), between cycles 2 and 3 (week 19). The post-treatment observation period included weeks 28 and 42. Peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll (Pharmacia, Uppsala, Sweden). Healthy volunteers served as control blood donors.

Flow cytometry analysis of lymphocytes The following FITC, PE, PerCP or Tricolour labelled antibody (Ab) specificities were used for analyses of cell surface markers: CD3, CD4, CD8, CD56, CD57, CD69 (Becton Dickinson, San Jose, CA, USA), CD25 (DAKO, Glostrop, Denmark). Flurochrome-labelled isotype matched irrelevant Abs served as controls in all experiments. Cells (1 104) within the live gate per sample were analyzed by FACS Calibur flow cytometry using Cell Quest software (Becton Dickinson, Immunocytometry Systems, San Jose, CA., USA).

Intracellular cytokine expression The method was previously described.14 In brief, PBMCs were fixed in 1% paraformaldehyde, washed in PBS and permeabilized with FACS Perm (Becton Dickinson). Cells were stained with flurochrome labelled anti-CD4 or -CD8 Abs followed by staining with anticytokine Abs specific for IFN-g, TNF-a, IL-2 or IL-4 (Pharmingen, San Diego, CA, USA). A minimum of 2 104 events per sample were acquired for analyses of expression of intracellular cytokines in CD4 þ or CD8 þ T-cell subpopulations after gating on respective cell population markers and lymphoid light scatter characteristics using Cell Quest software (Becton Dickinson).

NK cytotoxicity assay A standard 51Cr release assay was used to measure NK-cell cytotoxicity.15 Briefly, 2 106 K562 target cells (human erythroleukemia cell line) labelled with chromium-51 (sodium chromate, Amersham, Solna, The Hematology Journal


338 45

CD8+ CD3+

40

CD56+

CD4+ CD3+ CD57+

35 % of lymphocytes

Sweden) were plated at 1 104 cells per well in 96-well plates. Effector cells, (E) consisting of PBMCs from the patient and control donors, were added at different E:T ratios. After incubation at 371C for 4 h, supernatants were analyzed for radioactivity by a g-counter. Spontaneous release of 51Cr was determined by incubating target cells with medium, and maximum release was obtained by lysis of target cells in 1% Triton X-100. The percentage of specific 51Cr release was calculated according to the formula: % specific lysis ¼ 100 (experimental releasespontaneous release)/(maximum releasespontaneous release).

30 25 20 15 10 5 0 0

Lymphocyte counts in peripheral blood collected from the patient before, during and after cessation of immunotherapy, remained relatively constant and low during the first treatment cycles (approximated mean value 1.1 109/l; normal range, n: 1.2–4.5 109/l). The number of lymphocytes increased at two time points towards 1.5 109/l during and after the last cycle (weeks 19 and 42). Monocyte counts remained around 0.5 109/l (n: 0.1–1.0 109/l), while granulocytes fluctuated during treatment with an overall slight increase from pretreatment values of 2.0 109/l (n: 1.8–8 109/l) to approximately 2.3 109/l (data not shown).

6

14

17 18 Week

19

28

42

Figure 2 Variations of T- and NK-cells during treatment. Flow cytometry analyses of frequency of lymphocytes expressing T- and NK-cell markers before (week 0), during and after (weeks 28, 42) the treatment cycles (see also Figure 1).

1.6 1.4 Patient/Control ratio

Results

3

1.2 1 0.8 0.6 0.4 0.2

Peripheral T- and NK-cell distributions Flow cytometry analyses of lymphocytes showed fluctuations of the fraction of CD8 þ T cells (CD8 þ , CD3 þ , n: 19–34% of lymphocytes) from 33% pretreatment (mean value of two independent measurements), which increased during first cycle and decreased after the second cycle to 21.5%. measured 19 weeks after cessation of treatment (Figure 2). The frequency of CD4 þ T cells (CD4 þ , CD3 þ , n: 34–51%) was subnormal prior to treatment and remained so with limited fluctuations during the observation period (Figure 2). The frequency of NK cells (CD56 þ , CD3) varied during the course of treatment (Figure 2). The pretreatment percentage of CD56 þ lymphocytes was 11.5%, which increased to approximately 20% during the first cycle of treatment, then decreased to 12% prior to the initiation of cycle 2 (week 15), during which the NK frequency regained a level of approximately 20%. Post-treatment NK-cell frequencies had decreased to pretreatment values (Figure 2), thus indicating a treatment-dependent stimulation of NK cells. In contrast to the CD56 þ NK cells, CD57 þ lymphocytes (NK- and T-cell subsets) showed a pronounced decrease (Figure 2). The abnormally high pretreatment value of 40% CD57 þ cells in the patient’s PBMCs (n: 7–12%) decreased to approximately 20% after cycle 1, and remained at that level during the observation period (Figure 2). The Hematology Journal

0 0

6

14

17 18 Week

19

28

42

Figure 3

Enhancement of NK-cell cytotoxicity. NK-cell cytotoxicity of the patients PBMCs were assessed before (week 0), during and after the treatment (weeks 28, 42) in comparison with the NK-cellcytotoxicity of PBMCs from normal donors, which served as a control in each experiment. The NK-cell-mediated specific lysis of patients NK cells are expressed in relation to the control lysis: patient/control ratio ¼ % specific lysis of patient cells/% specific lysis of normal cells.

NK-cell cytotoxicity NK cytotoxicity was evaluated by a standard assay15 compairing cytolytic activity of the patient’s NK cells to that of NK cells from healthy donors (Figure 3). The pretreatment NK-cell-specific cytotoxicity level of the patient scored less than 20% of control values, which increased to approximately 60% during treatment cycle 1. Further normalization of the NK-cell cytotoxicity was measured during the remaining treatment cycles, and values close to or above controls was obtained after cycle 2, which remained during the post-treatment observation period (weeks 28 and 42, Figure 3).

T-cell intracellular cytokine expression A three color flow cytometry technique14 was used to quantify the proportion of CD4 þ and CD8 þ T cells


339

expressing intracellular IFN-g, TNF-a, IL-2 and IL-4.15 Measurements of CD4 þ or CD8 þ T cells expressing each cytokine in unstimulated PBMCs from eight healthy donors (mean value) served as controls (Figure 4).

IFN-g With the exception of one sample obtained

responses of such immunotherapy in a previously chemotherapy-treated NHL patient in remission with signs of treatment-induced peripheral immunosuppression. The patient was initially diagnosed with a grade II follicular lymphoma, which transformed into an aggressive lymphoma 6 months after chemotherapy and was considered to have high-risk features. HDT, total body

during cycle 2, the PBMCs of the patient showed an overall increase in the proportion of CD4 þ T cells expressing IFN-g as compared to pretreatment and control values. A maximum of 9% CD4 þ cells positive for intracellular IFN-g (control mean value 1.7%) was observed in the interval between cycles 1 and 2 (week 14). Analyses of CD8 þ T cells showed a similar pattern of increased IFN-g expression, however, at a lower degree (Figure 4a).

a

12 CD4/IFNγ

10

CD8/IFNγ

8 6 4

IL-4 A five- to tenfold increase in the number of IL-4

2

expressing CD4 þ cells, as compared to pretreatment values, was recorded at four time points during treatment. The pattern was similar to the IFN-g expression, since the increase occurred after cessation of cycle 1 and was much more pronounced in the CD4 þ as compared to the CD8 þ T-cells (Figure 4b).

0 0

b

18

19

28

CON CD4/IL-4 CD8/IL-4

8 6

IL-2 In contrast to the other cytokines analyzed, the

0

4 2

%

and CD8 cells expressing percentages of CD4 intracellular IL-2 (Figure 4d) were similar to pretreatment values and to the mean value of the healthy controls (o2% for each T-cell subset).

17

12

subset, reaching values between 7 and 9% of CD8 þ cells during and after treatment cycle 2. Pretreatment and control mean values were 2.5 and 3.4%, respectively (Figure 4c).

þ

14

10

TNF-a TNF-a expressing cells increased in the CD8 þ

þ

6

0

c

6

14

17

18

19

28

12

CON CD4/TNFα

10

CD8/TNFα

8

Discussion

6 4

The present report describes immunological effects of a combined treatment of low-dose rIL-2, rIFN-a and histamine to a patient with high-grade NHL. The study aimed at analyzing the feasibility and lymphocyte

2 0

d

0

6

14

17

18

19

28

CON

12 CD4/IL-2

Intracellular cytokine expression in CD4 þ and CD8 þ T cells. Flowcytometry analyses of intracellular cytokine expression before (week 0), during and after treatment (week 28). The frequency (%) of CD4 þ or CD8 þ cells positive for the respective cytokines among lymphocytes from the patient at different time points and the mean value of healthy control donors (CON) are plotted. (a) CD4 þ IFNg þ and CD8 þ IFNg þ cells, mean values and standard deviations (s.d.) of eight donors: 1.1870.95; 1.3871.38, respectively. (b) CD4 þ IL-4 þ and CD8 þ IL-4 þ cells, mean values and s.d. of seven donors: 1.7670.89; 1.7770.59, respectively. (c) CD4 þ TNFa þ and CD8 þ TNFa þ cells, mean values and s.d. of eight donors: 2.8671.91; 3.4871.41, respectively. (d) CD4 þ IL-2 þ and CD8 þ IL-2 þ cells, mean values and s.d. of eight donors: 1.1670.81; 1.3070.54.

Figure 4

10

CD8/IL-2

8 6 4 2 0 0

6

14

17 18 Week

19

28

CON

The Hematology Journal


340

irradiation and ASCT induced a remission as well as a longstanding immunosuppression as observed by peripheral lymphocyte analyses. At 8 months postASCT, prior to the initiation of immunomodulatory treatment, the disease remained in remission and the total numbers of lymphocytes, monocytes and granulocytes were subnormal or close to lower normal ranges, but the percentage of CD4 þ T cells was abnormally low. Furthermore, the NK-cell function was impaired as assessed by the degree of NK-cellmediated cytotoxicity, which was reduced to approximately 20% of control values before the initiation of treatment. The evaluation of lymphocyte functions during the immunotherapy showed a reconstitution of the NK-cell function, and improved cytotoxicity remained during the follow-up period (Figure 3). The frequency of CD56 þ NK cells varied during the treatment period (Figure 2), thus during cycles 1 and 2, the enhanced NK cytotoxicity as compared to pretreatment values, could be due to an increased frequency of effector cells. However, since the CD56 þ NK-cell frequency decreased to pretreatment values after cessation of treatment, while the NK-cell function remained normalized, the prolonged effects indicate at least a temporary restoration of the NK cytotoxicity function. The results are consistent with previous demonstrations of IL-2 as an activator of NK cells, both capable of enhancing the lytic activity of NK cells as well as expanding CD56 þ NK-cell populations in studies of patients undergoing rIL-2-based immunotherapies.16–18 NK cells can also be activated through the TNF-a pathway, thus the increase in CD8 þ T cells producing TNF-a as measured during treatment (Figure 4) may have contributed to the observed enhancement of NK activity. The CD57 surface glycoprotein (110 kDa), expressed on subsets of NK, CD4 þ and CD8 þ T cells, is proposed to play a role in cellular adhesion and leukocyte trafficking.19 CD8 þ , CD57 þ lymphocytes from bone marrow transplant patients may produce a soluble factor that can suppress cell-mediated cytotoxicity.20 Increased frequency of CD57 þ T cells have been observed in certain untreated malignancies including B-NHL patients.21 Similarly, prior to the treatment in the present study, the patient’s PBMC showed a significantly increased frequency of CD57 þ lymphocytes (Figure 2) of which 18% consisted of doublepositive CD57 þ , CD8 þ cells (n: 5%). The expanded CD57 þ population showed a pronounced reduction during the course of treatment (from approximately 40– 20%), which was partly due to a decrease of the CD57 þ , CD8 þ double-positive cells (data not shown). Since the reduction of the CD57 þ lymphocyte population was paralleled with enhancement of T- and NK-cell activities, the CD57 þ cells could hypothetically be implicated in immunoregulatory or immunosuppressive effects on other lymphocyte populations as previously suggested.20–22 The total numbers of CD8 þ T cells was reduced from approximately 0.36 to 0.31 109 cells/l (n: 0.33–0.82 109 cells/l) and remained low during the The Hematology Journal

post-treatment observation period, thus warranting a note of caution since these results may imply a risk of a treatment-induced aggravation of an a priori lymphopenic condition. The mechanism for such decay of CD8 þ T cells is unclear, lymphocyte activation-induced cell death might be a mechanism involved, since the results point to a pronounced treatment-dependent stimulation of lymphocytes. The numbers of CD4 þ T cells in peripheral blood remained subnormal during the study period, no further decrease in relative or absolute numbers was observed. The T lymphopenia was not accompanied by a measurable decrease in T-cell function, since the T-cell populations displayed pronounced inductions of effector function after initiation of the therapy, as assessed by intracellular cytokine expression (Figure 4). The flow cytometry analysis of T cells was performed directly on freshly isolated PBMCs without any prior in vitro activation thus the results are assumed to reflect a treatment-dependent in vivo activation of T lymphocytes. The overall cytokine response was most pronounced in the T helper (Th) cell subsets showing a high degree of inducible functional activity in the CD4 þ T-cell subset. The T-cell response to the treatment was not polarized, since it consisted both of type 1 and type 2 cytokines; however, with a tendency of a more longstanding effect on the Th1 type of response. A certain degree of selectivity of the T-cell response could be disclosed by analyses of the variable T-cell receptor repertoire (TcR Vb families) expressed during treatment; some TCR Vb families (eg Vb 2 and 3, data not shown) showed relative expansions during treatment which remained in the post-treatment observation period. Such results indicate a preactivation of certain T-cell clones in vivo based on their V-region specificity, a question that was not further addressed in the present study. The patient has now been followed for 3 years after the therapy, and remains in remission. Repeated CT of the abdomen have shown a further continuous reduction of the enlarged lymphoid tissue that remained after ASCT prior to the immunotherapy. This response could obviously be due to long-term effects of the prior HDT and ASCT, possibly potentiated by the combined immunotherapy, the causal relationship cannot be established here, but the results are encouraging for an extended investigation to address the question of treament specificity. In conclusion, the described combined immunotherapy induced significant increases in frequencies of cytokine producing T cells and an augmentation of NK-cell-mediated cytotoxicity. Such functional activation of lymphocytes, as observed in the present case might be due both to a direct cytokine effect on lymphocytes, as well as a diminished negative regulation by CD57 expressing lymphocytes, a population that was markedly reduced during the treatment. Within the limitations imposed by a risk of reducing CD8 þ T-cells, such synergism might be of importance for induction of immunological antitumor effects after chemotherapy in hematological malignancies.


341

Acknowledgements Supported by grants from the Swedish Cancer Society and the Torsten and Ragnar So¨derberg Foundation, Sweden. The

skillful CT evaluations by Associate Professor Hans Ohlse’n and the assistance of nurse Anette Larsson are gratefully acknowledged.

References 1 Salles G, Coiffier B. Autologous peripheral blood stem cell transplantation for non-Hodgkin’s lymphoma. Review. Baillieres Best Pract Res Clin Haematol 1999; 12: 151–169. 2 Hunault-Berger M, Ifrah N, Solal-Celigny P. Intensive therapies in follicular non-Hodgkin lymphomas. Review. Blood 2002; 15: 1141–1152. 3 Gisselbrecht C, Maraninchi D, Pico JL, Milpied N, Coiffier B, Divine M et al. Interleukin-2 treatment in lymphoma: a phase II multicenter study. Blood 1994; 83: 2081–2085. 4 Foon KA, Sherwin SA, Abrams PG, Longo DL, Fer MF, Stevenson HC et al. Treatment of advanced non-Hodgkin’s lymphoma with recombinant leukocyte A interferon. N Engl J Med 1984; 311: 1148–1152. 5 Leavitt RD, Ratanatharathorn V, Ozer H, Ultmann JE, Portlock C, Myers JW et al. Alfa-2b interferon in the treatment of Hodgkin’s disease and non-Hodgkin’s lymphoma. Semin Oncol 1987; 14: 18–23. 6 Smalley RV, Andersen JW, Hawkins MJ, Bhide V, O’Connell MJ, Oken MM et al. Interferon alfa combined with cytotoxic chemotherapy for patients with nonHodgkin’s lymphoma. N Engl J Med 1992; 327: 1336–1341. 7 Solal-Celigny P, Lepage E, Brousse N, Reyes F, Haioun C, Leporrier M et al. Recombinant interferon alfa-2b combined with a regimen containing doxorubicin in patients with advanced follicular lymphoma. Groupe d’Etude des Lymphomes de l’Adulte. N Engl J Med 1993; 329: 1608–1614. 8 Nagler A, Ackerstein A, Or R, Naparstek E, Slavin S. Immunotherapy with recombinant human interleukin-2 and recombinant interferon-alpha in lymphoma patients postautologous marrow or stem cell transplantation. Blood 1997; 89: 3951–3959. 9 Brune M, Hansson M, Mellqvist UH, Hermodsson S, Hellstrand K. NK-cell-mediated killing of AML blasts: role of histamine, monocytes and reactive oxygen metabolites. Eur J Haematol 1996; 57: 312–319. 10 Asea A, Hermodsson S, Hellstrand K. Histaminergic regulation of natural killer cell-mediated clearance of tumour cells in mice. Scand J Immunol 1996; 43: 9–15. 11 Naredi P. Histamine as adjunct to immunotherapy. Semin Oncol 2002; 29: 31–34. 12 Brune M, Hellstrand K. Remission maintenance therapy with histamine and interleukin-2 in acute myelogenous leukaemia. Br J Haematol 1996; 92: 620–626.

13 Hellstrand K, Naredi P, Lindner P, Lundholm K, Rudenstam CM, Hermodsson S et al. Histamine in immunotherapy of advanced melanoma: a pilot study. Cancer Immunol Immunother 1994; 39: 416–419. 14 Jung T, Schauer U, Heusser C, Neumann C, Rieger C. Detection of intracellular cytokines by flow cytometry. J Immunol Methods 1993; 159: 197–207. 15 Krensky AM, Ault KA, Reiss CS, Strominger JL, Burakoff SJ. Generation of long-term human cytolytic cell lines with persistent natural killer activity. J Immunol 1982; 129: 1748–1751. 16 Bosly A, Guillaume T, Brice P, Humblet Y, Staquet P, Doyen C et al. Effects of escalating doses of recombinant human interleukin-2 in correcting functional T-cell defects following autologous bone marrow transplantation for lymphomas and solid tumours. Exp Hematol 1992; 20: 962–968. 17 Gottlieb DJ, Prentice HG, Mehta AB, Galazka AR, Heslop HE, Hoffbrand AV et al. Malignant plasma cells are sensitive to LAK cell lysis: pre-clinical and clinical studies of interleukin 2 in the treatment of multiple myeloma. Br J Haematol 1990; 75: 499–505. 18 Peest D, Leo R, Deicher H. Tumour-directed cytotoxicity in multiple myeloma – the basis for an experimental treatment approach with interleukin 2. Stem Cells 1995; 13: 72–76. 19 Needham LK, Schnaar RL. The HNK-1 reactive sulfoglucuronyl glycolipids are ligands for L-selectin and P-selectin but not E-selectin. Proc Nat Acad Sci 1993; 90: 1359–1363. 20 Autran B, Leblond V, Sadat-Sowti B, Lefranc E, Got P, Sutton L et al. A soluble factor released by CD8 þ CD57 þ lymphocytes from bone marrow transplanted patients inhibits cell-mediated cytolysis. Blood 1991; 77: 2237–2241. 21 Van den Hove LE, Van Gool SW, Vandenberghe P, Boogaerts MA, Ceuppens JL. CD57 þ /CD28 T cells in untreated hemato-oncological patients are expanded and display a Th1-type cytokine secretion profile, ex vivo cytolytic activity and enhanced tendency to apoptosis. Leukemia 1998; 12: 1573–1582. 22 Serrano D, Monteiro J, Allen SL, Kolitz J, Schulman P, Lichtman SM et al. Clonal expansion within the CD4 þ CD57 þ and CD8 þ CD57 þ T cell subsets in chronic lymphocytic leukemia. J Immunol 1997; 158: 1482–1489.

The Hematology Journal