Long-Term Follow-Up on Phase II Study of High-Dose Proton/Photon Radiotherapy for Spine Sarcomas
Reporter: Abigail Berman Milby, MD
The Abramson Cancer Center of the University of Pennsylvania
Last Modified: May 21, 2012
Presenting Author/Institution: Thomas DeLaney, MD/Massachusetts General Hospital
The primary treatment of tumors of the spine and paraspinal tissues is challenging; surgical resection is the mainstay of treatment but negative margins are rarely obtained. As a result, high radiotherapy doses are needed as post-operative treatment.
Radiotherapy for spine sarcomas is constrained by the spinal cord, nerve, and viscera tolerance.
The local failure rates of chordomas and chondrosarcomas of the spine are very high. Chondrosarcomas have been shown to have up to 40% local failures rates (Bergh Cancer 2001) and chordomas up to a 55% local failure rate as shown in data from the Mayo Clinic (Keisch J Neurosurg 1991).
A pilot study was performed by the MGH showing high rates of local control (in 9 of 14 patients with chordomas and 6 of 6 patients which chondrosarcomas) with a mean target dose of 73.9 Gy (RBE) (Hug the Int. Journal Rad. Biol. Physics 1995).
A prospective phase II trial evaluating high-dose photon/proton RT for spine sarcomas accrued patients between 1997 and 2005. The first report was published in the Int. Journal Rad. Biol. Physics by Dr. DeLaney in 2009. This is a report of the long-term data.
Eligible patients had nonmetastatic, thoracic, lumbar, and/or sacral spine/paraspinal sarcomas.
Treatment included pre-and/or postoperative photon/proton RT with or without radical resection. Patients had no prior radiation. Some patients had dural plaque therapy.
Patients with osteosarcoma and Ewing's sarcoma received chemotherapy.
CTV1 was defined as subclinical disease and generally included the entire vertebral body and hemi-sacral segment, areas of extraosseous extension with at least 1-1.5 cm, and scar coverage for patients receiving post-op RT depending on grade of tumor and length of scar.
Shrinking fields delivered 50.4 Gy (RBE) to subclinical disease (CTV2), 70.2 Gy (RBE) to initial gross disease (CTV2), gross residual or biopsy alone (CTV3) to 77.4 Gy (RBE), all in in 1.8 cGy (RBE) fraction sizes. Ewing's sarcoma was treated to 61.2 Gy (RBE).
Doses were reduced for radiosensitive histologies, concurrent chemoradiation, or when diabetes or autoimmune disease were present.
If radiation was used pre-operatively a dose of 30.6 Gy was used except in the sacrum where 19.8 Gy was used.
Surgery was maximal safe resection. In the upper sacrum, several patients had biopsy only.
If plaque therapy was give, 90 Ytrium was used, delivered as 10 Gy for lesions abutting spinal cord.
All patients underwent CT planning with a myelogram and MRI fusion.
The following normal tissue constraints were used:
Spine: 63 Gy to the surface and 54 Gy to the center for up to 5 cm in superior-inferior length
For the rectum, use of an omental flap was used to displace the rectum. The posterior rectal wall was allowed to receive full radiation dose
Small bowel maximum dose of 50.4 Gy (MGH currently allows 57.6 Gy to 50% up to 10 cm)
For the skin, full dose was avoided over a large area of the sacrum.
Radiation therapy was delivered with a combined proton and photon technique. Proton therapy was passively scattered.
A total of 50 patients (29 chordoma, 14 chondrosarcoma, 7 other) were treated between 1997 and 2005. A total of 33 patients were still alive at the time of this analysis.
Sites within the spine were as follows: thoracic 11 lumbar 13, sacrum 26.
Primary tumors were treated in 36 and locally recurrent tumors in 14 patients.
Local control of primary tumors was 94% at 5 years and 84% at 8 years. Looking at all tumors, this was 80% and 73%, respectively.
There were 11 local recurrences with a median time of 34 months, with two late recurrences, one at 77 months another at 79 months. A total of 7 local recurrences were isolated and 4 were seen with distant metastases.
Local recurrences occurred in 7/14 of the locally recurrent vs. 4/36 primary tumor.
Local recurrences occurred in 5/29 chordomas vs. 5/14 chondrosarcomas and only 1 primary chordoma had a local recurrence.
There were no local failures for R0 resection.
Locoregional control was 77% at 5 years and 70% at 8 years
Overall survival was 83% at 5 years and 64% at 8 years
9 patients died of progressive tumor
8 died of other causes including 3 other tumors (oral cavity, bladder, CNS lymphoma) and 2 others have undergone treatment for other tumors (but are alive)
Relapse-free survival was 64% at 5 years and 52% at 8 years
The incidence of grade 3-5 complications was 13% at 5 years and 15% in 8 years.
An acute grade 3 complication was painful sacral stress fracture without late sequelae.
High dose photon/proton RT can be delivered to spine sarcomas. Morbidity is acceptable, and the local control is encouraging. Results are durable but occasional late recurrences do occur.
Local control is best if radiotherapy is performed immediately post-operatively rather than at recurrence.
The spinal cord constraints appear to be reasonable, however, late sacral nerve toxicity after 77.4 Gy is seen but is uncommon.
This is a report of a large phase II trial that investigated the role of high-dose proton/photon therapy for sarcomas of the spine. The results are impressive at 8 years following treatment with relatively high local control and low morbidity at 15%.
This report is one of a few that discuss the radiation tolerance of the sacral nerve roots, and, we see here that in some patients it appears to be below 77.4 Gy. Therefore, caution must be taken when treating the sacral nerve roots to a high dose.
This study used a mixed beam approach for several reasons: 1) skin sparing with photons and 2) decreased time on the proton therapy gantry permitting a greater total number of patients to be treated.
The authors do not specifically discuss the ratio of protons and photons and therefore it is difficult in this study to specifically address the utility of proton therapy. However, they were able to achieve the dose constraints outlined above.
Alternative approaches to sarcoma radiation include hypofractionation including stereotactic radiotherapy. The concern with this approach is that large fraction sizes may cause late toxicity to nerve roots more than smaller fraction sizes.
Future directions include incorporation of pencil beam scanning as the only modality of radiation or mixed with photon intensity modulated radiation therapy.
Sep 22, 2011 - Tumor necrosis factor-α and melphalan-based isolated limb perfusion therapy has a limb salvage rate of 81 percent in patients with locally advanced extremity soft tissue sarcomas who would have otherwise undergone amputation, according to a study published online Sept. 19 in the Journal of Clinical Oncology.