|Home | About | Journals | Submit | Contact Us | Français|
Although there have been a few large case series of giant cell tumor (GCT) in the spine and sacrum, the treatment of these lesions remains controversial. We are reporting 23 additional cases of giant cell tumor in the spine and sacrum gathered from our institution and the personal consultation files of the senior author. Ten lesions occurred in the sacrum with an average age of 31 years (range of 13-49) and 13 occurred in the mobile spine with an average age of 39.1 years (range of 13-64). Most patients presented with pain or neurologic deficit at the site of tumor involvement, and symptoms were usually present for many months prior to diagnosis. Six of the sacral GCT patients were treated with pre-operative arterial embolization and intralesional surgical resection, and two developed a recurrence. Two of the sacral GCT patients had an en bloc resection and neither developed a recurrence. One sacral GCT patient was treated only with serial arterial embolization with good disease control. One sacral GCT patient did not receive any treatment. Eleven spinal GCT patients were treated with en bloc surgical resection and two developed a recurrence, the other two spinal GCT patients were treated with intralesional surgical resection and both developed a recurrence. Giant cell tumors of the spine and sacrum should be managed with en bloc resections whenever possible as this provides the greatest chance for cure. When the risk of post-operative neurologic deficit after en bloc excision is high, as in most of our sacral lesions, conservative therapy involving arterial embolization and intralesional resection offers the best results.
Giant cell tumor (GCT) of bone is a rare neoplasm that accounts for approximately 5% of all primary bone tumors in adults.1 GCT most frequently occurs at the end of long bones,2 and the sacrum is the fourth most common site, accounting for between 1.7-8.2% of cases.3-6 Giant cell tumor also occurs in the mobile spine, but this location accounts for only 2-4% of cases.4,7,8 In all locations, the neoplasm occurs most commonly between the ages of 20-45 years of age, and it affects males and females with equal frequency.4
Various treatment methods have been advocated including arterial embolization, curettage, surgical excision, radiation, and cryotherapy.9-13 Treatment is very successful in long bone lesions, but the optimal treatment and medical management of GCT in the spine and sacrum has not been well established. We are presenting our experience with the diagnosis and management of 23 cases of giant cell tumors of the spine and sacrum. We discuss the clinical presentation, treatments received, and outcomes of therapy.
These cases were culled from the IRB approved database maintained by the senior author. A total of 23 cases of giant cell tumor of the spine and sacrum were found. The data collected consisted of clinic notes, operative notes, radiographic images, pathological reports, as well as gross and microscopic imaging.
Ten cases occurred in the sacrum (Table 1). Five of these were female and five were male. The mean age was 31 years (range of 13-49 years). All patients had pain for an average duration of 30 months (range of 1 week to 10 years). The pain was most frequently in the lower back and commonly radiated into the thighs. Neurological symptoms were common, occurring in seven of ten (70%) patients. Symptoms included bowel or bladder incontinence, muscle weakness, perineal hypoesthesia, and erectile dysfunction.
Thirteen cases occurred in the spine above the sacrum (Table 2). Eight of these were females and five were males. The mean age was 39.1 years (range of 13-64 years). Patients #12 and #15 were both treated with intralesional surgical resection at an outside institution and referred to our center following recurrence of their disease. Two spinal GCT patients were asymptomatic until the time of their presentation. Patient #19 was diagnosed following an elective total body CT scan, and patient # 14 presented acutely with compression fracture. The other patients presented with back pain. Neurologic symptoms occurred in six of thirteen (46%) patients. Symptoms included arm or leg weakness, parasthesias, diaphragmatic paralysis, and constipation.
Radiographs or CT scans were available for all patients except for patient #17. The sacral lesions were large, poorly defined lytic masses arising in the central part of the sacrum and spreading to involve both wings (Figure 1). Three patients (#6, #7, and #8) also had an associated soft tissue mass (Figure 2). The most frequent location was the upper two segments of the sacrum, although several lesions were so large that they involved the entirety of the sacrum.
The spinal lesions were poorly defined lytic masses confined to the vertebral body (Figure 3). Five patients (#12, #13, #21, #22, and #23) had vertebral compression fractures ranging from mild collapse (Figure 4) to a complete vertebral plana. Patient #23 had extension of her mass into the surrounding soft tissue. Six tumors occurred in the lumbar spine, four occurred in the cervical spine, and two occurred in thoracic spine. One tumor crossed the cervico-thoracic junction to involve two adjacent vertebrae (Figure 5).
Histologic review was performed on tissue from all cases. A CT guided needle biopsy diagnosis was attempted on six of the ten sacral giant cell tumors and was diagnostic in all six. The remaining four lesions required open biopsy. In the spinal lesions, a CT guided needle biopsy diagnosis was attempted in 7 of the 13 cases. The specimen was diagnostic in all seven cases. The remaining cases were not diagnosed until corpectomy.
Histologic features were typical of giant cell tumor of bone (Figure 6). Lesions in the sacrum occasionally had focal areas of aneurysmal bone cyst transformation. Lesions in the spine were frequently complicated by pathologic fracture. This resulted in considerable reactive bone formation which, on several occasions, led to consideration of an osteoblastic producing neoplasm (Figure 7). Osteoblastoma and aneurysmal bone cyst were not diagnostic considerations in the mobile spine lesions because the posterior elements are usually involved by these processes and our patients only had vertebral body involvement. There was no significant pleomorphism to lead to the consideration of malignancy.
Six out of ten patients with GCT of the sacrum were treated by pre-operative embolization with intralesional surgical resection (Table 1). Three of these patients received adjuvant radiation therapy (patient #2, #3, and #7). The remaining four patients were not treated with intralesional surgical resection. Patient #6 received neoadjuvant chemotherapy with doxorubicin and cis-platin followed by en bloc surgical resection. Patient #8 received an en bloc resection with adjuvant radiation therapy. Patient #9 declined to have surgery and instead elected to have serial arterial embolizations.
The most frequent treatment modality for GCT of the mobile spine was en bloc surgical resection, which occurred in eleven of thirteen (85%) of cases (Table 2). Three of these (patients #11, #16, and #18) also received pre-operative arterial embolization. Two others (patients #14 and #12) also received adjuvant radiation therapy. Patient #19 had had extensive disease and opted for pre-operative embolization and intra-lesional surgical resection. Patient # 23 had extensive soft tissue involvement of her tumor and was treated with intra-lesional resection only.
Average follow-up for the patients with sacral GCT was 31.9 months (range of 0-64 months). Average follow-up for patients with spinal GCT was 50.7 months (range of 0-144 months). No patient in this study died during the follow-up period, and no patients had metastases to the lungs. Four spinal GCT and two sacral GCT patients developed local recurrence during the follow-up period. At final follow-up, five sacral GCT and six spinal GCT patients had a complete neurologic recovery. The remainder of the patients had either chronic pain or residual neurologic deficits (Tables 1 and and22).
Few studies have presented large case series of GCT in the spine or sacrum.14-20 In this study, we have characterized the clinical, histologic, and radiographic presentation of this rare tumor and reported our experience with the treatment and management of 23 cases. Overall, two of nine (22%) sacral GCT and four of 13 (31%) spinal GCT patients who received surgical therapy at our institution developed a recurrence. In contrast, recurrence in long bone giant cell tumor is about 10%.1 Thus, spinal and sacral GCT carry a much worse prognosis.
Clinically, both spinal and sacral GCT patients in our study had a similar presentation. The most common symptoms were pain and neurologic deficit in the area affected by the tumor, which was present in nearly all patients. All but two patients in this study had pain for months prior to diagnosis, one of which was an incidental finding, while the other presented with an acute compression fracture. Like GCT of long bones, GCT of the spine and sacrum can develop the so-called “benign lung metastasis,” and this event has been reported to occur in up to 13.7% of the spinal lesions.21 However, none of our patients developed lung metastases.
Radiographically, sacral GCT has been reported to be an expansile lytic lesion involving both sides of the midline, without a sclerotic rim.22 Lesions in our study were similar, and three patients had adjacent soft tissue masses. Spinal GCT has been reported as an expansile lytic lesion that most often involves the vertebral body, with roughly equal incidence in all parts of the spine.23,24 In our study, all patients had lytic masses in the vertebral body, and only one patient had soft tissue involvement. Six lesions (46%) occurred in the lumbar spine, four (31%) occurred in the cervical spine, two (15%) occurred in the thoracic spine, and one (8%) crossed the cervico-thoracic junction. While the epiphyseal-metaphyseal location of giant cell tumor in the long bone is a clue to the radiographic diagnosis, these landmarks are not available in the spine and sacrum. As a result, the radiographic changes are less diagnostic in the spine.
Treatment of GCT in the long bones has typically involved curettage, sclerotherapy, and filling of the defect with bone cement.3 However, therapy of sacral and spinal lesions is less strait forward. Giant cell tumors in these locations frequently do not present until late in the course when the tumor is very large and when wide excision would leave the patient with unacceptable neurologic deficits. Although treatment algorithms have been proposed,12,13 the optimal treatment of GCT in the sacrum and spine is not well defined.
In our study, multiple different treatment modalities were applied with varying degrees of success. In the sacrum, two patients had tumors that were small enough to be amenable to complete en bloc surgical excision. Neither patient suffered a recurrence and both were completely disease free at final follow-up. This is consistent with prior studies that have shown en bloc excision to produce excellent rates of local control.25 Notably, both patients treated with this modality had residual neurologic deficit at final follow-up. This is consistent with prior studies that showed en bloc surgical excision to hold significant risk for permanent neurologic deficit.26-28
In the spine, 11 of our patients were treated with en bloc surgical excision, and two of 11 (18%) developed a local recurrence. The recurrence rate of spinal GCT following en bloc excision has been reported to range from ll-50%.15,24 Our results indicate that the true recurrence following en bloc excision is at the lower end of the reported ranges. Six of the 11 patients had no residual neurologic deficit at final follow-up, indicating that en bloc excision produces better results in the spine than it does in the sacrum. Given these results, en bloc surgical excision should be the treatment of choice for patients with spinal GCT unless the operation would result in significant post-operative morbidity.
Six of our sacral GCT patients were treated with pre-operative embolization followed by intralesional surgical resection. Two of these six (33%) had a recurrence of their disease. Previous studies have shown recurrence rates ranging from 29%,16 to 50%.29 Our results indicate that the true recurrence is at the lower end of these reported ranges. Only one patient treated with pre-operative embolization and intra-lesional resection had any residual neurologic deficits at follow-up. This is consistent with prior studies which have show intralesional resection of sacral GCT to be associated with lower post-operative morbidity than en-bloc resections.16 Thus, we have found that this modality produces a reasonably low recurrence rate with excellent preservation of neurologic function. Pre-operative embolization followed by intralesional resection should be the treatment of choice for sacral GCT patients with tumors that are too large for en bloc excision.
Only two of our spinal GCT patients were treated with intralesional resection. In both cases, the patients had extensive disease at the time of their presentation, and both patients developed a local recurrence. The rates of local recurrence following intralesional resection of spinal GCT are poorly established, but are reported to range from 0%-71%.17,20,24,30 It is likely that this high recurrence rate may be more a reflection of the initial disease burden than the surgical technique itself. For spinal GCT in the mobile spine, this modality should be reserved for patients with extensive disease who cannot be treated with en bloc excision.
A recent study showed no benefit of adjuvant radiation therapy following conservative surgical management of sacral GCT25 However, this remains controversial, and some authors still advocate its use following intralesional resections.10,12,13 In our study, none of three sacral GCT patients treated with intralesional resection and adjuvant radiation developed a recurrence; whereas two of three treated with intralesional resection alone developed a recurrence.
In summary, GCT of the spine and sacrum is a rare tumor that most frequently presents with pain and neurologic deficit at the site of involvement. None of our patients developed benign metastases, suggesting that the rates of this complication may be less than previously reported. Whenever possible, en bloc excision should be pursued as the surgical procedure of choice for management of spinal GCT. When en bloc excision is prohibited due to the high risk of post-operative morbidity, as is usually the case with sacral lesions, pre-operative embolization followed by intralesional resection should be the procedure of choice. Serial arterial embolization is an alternative modality that has a lower morbidity, and can be offered as an initial therapy in some patients. In contrast to the most recently published studies,25 we found an anecdotal benefit of adjuvant radiation therapy for GCT of the sacrum. We also found that CT guided needle biopsy is a valuable tool in establishing a preop-erative diagnosis.