A 39 YEAR‐OLD MAN WITH A CEREBELLAR MASS AND PANCYTOPENIA

CLINICAL HISTORY

A 39 year‐old‐man with no significant past medical history presented with recurrent syncopal episodes, some of which were accompanied by nausea and vomiting. He also complained of recurrent headaches, gait problems and bilateral hand tremor. The physical exam confirmed the bilateral and symmetric intentional tremor and gait difficulties, without any motor or sensory deficits, including his cranial nerve examination. A brain MRI revealed a non‐enhancing 2.4 × 1.7 cm mass within the right cerebellum; it was hypointense on T1 and hyperintense on T2 weighted sequences (Fig. 1a and 1b). Subtle leptomeningeal enhancement in the cortical sulci of both frontal lobes was suspicious for subarachnoid spread. Numerous T2 hyperintense lesions were also seen in the cervical and thoracic spine, the largest measuring 8 by 20 mm; these were similarly worrisome for CSF dissemination (Fig. 1c and 1d). Providing further support for this notion, a cerebrospinal fluid specimen revealed scattered atypical small to medium sized cells with hyperchromatic nuclei, highly suspicious for malignant neoplasm. The patient subsequently underwent a subtotal resection of the cerebellar mass. Intra‐operatively, the surgeon noted expanded cerebellar cortex with associated leptomeningeal coating or opacification suspicious for tumor involvement. The patient subsequently underwent craniospinal irradiation and temozolomide chemotherapy. Two months later, profound pancytopenia was noted on a routine cell blood count (white blood count = 0.4 K/mm³, hemoglobin = 9.0 g/dl and platelets = 42 K/mm³). A CT scan showed sclerotic foci within the right sacrum, left iliac and left pubic bone. A bone marrow biopsy was done.

Figure 1.

MICROSCOPIC PATHOLOGY

Histologically, there was a moderately to highly cellular neoplasm that extensively invaded the cerebellar cortex (Fig. 2) and, to a lesser extent, the underlying white matter, often entrapping both internal granular layer and Purkinje cell neurons, the latter involved in occasional vague perineuronal satellitosis. The tumor cells showed mild nuclear pleomorphism, with the majority of them containing round to oval nuclei with delicate chromatin and occasional clear perinuclear halos (Fig. 3a). Rare mitotic figures were seen, but there was no microvascular proliferation or necrosis. The tumor was also rich in basophilic mucin (Fig. 3b), confirmed by an Alcian blue stain, and had a microcystic growth pattern. No definitive rosettes, perivascular pseudorosettes, Rosenthal fibers, eosinophilic granular bodies, or regions of lipidization were seen. A bone marrow biopsy also revealed aggregates of malignant cells embedded within a fibrillary to collagenous matrix with tumor cells resembling those encountered previously in the cerebellum (Fig. 3c and 3d).

Figure 2.

Figure 3.

By immunohistochemistry, the tumor cells were strongly positive for S‐100 protein (Fig. 4a), and minor subsets were positive for glial fibrillary acidic protein (GFAP; Fig. 4b) and synaptophysin. A fraction of tumor cells also displayed cytoplasmic WT1 positivity (Fig. 4c). Both GFAP and WT1 stains highlighting thin perinuclear rims and tadpole‐like processes in the tumor cells. The Ki‐67 labeling index was low to moderate, focally reaching up to 6.4%. The tumor cells were negative for neurofilament protein and Neu‐N, although the latter highlighted entrapped internal granular cell neurons. FISH was performed on paraffin‐embedded tissue with probes localizing to 1p32, 1q42, 19p13 and 19q13 as previously described (2). Deletions of both 1p and 19q were identified (Fig. 4d and 4e). There was also evidence of 1q gain and monosomy 19 in a subset of tumor cells. Immunohistochemistry performed in the bone marrow biopsy showed that the tumor cells were also positive for S100 (Fig. 4f), synaptophysin, WT‐1 (Fig. 4g) and CD56, but negative for GFAP, CAM5.2, HMB‐45 pancytokeratin, and myeloperoxidase.

Figure 4.

DIAGNOSIS

Cerebellar oligodendroglioma with bone marrow metastasis

Prior to the dissemination to the bone marrow, the tumor was diagnosed as Oligodendroglioma WHO grade II, with 1p/19q co‐deletion, consistent with the genetically favorable variant of oligodendroglioma. However, the increased proliferative index was considered worrisome for the possibility of early anaplastic transformation and a comment suggesting close clinical follow‐up was made. FISH studies were also attempted in the bone marrow core biopsy, but were non‐informative, likely due to poor DNA preservation after specimen decalcification.

DISCUSSION

Oligodendrogliomas are diffusely infiltrating gliomas of unknown histogenesis with uniformly rounded nuclei that most closely resemble non‐neoplastic oligodendrocytes (1). They virtually always arise in the cerebral hemispheres, with most involving both neocortex and underlying white matter. In contrast, location in the cerebellum, brainstem, or spinal cord is so exceptional that the diagnosis often engenders appropriate skepticism. To date, less than 30 cases of primary infratentorial oligodendrogliomas have been published, with the vast majority reported prior to the genetic era and the diagnosis often based on little more than the presence of round nuclei with clear haloes. In general, these cases have been inadequately documented by current standards and would most likely be diagnosed as other entities today, including pilocytic astrocytoma, clear cell ependymoma, extraventricular neurocytoma, cerebellar liponeurocytoma, and rosette forming glioneuronal tumor. For instance, Westergaard et al reported 2 posterior fossa examples out of 96 oligodendrogliomas, but there was essentially no proof provided for this diagnosis (6). A more recent Korean series by Lee et al reported a remarkable 6 of 149 oligodendrogliomas presenting in the posterior fossa (3). Unfortunately, this study was focused on radiology issues and provided minimal pathology data with no genetic confirmation. We are aware of only one case report of a cerebellar oligodendroglioma examining 1p and 19q status by fluorescence in situ hybridization (FISH) and it was negative for deletions (5). A careful examination suggests that rosette‐forming glioneuronal tumor might have been another consideration based on the perivascular neuropil‐like material shown in Fig. 1F (4). Lastly, based on personal experience, one of the authors (AP) has received many consultation cases where the diagnosis of cerebellar oligodendroglioma was initially considered in the differential. With the exception of the current case, none showed a 1p/19q codeletion pattern by FISH and they were all eventually diagnosed as other entities. Herein, we report a highly unusual example of genetically confirmed primary oligodendroglioma of the cerebellum with associated CSF, bone marrow, and lung metastases. To our knowledge, this represents the first well supported case of cerebellar oligodendroglioma in the literature. Given its rarity however, it remains unclear whether the unusually aggressive clinical behavior encountered in our patient is typical of oligodendrogliomas in this location.

ABSTRACT

A 39 year‐old man presented with gait difficulties, headaches, bilateral tremor, and pancytopenia. An MRI revealed a T2 hyperintense 1.7 cm mass within the right cerebellum with associated leptomeningeal enhancement and additional lesions in the cervical and thoracic spine. A biopsy showed a moderately to highly cellular neoplasm composed of tumor cells with mild nuclear pleomorphism, round to oval nuclei, delicate chromatin and occasional clear perinuclear halos in a rich basophilic mucinous background. Rare mitotic figures were seen, but there was no microvascular proliferation or necrosis. A bone marrow biopsy also revealed aggregates of malignant cells resembling those encountered previously in the cerebellum. By immunohistochemistry the tumor was positive for S‐100 protein, GFAP and WT1. FISH also detected deletions of both 1p and 19q chromosomes. These findings confirmed a cerebellar oligodendroglioma with bone marrow metastasis. To date, less than 30 cases of primary infratentorial oligodendrogliomas have been published and none had genetic confirmation. To our knowledge, this represents the first well‐supported case of cerebellar oligodendroglioma in the literature.