Central neurocytoma (CN) is a well-differentiated tumor of neural cells occurring within the ventricles. It is composed of monomorphic cells with round, regular nuclei within clear cytoplasm and must be distinguished from other clear cell tumors. Immunohistochemical markers of CN that aid in diagnosis include synaptophysin and neuronal nuclear antigen. The molecular biology of these tumors is becoming increasingly elucidated, particularly with the use of microarray analyses. Several oncogenic pathways have been suggested by these studies. Although progress continues to be made, knowledge of CN has yet to dictate targeted therapies in treating patients with these tumors.
Key points
- •
Central neurocytoma (CN) is a well-differentiated tumor of neural cells occurring within the ventricles. It is composed of monomorphic cells with round, regular nuclei within clear cytoplasm and must be distinguished from other clear cell tumors.
- •
Immunohistochemical markers include synaptophysin and neuronal nuclear antigen.
- •
Chromosomal abnormalities that have been reported in CN include trisomy 7; chromosome 17 deletion; and gains in 2p, 10q, 13q, and 18q.
- •
Microarray analyses have identified many overexpressed genes in CN, including those of the insulin-like growth factor 2 (IGF2) and Wnt pathways; other pathways that may be involved in tumorigenesis include neuregulin 2, N-Myc, and platelet-derived growth factor. Pathways that have not been linked to CN include p53, epidermal growth factor receptor, and BCL-2.
Introduction
Neurocytoma is a rare tumor of mature neural cells within the central nervous system. It was first described arising from the ventricular system, although since then an extraventricular variant has been well documented. Central (intraventricular) neurocytoma is a low-grade tumor (World Health Organization [WHO] grade II) that occurs in young adults. It typically presents with signs of increased intracranial pressure. It is rare, representing less than 0.5% of brain tumors. Overall, the disease is associated with a favorable prognosis with treatment and in many cases surgical resection is curative, although a subset of central neurocytoma (CN) is clinically aggressive and necessitates adjuvant therapy. This article reviews the literature regarding the microscopic characteristics that define this tumor and the insights made into the molecular pathways by which tumorigenesis and progression occur.
Introduction
Neurocytoma is a rare tumor of mature neural cells within the central nervous system. It was first described arising from the ventricular system, although since then an extraventricular variant has been well documented. Central (intraventricular) neurocytoma is a low-grade tumor (World Health Organization [WHO] grade II) that occurs in young adults. It typically presents with signs of increased intracranial pressure. It is rare, representing less than 0.5% of brain tumors. Overall, the disease is associated with a favorable prognosis with treatment and in many cases surgical resection is curative, although a subset of central neurocytoma (CN) is clinically aggressive and necessitates adjuvant therapy. This article reviews the literature regarding the microscopic characteristics that define this tumor and the insights made into the molecular pathways by which tumorigenesis and progression occur.
Histology
Hassoun and colleagues first described CN in 1982, presenting 2 similar cases of tumors with extensive calcifications arising from the third ventricle and sharing histologic features of mature neuronal differentiation. Under light microscopy, clusters of tumor cells with regular, round nuclei and clear cytoplasm occurred in a matrix of fibrillary stroma. Electron microscopy revealed synapses, dense-core and clear vesicles, and parallel microtubules indicating differentiated neuronal tissue without evidence of glial cells.
The salient cytologic features on light microscopy (uniform, regular nuclei containing granular chromatin and accompanied by perinuclear halos) closely resemble those of oligodendroglioma ( Fig. 1 A ). Tumor cells are arranged in sheets, often interspersed between acellular fibrillary regions also known as neuropil islands. Perivascular pseudorosettes and Homer Wright rosettes have been reported but are rare. Calcifications are also rare, but when they occur they tend to be within the tumor as opposed to the peripheral calcifications seen in oligodendroglioma. However, without en bloc neurosurgical resection, it is difficult to appreciate this feature. Necrosis, hemorrhage, and mitoses are seen rarely.
Immunohistochemical staining shows positivity for synaptophysin within fibrillary areas (see Fig. 1 B). Neuron-specific enolase is universally positive in CN, as in many central nervous system tumors. Neuronal nuclear antigen (NeuN) immunoreactivity occurs diffusely in nuclei of tumor cells and has been noted as a more specific marker for CN. Neuronal cell adhesion marker is also reactive. Neuron-associated class III beta-tubulin (Tuj-1) and microtubule-associated protein 2 (MAP2) staining occurs within microtubules. S100 protein and retinal S-antigen have been reported to stain positively with some regularity. Glial fibrillary acidic protein (GFAP) staining occurs intermittently and is typically attributed to reactive astrocytes. However, immunoreactivity for GFAP has also been reported to occur in neoplastic cells in histologic variants. Rare staining for neurofilament protein has been documented as well, in association with ganglion cells. Focal OLIG2 staining may occasionally occur.
The histologic appearance of CN shares similarities with several other tumors of the central nervous system. Before a distinct entity was identified, these tumors were usually deemed to be oligodendroglioma or ependymoma. In relation to CN, oligodendroglioma cells share similar morphology, and they may even stain lightly for synaptophysin ; in addition, focal OLIG2 staining may be seen in CN. However, oligodendroglioma lacks the neuropil islands seen in CN, and oligodendroglioma cells have widespread immunoreactivity to OLIG2 and are negative for NeuN. Further, oligodendroglioma is usually hemispheric and rarely occurs as a ventricular mass. Non–clear cell ependymoma can usually be distinguished from CN based on morphology alone. However, clear cell ependymoma may require additional evaluation. Distinction can be made given that ependymoma cells are usually extensively positive for GFAP and are often positive for epithelial membrane antigen. Ultrastructural examination is occasionally used to diagnose CN, in which mature neural elements, including neurosecretory granules, may be seen (see Fig. 1 C). Other neoplasms that may be considered in the differential diagnosis include pineocytoma, supratentorial primitive neuroectodermal tumor, dysembryoplastic neuroectodermal tumor, and metastatic clear cell carcinoma.
Histologic Variants
Atypical CN is a clinically aggressive variant of neurocytoma that shows varying degrees of mitoses, vascular proliferation, and focal necrosis on histologic examination. MIB-1 labeling index is strongly correlated with vascular proliferation and has been used as a surrogate for atypia in these tumors. Attempts to correlate histologic findings with tumor phenotype have generally focused on this index. Several cutoffs have been used to differentiate typical from atypical CN in the most clinically relevant fashion, including 2%, 3%, and 4%.
Ganglioneurocytoma, a variant of neurocytoma containing ganglion cells as a constitutive component of the tumor, was first described in 1990 and was listed in the WHO classification of brain tumors beginning in 2000. The tumor shows similar clinical behavior to neurocytoma. Intraventricular ganglioneurocytomas are rare, with only 11 cases described in the literature.
Central liponeurocytoma is a neurocytoma variant with lipomatous differentiation that occurs within the ventricles. Cerebellar liponeurocytoma is a more common entity, first recognized by the WHO classification in 2000 and previously identified by more than a half-dozen names. Intraventricular (central) liponeurocytoma has been reported 9 times.
Genetics
In the last 15 years, there has been much insight gained into the molecular mechanisms that make up CN. Several chromosomal and genetic aberrations have been shown by a variety of methods. Several investigators have noted that the overall genetic aberrations in this tumor are fewer than in more aggressive central nervous system tumors, which likely helps explain the benign clinical course that is seen in most cases. These findings are helpful in understanding the biology of CN, but none of them is frequent or specific enough to be used as a diagnostic or prognostic marker.
Chromosomal Abnormalities
Taruscio and colleagues found nonrandom gains in chromosome 7 in 3 of 9 (33%) tumor specimens using fluorescent in situ hybridization. As noted by the investigators, trisomy 7 has been identified in a variety of intracranial tumors, including low-grade and high-grade gliomas and primitive neuroectodermal tumors, as well as in renal, lung, and bladder tumors. Trisomy 7 has also been associated with nonneoplastic diseases as well as normal tissues, including brain tissues.
Cerdá-Nicolás and colleagues reported deletion of chromosome 17 in a case study. Partial deletion of chromosome 17, commonly called 17p, has been associated with other central nervous system tumors, including medulloblastomas and low-grade and high-grade gliomas. In gliomas, 17p deletion is found approximately 30% of the time; more often in anaplastic astrocytoma than either glioblastoma or low-grade astrocytoma. In medulloblastoma, 17p deletion occurs in about 30% of cases and is associated with poor prognosis.
Yin and colleagues used comparative genomic hybridization to evaluate for chromosomal abnormalities in CN. In a study of 10 cases, 4 of 10 (40%) had gains in 2p and 10q, 3 of 10 (30%) had gains in 18q, and 2 of 10 (20%) had gains in 13q. Three cases had combined 2p, 10q, and 18q abnormalities, and the fourth had both 2p and 10q abnormalities, suggesting a nonrandom sequence in tumor development. The investigators noted that gain of 2p suggested MYCN amplification, although specific testing for N-Myc was not performed in the study. The gain seen in 18q was followed by immunohistochemistry targeting Bcl-2, an oncoprotein that inhibits apoptosis and whose gene maps to 18q21; however, staining was not seen in tumor cells.
A controversial finding in CN is loss of heterozygosity (LOH) at 1p/19q. LOH 1p/19q is found in 50% to 70% of oligodendrogliomas and is associated with a favorable prognosis. The relative specificity of LOH 1p/19q for oligodendroglioma has been used to distinguish it from CN when the diagnosis is not clear from histologic features. However, in a series of CNs, Tong and colleagues found 1p deletions in 6 of 9 (67%) and 19q deletions in 5 of 9 (56%), with codeletions occurring in 3 of 9 (33%). The investigators noted that the individual deletions were small (most of the informative loci were retained in all cases) and lacked significant commonalities, which may call into question the significance of these findings. LOH 1p/19q has also been identified in extraventricular neurocytoma, although whether the mechanisms of central and extraventricular neurocytomas can be assumed to be similar remains to be seen. With that in mind, LOH 1p/19q was found in 5 of 21 (24%) and 2 of 12 (17%) cases of extraventricular neurocytoma, as well as in several other reports. In CN, studies screening for large deletions have been unable to find the 1p/19q codeletion, and several investigators have concluded that, at least among intraventricular tumors, this aberration may be unique to oligodendroglioma.
Korshunov and colleagues conducted array-based comparative genomic hybridization on 20 CN specimens, finding concordance with several of the investigations mentioned earlier. Similar to Cerdá-Nicolás and colleagues, they identified frequent deletions of 17p and 17q (80% and 70% of cases at 17q13.3 and 17q11-23, respectively). Like Yin and colleagues, they identified 2p (85% of cases at 2p24.1-22.1), 10q (70% of cases at 10q23.3-26.3), and 18q (50% of cases at 18q21.3-qter) gains. In addition, following Tong and colleagues, they identified small 1p deletions (60% and 70% of cases at 1p34.3 and 1pter-36.3, respectively), but only 1 small 19q deletion (5% of cases at 19q13.3). Other deletions included 6q13-21 (80% of cases), 12q23-qter (60% of cases), and 20pter-12.1 (60% of cases). An additional gain was identified at 11q23-25 (60% of cases). This study also examined mRNA expression levels of 24 genes found in regions with gains and deletions. MYCN , PTEN , and OR5BF1 were overexpressed at least 2-fold, and BIN1 , SNRPN , and HRAS were decreased at least 2-fold.
Microarray Analyses
Two studies have used microarray analysis to determine gene overexpression in CN compared with normal brain tissue: Sim and colleagues in 2006 and Vasiljevic and colleagues in 2013. The studies each identified hundreds of overexpressed genes. From these lists, the investigators identified high-interest genes that may be involved in tumor development, phenotype, and progression ( Table 1 ).
| Sim et al, 2006 | Both Studies | Vasiljevic et al, 2013 |
|---|---|---|
| CHRNA3 FGFR3 FOXG1B FZD1 a GABRB1 GPR1 JAG1 NHLH2 NR2E1 NRG2 PDGF-D PIK3R3 b RXRA SOX2 SOX11 TCF7L2 (TCF4) a | ADCYAP1/PACAP AQP6 BTG1 c GDF8 IGF2 b NHLH1 c RC9 c RGS16 c SCGN c SERPINF1 c SLIT1 TCF4 a TOX3/CAGF9/TNRC9 SOX4 c ZHX2 | CAL2/CALB2 CHRDL2/BNF1 d CRYAB/CTPP2 d FABP7 GPR103 KiSS-1 d NEUROD4 c NTS RL/RELN WNT4 a WNT11 a TTF1 |
c Also overexpressed in pineal tumors.
Sim and colleagues compared gene expression of 3 CNs with ventricular zone (VZ) cells and isolated VZ neural progenitor cells taken from 4 temporal lobectomy specimens. Of 254 genes overexpressed by CN at least 2-fold compared with both VZ and VZ neural progenitor cells, 15 genes code for ligands and growth factors, of which the investigators noted GDF8, platelet-derived growth factor (PDGF) D, neuregulin 2 (NRG-2), and insulin-like growth factor 2 (IGF2) in particular. Thirty-five genes have been identified that code for receptors and intracellular signaling components, including proteins involved in the IGF2 and Wnt signaling pathways. In addition, 40 overexpressed genes coding transcription factors were noted in the CN cohort, which include downstream effectors of the Wnt pathway as well as FOXG1B/BF1; nascent helix-loop-helix 1 and 2 (NHLH1 and NHLH2); and Sry-box proteins 2, 4, and 11 (SOX2, SOX4, and SOX 11).
Vasiljevic and colleagues conducted microarray analysis of 3 primary and 2 recurrent CNs, comparing gene expression with that seen in 1 normal brain specimen and 4 pineal parenchymal tumors. A total of more than 7500 genes were overexpressed or underexpressed by at least a factor of 2. Sorting these genes into functional categories yielded several pronounced pathways, including the following signaling pathways with at least 30 genes involved each: mitogen-activated protein kinase (MAPK), calcium, chemokine, Wnt, neurotrophin, and ErbB.
Next, they identified 464 genes overexpressed at least 5-fold compared with normal brain tissue; of these, 45 were overexpressed in CN and not in pineal tumors and 40 were overexpressed at least 5-fold in both CN and pineal tumors. Of the 45 unique to CN, 7 were also found by Sim and colleagues to be upregulated: IGF2 , TCF7L2 , PACAP , AQP6 , GDF8 , SLIT1 , and ZHX2 . Of the 40 shared with pineal tumors, 7 ( NHLH1 , CAGF9 , SOX4 , RGS16 , SERPINF1 , BTG1 , and SCGN ) were also overexpressed in the Sim and colleagues study, and these pathways may represent shared mechanisms of tumorigenesis between CN and pineal tumors.
In addition, Vasiljevic and colleagues compared gene expression in 3 primary CNs with that of 2 recurrent CNs, investigating possible pathways of tumor progression, and identified 54 genes with expression levels at least twice as high in recurrent as in primary tumors. Of these, KiSS-1 , CRYAB/CTPP2 , and CHRDL2/BNF1 were greatly overexpressed (>10×) in primary tumors compared with normal tissue and are particularly interesting for their stepwise increase in expression level from normal tissue to recurrent CN.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
