Advertisement

Indoleamine 2,3-Dioxygenase Expression in Primary Cutaneous Melanoma Correlates with Breslow Thickness and Is of Significant Prognostic Value for Progression-Free Survival

Open ArchivePublished:October 17, 2017DOI:https://doi.org/10.1016/j.jid.2017.09.036
      The enzyme indoleamine 2,3-dioxygenase (IDO) is emerging as a facilitator of cancer development through its effects on cancer-associated inflammation. Recent studies report a significant improvement of the response rates in melanoma patients to PD-1 antibodies when IDO inhibitors were added to the regimen. Data on IDO expression in primary human melanomas are, however, incomplete and conflicting. Here, we show that the level of IDO expression in primary human melanoma cells significantly correlates with Breslow thickness (P = 0.003), the presence of tumor-infiltrating lymphocytes (P = 0.029), and the intensity of the peritumoral inflammatory infiltrate (P = 0.001). The expression of IDO in melanoma cells predicted independently of Breslow thickness and tumor stage (P = 0.04). We further show that CD11c+ dendritic cells and CD68+ macrophages in the microenvironment of melanomas express IDO. The level of IDO expression in antigen-presenting cells correlated positively to peritumoral inflammation (P = 0.001) but not to tumor-infiltrating lymphocytes. Significant negative correlation with progression-free survival was found for patients for whom antigen-presenting cells were very strongly IDO positive. These results suggest that IDO induction within melanoma cells may directly reflect tumor progression, whereas IDO in antigen-presenting cells may determine immune surveillance with impact on local and systemic tolerance.

      Abbreviations:

      APC (antigen-presenting cell), DC (dendritic cell), IDO (indoleamine 2,3-dioxygenase), mAb (monoclonal antibody), OS (overall survival), PFS (progression-free survival), TIL (tumor-infiltrating lymphocyte)

      Introduction

      The interaction between melanoma cells and infiltrating immune cells or dermal cells is an important determinant in the development of malignant melanoma (
      • Das M.
      • Zhu C.
      • Kuchroo V.K.
      Tim-3 and its role in regulating anti-tumor immunity.
      ,
      • Ziani L.
      • Safta-Saadoun T.B.
      • Gourbeix J.
      • Cavalcanti A.
      • Robert C.
      • Favre G.
      • et al.
      Melanoma-associated fibroblasts decrease tumor cell susceptibility to NK cell-mediated killing through matrix-metalloproteinases secretion.
      ). Chronic inflammation drives the development of many cancers, but an exact definition of what these specific factors are has been difficult to determine (
      • Holzel M.
      • Tuting T.
      Inflammation-induced plasticity in melanoma therapy and metastasis.
      ). The two situations, progression of treatment-naïve melanoma and relapse of treated melanoma (e.g., after checkpoint inhibitor therapy) are both under the influence of the immune system. In both cases, the selection of tumor cell clones with high antigen plasticity and the induction of T-cell tolerance may, in part, account for the progression of the disease (
      • Landsberg J.
      • Kohlmeyer J.
      • Renn M.
      • Bald T.
      • Rogava M.
      • Cron M.
      • et al.
      Melanomas resist T-cell therapy through inflammation-induced reversible dedifferentiation.
      ). Regulatory signals from the tumor, in turn, are important for stimulating or inhibiting T cells.
      One such regulatory signal is the expression of the tryptophan-catabolizing enzyme indoleamine 2,3-dioxygenase (IDO) (
      • Prendergast G.C.
      • Metz R.
      • Muller A.J.
      Towards a genetic definition of cancer-associated inflammation: role of the IDO pathway.
      ). IDO is an intracellular heme-containing enzyme that initiates the first and rate-limiting step of tryptophan degradation along the kynurenine pathway (
      • Routy J.P.
      • Routy B.
      • Graziani G.M.
      • Mehraj V.
      The kynurenine pathway is a double-edged sword in immune-privileged sites and in cancer: implications for immunotherapy.
      ). The expression of IDO by mesenchymal stromal cells, fibroblasts, and various myeloid-derived antigen-presenting cells (APCs) such as dendritic cells (DCs) and macrophages has been shown to block T-cell responses in a number of situations such as in immune tolerance toward allogeneic fetuses and the inhibition of graft rejection (
      • Chen G.
      • Kheradmand T.
      • Bryant J.
      • Wang S.
      • Tasch J.
      • Wang J.J.
      • et al.
      Intragraft CD11b(+) IDO(+) cells mediate cardiac allograft tolerance by ECDI-fixed donor splenocyte infusions.
      ,
      • Li Y.
      • Tredget E.E.
      • Ghaffari A.
      • Lin X.
      • Kilani R.T.
      • Ghahary A.
      Local expression of indoleamine 2,3-dioxygenase protects engraftment of xenogeneic skin substitute.
      ,
      • Munn D.H.
      • Zhou M.
      • Attwood J.T.
      • Bondarev I.
      • Conway S.J.
      • Marshall B.
      • et al.
      Prevention of allogeneic fetal rejection by tryptophan catabolism.
      ,
      • von Bubnoff D.
      • Bausinger H.
      • Matz H.
      • Koch S.
      • Hacker G.
      • Takikawa O.
      • et al.
      Human epidermal Langerhans cells express the immunoregulatory enzyme indoleamine 2,3-dioxygenase.
      ). On the contrary, IDO expression by cancer cells can substantially add to immune evasion by tumors (
      • Godin-Ethier J.
      • Hanafi L.A.
      • Piccirillo C.A.
      • Lapointe R.
      Indoleamine 2,3-dioxygenase expression in human cancers: clinical and immunologic perspectives.
      ,
      • Prendergast G.C.
      • Smith C.
      • Thomas S.
      • Mandik-Nayak L.
      • Laury-Kleintop L.
      • Metz R.
      • et al.
      Indoleamine 2,3-dioxygenase pathways of pathogenic inflammation and immune escape in cancer.
      ).
      IDO expression primarily affects T-cell activity but also APC function. Tryptophan deficiency specifically activates the GCN2 kinase in T cells and prevents T-cell activation (
      • Munn D.H.
      • Sharma M.D.
      • Baban B.
      • Harding H.P.
      • Zhang Y.
      • Ron D.
      • et al.
      GCN2 kinase in T cells mediates proliferative arrest and energy induction in response to indoleamine 2,3-dioxygenase.
      ). DCs, when generated in low tryptophan conditions in vitro, show low T-cell stimulatory capacity because of the induction of CD4+CD25+Foxp3+ T regulatory cells (
      • Brenk M.
      • Scheler M.
      • Koch S.
      • Neumann J.
      • Takikawa O.
      • Hacker G.
      • et al.
      Tryptophan deprivation induces inhibitory receptors ILT3 and ILT4 on dendritic cells favoring the induction of human CD4+CD25+ Foxp3+ T regulatory cells.
      ). In vitro, after activation with CD40L, IDO-positive APCs express increased levels of B7 family molecules such as CD40 and CD80, as well as PD-L1 (B7-H1) and PD-L2 (B7-DC), compared with IDO-negative DC subset (
      • von Bubnoff D.
      • Scheler M.
      • Wilms H.
      • Fimmers R.
      • Bieber T.
      Identification of IDO-positive and IDO-negative human dendritic cells after activation by various proinflammatory stimuli.
      ). It is therefore conceivable that in the tumor microenvironment, a complex process of immune editing takes place where IDO+CD80+CD86+PD-L1/PD-L2+ APCs and tumor cells contribute to the induction of tolerance in CTLA-4+PD1/2+ T cells.
      In melanoma but also in other cancers, it has been shown that these inhibitory pathways are strongly up-regulated and affect the development of an immune response in vivo. For instance, injection of stimulatory CTLA-4 Ig into mice induced IDO expression in specific DC subsets of the spleen by reverse signaling via CD80/86 into DCs. The induction of IDO in this setting prevented the development of a specific CD8 T-cell response (
      • Mellor A.L.
      • Baban B.
      • Chandler P.
      • Marshall B.
      • Jhaver K.
      • Hansen A.
      • et al.
      Cutting edge: induced indoleamine 2,3 dioxygenase expression in dendritic cell subsets suppresses T cell clonal expansion.
      ). When the CTLA-4–blocking antibody ipilimumab was injected into B16 melanoma-bearing mice, IDO-deficient animals showed a striking delay in B16 melanoma tumor growth (
      • Holmgaard R.B.
      • Zamarin D.
      • Munn D.H.
      • Wolchok J.D.
      • Allison J.P.
      Indoleamine 2,3-dioxygenase is a critical resistance mechanism in antitumor T cell immunotherapy targeting CTLA-4.
      ). The delay was associated with an increased number of tumor-infiltrating CD8+ and CD4+ effector T cells. Blockade of CTLA-4 in melanoma thus not only releases the checkpoint blockade on T cells but may also block induction of IDO in APCs.
      In human melanoma patients and in mouse models, IDO has been shown to be expressed by APCs in tumor draining lymph nodes (
      • Gerlini G.
      • Di Gennaro P.
      • Mariotti G.
      • Urso C.
      • Chiarugi A.
      • Pimpinelli N.
      • et al.
      Indoleamine 2,3-dioxygenase+ cells correspond to the BDCA2+ plasmacytoid dendritic cells in human melanoma sentinel nodes.
      ,
      • Sharma M.D.
      • Hou D.Y.
      • Liu Y.
      • Koni P.A.
      • Metz R.
      • Chandler P.
      • et al.
      Indoleamine 2,3-dioxygenase controls conversion of Foxp3+ Tregs to TH17-like cells in tumor-draining lymph nodes.
      ). An independent negative prognostic effect of IDO expression in the sentinel lymph node on progression-free survival (PFS) and overall survival (OS) in patients with stage I and II melanoma has been reported (
      • Speeckaert R.
      • Vermaelen K.
      • van Geel N.
      • Autier P.
      • Lambert J.
      • Haspeslagh M.
      • et al.
      Indoleamine 2,3-dioxygenase, a new prognostic marker in sentinel lymph nodes of melanoma patients.
      ). IDO expression could be a very early determinant of immune instruction in melanoma development, because IDO expression in tumor negative sentinel lymph nodes confers a negative prognostic value for patients (
      • Chevolet I.
      • Speeckaert R.
      • Haspeslagh M.
      • Neyns B.
      • Kruse V.
      • Schreuer M.
      • et al.
      Peritumoral indoleamine 2,3-dioxygenase expression in melanoma: an early marker of resistance to immune control?.
      ).
      We provide a detailed semiquantitative analysis of IDO expression in primary human melanoma cells and in cells of the peritumoral microenvironment from skin biopsy samples of melanoma patients. Findings were correlated to melanoma Breslow index, histopathologic parameters, peritumoral inflammation, PFS, and OS. Because IDO inhibitors have been shown to enhance tumor immunity in vivo (
      • Brochez L.
      • Chevolet I.
      • Kruse V.
      The rationale of indoleamine 2,3-dioxygenase inhibition for cancer therapy.
      ), these data are of potential value should immunotherapy be established at earlier stages in melanoma.

      Results

      Patient characteristics according to tumor thickness (groups 1–3)

      A total of 99 patients with primary melanoma were divided into three groups according to tumor thickness as outlined in the Materials and Methods section; characteristics are shown in Table 1. The follow-up time was 15 years (mean = 8 years). Patient numbers, sex, and mean age did not differ significantly, although female patients tended to have earlier melanoma development than male patients. As expected, ulceration, microsatellitosis, lymphangioinvasion and tumor-related death were most often seen in group 3.
      Table 1Patient characteristics according to melanoma thickness
      Group 1, tumor thickness < 0.75 mm; group 2, tumor thickness = 0.75–1.00 mm; group 3, tumor thickness ≥ 4 mm. A total of 99 patients were included in this study
      Clinical FeaturesAll GroupsGroup 1Group 2Group 3
      Number of patients, n (%)99 (100)32 (100)31 (100)36 (100)
      Sex, n (%)
       Male46 (46)12 (38)15 (48)19 (53)
       Female53 (53)20 (63)16 (52)17 (47)
      Age in years
       Range35–9846–9247–9835–90
       Mean68717067
      Follow-up time in months
       Range25.6–185.288.4–132.172.2–185.225.6–81.1
       Mean95.6111.1111.368.3
      Location of primary melanoma, n (%)
       Head and neck20 (20)7 (22)3 (10)10 (28)
       Trunk41 (41)16 (50)14 (45)11 (31)
       Extremities38 (38)9 (28)14 (45)15 (42)
      Melanoma characteristics
      P < 0.05.
       Breslow thickness in mm
      Range0.2–17.00.2–0.70.75–1.04.0–17.0
      Mean2.820.440.877.16
       Ulceration
      P < 0.05.
      , n (%)
      18 (18)0 (0)1 (3)17 (47)
       Regression, n (%)13 (13)2 (6)10 (32)1 (3)
       Lymphangioinvasion
      P < 0.05.
      , n (%)
      8 (8)0 (0)0 (0)8 (22)
       Microsatellitosis
      P < 0.05.
      , n (%)
      10 (10)0 (0)0 (0)10 (28)
      Evolution during follow-up, n (%)
       Alive74 (74)31 (97)27 (87)16 (44)
       Disease progression26 (26)2 (6)3 (10)21 (58)
       Dead25 (25)1 (3)4 (13)20 (56)
       Tumor-related death
      Yes10 (10)0 (0)1 (3)9 (25)
      No15 (15)1 (3)3 (10)11 (31)
      1 Group 1, tumor thickness < 0.75 mm; group 2, tumor thickness = 0.75–1.00 mm; group 3, tumor thickness ≥ 4 mm. A total of 99 patients were included in this study
      2 P < 0.05.

      The level of IDO expression in primary human melanoma significantly correlates to Breslow tumor thickness

      The expression of IDO was analyzed in all samples from groups 1, 2, and 3 (Figure 1a). In melanoma cells, very strong IDO expression (IDO+++) was seen in 25% of group 3 specimens and in 13% of group 2 and 6% of group 1 samples. Strong IDO expression (IDO++) was seen most frequently in samples of group 3 (36%); 39% of samples in group 3 were IDO negative (IDO) or of low IDO expression (IDO+), whereas this number was 64% in group 2 and 66% in group 1. The correlation of the level of IDO expression in primary melanoma cells to Breslow thickness was highly significant (P = 0.003). IDO expression in APCs did not correlate with Breslow thickness of primary melanoma (P = 0.35, data not shown). Thin melanomas with tumor thickness of less than 0.75 mm expressed IDO right at the site of cancer development (Figure 1b).
      Figure 1
      Figure 1IDO expression in primary human melanoma cells correlates with Breslow thickness. (a) The level of IDO expression in melanoma cells (IDO, IDO+, IDO++, IDO+++) correlates significantly and positively with the tumor thickness of the primary melanoma. P = 0.003. (b) Skin sections with cutaneous melanoma (n = 99) of various tumor thickness were stained for IDO. Representative images of immunohistochemistry are shown. Scale bar = 100 μm. IDO, indoleamine 2,3-dioxygenase; TD, tumor thickness.

      Patient characteristics, clinical and histopathological data, and their correlation with the level of IDO expression in melanoma cells and APCs

      Patients were stratified according to the level of IDO expression in melanoma cells to assess its association with patient characteristics and histopathological and clinical parameters (Table 2). Parameters that were positively and significantly correlated with the level of melanoma IDO expression were ulceration (P = 0.0023), lymphangioinvasion (P = 0.0068), and microsatellitosis (P = 0.01). The presence of tumor-infiltrating lymphocytes (TILs) was highly positively correlated with the level of IDO expression in tumor cells (P = 0.029) but not with the level of IDO APC expression (P = 0.298). The level of IDO APC expression correlated only positively with ulceration (P = 0.015). T-stage, N-stage, clinical stage category (American Joint Committee on Cancer 2009 staging classification), and tumor-related death was highest in IDO+++ patients. The level of IDO expression in primary melanoma cells appears to correlate with the established TNM classification for melanomas for the investigated stages.
      Table 2Patient characteristics and clinical and histological data according to the level of IDO expression in melanoma cells
      Clinical FeaturesIDOIDO+IDO++IDO+++
      Patients, n (%)35 (100)20 (100)29 (100)15 (100)
      Sex, n (%)
       Male13 (37)11 (55)15 (52)7 (47)
       Female22 (63)9 (45)14 (48)8 (53)
      Breslow thickness
      At time of primary diagnosis.
      ,
      P < 0.05.
      in mm
       Range0.2–12.010.25–10.00.2–17.00.45–15.1
       Mean2.012.203.585.38
      Mitotic rate
      At time of primary diagnosis.
      per mm2
       Range0–240–70–160–20
       Mean6.94.75.89.2
      Ulceration
      At time of primary diagnosis.
      ,
      P < 0.05.
      , n (%)
      3 (9)1 (5)8 (28)6 (40)
      Regression
      At time of primary diagnosis.
      , n (%)
      5 (14)3 (15)4 (14)1 (7)
      Lymphangioinvasion
      At time of primary diagnosis.
      ,
      P < 0.05.
      , n (%)
      0 (0)1 (5)4 (14)3 (20)
      Microsatellitosis
      At time of primary diagnosis.
      ,
      P < 0.05.
      ,
      Eight patients were not specified.
      , n (%)
      1 (3)0 (0)6 (21)3 (20)
      Tumor-infiltrating lymphocytes
      P < 0.05.
      , n (%)
       Absent12 (34)5 (25)3 (10)2 (13)
       Non-brisk13 (37)8 (40)12 (41)6 (40)
       Brisk10 (29)7 (35)14 (48)7 (47)
      IDO expression in APZs
      At time of primary diagnosis.
      , n (%)
       IDO19 (54)2 (10)2 (7)0 (0)
       IDO+5 (14)11 (55)4 (14)1 (7)
       IDO++9 (26)7 (35)18 (62)3 (20)
       IDO+++2 (6)0 (0)5 (17)11 (73)
      T stage
      At time of primary diagnosis.
      ,
      P < 0.05.
      , n (%)
       T126 (74)15 (75)16 (55)6 (40)
       T20 (0)0 (0)0 (0)0 (0)
       T31 (3)0 (0)1 (3)0 (0)
       T48 (23)5 (25)12 (42)9 (60)
      N stage
      At time of primary diagnosis.
      ,
      P < 0.05.
      , n (%)
       N034 (97)20 (100)22 (76)10 (67)
       N11 (3)0 (0)4 (14)2 (13)
       N20 (0)0 (0)3 (10)3 (20)
      M stage
      At time of primary diagnosis.
      , n (%)
       M035 (100)20 (100)29 (100)15 (100)
      AJCC
      At time of primary diagnosis.
      ,
      P < 0.05.
      , n (%)
       I26 (74)15 (75)16 (55)6 (40)
       II8 (23)5 (25)6 (21)4 (27)
       III1 (3)0 (0)7 (24)5 (33)
      Clinical outcome
      At the end of follow-up time.
      , n (%)
       Alive26 (74)17 (85)24 (83)7 (47)
       Disease progression8 (23)3 (15)6 (21)9 (60)
       Dead9 (26)3 (15)5 (17)8 (53)
       Tumor-related death
      Yes2 (6)2 (10)2 (7)4 (27)
      No7 (20)1 (5)3 (10)4 (27)
      Abbreviations: AJCC, American Joint Committee on Cancer; APC, antigen-presenting cell; IDO, indoleamine 2,3-dioxygenase.
      1 At time of primary diagnosis.
      2 P < 0.05.
      3 Eight patients were not specified.
      4 At the end of follow-up time.

      The intensity of IDO expression in primary melanoma cells and in APCs significantly correlates with the strength of the peritumoral infiltrate

      Because IDO is mainly induced by proinflammatory stimuli, we correlated the intensity of IDO expression in melanoma cells and APCs to the strength of the peritumoral inflammatory infiltrate (Figure 2a and b). First, most samples with moderate and intense infiltrates also showed substantial IDO expression (IDO++ and IDO+++) in melanoma cells (moderate, 62%; intense, 70%) and APCs (moderate, 76%; intense, 60%). Accordingly, 78% of samples with scarce infiltrate showed no IDO expression in melanoma cells and in APCs, respectively. The remaining 22% of samples with scarce infiltrate were associated with only mild (IDO+) melanoma cell IDO expression or with mild (11%) or moderate (11%) IDO expression in APCs. The correlation between the IDO expression in melanoma cells and in APCs with the density of the peritumoral inflammatory infiltrate was highly significant (P = 0.001).
      Figure 2
      Figure 2The intensity of IDO expression in melanoma cells and in APCs is significantly correlated to the strength of the peritumoral inflammatory infiltrate. Samples were stratified according to the strength of the inflammatory infiltrate and intensity of IDO expression. (a) IDO expression in melanoma cells significantly and positively correlates with the intensity of the peritumoral inflammatory infiltrate. P = 0.001. (b) IDO expression in APCs significantly and positively correlates with the strength of the inflammatory infiltrate. P = 0.001. APC, antigen-presenting cell; IDO, indoleamine 2,3-dioxygenase.

      CD68+ macrophages and CD11c+ DCs express IDO in the melanoma microenvironment

      The cells in the peritumoral inflammatory infiltrate that expressed IDO were aligned beneath the invasive front of the tumor cells (Figure 3a and g) and were peritumoral in foci of inflammation (Figure 3a and h). CD68+ macrophages and CD11c+ myeloid APCs most often encircled the abundant CD3+ T-cell inflammatory infiltrate (Figures 3c–e). Melanoma cells, epidermal Langerhans cells, and some myeloid cells in the infiltrate stained S100 positive (Figure 3f).
      Figure 3
      Figure 3Immunohistochemical staining for IDO, CD68, CD11c, CD3, and S100 in primary cutaneous melanoma. (a–f) Representative immunohistochemical pictures of an IDO-negative tumor but IDO-positive inflammatory infiltrate in primary melanoma. Note the IDO+ APCs located along the border of invasive melanoma. Original magnification × 40. Scale bar = 100 μm. Representative IDO+ primary melanoma with IDO-positive APCs (g) in inflammatory foci aside the vertical growth phase (h) beneath the vertical growth phase. Original magnification ×40. Scale bar = 100 μm. APC, antigen-presenting cell; IDO, indoleamine 2,3-dioxygenase.
      Double-immunofluorescence staining showed that CD68+ macrophages and CD11c+ myeloid DCs strongly expressed IDO (Figure 4).
      Figure 4
      Figure 4CD68+ macrophages and CD11c+ DCs express IDO in the peritumoral inflammatory infiltrate of primary cutaneous melanomas. Double immunofluorescence of primary melanoma for IDO with (a) CD11c and (b) CD68. Representative photographs of the single stainings are shown, as is the overlay of IDO (green) with individual markers (red). Double-positive cells are shown in yellow. Scale bar = 50 μm. DC, dendritic cell; IDO, indoleamine 2,3-dioxygenase.

      The level of IDO expression in primary melanoma cells and in peritumoral APCs shows significant negative correlation with PFS

      Clinical progression of melanoma within the first 4 years was especially pronounced in the patient group with very high IDO expression (IDO+++) in melanoma cells (Figure 5a). In line with this, the median PFS was reached by this patient group at 4 years of follow-up, whereas median PFS was not reached by the other groups. The between-group difference in PFS related to the level of IDO expression in melanoma cells was highly significant (P = 0.02). Because IDO expression positively correlates to tumor thickness and tumor thickness is known to predict PFS, we determined if IDO expression independently correlates with PFS. Covariate Cox regression analyses showed an independent correlation of IDO expression in melanoma cells with PFS (P = 0.04).
      Figure 5
      Figure 5The level of IDO expression in melanoma cells and APCs correlates with PFS. All 99 patients are shown. Of these, 26 patients showed progression during follow-up. (a) Patients with IDO+++ melanoma cells (20%) showed fast and early progression. Patients with progression and IDO ++ (19%), IDO + (23%), and IDO (27%) melanoma cells had slower tumor progression. P = 0.02. (b) Disease in patients with IDO+++ APCs (34%) progressed significantly faster compared with all other groups. P = 0.0095. The between-group difference in PFS related to the intensity of APC IDO expression was significant. P = 0.02. APC, antigen-presenting cell; IDO, indoleamine 2,3-dioxygenase; PFS, progression-free survival.
      With regard to the level of IDO expression in APCs, neither patient curve reached median PFS (Figure 5b). The patient curves with the strongest IDO expression (IDO+++) and with no IDO expression (IDO) in peritumoral APCs showed a fast decline during the first 2 years of follow up. The difference between the Kaplan-Meyer curve for patients with IDO+++ APCs compared with all other groups was statistically highly significant (P = 0.0095). In contrast, there was no significant difference in PFS for patients with IDO APCs compared with all other groups (P = 0.3078). Overall, the between-group differences in PFS related to the level of IDO expression in peritumoral APCs was highly significant (P = 0.02).

      The level of IDO expression in primary melanoma cells shows a strong trend toward correlation with OS

      Next, we evaluated the correlation of OS with IDO expression levels in primary melanoma and peritumoral APCs (Figures 6a and b). The point of median OS was not reached in any subgroup. Probability for OS decreased fastest during the first 5 years in patients with IDO+++ melanoma cells (Figure 6a). Statistical analysis just failed to show significance of a between-group difference for OS regarding IDO expression in melanoma cells (P = 0.08). Accordingly, Cox regression analyses did not show independent correlation of IDO expression in melanoma cells with OS, although a strong trend was observed (P = 0.08).
      Figure 6
      Figure 6Correlation of the level of IDO expression in melanoma cells and APCs with OS. Of 99 patients in total, 10 patients died of melanoma-related causes. (a) Four patients died in the IDO+++ melanoma group, and from each of the IDO++, IDO+, and IDO melanoma cell groups, two patients died. P = 0.08. (b) The numbers of patients who died during follow-up: three from the IDO+++, four from the IDO++, one from the IDO+, and two from the IDO APC groups. This difference was not statistically significant. P = 0.33. APC, antigen-presenting cell; IDO, indoleamine 2,3-dioxygenase; OS, overall survival.
      OS in patients with respect to varying IDO expression in APCs did not differ much within the first 5 years of follow-up but was lowest in the IDO+++ APC group after 6 years (80%) (Figure 6b). There was no significant difference between the varying degrees of IDO expression in APCs on OS (P = 0.33).

      Discussion

      In this study, we show a significant and positive correlation of the level of IDO expression in primary cutaneous melanoma cells with Breslow tumor thickness. In addition, the level of melanoma cell IDO expression was strongly correlated to well-known prognostic histopathologic parameters such as ulceration, mitotic rate, lymphangioinvasion, microsatellitosis, and the presence of TILs (
      • Rose C.
      [Diagnostics of malignant melanoma of the skin: Recommendations of the current S3 guidelines on histology and molecular pathology].
      ). Furthermore, IDO in melanoma cells carried a significant prognostic association with PFS, which was independent of Breslow thickness and tumor stage. Thus, our data further substantiate the literature showing the important immunoregulatory role of IDO in melanoma.
      Earlier observations showed that IDO expression in the sentinel node and in peritumoral endothelium of melanoma patients is associated with a lower PFS (
      • Speeckaert R.
      • Vermaelen K.
      • van Geel N.
      • Autier P.
      • Lambert J.
      • Haspeslagh M.
      • et al.
      Indoleamine 2,3-dioxygenase, a new prognostic marker in sentinel lymph nodes of melanoma patients.
      ,
      • Chevolet I.
      • Speeckaert R.
      • Haspeslagh M.
      • Neyns B.
      • Kruse V.
      • Schreuer M.
      • et al.
      Peritumoral indoleamine 2,3-dioxygenase expression in melanoma: an early marker of resistance to immune control?.
      ). In different cancer entities, IDO expression within tumor cells has been increasingly identified to be associated with poor prognosis (
      • Jia Y.
      • Wang H.
      • Wang Y.
      • Wang T.
      • Wang M.
      • Ma M.
      • et al.
      Low expression of Bin1, along with high expression of IDO in tumor tissue and draining lymph nodes, are predictors of poor prognosis for esophageal squamous cell cancer patients.
      ,
      • Kim J.W.
      • Nam K.H.
      • Ahn S.H.
      • Park do J.
      • Kim H.H.
      • Kim S.H.
      • et al.
      Prognostic implications of immunosuppressive protein expression in tumors as well as immune cell infiltration within the tumor microenvironment in gastric cancer.
      ,
      • Zhang T.
      • Tan X.L.
      • Xu Y.
      • Wang Z.Z.
      • Xiao C.H.
      • Liu R.
      Expression and prognostic value of indoleamine 2,3-dioxygenase in pancreatic cancer.
      ).
      • Brody J.R.
      • Costantino C.L.
      • Berger A.C.
      • Sato T.
      • Lisanti M.P.
      • Yeo C.J.
      • et al.
      Expression of indoleamine 2,3-dioxygenase in metastatic malignant melanoma recruits regulatory T cells to avoid immune detection and affects survival.
      showed strong melanoma cell IDO expression in metastatic lymph nodes and its association with poor survival (
      • Brody J.R.
      • Costantino C.L.
      • Berger A.C.
      • Sato T.
      • Lisanti M.P.
      • Yeo C.J.
      • et al.
      Expression of indoleamine 2,3-dioxygenase in metastatic malignant melanoma recruits regulatory T cells to avoid immune detection and affects survival.
      ).
      The presence of TILs in primary melanoma remains controversial regarding its prognostic value (
      • Weiss S.A.
      • Han S.W.
      • Lui K.
      • Tchack J.
      • Shapiro R.
      • Berman R.
      • et al.
      Immunologic heterogeneity of tumor-infiltrating lymphocyte composition in primary melanoma.
      ). The composition of the TIL infiltrate, either immunosupportive or immunosuppressive, may be one determinant of the prognosis of primary melanoma. In our study, the level of melanoma IDO expression, but not APC IDO expression, positively correlated with the gradation of TILs. APC IDO expression (and melanoma IDO expression), however, did correlate with the peritumoral infiltrate, which we determined as being located around the tumor.
      These results may have two implications. First, the quality of TILs, either immunosupportive or immunosuppressive, might be reflected by the expression of IDO in melanoma cells. In other words, the level of IDO expression in cancer cells could be an important factor to predict immunosuppressive TILs and indicate that loss of immune control has been initiated (
      • Inozume T.
      • Yaguchi T.
      • Furuta J.
      • Harada K.
      • Kawakami Y.
      • Shimada S.
      Melanoma cells control anti-melanoma CTL responses via Interaction between TIGIT and CD155 in the effector phase.
      ).
      Second, IDO in cancer might have two components: IDO expression in cancer cells and IDO expression in peritumoral inflammation. Induction of IDO in melanoma cells is probably through direct effects of IFN-γ from adjacent TILs (
      • Brandacher G.
      • Perathoner A.
      • Ladurner R.
      • Schneeberger S.
      • Obrist P.
      • Winkler C.
      • et al.
      Prognostic value of indoleamine 2,3-dioxygenase expression in colorectal cancer: effect on tumor-infiltrating T cells.
      ). IDO expression in APCs, however, might be induced by factors such as neoantigen presentation, tumor cell death, or IL-6/TNF-α from bystander macrophages in the peritumoral infiltrate (
      • Grohmann U.
      • Orabona C.
      • Fallarino F.
      • Vacca C.
      • Calcinaro F.
      • Falorni A.
      • et al.
      CTLA-4-Ig regulates tryptophan catabolism in vivo.
      ,
      • Huang L.
      • Li L.
      • Lemos H.
      • Chandler P.R.
      • Pacholczyk G.
      • Baban B.
      • et al.
      Cutting edge: DNA sensing via the STING adaptor in myeloid dendritic cells induces potent tolerogenic responses.
      ,
      • von Bubnoff D.
      • Scheler M.
      • Wilms H.
      • Fimmers R.
      • Bieber T.
      Identification of IDO-positive and IDO-negative human dendritic cells after activation by various proinflammatory stimuli.
      ). IDO induction of peritumoral APCs would therefore describe the recognition of the tumor by the immune system rather than directly influence tumor growth. This is supported by the fact that we did not detect a correlation of Breslow tumor thickness with the level of APC IDO expression in our study.
      In our patient population, we mainly detected IDO in peritumoral myeloid CD11c+ DCs and CD68+ macrophages by double immunofluorescence. A very strong IDO expression of APCs was linked to a significantly shorter PFS compared with all other IDO-expressing APC groups. One interpretation is that IDO+ APCs may account for local and systemic immunosuppression in specific situations (
      • Johnson T.S.
      • Munn D.H.
      Host indoleamine 2,3-dioxygenase: contribution to systemic acquired tumor tolerance.
      ,
      • Lemos H.
      • Mohamed E.
      • Huang L.
      • Ou R.
      • Pacholczyk G.
      • Arbab A.S.
      • et al.
      STING promotes the growth of tumors characterized by low antigenicity via IDO Activation.
      ,
      • von Bubnoff D.
      • Scheler M.
      • Wilms H.
      • Wenzel J.
      • von Bubnoff N.
      • Hacker G.
      • et al.
      Indoleamine 2,3-dioxygenase-expressing myeloid dendritic cells and macrophages in infectious and noninfectious cutaneous granulomas.
      ). Increased serum IDO activity and increased numbers of circulating IDO+ monocytic myeloid-derived suppressor cells have been linked to disease progression or decreased survival in melanoma patients and might limit the clinical efficacy of adjuvant IFN-α therapy in melanoma stage III (
      • Chevolet I.
      • Speeckaert R.
      • Schreuer M.
      • Neyns B.
      • Krysko O.
      • Bachert C.
      • et al.
      Clinical significance of plasmacytoid dendritic cells and myeloid-derived suppressor cells in melanoma.
      ,
      • de Lecea M.V.
      • Palomares T.
      • Al Kassam D.
      • Cavia M.
      • Geh J.L.C.
      • de Llano P.
      • et al.
      Indoleamine 2,3 dioxygenase as a prognostic and follow-up marker in melanoma. A comparative study with LDH and S100B.
      ,
      • Kuales M.A.
      • Wenzel J.
      • Schmid-Wendtner M.H.
      • Bieber T.
      • von Bubnoff D.
      Myeloid CD11c+ S100+ dendritic cells express indoleamine 2,3-dioxygenase at the inflammatory border to invasive lower lip squamous cell carcinoma.
      ,
      • von Bubnoff D.
      • Scheler M.
      • Wilms H.
      • Fimmers R.
      • Bieber T.
      Identification of IDO-positive and IDO-negative human dendritic cells after activation by various proinflammatory stimuli.
      ). These data underscore the important prognostic effect of IDO on the disease.
      There exists a controversy in the literature regarding which cells exactly express IDO. For double immunofluorescence, a polyclonal IDO antibody has been used, and as such, specificity for IDO1 is somewhat uncertain. However, we have used this polyclonal antibody in previous studies in which we identified macrophages as IDO negative around squamous cell carcinoma and in autoimmune skin disorders (
      • Kuales M.A.
      • Wenzel J.
      • Schmid-Wendtner M.H.
      • Bieber T.
      • von Bubnoff D.
      Myeloid CD11c+ S100+ dendritic cells express indoleamine 2,3-dioxygenase at the inflammatory border to invasive lower lip squamous cell carcinoma.
      ,
      • Scheler M.
      • Wenzel J.
      • Tuting T.
      • Takikawa O.
      • Bieber T.
      • von Bubnoff D.
      Indoleamine 2,3-dioxygenase (IDO): the antagonist of type I interferon-driven skin inflammation?.
      ). However, we cannot exclude a cross-reaction to pigment-ingesting antigens in macrophages by this antibody, and this is clearly a limitation of our study.
      It is obvious that IDO expression alone is neither sufficient nor a prerequisite for tumor development. In melanoma phase III studies, no effect was seen as monotherapy using an IDO inhibitor such as epacadostat or indoximod (

      Zakharia Y, McWilliams R, Shaheen M, Grossmann K, Drabick J, Milhem M, et al. Interim analysis of the phase 2 clinical trial of the IDO pathway inhibitor indoximod in combination With pembrolizumab for patients With advanced melanoma (Abstract CT11). Paper presented at: 107th Annual Meeting of the American Association for Cancer Research (AACR). 1-5 April 2017; Washington, DC.

      ). However, when an IDO-inhibitor was combined with anti-PD1 antibodies, response rates were significantly increased to 52–59% compared with anti-PD1 therapy only (30–40% response rates) (Indoximod combo triggers responses in
      • Indoximod combo triggers responses in melanoma
      ). It will be very interesting to correlate IDO expression in melanoma cells and APCs with therapy response in these groups.
      Our study implicates that investigations targeting the expression of IDO in melanoma may be of pivotal future interest. The use of IDO inhibitors may be of particular use in combination with immune-stimulatory protocols such as PD-1 antibodies to subvert melanoma-associated antigen tolerance and to sustain T-cell effector function for the eradication of melanoma cells.

      Materials and Methods

      Patients

      We analyzed samples from 99 patients with primary cutaneous melanomas with tumor thickness of less than 0.75 mm (n = 32, group 1, low-risk group), tumor thickness of 0.75–1 mm (n = 31, group 2, intermediate-risk group), and tumor thickness of 4 mm or greater (n = 36, group 3, high-risk group) collected at the Department of Dermatology–Medical Center University of Freiburg from 1998 through 2013. These groups were used to investigate correlation of IDO expression with tumor thickness. Written consent was obtained from all patients at the time of tissue rejection in accordance with the study protocol approved by the local ethics committee of the University Medical Center Freiburg. All patients came to the clinic for surgical treatment of primary melanoma. Clinical data (sex, age) and follow-up information in 2016 were obtained from medical follow-up care records. The study was performed according to local ethical guidelines and approved by the local regulatory committee.

      Histology and immunohistochemistry

      Serial sections were prepared from formalin-fixed, paraffin-embedded skin biopsy samples. Standard hematoxylin and eosin staining and periodic acid Schiff reactions were performed for diagnostic purposes. Tumor thickness, ulceration, mitotic rate, lymphangioinvasion, regression, and microsatellitosis were determined by a dermatopathologist (FM) (
      • Jia Y.
      • Wang H.
      • Wang Y.
      • Wang T.
      • Wang M.
      • Ma M.
      • et al.
      Low expression of Bin1, along with high expression of IDO in tumor tissue and draining lymph nodes, are predictors of poor prognosis for esophageal squamous cell cancer patients.
      ,
      • Kim J.W.
      • Nam K.H.
      • Ahn S.H.
      • Park do J.
      • Kim H.H.
      • Kim S.H.
      • et al.
      Prognostic implications of immunosuppressive protein expression in tumors as well as immune cell infiltration within the tumor microenvironment in gastric cancer.
      ). Anti-IDO monoclonal antibody (mAb) (clone 10.1, dilution 1:200; Millipore, Billerica, MA) labeling was performed on paraffin-embedded tissue sections (1 μm) after heat pretreatment for 10 minutes in Target Retrieval Solution, pH 9 (DAKO, Hamburg, Germany). Appropriate isotype-matched controls were included. Additional antibodies for immunohistochemistry included anti-CD11c mAb (5D11, dilution 1:100; Novocastra, Newcastle, UK), anti-S100 antibody (Z0311, dilution 1:1,000; DAKO), anti-CD68 mAb (PGM1, dilution 1:50; DAKO), anti-CD1a mAb (O10, dilution 1:10; Zytomed Systems, Berlin, Germany), and anti-CD3 mAb (F7.2.38, dilution 1:50; DAKO). Visualization was performed using the REAL detection system (4th edition, DAKO) with fast red as the chromogen. Photographs of stainings (microscope: Eclipse 80i, Nikon, Düsseldorf, Germany) were visualized with the program AxioVision (Zeiss, Feldbach, Switzerland).

      Immunofluorescence staining

      Immunofluorescence double staining was carried out using formalin-fixed, paraffin-embedded sections after heat antigen retrieval following standard protocols. The following primary antibodies were used: anti-IDO polyclonal antibody (AHP833, dilution 1:100; Serotec, Düsseldorf, Germany), anti-CD11c mAb (5D11, dilution 1:50; Novocastra), and anti-CD68 mAb (PGM1, dilution 1:50; Novocastra). As secondary antibody we used a FITC-conjugated donkey-anti-sheep-IgG (dilution 1:100; Serotec, Düsseldorf, Germany) to detect the IDO antibody. To detect the CD11c and CD68 antibodies we used Alexa Fluor 568 (dilution 1:500; Invitrogen, Waltham, MA). Controls included staining with isotype-matched antibodies. Analysis was performed using a microscope (Eclipse 80i; Nikon, Düsseldorf, Germany) interfaced with a 12-V, 100-W halogen lamp (Osram, München, Germany). Photographs (camera JVC KY-F75U; JVC Professional Products, Wayne, NJ) were visualized with the program AxioVision (Zeiss).

      Scoring of IDO+ melanoma and APCs

      IDO+ melanoma cells or IDO+ APCs were scored as IDO negative (IDO), IDO+ (<25%, mild), IDO++ (25–50%, strong), or IDO+++ (>50%, very strong) if these percentages of melanoma cells or APCs in 10 representative high-power fields (magnification ×200) were IDO+.

      Scoring of the density of the peritumoral inflammatory infiltrate

      Cells of 10 representative high-power fields (magnification ×200) around the melanoma (vertical and horizontal growth phase) were calculated. The density of the peritumoral inflammatory infiltrate was scored as follows: none, if less than 25% of all cells were inflammatory cells in all high-power fields; mild, if 25–50% of all cells were inflammatory cells; moderate, if 50–75% of all cells were inflammatory cells; or intense, if nearly all (>75%) cells were of inflammatory origin and only a few stromal cells were seen. All samples were evaluated independently by two investigators (DvB and FM).

      Grading of TILs

      The Clark method for grading TILs was used (
      • Fukuno K.
      • Hara T.
      • Tsurumi H.
      • Shibata Y.
      • Mabuchi R.
      • Nakamura N.
      • et al.
      Expression of indoleamine 2,3-dioxygenase in leukemic cells indicates an unfavorable prognosis in acute myeloid leukemia patients with intermediate-risk cytogenetics.
      ). TILs were graded as brisk when present throughout the vertical growth phase or infiltrating the entire base thereof, nonbrisk if present in one or more foci of the vertical growth phase, or absent if lymphocytes had no contact with the vertical growth phase but were present in perivascular or fibrotic areas.

      Statistical methods

      Statistical analyses were performed using Microsoft Excel 2010 (Redmond, CA) and IBM SPSS Statistics Version 24 (IBM, Chicago, IL). To compare clinical and histopathological parameters between groups (Table 1), the analysis of variance test was used. The correlations of IDO to Breslow thickness, histopathologic parameters, and the density of the peritumoral inflammatory infiltrate (Table 2) were performed with the linear-by-linear trend test and analysis of variance test. Follow-up time was defined from the date of primary excision to the date of last follow-up or death (OS). Survival analyses included only deaths caused by melanoma. Multivariate Cox regression analyses with IDO expression and tumor thickness as covariates was performed to show significance of IDO on PFS. Kaplan-Meier analyses were performed to estimate OS and PFS. Probabilities of PFS and OS are shown as contingent upon IDO expression in melanoma cells and APCs in the peritumoral inflammatory infiltrate. The log rank test was applied to determine significance between the varying IDO-expressing survival curves of patients (between-group difference) or to differ between two selected survival groups. A P-value less than 0.05 was considered statistically significant.

      Conflict of Interest

      The authors state no conflict of interest.

      Acknowledgments

      We would like to thank Harald Binder, Department of Medical Biometry and Statistics, Medical Center–University of Freiburg, who greatly advised us with the statistics.

      References

        • Brandacher G.
        • Perathoner A.
        • Ladurner R.
        • Schneeberger S.
        • Obrist P.
        • Winkler C.
        • et al.
        Prognostic value of indoleamine 2,3-dioxygenase expression in colorectal cancer: effect on tumor-infiltrating T cells.
        Clin Cancer Res. 2006; 1: 1144-1151
        • Brenk M.
        • Scheler M.
        • Koch S.
        • Neumann J.
        • Takikawa O.
        • Hacker G.
        • et al.
        Tryptophan deprivation induces inhibitory receptors ILT3 and ILT4 on dendritic cells favoring the induction of human CD4+CD25+ Foxp3+ T regulatory cells.
        J Immunol. 2009; 183: 145-154
        • Brochez L.
        • Chevolet I.
        • Kruse V.
        The rationale of indoleamine 2,3-dioxygenase inhibition for cancer therapy.
        Eur J Cancer. 2017; 76: 167-182
        • Brody J.R.
        • Costantino C.L.
        • Berger A.C.
        • Sato T.
        • Lisanti M.P.
        • Yeo C.J.
        • et al.
        Expression of indoleamine 2,3-dioxygenase in metastatic malignant melanoma recruits regulatory T cells to avoid immune detection and affects survival.
        Cell Cycle. 2009; 8: 1930-1934
        • Chen G.
        • Kheradmand T.
        • Bryant J.
        • Wang S.
        • Tasch J.
        • Wang J.J.
        • et al.
        Intragraft CD11b(+) IDO(+) cells mediate cardiac allograft tolerance by ECDI-fixed donor splenocyte infusions.
        Am J Transplant. 2012; 12: 2920-2929
        • Chevolet I.
        • Speeckaert R.
        • Haspeslagh M.
        • Neyns B.
        • Kruse V.
        • Schreuer M.
        • et al.
        Peritumoral indoleamine 2,3-dioxygenase expression in melanoma: an early marker of resistance to immune control?.
        Br J Dermatol. 2014; 171: 987-995
        • Chevolet I.
        • Speeckaert R.
        • Schreuer M.
        • Neyns B.
        • Krysko O.
        • Bachert C.
        • et al.
        Clinical significance of plasmacytoid dendritic cells and myeloid-derived suppressor cells in melanoma.
        J Transl Med. 2015; 13: 9
        • Das M.
        • Zhu C.
        • Kuchroo V.K.
        Tim-3 and its role in regulating anti-tumor immunity.
        Immunol Rev. 2017; 276: 97-111
        • de Lecea M.V.
        • Palomares T.
        • Al Kassam D.
        • Cavia M.
        • Geh J.L.C.
        • de Llano P.
        • et al.
        Indoleamine 2,3 dioxygenase as a prognostic and follow-up marker in melanoma. A comparative study with LDH and S100B.
        J Eur Acad Dermatol Venereol. 2017; 31: 636-642
        • Fukuno K.
        • Hara T.
        • Tsurumi H.
        • Shibata Y.
        • Mabuchi R.
        • Nakamura N.
        • et al.
        Expression of indoleamine 2,3-dioxygenase in leukemic cells indicates an unfavorable prognosis in acute myeloid leukemia patients with intermediate-risk cytogenetics.
        Leuk Lymphoma. 2015; 56: 1398-1405
        • Gerlini G.
        • Di Gennaro P.
        • Mariotti G.
        • Urso C.
        • Chiarugi A.
        • Pimpinelli N.
        • et al.
        Indoleamine 2,3-dioxygenase+ cells correspond to the BDCA2+ plasmacytoid dendritic cells in human melanoma sentinel nodes.
        J Invest Dermatol. 2010; 130: 898-901
        • Godin-Ethier J.
        • Hanafi L.A.
        • Piccirillo C.A.
        • Lapointe R.
        Indoleamine 2,3-dioxygenase expression in human cancers: clinical and immunologic perspectives.
        Clin Cancer Res. 2011; 17: 6985-6991
        • Grohmann U.
        • Orabona C.
        • Fallarino F.
        • Vacca C.
        • Calcinaro F.
        • Falorni A.
        • et al.
        CTLA-4-Ig regulates tryptophan catabolism in vivo.
        Nat Immunol. 2002; 3: 1097-1101
        • Holmgaard R.B.
        • Zamarin D.
        • Munn D.H.
        • Wolchok J.D.
        • Allison J.P.
        Indoleamine 2,3-dioxygenase is a critical resistance mechanism in antitumor T cell immunotherapy targeting CTLA-4.
        J Exp Med. 2013; 210: 1389-1402
        • Holzel M.
        • Tuting T.
        Inflammation-induced plasticity in melanoma therapy and metastasis.
        Trend Immunol. 2016; 37: 364-374
        • Huang L.
        • Li L.
        • Lemos H.
        • Chandler P.R.
        • Pacholczyk G.
        • Baban B.
        • et al.
        Cutting edge: DNA sensing via the STING adaptor in myeloid dendritic cells induces potent tolerogenic responses.
        J Immunol. 2013; 191: 3509-3513
        • Indoximod combo triggers responses in melanoma
        Cancer Discov. 2017; 7: 542-543
        • Inozume T.
        • Yaguchi T.
        • Furuta J.
        • Harada K.
        • Kawakami Y.
        • Shimada S.
        Melanoma cells control anti-melanoma CTL responses via Interaction between TIGIT and CD155 in the effector phase.
        J Invest Dermatol. 2016; 136: 255-263
        • Jia Y.
        • Wang H.
        • Wang Y.
        • Wang T.
        • Wang M.
        • Ma M.
        • et al.
        Low expression of Bin1, along with high expression of IDO in tumor tissue and draining lymph nodes, are predictors of poor prognosis for esophageal squamous cell cancer patients.
        Int J Cancer. 2015; 137: 1095-1106
        • Johnson T.S.
        • Munn D.H.
        Host indoleamine 2,3-dioxygenase: contribution to systemic acquired tumor tolerance.
        Immunol Invest. 2012; 41: 765-797
        • Kim J.W.
        • Nam K.H.
        • Ahn S.H.
        • Park do J.
        • Kim H.H.
        • Kim S.H.
        • et al.
        Prognostic implications of immunosuppressive protein expression in tumors as well as immune cell infiltration within the tumor microenvironment in gastric cancer.
        Gastric Cancer. 2016; 19: 42-52
        • Kuales M.A.
        • Wenzel J.
        • Schmid-Wendtner M.H.
        • Bieber T.
        • von Bubnoff D.
        Myeloid CD11c+ S100+ dendritic cells express indoleamine 2,3-dioxygenase at the inflammatory border to invasive lower lip squamous cell carcinoma.
        Histol Histopathol. 2011; 26: 997-1006
        • Landsberg J.
        • Kohlmeyer J.
        • Renn M.
        • Bald T.
        • Rogava M.
        • Cron M.
        • et al.
        Melanomas resist T-cell therapy through inflammation-induced reversible dedifferentiation.
        Nature. 2012; 490: 412-416
        • Lemos H.
        • Mohamed E.
        • Huang L.
        • Ou R.
        • Pacholczyk G.
        • Arbab A.S.
        • et al.
        STING promotes the growth of tumors characterized by low antigenicity via IDO Activation.
        Cancer Res. 2016; 76: 2076-2081
        • Li Y.
        • Tredget E.E.
        • Ghaffari A.
        • Lin X.
        • Kilani R.T.
        • Ghahary A.
        Local expression of indoleamine 2,3-dioxygenase protects engraftment of xenogeneic skin substitute.
        J Invest Dermatol. 2006; 126: 128-136
        • Mellor A.L.
        • Baban B.
        • Chandler P.
        • Marshall B.
        • Jhaver K.
        • Hansen A.
        • et al.
        Cutting edge: induced indoleamine 2,3 dioxygenase expression in dendritic cell subsets suppresses T cell clonal expansion.
        J Immunol. 2003; 171: 1652-1655
        • Munn D.H.
        • Sharma M.D.
        • Baban B.
        • Harding H.P.
        • Zhang Y.
        • Ron D.
        • et al.
        GCN2 kinase in T cells mediates proliferative arrest and energy induction in response to indoleamine 2,3-dioxygenase.
        Immunity. 2005; 22: 633-642
        • Munn D.H.
        • Zhou M.
        • Attwood J.T.
        • Bondarev I.
        • Conway S.J.
        • Marshall B.
        • et al.
        Prevention of allogeneic fetal rejection by tryptophan catabolism.
        Science. 1998; 281: 1191-1193
        • Prendergast G.C.
        • Metz R.
        • Muller A.J.
        Towards a genetic definition of cancer-associated inflammation: role of the IDO pathway.
        Am J Pathol. 2010; 176: 2082-2087
        • Prendergast G.C.
        • Smith C.
        • Thomas S.
        • Mandik-Nayak L.
        • Laury-Kleintop L.
        • Metz R.
        • et al.
        Indoleamine 2,3-dioxygenase pathways of pathogenic inflammation and immune escape in cancer.
        Cancer Immunol Immunother. 2014; 63: 721-735
        • Rose C.
        [Diagnostics of malignant melanoma of the skin: Recommendations of the current S3 guidelines on histology and molecular pathology].
        Pathologe. 2017; 38: 49-61
        • Routy J.P.
        • Routy B.
        • Graziani G.M.
        • Mehraj V.
        The kynurenine pathway is a double-edged sword in immune-privileged sites and in cancer: implications for immunotherapy.
        Int J Tryptophan Res. 2016; 9: 67-77
        • Scheler M.
        • Wenzel J.
        • Tuting T.
        • Takikawa O.
        • Bieber T.
        • von Bubnoff D.
        Indoleamine 2,3-dioxygenase (IDO): the antagonist of type I interferon-driven skin inflammation?.
        Am J Pathol. 2007; 171: 1936-1943
        • Sharma M.D.
        • Hou D.Y.
        • Liu Y.
        • Koni P.A.
        • Metz R.
        • Chandler P.
        • et al.
        Indoleamine 2,3-dioxygenase controls conversion of Foxp3+ Tregs to TH17-like cells in tumor-draining lymph nodes.
        Blood. 2009; 113: 6102-6111
        • Speeckaert R.
        • Vermaelen K.
        • van Geel N.
        • Autier P.
        • Lambert J.
        • Haspeslagh M.
        • et al.
        Indoleamine 2,3-dioxygenase, a new prognostic marker in sentinel lymph nodes of melanoma patients.
        Eur J Cancer. 2012; 48: 2004-2011
        • von Bubnoff D.
        • Bausinger H.
        • Matz H.
        • Koch S.
        • Hacker G.
        • Takikawa O.
        • et al.
        Human epidermal Langerhans cells express the immunoregulatory enzyme indoleamine 2,3-dioxygenase.
        J Invest Dermatol. 2004; 123: 298-304
        • von Bubnoff D.
        • Scheler M.
        • Wilms H.
        • Fimmers R.
        • Bieber T.
        Identification of IDO-positive and IDO-negative human dendritic cells after activation by various proinflammatory stimuli.
        J Immunol. 2011; 186: 6701-6709
        • von Bubnoff D.
        • Scheler M.
        • Wilms H.
        • Wenzel J.
        • von Bubnoff N.
        • Hacker G.
        • et al.
        Indoleamine 2,3-dioxygenase-expressing myeloid dendritic cells and macrophages in infectious and noninfectious cutaneous granulomas.
        J Am Acad Dermatol. 2011; 65: 819-832
        • Weiss S.A.
        • Han S.W.
        • Lui K.
        • Tchack J.
        • Shapiro R.
        • Berman R.
        • et al.
        Immunologic heterogeneity of tumor-infiltrating lymphocyte composition in primary melanoma.
        Hum Pathol. 2016; 57: 116-125
      1. Zakharia Y, McWilliams R, Shaheen M, Grossmann K, Drabick J, Milhem M, et al. Interim analysis of the phase 2 clinical trial of the IDO pathway inhibitor indoximod in combination With pembrolizumab for patients With advanced melanoma (Abstract CT11). Paper presented at: 107th Annual Meeting of the American Association for Cancer Research (AACR). 1-5 April 2017; Washington, DC.

        • Zhang T.
        • Tan X.L.
        • Xu Y.
        • Wang Z.Z.
        • Xiao C.H.
        • Liu R.
        Expression and prognostic value of indoleamine 2,3-dioxygenase in pancreatic cancer.
        Chin Med J. 2017; 130: 710-716
        • Ziani L.
        • Safta-Saadoun T.B.
        • Gourbeix J.
        • Cavalcanti A.
        • Robert C.
        • Favre G.
        • et al.
        Melanoma-associated fibroblasts decrease tumor cell susceptibility to NK cell-mediated killing through matrix-metalloproteinases secretion.
        Oncotarget. 2017; 8: 19780-19794

      Linked Article

      • Erratum
        Journal of Investigative DermatologyVol. 138Issue 6
        • Preview
          Indoleamine 2,3-Dioxygenase Expression in Primary Cutaneous Melanoma Correlates with Breslow Thickness and Is of Significant Prognostic Value for Progression-Free Survival
        • Full-Text
        • PDF
        Open Archive