While leukemias are the most common type of malignancy to afflict the pediatric population, brain tumors are the most common solid tumors in this age group (Gilbertson, 2004). Medulloblastoma is the most common malignant brain tumor in children constituting nearly 20% of all pediatric brain tumors (Rossi et al., 2008). It is categorized as a primitive neuroectodermal tumor (PNET) of the cerebellum. Within the first few years of diagnosis, mortality approximates 15%; however, cure rates can reach as high as 60% with current therapeutic modalities (Pfister et al., 2009; Wu et al., 2011). Surgical resection, preceded and/or followed by radiation and chemotherapy, is the mainstay of therapy, with five-year survival rates of between 50% to 90%. This wide range is multifactorial, owing in part to age at diagnosis, the presence of metastases at diagnosis, and a histologic variant of medulloblastoma (Gilbertson, 2004; Huse and Holland, 2009). Regardless of long-term survival, treatment-related cognitive, neurologic, and endocrinologic effects remain a debilitating concern and an impetus for the search for further therapeutic modalities.
There is no clear etiology to medulloblastoma. Some studies have found a link between maternal diet and blood/immune disorders during pregnancy (Bunin et al., 2005). Others report an association with viral infections, for example, early John Cunningham (JC) viral infections or human cytomegalovirus (CMV) infections in childhood (Baryawno et al., 2011; Krynska et al., 1999).
Examining data from the Surveillance, Epidemiology, and End-Results (SEER) database from 1973 through 2007, annual incidence for medulloblastoma was reported at six per million children, in other words, approximately 450 new pediatric cases per year. Children age 4 to 9 years old had the highest incidence at 44%, followed by adolescents (10 to 16 years old) at 23%, and only a 12% incidence in infants/toddlers (0 to 3 years old). Pediatric incidence was calculated to be ten-fold higher than adult incidence. Medulloblastoma affected males 1.5 times more than females (Smoll and Drummond, 2012; Taylor et al., 2012).
Thought to originate from the granule cell precursors in the external germinal layer (EGL) of the developing cerebellum, tumor growth starts in the fourth ventricle and can grow to occupy it completely. After that, he tumor spreads to the cerebellar vermis and the brainstem, seeding the craniospinal axis. Medulloblastoma is a highly malignant tumor with a propensity for local invasion and distant metastatic spread through the subarachnoid system, (i.e., within the brain and along the spinal cord, also known as "drop mets") (Raffel, 2004).
Extraneural metastases in pediatric medulloblastomas are an infrequent occurrence (approximately 7%). The most frequent sites of extraneural metastasis in children include bone (78%), lymph nodes (33%), liver (15%), and lungs (11%). The average time to develop after maximal surgical resection is approximately 20 months. Survival in these cases is dismal and in most cases can be less than six months (Rochkind et al., 1991).
Notably, the most common cytogenetic mutation encountered in medulloblastoma is isochromosome 17q, resulting from the loss of the short arm (p) with a resultant gain of genetic material from the long arm (q). Also, deletions in the short arm have also been frequently reported, leading to loss of heterozygosity of 17p, (i.e., 17pLOH) (Biegel et al., 1989; Bigner et al., 1988; Cogen and McDonald, 1996). Interestingly, the tumor suppressor gene, TP53, which is located on chromosome 17p, is rarely mutated in medulloblastoma (Adesina et al., 1994; Batra et al., 1995). Thus, the search for putative tumor suppressor genes on chromosome 17p in the context of medulloblastoma is ongoing.
Given its origin in the cerebellum with the propensity of locally spreading into the fourth ventricle, patients most often present with a combination of cerebellar signs like clumsiness, gait disturbances, and obstructive hydrocephalus, for example, early morning headaches, nausea/vomiting, double vision, or blurry vision. Time from symptom onset to diagnosis is usually short, usually two to three months (Raffel, 2004; Ryzhova et al., 2013).
Patients with medulloblastoma were traditionally stratified into average-risk or high-risk groups based on three primary characteristics: (1) age, (2) extent of residual tumor post-op, and (3) degree of dissemination at time of diagnosis. Patients who were at least three years old at diagnosis, had, at most, 1.5 cm of postoperative residual disease by MRI, and those who had no extraneural metastases were categorized as average-risk. The rest were high-risk (Gilbertson, 2004; Gottardo et al., 2014).
This traditional risk stratification discounted the influence of tumor histology on tumor behavior and impact on patient prognosis. The five histologic subtypes encountered in medulloblastoma include:
In classic medulloblastoma, sheets of small round cells, possessing high nuclear-to-cytoplasmic ratio, are noted. They have a high invasive tendency and possess occasional neuroblastic differentiation. The classic type constitutes approximately 70% of medulloblastomas. Nodules of tumor cells displaying neurocytic differentiation on a collagen-rich matrix characterize the desmoplastic variant; these tumors are less aggressive than the classic variant and account for 15% of medulloblastomas. Large-cell anaplastic medulloblastomas, as the designation suggests, demonstrate features of anaplasia. These features are large tumor cells with abundant cytoplasm, pleomorphic nuclei, and prominent nucleoli. These tumors are typically located in the cerebellar vermis and are highly aggressive, demonstrating high mitotic and apoptotic activity with large areas of necrosis. Consequently, the prognosis is especially poor with short survival times after diagnosis. They constitute approximately 10% of medulloblastomas (Kleihues et al., 2002). The last two variants are rare and make up the remaining 5% of medulloblastomas (Kool et al., 2012).
Recent transcriptional profiling studies derived from microarray data of large numbers of patients with medulloblastoma revealed clusters of patients with similar transcriptomes, proteomes, and cytogenetic signatures involving unique signal transduction pathways. This has led to the currently accepted classification system of medulloblastoma into four primary subgroups: (1) SHH, sonic hedgehog; (2) WNT, wingless; (3) non-SHH/WNT group 3; and (4) non-SHH/WNT group 4 (Gottardo et al., 2014; Taylor et al., 2012). Moreover, this system has not only augmented prognostication but has also facilitated the development of novel therapeutic options (discussed below).
Using a combination of molecular profiling and histology, the World Health Organization (WHO) presently divides medulloblastoma into five subtypes:
Current treatment modalities for medulloblastoma combine surgical resection with chemotherapy and radiation. By traditional risk stratification, cure rates in the average-risk group reached three-quarters of patients. However, post-surgical treatment-related neurologic, cognitive, and endocrinologic sequelae, including intellectual retardation and growth hormone deficiency, remain a source of morbidity in up to 80% of survivors. The high-risk group experienced up to 50% mortality due not only to the presence of extraneural metastases at diagnosis but also due to their young age at diagnosis, which poses significant limitations to their therapeutic options, namely lower doses of radiation and chemotherapeutic agents (Gilbertson, 2004).
Newer subgroup classification systems have facilitated the development of more targeted therapeutics aimed at disrupting signal transduction pathways critical to phenotypic transformation. These are currently under clinical investigation:
The SHH pathway is activated by the binding of Sonic Hedgehog to its receptor Patched 1 (PTCH1) which activates downstream signaling via a key mediator Smoothened (SMO). The most widely studied targeted therapeutic agents today are SMO inhibitors, i.e. cyclopamine, HhAntag, vismodegib, saridegib, and sonidegib (Lee et al., 2012; Romer et al., 2004; Rudin et al., 2009; Von Hoff et al., 2009). In fact, sonidegib is currently in phase 2/3 trials in both adult and pediatric patients.
The key step in WNT signal transduction leading to malignant transformation is a lack of degradation of beta-catenin due to mutations at key amino acid residues that are normally destined for phosphorylation. Hence, new drugs have been developed to target steps in downstream signaling by beta-catenin. These include naturally-occurring protein phosphatase inhibitors, cantharidin, norcantharidin, and ginkgetin (Yang et al., 2003; Ye et al., 2015).
Unfortunately, not much is known about the signaling pathways implicated in non-SHH/WNT subgroup medulloblastoma. As a result, targeted therapeutics have yet to be developed for this type of medulloblastoma. Promising strategies currently under investigation include myc inhibition, cell cycle checkpoint inhibitors, and histone deacetylase (HDAC) inhibitors.