A phase II trial of proton radiotherapy for medulloblastoma: Preliminary results

Reporter: Arpi Thukral, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: June 6, 2010

Presenter: T. I. Yock, Massachusetts General Hospital, Boston, MA; Children's Hospital Medical Center, Boston, MA; Harvard Medical School, Boston, MA


  • Medulloblastoma (MB) is the most common malignant brain tumor occurring in children, accounting for 20% of all pediatric CNS tumors. It commonly arises in the posterior fossa, and has tendency toward subarachnoid dissemination.
  • Treatment for MB typically involves surgical resection of tumor, chemotherapy and radiation.
    • Craniospinal irradiation (CSI) followed by a boost to the tumor bed and/or posterior fossa is the current standard of care.
    • Most institutions have adopted an intensity-modulated radiotherapy (IMRT)-based photon therapy approach for the partial brain elements of radiotherapy for MB.
  • Radiation therapy for medulloblastoma has been shown in many studies to cause disabling long-term side effects in children. These may include hearing abnormalities, growth deficits, endocrine abnormalities, neurocognitive decline, and decline in psychosocial function.
  • Proton radiotherapy allows the delivery of radiation dose with improved sparing of normal tissues. Protons are able to deposit maximal energy at their stopping point. This results in production of a Bragg peak, and therefore, no exit dose.
  • Protons may allow for greater conformality around the tumor compared with traditional 3D or IMRT photon plans, which will in turn spare more normal tissues.
  • Therefore, the use of proton radiotherapy in children may theoretically allow for an improvement in therapeutic ratio, allowing delivery of high dose and high volume radiation, while sparing normal tissues and reducing late and long-term toxicity.
  • This may also allow future escalation of both radiation and chemotherapy doses.
  • The purpose of the study presented here was to describe the clinical outcomes of patients undergoing proton radiotherapy for MB treated at Massachusetts General Hospital.


  • This is a Phase II trial examining the effects of proton irradiation in medulloblastoma patients.
  • The primary objective was to report DFS and OS in this cohort. The secondary objective was to report the severity of radiation toxicities encountered.
    • The 3 specific outcomes which were measured included: 1) hearing deficiency, 2) endocrine abnormalities, and 3) neurocognitive decline.
  • 60 patients with MB were prospectively enrolled from May 2003 to December 2009, and 59 were evaluable for this analysis.
  • Eligibility criteria:
    • Histologic diagnosis of medulloblastoma: standard-risk or high-risk, with imaging and pathology reviewed at MGH.
    • All patients underwent surgical resection of the primary tumor and adjuvant platinum-based chemotherapy.
    • Patients were allowed to enroll in other Children’s Oncology Group (COG) protocols, and 10% of patients were co-enrolled.
  • Radiation Therapy
    • CSI doses ranged from 18-36 Gray-Equivalents (GyE), with median dose 23.4 GyE. Patients received tumor bed boost to a total of 54 or 55.8 GyE.
  • Assessments: Several standard measurement tools were used to measured audiologic, endocrine and neurocognitive outcomes.
  • Audiology exams were performed at baseline (BL) and yearly thereafter.
  • Neurocognitive and IQ testing was done at BL and at years 1, 3, and 5 following XRT.


  • 60 patients were prospectively enrolled on this trial from May 2003 to December 2009, and 59 were evaluable for this analysis.
  • Patient demographics:
    • Median age: 6.6 years (range 3.5 -22 years)
    • Male/female ratio: 1.3
    • 45 patients were diagnosed with standard-risk (SR) disease; 14 patients had high-risk (HR) disease.
  • Median follow up: 16 months
  • 3 year OS for entire group: 90%
  • 3 year PFS for entire group: 80%
  • 2 year OS for entire group: 91% (92% for SR MB and 87% for HR MB)
  • 2 year PFS for entire group: 87%
  • Neurocognitive decline:
    • 54 pts had baseline (BL) neurocognitive testing
    • Of these, 19 pts (14 SR, 5 HR) had follow-up (FU) evaluations including Full Scale Intelligence Quotient (FSIQ) assessment at an average of 2 years. Mean FSIQ at BL and FU were 107 and 102 (paired t-test: 0.046).
    • BL FSIQ in this population was 11 points higher that that seen in the literature with standard photon radiation.
    • No significant drop in FSIQ, Verbal IQ (VIQ), or Performance IQ (PIQ) (all standard measurements) was seen for this cohort.
    • However, a statistically significant decline in processing speed from BL to follow up was seen, with a mean score decline from 102 to 88 (n=20).
  • Endocrine dysfunction:
    • 29% of patients had at least 1 endocrine deficiency, and there were significant differences seen in the 2 risk groups
      • § 22% in SR MB and 50% in HR MB
    • Late effects included 9 patients with growth hormone deficiency, 9 patients with central hypothyroidism and 3 patients with cortisol deficiency.
  • Audiology:
    • 31 patients had BL and follow-up audiograms (median f/up at 2.5 years) that were evaluable.
    • 8% of patients had grade 2 ototoxicity at baseline.
    • Dosimetric analysis showed that median doses to the cochleas were found to be significantly lower compared to IMRT plans for SR risk patients (25.3 GyE for protons vs. 29.5 Gy for IMRT). The differences were not statistically different for HR patients due to the higher CSI doses required in this setting.
    • 16% of patients had Pediatric Oncology Group grade 3 or 4 ototoxicity. (Paulino, et. al. 2010 showed 25% rate with IMRT)
    • Statistically significant hearing loss was seen by year 1 (median score of 0 at BL compared to 2 at 5 years: p<0.001)

Authors’ Conclusions

  • The authors concluded that disease control in these patients is equivalent to IMRT photon treatment.
  • They note that the early results of this phase II trial are promising for improving clinical outcomes in patients with medulloblastoma.
    • Neurocognitive outcomes were excellent, although the authors acknowledge that this study has a short-follow up at this time.

Clinical Implications

  • Radiation therapy for medulloblastoma is associated with debilitating long-term side effects in children, and novel ways to decrease these toxicities are necessary.
  • Proton beam therapy has been used for 35 years, and the biology of these beams is similar to that of photon beams. The benefit that protons may provide is aimed at minimizing normal tissue toxicity due to the better tumor conformality of this treatment.
  • As proton beam therapy is being implemented in a growing number of radiation therapy centers in the world, evidence of enhanced efficacy should be present before this expensive treatment is widely employed.
  • Dosimetric comparisons seen in the literature allow us to theorize that protons would offer more normal tissue sparing and benefit to some patients, especially children. However, the magnitude of clinical benefit is not known.
  • The authors of this trial have executed a well-designed trial, which is a great example of an evaluation of the magnitude of potential benefit that protons may provide, and have effectively evaluated the late-effects seen in patients treated with proton radiotherapy for MB.
  • They performed a very thorough clinical evaluation of toxicity with standardized measurement tools. Their clinical outcomes appear to be quite good, with limited acute and late toxicity. However, the follow-up at this time is short, which the authors do acknowledge.
  • Limitations:
    • No valid control groups were presented.
    • The number of patients undergoing neurocognitive and IQ evaluation after treatment is relatively small.
    • This cohort may be a selected subset of patients with higher IQs given that they are seeking out treatment at proton centers. They may also have more access to resources to help enrich or compensate for neurocognitive, hearing, and endocrine deficiencies.
    • Follow up is fairly short. The differential in outcomes may be greater or less with future follow-up as we know that radiation adverse effects can often take longer to manifest themselves. Neurocognitive decline in particular has been demonstrated to persist for many years after completion of radiotherapy in patients treated with IMRT (Merchant TE, 2007). This may or may not be the case in this cohort of patients treated with protons.
  • In the future, more detailed information regarding reduction of risk of secondary malignancies would be interesting. Also, the benefit of avoiding irradiation of thoracic and abdominal contents in patients receiving proton CSI should be explored.
  • Although randomized data for proton therapy may be desirable, this may be difficult to perform in the pediatric population due to ethical concerns and ability to accrue to trials. Will families accept randomization if they perceive no increased risk, but do perceive potential benefit from protons? Is it ethical to deny patients protons when we know there is a dosimetric advantage?
  • We do have a responsibility to society due to social costs of proton treatment, and there should be some clinical evidence of incremental value. These authors are making first attempts to provide this type of data and their work should guide further investigation in this area.


Understanding Proton Re-Irradiation
by OncoLink Team
January 28, 2016

Stay informed with the latest information from OncoLink!   Subscribe to OncoLink eNews
View our newsletter archives