Anatomic Pathology / DENDRITIC CELLS

Myxofibrosarcoma is a malignant tumor with distinctive histologic features and is believed to be derived from fibroblasts. The function of infiltrating myeloid cells in myxofibrosarcoma is poorly understood. It previously has been shown that a combination of dendritic morphologic features and expression of the C-type lectin, dendritic cell–specific intercellular adhesion molecule-3 grabbing nonintegrin (DC-SIGN), is useful for identifying DC populations in tissue sections. In the present study, we found that 3% to 61% (median, 22%) of cells in myxofibrosarcomas express DC-SIGN and have dendritic morphologic features. These DC-SIGN–positive cells are not in cell cycle and are consistent with infiltrating DCs. The percentage of DCs in myxofibrosarcomas is independent of tumor grade. It previously has been shown that DC-SIGN–positive cells are either immature DCs or DCs that predominantly activate TH2 cells, both subsets likely to give rise to ineffective antitumor responses. The DC-SIGN–positive DCs that we have identified in myxofibrosarcoma may, therefore, be involved in the induction of ineffective immune responses or even tolerance to tumor antigens. Myxofibrosarcomas are malignant tumors with distinctive histologic features and are believed to be derived from fibroblasts.1 They commonly show a nodular growth pattern and are composed of round or stellate tumor cells with indistinct cell margins, pale eosinophilic cytoplasm, and hyperchromatic atypical nuclei embedded in a myxoid matrix.1 Myxofibrosarcomas vary from a hypocellular, mainly myxoid, and purely spindle-cell appearance (low-grade neoplasms) to a high-grade, pleomorphic (malignant fibrous histiocytoma–like) tumor with multinucleated giant cells, high mitotic activity, and areas of necrosis.1 The presence of myeloid cells in myxofibrosarcoma is well described, although the function of such cells remains unclear. Myxofibrosarcoma may therefore be regarded as a useful model for studying tumor infiltration by myeloid cells. Determination of whether infiltrating myeloid cells in a neoplasm are macrophages or dendritic cells (DCs) may have important consequences for our understanding of antitumor immunity. DCs are specialized antigen-presenting cells capable of activating naive T lymphocytes and, thus, initiating an immune response.2 Macrophages, on the other hand, are likely to perpetuate only an existing immune response.2 Furthermore, the phenotype of a DC may vary, and this variation may permit the initiation of functionally different immune responses ranging from a cytotoxic response to immune tolerance.2 It has proven difficult to identify a molecule that labels the majority of DCs but does not immunostain macrophages.2 The recently identified C-type lectin, DC-specific intercellular adhesion molecule-3 grabbing nonintegrin (DC-SIGN), has emerged as a useful marker of DCs, and recent work has suggested that a combination of DC-SIGN immunostaining Anatomic Pathology / ORIGINAL ARTICLE Am J Clin Pathol 2003;119:540-545 541 541 DOI: 10.1309/JEB7DGHH01J11VUM 541 © American Society for Clinical Pathology and dendritic morphologic features identifies the majority of DCs in peripheral tissues.3 DC-SIGN also may be present on specialized populations of macrophages in the lung and placenta,4,5 but we have failed to identify expression of DCSIGN by macrophages at any other location, including sites of inflammation. DC-SIGN expression also provides information about DC function.4,6 DC-SIGN is present predominantly on immature DCs, which express few costimulatory molecules and are likely to be poor activators of T lymphocytes.4 Several DC lineages have been identified in humans,7 and recent work has demonstrated that DC-SIGN also is expressed by the DCs that predominantly activate TH2 cells (DC2) lineage.4 This finding is consistent with data demonstrating that DC-SIGN expression is up-regulated by both interleukin-4 and interleukin-13, cytokines intimately involved in the TH2 axis of immunity. 4,8 Both immature DCs and DC2s are likely to lead to relatively ineffective antitumor immune responses.7 In the present study, we demonstrated extensive DCSIGN expression on cells with dendritic morphologic features in all grades of myxofibrosarcoma. By using antibodies against the cell cycle marker minichromosome maintenance protein-2 (Mcm2),9-11 we performed double-labeled fluorescence confocal microscopy to show that the DC-SIGN–positive cell population is not in cell cycle, even in high-grade myxofibrosarcoma, and is, therefore, consistent with an infiltrating DC population. Our findings raise pertinent questions about the role of DC-SIGN–positive DCs in myxofibrosarcoma and potentially in other malignant neoplasms. Materials and Methods Selection of Tissue Samples Archival paraffin blocks were retrieved from 43 cases of myxofibrosarcoma from the Department of Histopathology, Royal Marsden Hospital, London, England. Sections were reviewed histologically, and tumors were stratified by grade, using the criteria of Trojani and colleagues.12 There were 23 low-grade tumors, 8 intermediate-grade tumors, and 12 high-grade tumors. Lymph node tissue was obtained from Addenbrooke’s Hospital, Cambridge, England. Tissue was obtained after receiving approval from the Local Research Ethics Committee. Single Immunostaining of Paraffin Sections We stained 5-μm paraffin sections by the indirect immunoperoxidase technique using rabbit anti–DC-SIGN polyclonal serum and preimmune serum on serial sections as a negative control, as described previously.4 Immunofluorescent Staining for Confocal Microscopy To perform double immunostaining for confocal microscopy, paraffin sections were immunostained with rabbit polyclonal anti–DC-SIGN serum4 and mouse monoclonal anti-Mcm2 antibody.10,11 Pressure cooking was used for antigen retrieval, as described previously.4,9 Sections were incubated in both primary antibodies overnight in a mixture of tris(hydroxymethyl)aminomethane (Tris)buffered saline (TBS), 1% bovine serum albumin, and 10% normal goat serum. Following rinsing in TBS, fluorescein isothiocyanate–conjugated goat antimouse antibody (Sigma, Poole, England) and Alexa 594 conjugated goat antirabbit antibody (Molecular Probes, Eugene, OR) were used for detection. Sections were mounted in fluorescence mounting medium (DAKO, Glostrup, Denmark), and images were obtained using serial scanning techniques with a confocal laser scanning microscope TCS 4D (Leica Lasertechnik, Heidelberg, Germany). Lymph node tissue was used as a positive control for both antibodies. As a negative control, serial sections of each tumor were immunostained with preimmune rabbit serum and/or omission of the anti-Mcm2 mouse primary antibody. Cell Counting and Statistical Analysis For each tumor, the number of cells positive for DCSIGN and the number of cells negative for DC-SIGN were recorded by each observer (K.L., B.R.) in 5 microscopic fields (×400 magnification) representative of the tumor as a whole, in terms of overall appearance and cellularity. No areas showing acute or chronic inflammation were included. We counted 600 to 1,500 cells in each case. For each tumor, the percentage of DC-SIGN–expressing cells was calculated by dividing the total number of cells positive for DC-SIGN by the total number of cells counted. All counting was performed by 2 independent observers who were blinded to tumor grade and to the findings of the other observer. The overall DC-SIGN labeling indices represented the means of the counts of the 2 observers. The Kruskal-Wallis test (Sigma Stat, version 10, SPSS, Chicago, IL) was used to determine whether there was a difference in the percentage of DC-SIGN–expressing cells between different grades of myxofibrosarcoma.