ABSTRACT: The risk of serious late complications in Hodgkin lymphoma (HL) survivors has led to a variety of strategies for reducing late treatment effects from both chemotherapy and radiation therapy. With radiation therapy, efforts have included reductions in dose, reductions in the size of the target volume, and most recently, significant reductions in the dose to nontargeted normal tissues at risk for radiation damage, achieved by using the emerging technologies of intensity-modulated radiation therapy and proton therapy (PT). PT is associated with a substantial reduction in radiation dose to critical organs, such as the heart and lungs, and has the potential to improve not only the therapeutic ratio, but also both event-free and overall survival. This review addresses the rationale and evidence for—and the challenges, cost implications, and future development of—PT as an important part of the treatment strategy in HL.
Introduction
RT and chemotherapy can cause agent-specific and dose-specific collateral damage to normal tissue. For example, alkylating agents cause bone marrow suppression and leukemia, while doxorubicin(Drug information on doxorubicin) causes cardiac toxicity; both chemotherapy toxicities are dose-related, so strategies for reducing these risks include either completely eliminating the agents from the treatment regimen or reducing their doses. Disease control may be compromised when an agent is eliminated[4,5]; therefore, the treatment offering the highest therapeutic ratio—that is, the greatest chance of disease control with the least chance of toxicity—may involve a combination of the minimal effective doses of maximally effective agents.
Compared with chemotherapy, RT offers more potential strategies for reducing normal-tissue effects because it is a targeted therapy rather than a systemic therapy. As with chemotherapy, the dose of radiation can be reduced. But in contrast to chemotherapy, the target volume for radiation can also be reduced, eg, from total nodal irradiation to extended field irradiation (EFRT) to involved field irradiation (IFRT) to involved node irradiation (INRT).[6] Furthermore, with new radiation modalities, such as intensity-modulated radiotherapy (IMRT) and proton therapy (PT), the radiation dose inadvertently delivered to nontargeted normal tissues can also be redistributed or lowered to reduce the probabilities of particular long-term normal-tissue toxicities.
Recent attempts to eliminate RT from the initial management of HL in adults have been unsuccessful—as with the European Organisation for Research and Treatment of Cancer (EORTC)/Groupe d’Étude des Lymphomes (GELA) HD10’s early closure of the chemotherapy-alone arm, reported at the 8th International Symposium on Hodgkin Lymphoma, Cologne, Germany 2010—confirming the efficacy of RT in the control of HL. In an analysis of the patterns of failure in the National Cancer Institute of Canada Clinical Trials Group/Eastern Cooperative Oncology Group (ECOG) HD6 trial—a study randomizing patients to EFRT +/− ABVD (Adriamycin [doxorubicin], bleomycin, vinblastine, dacarbazine(Drug information on dacarbazine)) vs ABVD chemotherapy alone—patients who did not receive EFRT had a lower rate of disease-free survival at 12 years (87% vs 92%; P = .05) and significant increased failure rates within the expected EFRT field (20/23 vs 3/10; P = .002) and within the expected IFRT field (16/23 vs 2/10; P = .02).[7,8] Unfortunately, the long-term toxicity of the generous EFRT field led to an increase in late side effects that translated into equivalent 12-year event-free survival rates of 85% vs 80% (P = .6) and lower overall survival rates. However, late toxicities can be avoided, providing a survival advantage, because of the evolution of RT in lymphoma, which currently uses smaller radiation fields, lower radiation doses, higher-energy linear accelerators, and modern radiation techniques.
Attempts to eliminate RT from pediatric HL protocols have met with mixed success, with a number of trials showing reduced disease-free survival (DFS) in chemotherapy-alone arms. While chemotherapy-intensive trials focusing on high-risk patients have had better success in providing adequate DFS without RT, DFS has come at the cost of both acute and chronic toxicities related to the use of alkylating agents, anthracyclines, and epidophylotoxins. The greatest success has resulted from coupling an increase in dose intensity with a response-based chemotherapy approach. Similar attempts to reduce chemotherapy and radiation exposure in low-risk pediatric patients have been less successful, and there is no single accepted standard of care for patients with low-risk nodular sclerosing HL.
Normal-Tissue Radiation Dose Effects
Much of our knowledge of late radiation effects on normal tissue comes from studies of patients with HL, who typically are young at presentation, are cured with RT alone, and are long-term survivors, with a unique opportunity to develop late toxicities decades after treatment. The two most commonly reported and most critical toxicities from RT in HL survivors are secondary cancers and cardiovascular disease. In fact, these are the two leading causes of death in 10-year survivors of HL.[9]
The most common secondary malignancies in HL survivors include lung cancer, breast cancer (for women), gastrointestinal cancer, and thyroid cancer [10]; however, other rarer cancers have also been reported, such as bone sarcoma[11] and mesothelioma.[12] Over the last decade, several studies have attempted to quantify the risk of developing cancer based on the radiation dose and the use of chemotherapy. In a nested case-control study of HL survivors of at least 1 year who developed or did not develop breast cancer (1:2 match, based on registry, year of diagnosis, age at diagnosis, and follow-up time period), Travis et al[13] demonstrated that increasing the radiation dose to the breast above 4 Gy was associated with an increased risk of subsequent tumor development compared with no RT. The relative risk (RR) of 1.8 for doses of 4 to 7 Gy further increased to a RR of 8 for doses of 40 to 60 Gy, and the median time to development of a secondary breast cancer was 18 years (range, 7 to 30 years) after HL diagnosis. In a similarly designed study of HL survivors (1:2 match, based on registry, sex, year of diagnosis, age at diagnosis, and follow-up period—but not smoking status), Travis et al[14] reported an increasing risk of lung cancer with increased radiation doses to the lung of 5 Gy or more, with a median time to development of lung cancer of 10 years (range, 1 to 28 years) after HL diagnosis. The RR for developing lung cancer was 4.1 for doses of 5 to 15 Gy, but 8.6 for doses of 30 Gy or more. In a third similarly designed study, 42 survivors of HL or testicular cancer who developed gastric cancer were matched with 126 other survivors of HL or testicular cancer who did not develop gastric cancer. The study demonstrated an increased risk of gastric cancer with increasing mean stomach dose, with a median interval between RT and development of secondary stomach cancer of 15.7 years (range, 9 to 28 years). The RR for developing a subsequent gastric cancer was 9.9 for a mean stomach dose of 20 Gy or higher compared with a dose below 11 Gy.[15] Lastly, in an analysis from the Childhood Cancer Survivor Study (CCSS) evaluating the impact of radiation dose on the development of thyroid cancer, a linear relationship was seen, with an increase in the radiation dose to 20 Gy resulting in an RR of 14.6 compared with no RT. However, unlike with other organs, radiation doses greater than 20 Gy resulted in a lower risk of thyroid cancer.[6]
Along with cumulative anthracycline dose, cardiac RT has been implicated in various cardiac problems in HL survivors. Cardiomyopathy, coronary artery disease, valvular disease, and pericarditis specifically have been found to be associated with increased RT dose to the heart. Mulrooney et al[16] evaluated cardiac complications among the patients in the CCSS who were followed for a minimum of 10 years. They found an increased hazard ratio (HR) for congestive heart failure (2.2), myocardial infarction (2.4), and valvular disease (3.3) with mean cardiac doses of 15 Gy and higher compared with no radiation. Tukenova et al[17] evaluated 4,122 people who were 5-year survivors of a childhood cancer diagnosed before 1986 in France or the United Kingdom. The risk of dying as a result of cardiac diseases (n = 21) was significantly higher in patients who had received a cumulative anthracycline dose greater than 360 mg/m2 (RR, 4.4; 95% confidence interval [CI], 1.3 to 15.3) and in those who received an average radiation dose that exceeded 5 Gy to the heart (RR was 12.5 for 5 to 14.9 Gy and 25.1 for > 15 Gy).
It is therefore to be expected that the risks of these serious late radiation-related toxicities could be reduced or eliminated by further reducing the radiation dose to nontargeted critical structures, such as the heart, thyroid, breasts, and lungs.
Smaller Radiation Fields and Lower Doses
Prior to the routine use of chemotherapy, radiation fields encompassed areas known to be “involved” as well as all areas at risk for subclinical involvement. Clinical trials have assisted with the establishment of efficacious yet moderate doses of chemotherapy for control of subclinical disease, permitting a reduction in the target volume for radiation treatment from all areas at risk for subclinical disease to only the area of involvement. Over the last 20 years, the typical radiation target volume has evolved from total nodal to subtotal nodal (or EFRT) to IFRT, without compromising event-free survival or overall survival. By reducing the radiation field, the volume of nontargeted normal tissue being irradiated has also been reduced, which has translated into lower rates of toxicity.
There is substantial evidence of toxicity reduction with reduction in radiation target volume. In 1972, Stanford University began to use cardiac and subcarinal blocking after 15 Gy, which resulted in a decrease in RR for non–myocardial infarction-related cardiac death, from 5.3 to 1.4. In the HD8 study by the German Hodgkin Study Group (GHSG), patients with early-stage unfavorable HL were randomized to 4 cycles of chemotherapy followed by either EFRT or IFRT to 30 Gy (with a 10-Gy boost to bulky disease).[18] Compared with IFRT, EFRT was associated with significantly increased rates of nausea (62.5% vs 29.1%, P < .001), pharyngitis (49.1% vs 40.5%, P = .001), leucopenia (49.1 vs 33.3%, P < .001), thrombocytopenia (16.7% vs 5.5%, P < .001), and gastrointestinal toxicity (17.5% vs 4.1%, P < .001), without any significant improvement in freedom from treatment failure or overall survival for the entire cohort because of the efficacy of 4 cycles of chemotherapy in controlling subclinical disease. In a subgroup analysis of 89 patients who were 60 years old or older, World Health Organization (WHO) grade 3/4 toxicity was remarkably higher with EFRT compared with IFRT (26.5% vs 8.6%). In particular, secondary cancers, grade 3/4 leucopenia and nausea, and grade 1/2 esophagitis and pharyngitis were considerably increased in the EFRT arm.[19] Lastly, in a meta-analysis evaluating the risk of secondary cancers in randomized controlled studies of patients with HL, there was a significantly higher risk of developing breast cancer following EFRT (odds ratio [OR] = 3.25; P = .04) vs following IFRT.[20] This difference in risk is due to omission of the axillary fields in patients for whom the axilla was uninvolved, resulting in a large component of breast dose from the mantle field.
With the development of more sensitive imaging modalities (such as positron-emission tomography [PET]-computed tomography [CT]), there is greater interest in radiation fields that include only the “involved nodes.” The long-term impact of further field reduction to only an “involved-node” field has been modeled in various studies and estimated to reduce the absolute risk of a cardiac event by as much as 5.1% compared with a mantle field,[21] and to produce lower estimated RRs of breast, lung, and thyroid cancers compared with IFRT.[22]
Chemotherapy has facilitated the use of both smaller RT fields and lower radiation doses. Lower radiation doses to the target also result in lower doses to in-field nontargeted normal tissue, which should translate into less toxicity. The recent GHSG HD10 study of patients with favorable-risk stage I/II HL demonstrated that in conjunction with 2 cycles of ABVD, 20 Gy of IFRT was equivalent to 30 Gy of IFRT, with similar rates of freedom from treatment failure and overall survival. Additionally, there were more adverse events and more severe acute toxicity (grade 3/4) in patients who received 30 Gy (8.7%) than there were in those who received 20 Gy (2.9%).[23] The GHSG HD11 study of patients with unfavorable-risk stage I/II HL was a four-armed study comparing two radiation dose levels (20 Gy vs 30 Gy) and two chemotherapy regimens (4 cycles of ABVD vs BEACOPP [bleomycin, etoposide, Adriamycin, cyclophosphamide(Drug information on cyclophosphamide), Oncovin, procarbazine(Drug information on procarbazine), and prednisone(Drug information on prednisone)]).[24] Similar rates of freedom from treatment failure, overall survival, and progression-free survival were observed with 20 Gy and 30 Gy in patients receiving BEACOPP chemotherapy; however, grade 3/4 toxicity was reduced with the lower dose of RT, from 12% to 5.7%. Furthermore, 20 Gy of IFRT following 4 cycles of ABVD chemotherapy was inferior to 30 Gy of IFRT following the same chemotherapy. Thus, the efficacy of the radiation dose appears to be influenced in part by the aggressiveness of the chemotherapy regimen used. In pediatric HL, combined-modality therapy has facilitated radiation dose reduction to 15 Gy and 25 Gy, doses likely to result in significantly reduced risks of late effects from RT.
In addition to reducing the size of the radiation target and the radiation dose, using higher-energy photons and integrating advanced radiation technologies, such as IMRT and PT, can also help reduce the volume of nontargeted normal tissue exposed to radiation.
