Proton therapy of ocular tumors for very young children under general anesthesia

Reviewer: Christine Hill-Kayser, MD
Abramson Cancer Center of the University of Pennsylvania
Last Modified: October 14, 2009

Presenter: J. Heufelder
Presenter's Affiliation: Berlin, Germany
Type of Session: Scientific


  • Ocular tumors represent an important subset of pediatric malignancies, often affecting very young children.
  • Most children with retinoblastoma and other ocular tumors go on to be cured of their disease; however, treatments are not without potential significant morbidity. This may include need for enucleation or exoneration from surgical approaches; radiotherapy to the globe and orbit may affect development of facial bones, resulting in disfigurement, and may also result in damage to lacrimal glands and optic structures.
  • In addition, children affected by retinoblastoma have a recognized substantially increased risk of development of second malignant neoplasm, most commonly osteosarcoma. Certainly, exposure of bone and soft tissue to radiotherapy may increase this risk.
  • Proton therapy offers sharper dose gradients than does proton radiation, improving conformality and dose delivery with sparing of normal tissue. These qualities make it appealing for treatment of tumors, particularly those in close proximity to vital structures.
  • In addition, use of a single frontal proton beam to deliver dose to the eye may allow treatment to be delivered without traversing bone; such an approach cannot be used for photon-based radiation delivery.
  • For all of these reasons, proton treatment for ocular tumors in children may potentially provide benefit. Such treatment is complicated because treatment of the eye requires a great deal of cooperation on the part of the patient. This is clearly not possible for very young children, who require general anesthesia to receive radiotherapy.
  • The present study describes a new methodology for ocular treatment of the eye in an anesthetized child.

Materials and Methods

  • All patients were treated with general anesthesia for simulation and daily treatments.
    • A dedicated mobile anesthesia workstation was installed in the treatment room of the proton therapy unit at the Helmholtz Institute Berlin.
    • Anesthesia induction was performed on a separate couch.
    • After induction, patients were transferred to treatment chairs, and stabilized with seat belts.
    • Following treatment, patients were transferred to a separate couch and then to the recovery room.
  • Patients were immobilized in conventional car seats, modified to serve as treatment chairs.
    • After transfer to the treatment chair, the child’s head was immobilized via a thermoplasic mask fixed to the car seat.
    • The car seat was then mounted to the treatment chair and moved to an upright position.
  • In treatment position, the eyelids were moved out of the radiation fields by lid retractors, and a suction cup was attached to the cornea to adjust the gazing angle of the eye for treatment.
  • A simulation session was performed prior to the treatment course.
  • Treatment was planned using CT simulation, with collimating and wedges used to sculpt one frontal beam for dosing of the tumor within the globe.
    • Treatment planning corrections for wedges and suction cup were performed.
    • Monte Carlo calcifications were employed to account for intratumor calcifications.
    • Dose to orbital bone was constrained to less than 10% the total dose.
  • Daily doses were delivered as follows:
    • Retinoblastoma: 31.62 CGE were delivered in 6 fractions of 5.27 CGE each.
    • Choroidal osteoma: 20 CGE were delivered in 4 fractions of 5 CGE each.
  • Eye position was verified daily prior to treatment administration.


  • Treatment was delivered using the described approach to three children:
    • 2 patients, aged 10 months and 7 months, were treated for retinoblastoma
    • 1 patient, aged 5 years, was treated for choroidal osteoma.
  • The treatment procedure was found to require a total of 2 hours daily as follows:
    • Anesthesia induction and positioning: 1 hour
    • Irradiation: 1 minute
    • Dismounting: 1-5 minutes
    • Recovery: 45 minutes
  • Emergency situations were simulated, and dismount was able to be performed in less than one minute to allow access to patients requiring emergency procedures.
  • Treatment was tolerated well and without significant side effects by all patients.
  • Frontal irradiation was found to allow sharp lateral penumbra and sharp distal fall off, so that the bones of the skull could be spared completely.

Author's Conclusions

  • The authors conclude that proton therapy of ocular tumors for very young children under general anesthesia on a horizontal eye beam is feasible.
  • They describe sparing of the bones of the skull and orbit using this approach.

Clinical/Scientific Implications

  • The authors present an interesting feasibility study demonstrating that a frontal proton beam may potentially be used to treat pediatric ocular tumors even when general anesthesia is required.
  • This study does not include presentation of outcomes data, which would of course be of great interest.
  • With regard to the treatment protocol used, information regarding tumor size and location would be useful; larger tumors may potentially require higher doses than were used during treatment on this study. In addition, tumor location within the globe could potentially impact ability to deliver adequate dose from a single frontal beam.
  • Certainly, larger studies with higher numbers of patients and more comprehensive clinical information are necessary for proper assessment of the technique described here. Long–term outcomes data will also be of great interest.
  • In the meantime, the authors have demonstrated that the treatment approach described is feasible. This is of particular interest with regard to the pediatric retinoblastoma population, for whom drastically increased rates of second malignant neoplasm have been described. This risk is increased by the presence of radiotherapeutic treatment, and use of a single proton beam may well serve to minimize it.
  • If the authors continue to treat patients in this manner and long-term data becomes available, the technique described here has potential to provide benefit to patients without the long-term sequelae associated with photon radiation to the pediatric orbit. This would certainly be a major advancement in the field of pediatric radiation oncology.