Traditionally, the milestone for treatment of standard-risk patients has consisted of complete or near complete surgical resection followed by postoperative CSRT. Using conventional doses of radiotherapy (36 Gy to the craniospinal axis together with a boost of 18–20 Gy to the posterior fossa, with a total dose of 54–56 Gy), different studies reported that after 5 years from diagnosis, between 55 % and 70 % of children are alive and free of progressive disease [21]. It is noteworthy that an elevated percentage of survivors have significant long-term sequelae, mostly neuropsychological deficits, linear growth deficit, and endocrine dysfunction due to irradiation of the pituitary gland and hypothalamic regions together with the effects of whole spine radiotherapy. In order to decrease the long-term neurocognitive effects of radiation and to control tumor growth, especially in young children, the dose administered to the brain and spine has been reduced.

At present, there is general agreement that after surgical resection, the gold standard treatment for patients older than 3 years at diagnosis is “reduced-dose” CSRT: a total dose of 23.4 Gy to the craniospinal axis within 40 days from surgery plus a localized boost to the posterior fossa (total dose of 54–55.8 Gy). The treatment is usually combined with weekly concurrent single administration of vincristine and followed by a multidrug regimen (cisplatin, vincristine, and lomustine or cisplatin, vincristine, and cyclophosphamide) [17] that provided a 5-year event-free survival of 80 %. The reduced-dose CSRT (23.4 Gy vs. 36 Gy) without the adjuvant chemotherapy has shown a higher number of early failures. The Children Oncology Group (COG) is evaluating in a randomized study further reduction of CSRT and posterior fossa boost dose, but at present, this treatment is not recommended. The HIT-SIOP PNET 4 trial, recently closed in Europe, compared the use of conventionally fractionated radiotherapy at a dose of 23.4 Gy plus boost versus hyperfractionated radiotherapy (HFRT) (1 Gy/fraction, 2 fractions/day) at a dose of 36 Gy plus boost, both followed by chemotherapy schedule with eight courses of vincristine (1.5 mg/m² for three doses), cisplatin (70 mg/m²), and lomustine (CCNU, 75 mg/m²). EFS and OS for HFRT were not better than the conventional one, which therefore remains standard of care in this disease [22]. However, a French study on standard-risk MB patients treated by hyperfractionated radiotherapy without adjuvant chemotherapy showed a 3-year PFS of 83 % with a good neuropsychological outcome at follow-up [23]. Future trials are needed to further evaluate the efficacy and safety of these treatment modalities.

Molecular biology has changed our knowledge of MB and has implications for diagnostic stratification and treatment. As newer biological agents are translated from the lab to the bedside, clinicians need to understand the fundamental signalling pathways that are targeted during therapy. More knowledge of the molecular biology of MB is needed so that more children will be cured or have an improved quality of life. Future protocols will stratify treatments on the basis of biological factors, such as beta-catenin, so a subgroup of “low-risk” patients will be identified to be addressed to only surgery and radiotherapy or chemotherapy without risk of relapse.

Prospective randomized studies performed as early as the late 1970s by the CCG and the SIOP, which compared PFS and OS in children treated with radiotherapy alone to those receiving radiotherapy plus chemotherapy given during and after radiation, demonstrated a statistically significant 15–20 % survival advantage for patients with poorer risk disease (defined as those with disseminated disease or larger and more infiltrative tumors and/or significant amounts of tumor left after surgery, as determined by postoperative MRI). These results were the basis of a series of trials performed over the ensuing two decades evaluating the efficacy of chemotherapy given at higher dose, coupled with other agents, or given in various sequences with radiotherapy [15, 17, 24]. Once again, studies using chemotherapy prior to radiotherapy have not shown benefit and possibly may have demonstrated inferior overall disease control rate [15]; on the other hand, chemotherapy treatment during and after radiotherapy has conducted to an improvement of survival rate [12]. Chemotherapy is therefore part of adjuvant treatment in this group of patients but optimal timing and schedule are not yet established.

A mono-institutional trial considering the use of CSRT followed by vincristine, cisplatin, and CCNU in high-risk patients reported a survival rate of around 85 % [25]. In an SIOP trial published with a 76-month follow-up, 27 metastatic patients treated with standard-dose RT followed by CCNU and vincristine obtained a 5-year PFS of 43 % [13]. Similar results have been reported by the SFOP study, which treated high-risk patients with the “eight-drugs-in-one-day” chemotherapy regimen, followed by two cycles of high-dose MTX, radiotherapy, and then further “eight-in-one” chemotherapy. The French national study confirmed the low rate of response to the “sandwich” chemotherapy without significant improvement in either M1 or M2/M3 patients, showing a 5-year EFS of 58.8 % and 43.1 %, respectively [26]. The “eight-in-one” chemotherapy regimen before and after RT has been also proposed by the CCG 921 randomized phase III trial, which reported a significantly lower PFS for metastatic patients [57 % M1; 40 % M2; 78 % M0, (p = 0.0006)] [27]. The randomized prospective multicenter trial HIT’91 compared two treatments: the postoperative neoadjuvant chemotherapy (ifosfamide, etoposide, high-dose methotrexate, cisplatin, and cytarabine given in two cycles) followed by CSRT and maintenance chemotherapy after immediate postoperative RT. For all randomized patients, the PFS resulted 65 % for M1 patients and 30 % for M2–M3 patients after 3 years of follow-up confirming that metastases stage plays a prognostic role [28]. Recently, high-dose chemotherapy and autologous stem cell transplantation provided encouraging results. Strother et al. reported that after 2 years, the PFS was 73.7 % ± 10.5 % in 19 enrolled patients with metastases treated by surgical resection and topotecan in a 6-week face window followed by CSRT and four cycles of high-dose cyclophosphamide (4,000 mg/m² per cycle), with cisplatin (75 mg/m² per cycle) and vincristine (two 1.5 mg/m² doses per cycle). Support with SCR or bone marrow rescue was administered after each cycle of high-dose CT [29]. Results of a study conducted on nine patients affected by supratentorial primitive neuroectodermal tumors and metastatic MB treated with high-dose chemotherapy (cyclophosphamide, cisplatin, vincristine, etoposide, and high-dose MTX) for two to three cycles before radiotherapy showed, after a median follow-up of 27 months, that seven patients remained in continuous complete remission [30]. More recently, 21 young patients with high-risk or disseminated MB were enrolled in order to evaluate the response rate to an intensified induction chemotherapy regimen and single myeloablative chemotherapy cycle with autologous SCR. This was followed by RT for patients more than 6 years of age or with evidence of residual disease on completion of the induction chemotherapy if under 6 years old. The 3-year EFS and OS were 49 and 60 %, respectively [31]. The European phase III clinical trial SIOP/UKCCSG PNET 3 demonstrated an unsatisfactory outcome in metastatic patients treated with neoadjuvant CT (vincristine, cisplatin, etoposide, and cyclophosphamide) followed by a standard CSRT dose with a posterior fossa boost and/or a boost to metastases, obtaining a 5-year PFS of less than 40 % [15]. Recently, Gandola et al. [32] reported 33 consecutive patients receiving postoperative high-dose chemotherapy (HDCT) with methotrexate plus vincristine, etoposide, cyclophosphamide, and carboplatin in a 2-month schedule. Hyperfractionated accelerated radiotherapy (HART) was delivered at a total dose to the neuraxis of 39 Gy (1.3 Gy/fraction, 2 fractions/day) with a posterior fossa boost up to 60 Gy (1.5 Gy/fraction, 2 fractions/day). In addition, patients with persistent disseminated disease before HART underwent two courses of myeloablative chemotherapy and circulating progenitor cell rescue. Otherwise, patients in complete remission received maintenance chemotherapy with vincristine and CCNU for 1 year. Of 32 evaluable patients (one patient dead by sepsis before radiotherapy), 22 responded to chemotherapy. 12/22 patients achieved complete remission (CR) before the radiation phase and 10/22 had radiological partial response (PR). In the remaining patients, disease was stable in five and progressed in five. All patients underwent radiotherapy. Fourteen of the 32 patients received consolidation therapy after HART. Eight patients relapsed after a median of 12 months after beginning chemotherapy. With a median follow-up of 82 months, the 5-year EFS, PFS, and OS were 70, 72, and 73 %, respectively. No severe clinical complications of HART have been reported. HART with intensive postoperative chemotherapy and myeloablative chemotherapy seems to be feasible without limiting major toxicity in children with metastatic MB.