CD38 Expression by Circulating and Skin-Infiltrating Lymphocytes from Sezary Syndrome Patients: A Flow Cytometry and Immunohistochemistry Study

Background Reports on the expression of CD38 in Sézary syndrome (SS), erythrodermic primary cutaneous T cell lymphoma with leukemic involvement, are limited. The aim of the present study is the analysis of the expression of CD38 by skin-infiltrating mononuclear cells and circulating T lymphocytes in a cohort of SS patients. Methods SS patients diagnosed since 1985 in our clinic were retrospectively analyzed for CD38 expression in biopsy and blood samples by immunohistochemistry and flow cytometry, respectively. Results SS patients show a predominant CD38-negative phenotype on both skin and blood. A subgroup of patients was found expressing CD38 (12 cases) in either the skin (>25% cell infiltrate) or blood (CD4+CD38+ >50%), among whom 4 in the blood, 7 in the skin, and 1 in both blood and skin. Conclusion The implications of these observations may be twofold: the relevance in basic science is related to a potential role in immune defense regulation, whilst in perspective CD38 may become a target for antibody therapy, considering the availability of different anti-CD38 monoclonal antibodies.

CD38 is a type II transmembrane glycoprotein expressed by different cells without lineage restriction [8]. The function of the CD38 molecule is pleiotropic, initially defined as receptor able to transduce activation signals; it was then considered as an adhesion molecule with CD31 (PECAM-1) as counterreceptor. Another function of CD38 is to act as an ectoenzyme, able to metabolize extracellular nicotinamide adenine dinucleotide (NAD+) and to produce adenosine diphosphate ribose (ADPR) and cyclic adenosine diphosphate ribose (cADPR), both cytoplasmic second messengers. Through the generation of ADPR, CD38 takes part in the production of extracellular adenosine (ADO), a potent tolerogenic compound acting as potent immune suppressor [8].
Reports on the CD38 expression by circulating or skininfiltrating lymphoid cells in SS are very limited. In our previous report [2], CD38 expression on peripheral blood lymphocytes (PBL) from SS patients was significantly lower than in healthy subjects.

Question Addressed
The aim of the present study is to analyze the expression of CD38 in skin-infiltrating mononuclear cells and circulating T lymphocytes from SS patients.
The relevance of the analysis of CD38 in SS patients is twofold. The expression of CD38 is associated with the production of the immune-suppressor molecule adenosine, which could therefore represent one of the mediators of immune downregulation in these patients [8]. Moreover, it is well known that SS cells lack CD26 expression [2][3][4][5][6]. CD26, a transmembrane glycoprotein with dipeptidyl peptidase IV (DPPIV/CD26) activity, represents the main cellular binding protein for adenosine deaminase (ADA), the enzyme involved in adenosine degradation [9]. The absence of CD26 may determine a reduced degradation of adenosine with upregulation of immune-suppressive signals.
The expression of CD38 could be also relevant as a potential therapeutic target (anti-CD38 antibodies daratumumab and isatuximab approved in the US, Europe, and Italy for the treatment of multiple myeloma) [10,11].

Materials and Methods
3.1. Patients. Seventy-six SS patients were extracted from the Cutaneous Lymphoma Database of the Dermatologic Clinic, University of Turin, which includes all the cases of primary cutaneous lymphoma diagnosed since 1985.
SS diagnosis was established based on the criteria available at the time of each patient presentation using a combination of clinical, histologic, and immunohistochemical findings; all the cases were reclassified according to the WHO/EORTC classification [1,6]. Inclusion criteria for this retrospective study were represented by the availability of the analysis of CD38 expression at diagnosis either in the blood and/or in the skin and at least 30% of circulating CD4+CD26-cells, to avoid potential biases associated with the presence of limited blood involvement [12].
The study was performed according to the principles of good clinical practices and the Declaration of Helsinki principles. Study approval was obtained from the local ethical committee of the "Azienda Ospedaliera Città della Salute e della Scienza di Torino, Italy." 3.2. Immunohistochemistry. Immunohistochemistry on frozen sections was performed by the standard streptavidinbiotin-peroxidase method (LSAB2 plus kit; Dako, Glostrup, Denmark) using a wide panel of monoclonal antibodies. CD38 expression was detected using the Leu-17 clone (BD Biosciences Europe). CD38 expression in biopsy samples taken at diagnosis was available in 63 out of 76 patients. All samples were evaluated by two independent pathologists and scored as follows: CD38 expression in ≤5% of the lymphocyte infiltrate = negative (-); in >5-10% = almost negative (+/-); in >10%-25% = weakly positive (+); in >25% = positive (++). All 76 patients had blood sample available at diagnosis. A total of 87 blood samples taken from age-and sexmatched healthy donors were included as normal controls.
The procedures for flow cytometry tests and analyses have been previously reported [2]. For CD38 expression, anti-CD38-PE or PE-Cy7 (clone HB-7) was purchased by BD Biosciences, San Josè, CA, USA. Samples were considered CD38-positive for this study purposes if the CD4+CD38+ cell subset was >50% of total circulating lymphocytes, according to the maximum percentage of CD4+CD38positive cells from healthy donors (48%) and to literature data [13].
To investigate simultaneously the CD38 expression in the CD4+CD26-circulating subset, blood samples from 4 SS patients in follow-up were stained with a combination of anti-CD4-FITC (clone MT-130, Dako), anti CD26-PE (clone L272, BD), and anti-CD38-APC (clone IB4, Aczon) for 30 min at 4°C. Then, erythrocytes were lysed using a standard procedure and samples were acquired and analyzed by a FACSCanto 1 and Diva software. In 3 of these patients, PD-1 expression (CD279/MIH4 clone, PE conjugated, BD Pharmingen) was evaluated on the CD4+ subset.

Statistical Analysis.
The results are presented as median and 25th and 75th percentile. The Mann-Whitney U test and the Kruskal-Wallis with Dunn post hoc test were used to compare data, giving similar results. P < 0:05 was considered statistically significant.

CD38 Expression in Skin
Biopsies. In 17 out of 63 skin samples available for analysis, CD38 was expressed by less than 5% of infiltrating lymphocytes, in 22 samples CD38 was expressed by 5% to 10%, in 16 by 10% to 25%, and in 8 by more than 25% of infiltrating lymphocytes. Therefore, in most patients (39/63; 62%), CD38 was absent or expressed in less than 10% of the lymphocyte infiltrates. Only 8 patients (10.5%) had a significant amount of CD38+ cells (>25%) in the context of the cell infiltrate ( Figure 1; Table 1).
To compare CD38 expression levels in patients and healthy controls, the expression of CD38 was measured as median fluorescence intensity (MFI) in CD4+CD26-CD38+ cells from 13 SS patients enrolled from 2015 onwards and 23 healthy subjects. The MFI of CD38 was slightly lower in 12 out of 13 patients as compared with that of CD4+CD26cells from healthy subjects (Figure 1 suppl).
The subpopulation of CD4+CD26-CD38+ cells was further analyzed by multiparametric flow cytometry in 4 SS patients during follow-up (Figure 2 suppl). All these patients had more than 90% CD4+CD26-circulating lymphocytes: two had less than 5% CD4+CD26-CD38+ cells, whilst the other two had 10.8% and 16% CD4+CD26-CD38+ cells, respectively, confirming the expression of CD38 in a fraction of neoplastic cells. PD-1 expression was constantly detected in CD4-positive Sézary cells in the same patients (Figure 3 suppl).
No differences in clinical features were found between the two groups according to age and gender: patients with CD38+ cells tended to show higher numbers of circulating Sézary cells as well as CD4/CD8 ratios, even if this difference was not statistically significant. No difference in overall survival was found between the two groups (median overall survival: 22.2 months, range 1.6 months-9 years in the CD38+ group; 25.2 months, range 2.9 months-12 years in the CD38-patients) ( Table 1 Supplementary).

Discussion
This paper reports the results of a retrospective analysis of CD38 expression in skin and blood samples from a series of 76 SS patients diagnosed, treated, and followed up at the Dermatologic Clinic of the University of Turin since 1985.
Our results confirmed that CD38 is not generally expressed by lymphoid cells in this group of patients, as reported in our previous study [2]. Moreover, a decreased expression of this marker was clearly confirmed by the significantly lower MFI of CD38 on the CD4+CD26-cells in SS patients with respect to HD. This observation is in keeping with the results recently shown also reported by Najidh et al. in 16 SS patients [6].
However, a new finding of the present work is that a small group of SS patients may express CD38, in both skin and blood lymphocytes at diagnosis (Table 1). Indeed, 12 patients expressed CD38 in more than 50% of CD4+ circulating lymphoid cells and/or 25% of skin infiltrates, among whom 4 in the blood, 7 in the skin, and 1 in both blood and skin. However, as the multiparametric flow cytometry evaluations of CD4+CD26-CD38+ cells were not routinely performed in these patients at the time of enrollment, we   Disease Markers are aware that a major limitation of this study is the lack of complete data concerning CD38 expression by neoplastic CD4+CD26-or TCR Vbeta-positive cells. The literature reports limited amount of data. A CD38+ phenotype was described in a single case who remarkably had no skin lesions with blood and bone marrow infiltration by atypical cells with ultrastructural features resembling Sézary cells lacking most T lymphoid markers (CD2, CD4, CD3, CD5, and CD8) [14]. In a small series of MF patients, CD38 expression was suggested to correlate with the stage, being higher in the skin infiltrate of patients with tumor stage with respect to that found in patients with patch/plaque stage disease [15]. Torrealba et al. found an increase in the levels of soluble CD38 in SS patients [16].

Disease Markers
The implications of the conclusions inferred from the results of this study are twofold. Firstly, from a biologic point of view, special attention deserves the relationship between CD38 and CD26 and immunity. The expression of CD38 (potentially involved in increased adenosine production) coupled with the lack of CD26 (leading to reduced adenosine degradation) [9] could generate an imbalance between the generation and degradation of adenosine contributing to marked local extracellular ADO increase thus switching the microenvironment towards immunosuppression. ADO is an immunosuppressive molecule with multiple functions [17], which is generated through the concerted action of surface molecules endowed with enzymatic functions. Two adenosinergic pathways exist: the canonical pathway is started by nucleoside-triphosphate diphosphohydrolase/CD39 that converts ATP to AMP. AMP is then fully converted to ADO by ecto-5 ′ -nucleotidase/CD73 [18], which mediates its immunoregulatory functions by binding to one of four adenosine receptors (AR). The noncanonical pathway is started by nicotinamide adenine dinucleotide (NAD+)-glycohydrolase/CD38 that converts NAD+ to ADPR. ADPR is then converted by CD203a (PC-1) to AMP that is subsequently metabolized to ADO by CD73 [19]. Based on the results of this study, it is tempting to speculate that in a subgroup of SS patients CD38 could mediate ADO production by the noncanonical pathways thus contributing to local immunosuppression. SS patients are characterized by a severe impairment of the immune system, which is intrinsic to the disease, is worsened by chemotherapeutic treatments, and gives rise to a high incidence of infections [1]. In this scenario, the high expression of PD-1 by neoplastic cells which is an established feature of this disease [3] also found in our study could represent a further mechanism contributing to the generation of an immune-suppressive microenvironment. The PD-1/ PD-L1 axis is associated with the upregulation of an immune-suppressive T cell response; hence, its inhibition is currently being explored to enhance antitumor T cell immunity with certain successful outcomes. However, PD-1 has also emerged as a central tumor suppressor in T cell lymphomas [20], suggesting that checkpoint inhibitors might have the potential to accelerate expansion of tumoral T cell clones. Similarly, we cannot exclude that an adenosine-rich milieu-due to CD38 expression in a relevant fraction of neoplastic cells-exerts its inhibitory effects against SS themselves in an autocrine fashion. This aspect requires careful evaluation in order to understand the clinical significance of CD38 in SS, although increasing evidence indicates that, regardless of the tumor cell type expressing CD38, its effect is immunosuppressive and tumor-promoting [21].
The second point of interest is related to CD38 expression by SS as potential therapeutic target in a subgroup of patients. Today, the armamentarium of anti-CD38 antibodies includes daratumumab and recently isatuximab, both FDA approved in myeloma therapy [9,10]. It is tempting to envision that the blockade of the CD38 enzymatic pathway could exert a positive impact on the immune regulation processes and promote the restoring of the immune response.
The results of the present study should be confirmed in larger patient cohorts studying CD38 expression in the neoplastic clonal cells characterized by the lack of CD7/CD26 and expression of a single Vbeta chain. Another important issue to address in the future is the detailed characterization of the phenotypical features of the CD38-positive subgroup of patients and the definition of the potential biological significance of CD38 expression in the production of ADO. Overall, these preliminary data could represent the starting point for the development of clinical trials using anti-CD38 antibodies alone or in combination with anti-PD-1 for the treatment of the small subgroup of SS patients who express CD38 and would likely benefit from such targeted therapy.

Data Availability
The Excel data used to support the findings of this study are available from the corresponding author upon request.