Head/Neck/Endocrine Pathology

Follicular Dendritic Cell Sarcoma, Tonsil

Virginia A. LiVolsi
University of Pennsylvania Medical Center
Philadelphia, PA


Case History:
This 48 year old man, former smoker, presented with difficulty swallowing. On exam, a 1.5-2 cm mass was noted in the tonsil area.

Biopsy with frozen section was done and a diagnosis of "squamous cell carcinoma" was rendered. He received radiation and chemotherapy but was not feeling better and referred himself to our ENT oncology service.

Slide 1
Click to view with ImageScope
Click to view with a Web-Based Viewer



Figure 1
Figure 2
Figure 3

Figure 4
Figure 5
Figure 6


Diagnosis:
Follicular Dendritic Cell Sarcoma, Tonsil

The lesion in both biopsies is similar and is an obviously malignant tumor based on infiltrative growth and cytologic atypia and pleomorphism. The differential diagnosis includes: carcinoma, either poorly differentiated nonkeratinizing squamous carcinoma or a high grade salivary gland carcinoma; melanoma; sarcoma (especially rhabdomyosarcoma, a hematolymphoid malignancy or other.

Before giving what I believe is the correct diagnosis, I will discuss the major categories of possibilities and how the workup of the tissue helped to rule each in or out.

Carcinoma:
Could this be an undifferentiated squamous carcinoma of the tonsil/base of tongue region. That was the initial diagnosis and the reason for the therapy the patient was given. However, there are several morphological clues against that diagnosis: first there is no evidence in the overlying squamous mucosa of any dysplasia or carcinoma-in situ; secondly although there are cell groupings focally, the tumor cells appear disassociated and do not form syncytia or sheets as a carcinoma would. The immunoprofile was strongly against any carcinoma since numerous antibodies against cytokeratins of various molecular weights and keratin cocktails were negative. Although undifferentiated squamous carcinomas can be problematic sometimes, none of the features (histologic or immunohistologic) support a diagnosis of carcinoma (squamous or salivary gland types).

Melanoma
Notoriously difficult lesions to diagnose in the head and neck mucosa, melanoma is often nonpigmented in this location (whether that be in the oral cavity, sinonasal cavity or other sites); if pigment is present it is often focal. In addition, it is uncommon to find a "precursor" intramucosal lesion (although some cases do show "melanosis"). Having said that, based on morphology alone, melanoma is a diagnosis that should be considered. The cells are pleomorphic and some giant forms with bizarre nuclei are present. In addition, many cells have prominent eosinophilic nucleoli (a useful feature of melanoma); however, immunostains were not supportive of this diagnosis and there was no other histologic clue to confirm this tumor as a melanoma.

Rhabdomyosarcoma.
Although this tumor is seen in the head and neck in the pediatric population, it is rare in adults. However, it does occur and our group has recently published our experience with such cases. When one thinks of this diagnosis, appropriate immunostains for muscle markers (e.g. desmin, myogenin) can be done; confirmatory molecular studies such as FISH for the translocation seen in these tumors can also be helpful.

Hematolymphoid Malignancy
Because of the marked atypia of the tumor cells and their dissociative pattern of growth, the possibility of a high grade diffuse lymphoma or even Hodgkin disease could be considered. There was no supportive evidence in the workup of these biopsies to conclude that we were dealing with any of these entities.

Other
So we are left with the "Other" category. I suppose in the broadest sense we are dealing with a tumor in the hematolymphoid group. This lesion's immunoprofile was compatible with a follicular dendritic cell sarcoma---CD 21 and CD 35 positive.

Lesions of the dendritic cells are quite rare and have been increasingly recognized in the past two decades. Although the majority apparently present in lymph nodes, they can present in almost any location. The tonsil and base of tongue region is one of those "extranodal" sites. Therefore, surgical pathologists who examine specimens from the head and neck must be aware of this entity.

Discussion:
The International Lymphoma Study Group published a study in 2002 which addressed the classification of "histiocytic" neoplasms and suggested a five-tier category which includes:
  1. histiocytic sarcoma

  2. Langerhans cell histiocytosis

  3. Langerhans cell sarcoma (Letterer Siwe disease)

  4. Follicular dendritic cell sarcoma

  5. Interdigitating dendritic cell sarcoma.
(This paper referred to the last two categories as "tumor/sarcoma". However, the literature on these unusual tumors seems to indicate they are malignancies and so the term "sarcoma" seems appropriate.)

These lesions have different histologic, immunohistochemical and clinicopathologic features and these are briefly summarized below.

a. histiocytic sarcoma occurs in adults, presents in extranodal sites (70%) and is highly malignant (almost 60% die of disease). Immunohistochemically, the tumor cells are positive for CD68 and lysozyme, one third are positive for S100 protein, and there is negative reaction for CD1a, CD21, and CD35.

b. Langerhans cell histiocytosis and sarcoma occur in children and young adults and have a 30 to 50% mortality rate. Immunohistochemically, the tumor cells are positive for CD68, S100 protein and CD1a whereas they are negative for CD21 and CD 35; fewer than half stain for lysozyme.

c. Follicular dendritic cell sarcoma occurs in adults, and although most appear as nodal based tumors, a significant minority occur in extranodal sites. They mark with CD21 and CD35 as well as Podoplanin (antibody D 2-40) and a few have variable staining with CD68, S100 protein and lysozyme; 40% stain for EMA. Mortality rates reported are variable (from 9% to 43 %).

d. Interdigitating dendritic cell sarcoma occurs in older adults with the majority being node based lesions. These are not associated with significant mortality ( as reported by this group although other reports indicate it can be an aggressive neoplasm).

They mark with CD68 and lysozyme (about 50%), all mark with S100 protein and there is no staining with CD21, CD 35 or CD 1a.

Additional markers have been studied including clusterin and podoplanin. In the appropriate setting of trying to distinguish among tumors of histiocytic lineage, these can be useful as they seem to stain preferentially the follicular dendritic cells. However, these are not helpful if the question is not specific as they stain many other cell types.

Back to our case which showed the following immunoprofile: Negative cytokeratins, focal S100 and strongly positive CD 21 and CD35.

The occurrence of this tumor in the head and neck region has been increasingly recognized and it is found most often in the tonsil/base of tongue and nasopharynx (with other sites less frequently affected).

From a morphological point of view the tumor showed several patterns. The dominant pattern was that of discohesive cells with marked muclear abnormalities. Small foci of clusters of cells as well as rare spindling were noted. In all areas of the tumor there were infiltrating lymphocytes and occasional plasma cells; the lesional cells however were obviously much larger than these inflammatory elements. The nuclei were truly interesting. They were large in relationship to the size of the cells, and they often have marked convolutions. Some embryoid nuclei were noted. In many of the tumor cells there were prominent eosinophilic nucleoli. In some areas of the tumor, prominent intranuclear cytoplasmic inclusions were noted. Mitoses were seen but were not numerous and there was no tumor necrosis in the biopsies (in the ultimate resection specimen areas of necrosis and fibrosis were seen, but the lesion had received both radiation and chemotherapy and it is difficult to know whether to attribute the changes to the tumor itself or to the treatment). Another finding in our case was the presence of multinucleated giant cells some of which had the appearance of "floret" cells; such cells described as "osteoclast like" have been noted before in these tumors and the first case I could find of this was a lesion of the tonsil.

The normal counterpart of the tumor cell is the follicular dendritic cell which resides in germinal centers in lymph nodes and lymphoid tissue and skin; these cells are believed to be antigen producing cells to B lymphocytes. The immunoprofile of the normal follicular dendritic cell is similar to that of its tumor—CD21 and CD 35 positive. By ultrastructural studies the cells both normal and neoplastic show marked nuclear convolutions as reflected in our case.

As cases of this tumor are studied some interesting features have been noted. Thus some tumors have been associated with Castleman disease and a few authors have postulated that this association may be causal, ie, in some patients with Castleman disease, there is a hyperplasia followed by neoplasia of the follicular dendritic cells. Some patients have presented with paraneoplastic pemphigus—thus dermatologists and dermatopathologists need to be aware of these tumors. It is important to remember that follicular dendritic cells normally reside in the skin. There are also apparent differences in etiologies and/or associations depending upon the clinical presentation. Hence follicular dendritic cell sarcomas presenting in the abdomen (especially liver and spleen) have a more spindle cell appearance and a rich lymphoplasmocytic infiltrate (an inflammatory pseudotumor pattern); many of these patients have EBV positive tumors. Follicular dendritic cell sarcomas in other sites are EBV negative. Finally, at least one case has been reported of a peripheral T cell lymphoma in which there was a massive proliferation of follicular dendritic cells that confused the diagnostic picture.

In terms of therapy, since the number of cases is small, it is difficult to know what is the best treatment. However, what is known is that the prime mode of treatment should be surgical excision (or debulking if the lesion is very large), followed by either radiation and/or chemotherapy. Apparently approaching this lesion therapeutically as a carcinoma (as was our patient initially) and offering nonsurgical treatment or as a lymphoma (and just giving chemotherapy) is considered inappropriate by oncologists who have had the opportunity to care for affected patients and see the responses to different modes of therapy.

In conclusion, this case illustrates the problem with the recognition of a rare tumor in an unusual location wherein we as surgical pathologists are used to seeing squamous cell carcinoma so commonly. Herein an experienced surgical pathologist missed the diagnosis. Two quotes are pertinent:

A. "Recognition of extranodal FDC sarcoma requires a high index of suspicion…" (two of their three cases were initially misdiagnosed). (Biddle et al 2002).

B. "Although fairly well characterized histologically, with a distinct immunophenotype, it remains underrecognized, with as many as one third of cases initially misdiagnosed." (Youens and Waugh 2008).

References:
  1. Youens KE and Waugh MS. Extranodal follicular dendritic cell sarcoma. Arch Pathol Lab Med 132: 1683, 2008.

  2. Lee IJ, Kim SC< Kim HS et al Paraneoplastic pemphigus associated with follicular dendritic cell sarcoma arising from Castleman's tumor. J Am Acad Dermatol 40: 294, 1999.

  3. Kairouz S, Hashash J, Kabbara W et al. Dendritic cell neoplasms: an overview Am J Hematol 82: 924, 2007.

  4. Selves J, Meggetto F, Brousset P et al Inflammatory pseudotumor of the liver: evidence for follicular dendritic reticulum cell proliferation associated with clonal Epstein-Barr virus. Am J Surg Pathol 20: 747, 1996.

  5. Chan JK, Tsang WY, Ng CS. Follicular dendritic cell tumor and vascular neoplasm complicating hyaline vascular Castleman's disease. Am J Surg Pathol 18: 517, 1994.

  6. Chan AC, Chan KW, Chan JK et al Development of follicular dendritic cell sarcoma in hyaline-vascular Castleman's disease of the nasopharynx: tracing its evolution by sequential biopsies. Histopathology38: 510, 2001.

  7. Grogg KL, Lae ME, Kurtin PJ Macon WR. Clusterin expression distinguishes follicular dendritic cell tumors from other dendritic cell neoplasms. Am J Surg Pathol 28: 988, 2004.

  8. Yu H, Gibson JA, Pinkus GS, Hornick JL Podoplanin (D2-40) is a novel marker for follicular dendritic cell tumors. Am J Clin Pathol 128; 776, 2007.

  9. Dominquez-Malagon, H, Caneo-Valdez AM, Mosqueda-Taylor A et al. Follicular dendritic cell sarcoma of the pharyngeal region: histologic, cytologic immunohistochemical and ultrastructural study of three cases. Ann Diagn Pathol 8: 325, 2004.

  10. Pileri SA, Grogan TM, Harris NL et al. Tumours of histiocytes and accessory dendritic cells: an immunohistochemical approach to classification from the International Lymphoma Study group based on 61 cases. Histopathology 41: 1, 2002.

  11. Valdeeswar P, George SM, Kane SV et al. Extranodal follicular dendritic cell sarcoma of the tonsil—case report of an epithelioid cell variant with osteoclastic giant cells. Path Res Pract 2008 Epub

  12. Starkey CR, Corn AJ, Porensky RS et al. Peripheral T cell lymphoma with extensive dendritic cell tumor: a case report with pathologic immunophenotypic and molecular findings. Am J Clin Pathol 126: 230, 2006.

  13. Fonseca R, Yamakawa M, Nakamura S et al Follicular dendritic cell sarcoma and interdigitating reticulum cell sarcoma: A review Am J Hematol 59: 161, 1998.

  14. Soriano AO, Thompson MA, Admirand JH et al Follicular dendritic cell sarcoma: A report of 14 cases and a review of the literature. Am J Hematol 82: 725, 2007.

  15. Biddle DA, Ro JY, Yoon GS et al. Extranodal follicular dendritic cell sarcoma of the head and neck region: three new cases with a review of the literature. Mod Pathol 15: 50, 2002.

  16. Montone KT, Barr F, Shang P, Feldman M, LiVolsi V. Embryonal and alveolar rhabdomyosarcoma of parameningeal sites in adults: A report of 13 cases. Int J Surg Pathol 2008 (EPub)