Immunohistochemical Features of Indoleamine 2,3-Dioxygenase (IDO) in Various Types of Lymphoma: A Single Center Experience

Indolamine-2,3-dioxygenase (IDO) is an intracellular enzyme that catalyzes amino acid tryptophan to L-kynurenine. IDO is overexpressed in various cancers and several IDO inhibitors have been assessed in multiple clinical trials. If an IDO inhibitor is to be commercialized, IDO immunohistochemistry will be an important method. In this study, 80% (28/35) of mature T- and natural killer (NK)-cell neoplasms showed positivity for IDO protein (score 1: five, score 2: one, score 3: seven, score 4: fifteen). In addition, 29.9% (23/77) of mature B-cell lymphomas showed positivity for IDO protein (score 1: three, score 2: tewelve, score 3: four, score 4: four). In mature B-cell lymphomas, 95.7% (22/23) of IDO positive cases were diffuse B-cell lymphomas. Our study includes various types of lymphoma that were previously unreported and shows various patterns of IDO stain according to the type. When the results are accumulated, IDO immunohistochemistry will be a useful tool to diagnose lymphomas and to predict their prognosis.


Introduction
Tumors express the antigens that induce the host immune response. Progression of tumors requires avoidance of host immune surveillance [1,2]. Recent studies have shown that tryptophan catabolism is one means of avoiding immune surveillance [3,4]. Indolamine-2,3-dioxygenase (IDO) is a cytosolic enzyme that catalyzes tryptophan. IDO converts the amino acid tryptophan to L-kynurenine [5]. The depletion of tryptophan and the production of L-kynurenine induces the apoptosis of T-cells and natural killer (NK)-cells [6][7][8]. In addition, the IDO-expressing macrophages, dendritic cells, and tumor cells suppress T-cell proliferation [7][8][9][10]. In previous reports, IDO expression and the serum concentration of L-kynurenine were negative prognostic factors in diffuse large B-cell lymphomas and adult T-cell leukemia/lymphomas [11][12][13]. In a previous immunohistochemical analysis for IDO expression in diffuse large B-cell lymphomas treated with R-CHOP chemotherapy, the IDO-positive group showed resistance to the treatment and a poorer prognosis than the IDO-negative group [14]. Immunohistochemistry is a relatively fast and inexpensive utility in diagnostic surgical pathology. Immunohistochemistry is widely used for subtyping of lymphomas and plays an important role in hematopathology. There are very few recent immunohistochemical assays of

Immunohistochemistry for Indoleamine 2, 3-Dioxygenase
We reviewed all slides of the cases and selected one representative formalin-fixed, paraffinembedded (FFPE) block from each case for immunohistochemistry. The representative blocks were cut on 4 µm thick sections and immunohistochemical staining was performed for Indoleamine 2, 3-dioxygenase (rabbit recombinant monoclonal, clone EPR20374, Abcam, 1:2000 dilution). All immunohistochemical staining was performed using the BenchMark XT autostainer (Ventana Medical Systems, Tucson, Arizona), according to the manufacturer's protocol. The results were evaluated by two authors (H.Y.L and M.S.K) independently, and any discrepancies were reviewed to achieve a consensus. In previous studies, the reactive immune cells around the tumors were stained by IDO protein [17,18]. It was difficult or impossible to distinguish reactive immune cells from tumor cells. Considering the possibility of containing reactive cells in the tumors, percentage of stained cells equal to or higher than 5% were regards as positive. All lymphomas were scored based on percentage of tumor cell staining: 0 = <5%, 1 = 5-25%, 2 = 26-50%, 3 = 51-75%, and 4 = >75%. The moderate-to-strong cytoplasmic staining was regards as positive. In Hodgkin lymphomas, due to the scant cellularity of tumor cells, the expressions were evaluated to be positive or negative. The positivity for IDO in the vessels was judged in comparison with hematoxylin and eosin (H&E) stain. The vessels were lined by endothelium and covered by pericytes in H&E stain.

Hodgkin Lymphoma
The clinicopathologic characteristics and the immunohistochemical results of IDO are summarized in Table 1. The male:female ratio was 1.3:1, and the age distribution was 35-79 years (mean = 61 years, median = 70 years). Four cases of classic Hodgkin lymphoma (three mixed cellularity and one nodular Diagnostics 2020, 10, 275 3 of 15 sclerosis), and three cases of nodular lymphocyte predominant Hodgkin lymphoma were included in the study. There was no difference of IDO staining pattern between classic type and nodular lymphocyte predominant type. In the positive cases, the tumor cells and vessels showed positivity for IDO protein (Figure 1). In the negative cases, only dendritic cells and macrophages showed positivity for IDO ( Figure 2). Three cases were positive, and four cases were negative. In past studies [15,16], IDO had been expressed in the dendritic cells and the macrophages only. However, our results showed tumor cell positivity for IDO in several cases. (mean = 61 years, median = 70 years). Four cases of classic Hodgkin lymphoma (three mixed cellularity and one nodular sclerosis), and three cases of nodular lymphocyte predominant Hodgkin lymphoma were included in the study. There was no difference of IDO staining pattern between classic type and nodular lymphocyte predominant type. In the positive cases, the tumor cells and vessels showed positivity for IDO protein (Figure 1). In the negative cases, only dendritic cells and macrophages showed positivity for IDO ( Figure 2). Three cases were positive, and four cases were negative. In past studies [15,16], IDO had been expressed in the dendritic cells and the macrophages only. However, our results showed tumor cell positivity for IDO in several cases.     Table 2. The male:female ratio was 2.7:1, and the age distribution was 10-84 years (mean = 61 years, median = 65.5 years). Six cases were germinal center type and 20 cases were activated B-cell type. There was no difference in the IDO staining pattern between germinal center type and activated Bcell type. The lymphomas were scored based on percentage of tumor cell staining: 0 = <5%, 1 = 5-25%, 2 = 26-50%, 3 = 51-75%, and 4 = >75% ( Figure 3A: score 0; 3B: 1; 3C: 2; 3D: 3; 3E: 4). The staining distribution was scored from 0 to 4. Eight cases were scored 0, three cases were 1, 10 cases were 2, three cases were 3, and two cases were 4.

Diffuse Large B-Cell Lymphoma (DLBCL), Subtypes
The clinicopathologic characteristics and the immunohistochemical results of IDO are summarized in Table 2. A total of 30.8% (8/26) of DLBCL, NOS were scored 0 for IDO immunohistochemistry, but four cases of primary DLBCL of the central nervous system were scored 0 ( Figure 4). A T-cell/histiocyte-rich large B-cell lymphoma was scored 2 ( Figure 3F), a primary cutaneous DLBCL, leg type was scored

B-Cell Lymphomas with Low IDO Expression
A total of 15 cases of follicular lymphoma (grade 1-2: 11, 3A: three, 3B: one) and 16 cases of marginal zone B-cell lymphoma of the mucosa-associated lymphoid tissue (MALT lymphoma) were enrolled in the study. The clinicopathologic characteristics and the immunohistochemical results of IDO are summarized in Table 3. The male:female ratio was 1.14:1, and the age distribution was 36-79 years (mean = 52.3 years, median = 50 years) in follicular lymphomas. The male:female ratio was 1:1, and the age distribution was 42-84 years (mean = 64.1 years, median = 64 years) in MALT lymphomas. A total of 93.3% (14/15) of follicular lymphomas were scored 0 ( Figure 5A). One 3B case showed positivity for IDO in the centroblasts of follicles and the diffuse patterned areas ( Figure 5B, submandibular lymph node and, 16 months later, the lymphoma transformed to DLBCL of the brain (IDO score 4). One pediatric follicular lymphoma was scored 0 for IDO expression. A total of 16 cases of MALT lymphoma ( Figure 5C) and four cases of nodal marginal zone lymphoma were scored 0 for IDO. Three mantle lymphomas ( Figure 5D), two chronic lymphocytic leukemia/small lymphocytic lymphomas ( Figure 5E), and one B-lymphoblastic lymphoma/leukemia, NOS ( Figure 5F) were scored 0 for IDO expression. Two Burkitt lymphomas were scored 0 for IDO ( Figure 6).

B-Cell Lymphomas with Low IDO Expression
A total of 15 cases of follicular lymphoma (grade 1-2: 11, 3A: three, 3B: one) and 16 cases of marginal zone B-cell lymphoma of the mucosa-associated lymphoid tissue (MALT lymphoma) were enrolled in the study. The clinicopathologic characteristics and the immunohistochemical results of IDO are summarized in Table 3. The male:female ratio was 1.14:1, and the age distribution was 36-79 years (mean = 52.3 years, median = 50 years) in follicular lymphomas. The male:female ratio was 1:1, and the age distribution was 42-84 years (mean = 64.1 years, median = 64 years) in MALT lymphomas. A total of 93.3% (14/15) of follicular lymphomas were scored 0 ( Figure 5A). One 3B case showed positivity for IDO in the centroblasts of follicles and the diffuse patterned areas ( Figure 5B, submandibular lymph node and, 16 months later, the lymphoma transformed to DLBCL of the brain (IDO score 4). One pediatric follicular lymphoma was scored 0 for IDO expression. A total of 16 cases of MALT lymphoma ( Figure 5C) and four cases of nodal marginal zone lymphoma were scored 0 for IDO. Three mantle lymphomas ( Figure 5D), two chronic lymphocytic leukemia/small lymphocytic lymphomas ( Figure 5E), and one B-lymphoblastic lymphoma/leukemia, NOS ( Figure 5F) were scored 0 for IDO expression. Two Burkitt lymphomas were scored 0 for IDO ( Figure 6).

Discussion
Indolamine-2,3-dioxygenase (IDO) is an intracellular enzyme that catalyzes amino acid tryptophan to L-kynurenine [19]. IDO is reportedly expressed by dendritic cells in tumor-draining lymph nodes [10]. In our study, dendritic cells and macrophages of lymph nodes or tonsils showed positivity for IDO protein. All lymphomas were scored based on percentage of tumor cell staining: score 0 = <5%, 1 = 5%-25%, 2 = 26%-50%, 3 = 51%-75%, and 4 = >75%. It was difficult or impossible to distinguish reactive immune cells from tumor cells. To clarify the distinction between the tumor cells and reactive cells, multiple immunofluorescence stainings against IDO and other antibodies can be helpful as has been done in previous research [17]. The staining distribution of IDO was scored 0-4. A total of 80% (28/35) of mature T-and NK-cell neoplasms showed positivity for IDO protein (score 1: five, score 2: one, score 3: seven, score 4: 15). There was no different staining pattern or positive rate between the histopathological types of mature T-and NK-cell neoplasms. A total of 78.6% (22/28) of positive cases were scored 3 or 4; 83.3% (10/12) of extranodal NK-/T-cell lymphomas showed diffuse positivity for IDO protein. The remaining two negative cases were of a relatively young age (19 and 26 years) than the positive extranodal NK-/T-cell lymphomas (aged 43-83). In anaplastic large cell lymphomas, all three ALK-positive cases were positive (scoring 2, 3, or 4) for IDO and both ALK-negative cases were negative for IDO. A total of six peripheral T-cell lymphomas, NOS showed positivity for IDO (five scoring 3 and one scoring 1). In contrast to mature T-and NK-cell neoplasms, 29.9% (23/77) of mature B-cell lymphomas showed positivity for IDO protein (score 1: three, score 2: 12, score 3: four, score 4: four). In mature B-cell lymphomas, 95.7% (22/23) of IDO positive cases were DLBCL, NOS, or DLBCL subtypes. There was no different staining pattern or positive rate between DLBCL, NOS, and DLBCL subtypes. All small B-cell neoplasms except one were negative for IDO protein. The remaining positive case was follicular lymphoma, grade 3B. The positive follicular lymphoma was scored 3 at centroblasts in follicle and diffuse areas. In contrast to other subtypes of DLBCL, all four primary DLBCLs of the CNS were negative for IDO protein. IDO proteins have been overexpressed in various cancers [20][21][22][23] and several IDO inhibitors have been assessed in multiple clinical trials. If an IDO inhibitor is to be commercialized, IDO immunohistochemistry will be important method. However, there are insufficient studies of immunohistochemistry for IDO in lymphomas. Previous studies have demonstrated only Hodgkin lymphoma and diffuse large B-cell lymphoma [14][15][16]. There was no immunohistochemical study of IDO protein for mature T-and NK-cell neoplasms. However, our study included various types of mature T-and NK-cell lymphoma (that have previously been unreported) and showed various patterns of IDO staining according to the type. Our study used scant sample size to determine the IDO pattern of various lymphomas and the relationship between IDO expression and prognosis. When the results are accumulated, IDO immunohistochemistry will be a useful tool to diagnose lymphomas and predict their prognosis.