Proton-Based Radiotherapy for Unresectable or Incompletely Resected Osteosarcoma

Reviewer: Christine Hill-Kayser
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: May 28, 2010

Presenter: Dr. Thomas F. Delaney
Presenter Affiliation: Massachusetts General Hospital, Francis H. Burr Proton Therapy Center, Boston, MA


  • Osteosarcoma represents the most common primary tumor of bone affecting adolescents and young adults.
  • Early treatment paradigms in the 1960's consisted of aggressive local control with amputation; since that time, the necessity of a multi-disciplinary approach including chemotherapy has come to be recognized. This approach has been demonstrated to improve both overall survival and preservation of organ/limb function.
  • Current treatment approaches general consist of neoadjuvant, intensive chemotherapy followed by resection. Radiation is not always part of the modern approach to treatment of osteosarcoma.
    • Osteosarcomas most commonly affect the extremities, and the majority of tumors are amenable to surgical resection.
    • Following this type of treatment regimen, limb preservation and local control rates approach 90%, with long-term local control rates of 60-70%.
    • Worse local control and overall survival may be associated with poor histologic/pathologic response, positive margins, or inadequate surgeries.
  • Osteosarcomas occurring in sites other than the extremities, including the craniofacial region, spine, sacrum, and pelvis, may pose particular difficulty in terms of resectability, and are associated with worse overall prognosis.
    • Other groups have reported local failure rates as high as 50% in the head and neck regions, and 70% in the pelvic and spinal regions, even after maximal surgical resection. Local failure rates are > 90% in these sites when tumors are unresected.
    • Along the same lines, the 5-year failure-free survival and overall survival are as low as 14% and 30%, respectively, for pelvic osteosarcoma overall.
  • Proton radiotherapy offers unique dosimetric properties that may allow safe delivery of high radiation doses with relative sparing of normal tissues. As a result, this type of treatment is particularly appealing in the setting of tumors in close proximity to vital normal structures -- among these are the craniofacial, spinal, and pelvic regions.
  • This study was undertaken to evaluate the role of proton radiotherapy for treatment of unresectable or incompletely resected osteosarcoma, and to analyze local control and overall survival after this type of treatment.


  • This study was carried out in a retrospective fashion, after approval from the Institutional Review Board had been obtained.
  • The records of all patients who received proton radiotherapy or mixed photon-proton radiation at the Massachusetts General Hospital for osteosarcoma between 1983 and 2009 were reviewed.
  • Criteria for receiving proton radiation at the time of treatment included:
    • Need for high dose and highly conformal radiation for unresected or partially resected osteosarcoma
    • Positive post-operative margins
    • Macroscopic disease on post-operative imaging studies
    • Incomplete resection as defined by the surgeon
  • Patients receiving proton radiation for the primary tumor at time of diagnosis or at the time of local recurrence, or for isolated metastatic disease, were included in this study.
  • Local control at the site treated with protons was defined as the primary endpoint of this study.
  • Secondary endpoints included:
    • Disease-free survival (DFS)
    • Overall survival (OS)
    • Long-term/late toxicity
    • Prognostic factors associated with clinical outcome
  • Median follow-up was 24.8 months.


  • A total of 55 patients, representing all of those treated with proton radiation at MGH for osteosarcoma from 1983 – 2009, were considered as part of this study.
  • Median age was 27 years (range 2 years – 76 years).
  • Disease sites included craniofacial and axial skeletal sites in most cases:
    • Head, n = 22
    • Cervical spine, n = 5
    • Thoracic spine, n = 8
    • Lumbar spine, n = 4
    • Sacrum/pelvis, n = 13
    • Femur, n = 1
    • Hip, n = 1
    • Chest wall, n = 1
  • Of these patients, 66% (n = 36) were treated with protons to the primary tumor site at the time of initial diagnosis, 30% (n = 17) were treated to the primary site at the time of local recurrence, and 2 patients (4%) were treated for isolated metastatic disease.
  • All patients had unresected or incompletely resected disease at the time of proton treatment, with 42% having undergone biopsy only, 9% having gross residual disease after maximum total resection, and 49% having positive surgical margins.
  • Most patients received neoadjuvant chemotherapy prior to radiation: 35% received intensive chemotherapy, and 56% some chemotherapy. Nine percent received no chemotherapy.
  • The median radiation dose delivered was 70 Cobalt-Gray Equivalents (CGE) (mean 68.4 CGE, standard deviation 5.4 CGE).
    • In 20% of patients, the entire radiation dose was delivered using protons. The remainder received mixed proton/photon radiation, with a mean of 58.2% of the total dose being delivered with protons (range 11% - 100%).
    • Treatment planning was performed such that 45 – 50.4 CGE was delivered to the clinical target volume (CTV), defined as the gross tumor volume (GTV) plus a margin to account for microscopic disease. This was followed by a boost to the GTV to a median total dose of 70 CGE.
    • 13% of patients received 19.8 CGE delivered pre-operatively.
    • 3 patients received intra-operative radiation via either dural plaque or electron beam treatment.
  • Local control at 3- and 5-years, respectively, was 82% and 72%. In total, 12 patients experienced local recurrence.
    • Of local recurrences, 10 of 12 were classified as in-field recurrences. Of these, 8 patients had craniofacial tumors, 1 had a sacral tumor, and 1 had a vertebral tumor.
    • The remaining 2 local recurrences were classified as marginal failures, one in the spinal region, and 1 in the sacral.
  • The rate of development of distant metastatic disease was 25%.
  • Five-year disease-free survival was 70%, and 5-year overall survival was 58%. Four patients died without osteosarcoma; of these, one died of acute lymphocytic leukemia, and the other of a secondary cancer of the maxillary sinus after chemoradiation. The remaining two patients died of non-cancer-related causes.
  • Extent of surgical resection did not correlate with outcome. A trend towards increased local failure was demonstrated in patients with craniofacial disease (HR 2.6, P = 0.09) but this did not translate to increased risk of death or distant metastasis. Risk of distant metastasis was positively correlated with intermediate/high grade disease, with 10 of 11 patients with distant disease having grade 2 or higher grade tumors.
  • Late toxicity data was analyzed and demonstrated the following:
    • 12 patients experienced grade 1 late toxicity.
    • 12 experienced grade 2 late toxicity, defined as minimal interference with activities of daily living or pain requiring non-narcotic medications.
    • 8 experienced grade 3 late toxicity including pain requiring opioids (3), limb immobility (2), cranial nerve deficit causing diplopia (1), and functional bowel denervation (1).
    • 9 experienced grade 4 late toxicity, defined as total loss of an organ or its function. These included:
      • Enucleation (4)
      • Limb immobility causing handicap (2)
      • Ipsilateral blindness (1)
      • Ipsilateral hearing loss (1)
      • Maxillary deformity requiring multiple prosthesis (1)
      • Additionally, 2 patients died of second malignancies.
  • The authors described in detail one case of an adolescent female requiring proton radiotherapy for sacral osteosarcoma. Sparing of her reproductive organs was possible, and she retains ovarian function currently.

Authors' Conclusions

  • The authors conclude that high dose radiotherapy may compensate for the extent of surgical resection that is possible.
  • They state that particle treatment allows organ-preserving locally curative treatment, and may be a standard of care for unresectable or incompletely resected osteosarcoma.
  • They note that further work is needed to decrease toxicity, noting that 30% of patients in this study developed grade 3-4 late toxicity.

Scientific/Clinical Implications

  • The authors present one of the largest series of osteosarcoma patients treated with proton radiotherapy, and the presentation is certainly a valuable contribution to the literature.
  • The authors have done a systematic review of all of their patients treated for this disease with proton radiation, and describe local control and overall survival rates that are much better than those reported elsewhere in the literature. These findings demonstrate that proton radiation may be a treatment technique offering improved outcomes in this difficult subset of osteosarcoma patients.
  • Having said this, the study reported here is limited by several aspects:
    • First, the patient group is quite heterogeneous, including patients with newly diagnosed disease, locally recurrent disease, and distant metastatic disease. These patients may have very different prognoses at the start of radiation treatment, and the authors did not comment upon this.
    • Next, the radiation delivered to the patients varied considerably, with only 20% receiving the full radiation dose from proton radiotherapy. Indeed, the median percentage of treatment delivered with protons was 58%, and the lowest percent dose delivered with protons was 11%. The authors did not comment on whether prognosis was impacted by the percentage of proton therapy versus photon therapy delivered, but this might provided interesting insight.
    • Finally, although the patients in this study were treated as early as 1983, the median follow-up of 24 months demonstrates that long-term follow-up was not available for many patients. Again, longer follow-up might provide interesting insight.
  • The project presented here provides interesting data and supports the use of particle therapy in treatment of unresectable or partially resected osteosarcomas despite its limitations. Further data that will be of interest may include examination of more homogenous groups of patients, and those with longer clinical follow-up. Dose escalation in specific patients could also prove to be interesting.
  • Finally, as the authors point out, the toxicity experienced in this patient population was quite high. Certainly, in central sites such as the craniofacial region and axial skeleton, morbidity from surgery would also be expected to be quite high. Further work to minimize toxicity through complex planning and delivery, as well as image-guided radiotherapy, may prove to be interesting in the future.


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