Follicular dendritic cell sarcoma

INTRODUCTION — Follicular dendritic cell sarcoma (FDCS) is a rare low-grade sarcoma that most commonly presents as a slowly growing, painless mass with the histologic appearance of spindle-shaped cells in a whorled pattern. Several hundred cases have been reported since it was first described in 1986 [1]. Occasional cases have an inflammatory presentation associated with fever and weight loss, and high-grade histologic features with cellular atypia.

FDCS is now recognized to be a low-grade sarcoma of mesenchymal dendritic cell origin [2]. Earlier classification schemes had erroneously categorized it as a histiocytic or dendritic cell neoplasm of myeloid origin.

The term histiocytic neoplasm has historically been used to refer to tumors associated with monocyte/macrophage and dendritic cell phenotypes, including entities that are malignant (eg, FDCS, histiocytic sarcomas, interdigitating dendritic cell sarcoma, Langerhans cell sarcoma, indeterminate cell sarcomas) and benign (eg, Langerhans cell histiocytosis, Erdheim-Chester disease). FDCS is distinct because unlike the others, it is not of hematopoietic origin and its management more closely resembles that of other soft tissue sarcomas.

The epidemiology, clinical manifestations, pathologic features, diagnosis, and management of FDCS will be described here. The diagnosis and management of other low-grade sarcomas, histiocytic sarcomas, and nonmalignant histiocytic neoplasms are addressed separately. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma" and "Histiocytic sarcoma" and "Clinical manifestations, pathologic features, and diagnosis of Langerhans cell histiocytosis" and "Erdheim-Chester disease".)

EPIDEMIOLOGY — Follicular dendritic cell sarcoma (FDCS) is a rare disorder, but the precise incidence is unknown. Only a few hundred cases have been reported in the medical literature. FDCS constitutes <0.4 percent of soft tissue sarcomas [3]. In a pooled analysis that included 462 cases of FDCS, the median age was in the fifth decade, although cases have been reported in children [4]. Males and females are affected equally, but the inflammatory variant is more common in females [5]. (See 'Pathologic features' below.)

There are no known associations with inherited susceptibility, and most reported cases appear to be sporadic.

PATHOGENESIS — Histiocytic and dendritic cells play a critical role in the immune system, contributing to phagocytosis, and antigen processing and presentation to B and T cells. The broad term histiocytic neoplasm refers to tumors associated with monocyte/macrophage and dendritic cell phenotypes. They include entities that are malignant (histiocytic sarcomas, interdigitating dendritic cell sarcoma, follicular dendritic cell sarcoma [FDCS], Langerhans cell sarcoma, indeterminate cell sarcomas) and benign (eg, Langerhans cell histiocytosis, Erdheim-Chester disease) [2,6].

Histiocytic and dendritic cell neoplasms are generally grouped together based on the functional properties of their normal counterpart (eg, antigen processing and presentation) rather than the cell of origin [6]. Most histiocytic and dendritic cell neoplasms arise from a CD34+ hematopoietic progenitor. In contrast, follicular dendritic cells, which are found in primary and secondary lymphoid follicles [7], are mesenchymal in origin [8-10]; FDCS have an immunophenotype that is distinct from other malignant histiocytic tumors. In addition, accumulating evidence suggests that FDCS are responsive to systemic therapy regimens that are used for soft tissue sarcoma. As a result, FDCS is now recognized to be a low-grade sarcoma of mesenchymal dendritic cell origin.

The process by which follicular dendritic cells develop neoplastic potential is unclear. Comprehensive genetic analysis has not revealed a universal driver mutation or translocation. Rather, FDCS is associated with widespread chromosomal instability, along with dysregulation of cell cycle progression, nuclear factor kappa beta (NF-kB) activation, mitogen-activated protein kinase (MAPK) activation, and immune evasion [11-13].

The BRAF V600E mutation, which is characteristic of many histiocytic disorders, has been identified in approximately 20 percent of FDCS; there may be a higher incidence of BRAF V600E mutations in the inflammatory variant [11]. The inflammatory variant is consistently associated with Epstein-Barr virus [14], but there is no known association with human herpesvirus 8 (HHV-8) [15]. (See 'Pathologic features' below.)

Mutations in PTEN, TP53, and components of the NF-kB pathway have been described [12,16]. Copy number gains in 9p24 and expression of programmed death ligand-1 and -2 (PD-L1 and PD-L2) suggest that escape from immune surveillance might contribute to tumor pathogenesis [12]. Some cases harbor BRCA2 mutations, show MDM2 amplification, or express somatostatin receptor 2A [17,18]. Another study reported that 67 percent of FDCS tumors overexpressed EZH2, and that 80 percent strongly expressed phospho-extracellular signal-related protein kinases 1 and 2 (p-ERK1/2) [19].

Notably, despite being classified as a sarcoma, nearly all cases of FDCS overexpress epidermal growth factor receptor (EGFR), a marker that is typical of epithelial cancers [16,20,21]; signaling through EGFR initiated by ligands in the microenvironment may be critical to the survival and proliferation of FDCS cells [22].

CLINICAL MANIFESTATIONS — Follicular dendritic cell sarcomas (FDCS) generally present as a slowly growing, painless mass. In one large analysis, 31 percent of patients presented with isolated nodal disease, 58 percent had isolated disease outside of a nodal structure, and only 10 percent had both nodal and extranodal involvement [4]. Other studies estimate that approximately one-third of cases present in extranodal sites [23-26]. The most commonly affected lymph node groups are cervical, mediastinal, axillary, and intra-abdominal sites, while the liver, lung, and spleen were the most commonly affected extranodal sites [4,27]. Involvement of the mesentery [28], mediastinum [29], omentum [30], skin [31], neck [32], and tonsils [33] has been described.

Most cases present with localized disease. In one series, only 8 percent of the FDCS cases presented with disseminated disease [27].

Systemic symptoms such as fever and weight loss are generally uncommon but may be found in patients with liver or abdominal involvement; such systemic symptoms are especially prominent in the inflammatory variant of FDCS, which usually presents in the spleen and/or liver and has a slight female predominance [5,34]. (See 'Pathologic features' below.)

Clonal expansion of follicular dendritic cells is also seen in Castleman disease, particularly in the hyaline vascular variant; 10 to 20 percent of cases of FDCS can present before, concurrently, or after a diagnosis of Castleman disease [35-38]. Shared overexpression of epidermal growth factor receptor (EGFR) has been suggested as the common pathophysiologic link [21]. (See "Unicentric Castleman disease" and "HHV-8/KSHV-associated multicentric Castleman disease".)

In some cases, FDCS is associated with paraneoplastic pemphigus or myasthenia gravis [38-42]. (See "Paraneoplastic pemphigus" and "Clinical manifestations of myasthenia gravis".)

PATHOLOGIC FEATURES — Follicular dendritic cell sarcoma (FDCS) usually has the histologic features of a low-grade sarcoma, with spindle-shaped cells containing weakly eosinophilic cytoplasm organized in a storiform pattern or in whorled bundles with fascicles, trabeculae, or diffuse sheets (picture 1A-B) [24]. High-grade features (eg, atypical spindle cells with indistinct cell borders, large vesicular nuclei, and distinct nucleoli) may be present in the inflammatory variant of FDCS. The mitotic index is characteristically low (0 to 10), and the Ki-67 is lower than 25 percent, although higher grade lesions can occur [23]. FDCS often has a robust and sometimes clonal infiltrate of B and T cells, including T lymphocytes that express terminal deoxytransferase (TdT+) [43,44].

Immunohistochemically, the neoplastic cells of FDCS express the dendritic cell markers CD21 (C3b complement receptor) (picture 2), CD23, and CD35 (C3d complement receptor), as well as vimentin, fascin, human leukocyte antigen (HLA)-DR, and desmoplakin [23,45-49]. Clusterin, CXCL13, podoplanin, and gamma-synuclein are expressed in 60 to 90 percent of cases of FDCS and may be helpful in distinguishing it from lymphoid and other cancers [50-53]. FDCS may express S100 and CD68, which can lead to confusion with histiocytic disorders [54]. (See "Clinical manifestations, pathologic features, and diagnosis of Langerhans cell histiocytosis" and "Erdheim-Chester disease".)

In one study, expression of follicular dendric cell-secreted protein (FDCSP) and serglycin had high specificity (99 and 100 percent, respectively) and sensitivity (73 and 68 percent, respectively) for FDCS [53]. In this study, a panel of CXCL13, CD21, CD35, FDCSP, and serglycin could discriminate 21 of 22 FDCS from other mesenchymal tumors.

As noted above, despite being classified as a sarcoma, nearly all cases of FDCS overexpress epidermal growth factor receptor (EGFR), a marker that is typical of epithelial cancers [16,20,21]; a subset express epithelial membrane antigen (EMA) [27,47,55].

The uncommon inflammatory variant of FDCS (also called inflammatory pseudotumor-like FDCS) is an Epstein-Barr virus (EBV)-associated neoplasm that typically arises in the liver or spleen; there is a female predominance [5,34,56-58]. Histologically, it is characterized by extensive coalescent epithelioid granulomas, eosinophilic infiltrates, and atypical spindle cells with indistinct cell borders, large vesicular nuclei, and distinct nucleoli [5]. The inflammatory variant expresses FDCSP, CD21, and CD35 (as do the cells of the more common form) but is distinguished by the prominent inflammatory component and positivity for EBV by in situ hybridization.

DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS — Follicular dendric cell sarcoma (FDCS) may be suspected in asymptomatic patients who have painless lymphadenopathy and/or masses in extranodal sites (eg, tonsils, lung, liver, spleen), especially in the setting of Castleman disease, pemphigus, or myasthenia gravis. FDCS may also be suspected in individuals with unexplained masses of the liver and/or spleen in the setting of fever and weight loss.

The diagnosis of FDCS is based on pathologic evaluation of involved tissue and demonstration of characteristic immunohistologic markers of dendritic cells (eg, CD21, CD23, CD35, follicular dendric cell-secreted protein [FDCSP], and serglycin). (See 'Pathologic features' above.)

Patients with suspected FDCS should undergo evaluation at a center with expertise in evaluating and managing sarcomas before biopsy and attempted resection. The biopsy specimen should provide sufficient tissue to permit histologic review and to reserve additional material for genetic testing. A core biopsy will usually be needed to obtain adequate material. Although fine needle aspiration biopsy may suffice to establish the preoperative diagnosis, the risk of a nondiagnostic or nonrepresentative sample must be considered. Expert pathology review is essential due to the rarity of FDCS and the difficulty distinguishing it from other entities [23].

Histologically, typical FDCS must be distinguished from other low-grade sarcomas, other histiocytic neoplasms, melanomas, thymomas, and other tumors.

●Soft tissue sarcomas – Low-grade sarcomas, including gastrointestinal stromal tumors (GIST) and other soft tissue tumors (eg, spindle cell sarcomas, liposarcomas, leiomyomas, desmoid tumors, schwannomas, and peripheral nerve sheath tumors), may resemble FDCS histologically.

GISTs generally arise within the gastrointestinal tract or associated tissues, 90 percent exhibit expression of CD117 (KIT), and they lack expression of the dendritic cell markers CD21, CD23, and CD35. (See "Clinical presentation, diagnosis, and prognosis of gastrointestinal stromal tumors".)

Other types of soft tissue sarcomas are identified on the basis of morphologic pattern, and immunophenotype may aid in identification of the presumptive tissue of origin. Expression of characteristic immunohistologic markers of FDCS (eg, CD21, CD23, CD35, FDCSP, serglycin) should distinguish this disorder from other soft tissue sarcomas. (See "Clinical presentation, histopathology, diagnostic evaluation, and staging of soft tissue sarcoma".)

Interdigitating dendritic (or reticulum) cell sarcomas have a paracortical distribution in lymph nodes and lack the storiform and whorled patterns of FDCS [59]. Unlike FDCS, interdigitating dendritic cell sarcoma does not express dendritic cell markers (CD21, CD35), is diffusely positive for S100, and lacks desmosomes on electron microscopy.

FDCS occurring within the retroperitoneum and/or the abdominal cavity may closely mimic dedifferentiated liposarcoma, particularly if they are MDM2 positive and/or amplified [18].

●Histiocytic neoplasms – Langerhans cell histiocytosis (LCH) and Erdheim-Chester disease are histiocytic disorders that can involve skin, bone, or other sites.

LCH demonstrates heterogeneous collections of Langerhans cells with associated eosinophils, neutrophils, small lymphocytes, and histiocytes that may form multinucleated giant cells. Langerhans cells can be recognized by morphologic criteria (ie, large, oval, mononuclear cells with few cytoplasmic vacuoles, little or no phagocytic material, and moderately abundant, slightly eosinophilic cytoplasm), and their identity must be confirmed by immunohistochemical staining for CD1a, S100, and CD207, or by the presence of Birbeck granules by electron microscopy. Both FDCS and LCH can have variable expression of CD68 and S100. (See "Clinical manifestations, pathologic features, and diagnosis of Langerhans cell histiocytosis", section on 'Pathologic features'.)

Erdheim-Chester disease is a rare disorder that is manifest as xanthogranulomatous infiltrates. Biopsy reveals histiocytes with non-Langerhans morphologic features, and the cells are negative for CD1a and S-100 protein expression. (See "Erdheim-Chester disease", section on 'Pathology'.)

Histiocytic sarcoma is a rare, malignant non-Langerhans histiocyte neoplasm of unknown cause that most commonly presents with symptoms due to unifocal or multifocal extranodal tumors. Histiocytic sarcoma is much more likely than FDCS to present with disseminated disease (46 versus 8 percent in one series) [27]. On immunohistochemistry, the cells typically express CD68, lysozyme, CD4, and CD163, but they do not express specific T and B cell markers, myeloid cell markers (eg, myeloperoxidase, CD33), CD1a, S100 protein, or epithelial markers [27]. (See "Histiocytic sarcoma".)

●Melanoma – The histopathologic diagnosis of melanoma is based on characteristic cytologic features (ie, large melanocytes with hyperchromatic nuclei, irregular nuclear shape, nuclear polymorphism, abnormal chromatin pattern, and prominent nucleoli) and architectural details (ie, asymmetry, poor circumscription, nests of melanocytes of various sizes and shapes in the lower epidermis and dermis). Both FDCS and melanoma may express S100, but melanomas are also positive for MART-1 and HMB-45. (See "Melanoma: Clinical features and diagnosis", section on 'Diagnosis confirmation'.)

●Thymoma – Thymomas are rare neoplasms that arise in the anterior mediastinum. The histologic presentations of thymomas are protean and may include features that resemble FDCS, including nests or swirls of polygonal, spindle-shaped, and atypical cells. Both FDCS and thymoma can occur in the mediastinum and contain lymphocytes that express terminal deoxytransferase (TdT+). Thymic epithelial cells stain for epithelial markers such as keratin and epithelial membrane antigen (EMA), as well as Leu-7, while thymic lymphocytes stain for leukocyte common antigen (LCA), CD3, CD1, and CD99 (MIC2). Thymomas are distinguished by the presence of Hassall corpuscles, and they do not express follicular dendritic cell markers (eg, CD21, CD23, CD35). (See 'Pathologic features' above and "Pathology of mediastinal tumors", section on 'Thymoma'.)

●Nasopharyngeal carcinoma – FDCS can be confused for nasopharyngeal carcinoma when it occurs in the tonsils, or head and neck region [60]. Nasopharyngeal carcinoma arises from the epithelial lining of the nasopharynx and may include keratinizing, nonkeratinizing, or basaloid squamous cells, but it should be distinguishable from FDCS based on the characteristic immunophenotype of the latter. (See "Epidemiology, etiology, and diagnosis of nasopharyngeal carcinoma", section on 'Histology'.)

The inflammatory variant of FDCS can be confused with the following disorders:

●Hodgkin lymphoma – The inflammatory variant of FDCS may have Reed-Sternberg-like cells, express Epstein-Barr virus, and have a clinical presentation that includes liver and spleen involvement with systemic symptoms [61]. (See "Hodgkin lymphoma: Epidemiology and risk factors".)

●Angioimmunoblastic T cell lymphoma – FDCS may be misdiagnosed as angioimmunoblastic T cell lymphoma, particularly in the presence of a robust, clonal T cell infiltrate [62]. (See "Clinical manifestations, pathologic features, and diagnosis of angioimmunoblastic T cell lymphoma".)

●Inflammatory myofibroblastic tumor – FDCS can also be confused for an inflammatory myofibroblastic tumor, although the latter expresses smooth muscle actin, and often expresses anaplastic lymphoma kinase (ALK) and/or harbors rearrangements of ALK [63,64]. (See "Inflammatory myofibroblastic tumor (plasma cell granuloma) of the lung".)

The characteristic immunohistologic markers of dendritic cells (eg, CD21, CD23, CD35, FDCSP, and serglycin) should help distinguish FDCS from the disorders described above.

PRETREATMENT EVALUATION — The goals of the initial evaluation are to confirm the histologic diagnosis, establish the site(s) and extent of disease, and determine the performance status and overall medical condition of the patient. In addition to a history and physical examination, it is our practice to perform the following pretreatment studies in patients with suspected follicular dendric cell sarcoma (FDCS):

●Laboratory studies include a complete blood count with differential, and serum chemistries, including electrolytes, liver enzymes, and renal function tests.

●Biopsy specimens should undergo standard histologic analysis, immunohistochemical assessment for CD21, CD23, CD35, epidermal growth factor receptor (EGFR), KIT, and programmed death ligand-1 (PD-L1) expression, and molecular testing for the BRAF V600E mutation.

●Imaging studies should include an integrated fluorodeoxyglucose (FDG)-positron emission tomography (PET)/computerized tomography (CT) scan, as FDCS is reported to be FDG avid [34,65-67]. CT scans of the neck, chest, abdomen, and pelvis are acceptable if PET/CT is not feasible.

●Clinicians should maintain a high degree of suspicion for concomitant Castleman disease; biopsy of multiple lesions may be necessary to exclude this diagnosis. (See "Unicentric Castleman disease" and "HHV-8/KSHV-associated multicentric Castleman disease".)

●Patients should be assessed for signs and symptoms suggestive of paraneoplastic pemphigus or myasthenia gravis. (See "Paraneoplastic pemphigus" and "Clinical manifestations of myasthenia gravis".)

MANAGEMENT — Given its rarity, optimal treatment recommendations for follicular dendritic cell sarcoma (FDCS) are not well defined, as robust studies comparing various treatment approaches are not available.

Management of localized FDCS primarily involves surgical resection. A role for adjuvant therapy has not been established for either radiation therapy (RT) or chemotherapy. Although FDCS typically has an indolent course, local recurrence develops in approximately one-half of patients, and an aggressive clinical course is possible, with metastases to the lung, liver, and lymph nodes [47,68-70].

RT, systemic chemotherapy, or treatment with targeted agents may be useful when surgery is not feasible or in the setting of recurrent or refractory disease.

Localized disease

Surgical resection — Surgical resection of localized FDCS is the mainstay of treatment [4,26]. In one single-institution study of 66 patients with FDCS, patients who underwent gross total resection (with or without microscopically positive margins) had better outcomes than those who did not [38]. Complete resection, if feasible, should be the goal of surgery for localized disease, although the optimal width of uninvolved surgical margins has not been formally defined in this disorder. For patients with a primary tumor located in the head and neck region, neck dissection may be beneficial for locoregional control [71].

Because the diagnosis of FDCS is usually unsuspected, unplanned and/or inadequate excisions frequently occur before a proper pathologic diagnosis has been made. Reresection should be attempted, if feasible, although it may entail a larger procedure than the de novo procedure and could affect the functional result. This underscores the importance of referring patients with soft tissue masses of uncertain identity to centers that specialize in treating sarcomas to enable adequate initial biopsy and resection.

Although there are rare reports of spontaneous regression, observation is not recommended unless the patient has comorbidities that preclude treatment [72].

Role of adjuvant therapy — The role of adjuvant therapy is undefined, and there are no prospective trials that have addressed this issue. For most patients who have undergone a gross total resection, we suggest not pursuing any form of postoperative therapy. However, there is no consensus on this point, and other experts disagree.

In several retrospective analyses, adjuvant RT and/or chemotherapy do not appear to improve progression-free or overall survival following complete surgical resection [4,73,74].

On the other hand, potential benefit for consolidative RT has been suggested in at least two separate series:

●One single-institution study reported lower locoregional recurrence rates for patients with FDCS of the head and neck region who underwent surgery plus adjuvant RT (15 percent recurrence rate among 21 patients) compared with those who underwent surgery alone (45 percent recurrence rate among 52 patients) [71]. However, it is difficult to draw conclusions about the potential benefits of adjuvant RT from this retrospective study because of the variety of locations and sizes of tumors as well as the diverse surgical treatments (eg, neck dissection versus resection alone).

●In a second report of 27 patients with FDCS at a variety of sites undergoing a gross total resection, consolidative RT (administered to 14 of 27) was associated with improved local control, which translated into improved progression-free and overall survival [38].

In another small series, 23 patients with localized or locally advanced FDCS who underwent surgery with neoadjuvant chemotherapy and/or adjuvant radiation had a median recurrence-free survival of 2.9 years and a five-year recurrence-free survival rate of 34 percent (p = 0.04) [73]. However, there was no comparison group of patients treated with surgical resection alone, making it difficult to ascertain the contribution of multimodality therapy.

As a result, some experts advocate multimodal therapy for FDCS, including adjuvant RT after total gross resection [33,38,71,75,76]. This issue remains unsettled, and the role of RT should be evaluated on a case-by-case basis in light of the risk of recurrence, acute and long-term morbidity associated with RT, and surgical options at the time of recurrence.

The use of neoadjuvant therapy for initially unresectable but localized tumors is discussed below. (See 'Advanced disease' below.)

Potentially resectable recurrent disease — The predominant pattern of relapse is locoregional recurrence [38]. If feasible, we offer surgical resection for patients with a potentially resectable, solitary site of disease [77]; we have observed long-term disease control with this approach in some patients.

Advanced disease — Advanced disease includes FDCS that is considered unresectable due to its size and/or location, as well as disseminated (ie, metastatic) disease. A subset of patients with locally advanced disease may be converted into surgical candidates following initial (ie, neoadjuvant) chemotherapy [38]. RT can be used for local palliation in patients with advanced disease or as primary treatment of a solitary site of disease for which surgery is not feasible. Patients with oligometastatic (ie, multiple lesions that are limited in number and/or location) FDCS may also be candidates for surgical resection. For other cases, the treatment of choice is palliative systemic chemotherapy. The optimal regimen is not established. Preferred options include gemcitabine plus docetaxel or an anthracycline-based regimen such as that used for other metastatic soft tissue sarcomas.

Surgery — A subset of patients with locally advanced but nonmetastatic disease may be converted into surgical candidates following initial chemotherapy [38]. The optimal selection of candidates for this approach is not established, and treatment must be individualized. (See 'Potentially resectable recurrent disease' above.)

Patients with oligometastatic (ie, multiple lesions that are limited in number and/or location) FDCS may also be candidates for surgical resection [77]. Long-term disease control has been reported using this approach. However, there is no agreed-upon definition of what constitutes oligometastatic disease in FDCS, and the number and location of resectable lesions may vary with the experience and expertise of the surgeon.

Radiation therapy — RT can be used for local palliation in patients with advanced disease or as primary treatment of a solitary site of disease for which surgery is not feasible. We generally use doses of 50 to 55 Gy when feasible, but the optimal dose and field size for primary treatment are not well defined.

Cytotoxic chemotherapy and VEGFR-targeted therapy — For initial management of disseminated disease, we suggest either an anthracycline-based regimen or gemcitabine and docetaxel rather than other chemotherapy regimens. Other regimens used for soft tissue sarcomas (eg, pazopanib as a single agent or in combination with a mechanistic [previously termed mammalian] target of rapamycin [mTOR] inhibitor such as sirolimus) are acceptable alternatives. (See "Overview of the initial treatment of metastatic soft tissue sarcoma".)

The efficacy of systemic treatment with chemotherapy and targeted agents for the management of disseminated FDCS is not well defined. Reports detailing outcomes with chemotherapy consist primarily of small retrospective series and case reports; there are no trials that directly compare various combination chemotherapy regimens in FDCS.

The available evidence suggests that the same regimens as are used in the treatment of advanced soft tissue sarcomas, particularly gemcitabine plus docetaxel, are effective in FDCS [78,79]. In the largest single-center study of 66 patients with FDCS, an overall response rate of 80 percent was reported in 10 patients who were treated with gemcitabine and a taxane (two complete responses); the median response duration was 13.4 months (range 3 to 83 months) [38]. Partial responses were attained in the three patients treated with ifosfamide-based regimens and in one of the two patients treated with an anthracycline-based regimen (CHOP [cyclophosphamide, doxorubicin, vincristine, and prednisone]). However, the modest number of patients, lack of random assignment to treatment, and variable use of surgery and radiation limit the interpretation of these results.

Responses have also been reported following other anthracycline-based regimens (eg, VAC [vincristine, doxorubicin, cyclophosphamide], AIM [doxorubicin, ifosfamide, mesna]) that are more typically used for treatment of metastatic soft tissue sarcoma [73]. There are no data on the efficacy of olaratumab, a monoclonal antibody targeting platelet-derived growth factor receptor alpha (PDGFRA), in FDCS. (See "Overview of the initial treatment of metastatic soft tissue sarcoma".)

There are case reports indicating activity for multitargeted kinase inhibitors, such as pazopanib and sorafenib, that target the vascular endothelial growth factor receptor (VEGFR) [73]. Sirolimus, an inhibitor of mTOR, can also be used along with sorafenib or pazopanib. Despite the paucity of data, the broad approval of pazopanib for advanced nonadipocytic soft tissue sarcomas includes FDCS, and this agent is another option for treatment of patients with chemotherapy-refractory FDCS. (See "Overview of the initial treatment of metastatic soft tissue sarcoma".)

Combination chemotherapy regimens that are commonly used in aggressive lymphomas, such as CHOP, ABVD (doxorubicin, bleomycin, vinblastine, dacarbazine), and ICE (ifosfamide, carboplatin, etoposide) have been utilized with variable success in FDCS [20,45,46,80]. Such regimens may be less efficacious than gemcitabine-based regimens and were more commonly used in the past because of the mistaken belief that FDCS was derived from hematopoietic cells.

Investigational approaches — FDCS is associated with mutations of BRAF, overexpression of epidermal growth factor receptor (EGFR), overexpression of programmed death ligand-1 and -2 (PD-L1 and PD-L2), and possible involvement of the RAS/RAF/AKT/mTOR pathways, but therapeutic agents that target these abnormalities have not been tested in this disorder. (See 'Pathogenesis' above.)

Several agents that have potential in FDCS include the following:

●A durable partial response was reported in a patient with FDCS after treatment with ridaforolimus, an mTOR inhibitor [81].

●Vemurafenib is a potent inhibitor of the kinase domain in mutant BRAF V600E that is used in the treatment of metastatic or unresectable melanoma and of histiocyte disorders that harbor activating point mutations in BRAF [82-86]. Although potentially beneficial, BRAF V600E inhibitors such as vemurafenib and dabrafenib have not yet been tested in FDCS, and their use in this setting should ideally be reserved for patients being treated in the context of a clinical trial. (See "Systemic treatment of metastatic melanoma with BRAF and other molecular alterations", section on 'BRAF V600 mutant disease'.)

●MEK inhibitors such as trametinib and cobimetinib are also active in BRAF V600E-mutated tumors. Neither of these drugs has yet been tested in FDCS, and their use in this setting, either alone or in combination with vemurafenib, should ideally be reserved for patients being treated in the context of a clinical trial. (See "Systemic treatment of metastatic melanoma with BRAF and other molecular alterations".)

●We have anecdotal evidence of durable stable disease using single-agent nivolumab (an immune checkpoint inhibitor targeting the programmed death receptor-1 [PD-1]) in a patient with progressive and refractory FDCS that initially presented with paraneoplastic myasthenia gravis. (See "Initial management of advanced non-small cell lung cancer lacking a driver mutation", section on 'Limitations in biomarkers' and "Initial management of advanced non-small cell lung cancer lacking a driver mutation", section on 'Nivolumab plus ipilimumab, with or without chemotherapy'.)

However, in our view, at this time, immunotherapy using immune checkpoint inhibitors should only be considered in carefully selected patients in the context of a clinical trial.

Treatment of the inflammatory variant of follicular dendritic cell sarcoma — Treatment of the inflammatory pseudotumor-like variant is the same as the management of the conventional form of FDCS. However, surgical resection may not be possible because the inflammatory variant most commonly presents in the liver and spleen; thus, patients are more likely to require systemic therapy. (See 'Advanced disease' above.)

Management of associated diseases — FDCS may be associated with paraneoplastic syndromes and may occur in the setting of Castleman disease [38].

Treatment of paraneoplastic pemphigus and myasthenia gravis — When FDCS occurs in the presence of paraneoplastic pemphigus or myasthenia gravis, treatment should be directed at the underlying FDCS, with concomitant therapy directed towards the paraneoplastic phenomena. Case reports indicate that rituximab can be effective for paraneoplastic pemphigus and myasthenia gravis occurring in the setting of FDCS [41,87,88]. (See "Paraneoplastic pemphigus", section on 'Treatment' and "Overview of the treatment of myasthenia gravis".)

Certain medications, including the anti-PD-1 monoclonal antibodies nivolumab and pembrolizumab, have been reported to unmask or worsen myasthenia gravis (table 1).

Treatment of follicular dendritic cell sarcoma occurring with Castleman disease — When FDCS occurs together with Castleman disease, treatment should be directed towards both entities as appropriate for the extent of FDCS and the nature of Castleman disease. (See "Unicentric Castleman disease" and "HHV-8/KSHV-associated multicentric Castleman disease".)

POSTTREATMENT SURVEILLANCE — There is no uniformly agreed-upon approach to posttreatment surveillance, and practice is variable. We suggest that patients be clinically assessed at least every three to six months, and more frequently as required by symptoms and organ dysfunction. We generally follow patients indefinitely, but only see them annually after five years.

For patients with advanced, incurable disease, the frequency of imaging is dictated by baseline disease status, organ involvement, and requirement for ongoing treatment. Our approach is to image affected organs every three to six months until stability is documented.

PROGNOSIS — Prognosis in follicular dendritic cell sarcoma (FDCS) it is not well defined due to variable approaches to management and because most data in the medical literature come from small retrospective series and anecdotal case reports. Nevertheless, the available evidence suggests that the natural history can be highly variable, with some patients having an aggressive course. There is a tendency toward local recurrence after complete resection and a low rate of systemic metastases [27,38]. The following data are available:

●In the largest single institutional series of 66 patients with FDCS, the median progression-free survival and overall survival (OS) were 21 and 50 months, respectively [38]. However, this report included a heterogeneous patient population (including those with both localized and advanced disease) that was treated with a variety of surgical, chemotherapy, and radiation approaches. In general, better outcomes were seen in those with limited nodal (as compared with extranodal) disease, non-bulky tumors, and extra-abdominal (as compared with abdominal) disease. Those who underwent gross total resection of the tumor and were treated with consolidative RT had the best outcomes. (See 'Role of adjuvant therapy' above.)

●In another single institutional series of 31 patients with FDCS, those with localized and metastatic disease had median OS of 9.8 and 2.7 years, respectively, and five-year OS of 55 and 38 percent, respectively [73].

●In a Surveillance, Epidemiology, and End Results (SEER) database study of 54 patients with FDCS who were followed for a median of 28 months, median OS was not reached in patients with localized disease, while median OS was 48 months for those who presented with disseminated disease [3].

●In a report of 39 cases with pharyngeal involvement followed for a median of 27 months, the local recurrence, metastasis, and mortality rates were 23, 21, and 3 percent, respectively [25].

●In another large series with a median follow-up of 20 months after treatment for FDCS, local recurrences, distant metastases, and concurrent local and distant recurrences developed in 28, 27, and 10 percent of patients, respectively [4]. Median OS for localized disease was 14 years, whereas median OS was not reached in patients with locally advanced or distant disease (although the number of patients with advanced disease was small). In multivariate analysis, tumor size and lymphoplasmacytic infiltration were associated with worse prognosis.

●Other reports describe tumor size (eg, ≥5 or 6 cm), intra-abdominal location, coagulative necrosis, ≥5 mitoses per 10 high-power fields (HPF), and nuclear atypia as poor prognostic features [47,68]. As an example, one analysis of 60 informative cases of extranodal FDCS from a literature review found that size ≥5 cm and high-grade histology (which included mitotic count ≥5 per 10 HPF or Ki-67 labeling index ≥10 percent, the presence of coagulative necrosis, and cellular/nuclear atypia) were significantly associated with tumor recurrence [68]. The authors used these features to develop a model for recurrence risk assessment:

•Low risk – Tumor size <5 cm, low- or high-grade histology; local recurrence 16 percent.

•Intermediate risk – Tumor size ≥5 cm, low-grade histology; local recurrence 46 percent.

•High risk – Tumor size ≥5 cm, high-grade histology; local recurrence 73 percent.

Whether and how this information could be used to tailor the posttreatment surveillance strategy or select patients for adjuvant therapy has not been addressed.

CLINICAL TRIALS — Often there is no better therapy to offer a patient than enrollment in a well-designed, scientifically valid, peer-reviewed clinical trial. Additional information and instructions for referring a patient to an appropriate research center can be obtained from the United States National Institutes of Health (NIH) or the National Cancer Institute (NCI). Areas of active interest for metastatic soft tissue sarcoma include inhibitors of epidermal growth factor (EGFR), BRAF, and MEK, and immune checkpoint inhibitors targeting programmed death receptor-1 (PD-1) and its ligand (PD-L1). (See "Principles of cancer immunotherapy", section on 'Checkpoint inhibitor immunotherapy'.)

SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Soft tissue sarcoma".)

SUMMARY AND RECOMMENDATIONS

●Clinical manifestations – Follicular dendritic cell sarcoma (FDCS) is a low-grade sarcoma of mesenchymal dendritic cell origin that most commonly presents as a slowly growing, painless mass, most commonly involving the head and neck or abdominal lymph nodes. Occasional cases manifest an inflammatory pseudotumor-like variant involving the liver and spleen that may be associated with systemic symptoms such as fever and weight loss. (See 'Clinical manifestations' above.)

●Diagnosis – Diagnosis of FDCS is based on characteristic histology (eg, spindle-shaped cells in a whorled pattern) and demonstration of immunohistologic markers of dendritic cells (eg, CD21, CD23, CD35, follicular dendric cell-secreted protein [FDCSP], serglycin). Confirmed cases should be assessed for the presence of BRAF mutations and expression of epidermal growth factor receptor (EGFR) and programmed death ligand-1 (PD-L1). Accurate diagnosis requires expert pathology review to exclude other low-grade sarcomas, other histiocytic neoplasms, lymphoma, and other cancers. (See 'Pathologic features' above and 'Pathogenesis' above and 'Diagnosis and differential diagnosis' above.)

●Pretreatment evaluation – In addition to diagnostic biopsy, the pretreatment evaluation for suspected FDCS includes the following (see 'Pretreatment evaluation' above):

•Laboratory studies include a complete blood count with differential, and serum chemistries, including electrolytes, liver enzymes, and renal function tests.

•Imaging studies should include an integrated fluorodeoxyglucose (FDG)-positron emission tomography (PET)/computerized tomography (CT) scan. CT scans of the neck, chest, abdomen, and pelvis are acceptable if PET/CT is not feasible.

•Clinicians should maintain a high degree of suspicion for concomitant Castleman disease; biopsy of multiple lesions may be necessary to exclude this diagnosis. (See "Unicentric Castleman disease" and "HHV-8/KSHV-associated multicentric Castleman disease".)

•Patients should be assessed for signs and symptoms suggestive of paraneoplastic pemphigus or myasthenia gravis. (See "Paraneoplastic pemphigus" and "Clinical manifestations of myasthenia gravis".)

●Localized disease – For management of localized FDCS, we recommend surgical resection rather than initial observation (Grade 1C). (See 'Surgical resection' above.)

•The role of adjuvant chemotherapy or radiation therapy (RT) following surgical resection is not defined, and we suggest not pursuing this approach for most patients after a gross total resection (Grade 2C). However, there is no consensus on this point, and other experts advocate consolidative RT following resection. (See 'Role of adjuvant therapy' above.)

●Potentially resectable recurrent disease – If feasible, we offer surgical resection for an apparently isolated, potentially resectable local recurrence. (See 'Potentially resectable recurrent disease' above.)

●Advanced disease – Advanced disease includes FDCS that is considered unresectable due to its size and/or location, as well as disseminated (ie, metastatic) disease. A subset of patients with locally advanced disease may be converted into surgical candidates following initial (ie, neoadjuvant) chemotherapy. RT can be used for local palliation in patients with advanced disease or as primary treatment of a solitary site of disease for which surgery is not feasible. Patients with oligometastatic (ie, multiple lesions that are limited in number and/or location) FDCS may also be candidates for surgical resection. For other cases, the treatment of choice is palliative systemic chemotherapy. (See 'Advanced disease' above.)

•Initial therapy – For initial management of disseminated disease, we suggest treatment with gemcitabine and docetaxel or an anthracycline-based chemotherapy regimen, as is commonly used for treatment of advanced soft tissue sarcomas, over other regimens (Grade 2C). (See 'Cytotoxic chemotherapy and VEGFR-targeted therapy' above.)

•Subsequent therapy – Other regimens used for advanced sarcoma, such as pazopanib as a single agent or in combination with a mechanistic (previously termed mammalian) target of rapamycin (mTOR) inhibitor, are appropriate for chemotherapy-refractory or intolerant cases.

●Investigational approaches – The following therapies may be options for selected patients with more widespread recurrences who do not respond to systemic chemotherapy, ideally within the context of a clinical trial (see 'Investigational approaches' above):

•An EGFR inhibitor may be selected if EGFR is expressed by the tumor.

•A BRAF inhibitor, with or without a MEK inhibitor, may be selected if a BRAF mutation is present.

•The role of immunotherapy is not yet defined. In our view, immunotherapy using immune checkpoint inhibitors targeting the programmed death receptor-1 (PD-1) should only be considered in carefully selected patients in the context of a clinical trial.