Glioblastoma NOS (haemorrhagic)
Loading Stack -
0 images remaining
A large haemorrhagic occipital mass is present with surrounding oedema.
The patient went on to have surgery.
Paraffin sections show fragments of a densely hypercellular tumour. This is composed predominantly of cells with pleomorphic, eccentrically located round and oval vesicular nuclei and a moderate amount of dense eosinophilic cytoplasm which shows moderately strong immunostaining for GFAP. These cells have features of gliofibrillary oligodendrocytes and are admixed with a quantitatively smaller population of strongly GFAP immunoreactive fibrillary and gemistocytic astrocytes. Frequent mitotic figures are noted in the gliofibrillary oligodendrocytes. In addition there are several foci of vascular endothelial cell hyperplasia as well as scattered areas of necrosis.
Much of the tumour is situated as a layer superficial to the surface of cerebral cortex deep to the pia-arachnoid with extension down Virchow-Robins spaces. Perivascular aggregation of tumour cells is prominent more deeply in the cortex. In several fragments, however, tumour is present in white matter and is seen to merge gradually with adjacent brain parenchyma. Areas of recent haemorrhage are seen and there are collections of haemosiderin-filled macrophages, indicating past haemorrhage. One fragment is composed largely of low grade oligodendroglioma and shows prominent calcification. No immunostaining for cytokeratins (AE1/AE3, CAM 5.2), CD34 or the melanocyte markers, tysosinase and melan A is seen in tumour cells. In some areas, the gliofibrillary astrocytes show strong surface staining for CD138.
This is interpreted as cross-reactivity and not indicating a plasma cell tumour as no kappa or lambda light chain immunoglobulin restriction is demonstrated and there is no staining in these cells for CD20 or CD79a.
The overall features are of glioblastoma with a predominant oligodendrocyte component.
FINAL DIAGNOSIS: Glioblastoma with a promiant oligodendroglial component
Note: Unfortunately this case predates routine IDH and 1p19q codeletion assessment. Although this tumour is entirely consistent with an IDH wild-type glioblastoma without establishing IDH wild-type status (primary glioblastoma IDH wild-type), or IDH mutant but 1p19q not-co-deleted (secondary glioblastoma IDH mutant) the diagnosis can not be made with certainty according to the current (2016) WHO classification of CNS tumours.
This is particularly relevant in this setting as there is a component that appears oligodendroglial, and MGMT is methylated. It is, therefore, possible that if a non-R132H mutation was identified, and 1p19q was co-deleted, then this would represent an anaplastic oligodendroglioma rather than a GBM.
In this case, even the patient's age is not available, further muddying the waters. Lesson learned. Always collect as much data as possible. :(