Therapeutic priorities sharpen as medulloblastoma biology splits into very different risk groups
The second Paediatric Therapeutic Development Workshop on medulloblastoma centered on a disease that is both treatable and unforgiving. Surgery, radiotherapy, and chemotherapy allow 60% to 70% of patients to survive, yet the price of cure is often steep: long-term neurologic, cognitive, and endocrinological toxicity affects around 80% of survivors. At the other end of the spectrum, the highest-risk disease groups still carry a survival probability below 10%. Those two realities framed the workshop’s central task: identify the therapeutic vulnerabilities that matter most across a biologically heterogeneous pediatric brain tumor while avoiding another generation of treatment-related harm.
Medulloblastoma sits in the posterior fossa and remains the second most common malignant brain tumor in childhood. Its molecular classification is no longer a technical detail reserved for pathology reports; it now defines prognosis, relapse pattern, and the feasibility of de-escalation or intensification. The workshop therefore focused on the subtypes with the poorest outcomes, especially SHH-medulloblastoma with MYCN amplification or TP53 mutation, and Group 3 medulloblastoma with c-MYC amplification. At the same time, the group also addressed WNT-medulloblastoma, where survival is relatively favorable and the clinical problem is not under-treatment but over-treatment.
The therapeutic agenda followed the biology. A degrader directed at SRC emerged as a high-priority strategy, reflecting the appeal of attacking a signaling node that can support malignant behavior without relying only on conventional cytotoxic intensity. In parallel, the workshop identified inhibition of the tumor-relevant functions of c-MYC and MYCN as a priority for poor-prognosis disease. Those oncogenes are not easy targets, which is precisely why the discussion mattered: in medulloblastoma, the most dangerous tumors are often driven by dependencies that resist simple drug design. If those drivers can be functionally interrupted, the payoff could be substantial in the very groups where survival remains poorest.
One sentence carried special weight for the clinicians in the room:
High-risk medulloblastoma still lacks an approved targeted therapy.
Reducing toxicity in WNT disease while pursuing new targets in aggressive subtypes
The workshop did not treat all medulloblastoma as a single therapeutic problem. WNT-medulloblastoma was discussed as a good-prognosis tumor in which further toxicity reduction is a rational goal. Here, the priority was a radiolabelled theranostic antibody approach, an innovative strategy intended to preserve disease control while lowering collateral damage. That concept fits the clinical reality of WNT disease: when cure rates are already relatively high, the next advance should not be measured only in survival curves but also in preserved neurodevelopment, endocrine function, and long-term quality of life.
The same logic does not apply to the highest-risk groups. SHH tumors with TP53 mutation and MYCN amplification, as well as Group 3 tumors with c-MYC amplification, remain biologically aggressive and clinically devastating. The workshop pointed toward B7-H3 as a particularly attractive target because it offers multiple translational routes, including CAR T-cell and antibody-drug conjugate platforms. Those options matter because a target with broad therapeutic utility can be pursued along more than one development path, increasing the odds that one approach will prove workable in children with refractory disease.
The discussion also moved beyond single-agent development. Based on currently available evidence, combinations of central nervous system-penetrant selective PARP-1, CHK1/2, or CDK9 inhibitors with an ATR inhibitor could be evaluated in early-phase trials for high-risk patients. Even so, the workshop drew a firm line: these combinations still require robust preclinical evaluation before they enter children’s studies. That caution is appropriate. In medulloblastoma, a biologically plausible combination can still fail if penetration into the central nervous system is inadequate, if the target dependency is absent in the right subtype, or if overlapping toxicity overwhelms any therapeutic gain.
Small patient numbers demand international trials, biomarker logic, and better disease models
The workshop repeatedly returned to the practical barriers that slow progress in pediatric brain tumors. Medulloblastoma is rare, subtype diversity is high, and the most urgent questions involve narrow molecular subsets. Trial design therefore cannot mimic the development model used for common adult solid tumors. Early-phase studies need to be international, use novel designs suited to small patient numbers, and incorporate correlative biological studies from the outset. Without that structure, it becomes difficult to know whether a drug failed because the biology was wrong, the dose was wrong, the subgroup was wrong, or the model system failed to recapitulate human disease.
Preclinical model quality remains a major constraint. Disease models are scant for several medulloblastoma subtypes, and Group 4 in particular remains difficult to grow in the laboratory. Patient-derived xenograft models may help close that gap, but they are not a perfect substitute for native tumor biology. The workshop also emphasized a more recent conceptual shift: oncogene aberrations may not initiate the tumor in every case, yet they can remain essential for progression and therapy resistance. In a disease with marked intertumoral heterogeneity, that distinction shapes target selection. A pathway that is relevant late in tumor evolution may still be a valuable therapeutic vulnerability, but only if the right patients are identified and the right biological context is tested.
The practical consequence is straightforward. Future medulloblastoma development will need both targeted therapeutics built around specific vulnerabilities and rational combinations of existing medicines that can be moved carefully into high-risk disease. The workshop participants also positioned future development through the C-Further consortium, which can provide funding and laboratory support to de-risk early projects. What remains unresolved is not whether the field has identified enough candidate targets; it has. The harder task is deciding which targets will survive rigorous preclinical testing, which combinations will remain tolerable in children, and which biologically defined subgroup will be able to benefit without adding to the lifelong burden that medulloblastoma treatment already leaves behind.