Medulloblastoma: Systemic Chemotherapy and Future Applications of Chemoradiotherapy

Abstract

Medulloblastoma is the most common malignant brain tumor in children. Prognosis remains guarded in patients who at diagnosis are very young, have craniospinal metastatic disease, or after resection, have residual disease. Treatment incorporates chemotherapy and radiation therapy. Cancer survivors often have life-altering treatment effects. This chapter reviews clinical trials over the years and the efforts to improve survival and minimize sequelae along with challenges in performing clinical trials. Quality assurance of the radiation therapy provided worldwide monitors compliance. Advances in the risk stratification and targeted treatment based on the genomics and biology of medulloblastoma are highlighted in recent clinical trials. Through chemotherapy aligned with specific biomarkers, sophisticated radiation therapy strategies, and continued quality assurance, the future vision of managing medulloblastoma is presented.

1. Introduction

Although medulloblastoma (MB) will affect two patients for every million people per year, the disease remains the most common malignant brain tumor in children comprising 15% of primary brain tumors in children [1]. Forty percent (40%) of patients with MB are identified before the age of 5, with approximately 70% of patients identified before the age of 10. Prognosis remains guarded in patients who are exceptionally young at diagnosis (<36 months), patients with limited surgical resection with residual post-surgical disease, or patients with metastatic disease at presentation including disease in the cerebral spinal fluid axis. Because of the propensity of the disease to involve the cerebral fluid pathways, radiation therapy (RT) is delivered in this disease to the entire cranial spinal neuroaxis. There has been limited success when the RT volume and dose to target have been titrated; therefore, patients remain at risk for measurable sequelae of management despite treatment with curative intent, especially given the propensity of younger patients affected with this disease [2]. With cumulative survival rates at 20 years of 50% using traditional therapies including craniospinal radiation therapy (CSRT), efforts have been made in multiple clinical trial settings to improve tumor control and potentially address the development of sequelae by using more advanced technology RT coupled with systemic therapy. Because RT treats large segments of normal tissue including bone marrow reserves, integrating chemotherapy either before, during, or after RT has been challenging. Effort identifying a consistent and strategic pathway for the application and integration of systemic care in this patient population is the focus of several clinical trials. In this chapter, the chemotherapy management of patients with MB from a historical and protocol perspective is reviewed, as well as strategies for care for the next generation of clinical trials in this disease currently characterized by unique biomarker expression products. The future vision of modern systemic therapy aligned to specific biomarkers identified in subsets of patients affected with this disease will be presented.

2. Past and current management involving systemic therapy for medulloblastoma

From a historical perspective, surgery followed by postoperative RT to the neuroaxis including boost to the primary site of the disease including sites of metastatic disease in the central nervous system and spine had been considered a standard of care for patients with MB. It was recognized by early investigators that, despite clinical control of disease in selected patients, RT had the potential to impose significant neurocognitive, endocrine, hearing, and developmental sequelae as a consequence to curative management [3, 4, 5, 6]. The sequelae can often be life altering for the cancer survivors and impose significant long-term changes in function for living with the consequence of management. The goal of introducing systemic therapy for these patients was to improve upon survival outcomes and potentially limit normal tissue sequelae imposed by higher-dose RT.

There have been challenges integrating chemotherapy with RT in these vulnerable patients. Although CSRT is highly effective, it imposes measurable acute and late effects due to the volume of normal tissue unintentionally included in the RT treatment field. Because disease has access to the cerebral spinal fluid pathways, the entire cranial and spinal neuroaxis is considered a clinical target volume and requires RT for microscopic and macroscopic disease control [7, 8]. In the initial RT strategy for treatment, the anatomy of the central nervous system and the neuroaxis became what we refer to today as the clinical target volume. Chemotherapy can have an additive effect to tumor control as well as impose acceleration of injury to tissues being irradiated. With the fluid pathways at risk for disease, the entire central nervous system, by default, receives treatment [7]. For patients, especially younger children affected by the disease, this can impose injury with consequence. For example, the temporal lobe and temporal bone mature at a later time point in development than other areas of the central nervous system; therefore, neurocognition, endocrine, auditory, processing, and verbal skills can be negatively impacted by therapy at a time point important for growth and development during adolescence [3, 4, 5, 6]. Exit dose from thoracic spine fields can impose cardiopulmonary sequelae relative to field width size in an age-related manner with direct impact on growth and development of the spine as an additional problem. Exit dose can also influence bowel, renal, ovarian, and testicular function if dose is not managed as part of a conformal avoidance strategy. Vertebral body development and development of pelvis structures can likewise be influenced [9], especially in patients with high-risk features that require a higher dose to target, therefore leading to a higher dose to bone including joint spaces. Although the target volume for the RT boost plan can be titrated with confidence to spare normal tissue including the cochlea from high dose, studies to date have demonstrated that decreasing the dose to the cranial spinal axis, in selected favorable patients, cannot be safely accomplished for the general population. Therefore, even moving forward, protocols written for molecular-driven subsets of patients with this disease will require extended volume RT for the majority of patients, possibly excluding those with highly favorable features on biomarker analysis and the exceptionally young.

Chemotherapy application strategies have to be both comprehensive and thoughtful. Interdigitating chemotherapy either before, during, or after RT has to be accomplished in a manner that the patient receives the best of both disciplines and not the additive consequence of both therapies. Because of the complexity of interdigitating therapy as part of a composite treatment program with goals of improved tumor control and normal tissue outcome, studies have not uniformly shown benefits to combined modality therapy. Nevertheless, chemotherapy including the advent of biomarker-specific targeted therapy remains the best option to improve outcomes and balance the known effects of RT.

In an important historical paper written by Michiels and colleagues from the Netherlands in 2015, investigators identified seven retrospective and historical clinical trials for meta-analysis to determine the efficacy of treatment including and not including chemotherapy in patients treated for MB. The studies included 1080 patients and intended to review both event-free survival (EFS) and disease-free survival (DFS). As often the case with retrospective analysis, there were inconsistencies in applying metrics for disease progression, and essential information including stage at presentation and completion of surgical intervention was often less clear. Two of the studies without clear definition of disease progression revealed no statistical difference between those treated with chemotherapy and not treated with chemotherapy. In another study, improvements in the chemotherapy-treated patients were only evident during the course of treatment; however, the significance was no longer visible with longer follow-up. The review identified one retrospective clinical trial comparing standard-dose RT and reduced-dose RT coupled with chemotherapy without a difference in EFS or DFS between the groups. This trial did not evaluate quality of life nor had sub-group analysis. Multiple chemotherapy agents were used in these protocols and include cisplatin (CDDP), cyclophosphamide (CPM), vincristine (VCR), etoposide (VP-16), carboplatin, methotrexate (MTX), lomustine (CCNU), and topotecan. Chemotherapy was applied at multiple timepoints as part of the evolution of standard of care for this disease. Although the review has considerable strength and value, the review points to the need for trials to be conducted with rigor and discipline to optimize what we can learn from the trial and generate process improvements in clinical care for this vulnerable patient population. If outcomes for patients can be improved and applied directly into clinical care, the information from the trial must be trusted by the oncology clinical community. This is especially true in a disease such as MB, which requires hundreds of institutions to participate on a worldwide basis in order to accrue enough patients to trial and answer important study questions concerning both tumor control and normal tissue tolerance [10].

The National Clinical Trials Network (NCTN) of the National Cancer Institute (NCI) provides an extraordinary and robust administrative and quality assurance mechanism for the conduct of oncology-driven clinical trials. Data management centers coupled with important core services including the Imaging and Radiation Oncology Core (IROC) and tissue banks provide security that data will be acquired, processed, and maintained in a manner to support analysis of primary and secondary clinical trial endpoints in a rigorous and consistent manner. Data, including pathology objects, can be archived and repurposed for secondary analysis of additional study endpoints not anticipated at the time of the initial trial design. Through this prism and data management architecture, clinical trials are developed and vetted through NCI disease and discipline trial committees in order to (1) ask important and clinically relevant study questions and (2) support the infrastructure to ensure appropriate trial conduct and management. The Children’s Oncology Group (COG) is one of the major clinical cooperative groups and is an outstanding resource for this purpose as it includes all of the necessary administrative and clinical tools necessary to manage hundreds of clinical trials and thousands of trial participants. During this next section, we will review how chemotherapy has been applied in protocol settings and what we can anticipate as a treatment strategy as we move toward molecular classification of the disease [11, 12, 13].

3. Chemotherapeutic strategy in medulloblastoma

Current approaches in the treatment of MB include both adjuvant and maintenance chemotherapy regimens and RT. Protocols are working to determine combination chemotherapy regimens that improve overall and progression-free survival (PFS), sensitize tumors to RT, and limit toxicities.

3.1 A9961

This study was an intergroup study involving the Children’s Cancer Group (CCG) and the Pediatric Oncology Group (POG) prior to the merger of the groups into the COG. The study treated patients ages 3–21 with standard-risk MB who had negative cerebral spinal fluid and less than 1.5 cm² residual disease in the posterior fossa. Patients were treated to a dose of 23.4 Gy to the cranial spinal axis with a boost dose of 32.4 Gy to the anatomically defined posterior fossa generally with lateral fields and two-dimensional radiation planning techniques. The study tested if the use of post-RT application of cyclophosphamide would improve outcome compared to established regimens. This study enrolled 421 patients. After RT was completed, patients were randomized to treatment with either adjuvant CCNU (oral; 75 mg/m² on day 1), CDDP (75 mg/m² IV on day 1), and VCR (1.5 mg/m² IV bolus, max dose 2 mg on days 1, 7, 14) or adjuvant CPM (1000 IV mg/m² on days 21, 22), CDDP, and VCR (Figure 1). The overall survival at 5 years was 86%, and the overall 5-year EFS was 81%. The CCNU and CPM arm results were in similar PFS, overall survival, and toxicity. Ototoxicity was the most significant adverse event in both arms. Although likely a competing consequence from CDDP treatment, patients were treated with lateral-field RT for both primary management and the boost, therefore incorporating auditory structures including the cochlea in the treatment field receiving prescription dose, which affected hearing as a late effect in both study arms [15].

Figure 1.

Protocol A9961 treatment schematic [14].

The study gave us insight into developing strategies to manage clinical trials to make certain the requirements of the study and data management infrastructure can answer study questions including questions with chemotherapy. In retrospect, multiple patients were excluded from analysis as many had incomplete data including inability to assess post-surgery studies. The study was completed at a time when data were submitted through hard copy and courier; therefore, pre-treatment review of imaging and RT objects could not be performed. Because two cooperative groups were involved in the study (both later merged together as the COG), on-treatment review of objects was mandated for the POG institutions and not for CCG institutions as on-treatment review of imaging and RT objects was not an established component of the CCG data management process. The study foreshadowed the need for process improvements in clinical trial data management if the trial was going to be successful. Upon retrospective review of available imaging objects, more than 10% of the study patients had image-guided evidence of high-risk disease including a larger tumor burden of 1.5 cm² in the posterior fossa and/or disseminated disease in the cerebral spinal pathways. In the study, survival was only 36% in this subgroup of patients. Because half of the patient population had radiation oncology objects reviewed in a retrospective manner, there was a considerable number of treatment volume major study deviations largely centered on coverage of the cribriform plate, skull base, and posterior fossa boost. Interestingly, unlike other similar studies, the deviations in this study did not uniformly affect the study outcome upon review, particularly in the posterior fossa. This likely influenced the structure of COG ACNS0331, which sought to evaluate the use of an image-guided boost to the surgical cavity and not intentionally include the entire contents of the posterior fossa as in A9961 [16]. The assumption was if exclusion of segments of the anatomically defined posterior fossa were unintentionally excluded in the RT treatment field and did not directly affect local control or study outcome, perhaps a titrated volume directed to an image-guided disease target would be sufficient for protocol and clinical care management for patients with this disease [15].

The study also pointed out the importance of imaging in clinical trials and the application of imaging to the field of radiation oncology. If images were made available for central review prior to initiation of therapy coupled with the review for response assessment and pattern of failure, the number of ineligible patients on this study would have been significantly decreased and the study would potentially be powered to answer additional study questions. If patients are not treated to radiation volumes that are study compliant and less than the protocol dose is delivered to therapy targets, it becomes daunting to assess the success of chemotherapy. Each component of a trial plays an important role, and each treatment, response assessment, and pathology validation becomes essential to the mission if the results of the study are to be trusted and applied to clinical care. This study provided direction for how trials need to be managed to generate validation of the process and outcome. Imaging is our best tool to find common ground between investigators and institutions who participate in clinical trials, and it serves to support the effectiveness of chemotherapy in a quantitative manner [15].

3.2 ACNS0332

This MB study was designed to address the effectiveness of chemotherapy intensification with carboplatin used as a sensitizer to RT and isotretinoin as a proapoptotic therapeutic agent in patients with high-risk MB coupled with correlative biology studies. High-risk features included metastasis, residual disease, and diffuse anaplasia on pathology. This population received 36 Gy RT to the craniospinal axis with weekly VCR with/without carboplatin followed by six cycles of maintenance chemotherapy using CDDP, VCR, and CPM with/without 12 cycles of isotretinoin during and following maintenance chemotherapy (Figure 2). This study was important as it supported distinctions in response to therapy with biomarker analysis and opened the door to alternate approaches to chemotherapy and RT to molecular subgroups within the MB patient population [17, 18]. During the conduct of ACNS0332, four molecular subtypes of medulloblastoma were defined, namely Wingless-type mouse mammary tumor virus integration site (WNT), Sonic Hedgehog (SHH), Group 3, and Group 4 [17].

Figure 2.

Protocol ACNS0332 treatment schematic.

The isotretinoin component of the study was closed early due to futility; however, there was a significant 19% advantage at 5 years for the Group 3 patients receiving carboplatin supporting the use of intensification treatment strategies in this high-risk patient population. For this high-risk population, the 5-year disease-free survival was 66.4%. Of importance in this study was the identification that tumors that were part of the WNT (wingless and proto-oncogene Int1) pathway-activated biomarker demonstrated 100% 5-year survival, with the SHH biomarker disease demonstrating 53.6% 5-year survival, Group 3 demonstrating 73.7% 5-year survival, and Group 4 demonstrating 76.9% 5-year survival. Overall survival and EFS at 5 years in all groups were 73.4% and 62.9%, respectively [18].

3.3 ACNS0331

More studies continue to be done to evaluate the efficacy of lower CSRT doses with increased chemotherapy doses. One study did this by increasing chemotherapy doses using anti-leptomeningeal drugs while lowering the boost volume dose (focused dose to the area of the tumor bed plus an additional circumscribed margin, rather than treating the entire posterior fossa). Patients aged 3–7 were randomized to reduced-dose CSRT (18 Gy) or standard-dose CSRT and were then randomized to smaller-volume boost (RT to tumor bed +5.4 Gy to posterior fossa) or standard-volume boost. Patients received adjuvant VCR during RT. Patients then received maintenance chemotherapy for nine cycles, which followed an AAB cycling pattern. Therapy A is CDDP, CCNU, and VCR, and Therapy B is CPM, VCR, and mesna. Patients aged 8–21 received standard-dose CSRT and were then randomized to smaller-volume boost or standard-volume boost, with adjuvant VCR, followed by the same maintenance therapy (Figure 3) [16].

Figure 3.

ACNS0331 treatment schematic.

This was a study of considerable importance to the field of MB therapy. The study asked important questions of the RT community and, in turn, addressed issues associated with integration of chemotherapy with alterations in RT. Prior to the development of advanced technology volumetric imaging, the radiation oncology community used anatomic guidelines to treat the posterior fossa as part of a boost or final-phase strategy after completion of cranial spinal RT. In this study, patients had an imbedded randomization to receiving final-phase RT to the traditional target volume or the image-guided surgical cavity defined by the postoperative surgical cavity with consideration to include potential tissues at risk defined on pre-operative imaging. This was made possible because of the advent of volumetric magnetic resonance imaging coupled with process improvements in digital data transfer tools permitting rapid acquisition of data and pre-therapy review of objects at quality assurance centers. The RT treatment objects were reviewed in real time with pre-therapy feedback given to sites for adjustments in the plan if needed to ensure RT compliance to protocol guidelines. Therefore, the execution of RT on study was highly consistent with study objectives relative to target volumes and radiation dose and confirmed that volume-specific RT targets were not inferior to traditional posterior fossa treatment volumes. The study also asked an important question as to whether radiation dose to the craniospinal axis could be reduced in patients seven or younger in order to limit toxicity imposed by therapy. The study demonstrated that dose reduction for younger children was not feasible for the general population. However, due to the suggestion in other studies that select subgroups of patients have excellent outcomes with titrated therapy, there may be an opportunity to revisit RT dose titration in patients with established favorable prognosis.

3.4 Current trials

One of the challenges in performing clinical trials in selected subgroups in MB is meeting accrual objectives to study important questions within the subgroups. In this disease, there are a limited number of patients within the subgroups, and most trials have studied the subsets as part of a larger trial treating all patients in a uniform manner evaluating response within the subset. A recent protocol was designed in COG to devise an alternate care strategy to MB patients with the favorable WNT biomarker. WNT expression comprises 10% of the MB population. The goal is to study patients >3 years and <22 years with this expression product and who have standard risk features including negative cerebral spinal fluid and <1.5 cm² residual disease in the posterior fossa. The protocol uses a reduced-dose cranial spinal therapy (18 Gy) with a boost to the image-guided target (36 Gy) for a total dose of 54 Gy. Unlike most protocols, the patients on study do not receive vincristine VCR during RT. Maintenance chemotherapy is delivered at a limited dose in alternating A and B cycles. Cycle A includes CDDP (75 mg/m²), CCNU (75 mg/m² oral dose), and VCR (1.5 mg/m²) on day 1 and VCR (1.5 mg/m²) on days 8 and 15. Cycle B includes CPM (1000 mg/m²-days 1 and 2) and VCR and mesna with CPM [19].

The study titrates both RT and chemotherapy. The study has been activated, and the results will help evaluate the ability to titrate therapy to a favorable patient population and conduct a trial in a subgroup of patients with MB. This is important for clinical trials in this disease. As biomarker-associated therapies become more available for management of subgroups of patients affected with this disease, it will be important to establish protocol driven pathways as both individual and integrated pathways to make clinical translational progress in this disease.

Likewise, an additional current study seeks to evaluate the efficacy of sodium thiosulfate during CDDP-containing chemotherapy cycles in reducing ototoxicity in children with average-risk MB. One such study seeks to evaluate whether sodium thiosulfate treatment will reduce this toxicity and not negatively influence outcome. The average-risk cohort receives 23.4 Gy CSI and 30.6 Gy IFRT boost, while the low-risk cohort receives 18 Gy CSI and 36 Gy involved-field RT (IFRT) boost. Risk is based on molecular and karyotype profiling. Maintenance chemotherapy is given for nine cycles in an AAB cycling pattern, with Cycle A containing CDDP, CCNU, VCR, and sodium thiosulfate (12.8 g/m² at day 1), whereas Cycle B contains cyclophosphamide, mesna, and vincristine. The use of cavity-targeted RT will also help address issues associated with ototoxicity [20].

The protocols have matured over several decades, and each was built on strong science available at the time of the development of the study. In diseases such as MB, one of the challenges is meeting accrual goals to ensure the study is powered to meet the objectives of the study. For protocol A9961, 189 institutions were required to participate in the study to meet the accrual objective of 421 patients [21]. Therefore, trial guidelines need to be written in a manner to encourage worldwide participation. In turn, this requires that the quality assurance program and data management platforms be strong with the ability to provide feedback to sites and site investigators in real (same day) time to ensure the quality of the data and plan compliance to study objectives including validation that the correct patient has been enrolled in the correct study. As many as 10% of patients were found ineligible for study, largely due to upstaging patients to high-risk status following retrospective review of pre-study imaging, which upon review demonstrated evidence of disease in the spine and greater than 1.5 cm² residual disease in the posterior fossa. Improved quality assurance programs including real-time study review by study investigators supporting onsite investigators have greatly facilitated this process and have served the following protocols well as we begin to study subsets of medulloblastoma patients in clinical trials [11, 22].

4. Future vision

The classification of MB is both histological and genetic, and recent associations between the two may be the basis for future methods of treatment. Histologically, MB is classified into the classic, desmoplastic/nodular (DN), extensive nodularity (MBEN), large cell/anaplastic (LCA), and not-otherwise specified subtypes, with the classic variant being the most common (72% of all MBs) [15, 23, 24, 25]. These tumors are composed of normocytic cells with round nuclei with infrequent mitoses. Homer Wright rosettes are commonly seen [26, 27]. DN MB has areas of neurocytic differentiation with surrounding areas of embryonal character. These tumors also have collagen deposits around cells [28], which are detectable with reticulin staining. The MBs with extensive nodularity (MBEN) (variant of DN MB) are frequently found in infants and have more irregular nodules that merge [25]. The LCA MB is typically divided into large-cell and anaplastic subtypes. Anaplastic MB is marked histologically by larger, more pleiomorphic tumor cells with increased mitoses, prominent nucleoli, as well as apoptotic tumor cells, whereas the large-cell subtype has less pleiomorphism but maintains the high rate of mitosis and apoptosis [24, 29]. Since 2016, anaplastic and large-cell MBs have been combined into one histologic category, as the two are often difficult to distinguish from one another. The LCA MB is associated with worse prognosis/higher risk when compared with the classic subtype [29, 30].

Importantly, the histologic features of some MB subtypes were found to be correlated with certain molecular features, which may represent a future vision of more targeted therapies in MB. Specifically, desmoplastic/nodular MB was found to be associated with SHH-activated tumors. LCA MB is correlated with MYC and MYCN amplification, SHH tumors with TP53 variations, and the G3 molecular subtype [31]. While surgical resection, RT, and chemotherapy remain the mainstay of treatment in MB, current advances in MB include targeted molecular therapy based on these new biomarker/pathology correlates.

Recent advances have shown certain mutations correlate with certain subgroups of MB, providing more specific tumor classification. The WNT tumors were found to be associated with CTNNB1, TP53, DDX3X, MLL2/3 mutations, TNRC6C methylation, and OTX2 and CDK6 amplifications. Overall, the WNT molecular subgroup is associated with high 5-year survival (97% in WNT α and 100% in WNT β) [24, 31, 32, 33]; SHH tumors are associated with PTCH1, SMO, SUFU, TP53, DDX3X, CREBBP, MLL2/3, TERT, and KDM6A mutations; TNR6C6, MX11, and IL8 methylation; MYCN, CDK6, and GLI2 amplification; and PTEN loss [24, 31, 33]. SHH α and β have 69.8% and 67.3% 5-year survival, respectively. SHH γ and δ have higher 5-year survival, at 88% and 88.5%, respectively. Molecular subgroup 3 (Non-WNT/non-SHH) is associated with TERT and KDM6A mutations; TNRC6C, MX11, and IL8 methylation; MYC, OTX2, and CDK6 amplification; KDM6A loss; and KBTBD4 insertion. Group 3 is associated with worse prognosis and the highest metastatic potential, with 5-year survival of 66.2%, 55.8%, and 41.9% in subgroups 3α, 3β, and 3γ, respectively. Subgroup 3γ has the lowest 5-year survival of all subgroups [31, 34, 35], In addition, studies have found Group 3 MB pathogenesis to be regulated by PRUNE-1 amplification, with PRUNE-1 being regulated by LSD1/KDM1A [34, 36]. Lastly, Group 4 (non-WNT/non-SHH) is associated with TERT and KDM6A mutations; TNRC6C, MX11, and IL8 methylation; Lmx1A enhancer activation, PRDM6 induction; MYCN, OTX2, and CDK6 amplification; PTEN and KDM6A loss; and KBTBD4 insertion. Five-year survival is 66.8%, 75.4%, and 82.5% in the 4α, 4β, and 4γ subgroups, respectively [31, 35, 37].

Many MB tumors are associated with SHH activation, so current approaches to target this signaling pathways may prove beneficial. Vismodegib, sonidegib, and glasdegib are SMO inhibitors that may be effective on tumors with mutations in the SHH pathway; however, mutations downstream of SMO may not benefit from this treatment [31, 33, 38]. Phase I studies performed by LoRusso et al. and Gajjar et al. of vismodegib showed an acceptable safety profile [39, 40]. LoRusso et al. showed vismodegib had antitumor activity in 20/68 patients (including the one patient in the study with MB) [39, 40]. Gajjar et al. showed antitumor activity in 1 of 3 SHH-subtype MB patients [39]. Rodon et al. showed sonidegib was safe and had antitumor activity in 3/9 patients with MB [41]. Robinson et al. showed vismodegib was safe and had antitumor activity in 4/40 SHH-subgroup MB patients. Robinson et al. showed vismodegib may have antitumor effects in adults with recurrent SHH-MB in two phase II studies [42]. As stated, SMO inhibitors may only have efficacy for tumors with mutations upstream of SMO. In addition, GL1 transcription factors mediate the HH pathway, which is often activated alongside the SHH pathway in medulloblastoma [31, 38]. Glioma-associated oncogene antagonist-61 (GANT61) is a GL1 antagonist shown to cause apoptosis in MB tumor cells. Arsenic trioxide has also been shown to block GLI1 transcription, which may also prove beneficial in SHH-activated MB, but it is still being used in preclinical models [31, 38].

PI3K pathway modifications have also been proven to be implicated in MB pathogenesis. Novel methods to target this pathway include BEZ235 (PI3K inhibitor) combined with vismodegib, which may be a promising treatment strategy. Additionally, the PI3K inhibitor samotolisib is currently being used to treat MB in clinical trials [31].

Another future direction for WNT-MB is to target the WNT/β-catenin pathway with PRI-724, a CREBBP-CTNNB1-interacting antagonist; however, this is not currently being used to treat MB [31, 38, 43].

The CDK4/6 signaling pathway is also implicated in non-WNT MBs, and targeting it has shown some success. Palbociclib (CDK4/6 inhibitor) has been shown to be effective in the Group 3 molecular subgroup and is being used in other trials for other subgroups [31, 38, 44]. Additionally, loading CDK4/6 inhibitors on nanoparticles may increase delivery of drug to tumor tissues. This is currently being tested in mouse models [45].

As Group 3 MB is mediated by PRUNE-1 and LSD1/KDM1A, novel approaches targeting these pathways represent hope in treating this subtype given its poor prognosis [34, 36]. A study done by Bibbò et al. showed that the combination of a PRUNE-1 inhibitor (AA7.1) and a LSD1/KDM1A inhibitor (SP-2577) inhibited tumor cytotoxicity, down-regulated N-cadherin expression, and inhibited tumor oxidative phosphorylation through altered mitochondrial metabolism [34].

Lastly, MYC-pathway disruptors, like JQ1 and BRD4 (bromodomain inhibitor) are thought to be another strategy in treating MYC-activated tumors, which are associated with poor prognosis and Group 3 MB; however, this therapy is not currently being tested [31, 46]. SETD8 inhibitors may also target this pathway by targeting the chromatin in genes implicated in MYC-mutated tumors [38, 47].

Other therapeutic approaches that researchers are considering are the development of targeted miRNAs associated with certain subgroups, as well as other molecular mutations that are highly associated with MB including DDX3x [30]. One small molecule inhibitor of DDX3 is RK-33, which has shown promise in causing cell cycle arrest and decreased expression downstream genes in the MB-WNT pathway [48].

5. Conclusions

Considerable progress has been made over the past decades incorporating chemotherapy regimens into the care of patients with medulloblastoma, both as agents to improve outcome and likewise balance the sequelae imposed by radiation management and provide patients with the better of both worlds. With the advent and use of molecular biology classification of disease, future treatment strategies will include targeted agents associated with biomarkers, which may both further promote the use of systemic therapy in this disease and titrate the use of traditional chemotherapy. We are at the edge of a new age in this disease, and it is likely not long before the paradigm will shift toward target-specific therapies as part of systemic therapy in this disease. While traditional treatment has had success, change is evident, and patients will benefit.

Acknowledgments

The Children’s Oncology Group is supported by the National Institutes of Health (U10CA180886, U10CA180899) and the St. Baldrick’s Foundation.

Conflict of interest

The authors declare no conflict of interest.