FINAL DIAGNOSIS: Gliosarcoma with selective extraneural sarcoma metastases.
DISCUSSION:
In 1955 Feigin (1) defined gliosarcoma as a primary glioblastoma with secondary proliferating vessels and their sarcomatous transformation. Gliosarcomas are estimated to constitute 2% of all glioblastomas, and have a similar location, sex and age distribution and median survivals (2). Gliosarcomas consist of a variable number of GFAP positive neoplastic cells mixed with bundles of spindle cells, which are positive for vimentin, but negative for GFAP on immunostaining. The histiogenesis of the sarcomatous component is somewhat controversial and on the basis of an immunohistochemical study Grant et al. (3) suggest that the spindle cell elements are derived from fibroblasts or primitive pluripotent mesenchymal cells in the adventitia. On the other hand some studies (4) have suggested that advanced tumor progression results in loss of GFAP positivity and acquisition of sarcomatous characteristics, based on the finding of a p53 gene product in both tumor components. Finally, Biernat et al. (5) found in two cases an identical p53 mutation for both cellular components, strongly suggesting that they were of the same glial origin.
There have been reports of an increasing tendency for CNS-tumors to metastasize extracranially over the last decade (6,7,13). Pasquier et al. (6) reported 72 cases of glial tumors with extraneural metastases. In 10 of these craniotomy had not been performed, and in 8 cases the tumor spread along a ventriculo-shunt. After craniotomy, the glioma has direct contact to the leptomenigeal venous system and may give rise to blood-born tumor emboli with access to the venous system outside the CNS. In addition direct spread by invasion of the dural sinuses may cause distant metastases. Cerame et al. (7) attributed the increasing incidence both to prolonged survival as a result of multiple craniotomies, and to the use of special glial stains to improve diagnosis of the extraneural metastastic lesion. Furthermore, these authors claimed that tumor kinetics and cell generation studies could explain the occurence of extraneural metastases either independent of therapeutic interventions or within a short time interval following surgery.
We are aware of 15 cases of gliosarcoma with extraneural metastases. In 4 of these cases the metastases contained only a sarcoma component (1,7,8,9,10,11,13,14). We believe that our case represents the fifth case of pure sarcoma metastases from a primary gliosarcoma. Although we were convinced that a component of the brain tumor was astrocytic in nature and GFAP positive, we stained parallel sections with GFAP, MIB-1 and p53, to establish the diagnosis of a primary brain malignancy and to reject the possibility of an extracranial sarcoma metastasizing to the CNS with secondary reactive gliosis. Moreover the patient underwent extensive investigations between the first symptoms in January 1997 and the time of death in September 1997, and no other tumor was found, except for the pulmonary, oral and palpebral metastases. As sarcomas have a well known tendency to disseminate hematogenously and considering Biernat et al`s finding (5), one might imagine that in spite of the tumor being primarily gliomatous, the sarcomatous component might spread hematogenously and thereby behave like a primary sarcoma. The fact that it does metastasize proves its malignant potential and that it is an integral part of the tumor. On the other hand one might consider a difference in antigenic presentation of the two tumor components or that gliomas are not accepted immunologically by extraneural tissues. This would explain why a metastastic lesion containing only gliomatous tissue (9) has never been reported.
REFERENCES:
Contributed by Lise Mette Gjerdrum, MD and Marie Bojsen-Møller, MD