The study indicated a surprisingly high incidence of intraosseous benign notochordal cell tumours. The anatomical distribution of the tumours was identical to that of classic chordomas. The results support other evidence that classic chordomas develop from intraosseous benign notochordal cell tumours.

Intraosseous benign notochordal cell tumors are recently recognized conditions that may undergo malignant transformation to classic chordomas. This study attempts to define the morphologic and immunohistochemical characteristics of 34 benign notochordal cell tumors by contrasting them with classic chordomas and the notochordal vestiges in fetal intervertebral discs. Benign notochordal cell tumors were characterized by well-demarcated though unencapsulated sheets of adipocyte-like vacuolated and less vacuolated eosinophilic cells within axial bones. The round nuclei were mildly polymorphic but bland. The tumor cells often contained cytoplasmic eosinophilic hyaline globules and lack any intercellular myxoid matrix or necrosis. The involved bone trabeculae were often sclerotic without evidence of bone destruction. The histologic features were different from those of both notochordal vestiges in fetal intervertebral discs and classic chordomas. There was overlap in immunohistochemical reactivity of benign notochordal cell tumors and chordomas, but notochordal vestiges failed to demonstrate cytokeratin 18 positivity. A more appropriate term for the lesions is "benign notochordal cell tumor" rather than "notochordal rest" or "notochordal hamartoma" as they are not rests and do not fulfill the definition of hamartoma. Benign notochordal cell tumors do not need any surgical procedure and must be adequately recognized to prevent unnecessary operations.

Chordomas are rare malignant bone tumors primarily involving both ends of the axial skeleton that present as destructive bone lesions with a large soft tissue mass. Chordomas were previously believed to arise from notochordal remnants. However, recent studies suggest the possibility that chordomas arise from benign notochordal cell tumors. We present two cases of coccygeal incipient chordoma that strengthen the new hypothesis. The first case was an 83-year-old man who died of prostatic adenocarcinoma. The second case was a 79-year-old man who died of hepatocellular carcinoma. The coccygeal tumors were composed of intraosseous and extraosseous infiltrative lesions. The intraosseous lesions consisted of both benign notochordal cell tumor and incipient chordoma. The extraosseous lesions were consistent with incipient chordoma. In addition, two other small benign notochordal cell tumors were found at a different level in case 1. It is conceivable that pre-existing intraosseous benign notochordal cell tumors transform into incipient chordoma and then extend through the cortex into the surrounding soft tissue. The incidence of incipient chordoma appears much higher than expected because chordomas are rare tumors with an incidence of one case per 1 000 000 persons per year. We suspect that unknown factors transform incipient chordoma into classic chordoma.

No abstract