FOXP3+ regulatory T cells in cutaneous T-cell lymphomas: association ...

4 downloads 0 Views 315KB Size Report
Aug 23, 2007 - mycosis fungoides (MF) and cutaneous T-cell lymphoma. (CTCL) unspecified were analysed for the expression of FOXP3 on tumour cells and ...
Leukemia (2007) 21, 2512–2518 & 2007 Nature Publishing Group All rights reserved 0887-6924/07 $30.00 www.nature.com/leu

ORIGINAL ARTICLE FOXP3 þ regulatory T cells in cutaneous T-cell lymphomas: association with disease stage and survival LM Gjerdrum1, A Woetmann2, N Odum2, CM Burton1, K Rossen3, GL Skovgaard4, LP Ryder5 and E Ralfkiaer1 1 Department of Pathology, Rigshospitalet, University of Copenhagen, Copenhagen, Denmark; 2Department of Medical Microbiology and Immunology, Institute of Molecular Biology, University of Copenhagen, Copenhagen, Denmark; 3Department of Pathology, Bispebjerg Hospital, Copenhagen, Denmark; 4Department of Dermatology, Bispebjerg Hospital, Copenhagen, Denmark and 5Tissue Typing Laboratory, Department of Immunology, University of Copenhagen, Copenhagen, Denmark

FOXP3 is a unique marker for CD4 þ CD25 þ regulatory T cells (Tregs). In solid tumours, high numbers of Tregs are associated with a poor prognosis. Knowledge about the implications of Tregs for the behaviour of haematological malignancies is limited. In this study, skin biopsies from 86 patients with mycosis fungoides (MF) and cutaneous T-cell lymphoma (CTCL) unspecified were analysed for the expression of FOXP3 on tumour cells and tumour-infiltrating Tregs. Labelling of above 10% of the neoplastic cells was seen in one case classified as an aggressive epidermotropic CD8 þ cytotoxic CTCL. In the remaining 85 cases, the atypical neoplastic infiltrate was either FOXP3 negative (n ¼ 80) or contained only very occasional weakly positive cells (n ¼ 5). By contrast, all biopsies showed varying numbers of strongly FOXP3 þ tumour-infiltrating Tregs. MF with early or infiltrated plaques had significantly higher numbers of FOXP3 þ Tregs than CTCL unspecified or advanced MF with tumours or transformation to large cell lymphoma. An analysis of all patients demonstrated that increasing numbers of FOXP3 þ Tregs were associated with improved survival in both MF and CTCL unspecified. In conclusion, our data indicate that the presence of FOXP3 þ Tregs in CTCL is associated with disease stage and patient survival. Leukemia (2007) 21, 2512–2518; doi:10.1038/sj.leu.2404913; published online 23 August 2007 Keywords: cutaneous T-cell lymphoma; FOXP3; immunohistochemistry; regulatory T cells; survival

Introduction FOXP3 of the forkhead box transcription factor family is now recognized as a unique marker for regulatory T cells (Tregs). The majority of these cells are CD4 þ CD25 þ T cells. They comprise 5–10% of the peripheral CD4 þ T cells in normal mice and humans. They respond to self-antigens, infectious agents and transplantation antigens, and are defined as a T-cell population that can influence other cell types with suppression of the immune response.1,2 In both humans and mice a lack of functional FOXP3 leads to fatal autoimmune, allergic and inflammatory diseases. In humans, a multi-organ autoimmune disease, known as the IPEX syndrome (an inheritable X-linked syndrome characterized by immune deregulation, polyendocrinopathy and enteropathy) has been associated with inherited defects in the FOXP3 gene.3 Recent evidence has shown that Treg-mediated immune suppression is also an important method for immune evasions of Correspondence: Dr LM Gjerdrum, Department of Pathology, Rigshospitalet, Copenhagen University Hospital, Frederik V’s vej 11, Copenhagen 2100, Denmark. E-mail: [email protected] Received 12 March 2007; revised 2 July 2007; accepted 24 July 2007; published online 23 August 2007

tumours and a significant obstruction for successful immunotherapy against cancer.4,5 In keeping with this, there have been several studies showing that increased numbers of Tregs are associated with poor prognosis and decreased survival in several types of carcinoma.6 More controversial results have been obtained in haematological malignancies. For example, in Hodgkin’s lymphoma, it has been shown that low infiltration of FOXP3 þ Tregs together with high numbers of TIA-1 þ cells is associated with poor outcome.7 Similarly, high numbers of infiltrating FOXP3 þ Tregs has been associated with improved survival in follicular lymphoma.8 These studies have suggested that tumour-infiltrating Tregs in lymphoma may act to suppress not only tumour-infiltrating cytotoxic cells, but also the malignant cells, and that the implications of Tregs for the behaviour of these conditions are therefore more complex than for solid tumours. Knowledge about FOXP3 expression in cutaneous T-cell lymphoma (CTCL) is limited. Using an in vitro approach, Berger et al.9 showed that CD4 þ CTCL cells could be induced to adopt a Treg phenotype and function upon prolonged cocultivation with autologous immature dendritic cells loaded with apoptotic tumour cells. Subsequent reports have indicated that the neoplastic cells in biopsies from CTCL are usually FOXP3 negative.10,11 Interesting, however, preliminary data suggest that the proportion of tumour-infiltrating Tregs in CTCL correlates inversely with the disease stage and tumour burden, suggesting that Tregs may have implications for the course and behaviour of these diseases.10 To address this hypothesis, we analysed skin biopsies from 86 CTCL patients by immunohistology for FOXP3 expression in tumour cells and tumour-infiltrating Tregs. The results show that high numbers of tumour-infiltrating FOXP3 þ Tregs are associated with superior survival in patients with both mycosis fungoides (MF) and CTCL unspecified. By contrast the neoplastic cells in these diseases are generally FOXP3 negative.

Materials and methods

Patients and tissue samples Biopsies from all patients diagnosed with CTCL during the period 1979–2004 were drawn from the archives of the Departments of Pathology at Rigshospitalet and Bispebjerg Hospitals. A total of 135 cases were retrieved. Following review of the histological samples and the clinical records, 86 cases with either MF (n ¼ 69) or CTCL unspecified (n ¼ 17) were selected for further analysis. The patients were 45 males and 28 females with a median age of 66 years (range 33–88 years) at diagnosis. Survival data could be retrieved for 54 (63%) of the

FOXP3 in CTCL LM Gjerdrum et al

2513 patients, including 9 of 10 patients with early MF plaques, 23 of 42 MF patients with infiltrated plaques, 6 of 10 cases with tumour MF, 6 of 7 patients with transformed MF and 9 of 17 cases with CTCL unspecified.

Histological review of biopsies and control tissues From all 86 cases, formalin-fixed, paraffin-embedded tissue samples were reviewed by histology and immunohistochemistry, using as a minimum CD3, CD4, CD8, CD30, CD56, TIA-1 and granzyme B. Classification was performed by two of the authors (LMG and ER) in accordance with the World Health Organization and European Organization for Research and Treatment of Cancer (2005). A total of 69 patients had MF and 17 were primary CTCL unspecified. Early plaque lesions were found in 10 patients with MF, 42 had infiltrated plaques, 10 patients had tumours and 7 patients had MF with transformation to large T-cell lymphomas. Five of these were positive for CD30. Of the 17 cases with unspecified CTCL, 2 cases had primary cutaneous small- to medium-sized CD4 þ CTCL, 2 patients had primary cutaneous gd T-cell lymphoma and 1 case was consistent with a primary cutaneous aggressive epidermotropic CD8 þ cytotoxic T-cell lymphoma. The remaining CTCL, unspecified were a heterogeneous group of neoplasms consisting of CD30-negative, medium to large, pleomorphic cells with varying expression of CD4 and CD8. Four were CD4 þ , CD8; three were CD4 þ , CD8 þ ; two were CD8 þ , CD4 and three cases were double negative. A multi-block, containing samples of reactive lymph nodes and tonsils was used as control.

Cell lines Cell lines derived from patients with MF, normal peripheral T lymphocytes from an MF patient and two T cell lines from healthy donors (a CD4 þ CD25 þ Treg cell line and a CD4 þ TH1-like cell line) were as described in detail elsewhere.12–15 The Jurkat cell line was transfected with haemagglutinin (HA)tagged FOXP3 or a vector control as described.16

Western blotting for FOXP3 Cells were rapidly pelleted and lysed in ice-cold lysis buffer (1% NP-40, 20 mM TRIS-HCl (pH 8.0), 140 mM NaCl, 10% glycerol, including the following inhibitors: 1 mM phenylmethylsulfonyl fluoride, 1 mM Na3VO4, 10 mM NaF, 1 mM iodioacetamide, 5 mM EDTA, and 7.5 mg ml1 aprotinin). To ensure equal loading, the total protein concentration of each sample was determined by Bio-Rad Protein Assay (#500–0001). Sodium dodecyl sulphate (SDS) sample buffer was added to the lysates, which were subsequently subjected to 10% SDS-polyacrylamide gel electrophoresis (PAGE) and transferred onto nitrocellulose membranes as previously described.16 Blots were evaluated by using enhanced chemiluminescence, stripped and reprobed according to the manufacturer’s manual (Amersham, Piscataway, NJ, USA).

Immunohistochemistry for FOXP3 Affinity-purified monoclonal mouse anti-human FOXP3 236A/ E7 was purchased from eBioscience (14-4777, San Diego, CA, USA). Following microwave heat-induced epitope retrieval, sections were incubated for 60 min at room temperature with a 1:50 dilution of FOXP3 and then stained using the DAKO Real EnVision Detection System, Peroxidase/DAB, Rabbit/Mouse (K5007, DakoCytomation, Glostrup, Denmark) by the Techmate 500 Immunostainer.

Scoring of FOXP3 þ tumour-infiltrating Tregs was performed by two of the authors (LMG and ER) by counting the numbers of positive cells in three representative areas at  40 magnification. The intensity was scored as weak, moderate or strong. FOXP3 labelling of the tumour cells was scored as negative (no visible staining or positive staining in o10%), moderately positive (positive staining in 10–50% of the tumour cells) and positive in a majority (450%) of the tumour cells. The intensity of FOXP3 þ tumour cells was estimated as for the Tregs.

Statistical analysis Statistical analysis was performed using SPSS 13.0 and SAS 9.1 software packages. Statistical significance was attributed to a Pvalue o0.05. Unless indicated otherwise, all continuous data are presented as median (range). As Klemke et al.10 suggested an inverse correlation between FOXP3 þ Treg and disease stage, our a priori hypothesis was that the presence of FOXP3 is associated with a favourable prognosis and survival. The data obtained from FOXP3 immunostains were compared between the different disease categories, using a linear mixed model (multi-analysis of variance (multi-ANOVA) for repeated measurements, where group is the fixed effect and patient is the random effect.17 This statistical model also allows for analysis of total variation contributed to by variations between patients in the same group (variance 1) and variations secondary to random effects (variance 2). Survival data were assessed by Kaplan– Meier method. Patients were considered at risk of death from the day of diagnosis. Patients were censored at the date of the last follow-up. Covariates were initially assessed by log-rank analyses. A Cox proportional hazards model was performed using the number of FOXP3 þ Tregs and morphological group entered as continuous and class variable, respectively. Hazard ratios (HR) are provided with 5–95% confidence intervals (CI). To reach adequate numbers of censoring events, the groups with MF tumour and transformed MF were combined in the Cox regression analysis.

Ethics The study was approved by the Local Ethics Committee (journal no. 01 284225) and the Danish Data Protection Agency (journal no. 2005-41–5930).

Results

FOXP3 in cell lines and control tissues In reactive tonsils and lymph nodes, occasional T lymphocytes in the interfollicular area and in the germinal centres were positive for FOXP3. The staining was always nuclear and very strong in intensity. The specificity of the anti-FOXP3 antibody was demonstrated in western blotting by comparing FOXP3 expression in Jurkat T cells transfected with FOXP3 or a vector control. As shown in Figure 1 (upper right), a high expression of FOXP3 was observed in Jurkat T cells transfected with FOXP3 but not in cells transfected with a vector control (Figure 1, upper left). The identity was confirmed by blotting with an anti-HA-tag antibody (Figure 1, middle). Western blotting with an antiextracellular signal-regulated kinase (ERK) antibody showed comparable levels of expression of the mitogen-activated protein kinases EKR1/2 in cells transfected with either FOXP3 or a vector control (Figure 1, lower right vs left), indicating that equal amounts of cell lysates had been loaded onto the gel. In addition to wild-type FOXP3, the antibody also recognized a Leukemia

FOXP3 in CTCL LM Gjerdrum et al

2514 low-molecular splice variant of FOXP3 in CD4 þ T cells with a potent suppressive capacity in vitro12 (data not shown). To address whether FOXP3 was expressed in cell lines derived from CTCL patients, MF2000 and MF1885 were assayed for FOXP3 expression using western blotting and immunohistochemistry, but none of the cell lines displayed FOXP3 expression (data not shown).

FOXP3 immunostaining in CTCL: labelling of the malignant cells All CTCL biopsies were suitable for immunohistochemical analyses. Of 86 tumour specimens, only one displayed positive staining for FOXP3 in above 10% of the lymphoma cells. In this case FOXP3 showed nuclear staining in 20% of the neoplastic

cells (Figure 2d). The intensity was moderate and hence weaker than in the tumour-infiltrating Tregs. In the remaining 85 cases, the atypical neoplastic infiltrate was either FOXP3 negative (n ¼ 80) or contained only very occasional weakly positive cells (n ¼ 5). The FOXP3 þ lymphoma sample originated from a 71-yearold female patient who presented with multiple, centrally ulcerated tumours on the trunk and extremities. Histologically, these lesions showed infiltrates of small- to medium-sized, pleomorphic cells with marked involvement of the epidermis and dense, band-like infiltrates in the upper dermis (Figure 2a). In the deeper part of the dermis, nodular infiltrates around the vessels were seen. By immunohistology, the lymphoma cells were positive for CD3, CD8, TIA-1 and granzyme B, but not for CD4, CD30 or CD56 (Figure 2a–c). These features were considered consistent with a diagnosis of an aggressive, epidermiotropic CD8 þ cytotoxic T-cell lymphoma.18 The clinical behaviour of this case was, as expected for this category of CTCL, aggressive with dissemination to lung and eye at 7 months and death occurring at 9 months after initial presentation.

FOXP3 immunostaining in CTCL: labelling of the tumour-infiltrating Tregs Figure 1 Anti-human FOXP3 specifically recognizes FOXP3. Jurkat cells were transfected with FOXP3 (right) or a vector control (left) before western blotting with anti-FOXP3 antibody (E-bioscience 236A/ E7) (upper lanes), an antihaemagglutinin (HA)-tag antibody (middle lanes) and an antibody against extracellular signal-regulated kinase (ERK)1/2 (lower lanes).

Results from the staining of Tregs in MF and CTCL unspecified are summarized in Table 1 and Figure 3. Representative examples are illustrated in Figure 4. As shown, the tumourinfiltrating Tregs displayed a very strong positivity for FOXP3 and only nuclei were stained. The highest median numbers of FOXP3 þ Tregs was seen in MF with infiltrated plaques (median 68), followed by CTCL unspecified (median 43) and MF with

Figure 2 Primary aggressive epidermotropic CD8 þ cytotoxic CTCL (a, HE), positive for CD8 (b) and FOXP3 (d), and negative for CD4 (c). Leukemia

FOXP3 in CTCL LM Gjerdrum et al

2515 Table 1

Number of patients and FOXP3+ Tregs according to CTCL category

Diagnosis MF I MF II MF tumour MF trans CTCL unspecified

Patients (n)

FOXP3+ Tregs, median (range)

10 42 10 7 17

37.5 68 23.5 6 43

Follow-up in months, median (range)

(27–136) (5–251) (2–117) (1–116) (1–197)

92 120 73.5 60 23

(8–324) (2–360) (48–120) (48–182) (2–59)

Abbreviations: CTCL, cutaneous T-cell lymphoma; n, number; MF I, MF with early patch/plaques; MF II, MF with infiltrated plaques; MF trans, MF with transformation to large T-cell lymphoma. The data are displayed as the median values of the median observation for each patient in the different groups. For each group minimum and maximum values for FOXP3+ Tregs and the range of survival data in each group are shown.

300 ∗ Number of FOXP3 pos Tregs

early patch/plaque lesions (median 37.5). Patients with tumour MF and MF with transformation to large T-cell lymphoma displayed lower frequencies for FOXP3 þ Tregs with median values of 23.5 and 6, respectively. Among the patients with CTCL unspecified, two had features consistent with primary cutaneous small- to medium-sized CD4 þ T-cell lymphoma. These patients had localized disease and a favourable prognosis with survival of 204 and 302 months. Interestingly, both of these had high numbers of FOXP3 þ Tregs (medians 131 and 147, respectively), see Figure 3. Pair-wise inter-group comparison of the numbers of FOXP3 þ Tregs in different disease categories was performed using a multi-ANOVA approach for repeated measurements. Using MF with infiltrated plaques as reference, statistically significant differences were observed to all CTCL groups. Identical results were found when using either of the other disease categories as reference group (Table 2). This shows that the Treg numbers correlate specifically to disease subtype and stage with higher numbers of FOXP3 þ Tregs in early than advanced MF or CTCL unspecified. In addition, a Kruskal–Wallis test was performed and confirmed the statistical significant difference between the groups (P ¼ 0.0001). All groups displayed a wide range of values for FOXP3 þ Tregs and there was a distinct overlap between the ranges comparing the different groups (Table 1 and Figure 3). The total variation was mainly contributed to by variation between the patients in the different groups (variance 1 ¼ 91%), and to a lesser extent due to random effects (variance 2 ¼ 9%).

250

200 ∗

150

∗ 100

50

0 MF I

MF II

MF turmour MF trans CTCL unspec Groups

Figure 3 Box plot displaying the numbers of FOXP3 þ Tregs in the different mycosis fungoides (MF) categories and CTCL unspecified. Box plot displaying median (bold line), interquartile range (box), outliers (circle) and extreme observations (star) according to standard definitions. In the CTCL unspecified group, the outliers represent FOXP3 observations in two cases of primary cutaneous small- to medium-sized CD4 þ CTCL. Abbreviations: MF I, MF with early patch/plaques; MF II, MF with infiltrated plaques; MF trans, MF with transformation to large T-cell lymphoma; CTCL unspec, cutaneous T-cell lymphoma unspecified.

Discussion

Cox regression analysis of survival The survival of all patients with respect to diagnostic group is shown in Figure 5a. A Cox proportional hazards model with morphological type as a class variable demonstrated inferior survival in advanced MF compared to early MF lesions. Furthermore, as anticipated, overall survival for patients with CTCL unspecified was significantly worse than for all MF categories (Figure 5a). The relationship between FOXP3 quartiles and survival is shown in Figure 5b. An analysis of the entire MF and CTCL unspecified cohort demonstrated that increasing numbers of FOXP3 þ Tregs were associated with improved survival (P ¼ 0.028; HR ¼ 0.988; CI (0.978, 0.999)). A final Cox model containing both FOXP3 values and disease type as covariates demonstrated that both numbers of FOXP3 þ Tregs and CTCL disease categories were independent risk factors for death (Table 3). This indicates that having high numbers of FOXP3 þ Tregs is associated with improved survival in different CTCL subtypes.

Over the past few years the relevance of Tregs in tumour immunosurveillance has attracted considerable attention. Woo et al.19 were the first to report increased numbers of CD4 þ CD25 þ Tregs in human epithelial tumours. These cells secreted transforming growth factor (TGF)-b, expressed high levels of CTLA-4 and inhibited autologous T-cell proliferation in a dosedependent manner.20 The impact of Tregs on antitumour immune responses has also been demonstrated in a transgenic murine colon carcinoma model, where Tregs abrogated the CD8 þ T-cell-mediated tumour rejection by specifically suppressing the cytotoxicity of these cells.21 Furthermore, there is evidence for a significantly enlarged Treg pool in malignant melanoma and carcinomas of the breast, pancreas, lung, ovary and gastrointestinal tract.6,20,22,23 This increase in Tregs appears to be stage dependent and to correlate inversely with survival rates.23–25 In this study of 86 CTCL patients, it is shown that plaque lesions of MF contain increased numbers of tumour-infiltrating Leukemia

FOXP3 in CTCL LM Gjerdrum et al

2516

Figure 4 Immunohistology for FOXP3 in plaque stage of MF with high numbers (a) and tumour MF with low numbers of FOXP3 þ Tregs (b). At higher magnification, it can be appreciated that the tumour cells are not immunolabelled (c).

Table 2 Results of the multi-ANOVA comparisons of FOXP3+ Tregs according to CTCL category

a

MF II vs

Group

P-value

MF I MF tumour MF trans CTCL unspecified

0.0001 0.0001 0.0001 0.0001

Abbreviations: CTCL, cutaneous T-cell lymphoma; MF I, MF with early patch/plaques; MF II, MF with infiltrated plaques; MF trans, MF with transformation to large T-cell lymphoma; vs, versus. a As an example MF with infiltrated plaques is displayed as reference group. Identical results were found using either of the other subtypes as reference.

Survival (cumulative proportion)

Analysis

1.0

0.8

0.6

0.4

0.2

Diagnosis: MF I MF II MF tumour & MF trans CTCL unspec

0.0

Leukemia

0

5

10

15 20 Time (years)

25

30

1.0 Survival (cumulative proportion)

FOXP3 þ Tregs compared to more advanced cases with tumours or transformation to large T-cell lymphomas. Importantly, we provide the first evidence that increasing numbers of FOXP3 þ Tregs is associated with improved survival in both MF and CTCL, unspecified. In a smaller study, Klemke et al.10 also obtained preliminary evidence for stage-related differences in FOXP3 expression in CTCL, but they did not address whether FOXP3 expression was associated with patient survival. However, two reports of follicular lymphoma and Hodgkin’s lymphoma have both shown that increasing number of Tregs was a favourable prognostic indicator,7,8 similar to the observations in this study, and apparently in contrast to the situation in solid tumours. How Tregs are recruited to the tumour environment is not clear. Studies in solid tumours have demonstrated that the elevated numbers of Tregs in peripheral blood is related to active proliferation of the Treg pool rather than to redistribution from other lymphoid structures.26 It has also been shown that epithelial tumour cells and macrophages in the tumour microenvironment attract Tregs by secretion of CCL22, a ligand to CCR4, expressed on tumour-infiltrating Tregs.25 Taken together, it is possible that both proliferation and migration may contribute to the expansion of Tregs in cancer tissue. Whether similar mechanisms are operative in CTCL is not known. However, it has been shown that a majority of circulating Tregs express skin-homing receptors, including both

0.8

0.6

0.4

FOXP3 quartiles

0.2

1 2 3 4

0.0 0

5

10

15 20 Time (years)

25

30

Figure 5 Kaplan–Meier survival curves according to (a) disease category, and (b) the number of FOXP3 þ Tregs in quartiles. Abbreviations: MF I, MF with early patch/plaques; MF II, MF with infiltrated plaques; MF trans, MF with transformation to large T-cell lymphoma; CTCL unspec, cutaneous T-cell lymphoma unspecified.

FOXP3 in CTCL LM Gjerdrum et al

2517 Table 3 Results of the Cox proportional hazards model with number of FOXP3+ Tregs and CTCL disease category as covariates Parameter Disease type MFI MFII MF tumour and MF trans FOXP3

Hazard ratio

Confidence

Intervals

P-value

0.013 0.052 0.050

0.001 0.011 0.011

0.144 0.247 0.238

0.0001 0.0001 0.0001

0.987

0.976

0.998

0.023

Abbreviations: MF I, MF with early patch/plaques; MF II, MF with infiltrated plaques; MF trans, MF with transformation to a large T-cell lymphoma. CTCL unspecified is used as a reference category.

CLA and CCR4, which can interact with corresponding ligands (CD62E and CCL17) on the dermal microvasculature.27,28 Furthermore, in MF CCL17 is unregulated on keratinocytes in lesional skin compared to healthy controls.29 Thus, it seems that skin can actively recruit memory Tregs from the periphery. At the cellular level, Tregs suppress the activation and proliferation of CD4 þ and CD8 þ T cells,30–32 by inhibition of interleukin (IL)-2 production and induction of cell cycle arrest in these cells, which requires cell–cell contact.32 In contrast, in vivo studies report that the presence of cytokines, such as IL-10 and TGF-b, might also be necessary for a suppressive milieu.9,33 In a recent study Yang et al.34 demonstrate that intratumoural Tregs inhibit the proliferation and degranulation of CD8 þ T cells in B-cell lymphoma, thereby dampening the antitumour effect of these cells. It is unknown whether a similar mechanism is involved in CTCL. Since malignant T cells and Tregs are derived from the same progenitors it is possible that Tregs might directly suppress the function of malignant T cells. In support of this hypothesis, Tiemessen et al.11 observed that Tregs from healthy individuals inhibit in vitro proliferation of T cells acquired from CTCL patients. Regarding FOXP3 labelling of the neoplastic cells, our results indicate that this is seen only very rarely in CTCL. These findings are in keeping with other reports of T-cell malignancies, which have indicated that peripheral T-cell lymphomas, including CTCL, are generally FOXP3 negative with the notable exception of adult T-cell lymphoma/leukaemia.35–39 One recent study has indicated that transformation of MF to CD30-positive large T-cell lymphomas is associated with a Treg phenotype.40 However, the limited number of cases investigated (n ¼ 5) and the lack of controls for the immunolabelling restricts the interpretation of this study. Adult T-cell lymphoma/leukaemia is recognized as a T-cell lymphoma associated with immunodeficiency and opportunistic infections.41 This phenomenon has been linked to immune suppression by the neoplastic cells which express FOXP3 in approximately half of the cases.35–39 A similar mechanism has been proposed to be operative in CTCL by Berger et al.9 They found that CD4 þ CTCL cells stimulated with autologous, immature dendritic cells loaded with apoptotic tumour cells could be induced to express CD25, CTLA-4 and FOXP3. In addition, the cells showed upregulation and secretion of both IL10 and TGF-b, features which are all characteristic of Tregs. Although it cannot be precluded that CTCL cells may be induced to assume Treg characteristics under certain conditions in vitro, the lack of FOXP3 labelling of the neoplastic cells in biopsy samples both in this and previous reports10,11,38 suggests that immune suppression by the neoplastic cells is less likely to be a general phenomenon in CTCL in vivo.

In conclusion, we provide the first evidence that a high frequency of tumour-infiltrating FOXP3 þ Tregs is associated with improved survival rates in MF and CTCL unspecified. Furthermore, our data show increased numbers of FOXP3 þ Tregs in early as opposed to advanced MF. Thus, our data support the notion that Tregs might play a fundamentally different role in lymphoma as compared to solid tumours.

Acknowledgements This work was supported by grants from the Novo Nordisk Foundation, the Foundation of 17F12-1981, the Research Council, the Danish Foundation for Cancer Research and the Danish Cancer Society.

References 1 Taams LS, Smith J, Rustin MH, Salmon M, Poulter LW, Akbar AN. Human anergic/suppressive CD4(+)CD25(+) T cells: a highly differentiated and apoptosis-prone population. Eur J Immunol 2001; 31: 1122–1131. 2 Zou W. Regulatory T cells, tumour immunity and immunotherapy. Nat Rev Immunol 2006; 6: 295–307. 3 Bennett CL, Christie J, Ramsdell F, Brunkow ME, Ferguson PJ, Whitesell L et al. The immune dysregulation, polyendocrinopathy, enteropathy, X-linked syndrome (IPEX) is caused by mutations of FOXP3. Nat Genet 2001; 27: 20–21. 4 Shevach EM. CD4+ CD25+ suppressor T cells: more questions than answers. Nat Rev Immunol 2002; 2: 389–400. 5 Shevach EM. Fatal attraction: tumors beckon regulatory T cells. Nat Med 2004; 10: 900–901. 6 Beyer M, Schultze JL. Regulatory T cells in cancer. Blood 2006; 108: 804–811. 7 Alvaro T, Lejeune M, Salvado MT, Bosch R, Garcia JF, Jaen J et al. Outcome in Hodgkin’s lymphoma can be predicted from the presence of accompanying cytotoxic and regulatory T cells. Clin Cancer Res 2005; 11: 1467–1473. 8 Carreras J, Lopez-Guillermo A, Fox BC, Colomo L, Martinez A, Roncador G et al. High numbers of tumor infiltrating FOXP3positive regulatory T-cells are associated with improved overall survival in follicular lymphoma. Blood 2006; 108: 2957–2964. 9 Berger CL, Tigelaar R, Cohen J, Mariwalla K, Trinh J, Wang N et al. Cutaneous T-cell lymphoma: malignant proliferation of T-regulatory cells. Blood 2005; 105: 1640–1647. 10 Klemke CD, Fritzsching B, Franz B, Kleinmann EV, Oberle N, Poenitz N et al. Paucity of FOXP3+ cells in skin and peripheral blood distinguishes Sezary syndrome from other cutaneous T-cell lymphomas. Leukemia 2006; 20: 1123–1129. 11 Tiemessen MM, Mitchell TJ, Hendry L, Whittaker SJ, Taams LS, John S. Lack of suppressive CD4+CD25+FOXP3+ T cells in advanced stages of primary cutaneous T-cell lymphoma. J Invest Dermatol 2006; 126: 2217–2223. 12 Odum N, Hofmann B, Jakobsen B, Langhoff E, Morling N, Platz P et al. HLA-DP related suppression of mixed lymphocyte reaction with alloactivated lymphocytes. Tissue Antigens 1986; 27: 32–43. 13 Woetmann A, Lovato P, Eriksen KW, Krejsgaard T, Labuda T, Zhang Q et al. Non-malignant T cells stimulate growth of T-cell lymphoma cells in the presence of bacterial toxins. Blood 2007; 109: 3325–3332. 14 Kaltoft K, Bisballe S, Dyrberg T, Boel E, Rasmussen PB, ThestrupPedersen K. Establishment of two continuous T-cell strains from a single plaque of a patient with mycosis fungoides. In Vitro Cell Dev Biol 1992; 28A: 161–167. 15 Odum N, Bregenholt S, Eriksen KW, Skov S, Ryder LP, Bendtzen K et al. The CC-chemokine receptor 5 (CCR5) is a marker of, but not essential for the development of human Th1 cells. Tissue Antigens 1999; 54: 572–577. 16 Krejsgaard T, Vetter-Kauczok CS, Woetmann A, Lovato P, Labuda T, Eriksen KW et al. Jak3- and JNK-dependent vascular endothelial growth factor expression in cutaneous T-cell lymphomas. Leukemia 2006; 20: 1759–1766. Leukemia

FOXP3 in CTCL LM Gjerdrum et al

2518 17 Gjerdrum LM, Sorensen BS, Kjeldsen E, Sorensen FB, Nexo E, Hamilton-Dutoit S. Real-time quantitative PCR of microdissected paraffin-embedded breast carcinomaFan alternative method for HER-2/neu analysis. J Mol Diagn 2004; 6: 42–51. 18 Ralfkiaer E, Willemze R, Meijer CJ, Dummer R, Jaffe ES, Swerdlow SH et al. Primary cutaneous peripheral T-cell lymphoma, unspecified. In: Leboit PE, Burg G, Weedon D, Sarasin A (eds). Pathology and Genetics of Skin Tumors. IARC Press: Lyon, France, 2006, pp 184–188. 19 Woo EY, Chu CS, Goletz TJ, Schlienger K, Yeh H, Coukos G et al. Regulatory CD4(+)CD25(+) T cells in tumors from patients with early-stage non-small cell lung cancer and late-stage ovarian cancer. Cancer Res 2001; 61: 4766–4772. 20 Woo EY, Yeh H, Chu CS, Schlienger K, Carroll RG, Riley JL et al. Cutting edge: regulatory T cells from lung cancer patients directly inhibit autologous T cell proliferation. J Immunol 2002; 168: 4272–4276. 21 Chen ML, Pittet MJ, Gorelik L, Flavell RA, Weissleder R, von BH et al. Regulatory T cells suppress tumor-specific CD8T cell cytotoxicity through TGF-beta signals in vivo. Proc Natl Acad Sci USA 2005; 102: 419–424. 22 Wolf AM, Wolf D, Steurer M, Gastl G, Gunsilius E, GrubeckLoebenstein B. Increase of regulatory T cells in the peripheral blood of cancer patients. Clin Cancer Res 2003; 9: 606–612. 23 Sasada T, Kimura M, Yoshida Y, Kanai M, Takabayashi A. CD4+CD25+ regulatory T cells in patients with gastrointestinal malignancies: possible involvement of regulatory T cells in disease progression. Cancer 2003; 98: 1089–1099. 24 Kono K, Kawaida H, Takahashi A, Sugai H, Mimura K, Miyagawa N et al. CD4(+)CD25 high regulatory T cells increase with tumor stage in patients with gastric and esophageal cancers. Cancer Immunol Immunother 2006; 55: 1064–1071. 25 Curiel TJ, Coukos G, Zou L, Alvarez X, Cheng P, Mottram P et al. Specific recruitment of regulatory T cells in ovarian carcinoma fosters immune privilege and predicts reduced survival. Nat Med 2004; 10: 942–949. 26 Wolf D, Rumpold H, Koppelstatter C, Gastl GA, Steurer M, Mayer G et al. Telomere length of in vivo expanded CD4(+)CD25 (+) regulatory T-cells is preserved in cancer patients. Cancer Immunol Immunother 2006; 55: 1198–1208. 27 Hirahara K, Liu L, Clark RA, Yamanaka K, Fuhlbrigge RC, Kupper TS. The majority of human peripheral blood CD4+CD25highFoxp3+ regulatory T cells bear functional skin-homing receptors. J Immunol 2006; 177: 4488–4494. 28 Chong BF, Murphy JE, Kupper TS, Fuhlbrigge RC. E-selectin, thymus- and activation-regulated chemokine/CCL17, and intercellular adhesion molecule-1 are constitutively coexpressed in

Leukemia

29

30 31

32 33 34

35 36

37 38 39

40 41

dermal microvessels: a foundation for a cutaneous immunosurveillance system. J Immunol 2004; 172: 1575–1581. Kakinuma T, Sugaya M, Nakamura K, Kaneko F, Wakugawa M, Matsushima K et al. Thymus and activation-regulated chemokine (TARC/CCL17) in mycosis fungoides: serum TARC levels reflect the disease activity of mycosis fungoides. J Am Acad Dermatol 2003; 48: 23–30. Murakami M, Sakamoto A, Bender J, Kappler J, Marrack P. CD25+CD4+ T cells contribute to the control of memory CD8+ T cells. Proc Natl Acad Sci USA 2002; 99: 8832–8837. Trzonkowski P, Szmit E, Mysliwska J, Dobyszuk A, Mysliwski A. CD4+CD25+ T regulatory cells inhibit cytotoxic activity of T CD8+ and NK lymphocytes in the direct cell-to-cell interaction. Clin Immunol 2004; 112: 258–267. Miyara M, Sakaguchi S. Natural regulatory T cells: mechanisms of suppression. Trends Mol Med 2007; 13: 108–116. Levings MK, Bacchetta R, Schulz U, Roncarolo MG. The role of IL10 and TGF-beta in the differentiation and effector function of T regulatory cells. Int Arch Allergy Immunol 2002; 129: 263–276. Yang ZZ, Novak AJ, Ziesmer SC, Witzig TE, Ansell SM. Attenuation of CD8(+) T-cell function by CD4(+)CD25(+) regulatory T cells in B-cell non-Hodgkin’s lymphoma. Cancer Res 2006; 66: 10145–10152. Chen S, Ishii N, Ine S, Ikeda S, Fujimura T, Ndhlovu LC et al. Regulatory T cell-like activity of Foxp3+ adult T cell leukemia cells. Int Immunol 2006; 18: 269–277. Karube K, Ohshima K, Tsuchiya T, Yamaguchi T, Kawano R, Suzumiya J et al. Expression of FoxP3, a key molecule in CD4CD25 regulatory T cells, in adult T-cell leukaemia/lymphoma cells. Br J Haematol 2004; 126: 81–84. Matsubara Y, Hori T, Morita R, Sakaguchi S, Uchiyama T. Phenotypic and functional relationship between adult T-cell leukemia cells and regulatory T cells. Leukemia 2005; 19: 482–483. Roncador G, Garcia JF, Garcia JF, Maestre L, Lucas E, Menarguez J et al. FOXP3, a selective marker for a subset of adult T-cell leukaemia/lymphoma. Leukemia 2005; 19: 2247–2253. Walsh PT, Benoit BM, Wysocka M, Dalton NM, Turka LA, Rook AH. A role for regulatory T cells in cutaneous T-cell lymphoma; induction of a CD4 + CD25 + Foxp3+ T-cell phenotype associated with HTLV-1 infection. J Invest Dermatol 2006; 126: 690–692. Hallermann C, Niermann C, Schulze HJ. Regulatory T-cell phenotype in association with large cell transformation of mycosis fungoides. Eur J Haematol 2007; 78: 260–263. Jaffe ES, Ralfkiaer E. Mature T-cell and NK-cell neoplasms. In: Jaffe ES, Harris NL, Stein H, Vardiman JW (eds). Pathology and Genetics of Tumours of Haematopoietic and Lymphoid Tissues. IARC Press: Lyon, France, 2001, pp 189–235.