A comprehensive review was initiated on hemibody irradiation for patients with solid tumors with an emphasis on bone metastases to determine the efficacy and the toxicity of this treatment.
Lawrence B Berk, MD, PhD
A comprehensive review was initiated on hemibody irradiation for patients with solid tumors with an emphasis on bone metastases to determine the efficacy and the toxicity of this treatment.A literature search was done on PubMed for “radiotherapy” or “radiation therapy” and “hemi-body” or “half-body.” The studies were analyzed for their endpoints of efficacy, usually pain relief, and for toxicity. The treatment method, including toxicity prevention methods, were also extracted. Because of the lack of common endpoints and poor reporting methods of the authors in the time period involved, no statistical evaluation was done.A total of 29 relevant patient series reported in 33 publications were abstracted. There were 6 prospective trials (2 phase III, 3 phase II and 1 phase I), 5 prospective series and 18 retrospective series. This comprised a total of 1841 patients treated with hemibody irradiation. Of the 1627 defined primaries, 46% were prostate, 32% were breast and 8% were lung. Most patients were treated with 8 Gy lower or mid hemibody irradiation and/or 6 Gy upper hemibody. There was no uniform reporting among the articles, and often they reported only partial or complete relief. A weighted average among 1005 patients with sufficient information reported 32% complete and 53% partial pain relief. Hemibody irradiation can also delay progression of bone metastases. Toxicity was primarily nausea and hematologic.Hemibody irradiation is an apparently efficacious treatment with acceptable toxicity given in simple anterior-posterior fields. Its efficacy may be improved with even less toxicity using modern techniques.
Oncology has become increasing expensive without an equivalent return in value. For example, Ra-223 treatment has been adopted as a standard treatment for men with bone metastases with a full course of treatment costing about $70,000. In contrast, hemibody irradiation is an older treatment that may deliver the same efficacy for a very small fraction of the cost. Hemibody irradiation is a simple radiation therapy technique in which patients are treated with large anterior and posterior fields encompassing half of the body. The original publication on the technique was from Fitzpatrick at Princess Margaret Hospital in Toronto in 1978.1 He described the results of “half-body” irradiation of 5 Gy to 10 Gy (uncorrected for tissue inhomogeneity) given on Co-60 or a Clinac 35 MV accelerator. The researchers shielded the eyes and did not treat the skull if there were no metastases in the skull, and similarly they omitted the lower legs if there was no disease present. The reason the investigators chose to omit the skull was to avoid hair loss. One hemibody field was treated and then, after the blood counts returned to the pretreatment levels (which was generally about 28 days) the alternate half was treated. Outcomes of 82 patients were reported, primarily breast cancer patients, and found that lower hemibody irradiation had few or no side effects and the primary toxicities of upper hemibody irradiation were nausea, vomiting and diarrhea. These usually started about one hour after treatment. One week after irradiation stomatitis and vulvitis were noted in some patients. The researchers found hematologic problems with sequential, bilateral hemibody irradiation only when the time between treatments was less than 28 to 35 days. Two autopsy proven cases of radiation pneumonitis developed after the 16th week, 1 at 8 Gy and 1 at 10 Gy. The investigators concluded1:
There have been many subsequent retrospective and prospective studies of hemibody irradiation. However, the technique is now using very sparingly if at all. This review is intended to re-present the efficacy of hemibody irradiation as a cost-effective, well tolerated, palliative treatment for patients with extensive bone metastases.A PUBMED search was conducted for all languages with the keywords “radiotherapy AND hemibody OR half-body.” The articles were reviewed for other relevant references. If several articles referred to the same patients, then the most comprehensive article was used. Only trials designed for the palliation of solid tumors were accepted. Articles that focused primarily on myeloma or other non-solid tumors were not included in this review. The patients were then analyzed as either retrospective or prospective reports, and if prospective whether only prospective or a planned phase I, II or III trial. The primary endpoints for this analysis were pain relief and toxicity. No statistical analysis was done because of the poor reporting of both pain relief and toxicity in the majority of papers. Therefore, a semi-quantitative approach to the data analysis was used.A total of 29 series encompassing 1833 patients with treatment for symptomatic bone metastases were found in the literature. Table 1 shows the studies and their results. Nineteen retrospective trials reported on 1066 patients. Not all studies identified the primary tumor sites of the patients, but in those that did, the majority had prostate cancer (A%) with (B%) of the remaining patients having breast cancer, C% with non-small cell lung cancer and then D% of patients with other solid tumors. Consistently, the response of the patients was classified in terms of the patient’s pain, and the patients were classified as having no response, partial response or complete response. As shown in Table 1, the complete response rate was about 25% with a range of 61% to 5%. The partial response rate was about 70%.
The prospective trials are better designed and give more precise data. In these trials, shown in Table 1, the complete response rate is about 20% and the partial response rate is about 75%. The definitions of response among the trials differ, with a major difference being whether complete response includes the need to use pain medication or not.
Salazar et al reported on RTOG 78-10.2 This was a phase I (safety) and phase II (efficacy) trial. It utilized the pain measures which had been created for a previous RTOG trial for focal irradiation of bone metastases (RTOG 74-02). They used a self-designed endpoint by scoring the severity and frequency of the pain on a scale of 0 to 3, and then these were multiplied together (thus the total score was 0 to 9). A similar scale was made based on the type of pain medication used and the frequency of its use. Table 2 shows the pain scale. To be eligible for RTOG 78-10 patients had to have multiple painful bone metastases and a pain or narcotic score of 4 or greater. The patient could only have pain in the half of the body to be treated. The patients were followed for hematological toxicity, nausea and vomiting, diarrhea, pneumonitis, and fever. The patients were evaluated for toxicity and response (pain score and narcotic score) daily for the first week, weekly for the next 2 months and monthly thereafter. The study had 108 patients enrolled and 91 were evaluable. Patients receiving upper hemibody irradiation were hospitalized for 24 hours and received a premedication program with hydration, anti-emetics and corticosteroids. Lower and mid-hemibody irradiation patients were treated as outpatients and given prophylactic anti-emetics. The treatments were: lower hemibody—bottom of L4 to ankles; mid hemibody–top of the diaphragm to bottom of the obturator foramina; upper hemibody–bottom of L4 to above the top of the head. Patients were treated with anterior and posterior fields. The maximal subjective response was classified as: no response (NR), some relief with less than 50% relief (SR), partial response with more than 50% relief (PR) and complete pain relief (CR). In this trial, there was no dose effect between single doses of 6 Gy (upper hemibody only), 8 Gy, 9 Gy, or 10 Gy, with 21% complete relief, 45% partial relief, 11% some relief, and 23% no response. Forty percent of patients responding did so within 24 hours, 50% within 2 days, 80% within a week, and 97% by 2 weeks. There was a trend towards more rapid pain relief with higher doses. Toxicity was manageable, as shown in Table 3, with some mild-to-moderate nausea and mild-to-moderate hematologic depression. The follow up RTOG phase III trial (RTOG 8206) looked at treating asymptomatic patients with bone metastases with hemibody irradiation and is discussed below. There were a subsequent Phase I/II RTOG trial adding the putative radiosensitizing agent misonidazole to single fraction hemibody irradiation but pain relief was not a measured endpoint and it will not be discussed.3
Mill and colleagues published a trial of 21 patients (6 patients with breast cancer, 5 with lung cancer, 4 with myeloma, 3 with prostate, 2 with adenoid cystic and 1 with base of tongue).4 The upper hemibody field went from the iliac crests (generally the L4-L5 interspace) to above the vertex of the skull. The lower hemibody field went from the top of the iliac crest to the ankles. Seventeen patients had a total of 21 hemibody treatments; 9 upper hemibody treatments and 12 lower hemibody treatments. This included 3 patients with both upper and lower hemibody treatment and one with lower hemibody treatment twice. They did not have a good pain rating system which limits the results. Seventeen treatments (not patients) produced good response.
Nag and Shah published a prospective trial of 2 fractions of 8 Gy lower hemibody irradiation given one week apart for symptomatic bone metastases.5 Eighteen patients were treated, 6 with multiple myeloma, 6 with breast cancer, 4 with lung cancer and 3 with prostate cancer. They used the RTOG pain and narcotic response criteria described above. They reported that 10 patients had complete relief and 9 patients had partial relief with most patients having pain relief within 24 to 48 hours. The median duration of pain relief was 5 months. Ten of 15 patients were pain free at death. Four patients were still alive and were pain free.
Berg et al reported on a phase II trial of hemibody for patients with painful bone metastases.6 Upper hemibody extended from the second cervical vertebral body to the iliac crests (which is approximately the L4-L5 interspace). Lower hemibody was from the iliac crests to the knees. The upper hemibody was treated to 7 Gy in 1 fraction with partial shielding of the lungs and kidneys to reduce the dose to 6 Gy. The bowels were also protected starting below the ribs. Patients were hospitalized and premedicated. Patients were followed for quality of life with EORTC QLQ-C30 and with a 7-point pain scale. No significant change in the QLQ-C30 occurred after treatment. The major toxicity from treatment was diarrhea which was greatest at week 2 and returned to baseline by week 8. Maximal pain relief was seen at 4 weeks after treatment with 8% complete pain relief and 68% partial pain relief. One third of patients were able to reduce their pain medication use.
The final phase II trial is by Pal and colleagues.7 In this study from India, 23 patients (11 prostate, 6 breast, 6 lung) were treated on a Cobalt-60 unit. All patients had widespread bone metastases and were on morphine for pain. The upper hemibody field went from the angles of the mandible to the umbilicus. The lower hemibody field went from the umbilicus to just below the knees. Six patients had upper hemibody treatment, 5 patients had lower hemibody treatment and 10 patients had both upper and lower hemibody treatment. Pain assessment was with an 11 point visual analog scale (VAS) and a verbal rating scale (VRS) of 0 (none) to 4 (excruciating). The authors’ presentation of results is difficult to interpret. The investigators state that, overall, 67% of the patients had partial relief of their pain and 22% had complete relief. The pain relief lasted for the entire 6 months of the trial. There were no grade 3 or higher toxicities.
There have been 4 phase III trials involving hemibody irradiation. The first was a phase III study, RTOG 8206, which studied hemibody irradiation in patients with widespread bone metastases after local irradiation of symptomatic bone metastases.8 Patients had to have metastases limited to the area of an either upper, middle or lower hemibody irradiation field. All of the patients’ painful bone metastases were treated with 30 Gy in 10 treatments and then the patients were randomized to either no further radiation therapy or hemibody irradiation of 8 Gy. The primary endpoint was time to disease progression. A total of 499 patients were randomized and there were 450 analyzable cases. Thirty-three percent of the patients had prostate cancer, 27% had breast cancer and 24% had lung cancer. At one year [or Year 1?] 46% of the patients without hemibody irradiation had progression and only 35% of the patients with hemibody irradiation had progression (P = 0.032). Progression delay within the treated area was even more significant. At one year [is this Year 1?], 68% of the patients without hemibody irradiation had progression whereas on 50% of the treated patients had progression (P = 0.0039). There was no overall survival difference. Only 12 patients (5%) had grade 3 complications, including 10 with leukopenia and there were 2 cases of grade 3 nausea/vomiting and 2 cases of grade 3 diarrhea (multiple morbidities in the same patients accounts for a total of only 12 patients overall).
The second randomized trial was a comparison of hemibody and intravenous strontium-90 treatments for patients with prostate cancer.9 In this trial for patients with metastatic prostate cancer, patients were evaluated for appropriateness of local or hemibody treatment (local irradiation or hemibody irradiation) and the patients were randomized between local irradiation or strontium-89 injection or between hemibody irradiation or strontium-89 injection. The hemibody technique was not described. Six Gy was administered to the upper hemibody or 8 Gy to the lower hemibody. In the hemibody group of 80 patients, about 30% of the patients at 12 weeks had substantial or dramatic improvement of the overall pain score based on the assessment method used in the trial (ie, increased mobility, decreased pain, decreased narcotic use), and 67% had some pain relief. Pain relief was equivalent in the hemibody and strontium-89 arms.
The third phase III trial was from the International Atomic Energy Agency (IAEA) and reported on by Salazar et al.10 This was a trial of various fractionations of hemibody irradiation for patients with widespread, symptomatic bone metastases. The three arms were 15 Gy in 5 fractions over 5 days, 8 Gy in 2 fractions in one day, and 12 Gy in 4 fractions over 2 days. This international study randomized 156 patients. Pain response was measured in a similar fashion as in the early RTOG trials by multiplying severity and frequency. The distribution of cancers was 46% breast, 32% prostate, 9% lung, and the remaining 16% were miscellaneous cancers. Forty-four percent of the patients had upper hemibody irradiation, 66 had mid hemibody irradiation and 51% were lower hemibody irradiation. They found that 45% of the patients had complete pain relief and 46% had partial pain relief. The toxicity was graded as 40% none, 50% mild/moderate and 12% severe but transitory, and was more common with upper hemibody irradiation. During upper hemibody irradiation the most common toxicities were nausea and vomiting and hematologic. During lower hemibody irradiation the most common toxicities were diarrhea and hematologic. They found that the white blood cell counts and platelets nadirs were at 1 to 2 weeks and returned to pre-treatment baseline in 6 to 8 weeks. The 8 Gy in 2 fractions arm had the lowest biologically effective dose and the poorest response rate.
The final RTOG study was also studying delay of progression. RTOG 88-22 was a phase I/II study of fractionated hemibody irradiation for patients with a single symptomatic bone metastasis and multiple asymptomatic bone metastasis confined to a single hemibody field.11 The patients were then treated to the painful site and received hemibody irradiation in 2.5 Gy fractions of a total of 10, 12.5, 15, 17.5, or 20 Gy. There was no significant dose effect in toxicity. Overall, among the 142 patients 37 had grade 3 hematologic toxicities and 11 had grade 4 (no grade 5). Six patients had grade 3 GI toxicity, 2 had grade 4, and none had grade 5. There was 1 grade 5 (death) pulmonary toxicity at the 12.5 Gy dose but none at higher doses. Thus the higher doses were well tolerated, but there was no increased time to progression when compared to RTOG 8206.Sequential hemibody radiation therapy (ie, the planned sequential treatment with upper and lower hemibody irradiation to treat the entire body) was also explored for use as adjuvant therapy for small cell lung cancer. None showed a survival advantage for sequential hemibody irradiation. The trials will therefore be reviewed focusing on the toxicity of the treatments.
Hüttner reported in 1989 on a 3-arm randomized clinical trial for patients with extensive stage small cell lung cancer comparing sequential hemibody irradiation to systemic chemotherapy to local irradiation.12 The hemibody irradiation was 8 Gy upper hemibody given as a 6 Gy dose in the morning and a 2 Gy dose 5 hours later. Then lower hemibody irradiation was given 6 weeks later with the same dosing. The chemotherapy arm had the best survival. Among the 31 patients in the sequential hemibody arm they report that 6.5% of the patients had grade 3 leukopenia after upper hemibody irradiation and 16% had grade 3 leukopenia after lower hemibody irradiation. Many of the patients also had 1 to 2 days of generalized symptoms, such as nausea, weariness, dry mouth and anorexia.
Bonner reported in 1995 on a trial combining chemotherapy and hemibody irradiation.13 In this trial, quoting the article: “The 20 enrolled patients received 7 cycles of cyclophosphamide-based chemotherapy. The first cycle consisted of cyclophosphamide, doxorubicin, etoposide, vincristine, and lomustine. Subsequent cycles used a regimen of doxorubicin alternating with cisplatin. Thoracic radiation was delivered in a split-course fashion during the first week of chemotherapy cycles 5 and 6 (2000 cGy in five fractions during each week). Prophylactic cranial radiation was delivered in a split-course fashion during the first week of chemotherapy cycles 2 and 3 (1700 cGy in 5 fractions during each week). After the 7 cycles, patients received 600 cGy upper hemibody radiation followed by 800 cGy lower hemibody radiation.” Not surprisingly they found significant hematological toxicity and some neurological toxicity but otherwise the treatment was well tolerated. Among the 19 patients there were 4 with grade 3 nausea, 4 with grade 3 vomiting, 3 with grade 3 acute peripheral neurological toxicity, and 1 each with acute esophagitis and chronic cardiac toxicity and chronic central neurological toxicity.
Powell reported in 1986 on 24 patients in a trial for limited stage small cell lung cancer.14 Their patients received 2 cycles of induction cyclophosphamide, doxorubicin and vincristine (CAV), and then 6 Gy upper hemibody irradiation (above the skull to L4-L5 interspace) and then cyclophosphamide and vincristine followed 3 days later by 6 Gy lower hemibody irradiation (L4-L5 interspace to just below the knees). Patients with a partial or complete response then had further CAV. The protocol initially required a second upper hemibody dose during consolidation but only 2 patients received this second dose. The patients also received prophylactic cranial irradiation of 20 Gy in 10 treatments. Only one third of patients had a platelet nadir of less than 50,000 and 25% had white blood cell count nadirs of 1000 to 2000 and 8% had a nadir below 1000. A third of patients had moderate esophagitis and 4% (1 patient) had severe esophagitis. Only 13% of patients had severe nausea and vomiting. There was also mild to moderate temporary neurotoxicity due to the vincristine in 9 patients and severe neurotoxicity in 1 patient. There were no treatment related deaths.Forty years of published results document both the efficacy and safety of hemibody irradiation for the treatment of wide-spread symptomatic bone metastases and the prevention of progression of subclinical bone metastases. A weighted average of partial and complete pain relief for patients treated with hemibody irradiation yields a 32% complete pain relief rate and another 53% with partial pain relief. And yet, the treatment is rarely given and instead patients are treated with more expensive treatments, such as systemic radionuclides. This technique has been shown in a randomized trial to be equipotent for the relief of bone metastases with strontium-89 injection.9 Both gave about 65% sustained pain relief at 3 months after treatment. There was increased toxicity in the hemibody irradiation arm—nausea, vomiting, and diarrhea–but these toxicities were transitory and can be better prevented today than in the early 1990s. For example, as shown in Figure 1, the dose to the normal structures such as the bowel, kidneys, liver, lungs and brain, can be significantly lowered using modern treatment planning and treatment. Also, hemibody irradiation has been shown to significantly delay disease progression within the treated area and it has also been shown to be compatible with aggressive concurrent chemotherapy.
Even using the more expensive for of radiation therapy, IMRT, for treatment, the cost of the treatment with single fraction hemibody irradiation or sequential hemibody irradiation will be a small fraction of the currently most commonly used treatment radionuclide, Ra-223 (Xofigo). It is estimated that a single injection of radium-223 will cost over $11,000 per treatment and the recommended course of 6 treatments will be $69,000.15 In contrast, a single IMRT treatment including planning and treatment is approximately $1500 in the US. Thus, even a full course of sequential hemibody irradiation costs about one quarter the cost of a single treatment and less than 5% of the cost of a full course of Ra-223.
Six doses of Ra-223 have been shown in a randomized trial funded by the manufacturer gives a modest survival advantage of 3.6 months to patients with bone metastases from prostate cancer (14.9 month vs. 11.3 month median survival). This is a crude cost of $230,000 per increased year survival. A more detailed analysis including the use of quality adjusted life years (QALY) and ancillary costs of treatment concluded that the cost was approximately $79,000 per QALY.16 The data presented in this review show that sequential hemibody irradiation is a tolerable treatment option even when combined with high dose chemotherapy. It is reasonable to hypothesize that hemibody irradiation without concurrent chemotherapy will be well tolerated, and allow for future chemotherapy. Therefore, it may be a viable and much more cost effective approach than Ra-223 to this group of patients. If sequential hemibody irradiation is equivalent to Ra-223 in prolonging survival of men with bone metastases from prostate cancer, the crude cost per increased year of survival would be approximately $17,000 per year of life gained, a reduction of $210,000 from the crude cost of Ra-223.
Given the overall effectiveness and more specifically the cost-effectiveness of hemibody irradiation, its use should be re-introduced into general oncologic practice. Further, a clinical trial is warranted to compare the efficacy of sequential hemibody irradiation to Ra-223 for patients with bone-only metastatic disease.This review shows as well as possible with the limited available data that hemibody irradiation is an effective treatment for symptomatic widespread bone metastases from a variety of primary cancers. It is reasonable to expect 20% to 30% of the patients will have an excellent reduction in pain and 40% to 50% will have a significant reduction. The patients had minimal toxicity with the pre-treatment methods available during these trials, and with modern supportive care should have even less toxicity. The patients did not suffer permanent leukopenia with sequential (whole body) radiation therapy even when combined with aggressive concurrent chemotherapy and so hemibody irradiation should be compatible with prior or subsequent chemotherapy should be tolerable. Therefore, hemibody irradiation should be brought to the forefront in the palliative treatment of patients with widespread bone metastases. Also, there should be randomized trials of hemibody irradiation for the palliation of widespread bone metastases in modern oncology patients. And specifically, there should be a randomized trial of Ra-223 treatment versus sequential hemibody irradiation for metastatic prostate cancer with a primary endpoint of survival.