Unusual Epithelial Differentiation in a Sarcoma

Electron Microscopy:
My introduction to this case came in 1989, when a specimen arrived in the EM unit with a limited history that simply stated "recurrent sarcoma of lumbar vertebra". When I screened the thin sections, my immediate reaction was that this must be metastatic squamous carcinoma and that perhaps the original diagnosis might be erroneous. For in virtually every field, the tumor cells tended to be closely associated and were often joined by typical desmosomes (Figs. 4 and 5). Small to moderate numbers of tonofilaments were also present in many tumor cells (Figs. 4 and 5). However, review of the histology of the original tumor and the recurrence in 1988 confirmed the diagnosis of osteogenic sarcoma.

Figure 4. Tumor cells with nuclear irregularities and moderate amounts of endoplasmic reticulum in the cytoplasm. Arrow points to a desmosome into which tonofilaments are inserting and some cells contain small amounts of tonofilaments (arrowheads). The inset shows typical desmosomes (arrows) joining cell processes.	Figure 5. Higher magnification of tonofilament bundles in the cytoplasm of adjacent tumor cells. Typical tonofilament-bearing desmosmes (arrowhead) join some tumor cells.

In rescreening thin sections prepared from the original lesion in 1985 and the 1988 recurrence, desmosmes and tonofilaments were readily identified in both specimens. Furthermore, electron microscopy of the metastasis to the right ileum in 1989 and the recurrent tumor in 1991 again revealed evidence for aberrant epithelial differentiation.