Dendritic Cell Function in Sentinel Nodes

Intraoperative lymphatic mapping and sentinel lymphadenectomy has become an increasingly popular technique for staging the regional lymph nodes in early-stage melanoma. This operative technique allows for detailed pathologic analysis of the first (or sentinel) lymph node in direct connection with the primary tumor, and provides a unique opportunity for assessing potential immunologic interactions between the primary tumor and regional lymph node basin. We performed lymphatic mapping and sentinel lymphadenectomy on 25 patients with early-stage melanoma and resected an additional nonsentinel node in each case. Sentinel and nonsentinel nodes were evaluated by routine pathologic analysis. A portion of each node was processed for expression of the dendritic markers of activation CD80, CD86, and CD40, and their corresponding T-cell receptors CTLA-4 and CD28. Of 25 patients undergoing lymphatic mapping and sentinel lymphadenectomy, 20 (80%) had matched sentinel and nonsentinel nodes. A total of 26 matched lymph node sets were obtained: three pairs from one patient and two from an additional two patients. Reverse transcription polymerase chain reaction analyses of corresponding sections of the sentinel and nonsentinel nodes demonstrated a marked reduction in semiquantitative expression of CD80 (77%), CD86 (77%), and CD40 (85%), as well as CTLA-4 (88%) and CD28 (85%) in sentinel as compared to nonsentinel nodes. The diminished expression of the dendritic cell markers appeared to be unrelated to the B-cell (CD20) and T-cell (CD2) expression. Lymphatic mapping and sentinel lymphadenectomy allows for detailed pathologic and molecular characterization of sentinel nodes. Our results suggest a quantitative reduction in dendritic cell markers in sentinel as compared to nonsentinel nodes, which may be important in the immunologic interaction between the primary site and regional lymph node basin and may also serve as useful criteria for identifying sentinel nodes. [ONCOLOGY 16(Suppl 1):27-31, 2002]

The management of the regional lymph nodes in early-stage melanoma has been revolutionized by the development of lymphatic mapping and sentinel lymphadenectomy. This minimally invasive operative technique has replaced elective lymph node dissection and provides for precise identification and evaluation of the lymph nodes (sentinel node) in direct connection with the primary melanoma. Since the initial feasibility study by Morton and colleagues in 1992, multiple investigators have validated the accuracy of the technique and the ease with which this technology can be transferred to a variety of cancers, including colon and breast cancer.[1-3]

Background

While lymphatic mapping and sentinel lymphadenectomy provides for detailed pathologic analysis of the sentinel nodes, it also allows investigators a unique opportunity to evaluate direct interactions between the microenvironment of the primary tumor and its relationship to the regional lymph nodes. Several clinical studies have suggested that physical alterations in the microscopic appearance of the regional lymph nodes resected concurrently with the primary tumor may be important for determining patient prognosis.[4,5]

Cochran and associates first described that both a physical and a functional alteration in the regional lymph nodes exist in early-stage melanoma. Prior to the development of lymphatic mapping and sentinel lymphadenectomy, his group demonstrated a differential response of proximal vs distal lymph node-derived lymphocytes from surgical specimens after in vitro stimulation with interleukin-2 and phytohemagglutinin in mixed lymphocyte reaction. These results suggested a diminished T-cell response from lymphocytes found in closer proximity to the primary melanoma.[6-8] The mechanism of this attenuated T-cell response at that time was unknown.

Recent studies have pointed toward the importance of dendritic cells as the initiating event in the immune response to malignancy.[9] In melanoma, dendritic cell maturation, likely from passage of these cells from the skin to the regional lymph nodes, causes changes in the function of these cells from antigen processing to presentation. Associated with these changes are upregulation of the costimulatory molecule CD40 (a relatively early change) and the B7 molecules CD80 (B7.1) and CD86 (B7.2) necessary for T-cell activation.[10,11] Activation of dendritic cells occurs with promotion of T-cell maturation and expression of their corresponding receptors to the B7 molecules (CD28 and CTLA-4). The maturation of both dendritic and T cells is thought to be instrumental in the T-cell-driven response to tumor presentation.

While the mechanisms that control tumor-derived dendritic cell activation is yet unknown, the development of lymphatic mapping and sentinel lymphadenectomy provides a unique opportunity to analyze more specific relationships of dendritic cells directly connected to the primary (sentinel nodes) and those more distal to this site (nonsentinel nodes). Our hypothesis is that sentinel nodes would have a diminished expression of functional dendritic cell markers of activation as compared to nonsentinel nodes, based on the earlier studies demonstrating a suppressed T-cell response.

Methods

A total of 24 patients with American Joint Committee on Cancer (AJCC) stage I and II melanoma were considered for lymphatic mapping and sentinel lymphadenectomy after review of their pathology specimens and a thorough clinical exam demonstrating no sign of regional lymph node or distant metastases. No patient had a history of myeloproliferative disease or primary or secondary immunodeficiency.

All patients underwent lymphatic mapping and sentinel lymphadenectomy as previously described.[1,3] In brief, patients underwent preoperative cutaneous lymphoscintigraphy on the day of surgery.[12] Patients were taken to the operating room after informed consent. In brief, lymphatic mapping and sentinel lymphadenectomy was performed with the combined use of blue dye and a radiopharmaceutical for probe-directed sentinel node biopsy. After excision of the sentinel node, the surgical wound was explored for adjacent, secondary, nonblue, nonradioactive, nonsentinel nodes. These nonsentinel nodes were usually identified within several centimeters of the sentinel nodes. In three cases, nonsentinel lymph nodes could not be identified. In two cases, the sentinel nodes were less than 1 cm in size and, as written in our protocol, we elected not to use tissue that may have been important for routine pathology review.

A total of 26 paired sentinel and nonsentinel nodes were analyzed, each pair from a separate lymph node basin. One patient had three pairs and two patients had two pairs. Fresh lymph nodes were immediately processed with complete freezing and tangential sections cut approximately 4 mm thick from both sides of the nodes parallel to the longest axis of the specimens. A single specimen was processed for reverse transcription polymerase chain reaction (RT-PCR) analyses. The remaining specimens were stored for other studies.

In brief, RNA was isolated from lymph node extracts using TRI Reagent (Molecular Research Center, Inc, Cincinnati). Total RNA was converted to cDNA with M-MLV reverse transcriptase and random hexamer (Promeda, Madison, Wisconsin). To assess the amount of mRNA of markers for dendritic cell activation, PCR was performed for CD80, CD86, CD40, CTLA-4, and CD28, and for the constituitively expressed housekeeping gene coding for GAPDH.

GAPDH primers (sense: 5¢-TGA AGG TCG GAG TCA ACG GAT TTG G-3¢; antisense: 5¢-GTT CAC ACC CAT GAC GAA CAT GG -3¢) were used as controls for semiquantification. CD2 and CD20 also served as quantitative B-cell and T-cell controls: CD2 (sense: 5¢-GGT CAT CGT TCC CAG GCA CCT AGT-3¢; antisense: 5¢-TGG TGT GAT GGA GCT CTC TGA GGA-3¢) and CD20 (sense: 5¢-CGT GCT CCA GAC CCA AAT CTA ACA-3¢; antisense: 5¢-GCG TGA CAA CAC AAG CTG CAA-3¢). Primers included CD80 (sense: 5¢-GTG GCA ACG CTG TCC TGT GGT-3¢; antisense: 5¢-CCA GGA GAG GTG AGG CTC-3¢), CD86 (sense: 5¢-CCA AAG CCT GAG TGA GCT AGT-3¢; antisense: 5¢-CTT AGG TTC TGG GTA ACC GTG-3¢), CD40 (sense: 5¢-TGG GGC TGC TTG CTG ACC GC-3¢; antisense: 5¢-CCA AAG CCG GGC GAG CAT GA-3¢), CTLA-4 (sense: 5¢-AGT ATG CAT CTC CAG GCA AAG C-3¢; antisense: 5¢-CCA GAG GAG GAA GTC AGA ATC TG-3¢), and CD28 (sense: 5¢-GTT TGA GTG CCT TGA TCA TGT GC-3¢; antisense: 5¢-GGC GAC TGC TTC ACC AAA ATC-3¢).

Amplification of 40 cycles was utilized in all samples, with PCR performed as previously described.[13,14] These conditions allowed sufficient linearity of PCR amplification. All PCR products were then separated by 2% agarose gel electrophoresis and detected by ethidium bromide fluorescence. Semiquantitation of band intensity was performed by comparison of densitometric readings of each band, correcting for GAPDH band intensity and background. Results are presented as ratios of sentinel node to nonsentinel node band intensity for each marker. Samples were run in triplicate to validate the accuracy of the results. A P value of less than .05 was considered significant.

Results

A total of 25 consecutive patients were entered into the study. In three basins, nonsentinel nodes were not identified, and in two the lymph nodes were too small for analysis. Three patients had dual lymphatic drainage at lymphatic mapping and sentinel lymphadenectomy and another had drainage to three basins, for a total of 26 matched sentinel and nonsentinel node pairs from 20 patients. The primary melanomas had a mean depth of 1.46 mm (range, 0.30 to 3.30 mm). Four of the 20 patients (20%) were found to have metastatic disease in the sentinel node.

Laser densitometry was used to assess the relative gene expression of the dendritic- and T-cell markers as compared to the housekeeping gene GAPDH. A majority of the dendritic- and T-cell markers were expressed in lower levels in sentinel nodes as compared to nonsentinel nodes: CD80 (20/26, 77%), CD86 (20/26, 77%), CD40 (22/26, 85%), CTLA-4 (23/26, 88%), and CD28 (22/26, 85%) (Figure 1). Depressed expression of one marker always occurred along with depression of at least one other marker. Four patients had metastases to the sentinel nodes. In all four cases, gene expression was diminished in sentinel nodes as compared to nonsentinel nodes for each of the five markers. CD2 and CD20 marker expression was evaluated in all cases. The relative expression of these T-cell and B-cell markers was no different for sentinel and nonsentinel nodes (Figure 2).

Discussion

The local immune suppression associated with primary cutaneous melanoma has been recognized for many years. Previous studies demonstrated significant differences in T-cell activation from regional lymph nodes draining a primary melanoma when lymphocytes were examined as a function of the distance from the primary tumor. Farzad et al[7] have shown that lymph node lymphocytes harvested from nodes closest to the primary melanoma have a significantly decreased ability to inhibit melanoma cell growth in vitro when compared to lymphocytes harvested from nodes more distal from the primary. In addition, Hoon et al[8] have previously shown significant decreases in [3H]thymidine uptake in lymphocytes harvested from nodes closest to a primary melanoma when compared to those taken from more distal nodes.

The impact of a primary melanoma on local antigen-presenting cells has been previously examined in both the local and regional environments. Stene et al[15] examined the population of Langerhans’ cells in resected specimens of benign and dysplastic nevi, melanoma in situ, early invasive melanomas, and deeply invasive melanomas using immunohistochemical staining with antibody to Leu-6. Comparisons were made between the surrounding normal skin and the region immediately adjacent to the primary lesion. Significant decreases in Langerhans’ cell absolute number and density were noted in the deeply invasive melanomas, with no differences noted among the other pathologic subtypes.

Dendritic Cell Frequency

Alteration in dendritic cell frequency in the regional draining lymph nodes has also been previously suggested by Huang et al.[16] Interdigitating dendritic cell frequency was noted to be dependent on distance from the primary melanoma in nodal resection specimens. Both a generalized "zone" of immunosuppression and a specific impact on the local and regional dendritic cells appear to be associated with primary cutaneous melanoma, although the agents responsible for this phenomenon are not yet well defined.

Lymphatic mapping and sentinel lymphadenectomy have significantly altered the management of the regional lymph nodes of patients with early-stage (I and II) melanoma. In addition, the ability to identify the first-draining lymph nodes from a primary tumor allows a more detailed analysis of the single node in direct anatomic continuity with the primary and any secreted tumor products. Our initial analysis of dendritic cell alterations in sentinel compared to nonsentinel nodes revealed a significant decrease in the number and density of paracortical dendritic cells.[16] In addition, the dendrites on paracortical dendritic cells from the sentinel nodes were noted to be significantly shorter or entirely absent when compared to the dendrites on dendritic cells from nonsentinel nodes. These quantitative and morphologic changes have been noted in patients with both breast cancer and melanoma.[6-8,16]

Dendritic Cell Activation Markers

In addition to the striking morphologic changes noted in the sentinel node, we hypothesized that the dendritic cells in the sentinel node would exhibit a marked reduction in ability to present antigens to the paracortical T cells. Therefore, the identified defect in the sentinel node would have possible physiologic consequences along with the noted histologic changes in the dendritic cells. We chose to examine markers of dendritic cell activation, the costimulatory molecules CD40, CD80 (B7.1), and CD86 (B7.2), and the corresponding T-cell receptors CD28 and CTLA4. CD40, CD80, and CD86 are noted to be up-regulated as the dendritic cell matures from an antigen-processing cell in the regional skin to an antigen-presenting cell concentrated in the paracortical portion of the regional nodal basin.[10,11,17,18]

CD40 is a cell surface receptor in the tumor necrosis factor (TNF) receptor family, which is generally expressed in low levels in immature dendritic cells. Up-regulation of CD40 is associated with dendritic cell maturation and a subsequent increase in the costimulatory B7 molecules CD80 and CD86.[11,19] All three markers have been shown to be significantly up-regulated by multiple endogenous activators of dendritic cells, including stressed, virally infected or necrotic cells, Escherichia coli lipopolysaccharide, and interferon-alpha.[18]

CD40 ligation has also been shown to have a significant antitumor effect, which is completely abrogated with anti-CD40L antibodies.[19] In addition, dendritic cell-dependent T-cell activation has been shown to be significantly inhibited by both anti-CD80 and anti-CD86 antibodies and by the supernatant of autologous and allogeneic melanoma cell cultures.[20] While these markers are not absolutely specific to dendritic cells, they have significant physiologic and immunologic functions, are consistently associated with dendritic cell maturation, and serve as excellent markers for dendritic cell activation.

In our studies, RT-PCR analyses demonstrated a significant number of the matched samples to have a decreased expression across all the markers of activation. The down-regulation of dendritic cell activation within the sentinel node may lead to a preferential site of metastases, not only anatomically, but also physiologically. The clinical relevance of this selective sentinel node immunosuppression has yet to be defined; however, the possible reversal of this defect has important immunologic implications. We were concerned that these findings could represent only artifact from the skin biopsy or a result of lymphatic mapping and sentinel lymphadenectomy. Biopsies of the original melanoma were made up to 6 weeks prior to lymphatic mapping and sentinel lymphadenectomy, by either excision or shave techniques. The timing or type of biopsy procedure had no effect on our results (data not shown). We have also speculated that either the technetium-labeled sulfur colloid or isosulfan blue dye used to identify the sentinel node may have influenced our results. The dose of gamma radiation used for lymphatic mapping and sentinel lymphadenectomy is not tissue toxic, nor is the blue dye. The timing of injection of either agent did not correlate with our results.

Clinical Applications

The concept of reversing the morphologic and functional alterations in the sentinel node may be important for understanding the interaction of the primary tumor and regional lymph node immunity. A variety of cytokines, including granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin-4, and TNF, have been shown to be potent stimulators of dendritic cell growth and activation. We have recently completed a pilot study evaluating the impact of a single dose of peritumoral intradermal GM-CSF (Sargramostim, Immunex Corporation, Seattle) on the morphologic alterations already noted within the sentinel node. Fifteen patients with early-stage melanoma underwent lymphatic mapping and sentinel lymphadenectomy preceded by a single injection of GM-CSF (either 100, 150, or 250 mg/m²) from 2 to 5 days prior. Patients receiving GM-CSF had a complete abrogation of the morphologic dendritic cell alterations in the sentinel nodes with a return to a normal nonsentinel node dendritic cell population (manuscript in preparation).

The long-term clinical impact of this therapy is uncertain; however, GM-CSF has been successfully utilized as a tumor vaccine adjuvant and in gene therapy of advanced melanoma.[21,22] Of note, GM-CSF has also been recently shown to significantly improve overall and disease-free survival of stage III and IV melanoma patients, compared to matched historical controls.[23] Therefore, GM-CSF may have an overall systemic benefit, in addition to the ability to reverse the previously identified dendritic cell down-regulation in sentinel nodes. Further study of this agent’s impact on dendritic cell function in the locoregional tumor environment is warranted.

Conclusion

In conclusion, we have demonstrated decreased markers of dendritic cell activation in the microenvironment of the sentinel nodes compared to the nonsentinel nodes of patients with cutaneous melanoma. The factors responsible for these changes are not known, and represent an active area of our current research. In addition, the clinical significance of these alterations, and more importantly, the significance of reversing these alterations, require further study.

References:

1. Morton DL, Wen D-R, Wong JH, et al: Technical details of intraoperative lymphatic mapping for early stage melanoma. Arch Surg 127:392-399, 1992.

2. Giuliano AE, Haigh PI, Brennan MB, et al: Prospective observational study of sentinel lymphadenectomy without further axillary dissection in patients with sentinel node-negative breast cancer. J Clin Oncol 18:2553-2559, 2000.

3. Morton DL, Thompson JF, Essner R, et al: Validation of the accuracy of intraoperative lymphatic mapping and sentinel lymphadenectomy for early-stage melanoma. A multicenter trial. Ann Surg 230:453-465, 1999.

4. Furukawa T, Watanabe S, Kodama T, et al: T-zone histocytes in adenocarcinoma of the lung in relation to postoperative prognosis. Cancer 56:2651-2656, 1985.

5. Tsujitani S, Furukawa T, Tamada R, et al: Langerhans cells and prognosis in patients with gastric carcinoma. Cancer 59:501-505, 1987.

6. Cochran AJ, Wen D-R, Farzad Z, et al: Immunosuppression by melanoma cells as a factor in the generation of metastatic disease. Anticancer Res 9:859-864, 1989.

7. Farzad Z, McBride WH, Ogbechi H, et al: Lymphocytes from lymph nodes at different distances from human melanoma vary in their capacity to inhibit/enhance tumor cell growth in vitro. Melanoma Res 7:S59-S65, 1997.

8. Hoon DSB, Korn EL, Cochran AJ: Variations in functional immunocompetence of individual tumor-draining lymph nodes in humans. Cancer Res 47:1740-1744, 1987.

9. Almand B, Resser JR, Lindman B, et al: Clinical significance of defective dendritic cell differentiation in cancer. Clin Cancer Res 6:1755-1766, 2000.

10. Banchereau J, Steinman RM: Dendritic cells and the control of immunity. Nature 392:245-252, 1998.

11. Van Kooten C, Banchereau J: Functions of CD40 on B cells, dendritic cells, and other cells. Curr Opin Immunol 9:330-337, 1997.

12. Essner R, Bostick PJ, Glass EC, et al: Standardized probe-directed sentinel node dissection in melanoma. Surgery 127:26-31, 2000.

13. Murata K, Dalakas MC: Expression of the costimulatory molecule BB-1, the ligands CTLA-4 and CD28, and their mRNA in inflammatory myopathies. Am J Pathol 155:453-460, 1999.

14. Behrens L, Kerschensteiner M, Misgeld T, et al: Human muscle cells express a functional costimulatory molecule distinct from B7.1 (CD80) and B7.2 (CD86) in vitro and in inflammatory lesions. J Immunol 161:5943-5951, 1998.

15. Stene MA, Babajanians M, Bhuta S, et al: Quantitative alterations in cutaneous Langerhans cells during the evolution of malignant melanoma of the skin. J Invest Dermatol 91:125-128, 1988.

16. Huang R-R, Wen D-R, Guo J, et al: Selective modulation of paracortical dendritic cells and T-lymphocytes in breast cancer sentinel lymph nodes. Breast J 6:225-232, 2000.

17. Dieu-Nosjean M-C, Vicari A, Lebecque S, et al: Regulation of dendritic cell trafficking: A process that involves the participation of selective chemokines. J Leukoc Biol 66:252-262, 1999.

18. Gallucci S, Lolkema M, Matzinger P: Natural adjuvants: Endogenous activators of dendritic cells. Nature Med 5:1249-1255, 1999.

19. Nakajima A, Kodama T, Morimoto S, et al: Antitumor effect of CD40 ligand: elicitation of local and systemic antitumor responses by IL-12 and B7. J Immunol 161:1901-1907, 1998.

20. Chang JWC, Vaquerano JE, Zhou Y-M, et al: Characterization of dendritic cells generated from peripheral blood of patients with malignant melanoma. Anticancer Res 19:1815-1820, 1999.

21. Kurane S, Arca MT, Aruga A, et al: Cytokines as an adjuvant to tumor vaccines: Efficacy of local methods of delivery. Ann Surg Oncol 4:579-585, 1997.

22. Armstrong CA, Botella R, Galloway TH, et al: Antitumor effects of granulocyte-macrophage colony-stimulating factor production by melanoma cells. Cancer Res 56:2191-2198, 1996.

23. Spitler LE, Grossbard ML, Ernstoff MS, et al: Adjuvant therapy of stage III and IV malignant melanoma using granulocyte-macrophage colony-stimulating factor. J Clin Oncol 18:1614-1621, 2000.