Gangliogliomas belong to the group of glioneuronal tumors, morphologically characterized by a biphasic population of neuronal ganglion cell and glial cell components. According to the WHO 2007 classification, GGs correspond to benign WHO grade I. Some rare GGs presenting the same anaplastic features found in aggressive gliomas are considered WHO grade III. Criteria for defining GG grade II have been not yet established (ref [4]).

GGs are characterized by a mixed glioneuronal phenotype usually composed of a major glial morphotype associated with a neuronal ganglion cell component (Fig. 51.4). The histopathological identification of GG is notoriously challenging because both glial and neuronal populations may exhibit marked phenotypic heterogeneity and distribution. Consequently, GG morphological spectrum can range from a predominantly glial phenotype with exceptional ganglion cells to predominantly neuronal population. Glial elements are usually piloid astrocytic, but the glial cells can also present with an oligodendroglial, clear cell, gemistocytic, or ependymal morphology [4, 6, 37, 49]. Ganglion cells are characterized by their abnormal noncortical localization and distribution as well as by cytomegaly, bizarre shape, vesicular nuclei, and abnormally aggregated Nissl substance. All these criteria are not always associated. Consequently, a broad neuronal spectrum may be observed, leading to potential discrepancies between true neoplastic ganglion cells and dystrophic residual neurons, particularly in anatomic areas containing pyramidal or large neurons as it occurs in temporal or brainstem structures. Binucleated or multinucleated neurons are pathognomonic but unfortunately rare and heterogeneously distributed. Additional histopathological features such as Rosenthal fibers, eosinophilic granular bodies, perivascular lymphoid infiltrates, and abundant capillary networks with many parallel capillaries are also encountered frequently. These nonspecific features may alert and motivate a careful analysis of all the samples to search for a ganglion cell component. Desmoplasia (connective tissue component intermingled with glial lobules) is a common associated feature and should not be interpreted as an aggressive metastatic leptomeningeal extension. Histologically, as compared to pilocytic astrocytoma, GGs are not frequently well defined from the surrounding brain structures. They are associated with areas of marked dysplastic and/or atrophic parenchyma, leading in MRI to pseudo-infiltrating images.

The immunohistochemical profile of GG reflects the mixed glioneuronal nature, and many neuronal routine markers (such as chromogranin A, synaptophysin, NF70, and NeuN) are heterogeneously positive in ganglion cells. It is important to note that all these markers are not specific to neoplastic ganglion cells and are also positive in normal residual neurons. As a consequence, neuronal markers should be interpreted carefully and comparatively, preferentially in areas of tumoral tissue or in desmoplastic areas. Recently, studies suggested that the oncofetal CD34 transiently expressed during early neural development is frequently expressed in GG [ [6, 12]]. These intratumoral and peritumoral highly ramified CD34 immunopositive cells are encountered in 66–74 % of gangliogliomas and are negative or only exceptional in DNET, pilocytic astrocytoma, or the infiltrative areas of low-grade oligodendroglioma or astrocytoma. Although the exact nature of the CD34-positive cells is not known, it has been suggested that they represent a glioneuronal progenitor cell as the majority of CD34-positive cells co-localized with S100 protein and more rarely with neuronal antigens [ [6, 12]]. Thus, CD34 represents a valuable marker for the diagnostic evaluation of GG. MIB-1 index is generally low with a mean value ranging from 1 to 2.7 % [34].

Nonconsistent and recurrent allelic alterations have been identified, but gain of chromosomes 5, 7, 8, and 12 and loss of chromosomes 9, 10, and 22 have been described too [26]. No TP53 mutation, PTEN mutation, KRAS or IDH 1 mutation, or EGFR amplification [15, 50] has been documented. Mutational analysis in the coding region of the tuberous sclerosis 1 and 2 genes (TSC1, TSC2) did not discover any mutation on genes involved in the Reelin pathway [34]. However, in the two past years, the pathogenesis of GG is moving up by the involvement of an aberrant mitogen-activated protein kinase (MAPK) pathway activation due to the identification of BRAF activating mutation [46]. The mechanism of MAPK activation seems to be different than in pilocytic astrocytoma (PA). Many recent studies have demonstrated gene fusions between KIAA1549 and BRAF in the majority of the PA, whereas BRAF activating mutation, particularly V600E, represents the genetic hallmark of GG [15, 34, 39, 45, 47]. These recent data, described in a small number of GG, need further translational studies particularly to establish the prognostic significance and more interestingly whether children with GG may benefit from inhibitors of the MAPK signaling pathway.

The treatment of choice for gangliogliomas is total macroscopic excision whenever possible [3, 23, 27, 33]. Total removal is associated with the highest probability of a long-term disease-free survival.

Unfortunately, infratentorial gangliogliomas are amenable to total excision only when the tumor is confined to the cerebellum. The excision is often incomplete if the tumor extends to the cerebellar peduncle and partial in practically all cases of brainstem involvement. Despite the difficulties that might be encountered during the removal of such lesions in these particular locations, surgery should still be considered as first treatment. It is worth to note that the removal of the enhancing portion of the tumor can result in a long event free survival even in the absence of any further treatment. In patients with residual nonenhancing tumors, no evidence of tumor progression or enhancement has been reported, with a follow-up of at least 2.5 years [3].

Because of the characteristics of the tumor, however, a postoperative neurological deterioration has been reported in about 30 % of the cases with infratentorial GGs [33]. This elevated morbidity reflects the difficulties related to the infiltrative nature of these tumors and the risks of a too aggressive surgery within the brainstem. This morbidity can, in fact, be lowered by limiting the surgery to the enhancing component of the tumor [3].

There are only limited reports about the use of adjuvant therapy in patients with infratentorial GG. Scarce information is available from the experience with supratentorial GG as adjuvant therapy is commonly not utilized in these tumors usually amenable to radical excision, except in case of malignant subtypes. In case or residual tumors, adjuvant therapy has utilized either in the early postoperative period or at the moment of disease progression [29, 33]. The efficacy of adjuvant therapy to control tumor progression is, however, not clear from the literature. Its use may be also weighted by potentially fatal complications such as severe hematological aplasia [3]. Responses to chemotherapy of non-pilocytic glioma have usually been reported less frequently [21]. Frequently, chemotherapy alone fails to control on the long run the progression of the tumor requiring therefore additional radiotherapy. The absent or modest benefit of chemotherapy explains why it is not utilized even in case of residual tumor [3, 28, 48].

Radiotherapy has been used for treatment of some brainstem GGs, but, similarly, its usefulness has yet to be proven [32, 41]. The deleterious side effects on the developing nervous system have been a strong argument against the use of this treatment modality in children [29]. Moreover, some reports suggest that postoperative radiation may favor the malignant degeneration [25, 44]. Due to bias in patient selection for radiotherapy (more aggressive tumors, incomplete resections, early postoperative progressions, etc.), it is difficult to assign the transformation to the effect of radiation therapy. However, secondary tumors have been extensively documented after brain irradiation.