Group 3 and group 4 medulloblastoma (non-WNT/non-SHH)

Groups 3 and 4 arise from the upper rhombic lip and unipolar brush cell lineages and share enough that WHO 2021 groups them as non-WNT/non-SHH medulloblastoma with eight methylation subgroups. Group 3 (about a quarter of medulloblastoma) occurs in infants and young children, often with large-cell/anaplastic histology, MYC amplification in about 17 percent, GFI1 enhancer hijacking, and metastases in 40 to 45 percent at diagnosis. Group 4 (about 35 to 40 percent) affects older children and adolescents, three boys to one girl, with isochromosome 17q in most, KDM6A and PRDM6 alterations and MYCN amplification in some, and metastases in about a third. Metastatic stage, MYC amplification and residual tumour define high risk in both.

The trials that set standard therapy predate the groups. SIOP PNET3 showed pre-radiotherapy chemotherapy improved event-free survival; COG A9961 in 2006 established the average-risk regimen of 23.4 Gy craniospinal radiotherapy with a boost followed by cisplatin, vincristine and cyclophosphamide or lomustine, with five-year event-free survival of 81 percent; HIT-SIOP PNET4 found hyperfractionated radiotherapy no better than standard. ACNS0331, reported in 2021, tried to lower the craniospinal dose to 18 Gy in children aged three to seven with average-risk disease and found it inferior, five-year event-free survival 71.4 percent against 82.9 percent, while confirming that boosting the tumour bed rather than the whole posterior fossa is safe. ACNS0332, for high-risk disease, showed that carboplatin given daily during craniospinal radiotherapy improved five-year event-free survival in group 3 from 53.7 to 73.2 percent with no benefit in group 4, and that isotretinoin maintenance added nothing. Infants with group 3 disease receive intensive chemotherapy with high-dose consolidation and stem cell rescue to avoid or delay radiotherapy.

Relapse in groups 3 and 4 is almost always fatal, recurs in the same molecular group, and typically occurs in the leptomeninges rather than the tumour bed; temozolomide, irinotecan and bevacizumab, re-irradiation and high-dose chemotherapy buy time. The current trials assign therapy by methylation subgroup: SJMB12 and the SIOP PNET5 and HIT-MED studies reduce therapy for low-risk group 4 (chromosome 11 loss or whole chromosome 17 gain without metastases) and intensify it for MYC-amplified group 3. No targeted drug has worked: MYC and MYCN have no inhibitor, and immunotherapy trials to date have shown little in an immunologically cold tumour. The long-term price of craniospinal radiotherapy, in intellect, hearing, endocrine function, stroke and second tumours, is why proton therapy is the preferred modality where available and why the dose question keeps being asked.