Re-irradiation in children with recurrent central nervous system (CNS) tumors remains one of the most challenging areas of pediatric radiation oncology. Although a second course of radiotherapy may provide meaningful disease control and, in selected patients, another opportunity for cure, uncertainty persists regarding patient selection, target volume definition, cumulative radiation tolerance, and long-term toxicity.
These questions were addressed during a recent SIOP Global Health Network webinar on pediatric radiotherapy, featuring Professor Thankam Ajithkumar of Cambridge, United Kingdom, an expert in pediatric CNS tumors and re-irradiation.
Organized through an international collaboration involving the SIOP Global Health Network, the SIOP working group, and the West German Proton Center in Essen, the session examined current evidence, clinical experience, and the limitations that continue to influence treatment decisions.
Drawing on published studies, long-term follow-up data, and clinical cases, Professor Ajithkumar discussed three central questions: when should children be re-irradiated, what target volumes should be treated, and how can treatment be delivered while limiting cumulative toxicity?
When Is Re-Irradiation Justified?
The decision to re-irradiate a child with recurrent CNS disease depends on several factors, including tumor histology, recurrence pattern, previous treatment, expected clinical benefit, and the risk of radiation-related complications.
Professor Ajithkumar opened his presentation with the case of a patient diagnosed with localized non-germinomatous germ cell tumor who initially received induction chemotherapy followed by whole-ventricular radiotherapy and a focal boost to a total dose of 54 Gy.
Only two weeks after completing treatment in February 2023, the patient developed an L4 spinal recurrence accompanied by elevated alpha-fetoprotein (AFP).
The recurrence substantially altered the prognosis, raising an important clinical question: could further radiotherapy contribute to curative salvage treatment despite the extremely short interval since the initial course?
This case illustrated the difficulty of establishing a minimum safe interval between radiation courses.
A comprehensive review discussed during the webinar identified a weighted median interval of approximately 2.3 years between treatments, although some patients underwent re-irradiation after only 2.6 months.
There is currently no universally accepted minimum interval for pediatric CNS re-irradiation. Clinical decisions therefore require individualized consideration of the anticipated benefit, previous radiation exposure, and normal tissue tolerance.
Evidence Across Pediatric CNS Tumors: Where Does Re-Irradiation Offer Benefit?
The available evidence varies considerably between tumor types, with the largest body of pediatric re-irradiation data coming from ependymoma.
Ependymoma: Encouraging Results, but Persistent Uncertainty
Early prospective experience from St. Jude Children’s Research Hospital demonstrated the potential role of re-irradiation in recurrent pediatric ependymoma.
A study involving 38 patients reported a four-year event-free survival of 53%, with treatment approaches including focal re-irradiation and craniospinal irradiation (CSI).
Across published studies, three-year event-free survival has ranged from 28% to 56%, while three-year overall survival has ranged from 48% to 81%.
Most studies with relatively short follow-up reported grade 3 necrosis rates below 10%.
However, these findings must be interpreted cautiously because of substantial differences in patient selection, treatment volumes, dose prescriptions, and follow-up duration.
Germinoma: The Potential for a Second Cure
For recurrent intracranial germinoma, which occurs in fewer than 5% of patients, salvage radiotherapy may offer particularly favorable outcomes.
A meta-analysis discussed during the presentation reported approximately 90% five-year overall survival among patients receiving craniospinal irradiation.
Patients who did not receive CSI or radiotherapy experienced significantly worse outcomes.
These findings suggest that appropriately selected patients with recurrent germinoma may achieve durable survival following salvage treatment.
Medulloblastoma and High-Grade Glioma: More Limited Outcomes
Re-irradiation for recurrent medulloblastoma presents additional difficulties because many patients have already received craniospinal irradiation during their initial treatment.
Published studies have reported three-year event-free or progression-free survival ranging from approximately 25% to 48%.
For pediatric high-grade gliomas, the available evidence generally indicates more limited benefits, although selected patients may experience temporary disease control or symptomatic improvement.
Professor Ajithkumar emphasized that identifying reliable clinical predictors of benefit remains difficult because existing studies include heterogeneous populations and treatment approaches.
Focal Re-Irradiation or Craniospinal Irradiation: Defining the Appropriate Target
One of the central controversies in pediatric CNS re-irradiation concerns the selection of treatment volume.
Depending on the location and extent of recurrence, patients may receive focal radiotherapy, irradiation of a broader anatomical region, or craniospinal irradiation.
Although smaller treatment volumes may reduce normal tissue exposure, they may not adequately address the risk of subsequent disease dissemination.
Professor Ajithkumar highlighted evidence from recurrent ependymoma demonstrating differences in progression outcomes between focal treatment and CSI.
In a small Canadian series involving 31 patients, five-year freedom from progression was approximately 83% following CSI compared with 15% following focal re-irradiation.
However, the limited sample size and differences between treatment groups prevent definitive conclusions.
The importance of recurrence patterns was further illustrated through comparisons of treatment practices in pediatric ependymoma.
In the St. Jude series, approximately 85% of patients received 59.4 Gy during initial treatment, compared with 45% in a French series. The predominant patterns of recurrence also differed, raising questions about whether initial radiation dose influences subsequent disease distribution.
A separate clinical case involving a nine-year-old boy with anaplastic ganglioglioma demonstrated the limitations of focal salvage treatment.
Following initial surgery and postoperative radiotherapy to 54 Gy in 30 fractions, the patient developed a local recurrence after 30 months. Re-excision was followed by focal re-irradiation to 45 Gy in 25 fractions.
Despite treatment, imaging four months later demonstrated diffuse spinal dissemination.
The case reinforced an unresolved question: should salvage radiation volumes be determined solely by the visible recurrence, or should they also account for the biological behavior and potential dissemination of the disease?
Current evidence remains insufficient to establish a universally preferred approach.
Cumulative Dose and Organ-at-Risk Tolerance: What Do We Actually Know?
Determining acceptable cumulative radiation exposure remains a major limitation in pediatric re-irradiation planning.
Professor Ajithkumar reviewed findings from the PENTEC literature, including an analysis of 449 patients with a cumulative equivalent dose in 2-Gy fractions (EQD2), calculated using an α/β ratio of 2 Gy, of approximately 111 Gy.
At a median follow-up of 1.6 years, reported rates were:
- Brain necrosis: 4.9%
- Brainstem necrosis: 3.1%
- Median time to necrosis: 5.7 months
Although these results suggest that selected patients may tolerate relatively high cumulative doses, the findings cannot be translated into universally safe dose constraints.
Most studies lacked detailed organ-at-risk dose-volume metrics, standardized toxicity definitions, and reliable cumulative dose calculations.
Professor Ajithkumar also described the cumulative EQD2 objectives used in his clinical practice, informed by published literature.
| Organ at risk | Cumulative EQD2 discussed |
|---|---|
| Brain | Approximately 100 Gy |
| Brainstem | Approximately 100 Gy |
| Brainstem following intensive chemotherapy | Approximately 80 Gy |
| Optic nerves and chiasm | Approximately 75 Gy |
| Spinal cord | Approximately 60–70 Gy |
These figures were presented as experience-based planning objectives rather than validated pediatric tolerance thresholds.
The discussion also addressed uncertainty regarding normal tissue recovery between treatment courses.
Experimental studies suggest that some radiation-induced injury may recover over time, but the extent of recovery in the developing pediatric CNS remains unknown.
Professor Beate Timmermann, contributing to the discussion from Essen, described a more conservative approach, generally allowing approximately 20–30% recovery even after prolonged intervals between treatments.
Both experts agreed that reliable pediatric-specific recovery models are lacking.
The Technical Challenges of Re-Irradiating a Developing Brain
Accurate cumulative dose assessment is particularly complicated in children because the brain and surrounding anatomy change during growth.
A patient initially treated at two years of age may have substantially different intracranial anatomy when considered for re-irradiation several years later.
These anatomical changes create difficulties for image registration, dose mapping, and cumulative organ-at-risk evaluation.
Additional uncertainties include the selection of appropriate radiobiological parameters, fractionation effects, and the use of biological effective dose (BED) or EQD2 calculations.
Professor Ajithkumar emphasized that accurate dose accumulation requires more than simply adding the prescribed doses from two treatment courses.
Ideally, treatment planning should account for voxel-level dose distribution, anatomical changes, fractionation, and potential tissue recovery.
However, standardized methods for integrating these factors into pediatric clinical practice remain underdeveloped.
Proton Therapy and Radiation Necrosis: An Unresolved Debate
The role of proton therapy in pediatric CNS re-irradiation generated considerable discussion.
Proton therapy offers the potential to reduce radiation exposure to surrounding normal tissues, but questions remain regarding relative biological effectiveness (RBE), beam arrangement, and radiation injury.
Professor Ajithkumar discussed recent long-term ependymoma data in which all-grade necrosis was reported in approximately 50% of patients treated with protons compared with 31% of those receiving photons.
The reasons for this difference remain unclear.
Potential contributing factors include differences in patient selection, previous surgery, treatment planning, proton delivery techniques, and institutional experience.
Professor Timmermann emphasized that these retrospective findings should not be interpreted as definitive evidence that proton therapy carries a higher necrosis risk.
She described substantially lower necrosis rates in her institution’s experience and highlighted the importance of careful beam arrangement and consideration of RBE uncertainties, particularly near the brainstem.
The discussion also addressed the potential implications of proton distal-edge placement, dose distribution, and differences between older passive-scattering techniques and contemporary proton delivery.
The available evidence does not establish that proton therapy is inherently more toxic than photon therapy in pediatric re-irradiation. More consistent dosimetric reporting and prospective comparisons are needed.
Radiation Necrosis or Treatment-Related Imaging Changes?
Another important issue is the interpretation of post-treatment magnetic resonance imaging.
New contrast-enhancing lesions after re-irradiation may represent tumor progression, radiation necrosis, or transient treatment-related vascular changes.
Professor Ajithkumar returned to the initial germ cell tumor case to illustrate this diagnostic challenge.
Following salvage high-dose chemotherapy, the patient underwent additional radiotherapy approximately five months after the initial course, receiving 30 Gy in 20 fractions with a further boost of 20 Gy in eight fractions.
Treatment was completed in August 2023.
Subsequent MRI examinations demonstrated contrast-enhancing abnormalities that were initially interpreted as possible necrosis or recurrent disease.
However, the patient remained under observation, and the abnormalities gradually resolved over approximately 18 months.
By August 2026, imaging showed no residual abnormal contrast enhancement, while tumor markers remained stable.
This case demonstrated the importance of multidisciplinary imaging review and longitudinal assessment before attributing radiological changes to irreversible necrosis or disease progression.
The speakers emphasized that isolated imaging findings should be interpreted alongside clinical symptoms, previous radiation distributions, and serial imaging.
Long-Term Follow-Up Reveals a Different Risk Profile
One of the most important findings discussed during the webinar came from updated long-term outcomes in recurrent pediatric ependymoma.
A recent report from the St. Jude experience included approximately 150 children treated for recurrent disease over three decades.
Earlier follow-up had demonstrated encouraging outcomes, but longer observation revealed continuing disease progression and an increasing burden of late complications.
At ten years, the reported progression-free survival was approximately 9%, while overall survival was approximately 10%.
Grade 3 necrosis requiring intervention, including treatments such as bevacizumab or hyperbaric oxygen, approached 20%.
The ten-year cumulative incidence of death from treatment-related complications and second malignancies was approximately 8.5%.
Reported second malignancies included glioblastoma, acute myeloid leukemia, osteosarcoma, and meningioma.
These findings demonstrate why short-term toxicity results may underestimate the full consequences of pediatric re-irradiation.
Children who survive for extended periods remain at risk of progressive radiation injury, treatment-related complications, and secondary cancers.
The results also reinforce the need to evaluate not only disease control but long-term neurological function, quality of life, and treatment-related mortality.
Toward Standardized Pediatric Re-Irradiation Practice
Despite increasing clinical experience, pediatric CNS re-irradiation continues to lack consistent treatment recommendations.
Professor Ajithkumar presented preliminary findings from an international survey involving 86 respondents.
The survey identified substantial variation in target volume selection, prescribed doses, organ-at-risk constraints, and technical planning approaches.
Approximately two-thirds of respondents reported having no formal institutional guideline for pediatric re-irradiation.
Although most clinicians prioritized organ-at-risk protection over target coverage, there was considerable disagreement regarding acceptable cumulative radiation doses.
Differences were particularly pronounced in complex situations involving disseminated recurrence or uncertain disease extent.
The findings highlight the need for standardized clinical and dosimetric reporting.
Future studies should consistently document patient characteristics, previous surgery and chemotherapy, radiation techniques, treatment intervals, cumulative EQD2, organ-at-risk dose-volume parameters, and clearly defined toxicity outcomes.
Standardized imaging criteria and long-term follow-up are equally important for distinguishing transient treatment effects from clinically significant radiation injury.
A Second Opportunity for Treatment, but Not Without Consequences
The webinar concluded with a central message: re-irradiation can provide meaningful benefit for selected children with recurrent CNS tumors, but treatment decisions must account for substantial uncertainty regarding cumulative toxicity and long-term outcomes.
For recurrent germinoma and selected cases of ependymoma, salvage radiotherapy may provide an opportunity for prolonged disease control or cure. In other tumors, the principal benefit may be temporary disease stabilization or symptomatic relief.
The appropriate timing, target volume, treatment modality, and cumulative dose remain areas of active investigation.
Professor Ajithkumar emphasized the importance of international prospective collaboration, standardized reporting, and more reliable pediatric-specific radiobiological data.
As clinical experience continues to expand, the priority is not simply to determine whether another course of radiation can be delivered, but to establish which children are most likely to benefit, how treatment can be delivered safely, and what the long-term consequences may be.
For pediatric radiation oncology, these questions remain central to improving the balance between disease control and preservation of neurological function and quality of life.
More articles like this are available on OncoDaily
