Although death rarely can be prevented, the aims of treatment of metastatic cancer are to relieve pain, preserve mobility and function, maintain skeletal integrity, as well as prolong useful life. Asymptomatic lesions may be treated to prevent predictable future symptoms. Palliative therapy should provide control of the treated lesion for the remainder of a patient's life. Ideally, it should be rapid, convenient, and free of immediate side effects and delayed complications. It should not unduly disturb daily activities, be wasteful of a patient's limited time and money, or consume scarce medical resources better employed for curative treatment.
Pharmacologic treatment has improved in recent years with the more effective use of oral opiate analgesics--both short and long acting--as well as use of transdermal, epidural, and continuous infusions and patient-controlled systems. However, pharmacologic treatment does not arrest or reverse the underlying disease process and does not prevent structural deterioration; there is a threat of pathologic fracture and spinal cord or other neurologic compromise, which usually are irreversible.
Orthopedic intervention is valuable for mechanical stabilization of impending or recent pathologic fractures, and total joint arthroplasties in carefully selected patients may relieve pain dramatically and restore function. In skilled hands, spinal surgery for decompression and stabilization is sometimes worthwhile.
Hormonal therapy--both additive and ablative--and chemotherapy are of major value in patients whose tumors are responsive. However, resistance eventually occurs and with chemotherapy, toxicity is often significant and major responses uncommon.
Various modes of therapy frequently may be used in sequence or in combination. In oncology, decisions involving palliative care are often the most difficult and require considerable experience and expertise as well as various consultations among multiple specialist disciplines to achieve optimum benefit. Although endpoints are often lacking or poorly defined, it is hoped that the trend toward additional quality-of-life assessments will bring more rigorous evaluation of this important and often neglected area.
Radiation therapy remains the major treatment modality of bone metastases, frequently providing major pain relief and often arresting or reversing the destructive disease process. Despite the frequency and magnitude of the problem, there is remarkably little agreement as to the time, volume, and dose schedules for optimal palliation. This may reflect the diversity of the problem, the multitude of confounding variables, and the lack of clearly defined endpoints.
The causes of bone pain are numerous and poorly understood. Why should a small lesion sometimes cause excruciating pain, whereas other times involvement of almost the entire skeleton produces little, if any, pain?
Similarly, the response to treatment can be puzzling. Ablative hormonal therapy by orchiectomy or oophorectomy may result in dramatic pain relief by the time a patient awakens from anesthesia. Hemibody irradiation (HBI) also may produce pain relief within 24 hours. Such responses clearly do not depend on cell death and tumor shrinkage within such a brief period. Chemical mediators of the pain response must be involved, but there is little understanding of their mechanism. Furthermore, pain relief does not correlate well with tumor radiosensitivity. For example, in prostate cancer, which is not highly radiosensitive, pain often responds well to single-dose or fractionated radiation.
The largest randomized, prospective study of radiation therapy for bone metastases was conducted by the Radiation Therapy Oncology Group (RTOG) in the 7402 study. Of the more than 1,000 patients entered, nearly 750 patients could be evaluated. More patients in the protracted fractionation arms did not complete therapy as assigned, compared with the other arms.
Study entry was determined by the pain and narcotic medication scores. Study requirements included: a pain or narcotic score of at least four, expected survival of at least 3 months, no prior radiation to the study site, and no new chemotherapy within two weeks of study entry. The pain score was the primary response measure, with the narcotic score a secondary response measure (Table 1). The randomization study design used is shown in aFigure 1.
In the analysis of this study by Tong et al, 90% of patients experienced some relief of pain, and 54% eventually obtained complete pain relief . No significant differences were seen among the various treatment arms. Patients with breast and prostate cancer fared better than patients with lung or other cancer, and patients with pain scores under nine also fared better.
Of all patients who eventually experienced minimal pain relief, 96% did so within the first 4 weeks, and almost 50% of those who experienced complete pain relief did so within 4 weeks after the start of treatment. In 70% of patients, pain relief lasted until death. In the group of patients with a solitary metastasis, a dose of 4,050 cGy was associated with a higher incidence of pathologic fracture. In 1985, this study was reanalyzed by Blitzer using different statistical analysis. He concluded that more highly fractionated regimens with higher doses improved the outcome  (Table 2).
A nonrandomized study by Arcangeli et al in 1989 concluded that more complete pain control and less need for retreatment were obtained by fractionated courses of 40 Gy or more . They found no correlation between fraction size and pain relief. In a prospective, randomized study by Price et al, a single 8 Gy fraction was compared with 30 Gy in 10 daily fractions; equally good pain control was demonstrated, with no difference in response rate or speed of onset or duration . In 1989, Cole compared 24 Gy in 6 fractions over 2 or 3 weeks with a single 8 Gy fraction . No differences were seen in pain scores, but 25% of the patients in the 8 Gy group required retreatment.
Although no fractionated scheme of treatment has been shown to be clearly and consistently superior to another, single-dose treatments of 8 Gy may be equally effective, with the advantages of convenience for the patient and reduced work load.
In the 1970s, following the initial report of HBI by Fitzpatrick and Rider from the Ontario Cancer Institute , a number of reports on this single-dose, large-field technique appeared [7-10]. The RTOG began the 7810 study of single-dose HBI for the palliation of multiple bone metastases from solid tumors. This was a dose-escalating study (Table 3) that used the same pain and narcotic scores as did the 7402 study. It is important to note that dose corrections were applied for increased lung transmission. Earlier studies showed a significant incidence of fatal radiation pneumonitis before this was appreciated [11-13]. Furthermore, as developed in a Rochester pilot study, adequate hydration, steroids, and antiemetics were utilized . The results showed that 73% of patients achieved pain relief . Approximately 67% of patients experienced better than 50% pain relief, and 20% experienced complete pain relief. A total of 50% of responding patients experienced pain relief within 48 hours, and 80% experienced pain relief within 7 days. Retreatment was required in only 50% of patients for the rest of their lives. The safest and most effective doses were 600 cGy for upper HBI and 800 cGy for lower HBI and mid-body irradiation  (Figure 2).
The next study, RTOG 8206, evaluated whether the addition of single-dose HBI to standard fractionated radiation was more effective than local-field irradiation alone . It also explored the possibility that the addition of HBI to local-field irradiation might delay the appearance of new metastases and/or decrease the need for further treatments.
A total of 499 patients entered the study. The time to disease progression was delayed significantly in the local and HBI arm, with only 35% of patients showing disease progression at 1 year compared with nearly 50% of patients showing disease progression in the local only arm of the study (Figure 3). The time to appearance of new metastases was delayed considerably for patients in the local and HBI arms. A total of 50% of patients in the local only arm developed new metastases in 6 months compared with 18 months in patients in the local and HBI arm of the study (Figure 4). The time to new treatment in the targeted HBI was delayed even more significantly (Figure 5). At 12 months, 76% of the patients in the local only group had undergone retreatment compared with 60% of the patients in the local and HBI arms of the study. We concluded that an HBI dose of 8 Gy delays the appearance of new metastases and the progression of existing metastases.
All the outcomes examined showed a significant advantage for HBI. Although survival was not prolonged, no lethal toxicities were noted. As shown in other studies, patients with breast cancer and prostate cancer experienced better survival rates than patients with other malignancies.
This study indicates that HBI might be capable of eradicating micrometastatic disease and eventually could be used as adjuvant therapy in patients at high risk for developing bone metastases. Several nonrandomized reports and studies have provided supporting evidence [17-22].
In an attempt to improve the efficacy of HBI, fractionated, rather than single, doses are being explored . Fractionation may improve the therapeutic ratio over that seen in single-dose treatments. The RTOG recently completed a dose-escalation study (8822) of fractionated HBI in patients with breast and prostate cancer. These particular patients were selected because of their longer survival and improved ability to reach endpoints of significance. Final results are pending analysis.
Studies to date do not show a clear or consistent dose response for fractionated treatment of bone metastases. This applied to both completeness and duration of response. There is evidence that single-dose treatments (of the order of 8 Gy) provide equivalent relief to multidose regimens. This applies to both local and hemibody techniques. Single-dose treatments are obviously advantageous in terms of patient convenience and use of remaining life span. Local-field doses of 8 Gy may be repeated safely if necessary.
The use of sequential hemibody techniques to eliminate micrometastatic disease in an adjuvant setting requires further exploration. To improve the therapeutic ratio of HBI, fractionation studies and the use of biologic response modifiers also are being examined.