Radiation Oncology/Medulloblastoma/Therapy

Medulloblastoma Therapy

Treatment Overview
Surgery
 * Major goals of surgery are to achieve an oncologic resection and to restore normal flow of CSF
 * Complete resection (GTR/R0) results in significantly better prognosis
 * Insufficient as single modality (1/61 patients survived 3 years in Cushing's original paper)

Radiation
 * Radiation therapy is delivered to the entire craniospinal axis (deferred if <3 years old).
 * Dose of RT based upon risk stratification grouping:
 * High risk patients treated with 36 Gy CSI (craniospinal irradiation) with concurrent vincristine
 * Standard risk patients treated with 23.4 Gy CSI with concurrent vincristine
 * Posterior fossa boost to 54-56 Gy

Chemotherapy
 * Chemotherapy is standard adjuvant therapy for all risk groups
 * Chemotherapy can also be given for younger patients in order to delay RT as the toxicity profile for patients <3 y/o who get CSI is worse than for older children.

Radiation doses (full/fraction) and chemotherapy regimes (pre/post, agents) are varied in concrete protocols

Overview
 * SIOP I and CCG 942 demonstrated no benefit of adding chemo to standard CSI 36 Gy, though on subgroup analysis advanced stage did benefit
 * Standard risk group
 * French SFOP group attempted to reduce RT volume by not irradiating supratentorium, but results very disastrous. <20% EFS at 6 years, with 9/13 failures in supratentorium. CSI is necessary, despite long-term toxicity
 * Based on these results, POG 8631/CCG 923 attempted to lower CSI alone dose to 23.4 Gy, but had to be stopped early after showing higher failure rate (67% vs. 52%)
 * SIOP II in standard risk patients also attempted to lower CSI dose from 35 Gy to 25 Gy, and add adjuvant chemo. Neither worked, with patients getting post-op adjuvant chemo and low dose CSI doing particularly badly. High risk patients were randomized to +/- adjuvant chemo, standard 35 Gy CSI and additional chemo. Adjuvant chemo didn't work here either. 5-year EFS 68% for standard CSI
 * Most standard risk patients were nevertheless treated with adjuvant chemotherapy despite evidence of benefit. PNET 3 study published in 2006 finally demonstrated benefit to EFS (5-year 74% vs. 60%), but no difference in OS. Unfortunately, it also demonstrated a significantly worse health status in survivors treated with CSI + adjuvant chemo compared to CSI alone
 * German trial HIT 91 evaluated post-op chemo -> RT vs. post-op RT -> chemo. At 3 years, EFS was better for immediate RT followed by chemo
 * Meanwhile, CCG 9892 was a single-arm pilot study in standard risk patients evaluating concurrent vincristine and low dose CSI 23.4 Gy, followed by CCNU/vincristine/cisplatin. 80% PFS at 5 years was as good as historical control with CSI 35 Gy
 * Based on these results, low dose CSI 23.4 Gy with concurrent vincristine became standard in CCG study A9961, and randomization was between two different chemo arms (CCNU-based vs. cyclophosphamide-based). 5-year EFS was 81%, and there was no difference between the arms
 * High risk group
 * Since both SIOP I and CCG 942 demonstrated benefit for chemo in advanced stage disease, efforts have been aimed at determining correct sequence and drugs
 * German trial HIT 91 evaluated post-op chemo -> RT vs. post-op RT -> chemo. At 3 years, EFS was better for immediate RT followed by chemo(PMID: 10661332 )
 * POG 9031 similarly randomized patients to pre-RT vs. post-RT chemotherapy. Results at 2 years reported in abstract form showed no difference in outcome
 * Infants (<3 years old)
 * Baby POG I trial attempted to delay RT by 1-2 years of intensive chemotherapy. In children with GTR, CSI of 24 Gy and PF boost of 50 Gy was used. 5-year EFS for this subgroup was 69%, and comparable to standard risk older children. However, many children failed early, with overall 5-year EFS 32% and OS 40%
 * At the same time, CCG tried a different approach with 8 drugs in one day (8-in-1), and delaying RT. This approach did not work as well as Baby POG I, probably due to lesser intensity chemo
 * Two trials, a German and French, have both evaluated eliminating RT altogether in young patients. They have both shown it is possible in children with gross total resection (GTR/R0), without metastatic disease. The neuropsychological outcomes of these children were favorable when compared to those receiving RT
 * to be continued...

CSI +/- chemo

 * CCG A9961 (1996-2000) PMID 16943538 "Phase III study of craniospinal radiation therapy followed by adjuvant chemotherapy for newly diagnosed average-risk medulloblastoma" Packer et al.  JCO 24(25):4204-8, 2006.
 * Randomized. 421 patients, standard risk. Treated with reduced dose RT (23.4 Gy CSI + 55.8 Gy to posterior fossa) and concurrent vincristine, and then randomized to adjuvant:  1) CCNU, cisplatin and vincristine or 2) cyclophosphamide, cisplatin, vincristine. Median F/U 5 years
 * 5-year outcome: EFS 81%, OS 86%. No difference by chemo arm
 * Prognostic factors: anaplasia OS 75% vs. 89% (SS)
 * Relapse: posterior fossa alone 32%, disseminated alone 40%, PF+disseminated 25%
 * Toxicity: 25% with delayed onset mutism, hypotonia, cerebellar deficits, supranuclear CN deficits, extreme irritability, and/or emotional lability. ~50% residual deficits at 1 year
 * Conclusion: Reduced dose CSI with chemo can be done safely to avoid toxicities of higher dose RT


 * PNET 3 (1992-2000)
 * Study closed early due to lack of accrual in RT alone arm Randomized. 179/217 standard risk patients (including M1). Treated with neoadjuvant chemo (vincristine/etoposide/carboplatin/cyclophosphamide) vs. RT alone (CSI 35 Gy + 20 Gy posterior fossa boost). Median F/U 5.4 years.
 * 2003 PMID 12697884 -- "Results of a randomized study of preradiation chemotherapy versus radiotherapy alone for nonmetastatic medulloblastoma: The International Society of Paediatric Oncology/United Kingdom Children's Cancer Study Group PNET-3 Study." (Taylor RE, J Clin Oncol. 2003 Apr 15;21(8):1581-91.)
 * 5-year outcome: OS 70% (NS), EFS chemo-RT 74% vs. RT alone 60% (SS)
 * Conclusion: improved EFS with addition of chemo, no impact on survival
 * 2007 PMID 17878477 -- "Reduction of health status 7 years after addition of chemotherapy to craniospinal irradiation for medulloblastoma: a follow-up study in PNET 3 trial survivors on behalf of the CCLG (formerly UKCCSG)" (Bull KS, J Clin Oncol. 2007 Sep 20;25(27):4239-45.)
 * Prospective analysis. 73% of patients assessed. Evaluated health status (HUI3 Index: vision, hearing, speech, ambulation, dexterity, emotion, cognition and pain), behavior, and quality of life. Mean F/U 7.2 years
 * Outcome: health status significantly worse in CSI+chemo arm, trend to poorer outcomes in behavior and quality of life. Also significantly worse physical restrictions, with significantly more therapeutic and educational support
 * Conclusion: Addition of chemo to CSI leads to significant decrease in health status


 * CCG 9892 PMID 10561268 "Treatment of children with medulloblastomas with reduced-dose craniospinal radiation therapy and adjuvant chemotherapy" Packer et al.  JCO 17(7):2127-36, 1999
 * Pilot study evaluating reduced dose (23.4 Gy) CSI w/ concurrent vincristine as alternative to 35 Gy for standard risk patients. Post fossa boosted to 55.8
 * PFS 86% at 3 yrs, 79% at 5 yrs
 * This study served as the basis for CCG A9961 reduced CSI dose


 * SIOP II (1984-1989) PMID 7623725 "Prospective randomised trial of chemotherapy given before radiotherapy in childhood medulloblastoma. International Society of Paediatric Oncology (SIOP) and the (German) Society of Paediatric Oncology (GPO)" Bailey et al.  Med Pediatr Onc 25(3):166-78, 1995
 * Randomized. 364 patients. SIOP I low risk (total/subtotal resection, no brain stem invasion, M0) randomized to +/- adjuvant chemo (vincristine, methotrexate, procarbazine), and then randomized to standard CSI 35 Gy vs. reduced CSI 25 Gy. Boost posterior fossa to 55 Gy. High risk randomized to +/- adjuvant chemo, then standard CSI and additional post-RT chemo
 * Outcome: No advantage to pre-RT chemo. Standard CSI 35 Gy for low risk pts increased EFS (68% vs 55%). Low risk patients with adjuvant chemo and CSI 25 Gy did particularly badly
 * Conclusion: No benefit to adjuvant chemo. Standard dose of CSI is 35 Gy


 * SIOP I PMID 2141512 -- "Adjuvant chemotherapy for medulloblastoma: the first multi-centre control trial of the International Society of Paediatric Oncology (SIOP I)." (Tait DM, Eur J Cancer. 1990 Apr;26(4):464-9.)
 * Randomized. 286 patients, 15 countries. Treated with CSI, randomized to +/- chemo (concurrent vincristine, then CCNU/vincristine maintenance)
 * Survival: 5-year 53%, 10-year 45%. No difference between arms
 * Subgroup benefit: subtotal surgery, brainstem involvement, T3-T4 disease


 * CCSG 942 PMID 2319316 -- "The treatment of medulloblastoma. Results of a prospective randomized trial of radiation therapy with and without CCNU, vincristine, and prednisone." (Evans AE, J Neurosurg. 1990 Apr;72(4):572-82.)
 * Randomized. 233 patients. CSI +/- chemo (CCNU, vincristine, prednisone)
 * 5-year outcomes: EFS chemo-RT 59% vs. RT alone 50% (NS); OS 65% for both (NS)
 * Subgroup benefit: large tumors (EFS chemo-RT 48% vs. RT alone 0%)
 * Conclusion: no benefit for chemo in low stage, subgroup benefit for advanced stage

CSI + adjuvant chemo alternatives

 * CCG 921 (1986-1992) -- vincristine/lomustine/prednisone (VCP) vs. 8-in-1
 * Randomized. 427 children <21 years, with medulloblastoma, pineoblastoma, ependymoblastoma, central neuroblastoma, PNET, or malignant ependymoma, with unfavorable features. For ST-PNET required M+ staging. Children age >1.5 years (n=328) received post-op CSI with Arm A) vincristine, lomustine, prednisone vs. Arm B) 8-in-1 (cisplatin, procarbazine, lomustine, vincristine, cyclophosphamide, methylprednisolone, hydroxyurea, cytarabine). RT: age >3 received CSI 36 Gy, boost to 50.4-54 Gy spine mets and 54 Gy primary brain site; age <3 received CSI 23.4 Gy with boost to 45 Gy. Children age <1.5 years (n=99) were not randomized and received only Arm B
 * 1999 PMID 10071274 -- "Metastasis stage, adjuvant treatment, and residual tumor are prognostic factors for medulloblastoma in children: conclusions from the Children's Cancer Group 921 randomized phase III study." (Zeltzer PM, J Clin Oncol. 1999 Mar;17(3):832-45.) Median F/U 7.0 years
 * Subset analysis. 203 patients with medulloblastoma. Median OS 8-in-1 55% vs. VCP 54% (NS); PFS 63% vs. 45% (SS)
 * Prognostic factors: age <3 (who received lower CSI dose); if >3 y/o then M stage (PFS M0 70% vs. M1 57% vs. M2+ 40%, SS); if M0 tumors, then residual (PFS <1.5 cm2 78% vs. >1.5 cm2 54%)
 * Conclusion: VCP + XRT superior; if <3 years (with reduced RT) had lowest survival

Pre-RT vs. Post-RT Chemo

 * POG 9031 (1990-1996)
 * Randomized. 226 patients with high risk disease (residual >1.5 cm3, T3b or T4, or M1-M3). Treated with upfront chemo x3 cycles (cisplatin/etoposide) followed by CSI followed by chemo (vincristine/cyclophosphamide) vs. upfront CSI, followedy by cisplatin/etoposide, followed by vincristine/etoposide. RT CSI 35.2/22 + boost PF 18/10 Gy (53.2 Gy) if M0-M1 and CSI 40.0/25 + boost PF 14.4/8 Gy (54.4 Gy) if M2-M3. Spinal or brain mets boosted to 44.8 Gy
 * RT QA, 2006 PMID 16413699 -- "Radiotherapy in pediatric medulloblastoma: quality assessment of Pediatric Oncology Group Trial 9031." (Miralbell R, Int J Radiat Oncol Biol Phys. 2006 Apr 1;64(5):1325-30.)
 * 5-year outcome: EFS 68%, OS 75%
 * Major deviations: brain 26%, spine 7%, posterior fossa boost 40%, primary tumor bed 17%
 * Conclusion: no correlation between major treatment deviations and negative outcome
 * 2000 ASTRO Abstract 134 -- "Outcome for children with high stage medulloblastoma: Results of the pediatric oncology group 9031" (Tarbell NJ, International Journal of Radiation Oncology*Biology*Physics, Volume 48, Issue 3, Supplement 1, 2000, Page 179). Median F/U 4.5 years
 * 2-year outcome: EFS pre-RT chemo 78% vs. post-RT chemo 80%; OS 83% vs. 90% (NS)
 * Conclusion: No difference at 2 years; OS favorable compared to historical controls


 * German HIT 88/89 and HIT 91 (1991-97)
 * Pilot (HIT 88/89) followed by randomized (HIT 91 PMID 10661332 ). Trial enrolled pts with supratentorial PNET, medulloblastoma, and anaplastic ependymomas. All treated with extensive resection. In HIT 88/89, pts treated postoperatively with chemo x2 cycles (Ifosfamide, etoposide, MTX, cisplatin, cytarabine) -> RT (CSI 35.2 Gy + PF boost 20 Gy). In HIT 91, randomized to 1) chemo -> RT (same as HIT 88/89) or 2) immediate RT (35.2 + 20) followed by maintenance chemo x8 cycles (CCNU, cisplatin, vincristine).
 * Medullo prognostics, 2007 PMID 17473196 -- "Prognostic relevance of clinical and biological risk factors in childhood medulloblastoma: results of patients treated in the prospective multicenter trial HIT'91." (Rutkowski S, Clin Cancer Res. 2007 May 1;13(9):2651-7.)
 * Formalin-fixed paraffin-embedded tumor samples analyzed, c-myc DNA, N-myc DNA, c-myc mRNA, and trkC mRNA
 * Risk groups identified based on trkC mRNA and c-myc mRNA
 * Favorable group (incl. 2 patiens M+): high trkC and low c-myc mRNA: 7-year EFS 100%
 * Intermediate group: everyone not in favorable/unfavorable - 7-year EFS 65%
 * Unfavorable group: M+ and low trkC and high c-myc mRNA: 7-year EFS 33%
 * Medullo outcomes, 2000 PMID 10661332 -- "Postoperative neoadjuvant chemotherapy before radiotherapy as compared to immediate radiotherapy followed by maintenance chemotherapy in the treatment of medulloblastoma in childhood: results of the German prospective randomized trial HIT '91." (Kortmann RD, Int J Radiat Oncol Biol Phys. 2000 Jan 15;46(2):269-79.)
 * Subset report of 158 medulloblastoma patients (137 randomized). Median F/U 2.5 years
 * 3-year outcome: RFS RT->chemo 78% vs. chemo->RT 65% (SS), but no difference if age 3-6 years
 * Negative prognosis: M2/3 disease, age <8 years. M1 not a bad prognostic factor
 * Conclusion: RT upfront with maintenance chemo more effective. Neoadjuvant chemo caused higher RT myelotoxicity, and more RT interruptions
 * Pilot HIT 88/89 PMID 9743957 -- "Preradiation chemotherapy of children and young adults with malignant brain tumors: results of the German pilot trial HIT'88/'89." (Kuhl J, Klin Padiatr. 1998 Jul-Aug;210(4):227-33.)
 * RT delayed median 23 weeks
 * 5-year OS: medulloblastoma 57%, ependymoma 62%, malignant glioma 36%, ST PNET 30%
 * Conclusion: Well tolerated and efficacious. Based on this, HIT 91 developed

Chemotherapy to Delay XRT

 * CCG 921 (1986-1992) -- vincristine/lomustine/prednisone vs. 8-in-1
 * Randomized. 427 children <21 years, with medulloblastoma, pineoblastoma, ependymoblastoma, central neuroblastoma, PNET, or malignant ependymoma, with unfavorable features. For ST-PNET required M+ staging. Children age >1.5 years (n=328) received post-op CSI with Arm A) vincristine, lomustine, prednisone vs. Arm B) 8-in-1 (cisplatin, procarbazine, lomustine, vincristine, cyclophosphamide, methylprednisolone, hydroxyurea, cytarabine). RT: age >3 received CSI 36 Gy, boost to 50.4-54 Gy spine mets and 54 Gy primary brain site; age <3 received CSI 23.4 Gy with boost to 45 Gy. Children age <1.5 years (n=99) were not randomized and received only Arm B
 * Infants; 1994 PMID 8040673 &mdash; "Survival of infants with primitive neuroectodermal tumors or malignant ependymomas of the CNS treated with eight drugs in 1 day: a report from the Childrens Cancer Group." (Geyer JR et al. J Clin Oncol. 1994 Aug;12(8):1607-15.)
 * Infant (<1.5 year) subset of 82 children (46 posterior fossa, 15 ependymoma, 11 nonpineal ST-PNET, 8 pineoblastoma, 2 ATRT. Received Arm B only (8-in-1), RT omitted in most cases
 * 3-year PFS: ST PNET 55% vs. medullo 22% vs. ependymoma 22% vs. pineoblastoma 0%. Median time-to-progression 6 months
 * Conclusion: OS is poor, a subset remained disease-free with chemo only
 * Comment: Results inferior compared to Baby POG, probably due to less intensive chemo, but ~20% long-term survivors with chemo only
 * Infants failures; 2005 PMID 16007595 -- "Patterns of treatment failure in infants with primitive neuroectodermal tumors who were treated on CCG-921: a phase III combined modality study." (Hong TS, Pediatr Blood Cancer. 2005 Oct 15;45(5):676-82.)
 * Infant (<1.5 years) subset of 65 patients (19 supratentorial, 46 posterior fossa.)
 * Outcome: 5-year relapse rate M0 64% vs. M+ 71%. ST-PNET subset M0 64% vs. M+ 100%. Relapses within 2 years
 * Conclusion: Despite aggresive chemo, infants have high rates of treatment failure


 * Baby POG I (1986-90)
 * Prospective. 198 children < 3 yrs (132 < 2 yrs, 66 age 2-3 yrs), bx proven malignant brain tumors (low-grade astro excluded), treated with maximal surgery, postop chemo (CTX/VCR followed by cis/etopo) for 2 yrs (if age < 2 at dx) or 1 yr (age 2-3) or until disease progression, followed by RT.
 * Histologies: medulloblastoma 31%, ependymoma 24%, PNET 18%, malignant glioma 9%, brain-stem glioma 7%, other 9%. 27% M+. GTR in 38% of cases.
 * RT for medullo, PNETs, anaplastic ependymoma, or subarachnoid seeding CSI 35.2 Gy + boost to primary to 54 Gy. RT for ependymoma, gliomas local to 54 Gy. If no residual disease after chemo, reduced RT to CSI 24 Gy and primary site 50 Gy. Infants <2 years 90% of dose
 * 1993 PMID 8388548 -- "Postoperative chemotherapy and delayed radiation in children less than three years of age with malignant brain tumors." (Duffner et al., N Engl J Med.  1993; 328(24):1725-31)
 * Response rate: best in medulloblastoma, glioma, ependymoma. Poor/no response in brain-stem gliomas and embryonal tumors
 * CR to chemo PFS comparable those with GTR PFS
 * Conclusion: Chemo effective in infancy. If GTF, chemo alone provided excellent control. No clinically important neurotoxicity. However, cannot recommend for embryonal tumors
 * 10-years, 1999 PMID 11554387 &mdash; "The treatment of malignant brain tumors in infants and very young children: an update of the Pediatric Oncology Group experience." (Duffner PK et al. Neuro-oncol. 1999 Apr;1(2):152-61.)
 * Medullo: GTR 38%, 62% R+; of these 48% CR/PR to chemo. 5-year PFS 32%, OS 40%. Progression typically <6 mo, none >2 years. No difference if RT delayed by 1 or 2 years
 * Medullo: For children with GTR and M0, treated with CSI 24 Gy + 30 Gy PF boost. 5-year OS 69% (comparable to older standard risk children)
 * Medullo conclusion: chemo allowed for a delay in RT in children with GTR and possibly CR to chemo, which allows for avoidance of the toxicity of irradiating CNS in very young


 * CCG, 8-in-1
 * Children < 18 months. Included medulloblastoma (56%), ependymoma (18%), ST PNET (12%), pineoblastoma (10%), ATRT (2%). Eight drugs in 1 day: vincristine, carmustine, procarbazine, hydroxyurea, cisplatin, cytarabine, prednisone, and cyclophosphamide. Planned for delayed RT but was omitted in most cases.
 * 1994: PMID 8040673 &mdash; "Survival of infants with primitive neuroectodermal tumors or malignant ependymomas of the CNS treated with eight drugs in 1 day: a report from the Childrens Cancer Group." Geyer JR et al. J Clin Oncol. 1994 Aug;12(8):1607-15.
 * 3-year PFS: ST PNET 55% vs. medullo 22% vs. ependymoma 22% vs. pineoblastoma 0%. Median time-to-progression 6 months
 * Results inferior compared to Baby POG, probably due to less intensive chemo, but ~20% long-term survivors with chemo only


 * St. Jude's Patterns of Failure Study PMID 9300735, Hartsell et al.  "Patterns of failure in children with medulloblastoma: effects of preirradiation chemotherapy"  Int J Radiat Oncol Biol Phys.  1997; 39(1): 15-24
 * 53 pts tx'd w/ pre-RT chemotherapy (if <3, received chemo then delayed CSI; if >3 w/ advanced dz, received chemo followed by CSI)
 * 23% rate of failure w/ delayed xrt; site of failure most commonly posterior fossa; salvage possible in 50% of failures
 * Conclusion: In M0 pts who get chemo followed by delayed CSI, risk of neuraxis failure increases w/ duration of chemo.

Postoperative chemotherapy alone

 * German Pediatric Brain Tumor Study Group, 2005 (1992-1997) - PMID 15758008 &mdash; "Treatment of Early Childhood Medulloblastoma by Postoperative Chemotherapy Alone." (Rutkowski S et al. New Engl J Med 352:978-986, 2005. )
 * Prospective. 43 children <3, 28% had M2-M3 disease. Treated at 31 different centers. Treatment was maximum surgical removal of the tumor, with confirmation by MRI or CT. Intraventricular reservoir (Ommaya) for MTX. Chemotherapy started in 2-4 weeks, consisting of three 2-month cycles of cyclophosphamide, methotrexate (intraventricular and intravenous) with leucovorin, vincristine, carboplatin, and etoposide, divided into four administrations per cycle, with one week between each cycle. Methotrexate was intraventricular and intravenous. Treatment stopped after 3rd cycle. If not a complete response, then went on to XRT or other treatment.
 * 5-yr OS and PFS was 66% and 58%; GTR 93% and 82% vs. STR 56% and 50% vs. M2/M3 38% and 33%.
 * Remission (no RT): 14/17 (82%) with GTR; 21/31 (68%) with M0-M1; 3/12 (25%) with M2-M3 . Response rate to chemo was 62% in pts with measurable disease. Higher general intelligence in pts who did not receive XRT (compared to pts from prior trials).
 * Conclusion: Post-op chemo alone is promising in young children without mets


 * Baby Brain French Society of Paediatric Oncology, SFOP, 2005 (1990-2002) - PMID 16054568 &mdash; "Treatment of medulloblastoma with postoperative chemotherapy alone: an SFOP prospective trial in young children." Grill J et al. Lancet Oncol. 2005 Aug;6(8):573-80.
 * 79 pts. Age < 5. Combination chemotherapy (does not include MTX), 3 courses in 7 cycles. Carboplatin, procarbazine, etoposide, cisplatin, vincristine, cyclophosphamide. Classified as R0M0 (no residual disease, no mets), R1M0 (radiologic residual disease) or RXM+ (mets).
 * Median f/u 6.8 yrs. For R0M0 : 5-yr OS 73%, PFS 29% if gross total section and 0% for subtotal resection. For R1M0 : 5-yr OS 41%, PFS 6%. For RXM+ : 5-yr OS 13%, PFS 13%.
 * Post-op chemotherapy alone can be used for children with complete resection but not those with metastatic disease or incompletely resected tumor.

RT Technique

 * CSI
 * Daily fractions: 1.6-1.8 | 2*1.0 Gy, 5 days a week up to 18-40 Gy (18-25 Gy for AR only)
 * Posterior Fossa boost
 * Daily fractions: 1.8-2.0 | 2*1.0 Gy, 5 days a week up to 54-60 Gy | 36-40 & 55-68 (tumor bed) Gy
 * Upper border 1 cm above the mid-point between the occiput and vertex
 * Anterior border at the posterior clinoid
 * Posterior border at the inner table of the skull
 * Inferior border at the C2-C3 junction
 * Plain film determination:
 * Line drawn from foramen magnum to skull vertex is bisected with a perpendicular line
 * A point X is identified on the perpendicular line at one-third of the distance from the back of the skull
 * A line that connects the posterior clinoid process, X, and the internal occipital protuberance defines the superior border of the posterior fossa
 * Extent should be determined on MRI; plain films estimates as above not very good
 * Posterior fossa may not require full boost; a multi-institutional prospective study showed 5% PF failure rate with CSI 23.4 Gy, PF 36 Gy, and tumor bed boost 55.8 Gy
 * RT should be delivered in <=50 days (limit in PNET-3 protocol), otherwise negative impact on EFS and OS
 * Adjuvant vincristine chemotherapy may be administered during radiotherapy for a total of eight doses (weekly)


 * COG ACNS0331 (2004-2014) -- Full posterior fossa vs IFRT
 * Randomized. 549 patients. Average-risk MB (no anaplastic, no dissemination, <1.5 cm2 residual). 1A) Age 3-7: CSI 23.4 Gy (standard) vs 18 Gy (low). 1B) Age 7-21: CSI 23.4 Gy. 2) Everyone: followed by posterior fossa RT 54 Gy vs IFRT 54 Gy. Weekly vincristine with RT, followed by maintenance chemotherapy x 9 cycles
 * 2021 PMID 34110925 -- "Children's Oncology Group Phase III Trial of Reduced-Dose and Reduced-Volume Radiotherapy With Chemotherapy for Newly Diagnosed Average-Risk Medulloblastoma" (Michalski JM, J Clin Oncol. 2021 Aug 20;39(24):2685-2697. doi: 10.1200/JCO.20.02730. Epub 2021 Jun 10.)
 * Outcome: 5-year EFS IFRT 82% vs PFRT 80% (NS), low CSI 71% vs standard CSI 83% (SS, HR 1.7). Some molecular differences
 * Toxicity: Greater late IQ decline in standard CSI (p=0.02)
 * Conclusion: Reducing PF boost volume is safe in average-risk MB. Reducing CSI dose in young children results in inferior outcomes


 * Multi-Institutional; 2008 (1996-2003) PMID 17892918 -- "Multi-institution prospective trial of reduced-dose craniospinal irradiation (23.4 gy) followed by conformal posterior fossa (36 gy) and primary site irradiation (55.8 gy) and dose-intensive chemotherapy for average-risk medulloblastoma." (Merchant TE, Int J Radiat Oncol Biol Phys. 2008 Mar 1;70(3):782-7. Epub 2007 Sep 24.)
 * Prospective. 4 institutions in US and Australia. 86 patients, average-risk MB, treated with risk-adapted RT. RT technique: CSI 23.4 Gy, conformal posterior fossa RT 36 Gy, primary site (GTV + 2cm + 0.5cm) RT 55.8 Gy. Chemo (cyclophosphamide, cisplatin, vincristine) 6 weeks after RT x4 cycles. Median F/U 5.1 years
 * Outcome: 5-year EFS 83%, posterior fossa failure 5%
 * RT dosimetry: 13% reduction in volume of PF receiving >55 Gy; reductions to temporal lobes, cochleae, hypothalamus significant
 * Conclusion: Irradiation of less than entire PF results in disease control comparable to that after treating entire PF


 * St. Bartholomew's, 2004 (UK) PMID 15447964 -- "A comparison of conventional, conformal and intensity-modulated coplanar radiotherapy plans for posterior fossa treatment." (Breen SL, Br J Radiol. 2004 Sep;77(921):768-74.)
 * Goal: sparing of cochlea to reduce ototoxicity, particularly when RT given concurrently with cisplatin. 2D (opposed laterals) vs. 3D-CRT (2 wedged posterior obliques) vs. IMRT (4F coplanar)
 * Conclusion: 3D-CRT superior to both 2D and IMRT


 * POG 8631/CCG 923 (1986-1990) -- CSI 36 Gy vs. CSI 23.4 Gy
 * 2000, PMID 10944134 -- "Low-stage medulloblastoma: final analysis of trial comparing standard-dose with reduced-dose neuraxis irradiation." (Thomas PR, J Clin Oncol. 2000 Aug;18(16):3004-11.)
 * Randomized. 126 patients, low risk (>3 years, complete resection T1-T2, later also T3a, M0) standard CSI 36(20*1.8) vs. reduced CSI 23.4(13*1.8); both followed by posterior fossa boost to 54(30*1.8) Gy. Study closed prematurely due to high relapse in reduced arm
 * 5-year EFS: standard CSI 67% vs. reduced CSI 52% (p=0.080) 8-year EFS: standard CSI 67% vs. reduced CSI 52% (p=0.141) These data confirm the original one-sided conclusions but suggest that differences are less marked with time
 * Conclusion: Reduced CSI 23.4 alone is insufficient; may need concurrent chemo


 * SIOP II (1984-1989) -- CSI 35 Gy vs. CSI 25 Gy; also +/- adjuvant chemo
 * Randomized. 364 patients. SIOP I low risk (total/subtotal resection, no brain stem invasion, M0) randomized to +/- adjuvant chemo (vincristine, methotrexate, procarbazine), and then randomized to standard CSI 35 Gy vs. reduced CSI 25 Gy. Boost posterior fossa to 55 Gy. High risk randomized to +/- adjuvant chemo, then standard CSI and additional post-RT chemo
 * 1995 PMID 7623725 "Prospective randomised trial of chemotherapy given before radiotherapy in childhood medulloblastoma. International Society of Paediatric Oncology (SIOP) and the (German) Society of Paediatric Oncology (GPO)" Bailey et al.  Med Pediatr Onc 25(3):166-78, 1995
 * Outcome: No advantage to pre-RT chemo. Standard CSI 35 Gy for low risk pts increased EFS (68% vs 55%). Low risk patients with adjuvant chemo and CSI 25 Gy did particularly badly
 * Conclusion: No benefit to adjuvant chemo. Standard dose of CSI is 35 Gy
 * RT Review, 2004 PMID 15001263 -- "Impact of radiotherapy parameters on outcome in the International Society of Paediatric Oncology/United Kingdom Children's Cancer Study Group PNET-3 study of preradiotherapy chemotherapy for M0-M1 medulloblastoma." (Taylor RE, Int J Radiat Oncol Biol Phys. 2004 Mar 15;58(4):1184-93.)
 * Assessed accuracy of cribriform fossa, skull base, and PF field in 179/217 patients. Planning films reviewed in 131/179 patients.
 * Target deviation: 1) cribriform fossa: <3-mm margin between shielding edge and cribriform fossa; 2) skull base: <8-mm margin between shielding edge and skull base; and PF: >5-mm discrepancy between field edge and protocol definition.
 * 3-year outcome: OS RT <=50 days 84% vs. RT >50 days 71% (SS), EFS 78% vs. 54% (SS)
 * RT QA: 44% had 1+ deviations. PF recurrence in 34% with targeting deviation (and 54% with sup/ant/post deviation) vs. 16% without (SS)
 * Conclusion: RT should be delivered <50 days, and attention to PF field placement is important


 * U Penn, 1995 PMID 7753996 -- "Posterior fossa: analysis of a popular technique for estimating the location in children with medulloblastoma." (Solit DB, Radiology. 1995 Jun;195(3):697-8.)
 * 10 patients. Compared standard method for identifying posterior fossa on plain films vs. MRI
 * Conclusion: Underestimated volume in all 10 cases in some areas, but also significantly overestimated volume in different areas


 * SFOP M4, 1992 PMID 1512166 -- "M4 protocol for cerebellar medulloblastoma: supratentorial radiotherapy may not be avoided." (Bouffet E, Int J Radiat Oncol Biol Phys. 1992;24(1):79-85.)
 * Prospective. Surgical resection and risk-adapted post-op chemo ("8-in-1" + MTX). RT only to posterior fossa (54 Gy) and spinal axis (36 Gy), supratentorium spared. Mean F/U 6 years
 * 6-year outcome: DFS 18%; progression in 9/13 supratentorial
 * Conclusion: Entire neuroaxis (CSI) needs to be treated, despite long-term toxicity

Toxicity considerations

 * Also please see Pituitary toxicity


 * Surgery: Posterior Fossa Syndrome: difficulty swallowing, mutism, truncal ataxia, emotional lability (occurs 12-24 hours postop; can take months and RT should not be delayed)
 * Chemo: infertility, bone marrow tolerance, ototoxicity, poor CNS penetration
 * RT: ototoxicity; endocrine and neurocognitive dysfunctions (long-term effects of CSI)
 * Full posterior fossa RT results in 5 point IQ drop per year for <7 year old and -0.8 points for >7 year old
 * The supratentorial radiation dose is the principal risk factor associated with impaired intellectual outcome. An approximate 8-point per year decline for children treated with 36 Gy. One can forecast final IQ score based on the initial IQ score, dose of irradiation, and age at time of irradiation PMID 1517781
 * Even with the reduced radiation dose side effects can still be substantial. An average loss in IQ scores of over 4 points per year for the first 3 years after treatment by 23.4 Gy. In younger children (or those with a higher baseline IQ score at the start of treatment) this loss was even more severe PMID 11481352
 * The dose of radiotherapy applied to the brain strongly influences later verbal and non-verbal skills in children with medulloblastoma. Verbal fluency, immediate word list recall, block design, and fine motricity of the dominant hand were significantly lower in children irradiated at the standard doses than in those irradiated at reduced doses PMID 11104345
 * Neurocognitive development and outcome of children with cerebellar tumors diagnosed in infancy is very positive among those who were treated with surgery and chemotherapy. However, those who received CRT as part of their treatment are likely to have neurocognitive and psychosocial deficits that require remediational interventions PMID 10550145

Adult
More rarely, medulloblastoma may present in older patients.

PMID 16189725

References:

1. Levin VA, Vestnys PS, Edwards MS, et al.: Improvement in survival produced by sequential therapies in the treatment of recurrent medulloblastoma. Cancer 51(8): 1364-1370, 1983.

2. Carrie C, Lasset C, Alapetite C, et al.: Multivariate analysis of prognostic factors in adult patients with medulloblastoma: retrospective study of 156 patients. Cancer 74(8): 2352-2360, 1994.

3. Allen JC, Bloom J, Ertel I, et al.: Brain tumors in children: current cooperative and institutional chemotherapy trials in newly diagnosed and recurrent disease. Seminars in Oncology 13(1): 110-122, 1986.