The management of patients who have giant-cell tumor of bone primarily has consisted of either wide en bloc resection9,10 or curettage without, or followed by, either autologous bone-grafting or use of cement. The local rates of failure of these two regimens have ranged from 4 percent (one of twenty-six patients10) for appendicular lesions to 42 percent (ten of twenty-four patients12) for axial lesions. Radiation therapy often is recommended when a patient cannot be operated on for medical reasons, when a tumor is technically inoperable, or when an operation would result in major and unacceptable disfigurement. Many such tumors involve the axial skeleton. Postoperative radiation has been used in patients with residual disease, especially when the tumor has recurred.
We describe the efficacy of our treatment in terms of lack of progression of disease, which is defined as a visible decrease or no increase in the size of the clinically or radiographically evident lesion. Accordingly, an increase in the radiolucent area or any other evidence of tumor growth is defined as progression of disease.
The two objectives of the present study were to determine the frequency with which treatment with megavoltage radiation was followed by a lack of progression of giant-cell tumor of bone and to ascertain the rate of treatment-related morbidity, including the development of radiation-induced neoplasms.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
†Departments of Radiation Oncology and Orthopaedic Surgery (H. J. M.), Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts 02114. Please address requests for reprints to Dr. Chakravarti. E-mail address for Dr. Chakravarti: achakravarti@partners.org.
Between March 1973 and March 1992, 300 patients who had giant-cell tumor of bone were managed at our institution. Twenty patients (6.7 percent) were managed with megavoltage radiation, and they form the basis of our study. Fifteen of the twenty patients had an axial skeletal lesion that could not be completely or partially resected without major morbidity or functional impairment. Five patients had an appendicular lesion, which was either recurrent or too extensive to be treated with an operation alone without functional impairment in four of them. The fifth patient (Case 7) had a medical contraindication to operative management.
The clinical, radiographic, and histopathological findings in these twenty patients were reviewed. The diagnosis of giant-cell tumor of bone was made for each patient by the bone pathology section of our hospital. Patients who had Paget disease of bone or the so-called brown tumors of hyperparathyroidism (which resemble giant-cell tumor) had been excluded from the study. However, all patients who had a diagnosis of giant-cell tumor of bone and had been managed with megavoltage radiation during this time-period were included. The mean age of the patients at the time of diagnosis was thirty-nine years (range, fifteen to seventy-two years). Thirteen patients were female, and seven were male. The patients were followed for three to nineteen years (median, 9.3 years). Eleven patients were followed for at least nine years (Table I).
Radiation treatment was administered by high-energy photon beams, with use of a cobalt-60 or two to twenty-five-megaelectron-volt linear accelerator. In six patients, a portion of the treatment was given by 160-megaelectron-volt proton beams (Harvard Cyclotron Laboratory, Cambridge, Massachusetts). Treatment volumes were minimized with use of immobilization devices, Cerrobend blocks (Cerrobend Alloy; Cerrometal Products, Bellafonte, Pennsylvania), wedge filters, and multiple fields. Total tumor doses, administered at 1.8 to 2.0 gray per fraction, ranged from forty gray (Cases 3 and 14) to seventy gray (Case 20). The average total duration of treatment was five to seven weeks.
The first seven patients had a biopsy as the only operative procedure (Table I). A total resection was attempted in thirteen patients, but either the operative margins were histologically positive for disease or the resection ended up being only partial with gross residual disease. Of these thirteen patients, eleven received postoperative irradiation only and two (Cases 9 and 20) received both preoperative and postoperative irradiation.
Of the seven patients who had been managed with radiation alone, one (Case 2) had evidence of progression of the disease at the time of follow-up (Table I). Of the thirteen patients who had had resection and radiation, two (Cases 17 and 20) had progression of the disease. Thus, seventeen of the twenty patients did not have progression of the disease. The response to radiation often was not immediate. The lesions regressed at varying rates; regression was usually detected within several months after treatment.
Seven patients had a lesion of a cervical, thoracic, or lumbar vertebra, and six of them (all except Case 8) were first seen with neurological symptoms. Each of the six patients had a decrease in the neurological symptoms after radiation therapy. One patient (Case 1), who was first seen with pain and stiffness in the neck, had no progression of the tumor and no symptoms in the neck at nine years after treatment. Another patient (Case 15) was first seen with weakness in the left upper limb. She was managed with a posterior arthrodesis of the fourth, fifth, and sixth cervical vertebrae, partial resection of the tumor, and administration of fifty gray of radiation. At seven years after treatment, the patient had no evidence of progression of the tumor. One patient (Case 8) had a lesion of the fourth lumbar vertebra, which was treated with partial resection and fifty gray of radiation postoperatively. No progression of the tumor was evident at six years after treatment. Another patient (Case 9) had complete paraplegia when he was first seen. He was managed with administration of eighteen gray of radiation, partial resection, and administration of thirty-two gray postoperatively, so that the total tumor dose was fifty gray. At eight years after treatment, no progression of the tumor was evident and the patient was able to walk with use of a walker. Another patient (Case 10) had paresthesias in the hand on presentation. He was managed with partial resection followed by forty-nine gray of radiation therapy. At three years postoperatively, the patient had had no progression of the tumor and had no neurological symptoms. Another patient (Case 11) had numbness in the right lower extremity. A laminectomy and partial resection was performed, followed by administration of forty-five gray of radiation. At the three-year follow-up examination, the tumor had not progressed and the patient had good sensation in the lower extremities. Another patient (Case 12) was first seen with severe pain in the left shoulder and axilla. She was managed with an arthrodesis of the first, second, and third thoracic vertebrae and partial resection of the second thoracic vertebra, followed by administration of forty-six gray of radiation. At a follow-up examination thirteen years postoperatively, there was no evidence of progression of the tumor.
Six patients had a lesion in the pelvis. One of them (Case 13) had had a recurrence of a pubic lesion within one year after curettage and bone-grafting. She was managed with fifty gray of radiation, and she had no evidence of progression of the tumor at the time of the follow-up five years later. The other five patients (Cases 2, 3, 4, 19, and 20) had a lesion of the sacrum and were managed with radiation alone or with an operation followed by radiation because of evidence of residual tumor after the resection. Two patients (Cases 2 and 20) had a failure of treatment, which will be described later. One patient (Case 3) was managed with forty gray of radiation after laminectomy of the first, second, and third sacral vertebrae and biopsy. No progression of the tumor was evident at a follow-up examination performed more than ten years postoperatively. One patient (Case 4), who had a tumor with a large soft-tissue component (Fig. 1-A), was managed with radiation alone. At one year after treatment, a computed tomography scan showed complete eradication of the soft-tissue tumor with osseous remodeling (Fig. 1-B). At ten years after treatment, additional osseous remodeling and radiation-induced sclerotic changes were seen (Fig. 1-C). Another patient (Case 19) had partial resection of the second and third sacral vertebrae followed by radiation. At a follow-up examination performed three years after treatment, the patient was healthy and had no progression of the disease.
The lesions in the remaining seven patients were at various other sites. One patient (Case 14) had metastases to the lung from a recurrent lesion of the femur. There was no evidence of progression of the metastatic lesion at fifteen years after treatment with radiation. The recurrent lesion of the femur was treated with partial resection followed by radiation. There was also no evidence of progression of this lesion at fifteen years postoperatively.
One patient (Case 7), who had a lesion in the distal part of the femur, was originally scheduled to have an operative resection with application of bone allograft. However, on the day before the scheduled operation, he had a cerebrovascular accident that resulted in contralateral hemiparesis. Therefore, he was managed with radiation alone, and he had no progression of the disease for seventeen years. However, he sustained several pathological fractures through the site of the femoral lesion and a marked valgus deformity developed. Another patient (Case 16) was alive and well seventeen years after irradiation of a partially resected lesion of the tibia. Another patient (Case 5), who was managed with radiation alone for a lesion of the temporal bone that had recurred after curettage and bone-grafting, had no evidence of progression of the disease at a follow-up examination performed nine years after treatment. Another patient (Case 18) was managed with radiation after partial resection of a lesion of the femur that was a metastasis from a primary lesion of the ischium. She was alive without progression of the disease nineteen years after radiation therapy. Another patient (Case 6) had no progression of the disease seven years after irradiation of an unresected giant-cell tumor of the sphenoid sinus. One patient (Case 17), who had a lesion in the distal part of the radius, had a failure of treatment, as will be described.
Of the three patients who had progression of the disease, one (Case 2) had no evidence of regression at five months. The surgeon interpreted this finding as an absence of short-term response and proceeded to perform a sacrectomy; the resected specimen revealed residual tumor. A second patient (Case 17), in whom curettage of a primary giant-cell tumor in the distal end of the radius had failed, received sixty-four gray of radiation for treatment of a recurrent giant-cell tumor in the soft tissue of the volar aspect of the wrist. This treatment was also ineffective, and twelve months later an amputation of the forearm was performed. A third patient (Case 20) had recurrence after curettage of a giant-cell tumor involving the second and third sacral vertebrae. The recurrent tumor was irradiated with fifty gray before partial resection, which was followed by irradiation with twenty gray (proton beam), for a total of seventy gray. Eight months later, the patient had a repeat failure and a sacrectomy was performed. Five years after the second operation, the patient had no evidence of progression of the tumor.
Giant-cell tumors regress slowly after radiation treatment. This pattern was observed in our patients and has been fully described by investigators at other centers. We describe the outcome of radiation treatment in terms of a lack of progression of the disease, which is conceptually identical to the term local control, which has been commonly used in the literature. Likewise, the term progression of the disease, which is used in the present study, is conceptually the same as the terms recurrence or local failure and is based on clinical or radiographic evidence, or both, of an increase in the extent of the osseous destruction or the soft-tissue mass. Despite the ultimate disappearance of the tumor, restitution and remodeling of the affected bone is, with few exceptions, nearly always incomplete. Furthermore, because of destruction of some cortical bone by the tumor, a pathological fracture may occur before or after radiation treatment, which is an event that, coordinated with the process of fracture repair, may interfere with the assessment of either lack of progression or progression of the disease.
In the era of orthovoltage (approximately 250 kilovolts peak), radiation was commonly administered in multiple courses, each of which was low dose. Rates of disease progression after treatment with orthovoltage radiation have been reported to be as high as 63 percent (twenty-nine of forty-six patients8). The total accumulated doses were often very high and were not infrequently associated with severe radiation damage. Since the mid-1960s to the early 1970s, radiation machines producing x-ray beams with energy higher than one megavolt have been regularly used and the treatment has been administered as a single course with fifty to sixty gray as the total dose. Megavoltage units produce more penetrating beams than do orthovoltage units. Cobalt-60 and linear accelerator units, both yielding megavoltage radiation, have been shown to substantially lower the rates of disease progression. In recent studies, the rate of disease progression after treatment with megavoltage radiation has ranged from 7 percent (two of thirty patients) to three of ten patients, with a rate of 26 percent (nine of thirty-five patients) reported in the largest series (Table II)1,2,6,7,14,15.
Radiation treatment may result in the transformation of a benign giant-cell tumor to a malignant one, or it may initiate a primary bone sarcoma at the site. It frequently is not possible to distinguish between the transformation of giant-cell tumor to a malignant tumor and the induction of a malignant tumor in the mesenchymal tissue. Data on treatment with orthovoltage radiation alone (200 to 250 kilovolts peak) have indicated rates of malignant transformation of as high as four of twelve tumors8. As mentioned, during the era of treatment with orthovoltage radiation, patients usually received several courses of treatment, each of which consisted of a modest dose, but often the cumulative dose was high. The time-period for either malignant transformation or the development of radiation-induced sarcoma is remote from the time of treatment, with a median duration of ten years. Pooled data from more recent studies on the results of treatment with a single course of megavoltage radiation (approximately forty to seventy gray) have indicated a far lower rate of malignant transformation—less than 3 percent (one malignant transformation in thirty-seven patients)—after a mean of ten years of follow-up1,2,6,7,13-15 (Table III). The most likely explanation for this difference between orthovoltage and megavoltage radiation is attributable to the much higher absorbed dose in bone than is indicated by the nominal prescribed dose. This phenomenon results from the physics of absorption of low-energy photons, particularly the photoelectric mechanism whereby energy absorption is directly proportional to the third power (Z3) of the atomic number of the tissue. Hence, tissue with high Z components, such as the calcium of bone, absorbs much more energy per gram of irradiated tissue than does muscle. In contrast, with radiation in the megavoltage range, the Compton effect dominates and energy absorption is independent of Z. Hence, the radiation absorption in muscle, tumor, and bone does not vary as it does with the lower energy.
Two major concerns have been raised with regard to the use of any type of radiation for treatment of giant-cell tumor—that is, the efficacy of the treatment and the frequency of sarcomatous transformation. Giant-cell tumors of bone are uncommon, and few are treated with radiation. Hence, the database with which these points may be addressed is quite small. Studies in which megavoltage equipment, modern imaging techniques, and a single course of treatment with fifty or sixty gray were used have shown a considerably lower rate of local failure and malignant transformation than have studies of the early treatment methods with orthovoltage radiation (Table II). Our patients were managed with a single course of radiation for an average of five to seven weeks. The actuarial result with regard to the lack of progression in our patients after a mean duration of ten years was 85 percent (seventeen of twenty) (Fig. 2).
Progression of the disease after treatment with megavoltage radiation was reported in twenty-five (18 percent) of 136 patients from seven centers in the United States, Europe, and Asia (Table II). The rate of disease progression was 30 percent (nine of thirty patients) after total doses of less than fifty gray and 16 percent (four of twenty-five patients) after doses of fifty gray or more.
In comparison, in one series, 42 percent (ten) of twenty-four patients had progression of disease after curettage alone for treatment of giant-cell tumor of the axial spine12. Complete excision at these axial sites often is not feasible because the loss in function would be unacceptable. In another study, the recent technique of curettage and insertion of cement was followed by recurrence in one (4 percent) of twenty-six patients who had an appendicular lesion10.
Malignant transformation of giant-cell tumor was widely reported after treatment with orthovoltage equipment, especially when multiple courses of treatment were administered. As mentioned, radiation-induced sarcoma of bone often can be difficult to distinguish from this process. Recent reports on treatment with a single course of megavoltage photons have indicated a much lower rate of malignant transformation or induction of sarcoma1,2,6,7,14,15. In our study, in which the mean duration of follow-up was 9.9 years, we found no malignant transformations or radiation-induced sarcomas. In an analysis of pooled data on 136 patients described in recent reports, only one patient had radiation-induced sarcoma and none had malignant transformation (Table III). The patient who had the radiation-induced sarcoma was one (3 percent) of the thirty-seven patients who had been followed for more than nine years.
The median time-period for malignant transformation after treatment with radiation is thought to be approximately ten years. Therefore, although the available clinical data reveal a low rate of malignant transformation after a single course of megavoltage radiation, a longer duration of follow-up is needed to evaluate the true frequency. There is increasing evidence that malignant transformation of giant-cell tumor may occur in the absence of irradiation. It has been reported that giant-cell tumors may undergo malignant transformation after an operation alone4. Brien et al. reported on two patients who had malignant transformation or, in the authors' words, "dedifferentiation of GCT's into osteosarcoma in previously unirradiated sites."4 They hypothesized that certain cell lines in giant-cell tumor of bone are derived from primitive mesenchymal cells that occasionally transform into a sarcoma in the absence of previous irradiation. A recent report from our hospital described a patient who had late development of high-grade malignant fibrous histiocytoma at the site at which giant-cell tumor had been operatively treated eighteen years before12. It remains unclear what proportion of giant-cell tumors that are considered to have undergone radiation-induced malignant transformation actually represents the radiation-independent process of so-called dedifferentiation that was described by Brien et al.4.
Our results confirm those of previous reports that demonstrated high clinical efficacy of megavoltage radiation as therapy for giant-cell tumor1,2,6,7,14,15. As 85 percent of our patients had a lack of disease progression at the time of long-term follow-up, radiation was clearly shown to be effective treatment and merits consideration for patients in whom an operation may cause problems or may result in severe morbidity. This approach is especially applicable to lesions of the axial skeleton and the skull. Operative removal remains the treatment of choice for nearly all giant-cell tumors involving the appendicular skeleton.