DOES CHONDROID CHORDOMA EXIST

The existence of chondroid chordoma (CC), initially described in 1973, has remained controversial. Since the antigenic profiles of both chordoma (CD) and cartilaginous (chondroid) lesions have been well characterized, we decided to study chondroid chordoma immunohistochemically. Our hypothesis was that chondroid chordoma should display a hybrid or mixed pattern of staining: chordomatous areas with an epithelial phenotype and cartilaginous areas with a mesenchymal (non-epithelial) phenotype. An analysis of CC (seven cases) was performed and compared with results obtained on notochord, cartilage, classic CD (18 cases), peripheral chondromas (two cases), and peripheral chondrosarcomas (CS, eight cases). Four epithelial markers were employed: MKER and AE-1 (both monoclonal antibodies to cytokeratin); PKER (a polyclonal antibody to cytokeratin); and, EMA (epithelial membrane antigen). In addition, selected cases were tested for the presence of neurofilament (NF) and glial fibrillary acidic protein (GFAP). All 18 CD''s exhibited the expected epithelial immunophenotype.sbd.MKER+, AE-1+, PKER+, and EM+.sbd.a reaction pattern nearly identical to that found in fetal notochord. This reinforced the importance of the growth pattern in assessing the presence of chordomatous elements. All chondromas and CS''s failed to express any of the epithelial markers studied and contained only S-100 immunoreactivity, like cartilage. Chondroid chordoma resembled cartilaginous tumors immunohistochemically; no mixed pattern with even focal epithelial marker reactivity was identified. All CC tested were also NF and GFAP negative. We conclude that CC either does not exist or is extremely rare and that these tumors are cartilaginous in nature. We propose abandoning the term chondroid chordoma and replacing it with "low grade chondrosarcoma". Distinct from CD is still imperative due to the clear difference in natural history; we provide histological definitions and an immunophenotype to aid in this distinction.