Advanced-Technology Radiation Therapy for Bone Sarcomas

Samir Patel, MD; Thomas F. DeLaney, MD


Cancer Control. 2008;15(1):21-37. 

In This Article

Techniques of Radiotherapy

IMRT uses modulated field intensity across the photon radiation field to improve coverage of the tumor volume while reducing the radiation dose to selected normal tissues (Fig 1A).[27] Computer-controlled multileaf collimation provides high conformality, which makes it feasible to treat tumors of any shape, even those that are in close proximity to the spinal cord. Recent dosimetric studies[20] comparing IMRT and 3-D conformal RT for sarcomas have been reported. When evaluating sarcomas arising in the extremities, pelvis, trunk, and paranasal sinuses, IMRT plans were more conformal. In the extremities, bone and subcutaneous doses were reduced by up to 20%. A conformal IMRT comparative planning study has been reported for a large extraskeletal chondrosarcoma of the extremity.[28] Not surprisingly, IMRT produced excellent conformal treatment plans for this complex target volume, with a reduction of the maximum dose to the bone compared with the 3-D photon plan. In addition, IMRT can reduce hot spots in the surrounding soft tissues and skin. It should be noted, however, that IMRT treatment plans often include localized areas within the high-dose volumes where dose inhomogeneities can be in the range of 10% to 15% above the prescription dose. Because there can also be dose inhomogeneities in the range of 5% below the target dose, treatment plans are often normalized to the 95% isodose line, meaning that selected areas of the treatment volume are receiving daily fractions and total doses of 15% to 20% above the target dose. Depending on the location of these "hot spots," there can be unanticipated acute normal tissue toxicity.[29] Whether there are late effects attributable to these focal areas of high dose remains unclear.

Axial displays of an IMRT (A) and proton RT (B) plan for an 18-year-old patient with a large osteosarcoma of the left pelvis, which was managed by a combination of induction chemotherapy with methotrexate, cisplatin, and doxorubicin followed by concurrent ifosfamide/etoposide chemotherapy and RT. RT consisted of IMRT (25.2 Gy in 14 fractions, left) and protons (19.8 CGE in 11 fractions, right) to the clinical target volume (contoured with thin magenta line). This encompassed the gross tumor and areas at risk for microscopic extension followed by boost radiation of another 25.2 CGE in 14 fractions (C) with protons to the gross tumor volume (contoured with thin red line) visualized on the planning CT and MRI scans to bring the total radiation dose to 70.2 CGE (bold red isodose line on D). Note reduction in dose to bowel (contoured with thin green line) and other tissues anterior to the target, as well absence of dose to contralateral pelvis, with protons compared to IMRT. Figures courtesy of Judy Adams, CMD, Francis H. Burr Proton Therapy Center, Boston, Mass.

The rationale for the use of protons rather than photons (ie, x-rays, which have traditionally been used for RT) is the superior dose distribution that can be achieved with protons.[30] Protons and other charged particles deposit little energy in tissue until near the end of the proton range where the residual energy is lost over a short distance, resulting in a steep rise in the absorbed dose known as the Bragg peak.[31,32] The Bragg peak is too narrow for practical clinical applications, so for the irradiation of most tumors, the beam energy is modulated by superimposing several Bragg peaks of descending energies (ranges) and weights to create a region of uniform dose corresponding to the depth of the target. These extended regions of uniform dose are called spread-out Bragg peaks (Fig 2). Although the beam modulation to spread out the Bragg peaks increases the entrance dose, the proton dose distribution is still characterized by a lower-dose region in normal tissue proximal to the tumor,a uniform high-dose region in the tumor,and zero dose beyond the tumor.[31,32] Most proton facilities deliver the beam via passive scattering techniques that spread out and shape the beam with physical beam-modifying devices. On a beam-by-beam comparison, protons will always give less dose to normal tissue than photons. This will often result in a reduction in composite dose (integral dose) to normal tissue of 2- to 3fold depending on the location of the tumor and the choice of fields.[33] The primary physical interaction of protons with tissue is ionization of electrons and yields a biologic effect that is similar to that of x-rays, although the magnitude of the biologic effect is slightly greater.[34] Thus, the physical dose of protons in the clinic is corrected for this difference in relative biological effectiveness (RBE) by an RBE factor (generally 1.1 in most clinical centers).[34] Proton doses in the clinic have been expressed in cobalt Gray equivalents (CGEs),which represent the physical proton dose multiplied by the 1.1 RBE correction factor.

Depth-dose distributions for a spread-out Bragg peak (SOBP, red), its constituent pristine Bragg peaks (blue), and a 10-MV photon beam (black). The SOBP dose distribution is created by adding the contributions of individually modulated pristine Bragg peaks. The penetration depth, or range and measured as the depth of the distal 90% of plateau dose, of the SOBP dose distribution is determined by the range of the most distal pristine peak (labeled "Pristine Peak"). The modulation width, as measured as the distance between the proximal and distal 90% of plateau dose values, of the SOBP dose distribution is controlled by varying the number and intensity of pristine Bragg peaks that are added, relative to the most distal pristine peak, to form the SOBP. The dashed lines (black) indicate the clinical acceptable variation in the plateau dose of +2%. The dot-dashed lines (green) indicate the 90% dose and spatial, range and modulation width intervals. The SOBP dose distribution of even a single field can provide complete target volume coverage in depth and lateral dimensions, in sharp contrast to a single photon dose distribution. Only a composite set of photon fields can deliver a clinical target dose distribution. Note the absence of dose beyond the distal fall-off edge of the SOBP. From Levin WP, Kooy H, Loeffler JS, DeLaney TF. Proton beam therapy. Br J Cancer. 2005;93: 849-854. Reprinted with permission.

Before the advent of IMRT, protons were used to deliver higher radiation doses than could be given with photons. Hence, the major emphasis for proton therapy clinical research initially was dose escalation for tumors for which local control with conventional RT was poor,[35] including base of skull and spine tumors,[14] locally advanced prostate cancer,[36] hepatocellular carcinoma,[37] and non–small-cell lung cancer.[38] The recent development of hospital-based cyclotrons with higher energy beams capable of reaching deep-seated tumors (up to approximately 30 cm with a 235 MeV beam), field sizes comparable to linear accelerators, and rotational gantries has greatly facilitated proton RT. Hence, proton-beam RT is being used for a wider range of clinical sites than in the past.

In tumor sites in which tumor control with photons is good, there is increasing interest in protocols aimed at morbidity reduction by exploiting the reduction in exit dose with protons to reduce normal tissue radiation doses and thus the toxicity of treatment. Many pediatric tumors fall into this category, and thus there is much interest in the use of protons to reduce the risk of late effects of RT on developing normal tissues in children (Fig 3A-C).[39] It is to be emphasized that dose escalation and morbidity reduction are not mutually exclusive when using protons and that the opportunity for both might be present in any given patient (Fig 1A-D).

Axial (A) and sagittal (B) dose displays of the proton radiation treatment plan in a 12-year-old female with a lumbar epidural Ewing's sarcoma (C, arrow). Following emergent surgical decompression and gross total resection of extraosseous tumor, chemotherapy with vincristine, ifosfamide, doxorubicin, and etoposide was started. After several months of chemotherapy, proton radiation concurrent with cyclophosphamide/vincristine was started. RT was delivered exclusively with protons to eliminate exit dose and spare all of the normal tissues anterior to the vertebrae. Radiation was delivered via progressively shrinking fields that initially encompassed the tumor bed, surgical scar, drain sites, and tissues suspected of microscopic tumor involvement (30.6 CGE in 17 fractions), then reduced to cover the tumor bed, surgical scar, and tissues suspected of microscopic tumor involvement for another 14.4 CGE in 8 fractions. This was followed by further field shrinkage to encompass the tumor bed for another 5.4 CGE in 3 fractions, and finally a localized boost to a site of bony abnormality that was not resected for another 5.4 CGE in 3 fractions. The total dose to the entire original tumor bed was 50.4 CGE and to sites of unresected disease was 55.8 CGE. Apart from some moderate erythema in the skin overlying the lumbar spine, the patient experienced no acute morbidity from her radiation treatment; in particular, she had no nausea or diarrhea related to treatment.

Photon IMRT can now deliver radiation doses to the target that are often competitive with those achievable with protons. The distribution of 3-D conformal proton doses is generally more homogeneous than IMRT, although focal hot spots can also arise where fields are junctioned or "patched."[40] Integral doses are higher with photon IMRT than with protons; thus, most experts would suggest greater toxicity with IMRT photons than with protons. The magnitude of the difference in acute and late toxicity between IMRT and protons is not currently known.

The definition of the treatment target volume (ie,the clinical target volume that encompasses the tumor and areas at risk for subclinical spread of tumor, which is by definition below the resolution of the imaging studies and is generally irradiated to doses of 45 to 50.4 Gy) and the gross tumor volume (ie,the tumor visualized on imaging studies) will be the same for 3-D conformal RT, IMRT, 3-D conformal protons, or intensity-modulated proton radiation therapy (IMPT). All of these techniques rely on the accuracy of tumor imaging for the definition of the gross tumor volume and the judgment of the treating physician about the margin to use around the gross tumor volume to allow for subclinical spread of tumor. This margin depends on multiple factors including tumor type, natural history, risk of nodal involvement, and historical patterns of failure. The dose gradient at the edge of the target volumes, however, will be sharper with IMRT,[41] 3-D conformal protons,[42] or IMPT[43] than with 3D conformal photons. Hence, there is increased importance on confirmation of positional accuracy of the patient and the beam. For this reason, image guidance can be of considerable importance for these tumors. In practice, daily pretreatment, diagnostic-quality positioning radiographs have been taken prior to each proton treatment to guarantee the accuracy of treatment. Diagnostic quality radiographs, cone-beam CT, megavoltage CT and other image guidance tools have been used to guarantee the accuracy of setup for IMRT.[44]

Because of the defined range of the proton, a technique known as "smearing" has been used to compensate for insufficient proton range in the case of inaccuracy in daily patient setup (where, for example, bone interposed in the path of the proton when the treatment plan had only anticipated soft tissue would result in a shortfall in proton range and a geographic miss of the distal edge of the target). Smearing adds additional range to the proton in a given radius around the desired point, taking into account the different tissue heterogeneities that the proton is likely to encounter in traversing any point within that radius.[45] Without this technique, the risk for geographic miss with the protons would be magnified. While this has been a safe technique to ensure target coverage, it underscores the importance of setup accuracy when using this modality; higher degrees of setup accuracy can reduce the radius required for smearing.

A subtle but important area that is the subject of ongoing clinical investigations is the greater sensitivity of proton treatment plans compared with photon plans to changes in tissue densities such as those in paranasal sinuses where air,bone,and soft tissues form a complex geometry. Changes in tumor configuration in this location from tumor shrinkage or normal tissue density from sinusitis,for example,are expected to have a more profound impact on a proton than on a photon treatment plan[46] due to the exquisite range sensitivity of the proton. Thus, adapting the radiation treatment plan to the changes in tumor and normal tissue that occur over a course of treatment by repeated radiation planning CT scans with reevaluation and replanning as dictated by the clinical findings will be an even more important tool for protons than for photons for anatomic sites and tumor types where tumor regression and normal tissue changes occur over the course of treatment.

Some concern has been raised recently about the potential late carcinogenicity of secondary neutrons associated with the beam-modifying hardware used to deliver passively scattered protons.[47] The magnitude and relative contribution of these neutrons to late secondary malignancies are currently being debated.[48] This is not a major concern for magnetically scanned protons (such as would be used for intensity-modulated proton therapy, IMPT). Because irradiation of normal tissue offers no possible advantage to the patient, there has been much discussion about the ethics and acceptability to patients of participation in randomized clinical trials of protons vs IMRT.

One property of charged particles used to assess the biologic effect of a particular radiation is linear energy transfer (LET), the rate of energy loss by the particle in tissue. The LET influences the biologic impact of the energy deposited in tissue.[49] X-ray and gamma ray photons, protons, and helium ions are considered low-LET radiation. Heavier charged particles (neon, carbon) and fast neutrons are considered high-LET radiation. There is an initial increase in the RBE with an increase in LET. Higher-LET radiation is less affected by tissue oxygenation and less sensitive to variations in the cell cycle and DNA repair. For these reasons, the use of heavier, higher LET-charged particles is of interest.[50] These higher LET particles have the superior physical dose distribution characteristics of charged particles and also have a higher RBE, thus potentially having both a physical and biologic advantage over protons. They present challenges for use in the clinic, however, because of uncertainties about potential differences in the RBE in different tissue types and changes in RBE over the path of the clinical beam.[51,52] Higher rates of late normal tissue injury were seen in some studies with neutrons (which have a higher RBE than photons), although contributing to this was the fact that their physical dose distribution was much inferior to that which can be achieved with heavier charged particles.[53] Hence, these heavier charged particles have been introduced in careful phase I and II studies to minimize the risk of potential normal tissue complications.[54]

Heavier charged particles have been used in several centers for the treatment of sarcomas with promising results.[25,26,55] To date, they have not been compared in any randomized clinical trials against protons to assess whether the higher RBE offers any clinical advantages over protons. This will depend on whether the higher LET results in an improved therapeutic ratio by achieving either a higher percentage of tumor control for a given level of normal tissue toxicity or a similar percentage of tumor control with less normal tissue toxicity. A practical consideration is that heavier charged particles such as carbon ions will be more expensive than protons to deliver because of their substantially higher mass (12-fold for the case of carbon) and the higher-energy cyclotrons and more powerful magnets needed to accelerate and steer them. Hence, comparative trials between protons and higher-LET heavier charged particles will be important.

Intensity modulation can also be applied to proton beams, potentially optimizing the dose distribution even further. Dosimetric comparisons between photon IMRT and proton IMPT show a marked reduction in doses delivered to normal tissues that may prove to have a clinically significant impact on toxicity for the patient (Fig 4A-B).[43] Although it is currently unresolved whether this optimized physical dose distribution will be accompanied by an important clinical advantage, randomized clinical comparisons might be difficult to conduct because patients and their treating clinicians might have concerns about randomizing patients to the photon IMRT arm with its associated higher normal tissue radiation doses.

(A) Color wash dose distributions for IMRT, and (B) IMPT treatment plans for a patient with a paraspinal sarcoma. The isodose contours are represented by different colors (corresponding values, in Gy (or CGE) and are displayed on the lower border of each figure. From Weber DC, Trofimov AV, DeLaney TF, et al. A treatment planning comparison of intensity modulated photon and proton therapy for paraspinal sarcomas. Int J Radiat Oncol Biol Phys. 2004;58:1596-1606. Reprinted with permission by Elsevier.

IMPT with a spot-scanning beam has been used to treat a limited number of patients at the Paul Scherrer Institute in Switzerland.[56,57] Raster-scanned carbon-ion RT is used to treat patients at the GSI in Darmstadt, Germany.[26]

Even with IMRT or a particle beam, it is difficult to deliver tumoricidal doses of ≥70 Gy to the surface of the dura when, as is often the case, it is involved by spinal or paraspinal sarcomas. The physical properties of beam penumbra and the proximity of the surface of the spinal cord within 3 to 4 mm of the dura prevent delivery of adequate dose to the dural surface without exceeding spinal cord tolerance. A strategy developed at our institute is the use of a custom-designed yttrium90 plaque to boost the dose to the dural surface intraoperatively. The percent depth dose characteristics of the applicator are favorable for this application, with 100%, 27%, and 8% doses measured at depths of 0 mm (dural surface), 2 mm, and 4 mm, respectively. Thus, 10 to 15 Gy can be delivered to the dural at the time of surgical resection, with minimal dose to the cord surface and essentially no dose to the cord center or the contralateral side of the cord (Fig 5). Initial experience with this applicator has been favorable and has not shown any acute toxicity or any neurologic sequelae.[7]

Brachytherapy plaque on the posterior dural surface of a 55-yearold patient undergoing resection of a chondroblastic osteosarcoma involving the T6 vertebral body along with placement of spine stabilization hardware at the time of his posterior approach. A radiation dose of 10 Gy was delivered to the dural surface over 9 minutes 21 seconds. The patient had received 19.8 Gy of IMRT preoperatively and then a total of 50.4 CGE of combined IMRT and proton radiation postoperatively for a total dose of 70.2 CGE delivered via shrinking field technique. The dural brachytherapy was used to increase the dose to the dural surface, which received only 60 Gy with the external irradiation. Adjuvant chemotherapy was delivered after completion of surgery and RT.

With the exception of tumors recurrent after prior external-beam radiation, when it may be used as the sole form of RT in the treatment course, intraoperative RT (IORT) has generally been administered as a boost dose given in conjunction with external beam radiation. During surgery, 10 to 15 Gy is delivered with electrons or high-dose rate intraoperative brachytherapy to sites of microscopically positive margins which require doses of radiation in excess of those tolerated by surrounding critical normal tissues such as bowel which can kept out of the IORT field. The risk of wound-healing complications does not seem to be enhanced with IORT. It is generally administered after preoperative, neoadjuvant external-beam RT, although it has also been given prior to external-beam RT. This technique is particularly advantageous for retroperitoneal and pelvic sarcomas for which the tolerances of surrounding critical normal tissues, including bowel, liver, spinal cord, and kidneys, constrain the dose that can be delivered to the tumor with external radiation.[21]

Preoperative RT at 45 to 50.4 Gy (1.8 Gy per fraction) has been preferred for retroperitoneal soft tissue sarcomas because the tumor acts as a tissue expander and moves normal structures out of the radiation beam.[58] Because of the extent of surgery required for pelvic sarcomas that may include internal hemipelvectomy or partial sacral resection with a relatively high risk of surgical complications in the absence of any RT, we have explored the use of lower-dose preoperative RT for patients undergoing surgical resection when a positive margin is anticipated, as is frequently the case for sacral or pelvic bone sarcomas. We have delivered 19.8 to 20 Gy at 1.8 to 2.0 Gy per fraction, followed immediately by resection with intraoperative and additional postoperative radiation. The advantage of this technique is that the radiation fields for both preoperative and postoperative radiation are smaller compared with that of postoperative radiation alone. Surgical scars and drain sites do not need to be covered after this low-dose preoperative irradiation, which appears to be sufficient to prevent tumor seeding into the wound, surgical scars, and drain sites.[59] This appears to be a promising technique for high-risk sarcomas involving the bone, particularly for patients managed in this manner at the time of initial treatment, with a 5-year local control rate of 91.8%.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.