Skeletal metastasis is an increasingly frequent problem related to the longer life expectancy of patients with cancer. The spinal column is the most frequent site of bone metastasis; between 30% and 70% of patients with cancer will have evidence of spinal metastases at autopsy.1–3 Surgery performed on patients with spinal metastases has been shown to be an excellent treatment for spinal instability, neurological complications, and severe pain that cannot be managed conservatively or by local radiation.4,5 In palliative surgery for spinal metastasis, direct decompression and stabilization with instrumentation are recommended because the results of sole laminectomy are not favorable.6,7 However, direct decompression of the metastasis can be associated with a risk of excessive or life-threatening bleeding.8,9 Several studies have described the technical aspects of preoperative embolization9–11 and reported its efficacy on perioperative blood loss.12–15 However, no study has measured its effect on intraoperative blood loss in a single palliative surgery compared with a control group without preoperative embolization.
The current study evaluated the efficacy of preoperative embolization on intraoperative blood loss in palliative decompression and instrumented surgery using a posterior approach for thoracic and lumbar spinal metastasis. A posterolateral transpedicular approach with posterior instrumentation is widely used in the surgical treatment of vertebral body metastases.5,16
Material and Methods
This retrospective study included 46 consecutive patients (28 men and 18 women; mean age, 61.0 years [range, 18–80 years]) with thoracic and lumbar spinal metastasis who underwent surgery between 2000 and 2010. All patients had existing or impending neurological deficit due to spinal canal involvement.
Each patient underwent a palliative decompression and instrumented surgery using a posterior approach as the primary surgery for spinal metastasis. Patients who were treated with anterior or combined approaches or without instrumentation were excluded. No patient had bleeding diathesis. One to 4 vertebral levels were decompressed and 5 to 7 vertebrae were stabilized using posterior instrumentation.
Preoperative embolization was performed on 23 patients (embolization group: 15 men and 8 women; mean age, 62.8 years) (Table 1). In the embolization group, the primary tumors were in the lung (n=6), breast (n=4), kidney (n=4), prostate (n=2), stomach (n=2), and other organs (n=5). Surgery was performed within 3 days after embolization. The other 23 patients did not undergo preoperative embolization (no embolization group: 13 men and 10 women; mean age, 59.2 years) (Table 2). In the no embolization group, the primary tumors were in the lung (n=7), prostate (n=3), colon (n=3), breast (n=2), thyroid (n=2), kidney (n=2), and other organs (n=4). In the 2 groups, no significant differences existed in the factors that could affect the degree of intraoperative blood loss (Table 3).
Table 1: Clinical Data and Results in the Embolization Group
Table 2: Clinical Data and Results in the No-embolization Group
Table 3: Patient Data
In the authors’ institution, preoperative embolization in radical surgery for spinal tumors has been routinely performed and is considered effective in reducing intraoperative blood loss. However, embolization in palliative surgery for spinal metastasis has not been routinely performed, mainly because the palliative surgeries were performed in an emergency situation.
Patients underwent diagnostic angiography and embolization under local anesthesia using the femoral approach. Throughout the thoracic and lumbar levels, paired segmental arteries arise at each vertebral body level, with the exception of the upper thoracic spine, where the superior intercostal arteries should be assessed in addition to the supreme intercostal arteries (arising from the costocervical trunk). Selective catheterization of the corresponding segmental arteries, including 2 levels above and 2 levels below the tumor site, was performed, followed by selective angiography of the suspected feeding arteries. Depending on the tumor vascularization on the arterial angiogram, all lesions were graded by 2 independent radiologists (T.M., O.M.): mild for weak tumor blush after contrast injection, moderate for medium tumor blush, and extensive for significant tumor blush. Between 2000 and 2004, polyvinyl alcohol particles ranging in size from 125 to 500 μm and gelatin sponges, a temporary vascular occlusive agent, were used as embolic materials. Coils have been used in combination with polyvinyl alcohol to obliterate the segmental artery feeding the tumor in 14 patients since 2005. If the postembolization tumor stain was reduced by more than 90%, embolization was rated as fully successful (complete) and otherwise as partially successful (partial). The 9 embolizations without coils prior to 2005 were rated as partial.
All 46 patients underwent surgical decompression and instrumented stabilization using a single posterior approach. When the tumor involved the vertebral body, anterior decompression following laminectomy was performed by means of a posterior transpedicular approach. When the tumor involved only the posterior and lateral elements, posterolateral decompression was performed by laminectomy and tumor excision. Intraoperative blood loss estimated by the anesthesiologists was obtained from the surgical records.
Statistical analysis was performed with SPSS version 16.0 for Windows software (SPSS, Inc, Chicago, Illinois), and the significance level was set at a P value less than .05. The t test, chi-square test, and multiple comparison using the Dunnett T3 post-hoc test were used.
All embolization procedures were performed in a single session without causing neurologic deficit or skin or muscle necrosis. Tumor vascularization on the arterial angiogram was mild in 8 (34.8%) patients, moderate in 10 (43.5%) patients, and extensive in 5 (21.7%) patients. One to 3 segmental vessels were embolized in 4 (17.4%) patients, and 4 or more segmental vessels were embolized in 19 (82.6%) patients. Embolization was rated as complete in 7 (30.4%) patients and partial in 16 (69.6%) patients. In 7 (30.4%) patients, the tumor-feeding artery also supplied a radiculomedullary artery and was not embolized; this was a common reasons for partial embolization. Surgery was performed within 24 hours after embolization in 12 (47.8%) patients and more than 24 hours after embolization in 11 (52.2%) patients.
Pain and neurologic symptoms in all 46 patients were relieved postoperatively. Mean intraoperative blood loss in the embolization group was 520 mL (range, 140–1380 mL) and 1128 mL (range, 100–3260 mL) in the no embolization group (P<.05) (Table 4). In the embolization group, average intraoperative blood loss was 304 mL (range, 140–580 mL) with weak tumor vascularization, 544 mL (range, 380–1270 mL) with moderate tumor vascularization, and 820 mL (range, 440–1380 mL) with extensive tumor vascularization. These values were significantly lower than those in the no embolization group (P<.05). However, no significant differences existed between the groups. Average intraoperative blood loss was 347 mL (range, 150–480 mL) after complete embolization and 596 mL (range, 170–1380 mL) after partial (incomplete) embolization, but the difference was not statistically significant. Even in partial embolization, intraoperative blood loss was significantly lower than that in the no embolization group. Average intraoperative blood loss was 535 mL (range, 140–1270 mL) with embolization within 24 hours preoperatively and 505 mL (range, 170–1380 ml) with embolization more than 24 hours preoperatively, with no statistical difference.
Table 4: Summary of Results
A 57-year-old man presented with severe back pain and paraparesis due to spinal cord compression with T12 metastasis from renal cell carcinoma (Figure 1). In spinal angiography before embolization, tumor vascularization was extensive and the artery of Adamkiewicz was originating from the same pedicle as the tumor, which was not embolized (Figure 2). The other 4 segmental vessels were embolized successfully, but the embolization was rated as partial. Three days after embolization, palliative posterior surgery was performed. Intraoperative bleeding was well controlled, and circumferential decompression of the spinal cord and stabilization with posterior instrumentation at T10-L2 was successfully achieved (Figure 3). Operative time was 268 minutes, and intraoperative blood loss was 1380 mL, which was the highest in the embolization group. The patient’s pain and neurologic symptom improved immediately postoperatively. He died 6 months postoperatively without symptomatic recurrence.
Figure 1: T2-weighted sagittal magnetic resonance image showing T12 metastasis causing spinal cord compression (A). T2-weighted axial magnetic resonance image of the T12 pedicle (B).
Figure 2: Preoperative angiogram of the left 12th intercostal artery showing significantly hypervascular tumor stain and feeding to the artery of Adamkiewicz (arrow).
Figure 3: Intraoperative photograph (A) and postoperative anteroposterior (B) and lateral (C) radiographs of palliative circumferential decompression and instrumented T10-L2 surgery via a posterior approach.
The use of selective arterial embolization to treat bone tumors was first described in 1975 by Feldman et al.17 Many radiologists then reported the efficacy of preoperative arterial embolization on perioperative blood loss for spinal metastasis.9,10,12–15 The amount of intraoperative blood loss may vary according to the operative procedure, the vascularity of the tumor, the completeness of the embolization, and the timing between embolization and surgery. The limitations of most studies on preoperative embolization include the use of various surgical procedures and the lack of control groups.
In most patients with spinal metastasis, the metastatic site is the thoracic and lumbar spine, and palliative posterior surgery with instrumentation is widely applied. Therefore, the current authors evaluated the effectiveness of preoperative embolization in palliative decompression and instrumented surgery using a posterior approach for thoracic and lumbar spinal metastasis. They hypothesized that the effect of embolization would be most significant when tumor excision was performed from a posterior approach where prior ligation of the segmental vessels would not be performed before tumor excision.
As per general principles of transarterial embolization, the “back door” of the segmental artery must be occluded to deliver embolic particles to the target neovascular territory without risk of nontarget embolization to distal territories.10 This embolization is best done using embolic coils. Although segmental arteries tend to supply a single segmental distribution, bimetameric or trimetameric trunks can occur, typically in the thoracic region. In these situations, embolization of all segmental arteries downstream from the target embolic site using embolic coils is required before beginning particle embolization. This maneuver has been used in the authors’ institution since 2005 and has led to successful embolization.
Some studies have described excessive intraoperative bleeding after incomplete preoperative embolization.13,18 Other studies have described no significant association between the embolization grade and intraoperative blood loss.14,19 Complete embolization was achieved in 30% of cases in the current study, compared with 50% to 81% in comparable studies.9,13,18 These differences may be explained by the current authors’ strict criteria of complete embolization and cautious embolization technique: embolization was aborted when doubt existed concerning potential spinal feeders. This strategy is supported by the result that the intraoperative blood loss after partial embolization was significantly lower than that without embolization.
Guzman et al15 reported that no significant difference existed in intraoperative blood loss among 3 groups divided according to the degree of tumor vascularization. A similar finding was observed in the current study and further proves that preoperative embolization is effective.
Several authors have described the optimal timing between embolization and surgery.9,20 The main concerns are arterial recanalization and collateral blood flow establishment. Surgery performed within 72 hours of embolization (optimally within 24 hours) has been shown to result in decreased perioperative blood loss compared with surgery performed beyond this time period.9,20 In the current study, all surgeries were performed within 72 hours of embolization, and no difference existed in intraoperative blood loss between surgeries performed within 24 hours and at more than 24 hours.
This study had some limitations. Various primary tumor types were included in the 2 cohorts. However, all patients had a single posterior surgery, and no significant differences existed between the 2 groups in terms of demographics (Table 3). Surgeons knew whether the patients underwent preoperative embolization; therefore, an inherent bias existed where the surgeons may have paid more attention to intraoperative blood loss in the embolization group. In addition, the study was not randomized or prospective. It was a retrospective comparative study (level of evidence, III). Despite this design limitation, preoperative embolization was found to be significantly effective in reducing intraoperative blood loss during palliative surgery for spinal metastasis.
Preoperative embolization of spinal metastases is safe and effective in restricting preoperative perfusion; intraoperative blood loss can be significantly reduced. In patients who underwent palliative decompression and instrumented surgery using a posterior approach, intraoperative blood loss after preoparative embolization was less than the half that without preoperative embolization.
- Abrams AD. The management of cancer metastatic to bone. JAMA. 1994; 272(15):1206–1209. doi:10.1001/jama.1994.03520150074040 [CrossRef]
- Fornasier VL, Horne JG. Metastases to the vertebral column. Cancer. 1975; 36(2):590–594. doi:10.1002/1097-0142(197508)36:2<590::AID-CNCR2820360240>3.0.CO;2-F [CrossRef]
- Ortiz Gomea JA. The incidence of vertebral body metastases. Int Orthop. 1995; 19(5):309–311.
- Sundaresan N, Galicich JH, Bains MS, Martini N, Beattie EJ Jr, . Vertebral body resection in the treatment of cancer involving the spine. Cancer. 1984; 53(6):1393–1396. doi:10.1002/1097-0142(19840315)53:6<1393::AID-CNCR2820530629>3.0.CO;2-0 [CrossRef]
- Ibrahim A, Crockard A, Antonietti P, et al. Dose spinal surgery improve the quality of life for those with extradural (spinal) osseous metastases? An international multicenter prospective observational study of 223 patients. J Neurosurg Spine. 2008; 8(3):271–278. doi:10.3171/SPI/2008/8/3/271 [CrossRef]
- Hall AJ, Machay NN. The results of laminectomy for compression of the cord or cauda equina by extradural malignant tumor. J Bone Joint Surg Br. 1973; 55(3):497–505.
- Sorensen PS, Borgensen SE, Rohde K, et al. Metastatic epidural spinal cord compression. Cancer. 1990; 65(7):1502–1508. doi:10.1002/1097-0142(19900401)65:7<1502::AID-CNCR2820650709>3.0.CO;2-D [CrossRef]
- Hess T, Kramann B, Schmidt E, Rupp S. Use of preoperative embolization in spinal metastatic resection. Arch Orthop Trauma Surg. 1997; 116(5):279–282. doi:10.1007/BF00390053 [CrossRef]
- Berkefeld J, Scale D, Kirchner J, Heinrich T, Kollath J. Hypervascular spinal tumors: influence of the embolizaiotn technique on preoperative hemorrhage. AJNR. 1999; 20(5):757–763.
- Heran MK. Preoperative embolization of spinal metastatic disease: rationale and technical considerations. Semin Musculoskelet Radiol. 2011; 15:135–142. doi:10.1055/s-0031-1275596 [CrossRef]
- Shi HB, Suh DC, Lee CS, et al. Preoperative transarterial embolization of spinal tumors: embolization techniques and results. AJNR. 1999; 20(10):2009–2015.
- Gellad FE, Sadato N, Numaguchi Y, Levine AM. Vascular metastatic lesions of the spine: preoperative embolization. Radiology. 1990; 176(3):683–686.
- Manke C, Bretschneider T, Lenhart M, et al. Spinal metastases from renal cell carcinoma: effect of preoperative particle embolization on intraoperative blood loss. AJNR. 2011; 22(5):997–1003.
- Prabhu VC, Bilsky MH, Jambhekar K, et al. Results of preoperative embolization for metastatic spinal neoplasms. J Neurosurg Spine. 2003; 98(2 suppl):156–164. doi:10.3171/spi.2003.98.2.0156 [CrossRef]
- Guzman R, Dubach-Schwizer S, Heini P, et al. Preoperative transarterial embolization of vertebral metastases. Eur Spine J. 2005; 14(3):263–268. doi:10.1007/s00586-004-0757-6 [CrossRef]
- Cho DC, Sung JK. Palliative surgery for metastatic thoracic and lumbar tumors using posterolateral transpedicular approach with posterior instrumentation. Surg Neurol. 2009; 71(4):424–433. doi:10.1016/j.surneu.2008.02.049 [CrossRef]
- Feldman F, Casarella WJ, Dick HM, Hollander BA. Selective intra-arterial embolization of bone tumors. A useful adjunct in the management of selected lesions. Am J Roentgenol Radium Ther Nucl Med. 1975; 123(1):130–139.
- Roscoe MW, McBroom RJ, St Louis E, Grossman H, Perrin R. Preoperative embolization in the treatment of osseous metastases from renal cell carcinoma. Clin Orthop Relat Res. 1989; (238):302–307.
- Jackson RJ, Loh SC, Gokaslan ZL. Metastatic renal cell carcinoma of the spine: surgical treatment and results. J Neurosurg. 2001; 94(1 suppl):18–24.
- Barton PP, Waneck RE, Karnal FJ, Ritschl P, Kramer J, Lechner GL. Embolization of bone metastases. J Vasc Interv Radiol. 1996; 7(1):81–88. doi:10.1016/S1051-0443(96)70738-8 [CrossRef]
Clinical Data and Results in the Embolization Group
|Patient No./Sex/Age, y||Primary Tumor||Localization||Degree of Embol||Time From Embol to Surgery, h||Degree of Vascularization||Intraoperative Blood Loss, mL||Operative Time, min|
Clinical Data and Results in the No-embolization Group
|Patient No./Sex/Age, y||Primary Tumor||Localization||Intraoperative Blood Loss, mL||Operating Time, min|
|Characteristic||Embolization Group||No-embolization Group|
|No. of patients||23||23|
|No. of men/women||15/8||13/10|
|Mean age, y||62.8±11.1||59.2±15.3|
|No. of thoracic/lumbar||19/4||22/1|
|No. of hypervascular tumor typesa||5||6|
|Mean No. of decompressed levels||2.4±.8||2.6±.9|
|Mean No. of instrumented levels||5.3±.9||5.6±1.0|
|Mean operative time, min||272±54||291±71|
Summary of Results
|Intraoperative Blood Loss, mL||Operative Time, min|
|Completeness of embolization|
|Time from embolization to surgery|
| Within 24 h||12||535±340a||275±60|
| >24 h||11||505±311a||270±49|