Emerging Radiation Technologies: IMRT and Proton Therapy
IMRT
IMRT is a sophisticated x-ray therapy technique that employs multiple radiation beams aimed at the target from different directions, with the beams varying in size and shape during treatment delivery to create a radiation dose distribution that is highly conformal with the 3-dimensional (3D) volume of the target. Since an x-ray beam deposits radiation throughout its entire path, areas where the radiation beams intersect receive high doses, while areas traversed by only one beam receive much lower doses. With the increased number of beams, the volume of tissue receiving the highest doses conforms more precisely to the actual target volume than is the case with simpler conventional RT techniques. Because of the increased number of beams used, however, a larger volume of nontargeted tissue receives some radiation dose compared with the simpler conventional RT techniques. In essence, IMRT redistributes the radiation dose to nontargeted tissues in a way that can be favorable to tissues at risk for a particular toxicity. Nonetheless, with both conventional and IMRT x-ray–based techniques, most of the total radiation dose is actually deposited outside the target.
Color-Wash Dose Distribution for Three Plans for a Representative Patient With Mediastinal Involvement of Hodgkin Lymphoma
Several published studies have compared the dose distributions of conventional 3D conformal RT (3DCRT) plans (typically designed as anterior-posterior/posterior-anterior [AP/PA] plans) with IMRT plans in patients with HL. In one of the first studies, investigators from Memorial Sloan-Kettering Cancer Center compared IMRT vs 3DCRT (AP/PA) and demonstrated a 12% reduction in mean lung dose with IMRT.[25] In another study, by Girinsky et al, IMRT was better able to protect the heart and coronary arteries than was 3DCRT, but with IMRT there was more concern regarding increased volume of normal tissue receiving “low doses” of RT than there was with 3DCRT.[26] More recently, Weber et al reported that in a nonlinear model of the development of secondary malignancies, IMRT increased the risk of breast, lung, and thyroid cancers compared with 3DCRT, in part because of the redistribution of radiation dose with IMRT that leads to an increased volume of normal tissue—which would not have been irradiated at all with 3DCRT—receiving “low doses” of RT.[22] Similarly, a recent study from the GHSG[27] demonstrated reduced dose to the heart and spinal cord with the use of IMRT, but increased dose to the lungs and breasts compared with 3DCRT.
Only one study has reported on outcomes in HL patients treated with IMRT.[28] Paumier et al reported on 32 patients treated to an INRT field with IMRT following chemotherapy and demonstrated 5-year progression-free survival and overall survival of 91% and 95%, respectively, comparable to what is achieved with standard techniques. Only one patient developed an in-field relapse, and one developed grade 3 pneumonitis.
Protons
Unlike x-rays, protons are charged particles with mass. Protons travel a finite distance, determined by their acceleration and the composition of the matter through which they travel; thus, the actual range of protons in tissue can be controlled, thereby eliminating the “exit” dose to nontargeted tissues. In addition, protons deposit most of their radiation dose in tissue near the end of their range in a striking pattern called the Bragg peak, with relatively little dose deposited along the “entrance” path. Whereas with the IMRT technique radiation dose to nontargeted tissue is redistributed, with PT it can actually be significantly reduced.
Toxicity and Critical Doses to Structures by Treatment Technique in Patients With Mediastinal Hodgkin Lymphoma
PT reduces high-, medium-, and low-dose radiation levels compared with 3DCRT. Dosimetric studies evaluating the use of PT in HL date back to 1974, when Archambeau et al explored the use of PT for total nodal irradiation.[29] In that study, the investigators demonstrated that PT could reduce the irradiated volume by 50% compared with photons. More sophisticated treatment planning studies have since been published. In a prospective phase II study of involved-node radiotherapy in patients with mediastinal HL,[30] the first 10 patients enrolled underwent treatment planning with 3DCRT (AP/PA), IMRT, and PT and were offered treatment with the plan that best spared the organs at risk while maintaining appropriate target coverage. In all 10 cases, PT was associated with the best plan and all patients were offered treatment with PT. The Figure shows the color-wash isodose distributions for the 3DCRT, IMRT, and PT plans for one of these patients. The Table describes the expected dose-volume effects from EFRT (mantle radiation) using 3DCRT, from IFRT using 3DCRT, from INRT using 3DCRT, from IMRT, and from PT; these data demonstrate the considerable dose reductions achieved with each successive treatment approach.
Organ-Specific Radiation Dose Reduction With PT
Using modern RT techniques for the treatment of HL, radiation oncologists must balance the risks and benefits of RT pathways through the various organs involved. Although secondary cancers are considered the biggest RT-associated concern for HL survivors, RT toxicity to the heart is responsible for more deaths than any other specific organ malignancy. Thus, at the University of Florida (UF), we attempt to limit the dose to all nontargeted normal tissue, but we prioritize the organs at risk in the following order: heart, lungs, breasts (women), and esophagus.
Heart
In a prospective UF study of 10 patients with HL who received 30–39.6 Gy of INRT,[30] the mean dose to the heart was 19.4 Gy with 3DCRT, 12.2 Gy with IMRT, and 8.9 Gy (relative biological effectiveness [RBE]) with PT. With PT, the dose was reduced by > 5 Gy in 9 of the 10 patients compared with 3DCRT, and in 4 of the 10 patients compared with IMRT. In a study from MD Anderson Cancer Center (MDACC) of 10 patients with mediastinal lymphoma, PT similarly reduced the mean heart dose by 9 Gy compared with 3DCRT.[31] In a follow-up UF study evaluating the dose to the subunits of the heart,[32] the mean doses with 3DCRT, IMRT, and PT, respectively, were: left ventricle, 13 Gy, 5 Gy, and 0 Gy (RBE); right ventricle, 17 Gy, 11 Gy, and 9 Gy (RBE); left atrium, 28 Gy, 15 Gy, and 5 Gy (RBE); right atrium, 24 Gy, 17 Gy, and 11 Gy (RBE); mitral valve, 28 Gy, 9 Gy, and 0 Gy (RBE); tricuspid valve, 19 Gy, 13 Gy, and 0 Gy (RBE); aortic valve, 30 Gy, 18 Gy, and 9 Gy (RBE); left circumflex artery, 30 Gy, 16 Gy, and 5 Gy (RBE), and left anterior descending artery, 18 Gy, 10 Gy, and 5 Gy (RBE).
Lungs
In the UF study,[30] mean lung dose was 13.2 Gy for 3DCRT, 10.6 Gy for IMRT, and 7.1 Gy (RBE) for PT. In particular, PT reduced the dose to the lungs by > 5 Gy in six patients compared with 3DCRT and in two patients compared with IMRT. Similarly, MDACC demonstrated a reduction in mean lung dose of 3.3 Gy compared with 3DCRT.[31] Because of tissue density in the lungs, however, the doses for the PT plans might have been slightly underestimated.
Breasts
In the UF study,[30] the mean breast dose was not substantially reduced because of the limited treatment field, which resulted in mean breast doses that averaged 5.4 Gy for 3DCRT, 5.5 Gy for IMRT, and 4.6 Gy for PT. However, in two patients with residual bulky disease extending posterior to the breast, PT was able to reduce the mean breast dose by > 5 Gy compared with 3DCRT or IMRT. In the MDACC study, only a marginal benefit was seen in the mean dose to the breasts with PT compared with 3DCRT (5.9 Gy vs 6.1 Gy).[31] In a study by Andolino et al, however, PT reduced the mean dose to the breasts to 1 Gy—from 4.7 Gy with 3DCRT.[33] Because this study chose a posterior approach, with the beam entering the anterior mediastinum through the posterior chest and heart, the heart received a higher mean dose with PT (17 Gy) than with 3DCRT (14 Gy), highlighting the occasional tradeoff with central tumors of sparing anterior vs posterior structures. In a case report of a posterior mediastinal HL, the breast dose was reduced with PT from 5.7 Gy to 1.7 Gy, and the mean heart dose was reduced from 24 Gy to 11 Gy.[34]
Esophagus
In the UF study, the mean esophagus dose was 22.6 Gy for 3DCRT, 17.2 Gy for IMRT, and 17 Gy for PT. PT reduced the dose by > 5 Gy in 7 of the 10 patients compared with 3DCRT and in 3 patients compared with IMRT. In the MDACC study, PT reduced the dose to the esophagus, on average, by 13 Gy compared with 3DCRT.[31]
Other structures
PT can also reduce the dose to other important structures in patients with HL. In a case report of a prepubescent pediatric patient with HL, PT was able to reduce the number of thoracic vertebral bodies irradiated by five, which should result in less of an impact on patient growth.[35] Lastly, subdiaphragmatic HL is treated with target volumes and fields similar to those used for seminoma (para-aortic fields +/− pelvic fields); two studies have demonstrated that PT could reduce the dose to the stomach, bowel, bladder, and ipsilateral kidney, compared with 3DCRT and IMRT.[36,37]
Uncertainties
PT treatment planning is more complex than x-ray treatment planning. The depth that protons travel in tissue depends on their energy and the composition of their pathway. Minor variations in daily patient positioning may result in minor variations in the proton path length, which must be accounted for in the treatment planning process. Improved treatment planning and delivery systems will reduce this uncertainty and minimize adjustments needed in the treatment planning process, leading to even more conformal PT dose distributions in the future. While concerns have been raised about uncertainty related to secondary neutron scatter in patients receiving double-scatter PT, these effects are minimal when considered in the context of the reduction in dose achieved with PT vs x-ray treatment. Clinical experience suggests no increase in the risk of second malignancy with PT, meaning that the impact of neutrons has been immeasurably small.[38]
Financial Costs
As is the case with most medical advances, PT is more expensive than IMRT or 3DCRT. According to the Centers for Medicare and Medicaid fee schedule for local 99, reimbursements for a patient receiving 30.6 Gy at 1.8 Gy per fraction with 3DCRT, IMRT, and PT are approximately $6,000, $15,000, and $23,000, respectively. However, considering the CCSS report indicating that 40% of HL survivors have a grade 3 or higher chronic health condition 25 years following treatment, a reduction in these chronic health problems by even a fraction through the use of PT may result in a lessening of the cost discrepancy over time.[3]
Guidelines
Although clinical studies of HL have been published by investigators from UF and MDACC, a randomized study would be impossible due to the low incidence of HL in the population and the decades of follow-up needed to quantify a difference in late toxicity. The most recent National Compre-hensive Cancer Network (NCCN) guidelines (January 2012) stipulate that treatment with either photons or protons is acceptable. Moreover, upcoming pediatric HL studies being developed by the Children’s Oncology Group will consider including PT to help reduce long-term treatment toxicity.
In summary, advances in the treatment of HL over the last 30 years have dramatically reduced the RT dose to nontargeted normal tissue. Some of these advances, such as IFRT, are only now being fully appreciated, with recent studies demonstrating less long-term toxicity. However, many of the more recent technological developments, including INRT, IMRT, and PT, are unlikely to demonstrate improvements in toxicity for at least another decade. Despite these limitations, modern techniques should be used in future clinical studies in an attempt to reduce late chronic morbidity.
Financial Disclosure: Dr. Li is a member of the Speakers Bureau for IBA (Ion Beam Applications). The rest of the authors have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.
Acknowledgement: The authors would like to thank Jessica Kirwan for her help with preparation of the manuscript.
