Introduction

Human solid tumors are often infiltrated by CD4+ and CD8+ lymphocytes, and numerous tumor-associated antigens have been discovered that can stimulate these cells after presentation by antigen-presenting cells in the context of the appropriate MHC molecules, inducing tumor-specific cytokine production or tumor cell lysis [1]. However, tumor progression is frequently seen in the presence of substantial lymphocytic infiltration, suggesting that this immune response is apparently incapable of controlling tumor growth. Escape mechanisms that might explain the lack of effectiveness of antitumor immune reactions include down-regulation of tumor-associated antigens (TAA) or MHC class I molecules, insufficient presentation of antigenic peptides, lack of costimulation, or inhibition of tumor-reactive T cells by immune suppressive factors or cells (reviewed in ref. [2]).

Numerous suppressor cell types have been identified that could contribute to the decrease in antitumor immune functions, including, among others, regulatory T cells, myeloid-derived suppressor cells and plasmacytoid dendritic cells [3]. The best characterized cell type is CD4+CD25+FOXP3+ regulatory T (Treg) cells, which comprise a functionally unique population of T lymphocytes controlling proliferation and function of other CD4+ as well as of CD8+ T cells, thus mediating peripheral tolerance. Natural Tregs arise in the thymus, but adaptive Tregs may also develop by conversion of CD4+CD25−FOXP3− cells at the periphery. Regulatory T cells are crucial in preventing autoimmune responses, and are implicated in a broad spectrum of autoimmune diseases, as well as in allograft rejection, graft-versus-host disease and allergy.

The role of Treg cells in cancer has recently been the focus of intense investigation. These cells have been found in increased number in the peripheral blood of cancer patients compared to healthy controls, and their accumulation in the tumor tissue has been demonstrated in the case of many cancer types (reviewed in ref. [4]).

Although a high number of functional regulatory T cells in the tumor microenvironment may be indicative of a local suppression of T cell immunity, the pathophysiological importance of Treg accumulation in human malignancies has remained questionable. Reports on the prognostic role of tumor-infiltrating Tregs are inconclusive; while in some cancer types their increased level has been correlated with poor outcome, as in the case of ovarian, breast, pancreatic and hepatocellular carcinomas [5–9], in others including prostate and renal cancers no significant associations were found [10, 11], and in the case of head and neck carcinomas, colorectal cancer, as well as in several lymphoma types a marked infiltration by Tregs showed correlation with improved survival [12–17].

Few studies have addressed the presence of regulatory T lymphocytes and its potential clinical correlates in melanoma, most of them comprising data on a limited number of patients and yielding controversial results. Mourmouras et al characterized cutaneous melanocytic lesions including melanomas with respect to the prevalence of Tregs, and found a lower percentage of Tregs in vertical than in radial growth phase melanomas [18]. Contrary to this finding, another small-scale study described higher prevalence of these cells in VGP compared to RGP, and also in stage III–IV tumors compared to stage I ones [19]. Miracco et al, in a study performed on vertical growth phase melanomas, reported that high intra- and peritumoral percentage of Tregs predicted local recurrence [20]. On the other hand, no correlation of the number of FOXP3+ lymphocytes with Breslow index or patient survival was found by Hillen et al [21]. Thus far, no systematic studies were performed correlating the extent of Treg infiltration of primary cutaneous melanoma with several clinicopathologic factors in large patient cohorts, providing conclusive data on the clinical relevance of these cells.

In a previous study examining the prevalence of activated T cells in cutaneous melanomas, we found a decreased peritumoral infiltration by T lymphocytes expressing CD25 or OX40 in tumors developing distant metastases during a follow-up period of 5 years, compared to nonmetastatic or lymph node metastatic ones. Accordingly, high peritumoral densities of CD25+ or OX40+ lymphocytes were associated with longer survival of the patients [22]. Furthermore, we also identified the density of DC-LAMP+ mature DCs, as well as combinations of high peritumoral CD1a+ or DC-LAMP+ cell densities with high numbers of CD25+ or OX40+ lymphocytes as prognostic factors in melanoma patients [23]. This suggests that the presence of activated T cells and antigen presenting DCs at the primary site could be a marker of a functional immune response against melanoma progression and influence the outcome of the disease. In the present study we examined the role of regulatory T cells, another subset of T lymphocytes with a presumed opposing function, in these processes. We investigated the prevalence of lymphocytes expressing the FOXP3 marker by immunohistochemistry in primary tumor samples from 97 patients with cutaneous melanoma, and evaluated with regard to its association with tumor thickness, the development of metastases, patients’ survival, and other clinicopathologic parameters.

Materials and Methods

Tumor Samples

Archival tissue samples were obtained from 97 patients with primary cutaneous melanoma who underwent surgery between 1980 and 2001 at the Institute of Dermato-Venerology, Semmelweis University, and at the National Institute of Oncology, Budapest. Patients were selected in order to obtain a study group involving a higher number of intermediate-thickness or thick (>1.0 mm) melanoma samples than their normal ratio, which have a more uncertain prognosis than thin tumors. The study was approved by the ethics committees of both institutions. Patients did not receive any anticancer treatment prior to surgery. Clinical and pathological characteristics are summarized in Table 1. The tumors were grouped into four thickness categories based on the AJCC staging system [24] (≤1.0, 1.01–2.0, 2.01–4.0, >4.0 mm), and into three categories according to disease progression (nonmetastatic, lymph node metastatic, visceral metastatic). Distribution according to stages [24] was: st. IA, 10; IB, 11; IIA, 17; IIB, 28; IIC, 14; IIIA, 14; IIIB, 3. Surviving patients had follow-up data for at least 5 years; none of the patients died of melanoma-unrelated causes within 5 years. Forty-five patients had no metastases developed during the follow-up period, while eight had metastases confined to regional lymph nodes, which were excised. Fourty-four patients developed distant visceral metastases. Five-year survival of patients in both the nonmetastatic and the lymph node metastatic groups was 100%, while only two patients developing distant visceral metastases survived for more than 5 years (62 and 72 months). Tumors with clinical regression and/or histological signs of extensive regression were not included in the study.

Table 1 Patient and tumor characteristics
Full size table

Immunohistochemical Detection of Infiltrating Cells

Three-μm sections cut from formalin-fixed, paraffin-embedded cutaneous melanoma samples were used. Immunohistochemistry was performed as described earlier [22], using monoclonal anti-FOXP3 (236A/E7; Abcam Inc., Cambridge, MA) primary antibody, followed by biotin/streptavidin-peroxidase method (LSAB2 System, HRP; Dako, Glostrup, Denmark) and visualization with 3-amino-9-ethylcarbazole (Vector Laboratories, Inc., Burlingame, CA).

Evaluation of the Immune Reactions

Slides were examined using a graticule of 10 × 10 squares, calibrated as 0.25 mm2 at 200× magnification. Because the distribution of stained cells was heterogeneous, the entire tumor area was analyzed in every case, and density of positive cells/mm2 is given. Although a few larger cells with morphological features of melanoma cells also showed nuclear FOXP3 staining in some instances, the majority of labeled cells had lymphocytic morphology, and only these cells were counted. The number of FOXP3+ lymphocytes was registered separately in intratumoral (infiltrating melanoma cell nests) and peritumoral areas (distributed in the infiltrate along the margin and the base of melanomas). In 3 areas of the highest density of positive cells the ratio of FOXP3+ lymphocytes was also determined as a percentage of the total cell number in the given field, using high magnification (400×). The proportion of patients with significant density of FOXP3+ cells was calculated using cutoff values set up separately for intra- and peritumoral localization as well as for the ratio of positive cells (30 and 220 cells/mm2, and 11%, respectively), based on the median of the given variable in the whole patient group.

Statistical Analysis

Comparisons between cell densities in different tumor groups was made using the Mann-Whitney U test and Kruskal-Wallis test, while χ2 test was used for comparing the proportions of samples with high cell densities. Association between tumor thickness and cell density values was evaluated by the Pearson test. Univariate analysis of survival was performed by the Kaplan-Meier method, and the statistical analysis was carried out by the Mantel-Cox test. All statistics were calculated using the BMDP Statistical Software Pack.

Results

Detection of FOXP3+ Lymphocytes in Melanoma Samples

In the melanoma samples, nuclear staining for FOXP3 was detected predominantly in small lymphocytes (Fig. 1). Occasionally, larger cells with morphological features of melanoma cells with nuclear FOXP3 staining were noted. Few positive cells were found in the uninvolved skin or dermal areas.

Fig. 1
figure 1

FOXP3+ cells in primary melanoma sample. Picture was taken using ×20 objective

Full size image

The density of FOXP3+ lymphocytes was determined separately in intratumoral (inside melanoma cell nests) and peritumoral areas (surrounding the tumor deposits). Lymphocytes stained for FOXP3 were more abundant in the tumor stroma, with a six-fold difference in average compared to their intralesional density (mean±SD, 292.2 ± 211.5 vs. 48.8 ± 63.2 cells/mm2). In three areas of the highest density of positive cells the ratio of FOXP3+ lymphocytes was also calculated as a percentage of the total cell number in the given field, and was estimated as 11.9 ± 5.8% for the whole patient group.

Correlations Between FOXP3+ Cell Density and Clinicopathologic Parameters

The degree of either intra- or peritumoral infiltration by FOXP3+ lymphocytes, or the ratio of positive cells did not show significant correlation with Breslow index (r = −0.1285, r = −0.1193 and r = −0.0048, respectively; p > 0.05). Although intermediate thickness tumors (between 1 and 4 mm) displayed increased intratumoral density (p = 0.0039 by Kruskal-Wallis test), no such differences were found in the case of either peritumoral density or the ratio of positivity (Fig. 2a). Moreover, no difference was found in FOXP3+ cell number between primary tumors according to their metastasis formation during a follow-up period of 5 years (p = 0.8292, p = 0.2337 and p = 0.1803 for intra-, peritumoral density and positivity ratio, respectively; Fig. 2b).

Fig. 2
figure 2

Intra- and peritumoral density, and the highest ratio of FOXP3+ cells (mean±SE), according to melanoma thickness (a) and metastasis formation (b)

Full size image

For intra- and peritumoral densities, as well as for the ratio of FOXP3+ cells, we calculated the proportions of samples with values higher than the median of the given variable in the whole patient group (30 and 220 cells/mm2, and 11%, respectively). Analyzing the relationship between the ratio of samples with high FOXP3+ cell numbers and tumor thickness or the metastatic pattern of the tumor, no significant differences were observed, with the exception of a higher frequency of samples with elevated intratumoral density of labeled cells in the groups of intermediate thickness tumors (Table 2). The distribution of melanomas with high FOXP3+ cell densities was also analyzed according to other clinicopathologic factors as patient age and gender, tumor location and the presence or absence of ulceration, with no associations found in the case of any of these parameters (Table 2). Concerning histological type, SSM and NM cases were compared only in the >2.0 mm thickness categories, since only 2 nodular melanomas were included in the thinner tumor groups; in these cases no difference in FOXP3+ cell content was found.

Table 2 Proportion of patients with high FOXP3+ cell density according to patient and tumor characteristics
Full size table

Survival Analysis According to FOXP3+ Cell Density

To evaluate the prognostic impact of the infiltration by FOXP3+ lymphocytes, we performed Kaplan-Meier analysis using the medians as cutoff levels as described above. Neither intra- or peritumoral densities, nor the ratio of positive cells showed any significant association with the survival of the patients (p = 0.4890, p = 0.6245 and p = 0.2711, respectively). The percentage of patients with more than 5 years survival was similar irrespective of the amount of FOXP3+ cells (60% vs. 53%, p = 0.5327, 59% vs. 54%, p = 0.6181 and 51% vs. 62%, p = 0.2773, for high or low intra- and peritumoral cell densities, and high or low positive cell ratios, respectively).

Discussion

In the present report, we determined the density of FOXP3+ lymphocytes in primary tumors of patients with malignant melanoma, and analyzed in relation to clinicopathologic parameters.

FOXP3+ cells were more prevalent in the stromal, lymphocyte-rich areas than in tumor cell nests, similarly to observations on other tumor types [6, 12]. Occasional positive staining of larger cells with morphological features of melanoma cells could also be found, in accordance with observations by Ebert et al [25]. Nevertheless, the overwhelming majority of FOXP3 staining was detected in small lymphocytes. We did not make attempts to characterize the labeled cell type(s) more precisely by double immunohistochemical staining. Although the expression of FOXP3 could be induced in vitro in CD4+CD25− and CD8+ lymphocytes as well [26–29], in previous immunohistochemical studies on melanoma all FOXP3+ cells expressed CD3 and almost all of them CD4 and CD25 as well [18, 19], suggesting that the vast majority of tumor-infiltrating FOXP3+ cells are CD4+CD25+ T lymphocytes.

The use of archival samples in our study did not allow the examination of the functional characteristics of the labeled cells. Since CD25 is expressed on activated T lymphocytes beside Tregs, FOXP3 has been considered the most reliable marker for immunohistochemical determination of the latter cells. It is a transcription factor critical for the development and function of regulatory T cells [30–33], and its expression is generally considered to define Tregs with suppressive capability [31–34]. On the other hand, in recent studies it became clear that the role of human FOXP3 is different from that of its murine counterpart, being a consequence of activation status, and responsible for an anergic phenotype, but not necessarily for a regulatory function of cells that express it [4, 26, 28, 34–36]. Nevertheless, a suppressive capacity of CD4+CD25highFOXP3+ cells isolated from human tumors was demonstrated by Kryczek et al. [37].

In any case, FOXP3-positive lymphocytes infiltrating primary human melanomas do not appear to exert a suppressor activity that would be reflected in the later course of the disease. Although their presence seemed to make a difference in terms of local recurrence of VGP melanomas [20], it did not influence either the development of metastases or survival of melanoma patients according to our study. With regard to the lack of effect on survival, our results corroborate those of an earlier report [21]. To the best of our knowledge, however, ours is the first study that systematically analyzed the associations of Treg infiltration with several clinicopathologic factors in a large patient cohort, in an attempt to provide conclusive data on the clinical relevance of these cells in melanoma. Since the amount of FOXP3+ cells did not show association with any of the studied parameters: patient age or gender, tumor thickness, location, histological type, ulceration and, most importantly, the outcome of the disease, as mentioned above, FOXP3+ cell infiltration in the primary tumor does not seem to have a significant impact on the course of the disease in cutaneous melanoma. Previous studies aiming at the clinical relevance of tumor-infiltrating Tregs have given rise to contradictory results, showing correlation with poor outcome in some cancer types [5–9] but no association or even correlation with improved survival in others [10–17, 21]. It could be suggested that each tumor type behaves differently in this context [4], depending perhaps on other players on the scene of antitumor immune reactions and their complex functional interrelationships with Tregs as well as with tumor cells.

In conclusion, we have shown that the density of FOXP3+ lymphocytes in the primary tumor is not significantly associated with clinicopathologic parameters and has no prognostic value in patients with cutaneous melanoma.