Killer cell immunoglobulin-like receptor expression delineatesin situ S�zary syndrome lymphocytes

Journal of Pathology J Pathol 2003; 199: 77–83. Published online 21 October 2002 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/path.1251

Original Paper

Killer cell immunoglobulin-like receptor expression ´ delineates in situ Sezary syndrome lymphocytes Janine Wechsler,1 Martine Bagot,2,3 Maria Nikolova,2 Silvia Parolini,4 Nadine Martin-Garcia,1 Laurence Boumsell,2 Alessandro Moretta5 and Armand Bensussan2 * 1 D´ epartement de Pathologie, Hˆopital Henri-Mondor, Cr´eteil, France 2 INSERM U448, Facult´ e de M´edecine de Cr´eteil, France 3 Service de Dermatologie, Hˆ opital Henri-Mondor, Cr´eteil, France 4 Dipartimento di Scienze Biomediche e Biotecnologie, Universit` a di Brescia, 5 Dipartimento di Medicina Sperimentale, Universit` a di Genova, Italy

*Correspondence to: Dr Armand Bensussan, INSERM U448, Facult´e de M´edecine, 8 Rue du G´en´eral Sarrail, 94010 Cr´eteil, France. E-mail: Armand.Bensussan@im3. inserm.fr

Received: 22 January 2002 Revised: 22 March 2002 Accepted: 8 August 2002

Italy

Abstract p140/KIR3DL2 has been identified in malignant cell lines isolated from the skin and blood of patients with transformed mycosis fungoides (MF) and S´ezary’s syndrome (SS). For the first time, the expression of a cell membrane structure appeared to be able to distinguish CD4+ tumour lymphocytes from reactive lymphocytes in these small cutaneous T-cell lymphomas (CTCLs). This study has examined the in vivo expression of this receptor in various CTCL subtypes, which constituted a heterogeneous group. Tumour cells diffusely expressed KIR in SS, in lymphomatoid papulosis (LyP) and in CD4+CD30+ as well as CD8+ large cell pleomorphic CTCL. In contrast, the infiltrating lymphocytes did not express KIR in MF at the patch/plaque stage or in CD4+CD30− large cell pleomorphic CTCL, except for scattered small cells. One quarter of the transformed MF tested exhibited KIR+ tumour cells, suggesting heterogeneity in this subtype. KIR expression was also examined in inflammatory lesions characterized by a dense infiltrate of T cells, such as lupus erythematosus and lichen planus. Only scattered CD8+ cells in lichen planus expressed a significant amount of KIR3DL2. Taken together, these results show for the first time that KIR molecules are expressed in distinct subtypes of malignant CTCL. It is also shown for the first time that SS and MF, which are frequent variants of CTCL with similar histological features, can be distinguished by their KIR3DL2 expression analysis. The identification of this KIR also differentiates between lupus erythematosus and lichen planus, which are both diseases with dense benign lymphocytic infiltrates. Copyright  2002 John Wiley & Sons, Ltd. Keywords: S´ezary syndrome; mycosis fungoides; KIR3DL2; cutaneous T-cell lymphoma; immunophenotype

Introduction Killer-cell immunoglobulin-like receptors (KIRs) are a new family of MHC class I-specific receptors, which are mainly expressed by natural killer (NK) lymphocytes [1–3] and a small subset of CD8+ T cells [4]. Upon interaction with MHC class I molecules, these receptors provide inhibitory signals to the cytotoxic effector NK lymphocyte [1–3]. The decreased or altered expression of self-MHC class I molecules results in the absence of appropriate interaction with the inhibitory receptors, which allows NK lymphocytes to exhibit their cytotoxic activity. In T lymphocytes, KIR expression seems to be involved in the survival of a subset of memory-phenotype CD8+ cells [5]. We recently reported that CD4+ malignant cell lines isolated from the skin or the blood of CTCL patients express p140/KIR3DL2 [6]. We also described a novel KIR3DL2 allele, which is expressed by malignant cells isolated from CTCL patients. This new allele was subsequently reported Copyright  2002 John Wiley & Sons, Ltd.

and designated 3DL2*008 [7]. Interestingly, none of the other inhibitory natural killer receptors, including KIR and lectin-like receptors [1], has been detected on CD4+ CTCL malignant cells [6]. CTCLs are a heterogeneous group of lymphomas primarily involving the skin. They include an indolent group of lymphomas with mycosis fungoides (MF) and its variants, which are characterized by skin invasion of a clonal proliferation of CD4+ T lymphocytes that phenotypically resemble mature Thelper lymphocytes. CD30+ T-cell proliferations are another group of indolent T lymphomas characterized by a favourable evolution with a tendency to spontaneous regression. Finally, an aggressive group includes S´ezary syndrome (SS), with the presence of circulating malignant T cells, transformed MF, and CD30− pleomorphic CTCL. The pathophysiology of CTCL remains poorly understood and until now it has been difficult phenotypically to distinguish normal reactive lymphocytes

78

from malignant cells [8–10]. In the present study, we performed in situ analysis with a monoclonal antibody specific for KIR3DL2, in order to characterize the subtypes of cutaneous T-cell lymphomas.

Materials and methods Patients Skin and blood samples from patients were obtained after informed consent and approval by an ethical committee. Forty-eight cases investigated in our Department of Dermatology were included in this study. The cases were classified according to the EORTC classification of primary cutaneous lymphomas [11]. The study included seven cases of SS at the erythrodermic stage, two transformed SS (TSS), 11 cases of patch/plaque stage mycosis fungoides (MF), eight cases of transformed MF (TMF), two cases of lymphomatoid papulosis (LyP), three cases of CD30+ pleomorphic cutaneous T-cell lymphomas (PleoM CD30+), four cases of CD30− pleomorphic cutaneous T-cell lymphomas (two pleomorphic CD4+ T-cell lymphoma, one cytotoxic CD8+ subcutaneous T-cell lymphoma, one HTLV1-associated CD8+ T-cell lymphoma). In addition, 11 cases of inflammatory skin disease chosen for their dense lymphocytic infiltrates were included as controls (six cases of lupus erythematosus and five cases of lichen planus).

Isolation of tumour lymphocytes Fresh CTCL tumour cells were obtained from tumour fragments mechanically dispersed into single-cell suspensions as described elsewhere [12]. The mononuclear cells were then washed and frozen in human serum plus 10% dimethyl sulphoxide for later use. For SS patients, the mononuclear blood cells were isolated by the technique of Ficoll-Isopaque (Pharmacia Fine Chemicals, Piscataway, NJ, USA).

Monoclonal antibodies and flow cytometry studies One- and two-colour immunofluorescence analyses were performed as previously described [6]. Briefly, after staining the cells with the various monoclonal antibodies (MAbs) and washing them in PBS, gated lymphocytes were analysed by flow cytometry using a single argon flow cytometer analyser (Epics XL, Beckman-Coulter, Miami, FL, USA). MAbs Q66 (IgM, anti-p140/KIR3DL2) (Beckman-Coulter) and AZ158 (IgG2a, recognizing both p70/KIR3DL1 and p140/KIR3DL2) were produced as previously described [13,14]. Anti-CD4 MAb was purchased from Beckman-Coulter. For the immunostaining procedure, purified AZ158 MAb was used at 5 µg/ml; CD4 and CD8 MAbs (Dako A/S, Denmark) were used diluted at 1/50; and TIA1 (Beckman-Coulter) was diluted at 1/100. Copyright  2002 John Wiley & Sons, Ltd.

J Wechsler et al

Immunostaining procedure Six-micrometre-thick frozen sections were mounted on Superfrost Plus slides (CML, Angers, France), airdried overnight, and fixed in acetone for 10 min before storing at −20 ◦ C. Before use, the slides were postfixed in acetone for 5 min and air-dried. After rehydratation in Tris/NaCl buffer (TBS), an immunostaining procedure was performed using a Biotin/Avidin system conjugated to alkaline phosphatase (Vectastain ABC-AP kit from Vector, Burlingame, USA) and according to the manufacturer’s instructions. The alkaline phosphatase reaction was revealed by NaphtholFast Red (Sigma, Saint Quentin Fallavier, France) and sections were counterstained in blue with haematoxylin. Primary antibodies were incubated for 1 h at room temperature in a humid chamber.

Results CD4+ T lymphocytes freshly isolated from patients with CTCL are stained by anti-p140/KIR3DL2 MAb To determine whether p140/KIR3DL2 is expressed by freshly isolated cells, we first tested by flow cytometry the reactivity of Q66 MAb with gated lymphocytes obtained from the blood of five SS patients (cases 1, 2, 5, 6, and 7, Tables 1 and 2) and from tumoural skin fragments of one MF patient (case 24, Tables 1 and 2). All patients exhibited a significant T-lymphocyte subset labelled by both Q66 and CD4 MAbs. It should be noted that similar results were obtained when the AZ158 MAb (specific for both KIR3DL1 and KIR3DL2) was used instead of Q66 MAb, indicating that in this case AZ158 MAb recognizes p140/KIR3DL2 and not p70/KIR3DL1 (Table 1). Interestingly, the percentages of CD4+KIR3DL2+ cells obtained in each patient corresponded to the percentages of circulating malignant cells detected by cytomorphology (data not shown). We have already reported a long-term cultured malignant cell line derived from the peripheral blood of a patient named Pno (case 7, Table 2). We have demonstrated that the malignant cell line and the malignant cells in the blood were identical [15] and expressed KIR3DL2 [6]. Table 1. Flow cytometric analysis of lymphocytes isolated from the blood of five patients with SS and from the skin of a patient with TMF Percentage of positive cells with Case Anti-CD4 MAb Anti-CD4 MAb No Anti-CD4 MAb + anti-Q66 MAb + anti-AZ158 MAb 1 2 5 6 7 24

82 96 81 93 95 66

51 16 19 20 70 19

52 17 23 20 89 17

J Pathol 2003; 199: 77–83.

KIR expression in SS lymphocytes

79

Table 2. Immunohistochemical findings in 37 patients with CTCL Case

Diagnosis

CD8

Pattern

TIA1

Pattern

AZ158

Pattern

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

SS SS SS SS SS SS SS TSS TSS MF MF MF MF MF MF MF MF CD8 MF CD8 MF CD8 MF CD8 TMF TMF TMF TMF TMF TMF TMF TMF LyP LyP PleoM CD30+ PleoM CD30+ PleoM CD30+ PleoM CD30− PleoM CD30− Subcut PleoM PleoM HTLV1+

+∗ + + − − nd + + − nd − nd nd − + + ++ ++ ++ ++ ++ + − + − − − − ++ − ++ − − − − ++ ++

Scattered Scattered Scattered

+ + + − nd nd nd + nd nd − nd nd + − nd + + ++ ++ + + − + nd − − − ++ + + + ++ − − ++ +

Scattered Scattered Scattered

++ ++ ++ ++ +++ +++ ++ +++ ++ − − − − − − + − − − − + + ++ +++ − − − − +++ ++ ++ +++ +++ − + ++ ++

Cohesive Cohesive Cohesive Cohesive Cohesive Cohesive Cohesive Cohesive Cohesive

∗−

Scattered Scattered Scattered

Scattered Scattered Cohesive Cohesive Cohesive Cohesive Scattered Scattered Scattered

Scattered Scattered

Cohesive Cohesive

Scattered

Scattered

Scattered Scattered Cohesive Cohesive Scattered Scattered Scattered

Scattered Scattered Scattered Scattered Cohesive

Cohesive Scattered

Scattered

Scattered Scattered Cohesive Cohesive

Cohesive Cohesive Cohesive Cohesive Cohesive Scattered Cohesive Cohesive

= 75% of positive cells; nd = not done.

In situ analysis reveals that in S´ezary syndrome, the majority of the cutaneous lymphocytes are reactive with AZ158 MAb Next, various subtypes of CTCL were analysed by immunohistological staining of cryostat tissue sections for expression of KIR3DL2. These experiments, however, were performed with the AZ158 MAb (which recognizes both p140/KIR3DL2 and p70/KIR3DL1) since the only anti-KIR3DL2-specific MAb (Q66) available is not suitable for this type of assay. The results showed that all the SS cases tested, including the erythrodermic as well as the transformed variant, were reactive with AZ158 MAb (Table 2). As shown by representative case 5 in Figure 1A, most of the intradermal and epidermotropic cells were labelled. The positive cells showed membrane staining and a cohesive pattern. It should be noted that in cases 1, 2, 5, 6, and 7, we were able to reveal that peripheral blood contained a subset of CD4+ lymphocytes which was reactive with Q66 MAb (Table 1), suggesting that AZ158 MAb detects in situ p140/KIR3DL2 and not p70/KIR3DL1. Table 2 also shows that in Copyright  2002 John Wiley & Sons, Ltd.

4/9 cases of SS, the infiltrate contained a significant percentage of scattered cells expressing CD8 and TIA1. These cells correspond to the CD8+ tumourinfiltrating lymphocytes expressing an activated cytotoxic phenotype [16,17].

In MF at the patch/plaque stage, the lymphocytic infiltrate fails to react with AZ158 MAb, whereas heterogeneous reactivity is obtained in MF at the transformed stage In 11/11 cases of MF at the patch/plaque stage, the infiltrating cells were not stained with AZ158 MAb (Table 2 and Figure 1B with representative case 13). It is important to emphasize that within this MF group, four cases were CD8+ lymphomas. Interestingly, in the eight cases of tumour-stage mycosis fungoides, the results of the immunostaining with AZ158 MAb were heterogeneous. In four cases the neoplastic cells were negative, whereas in two cases the majority of the cells were strongly positive (cases 23 and 24, Table 2). The latter cases had a CD4 phenotype and exhibited cohesive positive staining with AZ158 MAb, J Pathol 2003; 199: 77–83.

80

J Wechsler et al

Figure 1. In situ analysis of AZ158 monoclonal antibody reactivity in various subtypes of CTCL. (A) S´ezary syndrome. Case 5; most of the intradermal and epidermotropic cells are labelled by AZ158 MAb. ×400. (B) Mycosis fungoides. Case 13; AZ158 MAb stains only a few cells in the dermis. ×250. (C, D) Transformed mycosis fungoides. Case 24; the large tumour cells are strongly stained by AZ158 MAb (C). ×400. CD8+ cells correspond to scattered reactive cells (D). ×250. (E, F) Lymphomatoid papulosis. Case 29; numerous cells of the dermal infiltrate are recognized by AZ158 MAb. ×100 (E); ×400 (F). (G, H) CD30+ large cell lymphoma. Case 32; all large atypical cells are strongly stained by AZ158 MAb (G) and show intracytoplasmic granules of TIA1 cytotoxic protein (H). ×400

as shown by representative case 24 in Figure 1C. Importantly, we demonstrate that this cohesive positive staining corresponds to the CD4+ malignant lymphocytes, since CD8 MAb revealed only scattered Copyright  2002 John Wiley & Sons, Ltd.

lymphocytes (Figure 1D). Flow cytometry analysis performed with lymphocytes isolated from lesional skin of case 24 revealed that malignant lymphocytes were p140/KIR3DL2+, as shown by the reactivity J Pathol 2003; 199: 77–83.

KIR expression in SS lymphocytes

81

of a CD4+ subset with Q66 MAb (Table 1). In two other cases of TMF (21 and 22), the cells stained with AZ158 MAb were scattered (Table 2) and had a similar distribution to the CD8+ infiltrating lymphocytes (data not shown).

Lymphomatoid papulosis as well as CD30+CD4+ and CD30−CD8+ pleomorphic large T-cell lymphomas are stained with AZ158 MAb In the two LyP cases, the infiltrates showed numerous large atypical CD30+ lymphocytes that were stained with AZ158 MAb, as shown by representative case 29 in Figures 1E and 1F. Similarly, in all cases of CD30+ pleomorphic large cell lymphomas, atypical cells exhibited diffuse cytoplasmic staining with AZ158 MAb (case 32 is shown in Figure 1G) and were TIA1+ (Figure 1H). Furthermore, in the two cases of CD8+ cytotoxic pleomorphic large Tcell lymphomas, the tumour cells were also strongly reactive with AZ158 MAb (Table 2). One of these cases was a subcutaneous cytotoxic T-cell lymphoma and the other an HTLV1-associated T-cell lymphoma from the Caribbean. It should be noted that in the two cases of CD30 − CD8− pleomorphic large T-cell lymphoma, the tumour cells failed to react with AZ158 MAb (Table 2).

Infiltrating CD8+ lymphocytes in lichen planus react with AZ158 MAb To determine further whether non-malignant skin T lymphocytes were stained by AZ158 MAb, we analysed cutaneous samples with dense benign lymphocytic infiltrates, such as lichen planus and lupus erythematosus. The results presented in Table 3 indicate that in the infiltrates of lichen planus, many scattered CD8+ cells reacted with AZ158 MAb. Interestingly, in lupus erythematosus, there was also evidence of many scattered CD8+ cells, but no significant staining with AZ158 MAb (Table 3).

Discussion Skin infiltrates in CTCL are composed of a mixture of malignant and non-malignant lymphocytes. As there is no phenotypic marker to identify malignant small-sized lymphocytes, it can be very difficult to distinguish them from reactive lymphocytes and consequently to differentiate some cases of CTCL from inflammatory lesions. A previous report demonstrated by flow cytometry that KIR3DL2/p140 expression was restricted to the malignant CD4+ lymphocytes in cases of CTCL [6]. This was the first time that KIR3DL2/p140 had been reported on CD4+ lymphocytes. Looking for CTCL-associated antigen, the present study investigated the in situ expression of KIR3DL1 in 37 CTCL skin specimens. These studies were carried out using AZ158 MAb (recognizing both KIR3DL1 and KIR3DL2 molecules) since Q66, the only KIR3DL2specific MAb, is not suitable for this type of analysis. Most skin infiltrating cells were positive in 9/9 cases of S´ezary syndrome (including both erythrodermic and tumour cases), in 2/2 cases of lymphomatoid papulosis, in 3/3 cases of CD30+ pleomorphic large T-cell lymphomas, and in 2/2 cases of CD8+ pleomorphic large CTCLs. The infiltrating cells were negative in the 11/11 cases of patch/plaque stage MF. In contrast, the results were heterogeneous in transformed MF: we observed two strongly and two weakly positive cases, the latter due to the labelling of CD8+ reactive lymphocytes. The four remaining cases were negative. In the group of inflammatory lesions, reactivity with AZ158 MAb was detected in lichen planus but not in lupus erythematosus, both including CD8+ lymphocytes. According to the results presented in Table 1, and the previous study performed on blood lymphocytes [6], the S´ezary syndrome cells seem to express KIR3DL2/p140 instead of KIR3DL1/p70. Independently of the KIR3DL2 or KIR3DL1 expression determination, it is important to emphasize that the significant reactivity obtained with AZ158 MAb in SS skin fragments allows the identification of the smallsized S´ezary malignant cells. We still, cannot exclude

Table 3. Immunohistochemical findings in 11 patients with inflammatory lymphocytic infiltrates Case No 1 2 3 4 5 6 7 8 9 10 11 ∗−

Diagnosis Lupus Lupus Lupus Lupus Lupus Lupus Lichen Lichen Lichen Lichen Lichen

CD8

Pattern

−∗ nd + + + − + ++ ++ ++ ++

Scattered Scattered Scattered Scattered Scattered Scattered Scattered Scattered

TIA1 − nd nd − − − nd nd nd ++ ++

Pattern

AZ158

Pattern

Scattered Scattered

− − − − − − + ++ + ++ +

Scattered Scattered Scattered Scattered Scattered

= 75% of positive cells; nd = not done.

Copyright  2002 John Wiley & Sons, Ltd.

J Pathol 2003; 199: 77–83.

82

the possibility of residual labelling obtained with the CD8+ reactive lymphocytes. In contrast, in transformed S´ezary syndrome, it is easier to identify malignant cells, as they are medium/large-sized and cohesive, whereas CD8+ reactive lymphocytes are small and scattered. Whereas MF belongs to the indolent group of CTCLs, SS is included in the aggressive group and is characterized by erythroderma, lymphadenopathy, and the presence of S´ezary cells in peripheral blood. The histological features in SS may be similar to those in MF, so that the distinct reactivity of AZ158 MAb appears crucial for distinguishing SS from MF. Large cell transformation in MF is associated with an aggressive clinical course [18]. There are neither diagnostic nor predictive biological markers of MF transformation, which is defined on a histopatholological basis. In transformed MF, the reactivity with AZ158 MAb was heterogeneous; thus, further studies are needed to interpret these results. CD30+ large T-cell lymphomas, lymphomatoid papulosis, and borderline proliferations belong to a spectrum of primary cutaneous CD30+ lymphoproliferations. In all cases of lymphomatoid papulosis and CD30+ pleomorphic large cell lymphomas, the large atypical cells were stained by AZ158 MAb. In a previous study, we demonstrated in these lymphomas that CD30 expression was associated with the cytotoxic protein TIA1 expression [19]. Here, by comparing serial sections respectively stained with AZ158, anti-CD30, and anti-TIA1 antibodies (data not shown), we found that the same malignant cells that expressed TIA1 (see Figures 1G and 1H) and CD30 were also stained by AZ158 MAb. At present, however, it is not yet possible to determine whether this co-expression is correlated with the favourable prognosis commonly associated with CD30+ proliferations in the skin. Human T-cell lymphotropic virus type-1 (HTLV1)associated lymphomas are frequent in the Caribbean, where this virus is endemic. They can result in any variant of CTCL, either chronic or aggressive. Cytotoxic cutaneous lymphomas in their panniculitislike variant may have either an aggressive or an indolent course. They may show either a CD8+ or a CD4+ phenotype. The reactivity with AZ158 MAb that we observed in the HTLV1-associated and subcutaneous cases of lymphomas in this study may be related to their CD8+ phenotype. However, as the four cases of patch/plaque stage CD8+ MF were found to be negative with AZ158 MAb, the expression of the inhibitory receptors must be related not only to the CD8+ phenotype. Lupus erythematosus and lichen planus are characterized by infiltrates composed of both CD4 and CD8 T lymphocytes. Although in these two entities CD8+ T lymphocytes were numerous, the scattered CD8 cells were reactive with AZ158 MAb only in lichen planus. This suggests that in lichen planus the lymphocyte infiltrate consists of a CD8+ cell subset. Copyright  2002 John Wiley & Sons, Ltd.

J Wechsler et al

Lupus erythematosus may present as a lichenoid variant that is not easily distinguishable from lichen planus. The differential reactivity obtained in this study with AZ158 MAb might also be helpful to distinguish lupus erythematosus from lichen planus. In conclusion, we demonstrate for the first time that expression of KIR3DL2 allows a distinction to be made between the two common variants of CTCL (SS vs MF) that differ in their prognosis. Additional studies are now required to verify the actual role of KIR3DL2 in the pathophysiology of SS, taking into account the potential role of this inhibitory receptor in the maintenance of lymphocyte survival [5].

Acknowledgements This work was supported by grants from INSERM, ARC (to MB and AB), and SRD.

References 1. Long EO. Regulation of immune responses through inhibitory receptors. Annu Rev Immunol 1999; 17: 875–904. 2. Moretta A, Biassoni R, Bottino C, et al. Major histocompatability complex class I-specific receptors on human natural killer and T lymphocytes. Immunol Rev 1997; 155: 105–117. 3. Lanier LL. NK cell receptors. Annu Rev Immunol 1998; 16: 359–393. 4. Mingari MC, Moretta A, Moretta L. Regulation of KIR expression in human T lymphocytes. A safety mechanism which may impair protective T cell responses. Immunol Today 1998; 19: 153–157. 5. Ugolini S, Arpin C, Anfossi N, et al. Involvement of inhibitory NKRs in the survival of a subset of memory-phenotype CD8+ T cells. Nat Immun 2001; 2: 430–435. 6. Bagot M, Moretta A, Sivori S, et al. CD4+ cutaneous T-cell lymphoma cells express the p140-killer cell immunoglobulin-like receptor. Blood 2001; 97: 1388–1391. 7. Gardiner CM, Guethlein LA, Shilling HG, et al. Different NK cell surface phenotypes defined by the DX9 antibody are due to KIR3DL1 gene polymorphism. J Immunol 2001; 166: 2992–3001. 8. Wood GS, Deneau DG, Miller RA, Levy R, Hoppe PT, Warnke RA. Subtypes of cutaneous T-cell lymphoma defined by expression of Leu-1 and Ia. Blood 1982; 59: 876–882. 9. Laetsch B, H¨affner A, D¨obbeling U, et al. CD4+/CD7− T cell frequency and polymerase chain reaction-based clonality assay correlate with stage in cutaneous T cell lymphomas. J Invest Dermatol 2000; 114: 107–111. 10. Nikolova M, Tawab A, Marie-Cardine A, Bagot M, Boumsell L, Bensussan A. Increased expression of a novel early activation surface membrane receptor in cutaneous T cell lymphoma cells. J Invest Dermatol 2001; 116: 731–738. 11. Willemze R, Kerl H, Sterry W, et al. EORTC classification for primary cutaneous lymphomas: a proposal from the Cutaneous Lymphoma Study Group of the European Organization for Research and Treatment of Cancer. Blood 1997; 90: 354–371. 12. Bagot M, Echchakir H, Mami-Chouaib F, et al. Isolation of tumor-specific cytotoxic CD4+ and CD4+CD8dim+ T cell clones infiltrating a cutaneous T-cell lymphoma. Blood 1998; 91: 4331–4341. 13. Pende D, Biassoni R, Cantoni C, et al. The natural killer cell receptor specific for HLA-A allotypes: a novel member of the p58/p70 family of inhibitory receptors that is characterized by three immunoglobulin-like domains and is expressed as a 140 kD disulphide-linked dimer. J Exp Med 1996; 184: 505–518. 14. Parolini S, Cantoni C, Castriconi R, et al. The AZ158 MAb specifically reacts with p70 and p140 inhibitory NK receptors for HLA-B and HLA-A alleles. In Leucocyte Typing VII, White Cell J Pathol 2003; 199: 77–83.

KIR expression in SS lymphocytes

Differentiation Antigens Mason D (ed.). Oxford University Press Inc: New York, 2002; 415–417. 15. Poszepczynska E, Bagot M, Echchakir H, et al. Functional characterization of an IL-7-dependent CD4 + CD8 alpha alpha+ Th3-type malignant cell line derived from a patient with a cutaneous T-cell lymphoma. Blood 2000; 96: 1056–1063. 16. Wood GS, Edinger A, Hoppe RT, Warnke RA. Mycosis fungoides skin lesions contain CD8+ tumor-infiltrating lymphocytes expressing an activated, MHC-restricted cytotoxic T-lymphocyte phenotype. J Cutan Pathol 1994; 21: 151–156. 17. Berger CL, Wang N, Christensen I, Longley J, Heald P, Edelson RL. The immune response to class I-associated tumor-specific

Copyright  2002 John Wiley & Sons, Ltd.

83

cutaneous T-cell lymphoma antigens. J Invest Dermatol 1996; 107: 392–397. 18. Vergier B, de Muret A, Beylot-Barry M, et al. Transformation of mycosis fungoides: clinicopathological and prognostic features of 45 cases. Blood 2000; 95: 2212–2218. 19. Boulland ML, Wechsler J, Bagot M, Pulford K, Kanavaros P, Gaulard P. Primary CD30-positive cutaneous T-cell lymphomas and lymphomatoid papulosis frequently express cytotoxic proteins. Histopathology 2000; 36: 136–144.

J Pathol 2003; 199: 77–83.