Maturation of dendritic cells and T-cell responses in sentinel lymph nodes from patients with breast carcinoma

Authors

  • Kazuo Matsuura M.D.,

    1. Department of Surgical Oncology, Research Institute for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan
    Search for more papers by this author
  • Yoshiyuki Yamaguchi M.D., Ph.D.,

    Corresponding author
    1. Department of Surgical Oncology, Research Institute for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan
    • Department of Surgical Oncology, Research Institute for Radiation Biology and Medicine, Hiroshima University, Kasumi 1-2-3, Minami-Ku, Hiroshima 734-8553, Japan===

    Search for more papers by this author
    • Fax: (011) 81 82 256 7109

  • Hideaki Ueno M.D., Ph.D.,

    1. Department of Surgical Oncology, Research Institute for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan
    Search for more papers by this author
  • Akihiko Osaki M.D., Ph.D.,

    1. Department of Surgical Oncology, Research Institute for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan
    Search for more papers by this author
  • Koji Arihiro M.D., Ph.D.,

    1. Department of Pathology, Hiroshima University, Hiroshima, Japan
    Search for more papers by this author
  • Tetsuya Toge M.D., Ph.D.

    1. Department of Surgical Oncology, Research Institute for Radiation Biology and Medicine, Hiroshima University, Hiroshima, Japan
    Search for more papers by this author

Abstract

BACKGROUND

Methods for identifying sentinel lymph nodes (SNs) and their clinical significance have been established. Recent advances in molecular immunology have enabled the analysis of precise immune responses. The objective of the current study was to clarify the dendritic cell (DC) maturation, T-helper type 1 (Th-1) and Th-2 responses, and regulatory T-cell responses of SNs in patients with breast carcinoma.

METHODS

SNs and non-SNs were identified by radioguided and blue dye-guided methods in 70 consecutive patients with clinically lymph node negative (N0) breast carcinoma. Lymphocytes were collected from SNs and non-SNs and were subjected to flow cytometric analysis (FCM) using antibodies of CD83-fluorescein isothiocyanate (FITC), CD80-phycoerythrin (PE), CD86-PE, CD40-PE, human leukemic D-related antigen (HLA-DR)-FITC, CD4-FITC, and CD25-PE. Total RNA was extracted from SNs and non-SNs, and the expression of CD83, interleukin 12p40 (IL-12p40), interferon γ (IFN-γ), IL-4, IL-10, and Foxp3 was evaluated by using quantitative real-time reverse transcriptase-polymerase chain reaction (RT-PCR) analysis. The immunologic status of SNs was analyzed further with regard to micrometastases, which were identified as negative microscopically but positive according to an RT-PCR analysis that was specific for mammaglobin.

RESULTS

SNs were detectable in 70 of 71 consecutive patients (98.6%) with clinically N0 breast carcinoma. Fourteen of 70 patients (20.0%) had positive metastasis in SNs. When SNs were compared with non-SNs in 56 metastasis-negative patients, FCM revealed that HLA-DR-positive, CD80-positive, CD86-positive, and CD40-positive cell populations were decreased significantly in SNs. RT-PCR analysis demonstrated that, among 44 patients with metastasis-negative SNs, the expression levels of CD83 and IFN-γ mRNA were significantly lower in SNs compared with non-SNs. Immunologic parameters also were compared between 44 metastasis-negative SNs and 14 metastasis-positive SNs. The metastasis-positive SNs demonstrated significantly higher expression of CD83, IL-12p40, IFN-γ, IL-10, and Foxp3 mRNA than the metastasis-negative SNs. Correction of micrometastasis detected by mammaglobin enhanced these differences consistently.

CONCLUSIONS

In patients with breast carcinoma, cellular immune responses, from DC maturation to Th-1 responses, appeared to be less active in SNs compared with non-SNs before metastasis developed. Once metastasis was established in SNs, DC maturation was triggered and was followed by the up-regulation of Th-1 responses, which may reflect antigen-specific immune responses in SNs. Unlike DC maturation and Th-1 responses after metastasis in SNs, up-regulation of Th-2 and regulatory T-cell responses developed in parallel. Cancer 2006. © 2006 American Cancer Society.

The machinery of antigen presentation and recognition has been understood at cellular and molecular levels in tumor antigen (Ag)-specific immunity.1 In such responses, it has been shown that dendritic cells (DCs) play an important role.2 Resident tissue DCs capture Ags and migrate to draining lymph nodes. DCs collect both material, in the form of Ag, and information, in the form of cellular activation signals, in peripheral tissues and sites of Ag accumulation.3 DCs constitutively transport this cargo to areas of Ag presentation, primarily the T-cell zones of secondary draining lymph nodes.4 During these processes, DCs differentiate from immature cells to mature cells.

Mature DCs express major histocompatability complex (MHC) Class II, CD86, CD54, and CD83 molecules. CD83 is the best known marker for mature DCs.5 Human leukemic D-related antigen (HLA-DR), which is the MHC Class II for man, is expressed by DCs; and it is known that the expression of CD80, CD86, and CD40 is necessary for DCs to activate T cells.6–8 Interleukin 12p70 (IL-12p70) is a heterodimer that consists of covalently linked p35 and p40 subunits encoded by 2 distinct genes. This heterodimer, which is produced at high levels by DCs, drives naive T cells to differentiate into interferon γ (IFN-γ)-producing T-helper 1 (Th-1) cells.9 The mature DCs contribute to antigen-presentation, which results in the emergence of antigen-specific immune responses, including Th-1 responses, that are desirable for antimetastatic responses.10 Conversely, it has been recognized that Th-2 responses, which represent IL-4 or IL-10 production,11 and regulatory T (T-reg) cell responses, which are identified by CD4-positive/CD25-positive cells and by Foxp3 expression,12 serve as anti-Th-1, antidefense responses.11, 13 Recent advances in molecular immunology have enabled us to analyze host immunologic status by using these molecular markers related to DC maturation and the subsequent T-cell responses.

Draining lymph nodes that are targeted to be reached first by tumor cell metastasis have been identified as sentinel lymph nodes (SNs).14 The SN concept is one of the most highlighted issues in recent surgical oncology. The staging merits and individualized surgical management have been proposed for patients with melanoma,15 breast carcinoma,16, 17 and upper gastrointestinal carcinoma18 based on the SN concept. This has been possible, because a method has been established for detecting SNs.15, 16, 18 SNs may be targeted not only by tumor cell metastasis but also by antigen-captured DCs for the emergence of antitumor immune responses.3, 4 However, it is surprising to note that the immunologic roles of SNs have not been understood fully, particularly with regard to antigen-specific immunity.

In the current study, we attempted to clarify the antigen-specific immunity of SNs based on the maturation of DCs, Th-1 and Th-2 responses, and T-reg cell responses. Immunologic status was compared between SNs and non-SNs in patients with metastasis-negative breast carcinoma and also was compared between metastasis-negative SNs and metastasis-positive SNs from patients with breast carcinoma. Moreover, the immunologic status of SNs was analyzed in further detail with regard to micrometastases that were analyzed by using a gene that encodes mammaglobin, which is the most informative marker for lymph node metastasis of breast carcinoma.19

MATERIALS AND METHODS

Patients Enrolled

Adult female patients with primary breast carcinoma who underwent SN biopsy as part of their surgical treatment at Hiroshima University Hospital between 2001 and 2004 were enrolled in the study. Each patient was clinically negative for lymph node metastasis (N0). The study was approved by the Institutional Review Board of Hiroshima University Hospital, and written informed consent was obtained from all enrolled patients.

Detection of SNs

Detection of SNs was performed by using two independent methods: the radioguided method and the blue dye-guided method. In brief, 1 mL of technetium 99m-phytate (total dose, 0.5 mCi) was injected subcutaneously at the site of the breast tumor 1 day before surgery. Ingigocalmine dye (3 mL) was injected subcutaneously at the alleola of the nipple under general anesthesia just before the operation was started. During surgery, the radiation levels for each lymph node were assessed by using a navigational global positioning system (Tyco Electronics Japan). Lymph nodes that displayed radioactivity in > of 10 cps were considered SNs. SNs also were detected as blue lymph nodes. Lymph nodes just beside the SNs were collected randomly as non-SNs, which displayed radioactivity in < 10 cps and were not blue lymph nodes.

Detection of Metastasis

The SNs were cut immediately into two pieces, one of which was subjected to histologic analysis, and the piece was divided further into another two pieces to collect the SN cells for cellular analysis, and those two pieces were frozen for mRNA analysis to detect the immunologic parameters. The tumor status of the SNs was diagnosed cytologically and histologically, and micrometastases were analyzed by reverse transcriptase-polymerase chain reaction (RT-PCR) analysis that was specific for mammaglobin.19

Immunohistochemical Staining

Mouse antihuman monoclonal antibody (MoAb) to CD83 (clone, HB15a; Immunotech SA, Marseille, France) and CD25 (clone, ACT-1; DakoCytomation, Glosstrup, Denmark) were used as primary antibodies. These MoAbs were diluted in phosphate buffered saline at 1:100. The second antibody was biotinylated goat antimouse immunoglobulin G (IgG)/IgM followed by streptavidin-horseradish peroxidase (HRP) (Dako Corporation, Carpinteria, CA). Prior to staining of CD83, the tissue slide was deparaffinized to remove embedded media and then rehydrated. Tissue sections were treated with pepsin (0.2 mol/L HCl; DAKO, Osaka, Japan) for 30 minutes at 37 °C and then with 0.3% methanol-hydrogen peroxide to inactivate endogenous peroxidase before staining for CD83 (clone, HB15a; Immunotech SA). Frozen sections were acetone-fixed for 15 minutes, and no antigen retrieval was performed for CD25 (clone, ACT-1; DakoCytomation). After each treatment, tissue sections were incubated overnight at 4 °C with mouse antihuman MAb to CD83 (clone, HB15a; Immunotech SA) and CD25 (clone, ACT-1; DakoCytomation), followed by incubation with biotinylated goat antimouse IgG/IgM and streptavidin-HRP (Dako Corporation). High-sensitivity nickel chloride-enhanced 3,3′ diaminobenzidine tetrahydrochloride techniques were used, and sections were counterstained with hematoxylin.20

Collection of Lymphocytes

An SN cell suspension was prepared immediately after the SN biopsy by mincing blocks of biopsied SNs using scissors and RPMI-1640 medium through a wire mesh (no. 2000). The SN cell suspension was layered on Ficoll-Conray. SN lymphocytes (SNLs) were isolated by gradient centrifugation (2000 revolutions per min for 30 min), washed twice, resuspended in medium containing 2% autologous serum at a density of 1 × 106/mL, and immediately subjected to flow cytometric (FCM) analysis.

Flow Cytometry

Fifty microliters of the SNL suspension (5 × 105 cells) were stained with antibodies, washed, and then analyzed within 24 hours on a FACSCalibur (10,000 events per sample; Becton Dickinson, San Jose, CA). The antibodies used were CD83-fluorescein isothiocyanate (FITC), CD80-phycoerythrin (PE), CD86-PE, CD40-PE (Immunotech SA), HLA-DR-FITC, CD4-FITC, and CD25-PE (Becton Dickinson). All data were obtained with gates set on living cells and were analyzed using CellQuest software (Becton Dickinson).

RT-PCR

Stored SN tissues were applied to total RNA extraction by using an RNeasy Mini Kit (QIAGEN, Hilden, Germany). DNase treatment was performed during the protocol of the RNeasy Mini Kit with the RNase-Free DNase Set (QIAGEN) to avoid amplification of pseudogene sequences from contaminating genomic DNA. The quality and quantity of the obtained total RNA preparation were determined by absorbance at 260 nanometers (nm) and 280 nm and were adjusted at a concentration of 0.10 μg/μL. In addition, the integrity of RNA in each preparation was tested by RT-PCR of glyceraldehydes-3-phosphate dehydrogenase (GAPDH). The prepared total RNA served as the template in the first-strand cyclic DNA (cDNA) synthesis using Ready-To-Go™ You-Prime-First-Strand Beads (Amersham Pharmacia Biotech Inc., Piscataway, NJ) with 0.5 μg of Oligo(dT)15 primer (Novagen, Darmstadt, Germany) according to the manufacturer's instructions. Each cDNA product was diluted 10-fold with RNase-free water to avoid inhibition of the PCR reaction by some reagent used in RNA extraction or reverse transcription. For relative quantification by PCR, each cDNA product was analyzed at a final Mg++ concentration of 3 mM in a LightCycler with software (version 3.5; Roche Molecular, Mannheim, Germany) by using a FastStart DNA Master SYBR Green I Kit (Roche Diagnostics, Mannheim, Germany). For each primer pair, a standard curve was developed. The PCR condition was 95 °C for 10 minutes for the initial denaturation, followed by 35 cycles of 1 second at 95 °C, 10 seconds at 60 °C, and 10 seconds at 72 °C, followed by a melting program (at 60–95 °C) to check the proper melting temperature of the product. Annealing temperatures and elongation times were optimized for primer generation and exclusion of artifacts. Primers for IL-12p40, IFN-γ, IL-4, IL-10, FoxP3, and GAPDH were from Search-LC GmbH (Heidelberg, Germany). The sequences of CD83 and the mammaglobin primer pairs used were as follows: for CD83, 5′-TGAAGGTGGCTTGCTCCGAAGAT-3′ and 5′-CAGGACAATCTCCGCTCTGTATT-3′; for mammaglobin, 5′-CGGATGAAACTCTGAGCAATGT-3′, and 5′-CTGCAGTTCTGTGAGCCAAAG-3′.21 All clinical samples were amplified by PCR for each diluted standard nucleotide, and the quality of amplification was checked with LightCycler® software (version 3.5; Idaho Technology Inc., Salt Lake City, UT) by calculating the ratio of the crossing-point (threshold cycle number to the exponential phase at the beginning of PCR) and the logarithm concentration of the amplified copy number of each standard sample.22

Statistical Analysis

Clinicopathologic features associated with SNs were evaluated by using the chi-square test and the Mann–Whitney U test. Because of the skewed nature of some of the PCR data, differences between unpaired groups were evaluated by using the Mann–Whitney U test. Differences between paired groups in the FCM analysis were evaluated by using the Student t test for paired data with StatView software (version 5 for Windows; Abacus Concepts, Inc., London, UK). P values < 0.05 were considered significant.

RESULTS

Detection of SNs

SNs could be detectable in 70 of 71 consecutive patients (98.6%) with clinical N0 status by using a radioguided method and a blue dye-guided method. The radioguided method had a detection rate of 98.6%, and the blue dye-guided method had a detection rate of 80.1%. Sixty patients (85.7%) had 1 SN, 9 patients (12.9%) had 2 SNs, and 1 patient (1.4%) had 3 SNs, for a mean of 1.2 SNs among the 70 patients studied.

Patient Characteristics

The characteristics of the 70 patients who were included in the study are shown in Table 1. The enrolled patients were age 56.0 ± 11.9 years (mean ± standard deviation; range, 33–79 yrs). Twenty-eight patients (40%) were premenopausal, and 42 patients (60%) were postmenopausal. Fifty-two patients (74%) underwent breast-conserving surgery, and 18 patients (26%) underwent mastectomy. Tumor size ranged from 0.8 cm to 4.6 cm in greatest dimension. Greater than 50% of tumors expressed estrogen (56%) or progesterone (51%) receptors. HER-2 status varied from negative to 3 +; 20%of patients had negative status, 26% of patients had 1 + status, 14% of patients had 2 + status, and 9% of patients had 3 + status. Nearly 50% of patients had Scarff-Bloom-Richardson Grade 1 tumors (40%). Forty-three percent of patients had no lymphatic invasion (ly0), and 86% had no vascular invasion (v0). Of the 70 patients enrolled, 56 patients had pathologic (p)N0 status, 13 patients had pN1 status, and 1 patient had pN2 status according to standard histopathology (hematoxylin and eosin staining).

Table 1. Clinicopathologic Features of the Sentinel Lymph Nodes Examined
FeatureTotal (n = 70)SN metastasis status: No. of patients (%)P value
Positive (n = 56 patients)Negative (n = 14 patients)
  1. SN: sentinel lymph node; SD: standard deviation; ER: estrogen receptor; PgR: progesterone receptor; SBR: Scarff-Bloom-Richardson.

Age in yrs    
 Mean ± SD56.0 ± 11.956.2 ± 12.355.0 ± 10.90.9532
 Range33–7933–7935–74 
Menopausal status    
 Premenopausal28 (40)23 (41)5 (36)0.9514
 Postmenopausal42 (60)33 (59)9 (64) 
Type of surgery    
 Breast-conserving surgery52 (74)47 (84)5 (36)0.0008
 Mastectomy18 (26)9 (16)9 (64) 
Tumor size in cm    
 Mean ± SD2.3 ± 1.02.2 ± 1.02.7 ± 1.20.1688
 Range0.8–4.60.8–4.41.0–4.6 
Hormone receptor status    
 ER    
  Positive39 (56)33 (59)6 (43)0.5223
  Negative19 (27)14 (25)5 (36) 
  Unknown12 (17)9 (16)3 (21) 
 PgR    
  Positive36 (51)32 (57)4 (29)0.1081
  Negative22 (31)15 (27)7 (50) 
  Unknown12 (17)9 (16)3 (21) 
HER-2 status (%)    
 Negative14 (20)12 (21)2 (14)0.9016
 1+18 (26)14 (25)4 (29) 
 2+10 (14)7 (13)3 (21) 
 3+6 (9)5 (9)1 (7) 
 Unknown22 (31)18 (32)4 (29) 
Tumor grade (SBR)    
 Grade 128 (40)25 (45)3(21)0.0921
 Grade 222 (31)15 (27)7(50) 
 Grade 312 (17)8 (14)4(29) 
 Unknown8 (11)8 (14)0 (0) 
Lymphatic invasion score    
 030 (43)29 (52)1 (7)< 0.0001
 123 (33)17 (30)6 (43) 
 27 (10)1 (2)6 (43) 
 31 (1)0 (0)1 (7) 
 Unknown9 (13)9 (16)0 (0) 
Vascular invasion score    
 060 (86)47 (84)13 (93)0.504
 11 (1)0 (0)1 (7) 
 Unknown9 (13)9 (16)0 (0) 

SN Status

When the 56 metastasis-negative SNs were compared with 14 metastasis-positive SNs from patients with breast carcinoma, no significant differences were indicated in mean age, menopausal status, tumor size, hormone receptor status, HER-2 expression, or venous invasion levels (Table 1), although the sample size of metastasis-positive SNs was too small to evaluate (n = 14 SNs). Patients with metastasis-negative SNs underwent breast-conserving surgery significantly more often than patients with metastasis-positive SNs (84% vs. 36%; P = 0.0008). Tumors tended toward a lower grade among patients who had metastasis-negative SNs compared with tumors among patients who had metastasis-positive SNs (Grade 1, 44% vs. 27%; Grade 2, 28% vs. 47%; and Grade 3, 15% vs. 27%; P = 0.0799). The potential of lymphatic invasion (ly) was significantly lower in tumors from patients who had metastasis-negative SNs than in tumors from patients who had metastasis-positive SNs (ly0, 52% vs. 7%; ly1, 31% vs. 40%; ly2, 2% vs. 47%; and ly3, 0% vs. 7%; P < 0.0001) (Table 1).

Immunohistochemical Staining

First, we confirmed the presence of CD83-positive cells in SNs. Hematoxylin and eosin staining clearly indicated a paracortical area known as the T-cell zone (Fig. 1A). Immunohistochemical staining using an anti-CD83 antibody indicated that brownish-stained CD83-positive cells were presented surrounding the paracortical T-cell area (Fig. 1B). Mononuclear cells surrounded a CD83-positive cell in the high-power field (Fig. 1C).

Figure 1.

These photomicrographs illustrate immunohistochemical staining of sentinel lymph nodes (SNs) for CD83 expression. SNs were analyzed with (A) hematoxylin and eosin staining and (B, C) immunohistochemical staining for CD83 expression. CD83-positive cells were identified as brownish cells in the paracortical area of lymph node (B). Under high-power magnification, mononuclear cells surrounded a brownish, CD83-positive cell.

Comparison of Immunologic Parameters between SNs and Non-SNs in Metastasis-Negative Patients

Mononuclear cells were isolated from SNs and non-SNs, and immunologic parameters were assessed with FCM analysis (Table 2) and RT-PR analysis (Table 3) among 56 lymph node-negative patients. FCM analysis revealed that there was no significant difference in CD83-positive cell populations between SNs and non-SNs. However, the HLA-DR-positive, CD80-positive, CD86-positive, and CD40-positive cell populations were decreased significantly in SNs compared with non-SNs (HLA-DR: 33.7% ± 13.8% vs. 43.3% ± 10.9% for SNs and non-SNs, respectively [P = 0.0111]; CD80: 14.1% ± 7.5% vs. 18.0% ± 8.3%, respectively [P = 0.0160]; CD86: 12.8% ± 6.8% vs. 16.3% ± 6.0%, respectively [P = 0.0087]; and CD40: 27.2% ± 10.6% vs. 35.6% ± 9.3%, respectively [P = 0.0009]). There were no significant differences noted in CD4-positive or CD4-positive /CD25-positive cell populations (Table 2). RT-PCR analysis demonstrated that the crossing points (CPs) of CD83 and IFN-γ were significantly higher, indicating their significantly lower expression levels in SNs compared with non-SNs (CD83: 24.3 ± 3.1 vs. 23.4 ± 1.2, respectively [P = 0.0338]; and IFN-γ: 29.2 ± 2.1 vs. 28.1 ± 1.7, respectively [P = 0.0425]) among metastasis-negative patients with breast carcinoma. There were no significant differences in the CPs of IL-12p40, IL-4, IL-10, or Foxp3 between SNs and non-SNs (Table 3).

Table 2. Comparison of the Immunologic Parameters Analyzed by Flow Cytometry between Sentinel and Nonsentinel Lymph Nodes in Patients with Lymph Node-Negative Breast Carcinoma
ExpressionNo. of patientsPercentage positive (mean ± SD)aP value
SNNon-SN
  • SD: standard deviation; SN: sentinel lymph node; plus;: positive; HLA-DR: human D-related leukemic antigen.

  • a

    Sentinel lymph nodes and nonsentinel lymph nodes were obtained using sentinel lymph node biopsy from lymph node-negative patients with breast carcinoma. Sentinel lymph node mononuclear cells were collected, and the immunologic parameters indicated were analyzed using flow cytometry. Percentages of the antigen-positive cell population (mean ± standard deviation) were compared between sentinel lymph nodes and nonsentinel lymph nodes.

CD83+182.4 ± 1.82.5 ± 1.50.8649
HLA-DR+733.7 ± 13.843.3 ± 10.90.0111
CD80+1414.1 ± 7.518.0 ± 8.30.0160
CD86+1212.8 ± 6.816.3 ± 6.00.0087
CD40+1527.2 ± 10.635.6 ± 9.30.0009
CD4+1951.2 ± 14.147.4 ± 12.60.2958
CD4+/CD25+194.2 ± 1.54.6 ± 1.70.2066
Table 3. Comparison of the Immunologic Parameters Analyzed by Reverse Transcriptase-Polymerase Chain Reaction between Sentinel Lymph Nodes and Nonsentinel Lymph Nodes in Patients with Lymph Node-Negative Breast Carcinoma
ExpressionCrossing points (mean ± SD)aP value
SN (n = 44)Non-SN (n = 16)
  • SD: standard deviation; SN: sentinel lymph node; IL: interleukin; IFN: interferon.

  • a

    Sentinel lymph nodes and nonsentinel lymph nodes were obtained using sentinel lymph node biopsy from lymph node-negative patients with breast carcinoma. RNA was extracted and subjected to quantitative reverse transcriptase-polymerase chain reaction analysis that was specific for the immunologic parameters indicated. Crossing points (mean ± standard deviation) were compared between sentinel lymph nodes and nonsentinel lymph nodes.

CD8324.3 ± 3.123.4 ± 1.20.0338
IL-12p4031.6 ± 2.531.4 ± 2.20.4739
IFN-γ29.2 ± 2.128.1 ± 1.70.0425
IL-431.1 ± 2.430.7 ± 1.40.3354
IL-1029.8 ± 3.029.4 ± 1.60.1892
FoxP327.0 ± 3.627.1 ± 1.80.7291

Comparison of Immunologic Parameters between Metastasis-Negative SNs and Metastasis-Positive SNs

Next, we compared the DC maturation status and T-cell responses between metastasis-negative SNs and metastasis-positive SNs from patients with breast carcinoma (Table 4). The CP of CD83 was significantly lower, indicating a significantly higher level of CD83 expression, in metastasis-positive SNs (22.8 ± 2.5) than in metastasis-negative SNs (24.3 ± 3.1; P = 0.0149). The CPs of IL-12p40 and IFN-γ also were significantly lower, indicating their significantly higher expression levels in positive SNs compared with negative SNs (IL-12p40: 30.0 ± 1.8 vs. 31.6 ± 2.5 [P = 0.0030]; and IFN-γ: 29.2 ± 2.1 vs. 27.9 ± 1.6 [P = 0.0018]). There was no difference noted with regard to the CP of IL-4 between metastasis-negative SNs and metastasis-positive SNs. However, the CPs of IL-10 and Foxp3 were significantly lower, indicating their significantly higher expression levels in metastasis-positive SNs compared with metastasis-negative SNs (IL-10: 28.4 ± 1.3 vs. 29.8 ± 3.0 [P = 0.0072]; and Foxp3: 25.0 ± 1.0 vs. 27.0 ± 3.6 [P = 0.0010]) (Table 4).

Table 4. Comparison of Immunologic Parameters Analyzed by Reverse Transcriptase-Polymerase Chain Reaction between Metastasis-Negative and Metastasis-Positive Sentinel Lymph Nodes
ExpressionCrossing points (mean ± SD)aP value
Metastasis-negative SN (n = 44)Metastasis-positive SN (n = 14)
  • SD: standard deviation; SN: sentinel lymph node; IL: interleukin; IFN: interferon.

  • a

    Sentinel lymph nodes were obtained using sentinel lymph node biopsy from patients with breast carcinoma. RNA was extracted and subjected to quantitative reverse transcriptase-polymerase chain reaction analysis that was specific for the immunologic parameters indicated. Crossing points (mean ± standard deviation) were compared between metastasis-negative sentinel lymph nodes and metastasis-positive sentinel lymph nodes.

CD8324.3 ± 3.122.8 ± 2.50.0149
IL-12p4031.6 ± 2.530.0 ± 1.80.0030
IFN-γ29.2 ± 2.127.9 ± 1.60.0018
IL-431.1 ± 2.430.5 ± 2.00.0539
IL-1029.8 ± 3.028.4 ± 1.30.0072
FoxP327.0 ± 3.625.0 ± 1.00.0010

Detection of Micrometastasis in SNs using RT-PCR Specific for Mammaglobin and Immunologic Analysis

SNs that were diagnosed histologically as metastasis-negative (N0) were analyzed further by RT-PCR analysis that was specific for mammaglobin to ensure microscopically negative but RT-PCR-positive micrometastasis. Among 56 SNs with N0 status, 7 SNs showed positive bands in the RT-PCR-specific analysis for mammaglobin. Those 7 SNs with micrometastasis were transferred from the metastasis-negative population to the metastasis-positive population, and their immunologic parameters of DC maturation status and T-cell responses were reanalyzed to compare between SNs and non-SNs among the corrected metastasis-negative populations (Table 5) and between the corrected metastasis-negative SNs and metastasis-positive SNs (Table 6). This correction for SNs with micrometastasis made the above-described differences in immunologic parameters remarkable. By using RT-PCR analysis, CPs of CD83 and IFN-γ were significantly much higher, indicating their significantly lower expression levels in SNs compared with non-SNs in the corrected metastasis-negative patients, and P values were augmented (CD83: 24.5 ± 3.1 vs. 23.4 ± 1.2 [P = 0.0206]; and IFN-γ: 29.2 ± 2.2 vs. 28.2 ± 1.5 [P = 0.0292]) (Table 5). When the 36 corrected metastasis-negative SNs were compared with the 21 metastasis-positive SNs, the CPs of CD83, IL-12p40, IFN-γ, IL-10, and Foxp3, but not IL-4, were significantly lower, indicating their significantly higher expression levels in metastasis-positive SNs compared with metastasis-negative SNs, and P values also were augmented (CD83: 22.9 ± 2.1 vs. 24.5 ± 3.1 [P = 0.009]; IL-12p40: 30.1 ± 2.1 vs. 31.7 ± 2.5 [P = 0.0021]; IFN-γ: 27.9 ± 1.8 vs. 29.2 ± 2.2 [P = 0.0008]; IL-10: 29.1 ± 1.7 vs. 29.9 ± 3.0 [P = 0.0192]; and Foxp3: 25.0 ± 1.6 vs. 27.3 ± 3.3 [P = 0.0011]) (Table 6).

Table 5. Comparison of Immunologic Parameters Analyzed by Reverse Transcriptase-Polymerase Chain Reaction Analysis between Sentinel Lymph Nodes and Nonsentinel Lymph Nodes in Patients with Micrometastasis-Negative with Breast Carcinomaa
ExpressionCrossing points (mean ± SD)bP value
SN (n = 36)Non-SN (n =15)
  • SD: standard deviation; SN: sentinel lymph node; IL: interleukin; IFN: interferon.

  • a

    Sentinel lymph nodes and nonsentinel lymph nodes were obtained using sentinel lymph node biopsy from patients with negative lymph node status with breast carcinoma. RNA was extracted and subjected to quantitative reverse transcriptase-polymerase chain reaction (RT-PCR) analysis that was specific for the immunologic parameters indicated. Micrometastases were identified by RT-PCR analysis that was specific for mammaglobin.

  • b

    Seven micrometastasis-positive sentinel lymph nodes were corrected by moving them from the metastasis-negative sentinel lymph node population to the metastasis-positive sentinel lymph node population.

CD8324.5 ± 3.123.4 ± 1.20.0206
IL-12p4031.7 ± 2.531.6 ± 2.00.4821
IFN-γ29.2 ± 2.228.2 ± 1.50.0292
IL-431.2 ± 2.930.9 ± 1.60.3468
IL-1029.9 ± 3.029.6 ± 1.50.2643
FoxP327.3 ± 3.327.0 ± 1.80.9311
Table 6. Comparison of the Immunologic Parameters Analyzed using Reverse Transcriptase-Polymerase Chain Reaction between Metastasis-Negative and Metastasis-Positive Sentinel Lymph Nodes after Correcting for Micrometastasisa
ExpressionCrossing points (mean ± SD)bP value
Metastasis-negative SNs (n = 36)Metastasis-positive SNs (n = 21)
  • SD: standard deviation; SN: sentinel lymph node; IL: interleukin: IFN: interferon.

  • a

    Sentinel lymph nodes were obtained using sentinel lymph node biopsy from patients with breast carcinoma. RNA was extracted and subjected to reverse transcriptase-polymerase chain reaction (RT-PCR) analysis that was specific for the immunologic parameters indicated. Micrometastases were identified by RT-PCR analysis that was specific for mammaglobin.

  • b

    Seven micrometastasis-positive sentinel lymph nodes were corrected by moving them from the metastasis-negative sentinel lymph node population to the metastasis-positive sentinel lymph node population. Crossing points (mean ± standard deviation) were compared between the positive and negative sentinel lymph node groups.

CD8324.5 ± 3.122.9 ± 2.10.009
IL-12p4031.7 ± 2.530.1 ± 2.10.0021
IFN-γ29.2 ± 2.227.9 ± 1.80.0008
IL-431.2 ± 2.930.6 ± 2.10.1123
IL-1029.9 ± 3.029.1 ± 1.70.0192
FoxP327.3 ± 3.325.0 ± 1.60.0011

DISCUSSION

Exploration of the immunologic status of SNs, including DC maturation and T-cell responses, is one of the most interesting issues in recent tumor immunology. We observed that, when comparing SNs and non-SNs from lymph node-negative patients, populations of HLA-DR, CD80, CD86, and CD40 cells by FCM analysis and expression levels of CD83 and IFN-γ by RT-PCR analysis were lower in SNs compared with non-SNs. When comparing metastasis-negative SNs with metastasis-positive SNs, the expression levels of CD83, IL-12p40, and IFN-γ by RT-PCR analysis were higher in metastasis-positive SNs than in metastasis-negative SNs. Moreover, these differences were augmented when the values were corrected for micrometastases that were detected as microscopically negative but were positive in the mammaglobin-specific RT-PCR analysis of SNs. These results clearly indicate that the immunologic status of SNs, including DC maturation and Th-1 responses, is depressed in SNs before metastasis but is up-regulated after metastasis occurs.

Previous studies have addressed DC status and T-cell responses in SNs in several types of carcinoma. In melanoma studies, markers of DC activation were analyzed by using immunohistochemistry for S-100 protein or RT-PCR analysis that was specific for CD80, CD86, and CD40 and their corresponding T-cell receptors (CTLA-4 and CD28), and the markers were decreased in the microenvironment of SNs compared with non-SNs.23–25 Those results suggest that the immunologic paralysis indicated in SNs may be induced by the release of immunosuppressive factors from primary melanoma and may facilitate the implantation and growth of melanoma cells in SNs. More recently, Lee et al.26 provided molecular evidence of cytokine-mediated SN immunosuppression associated with the presence of melanoma and showed that SN immunosuppression potentially may be reversed by cytokine therapy using granulocyte macrophage-colony stimulating factor.

In breast carcinoma studies, Huang et al.27 compared S-100-positive DCs in SNs and non-SNs and demonstrated that the SNs examined in their study had reduced paracortical areas, reduced densities of paracortical DCs, and reduced frequencies of S-100-positive DCs with a predominance of immature, poorly dendritic DCs, although those authors did not identify tumor-free SNs versus tumor-containing SNs in their study. They concluded that, based on the morphology and density of the paracortical area and the number of paracortical S-100-positive DCs, SNs are immunomodulated compared with non-SNs in patients with breast carcinoma. Schule et al.28 reported down-regulation of the CD3-ζ chain to varying degrees in SNs and suggested that SNs may be an appropriate location for detecting early-stage immunologic down-regulation, suggesting the possibility of selecting patients with breast carcinoma who may benefit from immunotherapy. These observations are consistent with our current results.

In contrast, Ishigami et al.29 investigated SN immunity by using immunohistochemistry for natural killer cells, DCs, Ki-67-positive cells, and CD3-ζ chain in patients with gastric carcinoma and indicated that SNs were not suppressed. Those authors proposed that attention should be paid to cell types of carcinoma when SN immunity is discussed. More recently, Poindexter et al.30 also investigated the immunologic status of banked SN samples by using immunohistochemistry for CD3, HLA Class II, CD83, CD1a, IL-10, and IL-12 in patients with breast carcinoma. They reported that metastasis-negative SNs are competent immunologically compared with metastasis-positive SNs and potentially are a site of tumor-specific T-cell activation. Those observations appear to be inconsistent with melanoma studies and with our own observations. They investigated SNs by using immunohistochemical analysis,30 which differs from the RT-PCR and FCM methods used in the current study. Akira et al.31 demonstrated that the information obtained by immunohistochemistry regarding intracellular protein synthesis was limited, because the protein may be secreted, deposited, or posttranslationally modified. Moreover, in the above-described melanoma studies and in our breast carcinoma studies, SNs from patients who with clinical N0 status were investigated. We attempted to elucidate the immune responses of early-stage SN metastasis, whereas the SN metastasis status of the patients studied by Poindexter et al.30 was not stated. More advanced-stage SN metastasis can destroy and depress further responses of the SN immune system. Advanced-stage SN metastasis may make the immunologic status of metastasis-negative SNs relatively higher. In addition, Poindexter et al. compared SNs between those from breast carcinoma and from normal breast and observed no difference in DC numbers, suggesting that SNs from breast carcinoma may not be immunocompetent compared with SNs from normal breast. In our study, non-SNs, which differ from SNs from normal breast, showed increased DC maturation and up-regulation of Th-1 responses compared with SNs when they were analyzed in patients with lymph node-negative breast carcinoma. Collectively, the differences in the results described above may be explained by the differences in the methods used to analyze SNs and possibly by the stage of SN metastasis. Further functional analyses, which already are in progress in our laboratory, may clarify the immunologic status of SNs.

It is noteworthy that, in addition to DC maturation and up-regulation of Th-1 responses, we observed the up-regulation of IL-10 in metastasis-positive SNs compared with metastasis-negative SNs. Moreover, we observed the up-regulation of Foxp3 expression in metastasis-positive SNs. These results appear to be conflicting, because IL-10 belongs to a Th-2 response,11 and Foxp3 expression indicates the presence of regulatory T-cells,12 and both can suppress DC activity and Th-1 responses.11, 13 Thus, there were two reciprocal immune responses—one antimetastatic and the other antidefense—that coexisted in the metastasis-positive SNs in our results. Poindexter et al.30 identified reciprocal IL-12-expressing cells and IL-10-expressing cells in metastasis-negative SNs from patients with breast carcinoma. One possible explanation is that DC maturation and Th-1 up-regulation by SN metastasis may trigger the up-regulation of a Th-2 response, including IL-10 and T-reg cells in the form of an immune network. Fujimoto et al.32 reported that the stimulation of peripheral blood mononuclear cells with OK-432 (a streptococcal preparation) induced a strong IL-12 response, which was followed by IL-10 expression. Therefore, we may be looking at both aspects of immune regulation. Although mature DCs, in contrast to immature DCs, are resistant to the effects of IL-10 and can present antigens to T-cells to achieve antitumor immunity,33 these antidefense responses may facilitate the growth of metastases in SNs and allow them to range from subclinical to clinical metastases. In the design of immunotherapy for SN activation, we should pay more attention not only to DC maturation and up-regulation of Th-1 but also to the down-regulation of Th-2 and T-reg responses.

Cellular immune responses, ranging from DC maturation to Th-1 responses, may be less active in SNs than in non-SNs in patients with breast carcinoma before they develop metastasis. Once metastasis is established in SNs, DC maturation is triggered and is followed by the up-regulation of Th-1 responses, which may reflect antigen-specific immune responses in SNs. Unlike DC maturation and Th-1 responses, the up-regulation of Th-2 and regulatory T-cell responses are developed in parallel in metastasis-positive SNs. Further functional analysis will be needed to design immunotherapy for SN activation and regulation of SN metastasis in patients with breast carcinoma.

Ancillary