Pediatric Radiation Oncology — Board Review Summary

PART I — NON-CNS MALIGNANCIES

Wilms Tumor (Nephroblastoma)

Most common abdominal tumor of childhood · 6% of pediatric cancers · ~650/yr · median age just under 4 yo · painless mass that moves with kidney on US.

Syndromes: WAGR, Denys-Drash, Beckwith-Wiedemann. Associated findings in 10–13%: aniridia, GU anomalies, hemihypertrophy.
Histology: triphasic embryonal (blastemal 39%, epithelial 18%, stromal 1%, mixed 41%). Unfavorable: anaplastic (4–5%), CCSK (bone scan + brain MRI), rhabdoid tumor of kidney (brain MRI).
Prognosis worse: higher stage, UH, age >24 mo, LOH 1p and/or 16q.
Do NOT biopsy if you suspect Wilms — upstages favorable histology to Stage III via spill risk. COG approach: surgery first, then chemo (vs. SIOP: biopsy → chemo → surgery).
RT indications: FH Stage III–V; UH all stages. Start RT within 10–14 days of surgery (NWTS-1).

Scenario	Dose / Fractionation
Flank (Stage III FH; Stage I–III focal anaplasia; Stage I–II diffuse anaplasia)	10.8 Gy / 6 fx
Flank (Stage III diffuse anaplasia)	19.8 Gy
Whole abdomen (positive cytology, tumor spill, peritoneal seeding)	10.5 Gy / 7 fx
Boost to gross residual (nodes or abdomen)	+10.5–19.8 Gy
Whole lung	12 Gy / 8 fx (10.5 Gy / 7 fx if <1 yr)
Whole brain	21.6 Gy / 12 fx
Focal liver	19.8 Gy / 11 fx
Bone mets	25.2 Gy (<16 yo) / 30.6 Gy (≥16 yo)
Lymph nodes	10.8 Gy resected / 19.8 Gy unresected

AREN0533 (Dix, JCO 2018): FH lung mets without LOH → if CR to upfront chemo, omit WLI. EFS 79% vs 85% expected (p=0.052), OS unchanged. AREN2231 moving all RT to after induction to avoid field overlap.

Flank field rules: pre-resection tumor + 1 cm; avoid splitting vertebral bodies; don't cross >1 cm into contralateral kidney; if LN+ include para-aortics from crus of diaphragm to L5/S1.

Neuroblastoma

2nd most common abdominal tumor of childhood · most common in <18 mo · neural crest origin · ~600–700/yr · acutely ill appearance, raccoon eyes, mass moves separately from kidney.

Workup: urine catecholamines, CT C/A/P, MIBG (PET if not MIBG-avid), bone marrow bx, echo/MUGA, audiogram/BAERS. Do NOT resect at diagnosis (commits to larger RT fields) — contrast with Wilms.
Histologic risk factors: MYCN amplification, diploid, age >17 mo, LOH 1p or 11q, grade of differentiation.
INRG staging → risk stratification; Low/Int risk rarely get RT, High risk all get RT.
Treatment sequence (high risk): induction chemo → surgery → ASCT → RT → immunotherapy (anti-GD2).

Site	Dose / Target
Primary site (upfront, regardless of gross residual)	21.6 Gy / 12 fx
Metastases (up to 5 functionally-avid sites at ASCT)	21.6 Gy / 12 fx
GTV1 definition	Extent of disease at time of surgery (post-induction), including involved nodes. CTV1 = GTV1 + 1 cm (anatomically confined).

Key trials: CCG-3891 (Haas-Kogan IJROBP 2003) — TBI conditioning 10 Gy reduced local failure 50%→20%. ANBL0532 (Liu JCO 2020) — don't boost gross residual post-op. ANBL1531: if vertebral body touches 10 Gy line, include it in 18 Gy target to prevent asymmetric growth.

Rhabdomyosarcoma

Most common pediatric soft tissue sarcoma · ~400/yr · radiosensitive. FOXO1 fusion replaces embryonal/alveolar as the dominant prognostic variable.

Favorable (FOXO1 neg): embryonal, botryoid, spindle cell — more common, younger. Unfavorable (FOXO1 pos): 80% of alveolar; older (adolescent); worse EFS. FOXO1-neg alveolar behaves like embryonal.
Nodal risk by site: orbit 0–1% · extremity 10–15% · paratesticular 25–30%. LN sampling required for all extremity primaries and males >10 yo with paratesticular.
Special workup: parameningeal → CSF cytology + neuraxis MRI. H&N/PM planning: T1+Gad and T2 both needed to distinguish tumor from mucus.
GROUP — not stage — drives RT decisions. Group is defined before chemo.

Group	Definition	RT Dose
I	Localized, completely resected (FOXO1 neg, non-orbit)	0 Gy (FOXO1-); 36 Gy if FOXO1+
IIA	Gross resection, microscopic + margin	36 Gy
IIB/C	Resection with involved nodes	41.4 Gy
III (orbit)	Gross residual, orbital primary	45 Gy
III (non-orbit)	Gross residual disease	50.4 Gy
Whole lung (mets)	—	15 Gy / 10 fx (>6 yo); 12 Gy / 10 fx (<7 yo)

Stage vs Group: Stage is driven by SITE (Stage 1 = favorable sites any size). Group is driven by extent after initial surgery. Boards love to test this distinction.

ARST0531 / ARST1431 (Casey Cancer 2019; Jackson IJROBP 2025): Dose escalation 50.4 → 59.4 Gy does NOT improve local control for >5 cm tumors. Delayed primary excision (DPE) with dose reduction had LF 5% vs 22% with RT alone — consider DPE after induction chemo when feasible.

Target volumes (ARST0531): GTV1 = visible disease pre-therapy (cut back for pushing margins, not infiltrating). CTV1 = GTV1 + 1 cm (1.5 cm threshold low), anatomically confined. PTV 3–5 mm. Chemo starts first, local therapy at week 13.

Cyclophosphamide dose matters: IRS-IV used CPM 26.4 g/m² and had best LC. Subsequent trials with lower CPM show higher local recurrence even with same RT dose. RT cannot compensate for reduced chemo.

Ewing Sarcoma

~300/yr · bone or soft tissue · t(11;22) EWS-FLI1 · small round blue cell tumor · "onion skinning" periosteal reaction · fairly radiosensitive.

Outcomes: localized (AEWS1031, Leavey JCO 2021) — 5y EFS 78%, OS 87%. Metastatic (AEWS1221, Dubois JCO 2023) — 3y EFS 38%, OS 58%.
Chemo: VDC/IE — vincristine/doxorubicin/cyclophosphamide alternating with ifosfamide/etoposide. Drop the "A" (doxorubicin) during RT.
Local therapy: surgery, RT, or both — no RCTs. Surgery preferred for "expendable" bones (rib, fibula, clavicle). Pelvis and large (>200 mL) primaries have worse LC. Larger tumors → poorer EFS (not just LC).
Even localized, the main long-term threat is mets — 80% of initial failures are distant. Small differences in LC rates matter less than adequate systemic therapy.

Scenario	Dose
Definitive RT (or post-op gross residual)	55.8 Gy
Post-op microscopic + margin	50.4 Gy
R0 resection (no tumor at ink)	0 Gy
Pre-op RT (AEWS1031, rare)	36 Gy
Large tumor dose escalation (trial only)	up to 63 Gy
Lung mets	15 Gy / 10 fx (12 Gy / 8 fx if <6 yo)

Target volumes (AEWS1031): GTV1 = pre-chemo disease in soft tissue AND bone. CTV1 = GTV1 + 1 cm anatomically confined. GTV2 = pre-chemo bone + post-chemo soft tissue. CTV2 = GTV2 + 1 cm. Cone-down after 45 Gy. Concurrent IE (no doxorubicin).

PART II — CNS MALIGNANCIES

Medulloblastoma

500–700/yr · posterior fossa · median age 5–6 · high CSF dissemination potential · Clinical "Rules of 20": 20% of peds cancer = brain tumors, 20% of those = medullo, 20% of medullo pts die of disease, 20% are >14 yo.

Posterior fossa syndrome: 15–25% post-resection — mutism, dysphagia, truncal ataxia, hypotonia, mood lability, gaze palsy, rare respiratory failure.
WHO 2021 molecular subtypes: WNT (best), SHH, Group 3, Group 4.
Workup: MRI brain AND spine pre-op (avoid blood artifact), max safe resection, post-op MRI brain, LP 10–14 days post-surgery, methylation profiling.
M staging: M0 · M1 CSF+ · M2 intracranial beyond primary · M3 gross spinal seeding · M4 beyond CNS.
Start RT within 31–35 days of surgery. RT followed by 6–9 mo adjuvant chemo.

Risk Group	Criteria	5y EFS	RT Regimen
Average	≥3 yo, M0, residual <1.5 cm² (NOT cm³)	81% (ACNS0331)	CSI 23.4 Gy → boost 54 Gy
High	M+ OR residual ≥1.5 cm² OR diffuse anaplasia	60% (ACNS0332)	CSI 36 Gy → boost 54–55.8 Gy
Infant	<3 yo	65%	Typically chemo-focused ± delayed RT

Boost/mets doses: primary → 54 Gy (1.8 Gy/fx, CTV = tumor bed + 1 cm, reduce into brainstem if not invaded per ACNS1422). Diffuse posterior fossa mets → 54–55.8 Gy. Diffuse spine mets → 39.6 Gy. Focal drop mets: 45 Gy at cord level, 50.4 Gy below cord. Supratentorial mets require clinical judgment.

ACNS0331 lesson: CSI dose reduction 23.4 → 18 Gy was inferior in average-risk medulloblastoma. ACNS1422 (WNT pathway only, completed accrual Sept 2024) tested 18 Gy CSI — results pending, do not reduce CSI off protocol. WNT pts still need CSI (Gupta CCR 2022 — omission led to high LM failure).

CSI contouring: thecal sac inferior border usually S2–S3 (1.5 cm below sac on sagittal MRI). Cover nerve roots to lateral edge of vertebral bodies. In skeletally immature patients: target entire vertebral body at lower dose to prevent asymmetric growth (or full CSI bony spine per ACNS1422).

Ependymoma

~300/yr · median age 5–6 (many <3) · 2/3 infratentorial, 1/3 supratentorial · low CSF dissemination risk.

#1 prognostic factor: extent of resection. 7y EFS GTR 77% vs STR 34%. Grade has smaller effect (Grade II 79% vs Grade III 61%). 1q gain = bad.
Workup identical to embryonal tumors (MRI brain + spine pre-op → max safe resection → post-op brain MRI → LP 10–14 d later).
Watch for extension through foramen of Magendie/Luschka, jugular foramen, IAC, cavernous sinus when defining GTV.

Target	Definition / Dose
GTV	Residual disease + resection bed
CTV1	GTV + 0.5 cm (tighter than most CNS tumors); limit 3 mm into brainstem
Dose (CTV1)	54 Gy / 30 fx
Boost to GTV	+5.4 Gy (total 59.4 Gy); omit if <18 mo
Metastatic disease (≥3 yo)	CSI 36 Gy

ACNS0121 (Merchant JCO 2019): supratentorial Grade II s/p GTR — adjuvant RT better than observation. Grade II myxopapillary (spine) typically surgical; RT role controversial.

CNS Germ Cell Tumors

2–4% of peds brain tumors (up to 11% in Japan/Taiwan) · suprasellar 30–35%, pineal 50–60% · pineal M>F, suprasellar F>M · 60% germinomas in 2nd decade.

Presentation: suprasellar → DI, neuroendocrine, visual deficits. Pineal → hydrocephalus, Parinaud's syndrome (↓ upward gaze, Argyll-Robertson pupil, convergence nystagmus).
Tumor markers (CSF > serum): Germinoma — β-hCG may be up to 50–75 IU/L, but AFP is ALWAYS WNL. NGGCT — choriocarcinoma: ↑↑ β-hCG; yolk sac: ↑ AFP; embryonal: may have either.
Bifocal (suprasellar + pineal only, 5–10%): treat as localized.
Prefer diagnosis via tumor markers — surgery is morbid in both regions.

Entity	Regimen
Localized germinoma (ACNS1123 Stratum B, Bartels 2022)	4 cycles carbo/etop → if CR: WV 18 Gy + boost 12 Gy (total 30 Gy). If <CR: WV 24 Gy + boost to 36 Gy. ACNS2321 testing WV 12 + boost to 24 (protocol only).
Metastatic germinoma	CSI (cure rate still 90–95%)
NGGCT (ACNS2021 current)	6 cycles carbo/etop ± ifos/etop → WV + spine 30.6 Gy → boost primary to 54 Gy. 2nd-look surgery if <CR.

NGGCT field evolution: ACNS0122 CSI 36 + boost 54 (PFS 92%) → ACNS1123A WVI 30.6 + boost 54 (PFS 88%, but all failures in spine) → ACNS2021 added spine back (avoiding whole brain).

Watch for "growing teratoma": malignant component may respond by markers while teratomatous component progresses — needs resection for mass effect and to exclude malignant progression.

Craniopharyngioma

Arises from Rathke pouch remnants · cysts + calcifications radiographically · suprasellar presentation similar to GCT but with characteristic imaging.

Management tradeoff: GTR 70–85% DFS but huge morbidity risk (DI, vision loss, hypothalamic dysfunction — "when surgery goes wrong, it really goes wrong"). Biopsy/STR + RT 85–90% DFS, with RT risks (2nd malig, neurocog, stroke, cataract, panhypopit, hypothalamic obesity).
Targets: GTV = residual disease including cysts. CTV = GTV + 3–5 mm. PTV = CTV + 3–5 mm.
Dose: 52.2–54 Gy at 1.8 Gy/fx.
Re-image during treatment — cyst/tumor swelling during RT can threaten target coverage and vision. Pituitary dysfunction is nearly universal.

Diffuse Midline Glioma / DIPG

Functionally a GBM of brainstem or thalamus · H3K27M rearrangement signature · diagnosis often by imaging (biopsy morbid) · median survival 9–12 months.

RT is the only proven life-extending therapy. Start ASAP — patients can progress quickly.
Dose: 54 Gy / 30 fx to MRI extent + 1 cm CTV (anatomically confined) + 0.3–0.5 cm PTV. Outside North America: 39 Gy / 13 fx common.
Patients often clinically improve before later progressing.
At recurrence: clinical trial (several CAR-T) vs re-irradiation 20–30 Gy (Janssens EJC 2017).
Do not mistake DIPG for LGG — treatment paradigms are entirely different.

CROSS-CUTTING HIGH-YIELD POINTS

Wilms vs Neuroblastoma bedside differentiation: Wilms → well-appearing, mass moves with kidney, resect upfront. NB → acutely ill, raccoon eyes, mass moves separately, biopsy only.
RMS "Stage" vs "Group": Stage = site + TNM (pre-treatment); Group = surgical result (extent of residual). Group drives RT.
LP timing post-op: always 10–14 days (too early = blood contamination, false positive) — applies to embryonal tumors, ependymoma, CNS GCT.
CSI doses memorized: Average-risk medullo 23.4 · High-risk medullo / metastatic ependymoma / metastatic germinoma 36 · NGGCT WV+spine 30.6.
PENTEC (Red Journal June 2024) — 19 organ-specific pediatric normal-tissue analyses; emerging gold standard for pediatric OAR constraints.
Growth-preserving techniques: if vertebral body touches 10 Gy isodose, include it in 18 Gy target (prevent asymmetric growth) — applies to neuroblastoma and CSI planning.

KEY LANDMARK TRIALS (MEMORIZE)

Trial	Disease	One-line takeaway
AREN0533 (Dix 2018)	Wilms FH + lung mets	Omit WLI if CR to induction; EFS marginal, OS same.
CCG-3891 (Haas-Kogan 2003)	High-risk neuroblastoma	TBI conditioning ↓ local failure 50%→20%.
ANBL0532 (Liu 2020)	High-risk neuroblastoma	Don't boost gross residual post-op.
ARST0531 (Casey 2019)	RMS intermediate risk	50.4 → 59.4 Gy doesn't help LC in >5 cm tumors.
ARST1431 (Jackson 2025)	RMS	Delayed primary excision → LF 5% vs 22%.
AEWS1031 (Leavey 2021)	Localized Ewing	5y EFS 78%, OS 87%; larger primaries → worse EFS.
ACNS0331 (Michalski 2021)	Average-risk medullo	CSI 18 Gy inferior to 23.4; boost to tumor bed OK vs whole post fossa.
ACNS0332 (Leary 2021)	High-risk medullo	5y EFS 60%; concurrent carbo benefits Group 3.
ACNS1123B (Bartels 2022)	Localized germinoma	Chemo + WV 18 Gy + boost 12 Gy = 30 Gy total if CR.
ACNS1123A (Fangusaro 2020)	Localized NGGCT	WVI 30.6 + boost 54 (PFS 88%, spine failures).
ACNS0121 (Merchant 2019)	Supratentorial ependymoma	Adjuvant RT > observation after GTR.
ACNS0831	Intracranial ependymoma	Adjuvant chemo role established then questioned.