DE NOVO MALIGNANT GANGLIOGLIOMA (GG) OF THE SPINAL CORD WITH UNUSUALLY PROMINENT DYSPLASTIC VESSELS ("MALIGNANT ANGIO-GANGLIOGLIOMA")
We present a detailed description of a rare de novo malignant spinal GG in an adult patient. Despite broad excision and radiotherapy at diagnosis, the tumor disseminated 12 months later to give rise to multiple spinal and cerebral predominantly periventricular deposits, and lead to a rapidly fatal outcome.
While there is general agreement that spinal GGs in adults are rare tumors, their true incidence is not known. The highest reported estimates of 6% of all spinal tumors of adulthood (9) are considered unacceptably high by some authors (12). Only 8 cases were listed in an exhaustive review of the older literature (4), 2 case reports appear to have been published ever since (11, 13), a large series of 326 GGs seemingly includes less than 10 adult cases (1). Our own search of the literature produced 5 more (3, 5, 6). It has been argued that over-diagnosis of GG in some series may have been due to misleading expression of synaptophysin by normal neurons overrun by an astrocytoma, a phenomenon reportedly prone to occur in the spine quite often (12). While this anatomical particularity may indeed be vexing in some specimens, it is clear that insufficiently representative biopsies are the main obstacle to establish the correct diagnosis. Spinal tumors, not always easily accessible, often yield small biopsies (9). The useful synaptophysin immunostaining should always be backed up by immunopositivities for other markers of neuronal cells. Interestingly, CD34 was recently shown to be expressed by the majority of GG (1), and is thus helpful to make the diagnosis.
Most GG in children as well as in adults are benign (9). Malignant transformation is rare, and when it occurs , it is almost invariably through a protracted evolution of the astrocytic component to a high grade tumor (1, 2). In an extensive review of older and more recent reports, McLendon et al (2) cited 12 cases of de novo GG/malignant astrocytoma (grade III-IV); none of them was spinal. In their large series, Blümcke & Wiestler (1) reported 17 de novo anaplastic GGs. Four of these tumors were spinal, and, to our knowledge, this is the sole report to have documented anaplastic GGs at this location.
In our case, endothelial proliferation was inconspicuous, but the presence of vast areas of ischemic necrosis and of many highly atypical and immature GFAP positive cells, were most compatible with a grade IV astrocytoma.
Angioglioma is a descriptive term long applied to composite lesions of tumoral or reactive glial proliferations associated to vessel proliferations (8). As indicated in a detailed study, this term, devoid of prognostic connotations, should probably best be retained for some excessively vascularized low grade astrocytomas, commonly of the pilocytic type (8). The vessels are likely malformative, and older theories of casual relationships between vascular and glial tumors or common viral etiology of both lesions although stimulating, are probably overambitious (8). The importance of this pattern is that angiogliomas may simulate a vascular tumor or malformation. In our case , this exophytic tumor evoked in the first place a hemangioblastoma both clinically and angiographically. Spinal cord hemangioblastomas , rare benign tumors that commonly arise in a dorsal and median position in the thoracic spine of middle aged males, are densely vascular, and produce a typical angiographic image (10). Rarely, vessels in GGs are malformative, in line with the presumed dysontogenetic origin of these tumors (1). Exceptionally, they may be conspicuous enough to engender the diagnosis of a mixed lesion as exemplified by a remarkable childhood case of a bihemispheric GG (7).
Contributed by Fabrice Chrétien MD, Michel Djindjian MD, PhD, Philippe Caramelle, Frederic Ricolfi MD, PhD, Christo Christov MD, PhD