Programmed death 1 is expressed in cutaneous infiltrates of mycosis fungoides and Sézary syndrome

Abstract

Expression of PD-1 (receptor Programmed Death 1, CD279) in some peripheral T-cell lymphomas has recently been demonstrated. Because antibody-based therapies have improved outcomes in non-Hodgkin lymphoma, and antibodies targeting PD-1 are in clinical development, expression of this molecule in cutaneous T-cell lymphoma may be of therapeutic interest. There are no studies to date specifically addressing the frequency and extent of PD-1 expression in mycosis fungoides or Sézary syndrome. We thus characterized PD-1 expression in mycosis fungoides and Sézary syndrome by immunohistochemistry on lesion skin biopsies and by flow cytometry on peripheral blood tumor cells. Fifteen of 30 cases of mycosis fungoides and 8 of 11 cases of Sézary syndrome were positive for PD-1 by immunohistochemistry. Circulating tumor cells from five of the immunohistochemistry-positive cases of Sézary syndrome were evaluated by flow cytometry and also found to be positive. These data indicate that a substantial proportion of patients with mycosis fungoides and Sézary syndrome are positive for PD-1. This result warrants further investigation of PD-1 as a potential therapeutic target. Expression of PD-1 (receptor Programmed Death 1, CD279) in some peripheral T-cell lymphomas has recently been demonstrated. Because antibody-based therapies have improved outcomes in non-Hodgkin lymphoma, and antibodies targeting PD-1 are in clinical development, expression of this molecule in cutaneous T-cell lymphoma may be of therapeutic interest. There are no studies to date specifically addressing the frequency and extent of PD-1 expression in mycosis fungoides or Sézary syndrome. We thus characterized PD-1 expression in mycosis fungoides and Sézary syndrome by immunohistochemistry on lesion skin biopsies and by flow cytometry on peripheral blood tumor cells. Fifteen of 30 cases of mycosis fungoides and 8 of 11 cases of Sézary syndrome were positive for PD-1 by immunohistochemistry. Circulating tumor cells from five of the immunohistochemistry-positive cases of Sézary syndrome were evaluated by flow cytometry and also found to be positive. These data indicate that a substantial proportion of patients with mycosis fungoides and Sézary syndrome are positive for PD-1. This result warrants further investigation of PD-1 as a potential therapeutic target. Mycosis fungoides and Sézary syndrome are two of the most common types of cutaneous T-cell lymphomas (CTCLs). Recently used therapeutic strategies for advanced-stage CTCL involve targeted treatment of the neoplastic cell population. The use of antibodies directed against markers expressed in the neoplastic infiltrates might greatly improve patient outcomes, as has been the case with rituximab for B-cell lymphomas. Programmed death 1 (PD-1; CD279) is an inhibitory receptor for B7-H1 and B7-DC, two members of the B7 family of costimulatory/coinhibitory molecules that have an important role in the inhibition of T-cell immunity [1]. Therapeutic strategies targeting B7-H1 and PD-1 are currently being investigated in the clinical trial setting [2]. PD-1 is expressed by germinal center–associated T cells in normal or reactive lymphoid tissue and by some activated T cells [3, 4]. PD-1 has also been detected in several types of T-cell non-Hodgkin lymphoma, including angioimmunoblastic T-cell lymphoma, adult T-cell leukemia lymphoma, and peripheral T-cell lymphoma/leukemia [4-7]. Because PD-1 expression by malignant cells in CTCL, particularly Sézary syndrome, has not been previously characterized, we aimed to determine the frequency of expression of this marker in mycosis fungoides and Sézary syndrome. For the study period of 1998–2008, skin biopsy specimens were available from 37 patients with CTCL (26 mycosis fungoides, 11 Sézary syndrome). Five of these patients (four mycosis fungoides, one Sézary syndrome) had two biopsies from distinct sites available; the results were consistent in both samples in all such cases (three positive, two negative). One case of Sézary syndrome and nine control cases of spongiotic dermatitis were excluded because they did not meet the inclusion criteria. Peripheral blood samples from five patients with Sézary syndrome were available and evaluated by flow cytometric analysis. All control tissues stained appropriately. All cases of nodal angioimmunoblastic T-cell lymphoma had clusters of medium-sized PD-1-positive lymphocytes associated with high endothelial venules, as previously described for this marker [4]. In reactive human tonsil tissue, scattered small lymphocytes in the germinal centers and occasional PD-1-positive T cells in interfollicular T-zones were PD-1 positive, as previously described [4]. Skin biopsies of spongiotic dermatitis samples contained variable percentages of PD-1-positive lymphocytes scattered throughout the infiltrate (mean, 35%; median, 30%; range, 10–80%). Of 41 samples of CTCL, 23 (56%) were positive for PD-1 by immunohistochemistry (see Fig. 1). The percentages of PD-1-positive cases of limited patch or plaque mycosis fungoides, generalized tumor mycosis fungoides, and Sézary syndrome were 40, 60, and 73%, respectively (Table I). In the PD-1-positive cases, the percentages of PD-1-positive cells within the infiltrate were well above (mean, 73%) the cutoff of 30%. Biopsy sample from a patient with Sézary syndrome (top, hematoxylin-eosin; original magnification ×100). Tumor cells stained positively with anti-CD4 antibody (inset, original magnification ×400) and anti-PD-1 antibody (bottom, original magnification ×400). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] The five cases of Sézary syndrome for which peripheral blood mononuclear cells were available were positive for PD-1 by immunohistochemistry and were also observed to express PD-1 on the cell surface by flow cytometry (see Fig. 2). Flow cytometry of peripheral blood samples from the same patient with Sézary syndrome (as in Fig. 1) showed circulating malignant CD4+, Vβ-restricted cells that expressed surface PD-1. All imaging was at room temperature. The current study demonstrated that malignant T cells in CTCL, particularly Sézary syndrome, highly express PD-1, as determined by immunohistochemistry and flow cytometry. In all five cases of Sézary syndrome analyzed by flow cytometry, PD-1 was expressed on circulating cells, which correlated with the immunohistochemistry results from those patients. Because PD-1 expression may potentially be restricted to intracellular sites, the detection of PD-1 on the cell surface by flow cytometry excludes the possibility that PD-1 positivity detected by immunohistochemistry is solely due to intracellular compartmentalization. PD-1 expression was observed in most CTCL cases examined, and surface expression was confirmed by flow cytometry, which also suggests that PD-1 may be a viable therapeutic target. Our findings are consistent with a previously published study centered on PD-1 expression in angioblastic T-cell lymphoma, which included a small series of patients with mycosis fungoides, some of whom were PD-1 positive (5/9 cases) [5]. PD-1 expression pattern was consistent within each patient at different locations and/or at different points in time in the five patients who had multiple biopsies. This suggests that this marker is constitutively expressed in CTCL, unlike the inducible expression described in activated T cells. Although few cases with multiple biopsies were tested, this is a significant finding for PD-1 expression in CTCL as a potential therapeutic target. We also found the percentage of PD-1-positive cells in positive cases to be much higher than proportions observed in inflammatory skin lesions (spongiotic dermatitis) in most cases. This illustrates that PD-1-positive tumor cells are present in sufficient quantities to suggest that targeting this population would have a potential impact on the tumor burden. In addition to its ability to regulate cytokine production, PD-1 may be upregulated on antigen-specific T cells in some chronic viral infections, including human T-lymphotropic virus type 1 infection associated with adult T-cell leukemia lymphoma, and suppress their proliferation [8]. Such PD-1-positive T cells are hypoproliferative (exhausted) and may contribute to viral persistence and chronic infection [9, 10]. Similar to PD-1-positive exhausted T cells, Sézary cells in vivo are less responsive to conventional mitogens used to stimulate normal T cells, raising the possibility that PD-1 receptor activation is associated with suppression of Sézary cell proliferation [11]. In addition to being hypoproliferative, Sézary cells are relatively resistant to apoptotic stimuli; however, the factors responsible for this resistance are incompletely understood [12]. Recent work has demonstrated that reverse signaling by B7-H1, after PD-1 engagement, may prevent the induction of apoptosis in tumor cells; thus, cross-talk between B7-H1- and PD-1-expressing malignant T cells may confer resistance to apoptotic stimuli via B7-H1 reverse signaling [13, 14]. Further studies to assess the expression and function of PD-1 and B7-H1 in CTCL may be warranted; blocking antibodies for B7-H1 and PD-1 are currently being tested in the clinical trial setting [15]. Furthermore, the potential use of PD-1 as a therapeutic target in CTCL warrants the investigation of antibodies capable of influencing complement-mediated cytotoxicity and antibody-mediated cytotoxic activity to directly deplete target tumor cells. In summary, we demonstrate by immunohistochemistry and flow cytometry that CTCL tumor cells frequently express PD-1. This finding has potential clinical impact because it presents a novel therapeutic target for CTCL. Additional studies are warranted to confirm these results with additional sampling and to characterize the biological and therapeutic effects of targeting the PD-1 molecule in patients with mycosis fungoides or Sézary syndrome. This study was approved by the Mayo Clinic Institutional Review Board. The Mayo Clinic archives were searched for skin biopsy specimens obtained between 1998 and 2008 from patients with CTCL. If available, peripheral blood samples were obtained for flow cytometric analysis. Twenty specimens from patients with spongiotic dermatitis were selected as controls. Tissue samples from three cases of nodal angioimmunoblastic T-cell lymphoma and three sections of reactive human tonsil tissue were also stained as positive controls [4]. For all study and control samples, cases were included only if the clinical and histologic findings were consistent with the diagnosis and if neoplastic cells, on the basis of histologic interpretation, were sufficient in quantity to score (to determine the proportion of cells demonstrating positive staining) by light microscopy. PD-1 immunostaining was performed by taking 5-μm-thick sections from formalin-fixed, paraffin-embedded tissue blocks. Antigen retrieval was done by heating tissue sections in 1 mmol l−1 EDTA (pH 8) to 121°C in a Digital Decloaking Chamber (Biocare Medical, Concord, CA). Sections were blocked for endogenous peroxidase for 5 min using Endogenous Blocking Solution (Dako, Carpinteria, CA), washed twice, and incubated for 5 min in Serum-Free Protein Block (Dako), followed by incubation for 60 min in purified goat anti-human PD-1 antibody (R&D Systems, Minneapolis, MN) diluted 1:40. The specificity of the antibodies has been validated previously [16]. Immunohistochemical stains for CD3 (Novocastra, Bannockburn, IL), CD4 (Novocastra), and CD8 (Dako) were used to assist with the identification of the T-cell populations in all cases. PD-1 expression was determined semiquantitatively by two pathologists (D.A.W. and N.I.C.). The proportion of PD-1-positive cells in the neoplastic cell population of each sample that stained at moderate intensity or greater was approximated to the nearest fifth percentile. Furthermore, cases were counted as positive for PD-1 if staining of the cell surface and cytoplasm was of at least moderate intensity in greater than 30% of the lymphoid infiltrate. This cutoff was chosen because, in reactive populations, demonstration of PD-1 expression by the neoplastic infiltrate might be relevant when considering this marker as a therapeutic target. Atypical T cells localized to the epidermis and dermis were included in the assessment. If the epidermal population of lymphoma cells was not prominent or if this assessment was not applicable (tumor mycosis fungoides, Sézary syndrome), the atypical dermal population was scored. Malignant T cells were identified by the expression of the relevant T-cell receptor Vβ chain using fluorochrome-conjugated anti-Vβ antibodies (Beckman Coulter, Fullerton, CA). For cases in which T-cell receptor Vβ usage was unknown, malignant T cells were identified as CD4+/CD7− cells. Fluorochrome-conjugated isotype control and PD-1 monoclonal antibodies were obtained from BD Biosciences (Bedford, MA). Cells were analyzed on a FACSCalibur instrument (BD Biosciences) and analyzed using CellQuest or FACSDiva Software (BD Biosciences).