Journal of Surgical Oncology 2014;110:123–128
Percutaneous Vertebroplasty for Painful Spinal Metastasis With Epidural Encroachment GANG SUN,
MD, PhD,*
LI LI, MD, PENG JIN,
MD,
XUN-WEI LIU, MD, AND MIN LI, MD
Department of Medical Imaging, Jinan Military General Hospital, Jinan, Shandong Province, China
Background and Objectives: Spinal metastasis with epidural encroachment is regarded by several authors to be a contraindication to percutaneous vertebroplasty (PVP) because of the risk of increasing symptomatic leakage of cement. This analysis aims to evaluate the safety and efficacy of PVP in patients with painful spinal metastasis and encroachment of epidural space. Methods: A retrospective study was conducted to review 43 consecutive patients with spinal metastasis that underwent PVP, for a total of 69 affected levels. All patients had at least 1 level associated with epidural encroachment related to metastasis. Among these patients, 14 had signs of spinal cord or cauda equina compression. Pain intensity was scored on a visual‐analog scale (VAS). The analgesic efficacy was defined as at least 50% improvement in pain score as compared with the pre‐procedure baseline and post‐procedure. Clinical improvement of neurological compressive symptoms was defined as a decrease in ASIA impairment scale from baseline of 1 point or more. Results: The analgesic efficacy was achieved in 89.7% of survival patients at 1 month, 87.5% at 3 months, 86.9% at 6 months, and 84.6% at 1 year. Small amount cement leakages were detected in 69.6% of treated levels without clinical complications. No deterioration of spinal cord or cauda equina compression symptoms was observed after a PVP in any patients. The different grade of epidural encroachment of the lesions was not correlated with filling volume or extraosseous leakage (P > 0.05). The treated levels with epidural encroachment showed a statistically significant relationship to spinal‐canal leakage (P < 0.05). Conclusions: PVP can be performed safely and effectively in patients with painful spinal metastasis and epidural encroachment.
J. Surg. Oncol. 2014;110:123–128. ß 2014 Wiley Periodicals, Inc.
KEY WORDS: osteolytic metastasis; spine; pain; interventional procedures; percutaneous vertebroplasty
INTRODUCTION Percutaneous vertebroplasty (PVP) involves the injection of a biomaterial, usually polymethyl methacrylate (PMMA), into the involved vertebral body under fluoroscopic guidance. Reinforcement of the bone and stabilization of the spine relieves pain and prevents pathological fractures [1–4]. The principal indications for this procedure are osteolytic vertebral metastasis and myeloma, vertebral hemangioma, and painful osteoporotic vertebral collapse. Symptomatic spinal cord compression is usually considered as a contraindication to PVP due to the potential risk of aggravating the symptoms. The destruction of the posterior vertebral wall with absence of direct‐compressive symptoms or evidence of tumor‐epidural encroachment is considered to be a relative contraindication to the procedure [5–7]. Despite new developments in surgery, medical oncology, and radiotherapy, the treatment option is still a matter of debate for painful spinal metastasis in patients with epidural encroachment in whom anticipated survival was limited [8]. Earlier investigations have evaluated the analgesic and stabilizing effects of PVP in spinal metastasis without epidural encroachment [6,7,9]. However, to our knowledge, there have been only three reports of PVP in vertebral lesions with epidural encroachment, with or without neurologic deficit [10–12]. The experience of the procedure in treating such patients remains limited. We have successfully performed PVP in spinal metastasis with epidural encroachment with or without neurologic deficit. The purpose of this study was to review our experience on painful spinal metastasis in 43 patients with epidural encroachment treated by PVP and evaluate the efficacy and safety of the procedure.
MATERIALS AND METHODS Selection Criteria We performed a retrospective review of all patients with painful spinal metastasis who underwent vertebroplasty in our hospital between
ß 2014 Wiley Periodicals, Inc.
March 2000 and May 2012. During this period, 412 consecutive patients with painful spinal metastasis were treated using PVP at our hospital. Approval for this study was obtained from the hospital review board at our hospital for this retrospective study, and all patients had previously consented to the use of their medical records for research purposes. The medical files from all cases were reviewed for evidence of effacement of the epidural space around the spinal cord and compression of the cord. Patients were excluded from consideration for this study if they were treated for a spinal metastasis without epidural effacement. From this group, 48 patients with at least 1 vertebral lesion associated with epidural effacement related to metastatic spinal disease were identified. Five of these patients were excluded because of lack of available imaging data for retrospective review. A total of 43 patients (24 women, 19 men, with mean age of 64.1 years, range of 34–84 years) were included in this study. Primary tumor types were lung (n ¼ 13), breast (n ¼ 12), kidney (n ¼ 5), stomach (n ¼ 4), bladder (n ¼ 3), uterine (n ¼ 3), colon (n ¼ 2), and ovary (n ¼ 1). The patients underwent PVP for management of persistent pain not responsive to conservative therapy, including various analgesics and chemotherapeutic regimens. The decision to perform
Grant sponsor: National Nature Science Foundation of China; Grant number: 51073173; Grant sponsor: National High Technology Research and Development Program of China; Grant number: 2013AA032203. Conflict of interest: All authors stated that they had no conflicts of interest. *Correspondence to: Gang Sun, PhD, MD, Department of Medical Imaging, Jinan Military General Hospital, No. 25, Shifan Road, Jinan, Shandong Province 250031, China. Fax: þ86‐531‐51666486. E‐mail: [email protected] Received 12 November 2013; Accepted 25 February 2014 DOI 10.1002/jso.23608 Published online 24 March 2014 in Wiley Online Library (wileyonlinelibrary.com).
124
Sun et al.
PVP was made by the attending radiologist in conjunction with referring physicians and surgeons after review of clinical and radiographic data in all patients who were not candidates for surgery due to the patient’s poor condition contraindicating the operation. Radiation therapy was not indicated in 27 patients who had already received the maximum radiation dose for local lesions. PVP procedures were performed pre‐ or post‐radiotherapy for rapid stabilization of the vertebrae in 16 patients due to delayed and incomplete action of radiation therapy.
Parameters for Analysis All patients underwent preoperative, unenhanced CT and MRI scans, focusing on the location, and extent of the lesions and their relationship to epidural space and spinal cord. The vertebral structure (osteolytic, osteosclerotic, or mixed) and presence of an osteolytic lesion on the posterior vertebral body wall were evaluated on plain CT scans. The evidence of the epidural space encroachment around the spinal cord, or compression of the spinal cord, was assessed on MRI scans. To analyze the relationship of the grade of epidural encroachment and post‐PVP changes in imaging data, we classified the grade of epidural encroachment or spinal cord involvement as follows: no epidural encroachment (group A), mild epidural encroachment without spinal cord deformity (group B), and epidural encroachment with spinal cord deformity (group C), depending on a modified epidural spinal cord compression scale suggested by Bilsky and Shimony [10,13]. This classification scheme is illustrated in Figure 1. The American Spinal Injury Association (ASIA) impairment‐scale designation was used in grading the degree of the severity of the neurologic deficit [14]: A (complete), no sensory or motor function is preserved in the sacral segments S4–S5; B (sensory incomplete), sensory but not motor function is preserved below the neurological level and includes the sacral segments S4–S5; C (motor incomplete), motor function is preserved below the neurological level and more than half of key muscle functions below the single neurological level of injury have a muscle grade less than 3 (grades 0–2); D (motor incomplete), motor function is preserved below the neurological level and at least half (half or more) of key muscle functions have a muscle grade of 3 or more; and E (normal), normal motor and sensory function. Pain level with activity was assessed according to the visual‐analog scale (VAS) score; a score of 0 indicated no pain and a score of 10 indicated the most pain imaginable.
PVP Procedure PVP procedures were performed under strict, sterile conditions using C‐arm fluoroscopy for guidance. The patients were placed in a prone position on the table and continuously monitored for blood pressure, electrocardiogram, and pulse‐oxygen saturation during the procedure. Conscious sedation and analgesia were obtained with IV flunitrazepam
(Versed, Roche Pharmaceuticals, Shanghai, China) and buprenorphine hydrochloride (Institute of Pharmaceutical Research, Tianjin, China). The patients were kept alert enough so they could state if any pain developed during the procedure. Under local anesthesia, a 13‐gauge, beveled trocar needle was used for vertebrae treated above T7 through a parapedicular (transcostovertebral) route, and an 11‐gauge, beveled trocar needle was for vertebrae treated between T8‐L5 through a transpedicular route. With fluoroscopic guidance, the needle was advanced to the center of the lesion at the vertebral body. When the needle placement was completed, vertebral venography through the needle was performed with 3–5 ml of half‐strength contrast agent (a 50:50 mixture of Iohexol Injection and normal saline) to evaluate needle positioning. If the needle was in, or immediately adjacent to, the basivertebral vein or another larger draining vein, it was repositioned. After confirming ideal needle placement, cement was prepared as described elsewhere [15]. In brief, the cement material was prepared by combining polymethylmethacrylate (PMMA) powder with sterile barium sulfate for opacification, followed by the addition of a liquid monomer to create a smooth‐liquid consistency. Cement was injected into the vertebral body with a screw‐type 10 ml syringe under fluoroscopy in the lateral projection. In the earlier, injected cement was the consistency of toothpaste for 58 levels of 32 patients, whereas in recent years, it was in the consistency of toothpaste for 5 levels with no epidural encroachment and a smooth‐liquid consistency for 11 levels with epidural encroachment in 11 patients. If cement leakage was detected, the injection was temporarily halted to allow the cement to harden, and resumed after changing the direction of the bevel and the position of the needle tip to fill as much of the lesion as possible. The stopping point was determined when the cement reached the posterior quarter of the vertebral body. Injection was immediately terminated when the patient complained of any pain, such as radicular pain, that could be due to pressure on posterior neurologic structures. If the cement mixture filled only one side of the vertebral body from a single pedicle injection, the contralateral pedicle was accessed for further filling of the vertebral body. A biopsy was routinely performed before cement injection. In patients with multiple spinal metastases, a PVP at corresponding levels was performed routinely in a one‐stage operation. Post‐procedural steroid therapy (1.5 mg per kilogram of body weight of prednisone equivalent, once per day for 5 days) was given for patients with symptoms of partial cord compression. A selective thoracolumbar nerve root block with lidocaine, ropivacaine, and dexamethasone was performed unilaterally or bilaterally in patients with pre‐existing and persistent nonrelieved post‐ procedure radiculopathy [16].
Clinical Evaluation All clinical, imaging, and laboratory data of patients who underwent PVP were reviewed by the operators (G.S. or P.J.). Patients were evaluated clinically before and immediately after treatment, and 1 week
Fig. 1. Scheme shows classification of epidural space encroachment. a: Scheme shows normal epidural space. b: Scheme shows mild ventral epidural space encroachment without spinal cord deformity. c: Scheme shows ventral epidural space encroachment with spinal cord deformity. Journal of Surgical Oncology
PVP for Spinal Metastasis Epidural Invasion later. During the period, patients’ acute post‐procedure adverse events, ASIA‐impair scale, and VAS score with activity were evaluated. Further, VAS scores and ASIA‐impair scale were obtained at 1, 3, 6, and 12 months. Effective pain relief was defined as at least a 50% improvement in VAS scores as compared with baseline. Clinical improvement of neurological compressive symptoms was defined as a decrease in ASIA impairment scale from baseline of 1 point or more. A CT scan of the spine was obtained immediately after PVP to assess cement distribution and extraosseous leakage in treated levels.
Statistical Analysis Statistical analysis was performed with statistic software (SPSS, version 16.0 for Windows; SPSS, Chicago, IL). Categorical variables are frequencies or percentages. A Fisher exact test and Pearson Chi‐ squared test were performed for qualitative variables, and a Mann– Whitney test was used for quantitative variables. A P value of less than 0.05 was considered to indicate a significant difference. The cumulative proportion of analgesic efficacy after PVP was evaluated by Kaplan– Meier analysis, with the exclusion of dead patients.
RESULTS Sixty‐nine levels (range T3‐L5 levels) in 43 patients were judged suitable for this study. The number of metastatic lesions in each patient varied from 1 to 4 levels: 24 patients (24/43, 56%) had 1 lesion level, 13
125
patients (13/43, 30%) had 2 lesion levels, 5 patients (5/43, 12%) had 3 lesion levels, and 1 patient (1/43, 2%) had 4 lesion levels. Each patient had at least 1 level with epidural encroachment (Fig. 2). In total, 23 levels with normal epidural space were classified into group A, 23 levels with mild ventral epidural space encroachment and no spinal cord deformity were classified into group B, and 21 levels with ventral epidural space encroachment and spinal cord deformity were classified into group C (Table I). Of the 43 patients, 14 had signs of spinal cord or cauda equina compression. One of these patients presented complete impairment (ASIA scale A) at the T3 level, and 13 presented incomplete impairment (ASIA scale C or D) at T3 (n ¼ 1), T11 (n ¼ 2), T12 (n ¼ 2), L1 (n ¼ 3), L4 (n ¼ 3), L5 (n ¼ 2), respectively. PVPs were performed successfully with a unipedicular approach in 32 levels and a bipedicular approach in 37 levels. The amount of cement injected per vertebral body varied between 2.0 and 7.0 ml (mean 5.2 1.3 ml). On post‐procedural CT images, the percentage of lesion filling with bone cement was more than 50% in all levels. Leakages of small amounts of cement were detected in 69.6% (48 of 69) levels, which were localized at the intervertebral disk in 21.7% (15/69), epidural space in 20.3% (14/69), and other location in 27.5% (19/69) (Table II). This different grade of epidural encroachment of the vertebral lesions was not correlated with filling volume or with the risk of extraosseous leakage (P > 0.05). The treated vertebral levels with epidural encroachment showed a statistically significant relationship to spinal canal leakage by univariate analysis (P < 0.05) (Table III). No cement leakage was symptomatic with onset of spinal cord or cauda equina compression
Fig. 2. Sagittal T1‐weighted (a), T2‐weighted MR images (b), Sagittal (c), and axial CT views (d) show metastatic pulmonary carcinoma of the spine and compression fracture of L1 vertebral body with epidural encroachment and destruction of posterior wall (arrows) in 61‐year‐old man. Patient had local pain and signs of spinal cord compression with complete impairment (ASIA scale D) at L1 level. Lateral fluoroscopic radiograph (e) and Sagittal CT images (f) after injection PVP showed a good deposition of the cement in the lesion (arrows). Note: cement leakage of the epidural space seen on axial CT image (g) after the procedure (arrowhead). Journal of Surgical Oncology
126
Sun et al.
TABLE I. Characteristics of Treated Levels Treated level
Group A (n ¼ 25)
Group B (n ¼ 23)
Group C (n ¼ 21)
1 2 1 0 1 1 0 0 3 1 3 1 3 2 6
0 1 0 1 1 2 2 1 2 2 2 2 2 4 1
1 0 0 2 1 1 1 1 2 2 3 1 1 3 2
TABLE III. Statistical Results of the Clinical, Radiologic, and Technical Parameters P value
Variable T3 T4 T5 T6 T7 T8 T9 T10 T11 T12 L1 L2 L3 L4 L5
Volume of cement Group A vs. B Group B vs. C Group C vs. A Cement leakage Group A vs. B Group B vs. C Group C vs. A Volume of cement Epidural leakage Group A vs. B Group B vs. C Group C vs. A
0.982a 0.845a 0.897a 0.838b 0.426b 0.539b 0.033c 1.000c 0.017c
a
Mann–Whitney test. Pearson Chi‐squared test. c Fisher exact test. b
symptoms after procedures. Systemic complications, such as bleeding, pulmonary embolism, infection, stroke, or cardiac arrest, were not encountered in this series. Selective nerve‐root blocks were performed before PVP procedures in five patients with pre‐existing radiculopathy. Three of the five reported persistent post‐procedure radiculopathy. They received other selective nerve‐root block on the second day after the PVP procedure. No patients required surgery. The mean VAS score with activity was 7.6 (range 6–10) preoperatively as a baseline. Effective pain relief, based on 50% or greater improvement in VAS scores as compared with baseline, was rated at 93% (40/43) at 1 day and 95% (41/43) at 1 week. In this period, six patients with incomplete neurological impairment had improved neurological compression symptoms. None of the 43 patients reported worsening spinal cord or cauda equina compression symptoms. Three patients reported no change in radiculopathy immediately following the procedure and were treated with selective nerve‐root blocks, which resulted in improvement in pain levels. In two patients, however, the radiculopathy eventually returned to pre‐PVP levels at 1 week following the procedure. The time of the follow‐up among the 43 patients was biased in favor of a short period because of the short life expectancies in this population. The final follow‐ups prior to patients’ death were made at the following times: 39 patients at 1 month, 32 at 3 months, 23 at 6 months, and 13 at 1 year. Among all patients, effective pain relief was achieved in 35 (89.7%) at 1 month, 28 (87.5%) at 3 months, 20 (86.9%) at 6 months, and 11 (84.6%) at 1 year (Table IV). The cumulative proportion of effective pain relief after PVP was evaluated by the Kaplan–Meier survival curve (Fig. 3). During the following‐up period, 19 patients reported recurring painful symptoms after a period of improvement. The painful symptoms in the spine were located at PVP‐treated levels in 12 patients, and at
TABLE II. Treatment Data, and Cement Leakage Parameter Vertebroplasty Approach Unipedicular Bipedicular Total cement leakage Discal Paravertebral Foramina Epidural Access track
Journal of Surgical Oncology
No. of vertebrae 69 46 23 48 15 12 3 14 4
(66.7) (33.3) (69.6) (21.7) (17.4) (4.4) (20.3) (5.8)
levels other than PVP‐treated areas in 7 patients. Of the six patients who had experienced improved neurological compression symptoms at 1 week after the procedure, two were unchanged and four had deteriorated at the final clinical follow‐up. The reason may have been related to disease progression.
DISCUSSION Patients with severe, intractable pain caused by spinal metastasis are left with few pain‐control options after the failure of conservative treatment. Although radiotherapy may result in partial or complete pain relief, many patients have either late onset of pain relief, or limited or no response to the therapy [17]. Open surgery may be not available in these patients with multiple metastases due to a poor general medical status, limited life expectancy because of the long recovery period, and high morbidity and mortality. PVP remains a useful option for eliminating or decreasing pain from vertebral lesions that cannot be controlled with conservative treatment [18,19]. However, the procedure bears increased risks for severe symptomatic complications from leakage of cement into the spinal canal and neural foramina, especially if the posterior cortex of the vertebral body is destructive. Usually, the presence of destruction of the posterior vertebral wall, with or without spinal compressive symptoms, has been considered as a contraindication, or relative contraindication, to the procedure [5–7]. In this retrospective study, we reviewed patients in which PVP was performed at levels with epidural encroachment from spinal canal metastases. The study showed that PVP was performed successfully in all patients. The analgesic efficacy of PVP in survival patients, defined as at least a 50% improvement in pain VAS scores as compared with baseline, was achieved in most survival patients in this study. The incidence of cement leakage in PVP reported in the literature ranges from 9.2% to 73% [20–22]. One study reported the detection of leakage was different based on different imaging modalities, and the cement leakage rate was often underestimated by radiography compared with CT [23]. The incidence of cement leakage in this study was 62.3% (42 of 69) in treated levels on post‐procedural CT images. The comparatively high leakage rate in the present study may be related to the fact that leakages were observed by CT and that many lesions involved cortical bone. Although no significant differences in extraosseous leakage were observed between levels with no epidural encroachment and levels with epidural encroachment in our study, spinal‐canal leakage was associated with epidural encroachment with and without spinal cord deformity. The spinal‐canal leakage observed in the present study could partly correspond to the fact that destruction of
PVP for Spinal Metastasis Epidural Invasion
127
TABLE IV. Follow‐Up of Patients After Vertebroplasty Parameters No. of cases Effective pain relief No effective pain relief
1‐Week follow‐up
1‐Month follow‐up
3‐Month follow‐up
6‐Month follow‐up
1‐Year follow‐up
43 41 (95) 2 (5)
39 35 (89.7) 4 (15.4)
32 28 (87.5) 4 (18.8)
23 20 (86.9) 3 (14.3)
13 11 (84.6) 2 (15.5)
the posterior vertebral body wall or epidural extension of the tumor reduces the ability to prevent cement releasing into spinal canal. Another possible reason is that the injected cement, with a smooth‐liquid consistency rather than a toothpaste consistency in the partially treated levels, may have diffused and deposed in the lesion of least resistance inside the vertebra and prevented it from pushing the epidural tumor mass to the spinal cord. In levels with a malignant soft‐tissue mass involved in the anterior part of the spinal canal, the cement commonly diffuses the soft‐tissue mass in the spinal canal. Cortical destruction and presence of an epidural soft‐tissue mass may be a factors that are likely to increase the rate of complications. However, Weill et al. [18] found that complications of PVP for vertebral tumors are not more frequent when there is destruction of the posterior cortex of the vertebral body. From the data in our study, no complication with clinical sequelae was observed after PVP in the patients with epidural encroachment, which is similar to other authors’ results in vertebral metastases with epidural encroachment [10,11]. Also noted is that the foraminal leakages of small amounts of cement observed in three cases in this study were not responsible for any complications, in line with results from the previous studies [10,23] in which no correlation was found between foraminal leakages and neuralgia. A possible explanation is that too little cement leaked to cause compression of anatomic structures and produce clinical symptoms. In general, the leakage rate of cement and the analgesic efficacy in this study are not substantially different from other studies in patients with metastatic disease and multiple myeloma that excluded patients with clinical evidence of spinal cord compression. For example, Cotten et al. [9] reported that pain status improved in 75% of patients at 6‐month follow‐up and cement leakage was detected in 72.5% of patients in a series of 37 patients with osteolytic metastases and myeloma. Our data
indicate that the increased risk of clinical complications related to cement leakage is small, and there are substantial benefits for patients with few other pain‐control options, severe pain, and a short life expectancy. On the basis of our data, we believe that painful spinal metastasis with epidural encroachment, and even spinal cord compression, should not preclude performance of PVP. With care, the procedure can be safely performed by experienced operators on this group of patients, even in clinical situations that have been cited in the literature as a contraindication or a relative contraindication. It is worth noting that the follow‐up results in this population indicated that many patients reported a pattern of early benefit followed by a subsequent deterioration in their condition near the end of life. Patients increased physical activity and improved quality of life temporarily. It is reasonable to assume that PVP is a palliative treatment for pain relief and spine stabilization concerning spinal metastasis but has a limited antitumor effect, which can not impede the disease progression. PVP combined with multidisciplinary procedures of anticancer modality have shown promising results [8]. The need of PVP for patients suffered spinal metastases with epidural encroachment should be consulted by a multidisciplinary team including orthopedic surgeons, neuroradiologists, medical oncologists, radiation oncologists, and interventional radiologists. Our study has several limitations. Most importantly, the study was retrospective in nature. Secondly the small size of patient samples and the short life spans of most patients might mask the benefits and risks of the procedure. A much larger group of patients might reveal a higher complication rate, especially treatment failure. In addition, patients with epidural encroachment in whom PVP is indicated are relatively few in number compared with the majority of cancer patients without epidural encroachment where PVP is sufficient. In conclusion, the PVP procedure is especially valuable in the treatment of malignant vertebral fractures. Our study confirms that epidural encroachment, and even spinal cord compression with epidural encroachment, should not preclude performance of PVP in patients with few other treatment options and a short life expectancy. Considering the risk of complications of the procedure in this type of patients, it is highly recommended that patients should be properly screened and the procedure be performed by an experienced specialist.
ACKNOWLEDGMENTS This study was supported by Grants No.51073173 from National Nature Science Foundation of China (NNSFC) and No.2013AA032203 National High Technology Research and Development Program of China (863 programme).
REFERENCES
Fig. 3. Kaplan–Meier curve of pain relief after PVP, with exclusion of dead patients. X‐axis corresponds to time after PVP procedure. Journal of Surgical Oncology
1. Pilitsis JG, Rengachary SS: The role of vertebroplasty in metastatic spinal disease. Neurosurg Focus 2001;11:e9. 2. Alvarez L, Perez‐Higueras A, Quinones D, et al.: Vertebroplasty in the treatment of vertebral tumors: Postprocedural outcome and quality of life. Eur Spine J 2003;12:356–360. 3. Legroux‐Gerot I, Lormeau C, Boutry N, et al.: Long‐term follow‐up of vertebral osteoporotic fractures treated by percutaneous vertebroplasty. Clin Rheumatol 2004;23:310–317. 4. Wenger M: Vertebroplasty for metastasis. Med Oncol 2003;20: 203–209.
128
Sun et al.
5. Cotten A, Boutry N, Cortet B, et al.: Percutaneous vertebroplasty: state of the art. Radiographics 1998;18:311–320. 6. Murphy KJ, Deramond H: Percutaneous vertebroplasty in benign and malignant disease. Neuroimaging Clin N Am 2000;10:535– 545. 7. Heini PF, Walchli B, Berlemann U: Percutaneous transpedicular vertebroplasty with PMMA: Operative technique and early results. A prospective study for the treatment of osteoporotic compression fractures. Eur Spine J 2000;9:445–450. 8. Bartels RH, van der Linden YM, van der Graaf WT: Spinal extradural metastasis: Review of current treatment options. CA Cancer J Clin 2008;58:245–259. 9. Cotten A, Dewatre F, Cortet B, et al.: Percutaneous vertebroplasty for osteolytic metastases and myeloma: Effects of the percentage of lesion filling and the leakage of methyl methacrylate at clinical follow‐up. Radiology 1996;200:525–530. 10. Shimony JS, Gilula LA, Zeller AJ, et al.: Percutaneous vertebroplasty for malignant compression fractures with epidural involvement. Radiology 2004;232:846–853. 11. Saliou G, Kocheidael M, Lehmann P, et al.: Percutaneous vertebroplasty for pain management in malignant fractures of the spine with epidural involvement. Radiology 2010;254:882–890. 12. Sun G, Jin P, Li M, et al.: Percutaneous vertebroplasty for pain management in spinal metastasis with epidural involvement. Technol Cancer Res Treat 2011;10:267–274. 13. Bilsky MH, Laufer I, Fourney DR, et al.: Reliability analysis of the epidural spinal cord compression scale. J Neurosurg Spine 2010;13: 324–328.
Journal of Surgical Oncology
14. American Spinal Injury Association. Reference manual of the International standards for neurological classification of spinal cord injury. Chicago, IL: American Spinal Injury Association; 2003. 15. Guo WH, Meng MB, You X, et al.: CT‐guided percutaneous vertebroplasty of the upper cervical spine via a translateral approach. Pain Physician 2012;15:E733–E741. 16. Fenton DS, Czervionke LF: Image‐guided spine intervention. WB Saunders Philadelphia; 2003. 17. Gilbert HA, Kagan AR, Nussbaum H, et al.: Evaluation of radiation therapy for bone metastases: Pain relief and quality of life. AJR Am J Roentgenol 1997;129:1095–1096. 18. Weill A, Chiras J, Simon JM, et al.: Spinal metastases: Indications for and results of percutaneous injection of acrylic surgical cement. Radiology 1996;199:241–247. 19. Peh WC, Gilula LA: Percutaneous vertebroplasty: Indications, contraindications, and technique. Br J Rdiol 2003;76:69–75. 20. Mathis JM: Vertebroplasty for vertebral fractures with intravertebral clefts. AJNR Am J Neuroradiol 2002;23:1619–1620. 21. Barragan‐Campos HM, Vallée JN, Lo D, et al.: Percutaneous vertebroplasty for spinal metastases: Complications. Radiology 2002;238:354–362. 22. Cotton A, Dewatre F, Cortet B, et al.: Percutaneous vertebroplasty for osteolytic metastases and myeloma: Effects of the percentage of lesion filling and the leakage of methyl methacrylate at clinical follow‐up. Radiology 1996;200:525–530. 23. Schmidt R, Cakir B, Mattes T, et al.: Cement leakage during vertebroplasty: An underestimated problem? Eur Spine J 2005;14: 466–473.
Percutaneous Vertebroplasty for Painful Spinal Metastasis With Epidural Encroachment GANG SUN,
MD, PhD,*
LI LI, MD, PENG JIN,
MD,
XUN-WEI LIU, MD, AND MIN LI, MD
Department of Medical Imaging, Jinan Military General Hospital, Jinan, Shandong Province, China
Background and Objectives: Spinal metastasis with epidural encroachment is regarded by several authors to be a contraindication to percutaneous vertebroplasty (PVP) because of the risk of increasing symptomatic leakage of cement. This analysis aims to evaluate the safety and efficacy of PVP in patients with painful spinal metastasis and encroachment of epidural space. Methods: A retrospective study was conducted to review 43 consecutive patients with spinal metastasis that underwent PVP, for a total of 69 affected levels. All patients had at least 1 level associated with epidural encroachment related to metastasis. Among these patients, 14 had signs of spinal cord or cauda equina compression. Pain intensity was scored on a visual‐analog scale (VAS). The analgesic efficacy was defined as at least 50% improvement in pain score as compared with the pre‐procedure baseline and post‐procedure. Clinical improvement of neurological compressive symptoms was defined as a decrease in ASIA impairment scale from baseline of 1 point or more. Results: The analgesic efficacy was achieved in 89.7% of survival patients at 1 month, 87.5% at 3 months, 86.9% at 6 months, and 84.6% at 1 year. Small amount cement leakages were detected in 69.6% of treated levels without clinical complications. No deterioration of spinal cord or cauda equina compression symptoms was observed after a PVP in any patients. The different grade of epidural encroachment of the lesions was not correlated with filling volume or extraosseous leakage (P > 0.05). The treated levels with epidural encroachment showed a statistically significant relationship to spinal‐canal leakage (P < 0.05). Conclusions: PVP can be performed safely and effectively in patients with painful spinal metastasis and epidural encroachment.
J. Surg. Oncol. 2014;110:123–128. ß 2014 Wiley Periodicals, Inc.
KEY WORDS: osteolytic metastasis; spine; pain; interventional procedures; percutaneous vertebroplasty
INTRODUCTION Percutaneous vertebroplasty (PVP) involves the injection of a biomaterial, usually polymethyl methacrylate (PMMA), into the involved vertebral body under fluoroscopic guidance. Reinforcement of the bone and stabilization of the spine relieves pain and prevents pathological fractures [1–4]. The principal indications for this procedure are osteolytic vertebral metastasis and myeloma, vertebral hemangioma, and painful osteoporotic vertebral collapse. Symptomatic spinal cord compression is usually considered as a contraindication to PVP due to the potential risk of aggravating the symptoms. The destruction of the posterior vertebral wall with absence of direct‐compressive symptoms or evidence of tumor‐epidural encroachment is considered to be a relative contraindication to the procedure [5–7]. Despite new developments in surgery, medical oncology, and radiotherapy, the treatment option is still a matter of debate for painful spinal metastasis in patients with epidural encroachment in whom anticipated survival was limited [8]. Earlier investigations have evaluated the analgesic and stabilizing effects of PVP in spinal metastasis without epidural encroachment [6,7,9]. However, to our knowledge, there have been only three reports of PVP in vertebral lesions with epidural encroachment, with or without neurologic deficit [10–12]. The experience of the procedure in treating such patients remains limited. We have successfully performed PVP in spinal metastasis with epidural encroachment with or without neurologic deficit. The purpose of this study was to review our experience on painful spinal metastasis in 43 patients with epidural encroachment treated by PVP and evaluate the efficacy and safety of the procedure.
MATERIALS AND METHODS Selection Criteria We performed a retrospective review of all patients with painful spinal metastasis who underwent vertebroplasty in our hospital between
ß 2014 Wiley Periodicals, Inc.
March 2000 and May 2012. During this period, 412 consecutive patients with painful spinal metastasis were treated using PVP at our hospital. Approval for this study was obtained from the hospital review board at our hospital for this retrospective study, and all patients had previously consented to the use of their medical records for research purposes. The medical files from all cases were reviewed for evidence of effacement of the epidural space around the spinal cord and compression of the cord. Patients were excluded from consideration for this study if they were treated for a spinal metastasis without epidural effacement. From this group, 48 patients with at least 1 vertebral lesion associated with epidural effacement related to metastatic spinal disease were identified. Five of these patients were excluded because of lack of available imaging data for retrospective review. A total of 43 patients (24 women, 19 men, with mean age of 64.1 years, range of 34–84 years) were included in this study. Primary tumor types were lung (n ¼ 13), breast (n ¼ 12), kidney (n ¼ 5), stomach (n ¼ 4), bladder (n ¼ 3), uterine (n ¼ 3), colon (n ¼ 2), and ovary (n ¼ 1). The patients underwent PVP for management of persistent pain not responsive to conservative therapy, including various analgesics and chemotherapeutic regimens. The decision to perform
Grant sponsor: National Nature Science Foundation of China; Grant number: 51073173; Grant sponsor: National High Technology Research and Development Program of China; Grant number: 2013AA032203. Conflict of interest: All authors stated that they had no conflicts of interest. *Correspondence to: Gang Sun, PhD, MD, Department of Medical Imaging, Jinan Military General Hospital, No. 25, Shifan Road, Jinan, Shandong Province 250031, China. Fax: þ86‐531‐51666486. E‐mail: [email protected] Received 12 November 2013; Accepted 25 February 2014 DOI 10.1002/jso.23608 Published online 24 March 2014 in Wiley Online Library (wileyonlinelibrary.com).
124
Sun et al.
PVP was made by the attending radiologist in conjunction with referring physicians and surgeons after review of clinical and radiographic data in all patients who were not candidates for surgery due to the patient’s poor condition contraindicating the operation. Radiation therapy was not indicated in 27 patients who had already received the maximum radiation dose for local lesions. PVP procedures were performed pre‐ or post‐radiotherapy for rapid stabilization of the vertebrae in 16 patients due to delayed and incomplete action of radiation therapy.
Parameters for Analysis All patients underwent preoperative, unenhanced CT and MRI scans, focusing on the location, and extent of the lesions and their relationship to epidural space and spinal cord. The vertebral structure (osteolytic, osteosclerotic, or mixed) and presence of an osteolytic lesion on the posterior vertebral body wall were evaluated on plain CT scans. The evidence of the epidural space encroachment around the spinal cord, or compression of the spinal cord, was assessed on MRI scans. To analyze the relationship of the grade of epidural encroachment and post‐PVP changes in imaging data, we classified the grade of epidural encroachment or spinal cord involvement as follows: no epidural encroachment (group A), mild epidural encroachment without spinal cord deformity (group B), and epidural encroachment with spinal cord deformity (group C), depending on a modified epidural spinal cord compression scale suggested by Bilsky and Shimony [10,13]. This classification scheme is illustrated in Figure 1. The American Spinal Injury Association (ASIA) impairment‐scale designation was used in grading the degree of the severity of the neurologic deficit [14]: A (complete), no sensory or motor function is preserved in the sacral segments S4–S5; B (sensory incomplete), sensory but not motor function is preserved below the neurological level and includes the sacral segments S4–S5; C (motor incomplete), motor function is preserved below the neurological level and more than half of key muscle functions below the single neurological level of injury have a muscle grade less than 3 (grades 0–2); D (motor incomplete), motor function is preserved below the neurological level and at least half (half or more) of key muscle functions have a muscle grade of 3 or more; and E (normal), normal motor and sensory function. Pain level with activity was assessed according to the visual‐analog scale (VAS) score; a score of 0 indicated no pain and a score of 10 indicated the most pain imaginable.
PVP Procedure PVP procedures were performed under strict, sterile conditions using C‐arm fluoroscopy for guidance. The patients were placed in a prone position on the table and continuously monitored for blood pressure, electrocardiogram, and pulse‐oxygen saturation during the procedure. Conscious sedation and analgesia were obtained with IV flunitrazepam
(Versed, Roche Pharmaceuticals, Shanghai, China) and buprenorphine hydrochloride (Institute of Pharmaceutical Research, Tianjin, China). The patients were kept alert enough so they could state if any pain developed during the procedure. Under local anesthesia, a 13‐gauge, beveled trocar needle was used for vertebrae treated above T7 through a parapedicular (transcostovertebral) route, and an 11‐gauge, beveled trocar needle was for vertebrae treated between T8‐L5 through a transpedicular route. With fluoroscopic guidance, the needle was advanced to the center of the lesion at the vertebral body. When the needle placement was completed, vertebral venography through the needle was performed with 3–5 ml of half‐strength contrast agent (a 50:50 mixture of Iohexol Injection and normal saline) to evaluate needle positioning. If the needle was in, or immediately adjacent to, the basivertebral vein or another larger draining vein, it was repositioned. After confirming ideal needle placement, cement was prepared as described elsewhere [15]. In brief, the cement material was prepared by combining polymethylmethacrylate (PMMA) powder with sterile barium sulfate for opacification, followed by the addition of a liquid monomer to create a smooth‐liquid consistency. Cement was injected into the vertebral body with a screw‐type 10 ml syringe under fluoroscopy in the lateral projection. In the earlier, injected cement was the consistency of toothpaste for 58 levels of 32 patients, whereas in recent years, it was in the consistency of toothpaste for 5 levels with no epidural encroachment and a smooth‐liquid consistency for 11 levels with epidural encroachment in 11 patients. If cement leakage was detected, the injection was temporarily halted to allow the cement to harden, and resumed after changing the direction of the bevel and the position of the needle tip to fill as much of the lesion as possible. The stopping point was determined when the cement reached the posterior quarter of the vertebral body. Injection was immediately terminated when the patient complained of any pain, such as radicular pain, that could be due to pressure on posterior neurologic structures. If the cement mixture filled only one side of the vertebral body from a single pedicle injection, the contralateral pedicle was accessed for further filling of the vertebral body. A biopsy was routinely performed before cement injection. In patients with multiple spinal metastases, a PVP at corresponding levels was performed routinely in a one‐stage operation. Post‐procedural steroid therapy (1.5 mg per kilogram of body weight of prednisone equivalent, once per day for 5 days) was given for patients with symptoms of partial cord compression. A selective thoracolumbar nerve root block with lidocaine, ropivacaine, and dexamethasone was performed unilaterally or bilaterally in patients with pre‐existing and persistent nonrelieved post‐ procedure radiculopathy [16].
Clinical Evaluation All clinical, imaging, and laboratory data of patients who underwent PVP were reviewed by the operators (G.S. or P.J.). Patients were evaluated clinically before and immediately after treatment, and 1 week
Fig. 1. Scheme shows classification of epidural space encroachment. a: Scheme shows normal epidural space. b: Scheme shows mild ventral epidural space encroachment without spinal cord deformity. c: Scheme shows ventral epidural space encroachment with spinal cord deformity. Journal of Surgical Oncology
PVP for Spinal Metastasis Epidural Invasion later. During the period, patients’ acute post‐procedure adverse events, ASIA‐impair scale, and VAS score with activity were evaluated. Further, VAS scores and ASIA‐impair scale were obtained at 1, 3, 6, and 12 months. Effective pain relief was defined as at least a 50% improvement in VAS scores as compared with baseline. Clinical improvement of neurological compressive symptoms was defined as a decrease in ASIA impairment scale from baseline of 1 point or more. A CT scan of the spine was obtained immediately after PVP to assess cement distribution and extraosseous leakage in treated levels.
Statistical Analysis Statistical analysis was performed with statistic software (SPSS, version 16.0 for Windows; SPSS, Chicago, IL). Categorical variables are frequencies or percentages. A Fisher exact test and Pearson Chi‐ squared test were performed for qualitative variables, and a Mann– Whitney test was used for quantitative variables. A P value of less than 0.05 was considered to indicate a significant difference. The cumulative proportion of analgesic efficacy after PVP was evaluated by Kaplan– Meier analysis, with the exclusion of dead patients.
RESULTS Sixty‐nine levels (range T3‐L5 levels) in 43 patients were judged suitable for this study. The number of metastatic lesions in each patient varied from 1 to 4 levels: 24 patients (24/43, 56%) had 1 lesion level, 13
125
patients (13/43, 30%) had 2 lesion levels, 5 patients (5/43, 12%) had 3 lesion levels, and 1 patient (1/43, 2%) had 4 lesion levels. Each patient had at least 1 level with epidural encroachment (Fig. 2). In total, 23 levels with normal epidural space were classified into group A, 23 levels with mild ventral epidural space encroachment and no spinal cord deformity were classified into group B, and 21 levels with ventral epidural space encroachment and spinal cord deformity were classified into group C (Table I). Of the 43 patients, 14 had signs of spinal cord or cauda equina compression. One of these patients presented complete impairment (ASIA scale A) at the T3 level, and 13 presented incomplete impairment (ASIA scale C or D) at T3 (n ¼ 1), T11 (n ¼ 2), T12 (n ¼ 2), L1 (n ¼ 3), L4 (n ¼ 3), L5 (n ¼ 2), respectively. PVPs were performed successfully with a unipedicular approach in 32 levels and a bipedicular approach in 37 levels. The amount of cement injected per vertebral body varied between 2.0 and 7.0 ml (mean 5.2 1.3 ml). On post‐procedural CT images, the percentage of lesion filling with bone cement was more than 50% in all levels. Leakages of small amounts of cement were detected in 69.6% (48 of 69) levels, which were localized at the intervertebral disk in 21.7% (15/69), epidural space in 20.3% (14/69), and other location in 27.5% (19/69) (Table II). This different grade of epidural encroachment of the vertebral lesions was not correlated with filling volume or with the risk of extraosseous leakage (P > 0.05). The treated vertebral levels with epidural encroachment showed a statistically significant relationship to spinal canal leakage by univariate analysis (P < 0.05) (Table III). No cement leakage was symptomatic with onset of spinal cord or cauda equina compression
Fig. 2. Sagittal T1‐weighted (a), T2‐weighted MR images (b), Sagittal (c), and axial CT views (d) show metastatic pulmonary carcinoma of the spine and compression fracture of L1 vertebral body with epidural encroachment and destruction of posterior wall (arrows) in 61‐year‐old man. Patient had local pain and signs of spinal cord compression with complete impairment (ASIA scale D) at L1 level. Lateral fluoroscopic radiograph (e) and Sagittal CT images (f) after injection PVP showed a good deposition of the cement in the lesion (arrows). Note: cement leakage of the epidural space seen on axial CT image (g) after the procedure (arrowhead). Journal of Surgical Oncology
126
Sun et al.
TABLE I. Characteristics of Treated Levels Treated level
Group A (n ¼ 25)
Group B (n ¼ 23)
Group C (n ¼ 21)
1 2 1 0 1 1 0 0 3 1 3 1 3 2 6
0 1 0 1 1 2 2 1 2 2 2 2 2 4 1
1 0 0 2 1 1 1 1 2 2 3 1 1 3 2
TABLE III. Statistical Results of the Clinical, Radiologic, and Technical Parameters P value
Variable T3 T4 T5 T6 T7 T8 T9 T10 T11 T12 L1 L2 L3 L4 L5
Volume of cement Group A vs. B Group B vs. C Group C vs. A Cement leakage Group A vs. B Group B vs. C Group C vs. A Volume of cement Epidural leakage Group A vs. B Group B vs. C Group C vs. A
0.982a 0.845a 0.897a 0.838b 0.426b 0.539b 0.033c 1.000c 0.017c
a
Mann–Whitney test. Pearson Chi‐squared test. c Fisher exact test. b
symptoms after procedures. Systemic complications, such as bleeding, pulmonary embolism, infection, stroke, or cardiac arrest, were not encountered in this series. Selective nerve‐root blocks were performed before PVP procedures in five patients with pre‐existing radiculopathy. Three of the five reported persistent post‐procedure radiculopathy. They received other selective nerve‐root block on the second day after the PVP procedure. No patients required surgery. The mean VAS score with activity was 7.6 (range 6–10) preoperatively as a baseline. Effective pain relief, based on 50% or greater improvement in VAS scores as compared with baseline, was rated at 93% (40/43) at 1 day and 95% (41/43) at 1 week. In this period, six patients with incomplete neurological impairment had improved neurological compression symptoms. None of the 43 patients reported worsening spinal cord or cauda equina compression symptoms. Three patients reported no change in radiculopathy immediately following the procedure and were treated with selective nerve‐root blocks, which resulted in improvement in pain levels. In two patients, however, the radiculopathy eventually returned to pre‐PVP levels at 1 week following the procedure. The time of the follow‐up among the 43 patients was biased in favor of a short period because of the short life expectancies in this population. The final follow‐ups prior to patients’ death were made at the following times: 39 patients at 1 month, 32 at 3 months, 23 at 6 months, and 13 at 1 year. Among all patients, effective pain relief was achieved in 35 (89.7%) at 1 month, 28 (87.5%) at 3 months, 20 (86.9%) at 6 months, and 11 (84.6%) at 1 year (Table IV). The cumulative proportion of effective pain relief after PVP was evaluated by the Kaplan–Meier survival curve (Fig. 3). During the following‐up period, 19 patients reported recurring painful symptoms after a period of improvement. The painful symptoms in the spine were located at PVP‐treated levels in 12 patients, and at
TABLE II. Treatment Data, and Cement Leakage Parameter Vertebroplasty Approach Unipedicular Bipedicular Total cement leakage Discal Paravertebral Foramina Epidural Access track
Journal of Surgical Oncology
No. of vertebrae 69 46 23 48 15 12 3 14 4
(66.7) (33.3) (69.6) (21.7) (17.4) (4.4) (20.3) (5.8)
levels other than PVP‐treated areas in 7 patients. Of the six patients who had experienced improved neurological compression symptoms at 1 week after the procedure, two were unchanged and four had deteriorated at the final clinical follow‐up. The reason may have been related to disease progression.
DISCUSSION Patients with severe, intractable pain caused by spinal metastasis are left with few pain‐control options after the failure of conservative treatment. Although radiotherapy may result in partial or complete pain relief, many patients have either late onset of pain relief, or limited or no response to the therapy [17]. Open surgery may be not available in these patients with multiple metastases due to a poor general medical status, limited life expectancy because of the long recovery period, and high morbidity and mortality. PVP remains a useful option for eliminating or decreasing pain from vertebral lesions that cannot be controlled with conservative treatment [18,19]. However, the procedure bears increased risks for severe symptomatic complications from leakage of cement into the spinal canal and neural foramina, especially if the posterior cortex of the vertebral body is destructive. Usually, the presence of destruction of the posterior vertebral wall, with or without spinal compressive symptoms, has been considered as a contraindication, or relative contraindication, to the procedure [5–7]. In this retrospective study, we reviewed patients in which PVP was performed at levels with epidural encroachment from spinal canal metastases. The study showed that PVP was performed successfully in all patients. The analgesic efficacy of PVP in survival patients, defined as at least a 50% improvement in pain VAS scores as compared with baseline, was achieved in most survival patients in this study. The incidence of cement leakage in PVP reported in the literature ranges from 9.2% to 73% [20–22]. One study reported the detection of leakage was different based on different imaging modalities, and the cement leakage rate was often underestimated by radiography compared with CT [23]. The incidence of cement leakage in this study was 62.3% (42 of 69) in treated levels on post‐procedural CT images. The comparatively high leakage rate in the present study may be related to the fact that leakages were observed by CT and that many lesions involved cortical bone. Although no significant differences in extraosseous leakage were observed between levels with no epidural encroachment and levels with epidural encroachment in our study, spinal‐canal leakage was associated with epidural encroachment with and without spinal cord deformity. The spinal‐canal leakage observed in the present study could partly correspond to the fact that destruction of
PVP for Spinal Metastasis Epidural Invasion
127
TABLE IV. Follow‐Up of Patients After Vertebroplasty Parameters No. of cases Effective pain relief No effective pain relief
1‐Week follow‐up
1‐Month follow‐up
3‐Month follow‐up
6‐Month follow‐up
1‐Year follow‐up
43 41 (95) 2 (5)
39 35 (89.7) 4 (15.4)
32 28 (87.5) 4 (18.8)
23 20 (86.9) 3 (14.3)
13 11 (84.6) 2 (15.5)
the posterior vertebral body wall or epidural extension of the tumor reduces the ability to prevent cement releasing into spinal canal. Another possible reason is that the injected cement, with a smooth‐liquid consistency rather than a toothpaste consistency in the partially treated levels, may have diffused and deposed in the lesion of least resistance inside the vertebra and prevented it from pushing the epidural tumor mass to the spinal cord. In levels with a malignant soft‐tissue mass involved in the anterior part of the spinal canal, the cement commonly diffuses the soft‐tissue mass in the spinal canal. Cortical destruction and presence of an epidural soft‐tissue mass may be a factors that are likely to increase the rate of complications. However, Weill et al. [18] found that complications of PVP for vertebral tumors are not more frequent when there is destruction of the posterior cortex of the vertebral body. From the data in our study, no complication with clinical sequelae was observed after PVP in the patients with epidural encroachment, which is similar to other authors’ results in vertebral metastases with epidural encroachment [10,11]. Also noted is that the foraminal leakages of small amounts of cement observed in three cases in this study were not responsible for any complications, in line with results from the previous studies [10,23] in which no correlation was found between foraminal leakages and neuralgia. A possible explanation is that too little cement leaked to cause compression of anatomic structures and produce clinical symptoms. In general, the leakage rate of cement and the analgesic efficacy in this study are not substantially different from other studies in patients with metastatic disease and multiple myeloma that excluded patients with clinical evidence of spinal cord compression. For example, Cotten et al. [9] reported that pain status improved in 75% of patients at 6‐month follow‐up and cement leakage was detected in 72.5% of patients in a series of 37 patients with osteolytic metastases and myeloma. Our data
indicate that the increased risk of clinical complications related to cement leakage is small, and there are substantial benefits for patients with few other pain‐control options, severe pain, and a short life expectancy. On the basis of our data, we believe that painful spinal metastasis with epidural encroachment, and even spinal cord compression, should not preclude performance of PVP. With care, the procedure can be safely performed by experienced operators on this group of patients, even in clinical situations that have been cited in the literature as a contraindication or a relative contraindication. It is worth noting that the follow‐up results in this population indicated that many patients reported a pattern of early benefit followed by a subsequent deterioration in their condition near the end of life. Patients increased physical activity and improved quality of life temporarily. It is reasonable to assume that PVP is a palliative treatment for pain relief and spine stabilization concerning spinal metastasis but has a limited antitumor effect, which can not impede the disease progression. PVP combined with multidisciplinary procedures of anticancer modality have shown promising results [8]. The need of PVP for patients suffered spinal metastases with epidural encroachment should be consulted by a multidisciplinary team including orthopedic surgeons, neuroradiologists, medical oncologists, radiation oncologists, and interventional radiologists. Our study has several limitations. Most importantly, the study was retrospective in nature. Secondly the small size of patient samples and the short life spans of most patients might mask the benefits and risks of the procedure. A much larger group of patients might reveal a higher complication rate, especially treatment failure. In addition, patients with epidural encroachment in whom PVP is indicated are relatively few in number compared with the majority of cancer patients without epidural encroachment where PVP is sufficient. In conclusion, the PVP procedure is especially valuable in the treatment of malignant vertebral fractures. Our study confirms that epidural encroachment, and even spinal cord compression with epidural encroachment, should not preclude performance of PVP in patients with few other treatment options and a short life expectancy. Considering the risk of complications of the procedure in this type of patients, it is highly recommended that patients should be properly screened and the procedure be performed by an experienced specialist.
ACKNOWLEDGMENTS This study was supported by Grants No.51073173 from National Nature Science Foundation of China (NNSFC) and No.2013AA032203 National High Technology Research and Development Program of China (863 programme).
REFERENCES
Fig. 3. Kaplan–Meier curve of pain relief after PVP, with exclusion of dead patients. X‐axis corresponds to time after PVP procedure. Journal of Surgical Oncology
1. Pilitsis JG, Rengachary SS: The role of vertebroplasty in metastatic spinal disease. Neurosurg Focus 2001;11:e9. 2. Alvarez L, Perez‐Higueras A, Quinones D, et al.: Vertebroplasty in the treatment of vertebral tumors: Postprocedural outcome and quality of life. Eur Spine J 2003;12:356–360. 3. Legroux‐Gerot I, Lormeau C, Boutry N, et al.: Long‐term follow‐up of vertebral osteoporotic fractures treated by percutaneous vertebroplasty. Clin Rheumatol 2004;23:310–317. 4. Wenger M: Vertebroplasty for metastasis. Med Oncol 2003;20: 203–209.
128
Sun et al.
5. Cotten A, Boutry N, Cortet B, et al.: Percutaneous vertebroplasty: state of the art. Radiographics 1998;18:311–320. 6. Murphy KJ, Deramond H: Percutaneous vertebroplasty in benign and malignant disease. Neuroimaging Clin N Am 2000;10:535– 545. 7. Heini PF, Walchli B, Berlemann U: Percutaneous transpedicular vertebroplasty with PMMA: Operative technique and early results. A prospective study for the treatment of osteoporotic compression fractures. Eur Spine J 2000;9:445–450. 8. Bartels RH, van der Linden YM, van der Graaf WT: Spinal extradural metastasis: Review of current treatment options. CA Cancer J Clin 2008;58:245–259. 9. Cotten A, Dewatre F, Cortet B, et al.: Percutaneous vertebroplasty for osteolytic metastases and myeloma: Effects of the percentage of lesion filling and the leakage of methyl methacrylate at clinical follow‐up. Radiology 1996;200:525–530. 10. Shimony JS, Gilula LA, Zeller AJ, et al.: Percutaneous vertebroplasty for malignant compression fractures with epidural involvement. Radiology 2004;232:846–853. 11. Saliou G, Kocheidael M, Lehmann P, et al.: Percutaneous vertebroplasty for pain management in malignant fractures of the spine with epidural involvement. Radiology 2010;254:882–890. 12. Sun G, Jin P, Li M, et al.: Percutaneous vertebroplasty for pain management in spinal metastasis with epidural involvement. Technol Cancer Res Treat 2011;10:267–274. 13. Bilsky MH, Laufer I, Fourney DR, et al.: Reliability analysis of the epidural spinal cord compression scale. J Neurosurg Spine 2010;13: 324–328.
Journal of Surgical Oncology
14. American Spinal Injury Association. Reference manual of the International standards for neurological classification of spinal cord injury. Chicago, IL: American Spinal Injury Association; 2003. 15. Guo WH, Meng MB, You X, et al.: CT‐guided percutaneous vertebroplasty of the upper cervical spine via a translateral approach. Pain Physician 2012;15:E733–E741. 16. Fenton DS, Czervionke LF: Image‐guided spine intervention. WB Saunders Philadelphia; 2003. 17. Gilbert HA, Kagan AR, Nussbaum H, et al.: Evaluation of radiation therapy for bone metastases: Pain relief and quality of life. AJR Am J Roentgenol 1997;129:1095–1096. 18. Weill A, Chiras J, Simon JM, et al.: Spinal metastases: Indications for and results of percutaneous injection of acrylic surgical cement. Radiology 1996;199:241–247. 19. Peh WC, Gilula LA: Percutaneous vertebroplasty: Indications, contraindications, and technique. Br J Rdiol 2003;76:69–75. 20. Mathis JM: Vertebroplasty for vertebral fractures with intravertebral clefts. AJNR Am J Neuroradiol 2002;23:1619–1620. 21. Barragan‐Campos HM, Vallée JN, Lo D, et al.: Percutaneous vertebroplasty for spinal metastases: Complications. Radiology 2002;238:354–362. 22. Cotton A, Dewatre F, Cortet B, et al.: Percutaneous vertebroplasty for osteolytic metastases and myeloma: Effects of the percentage of lesion filling and the leakage of methyl methacrylate at clinical follow‐up. Radiology 1996;200:525–530. 23. Schmidt R, Cakir B, Mattes T, et al.: Cement leakage during vertebroplasty: An underestimated problem? Eur Spine J 2005;14: 466–473.
