Abstract
Editor:
I read with interest the article by Craig, Julian, and Ferracone2 on synovial tumors in dogs. I would like to add to the discussion because I believe the situation regarding synovial sarcoma in veterinary pathology is unclear as a result of the extrapolation of information on human synovial sarcoma, a tumor lacking synovial differentiation, to veterinary pathology.
In human beings, synovial sarcoma occurs in a monophasic or biphasic form, and most tumors of either form share a characteristic chromosomal translocation. The human synovial sarcoma was named as such because earlier pathologists believed the biphasic synovial sarcoma recapitulated the normal synovium. Subsequently, a number of publications have disputed this conclusion, and two lines of evidence contradict the synovial differentiation of synovial sarcoma. The first is the location of these tumors. My understanding, different from the authors, is that most human synovial sarcomas arise within the connective tissues (para-articularly) and not from the synovium.4 They are also reported from locations such as the lung,7 pericardium,1 prostate,5 and elsewhere. Now it is true that mesenchymal tumors do not always arise where you might expect. Although most osteosarcomas arise in the skeletal system, they can arise outside the skeletal system. However, the reverse is true for human synovial sarcoma, with most arising outside of synovial structures. The second line of evidence is provided by immunohistochemical and electron microscopic studies of the biphasic tumors. In these tumors, the cells lining the glandlike spaces look like epithelial cells, and they have typical features of epithelial cells—they are positive for cytokeratin, have desmosomes, and lie on a basement membrane.3,6,8 These are not features of the synovial intima.
On the basis of this evidence, many authors in the medical literature now state that synovial sarcoma is an inappropriate name for this tumor. For example, the most recent fascicle by the Armed Forces Institute of Pathology (AFIP) on tumors of the bones and joints includes this statement: “The term, synovial sarcoma, has come to be recognized as a misnomer for a neoplasm that shows no clear-cut synovial differentiation and quite rarely occurs in an intra-articular location.”4 It seems that the human synovial sarcoma is a tumor of some unknown mesenchymal or pleuripotential cell that may occasionally arise within the synovium just as it may occasionally arise within other organs.
I believe that the integration of this information on synovial sarcoma from the medical literature has had an unintentional and perhaps unrecognized influence on the veterinary literature. Because of this influence, and because there have been a number of recent veterinary publications on synovial sarcoma, it seems timely to ask the question: in veterinary pathology, what is a synovial sarcoma? Is it the counterpart of the human synovial sarcoma, or is it a tumor with synovial differentiation? The influence of the medical literature is such that articles in the veterinary literature are now steering us toward the view that it is the counterpart of the human tumor. This is awkward if we wish the names of tumors to reflect their differentiation. As an example of its influence, the presence of cytokeratin-positive cells in mesenchymal tumors is now being used to help support a diagnosis of synovial sarcoma. This might be valid if you believe that the neoplasm is similar to the (nonsynovial) synovial sarcoma of humans, but it may not be (is likely not) valid for tumors with supposed synovial differentiation.
However, the issue of what is a synovial sarcoma has another angle to it: is there such a neoplasm in animals as a sarcoma with synovial differentiation? This unbelievably heretical question is not as ludicrous as it might first seem, for in the AFIP fascicle on tumors of the bones and joints4 no such entity is listed, nor could I find such an entity mentioned elsewhere in the medical literature. Of course, the apparent absence of such a tumor in human medicine does not preclude the possibility that it might occur in veterinary medicine, but it is an interesting point to ponder. Considering that the normal synovium is not a single population of cells, what would synovial differentiation look like? I will leave that question open, too, except to comment that some of Pool's9 photomicrographs seem reasonable as possibilities.
It is of interest, therefore, to ask what these synovial sarcomas described by Craig et al.2 are. Are they analogous to the human synovial sarcoma, or do they have synovial differentiation? If these tumors are analogous to the human synovial sarcoma, then I believe they need a more suitable name. I am against the unfortunate nomenclature of the medical literature becoming entrenched in our veterinary literature—if it is not already too late. Finally, I am interested to hear the views of other veterinary pathologists.